diff --git "a/checkpoint-61000/trainer_state.json" "b/checkpoint-61000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-61000/trainer_state.json" @@ -0,0 +1,427033 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.61, + "eval_steps": 500, + "global_step": 61000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 1e-05, + "grad_norm": 1.6317715205474108, + "learning_rate": 3e-06, + "loss": 10.867, + "step": 1 + }, + { + "epoch": 2e-05, + "grad_norm": 1.6173147870740345, + "learning_rate": 6e-06, + "loss": 10.8657, + "step": 2 + }, + { + "epoch": 3e-05, + "grad_norm": 1.6387509359885835, + "learning_rate": 9e-06, + "loss": 10.8658, + "step": 3 + }, + { + "epoch": 4e-05, + "grad_norm": 1.597552357815991, + "learning_rate": 1.2e-05, + "loss": 10.865, + "step": 4 + }, + { + "epoch": 5e-05, + "grad_norm": 1.6454471252189307, + "learning_rate": 1.5e-05, + "loss": 10.8617, + "step": 5 + }, + { + "epoch": 6e-05, + "grad_norm": 1.6407925722175996, + "learning_rate": 1.8e-05, + "loss": 10.8593, + "step": 6 + }, + { + "epoch": 7e-05, + "grad_norm": 1.6096088910322361, + "learning_rate": 2.1000000000000002e-05, + "loss": 10.8456, + "step": 7 + }, + { + "epoch": 8e-05, + "grad_norm": 1.4682698515009915, + "learning_rate": 2.4e-05, + "loss": 10.8184, + "step": 8 + }, + { + "epoch": 9e-05, + "grad_norm": 1.3934246427009196, + "learning_rate": 2.7e-05, + "loss": 10.8113, + "step": 9 + }, + { + "epoch": 0.0001, + "grad_norm": 1.3326087040550991, + "learning_rate": 3e-05, + "loss": 10.7969, + "step": 10 + }, + { + "epoch": 0.00011, + "grad_norm": 1.2173090118888668, + "learning_rate": 3.2999999999999996e-05, + "loss": 10.7801, + "step": 11 + }, + { + "epoch": 0.00012, + "grad_norm": 1.176457724285593, + "learning_rate": 3.6e-05, + "loss": 10.7688, + "step": 12 + }, + { + "epoch": 0.00013, + "grad_norm": 1.1304424318539597, + "learning_rate": 3.9e-05, + "loss": 10.7498, + "step": 13 + }, + { + "epoch": 0.00014, + "grad_norm": 1.1158244568462428, + "learning_rate": 4.2000000000000004e-05, + "loss": 10.739, + "step": 14 + }, + { + "epoch": 0.00015, + "grad_norm": 1.10654756415174, + "learning_rate": 4.4999999999999996e-05, + "loss": 10.7299, + "step": 15 + }, + { + "epoch": 0.00016, + "grad_norm": 1.0751770608444569, + "learning_rate": 4.8e-05, + "loss": 10.7126, + "step": 16 + }, + { + "epoch": 0.00017, + "grad_norm": 1.046143788290158, + "learning_rate": 5.1000000000000006e-05, + "loss": 10.6968, + "step": 17 + }, + { + "epoch": 0.00018, + "grad_norm": 1.0230391412556632, + "learning_rate": 5.4e-05, + "loss": 10.6806, + "step": 18 + }, + { + "epoch": 0.00019, + "grad_norm": 0.9869982733638543, + "learning_rate": 5.7e-05, + "loss": 10.6649, + "step": 19 + }, + { + "epoch": 0.0002, + "grad_norm": 0.9728818553338922, + "learning_rate": 6e-05, + "loss": 10.6526, + "step": 20 + }, + { + "epoch": 0.00021, + "grad_norm": 0.9439994749998407, + "learning_rate": 6.3e-05, + "loss": 10.639, + "step": 21 + }, + { + "epoch": 0.00022, + "grad_norm": 0.9279528910342688, + "learning_rate": 6.599999999999999e-05, + "loss": 10.6244, + "step": 22 + }, + { + "epoch": 0.00023, + "grad_norm": 0.9208663519602571, + "learning_rate": 6.9e-05, + "loss": 10.6103, + "step": 23 + }, + { + "epoch": 0.00024, + "grad_norm": 0.9178551557561957, + "learning_rate": 7.2e-05, + "loss": 10.5993, + "step": 24 + }, + { + "epoch": 0.00025, + "grad_norm": 0.9193923250060233, + "learning_rate": 7.500000000000001e-05, + "loss": 10.5847, + "step": 25 + }, + { + "epoch": 0.00026, + "grad_norm": 0.9190901609677985, + "learning_rate": 7.8e-05, + "loss": 10.5717, + "step": 26 + }, + { + "epoch": 0.00027, + "grad_norm": 0.913753327244254, + "learning_rate": 8.1e-05, + "loss": 10.5597, + "step": 27 + }, + { + "epoch": 0.00028, + "grad_norm": 0.9119625217070774, + "learning_rate": 8.400000000000001e-05, + "loss": 10.5467, + "step": 28 + }, + { + "epoch": 0.00029, + "grad_norm": 0.9131038863398008, + "learning_rate": 8.7e-05, + "loss": 10.5323, + "step": 29 + }, + { + "epoch": 0.0003, + "grad_norm": 0.9186172829723749, + "learning_rate": 8.999999999999999e-05, + "loss": 10.517, + "step": 30 + }, + { + "epoch": 0.00031, + "grad_norm": 0.9155453723962563, + "learning_rate": 9.3e-05, + "loss": 10.5024, + "step": 31 + }, + { + "epoch": 0.00032, + "grad_norm": 0.909575589137279, + "learning_rate": 9.6e-05, + "loss": 10.4882, + "step": 32 + }, + { + "epoch": 0.00033, + "grad_norm": 0.9102175928717151, + "learning_rate": 9.900000000000001e-05, + "loss": 10.4717, + "step": 33 + }, + { + "epoch": 0.00034, + "grad_norm": 0.9103991171564593, + "learning_rate": 0.00010200000000000001, + "loss": 10.4554, + "step": 34 + }, + { + "epoch": 0.00035, + "grad_norm": 0.9164468459870767, + "learning_rate": 0.00010500000000000002, + "loss": 10.4388, + "step": 35 + }, + { + "epoch": 0.00036, + "grad_norm": 0.9129850089149896, + "learning_rate": 0.000108, + "loss": 10.4213, + "step": 36 + }, + { + "epoch": 0.00037, + "grad_norm": 0.9029808327462479, + "learning_rate": 0.000111, + "loss": 10.4047, + "step": 37 + }, + { + "epoch": 0.00038, + "grad_norm": 0.9074583652458479, + "learning_rate": 0.000114, + "loss": 10.3859, + "step": 38 + }, + { + "epoch": 0.00039, + "grad_norm": 0.9101621604618185, + "learning_rate": 0.000117, + "loss": 10.3665, + "step": 39 + }, + { + "epoch": 0.0004, + "grad_norm": 0.9144345472354501, + "learning_rate": 0.00012, + "loss": 10.3443, + "step": 40 + }, + { + "epoch": 0.00041, + "grad_norm": 0.9045315909874942, + "learning_rate": 0.000123, + "loss": 10.3264, + "step": 41 + }, + { + "epoch": 0.00042, + "grad_norm": 0.9101518170592343, + "learning_rate": 0.000126, + "loss": 10.3059, + "step": 42 + }, + { + "epoch": 0.00043, + "grad_norm": 0.9147491310031046, + "learning_rate": 0.000129, + "loss": 10.2831, + "step": 43 + }, + { + "epoch": 0.00044, + "grad_norm": 0.9138166723808987, + "learning_rate": 0.00013199999999999998, + "loss": 10.2617, + "step": 44 + }, + { + "epoch": 0.00045, + "grad_norm": 0.9166634386783579, + "learning_rate": 0.000135, + "loss": 10.2388, + "step": 45 + }, + { + "epoch": 0.00046, + "grad_norm": 0.9061319871747918, + "learning_rate": 0.000138, + "loss": 10.2171, + "step": 46 + }, + { + "epoch": 0.00047, + "grad_norm": 0.9134144094551727, + "learning_rate": 0.000141, + "loss": 10.1906, + "step": 47 + }, + { + "epoch": 0.00048, + "grad_norm": 0.9111015850262806, + "learning_rate": 0.000144, + "loss": 10.1669, + "step": 48 + }, + { + "epoch": 0.00049, + "grad_norm": 0.9106086112424904, + "learning_rate": 0.000147, + "loss": 10.1439, + "step": 49 + }, + { + "epoch": 0.0005, + "grad_norm": 0.9135108001899231, + "learning_rate": 0.00015000000000000001, + "loss": 10.1178, + "step": 50 + }, + { + "epoch": 0.00051, + "grad_norm": 0.9152031108656089, + "learning_rate": 0.000153, + "loss": 10.0918, + "step": 51 + }, + { + "epoch": 0.00052, + "grad_norm": 0.9133043896843657, + "learning_rate": 0.000156, + "loss": 10.0658, + "step": 52 + }, + { + "epoch": 0.00053, + "grad_norm": 0.9039676544194273, + "learning_rate": 0.000159, + "loss": 10.0419, + "step": 53 + }, + { + "epoch": 0.00054, + "grad_norm": 0.9245050218484777, + "learning_rate": 0.000162, + "loss": 10.0099, + "step": 54 + }, + { + "epoch": 0.00055, + "grad_norm": 0.915653013423474, + "learning_rate": 0.000165, + "loss": 9.9858, + "step": 55 + }, + { + "epoch": 0.00056, + "grad_norm": 0.90743999026624, + "learning_rate": 0.00016800000000000002, + "loss": 9.9567, + "step": 56 + }, + { + "epoch": 0.00057, + "grad_norm": 0.9125740935300273, + "learning_rate": 0.000171, + "loss": 9.9292, + "step": 57 + }, + { + "epoch": 0.00058, + "grad_norm": 0.9103134473221595, + "learning_rate": 0.000174, + "loss": 9.9046, + "step": 58 + }, + { + "epoch": 0.00059, + "grad_norm": 0.91502039796166, + "learning_rate": 0.000177, + "loss": 9.8727, + "step": 59 + }, + { + "epoch": 0.0006, + "grad_norm": 0.902549468432534, + "learning_rate": 0.00017999999999999998, + "loss": 9.8467, + "step": 60 + }, + { + "epoch": 0.00061, + "grad_norm": 0.9011753674575653, + "learning_rate": 0.000183, + "loss": 9.8184, + "step": 61 + }, + { + "epoch": 0.00062, + "grad_norm": 0.9073829944096542, + "learning_rate": 0.000186, + "loss": 9.7865, + "step": 62 + }, + { + "epoch": 0.00063, + "grad_norm": 0.900277782228101, + "learning_rate": 0.000189, + "loss": 9.7594, + "step": 63 + }, + { + "epoch": 0.00064, + "grad_norm": 0.8964836174343672, + "learning_rate": 0.000192, + "loss": 9.7292, + "step": 64 + }, + { + "epoch": 0.00065, + "grad_norm": 0.9040253232587138, + "learning_rate": 0.00019500000000000002, + "loss": 9.6969, + "step": 65 + }, + { + "epoch": 0.00066, + "grad_norm": 0.8969700410935529, + "learning_rate": 0.00019800000000000002, + "loss": 9.6728, + "step": 66 + }, + { + "epoch": 0.00067, + "grad_norm": 0.8981910064021, + "learning_rate": 0.000201, + "loss": 9.643, + "step": 67 + }, + { + "epoch": 0.00068, + "grad_norm": 0.9049113812894196, + "learning_rate": 0.00020400000000000003, + "loss": 9.61, + "step": 68 + }, + { + "epoch": 0.00069, + "grad_norm": 0.8903730719674341, + "learning_rate": 0.00020700000000000002, + "loss": 9.5824, + "step": 69 + }, + { + "epoch": 0.0007, + "grad_norm": 0.9034192157313848, + "learning_rate": 0.00021000000000000004, + "loss": 9.5456, + "step": 70 + }, + { + "epoch": 0.00071, + "grad_norm": 0.8933343163190056, + "learning_rate": 0.00021299999999999997, + "loss": 9.5189, + "step": 71 + }, + { + "epoch": 0.00072, + "grad_norm": 0.8968593008835964, + "learning_rate": 0.000216, + "loss": 9.4914, + "step": 72 + }, + { + "epoch": 0.00073, + "grad_norm": 0.8960167287531013, + "learning_rate": 0.00021899999999999998, + "loss": 9.4574, + "step": 73 + }, + { + "epoch": 0.00074, + "grad_norm": 0.8965354004907367, + "learning_rate": 0.000222, + "loss": 9.4304, + "step": 74 + }, + { + "epoch": 0.00075, + "grad_norm": 0.8923958382533757, + "learning_rate": 0.000225, + "loss": 9.3946, + "step": 75 + }, + { + "epoch": 0.00076, + "grad_norm": 0.9011054851249423, + "learning_rate": 0.000228, + "loss": 9.3667, + "step": 76 + }, + { + "epoch": 0.00077, + "grad_norm": 0.8959973023769764, + "learning_rate": 0.000231, + "loss": 9.3374, + "step": 77 + }, + { + "epoch": 0.00078, + "grad_norm": 0.8901362352309407, + "learning_rate": 0.000234, + "loss": 9.3055, + "step": 78 + }, + { + "epoch": 0.00079, + "grad_norm": 0.8931237127830396, + "learning_rate": 0.00023700000000000001, + "loss": 9.2694, + "step": 79 + }, + { + "epoch": 0.0008, + "grad_norm": 0.8937972757596185, + "learning_rate": 0.00024, + "loss": 9.2355, + "step": 80 + }, + { + "epoch": 0.00081, + "grad_norm": 0.8984186266430717, + "learning_rate": 0.00024300000000000002, + "loss": 9.2033, + "step": 81 + }, + { + "epoch": 0.00082, + "grad_norm": 0.9060711515558022, + "learning_rate": 0.000246, + "loss": 9.1789, + "step": 82 + }, + { + "epoch": 0.00083, + "grad_norm": 0.8964035250431631, + "learning_rate": 0.00024900000000000004, + "loss": 9.1434, + "step": 83 + }, + { + "epoch": 0.00084, + "grad_norm": 0.8986377804082708, + "learning_rate": 0.000252, + "loss": 9.1119, + "step": 84 + }, + { + "epoch": 0.00085, + "grad_norm": 0.8933390971667627, + "learning_rate": 0.000255, + "loss": 9.0885, + "step": 85 + }, + { + "epoch": 0.00086, + "grad_norm": 0.8953786211031305, + "learning_rate": 0.000258, + "loss": 9.0521, + "step": 86 + }, + { + "epoch": 0.00087, + "grad_norm": 0.8876773666743288, + "learning_rate": 0.000261, + "loss": 9.0302, + "step": 87 + }, + { + "epoch": 0.00088, + "grad_norm": 0.8929152197347487, + "learning_rate": 0.00026399999999999997, + "loss": 8.9927, + "step": 88 + }, + { + "epoch": 0.00089, + "grad_norm": 0.8872457913370299, + "learning_rate": 0.000267, + "loss": 8.9669, + "step": 89 + }, + { + "epoch": 0.0009, + "grad_norm": 0.8862509419810688, + "learning_rate": 0.00027, + "loss": 8.9393, + "step": 90 + }, + { + "epoch": 0.00091, + "grad_norm": 0.8807880266520192, + "learning_rate": 0.000273, + "loss": 8.9089, + "step": 91 + }, + { + "epoch": 0.00092, + "grad_norm": 0.8831207589668301, + "learning_rate": 0.000276, + "loss": 8.876, + "step": 92 + }, + { + "epoch": 0.00093, + "grad_norm": 0.883866988348631, + "learning_rate": 0.000279, + "loss": 8.8462, + "step": 93 + }, + { + "epoch": 0.00094, + "grad_norm": 0.8824853161719922, + "learning_rate": 0.000282, + "loss": 8.8199, + "step": 94 + }, + { + "epoch": 0.00095, + "grad_norm": 0.8794572899807177, + "learning_rate": 0.000285, + "loss": 8.789, + "step": 95 + }, + { + "epoch": 0.00096, + "grad_norm": 0.8879520784944948, + "learning_rate": 0.000288, + "loss": 8.7571, + "step": 96 + }, + { + "epoch": 0.00097, + "grad_norm": 0.8798042444972031, + "learning_rate": 0.000291, + "loss": 8.7321, + "step": 97 + }, + { + "epoch": 0.00098, + "grad_norm": 0.8802597894834375, + "learning_rate": 0.000294, + "loss": 8.6954, + "step": 98 + }, + { + "epoch": 0.00099, + "grad_norm": 0.8806466323910314, + "learning_rate": 0.000297, + "loss": 8.6749, + "step": 99 + }, + { + "epoch": 0.001, + "grad_norm": 0.8777097574069823, + "learning_rate": 0.00030000000000000003, + "loss": 8.6485, + "step": 100 + }, + { + "epoch": 0.00101, + "grad_norm": 0.8786755147609817, + "learning_rate": 0.00030300000000000005, + "loss": 8.614, + "step": 101 + }, + { + "epoch": 0.00102, + "grad_norm": 0.8680143868447665, + "learning_rate": 0.000306, + "loss": 8.5949, + "step": 102 + }, + { + "epoch": 0.00103, + "grad_norm": 0.8748449452730288, + "learning_rate": 0.000309, + "loss": 8.5706, + "step": 103 + }, + { + "epoch": 0.00104, + "grad_norm": 0.8670215859388973, + "learning_rate": 0.000312, + "loss": 8.5498, + "step": 104 + }, + { + "epoch": 0.00105, + "grad_norm": 0.8687292008731472, + "learning_rate": 0.000315, + "loss": 8.5231, + "step": 105 + }, + { + "epoch": 0.00106, + "grad_norm": 0.8611486197845404, + "learning_rate": 0.000318, + "loss": 8.4945, + "step": 106 + }, + { + "epoch": 0.00107, + "grad_norm": 0.8521924348129856, + "learning_rate": 0.000321, + "loss": 8.4693, + "step": 107 + }, + { + "epoch": 0.00108, + "grad_norm": 0.8581933475380797, + "learning_rate": 0.000324, + "loss": 8.4407, + "step": 108 + }, + { + "epoch": 0.00109, + "grad_norm": 0.8524146875206363, + "learning_rate": 0.000327, + "loss": 8.421, + "step": 109 + }, + { + "epoch": 0.0011, + "grad_norm": 0.8682563584613229, + "learning_rate": 0.00033, + "loss": 8.3983, + "step": 110 + }, + { + "epoch": 0.00111, + "grad_norm": 0.8871469803064874, + "learning_rate": 0.000333, + "loss": 8.3685, + "step": 111 + }, + { + "epoch": 0.00112, + "grad_norm": 0.9236879668329372, + "learning_rate": 0.00033600000000000004, + "loss": 8.3463, + "step": 112 + }, + { + "epoch": 0.00113, + "grad_norm": 0.9129864456504505, + "learning_rate": 0.000339, + "loss": 8.3063, + "step": 113 + }, + { + "epoch": 0.00114, + "grad_norm": 0.8352263501003522, + "learning_rate": 0.000342, + "loss": 8.2966, + "step": 114 + }, + { + "epoch": 0.00115, + "grad_norm": 0.8592375580090957, + "learning_rate": 0.00034500000000000004, + "loss": 8.2718, + "step": 115 + }, + { + "epoch": 0.00116, + "grad_norm": 0.8674294753896091, + "learning_rate": 0.000348, + "loss": 8.2506, + "step": 116 + }, + { + "epoch": 0.00117, + "grad_norm": 0.8276917493567776, + "learning_rate": 0.000351, + "loss": 8.2188, + "step": 117 + }, + { + "epoch": 0.00118, + "grad_norm": 0.8476744963131545, + "learning_rate": 0.000354, + "loss": 8.2045, + "step": 118 + }, + { + "epoch": 0.00119, + "grad_norm": 0.844358071644388, + "learning_rate": 0.000357, + "loss": 8.1926, + "step": 119 + }, + { + "epoch": 0.0012, + "grad_norm": 0.8166594269287538, + "learning_rate": 0.00035999999999999997, + "loss": 8.1658, + "step": 120 + }, + { + "epoch": 0.00121, + "grad_norm": 0.8239930081024902, + "learning_rate": 0.000363, + "loss": 8.1389, + "step": 121 + }, + { + "epoch": 0.00122, + "grad_norm": 0.8099951959348987, + "learning_rate": 0.000366, + "loss": 8.1225, + "step": 122 + }, + { + "epoch": 0.00123, + "grad_norm": 0.830800320388625, + "learning_rate": 0.000369, + "loss": 8.1005, + "step": 123 + }, + { + "epoch": 0.00124, + "grad_norm": 0.8139169053139192, + "learning_rate": 0.000372, + "loss": 8.0791, + "step": 124 + }, + { + "epoch": 0.00125, + "grad_norm": 0.8112246790149765, + "learning_rate": 0.000375, + "loss": 8.0547, + "step": 125 + }, + { + "epoch": 0.00126, + "grad_norm": 0.7922278873371895, + "learning_rate": 0.000378, + "loss": 8.0424, + "step": 126 + }, + { + "epoch": 0.00127, + "grad_norm": 0.7755075943975184, + "learning_rate": 0.000381, + "loss": 8.0182, + "step": 127 + }, + { + "epoch": 0.00128, + "grad_norm": 0.8028212720713388, + "learning_rate": 0.000384, + "loss": 8.0, + "step": 128 + }, + { + "epoch": 0.00129, + "grad_norm": 0.8765129391436198, + "learning_rate": 0.00038700000000000003, + "loss": 7.9904, + "step": 129 + }, + { + "epoch": 0.0013, + "grad_norm": 1.1486399983200042, + "learning_rate": 0.00039000000000000005, + "loss": 7.9724, + "step": 130 + }, + { + "epoch": 0.00131, + "grad_norm": 1.0507777578095836, + "learning_rate": 0.000393, + "loss": 7.9382, + "step": 131 + }, + { + "epoch": 0.00132, + "grad_norm": 0.7511305165281239, + "learning_rate": 0.00039600000000000003, + "loss": 7.9248, + "step": 132 + }, + { + "epoch": 0.00133, + "grad_norm": 0.8108077692812641, + "learning_rate": 0.00039900000000000005, + "loss": 7.906, + "step": 133 + }, + { + "epoch": 0.00134, + "grad_norm": 0.8301313874662418, + "learning_rate": 0.000402, + "loss": 7.8952, + "step": 134 + }, + { + "epoch": 0.00135, + "grad_norm": 0.7123568807247732, + "learning_rate": 0.00040500000000000003, + "loss": 7.8651, + "step": 135 + }, + { + "epoch": 0.00136, + "grad_norm": 0.7697556904537746, + "learning_rate": 0.00040800000000000005, + "loss": 7.8515, + "step": 136 + }, + { + "epoch": 0.00137, + "grad_norm": 0.7190977621725152, + "learning_rate": 0.000411, + "loss": 7.8299, + "step": 137 + }, + { + "epoch": 0.00138, + "grad_norm": 0.7147305335216294, + "learning_rate": 0.00041400000000000003, + "loss": 7.8066, + "step": 138 + }, + { + "epoch": 0.00139, + "grad_norm": 0.747529428120578, + "learning_rate": 0.00041700000000000005, + "loss": 7.7954, + "step": 139 + }, + { + "epoch": 0.0014, + "grad_norm": 0.6748198408281931, + "learning_rate": 0.00042000000000000007, + "loss": 7.7774, + "step": 140 + }, + { + "epoch": 0.00141, + "grad_norm": 0.662142968172009, + "learning_rate": 0.000423, + "loss": 7.7644, + "step": 141 + }, + { + "epoch": 0.00142, + "grad_norm": 0.6859796391897652, + "learning_rate": 0.00042599999999999995, + "loss": 7.7534, + "step": 142 + }, + { + "epoch": 0.00143, + "grad_norm": 0.6828514310354903, + "learning_rate": 0.00042899999999999997, + "loss": 7.7255, + "step": 143 + }, + { + "epoch": 0.00144, + "grad_norm": 0.6490687938821236, + "learning_rate": 0.000432, + "loss": 7.7078, + "step": 144 + }, + { + "epoch": 0.00145, + "grad_norm": 0.6453156151137228, + "learning_rate": 0.000435, + "loss": 7.7035, + "step": 145 + }, + { + "epoch": 0.00146, + "grad_norm": 0.7166050341593803, + "learning_rate": 0.00043799999999999997, + "loss": 7.6672, + "step": 146 + }, + { + "epoch": 0.00147, + "grad_norm": 0.7443563124698165, + "learning_rate": 0.000441, + "loss": 7.6627, + "step": 147 + }, + { + "epoch": 0.00148, + "grad_norm": 0.7124453009383569, + "learning_rate": 0.000444, + "loss": 7.648, + "step": 148 + }, + { + "epoch": 0.00149, + "grad_norm": 0.6775306354557482, + "learning_rate": 0.00044699999999999997, + "loss": 7.6419, + "step": 149 + }, + { + "epoch": 0.0015, + "grad_norm": 0.8906380813028638, + "learning_rate": 0.00045, + "loss": 7.6135, + "step": 150 + }, + { + "epoch": 0.00151, + "grad_norm": 1.2892607147030477, + "learning_rate": 0.000453, + "loss": 7.605, + "step": 151 + }, + { + "epoch": 0.00152, + "grad_norm": 0.6492280537852009, + "learning_rate": 0.000456, + "loss": 7.5875, + "step": 152 + }, + { + "epoch": 0.00153, + "grad_norm": 0.8226748559002907, + "learning_rate": 0.000459, + "loss": 7.5783, + "step": 153 + }, + { + "epoch": 0.00154, + "grad_norm": 1.073498340899344, + "learning_rate": 0.000462, + "loss": 7.5662, + "step": 154 + }, + { + "epoch": 0.00155, + "grad_norm": 0.772416311968319, + "learning_rate": 0.000465, + "loss": 7.5481, + "step": 155 + }, + { + "epoch": 0.00156, + "grad_norm": 0.7999364848474875, + "learning_rate": 0.000468, + "loss": 7.5209, + "step": 156 + }, + { + "epoch": 0.00157, + "grad_norm": 0.9435570004081977, + "learning_rate": 0.000471, + "loss": 7.5196, + "step": 157 + }, + { + "epoch": 0.00158, + "grad_norm": 0.8364296006508578, + "learning_rate": 0.00047400000000000003, + "loss": 7.4997, + "step": 158 + }, + { + "epoch": 0.00159, + "grad_norm": 0.5596750978008624, + "learning_rate": 0.000477, + "loss": 7.4825, + "step": 159 + }, + { + "epoch": 0.0016, + "grad_norm": 0.5748856794890025, + "learning_rate": 0.00048, + "loss": 7.4616, + "step": 160 + }, + { + "epoch": 0.00161, + "grad_norm": 0.5986411025981033, + "learning_rate": 0.00048300000000000003, + "loss": 7.4419, + "step": 161 + }, + { + "epoch": 0.00162, + "grad_norm": 0.5117747222458712, + "learning_rate": 0.00048600000000000005, + "loss": 7.4429, + "step": 162 + }, + { + "epoch": 0.00163, + "grad_norm": 0.5509815158764758, + "learning_rate": 0.0004890000000000001, + "loss": 7.4259, + "step": 163 + }, + { + "epoch": 0.00164, + "grad_norm": 0.46393808675226217, + "learning_rate": 0.000492, + "loss": 7.4236, + "step": 164 + }, + { + "epoch": 0.00165, + "grad_norm": 0.5390679177469344, + "learning_rate": 0.000495, + "loss": 7.4006, + "step": 165 + }, + { + "epoch": 0.00166, + "grad_norm": 0.645354949444588, + "learning_rate": 0.0004980000000000001, + "loss": 7.3773, + "step": 166 + }, + { + "epoch": 0.00167, + "grad_norm": 1.0877276724000633, + "learning_rate": 0.000501, + "loss": 7.3741, + "step": 167 + }, + { + "epoch": 0.00168, + "grad_norm": 1.0909247587015876, + "learning_rate": 0.000504, + "loss": 7.3697, + "step": 168 + }, + { + "epoch": 0.00169, + "grad_norm": 0.5084368948335112, + "learning_rate": 0.0005070000000000001, + "loss": 7.3418, + "step": 169 + }, + { + "epoch": 0.0017, + "grad_norm": 1.5053615947271437, + "learning_rate": 0.00051, + "loss": 7.3492, + "step": 170 + }, + { + "epoch": 0.00171, + "grad_norm": 0.6120595065498764, + "learning_rate": 0.000513, + "loss": 7.3094, + "step": 171 + }, + { + "epoch": 0.00172, + "grad_norm": 0.9401489103136018, + "learning_rate": 0.000516, + "loss": 7.3189, + "step": 172 + }, + { + "epoch": 0.00173, + "grad_norm": 0.5558816443971099, + "learning_rate": 0.0005189999999999999, + "loss": 7.2906, + "step": 173 + }, + { + "epoch": 0.00174, + "grad_norm": 0.7810658223347703, + "learning_rate": 0.000522, + "loss": 7.314, + "step": 174 + }, + { + "epoch": 0.00175, + "grad_norm": 0.5409187817835382, + "learning_rate": 0.000525, + "loss": 7.2727, + "step": 175 + }, + { + "epoch": 0.00176, + "grad_norm": 0.630636606756854, + "learning_rate": 0.0005279999999999999, + "loss": 7.2484, + "step": 176 + }, + { + "epoch": 0.00177, + "grad_norm": 0.5206138592215499, + "learning_rate": 0.000531, + "loss": 7.2507, + "step": 177 + }, + { + "epoch": 0.00178, + "grad_norm": 0.5577166707239251, + "learning_rate": 0.000534, + "loss": 7.2428, + "step": 178 + }, + { + "epoch": 0.00179, + "grad_norm": 0.4980094016420717, + "learning_rate": 0.000537, + "loss": 7.2333, + "step": 179 + }, + { + "epoch": 0.0018, + "grad_norm": 0.5436132024149352, + "learning_rate": 0.00054, + "loss": 7.2053, + "step": 180 + }, + { + "epoch": 0.00181, + "grad_norm": 0.4850436660416719, + "learning_rate": 0.000543, + "loss": 7.1915, + "step": 181 + }, + { + "epoch": 0.00182, + "grad_norm": 0.686701132410735, + "learning_rate": 0.000546, + "loss": 7.1909, + "step": 182 + }, + { + "epoch": 0.00183, + "grad_norm": 0.68400997444651, + "learning_rate": 0.000549, + "loss": 7.1802, + "step": 183 + }, + { + "epoch": 0.00184, + "grad_norm": 0.8396027257351396, + "learning_rate": 0.000552, + "loss": 7.171, + "step": 184 + }, + { + "epoch": 0.00185, + "grad_norm": 0.7665365037410753, + "learning_rate": 0.000555, + "loss": 7.1476, + "step": 185 + }, + { + "epoch": 0.00186, + "grad_norm": 0.5359175667027454, + "learning_rate": 0.000558, + "loss": 7.1506, + "step": 186 + }, + { + "epoch": 0.00187, + "grad_norm": 0.5513800735908173, + "learning_rate": 0.000561, + "loss": 7.1261, + "step": 187 + }, + { + "epoch": 0.00188, + "grad_norm": 0.5352093004255375, + "learning_rate": 0.000564, + "loss": 7.1044, + "step": 188 + }, + { + "epoch": 0.00189, + "grad_norm": 0.5938457818726526, + "learning_rate": 0.000567, + "loss": 7.1004, + "step": 189 + }, + { + "epoch": 0.0019, + "grad_norm": 0.5273842405001533, + "learning_rate": 0.00057, + "loss": 7.0834, + "step": 190 + }, + { + "epoch": 0.00191, + "grad_norm": 0.47487115515279366, + "learning_rate": 0.000573, + "loss": 7.0721, + "step": 191 + }, + { + "epoch": 0.00192, + "grad_norm": 0.6036837698051599, + "learning_rate": 0.000576, + "loss": 7.0655, + "step": 192 + }, + { + "epoch": 0.00193, + "grad_norm": 0.3951184212196986, + "learning_rate": 0.000579, + "loss": 7.061, + "step": 193 + }, + { + "epoch": 0.00194, + "grad_norm": 0.442083567688087, + "learning_rate": 0.000582, + "loss": 7.0548, + "step": 194 + }, + { + "epoch": 0.00195, + "grad_norm": 0.439546882682468, + "learning_rate": 0.000585, + "loss": 7.0348, + "step": 195 + }, + { + "epoch": 0.00196, + "grad_norm": 0.46247531692771043, + "learning_rate": 0.000588, + "loss": 7.0228, + "step": 196 + }, + { + "epoch": 0.00197, + "grad_norm": 0.4140335072217301, + "learning_rate": 0.000591, + "loss": 7.0171, + "step": 197 + }, + { + "epoch": 0.00198, + "grad_norm": 0.3685986320410548, + "learning_rate": 0.000594, + "loss": 7.0081, + "step": 198 + }, + { + "epoch": 0.00199, + "grad_norm": 0.4020373564129086, + "learning_rate": 0.0005970000000000001, + "loss": 6.9898, + "step": 199 + }, + { + "epoch": 0.002, + "grad_norm": 0.37126410475941546, + "learning_rate": 0.0006000000000000001, + "loss": 6.9867, + "step": 200 + }, + { + "epoch": 0.00201, + "grad_norm": 0.3773154493828028, + "learning_rate": 0.000603, + "loss": 6.9617, + "step": 201 + }, + { + "epoch": 0.00202, + "grad_norm": 0.3540017416986532, + "learning_rate": 0.0006060000000000001, + "loss": 6.9491, + "step": 202 + }, + { + "epoch": 0.00203, + "grad_norm": 0.403279648640721, + "learning_rate": 0.0006090000000000001, + "loss": 6.9534, + "step": 203 + }, + { + "epoch": 0.00204, + "grad_norm": 0.5112949618253247, + "learning_rate": 0.000612, + "loss": 6.9385, + "step": 204 + }, + { + "epoch": 0.00205, + "grad_norm": 0.7200998739972175, + "learning_rate": 0.000615, + "loss": 6.931, + "step": 205 + }, + { + "epoch": 0.00206, + "grad_norm": 1.209379808685074, + "learning_rate": 0.000618, + "loss": 6.9351, + "step": 206 + }, + { + "epoch": 0.00207, + "grad_norm": 1.0040442357645134, + "learning_rate": 0.000621, + "loss": 6.9279, + "step": 207 + }, + { + "epoch": 0.00208, + "grad_norm": 0.7878464521205251, + "learning_rate": 0.000624, + "loss": 6.9129, + "step": 208 + }, + { + "epoch": 0.00209, + "grad_norm": 1.3096135062236434, + "learning_rate": 0.000627, + "loss": 6.9067, + "step": 209 + }, + { + "epoch": 0.0021, + "grad_norm": 0.910504669978176, + "learning_rate": 0.00063, + "loss": 6.8878, + "step": 210 + }, + { + "epoch": 0.00211, + "grad_norm": 1.3015817966038044, + "learning_rate": 0.000633, + "loss": 6.8987, + "step": 211 + }, + { + "epoch": 0.00212, + "grad_norm": 0.5587575104994011, + "learning_rate": 0.000636, + "loss": 6.8696, + "step": 212 + }, + { + "epoch": 0.00213, + "grad_norm": 0.8081412049208773, + "learning_rate": 0.000639, + "loss": 6.8608, + "step": 213 + }, + { + "epoch": 0.00214, + "grad_norm": 0.6397075273457759, + "learning_rate": 0.000642, + "loss": 6.8511, + "step": 214 + }, + { + "epoch": 0.00215, + "grad_norm": 0.5875043250740225, + "learning_rate": 0.000645, + "loss": 6.8482, + "step": 215 + }, + { + "epoch": 0.00216, + "grad_norm": 0.6060774535893669, + "learning_rate": 0.000648, + "loss": 6.8413, + "step": 216 + }, + { + "epoch": 0.00217, + "grad_norm": 0.5183751970166313, + "learning_rate": 0.000651, + "loss": 6.8119, + "step": 217 + }, + { + "epoch": 0.00218, + "grad_norm": 0.5539011900924167, + "learning_rate": 0.000654, + "loss": 6.8171, + "step": 218 + }, + { + "epoch": 0.00219, + "grad_norm": 0.5004312163301685, + "learning_rate": 0.000657, + "loss": 6.7882, + "step": 219 + }, + { + "epoch": 0.0022, + "grad_norm": 0.4762494747244133, + "learning_rate": 0.00066, + "loss": 6.8062, + "step": 220 + }, + { + "epoch": 0.00221, + "grad_norm": 0.353452768224107, + "learning_rate": 0.0006630000000000001, + "loss": 6.7814, + "step": 221 + }, + { + "epoch": 0.00222, + "grad_norm": 0.3879096289808107, + "learning_rate": 0.000666, + "loss": 6.7696, + "step": 222 + }, + { + "epoch": 0.00223, + "grad_norm": 0.4253518811476648, + "learning_rate": 0.000669, + "loss": 6.7658, + "step": 223 + }, + { + "epoch": 0.00224, + "grad_norm": 0.4076338977034843, + "learning_rate": 0.0006720000000000001, + "loss": 6.7609, + "step": 224 + }, + { + "epoch": 0.00225, + "grad_norm": 0.4491881195721512, + "learning_rate": 0.000675, + "loss": 6.7489, + "step": 225 + }, + { + "epoch": 0.00226, + "grad_norm": 0.4195112098951784, + "learning_rate": 0.000678, + "loss": 6.7444, + "step": 226 + }, + { + "epoch": 0.00227, + "grad_norm": 0.371663908330708, + "learning_rate": 0.0006810000000000001, + "loss": 6.7174, + "step": 227 + }, + { + "epoch": 0.00228, + "grad_norm": 0.3462952066328263, + "learning_rate": 0.000684, + "loss": 6.7197, + "step": 228 + }, + { + "epoch": 0.00229, + "grad_norm": 0.5288966810878937, + "learning_rate": 0.000687, + "loss": 6.7178, + "step": 229 + }, + { + "epoch": 0.0023, + "grad_norm": 0.7623592789505088, + "learning_rate": 0.0006900000000000001, + "loss": 6.6993, + "step": 230 + }, + { + "epoch": 0.00231, + "grad_norm": 0.9880780315432149, + "learning_rate": 0.000693, + "loss": 6.6923, + "step": 231 + }, + { + "epoch": 0.00232, + "grad_norm": 1.028691756937799, + "learning_rate": 0.000696, + "loss": 6.7142, + "step": 232 + }, + { + "epoch": 0.00233, + "grad_norm": 0.9501761132787978, + "learning_rate": 0.0006990000000000001, + "loss": 6.6946, + "step": 233 + }, + { + "epoch": 0.00234, + "grad_norm": 0.9999298841530961, + "learning_rate": 0.000702, + "loss": 6.6948, + "step": 234 + }, + { + "epoch": 0.00235, + "grad_norm": 0.6945930996245869, + "learning_rate": 0.000705, + "loss": 6.6675, + "step": 235 + }, + { + "epoch": 0.00236, + "grad_norm": 0.6422608041910052, + "learning_rate": 0.000708, + "loss": 6.6513, + "step": 236 + }, + { + "epoch": 0.00237, + "grad_norm": 0.47985797682797676, + "learning_rate": 0.0007109999999999999, + "loss": 6.657, + "step": 237 + }, + { + "epoch": 0.00238, + "grad_norm": 0.6191094472441181, + "learning_rate": 0.000714, + "loss": 6.6482, + "step": 238 + }, + { + "epoch": 0.00239, + "grad_norm": 0.5181523072026278, + "learning_rate": 0.000717, + "loss": 6.628, + "step": 239 + }, + { + "epoch": 0.0024, + "grad_norm": 0.5894783578801835, + "learning_rate": 0.0007199999999999999, + "loss": 6.645, + "step": 240 + }, + { + "epoch": 0.00241, + "grad_norm": 0.5639698176641863, + "learning_rate": 0.000723, + "loss": 6.6279, + "step": 241 + }, + { + "epoch": 0.00242, + "grad_norm": 0.9095170652542525, + "learning_rate": 0.000726, + "loss": 6.6087, + "step": 242 + }, + { + "epoch": 0.00243, + "grad_norm": 1.3373514416355459, + "learning_rate": 0.000729, + "loss": 6.6191, + "step": 243 + }, + { + "epoch": 0.00244, + "grad_norm": 0.5837452319331187, + "learning_rate": 0.000732, + "loss": 6.5991, + "step": 244 + }, + { + "epoch": 0.00245, + "grad_norm": 1.0261084178764917, + "learning_rate": 0.000735, + "loss": 6.6035, + "step": 245 + }, + { + "epoch": 0.00246, + "grad_norm": 0.5663979442820745, + "learning_rate": 0.000738, + "loss": 6.574, + "step": 246 + }, + { + "epoch": 0.00247, + "grad_norm": 0.5896636858778472, + "learning_rate": 0.000741, + "loss": 6.5719, + "step": 247 + }, + { + "epoch": 0.00248, + "grad_norm": 0.47326080430149503, + "learning_rate": 0.000744, + "loss": 6.5757, + "step": 248 + }, + { + "epoch": 0.00249, + "grad_norm": 0.6248164732961499, + "learning_rate": 0.000747, + "loss": 6.5582, + "step": 249 + }, + { + "epoch": 0.0025, + "grad_norm": 0.48362025373458484, + "learning_rate": 0.00075, + "loss": 6.5567, + "step": 250 + }, + { + "epoch": 0.00251, + "grad_norm": 0.4858269118610639, + "learning_rate": 0.000753, + "loss": 6.5227, + "step": 251 + }, + { + "epoch": 0.00252, + "grad_norm": 0.41862369731289734, + "learning_rate": 0.000756, + "loss": 6.5302, + "step": 252 + }, + { + "epoch": 0.00253, + "grad_norm": 0.47200180330590646, + "learning_rate": 0.000759, + "loss": 6.522, + "step": 253 + }, + { + "epoch": 0.00254, + "grad_norm": 0.3651546358356223, + "learning_rate": 0.000762, + "loss": 6.5136, + "step": 254 + }, + { + "epoch": 0.00255, + "grad_norm": 0.4285192093499382, + "learning_rate": 0.0007650000000000001, + "loss": 6.5149, + "step": 255 + }, + { + "epoch": 0.00256, + "grad_norm": 0.42767490819993825, + "learning_rate": 0.000768, + "loss": 6.4917, + "step": 256 + }, + { + "epoch": 0.00257, + "grad_norm": 0.48143237273687123, + "learning_rate": 0.000771, + "loss": 6.4974, + "step": 257 + }, + { + "epoch": 0.00258, + "grad_norm": 0.5467090762573645, + "learning_rate": 0.0007740000000000001, + "loss": 6.4744, + "step": 258 + }, + { + "epoch": 0.00259, + "grad_norm": 0.6750367098607939, + "learning_rate": 0.000777, + "loss": 6.4781, + "step": 259 + }, + { + "epoch": 0.0026, + "grad_norm": 0.8320211245129605, + "learning_rate": 0.0007800000000000001, + "loss": 6.4681, + "step": 260 + }, + { + "epoch": 0.00261, + "grad_norm": 0.905841412497731, + "learning_rate": 0.0007830000000000001, + "loss": 6.4809, + "step": 261 + }, + { + "epoch": 0.00262, + "grad_norm": 1.0687360869781928, + "learning_rate": 0.000786, + "loss": 6.4647, + "step": 262 + }, + { + "epoch": 0.00263, + "grad_norm": 1.2051408681263374, + "learning_rate": 0.0007890000000000001, + "loss": 6.4874, + "step": 263 + }, + { + "epoch": 0.00264, + "grad_norm": 0.8690142485653533, + "learning_rate": 0.0007920000000000001, + "loss": 6.455, + "step": 264 + }, + { + "epoch": 0.00265, + "grad_norm": 1.2774066489819682, + "learning_rate": 0.000795, + "loss": 6.4641, + "step": 265 + }, + { + "epoch": 0.00266, + "grad_norm": 0.7288978979341997, + "learning_rate": 0.0007980000000000001, + "loss": 6.4454, + "step": 266 + }, + { + "epoch": 0.00267, + "grad_norm": 0.6506287971604123, + "learning_rate": 0.0008010000000000001, + "loss": 6.4371, + "step": 267 + }, + { + "epoch": 0.00268, + "grad_norm": 0.6866545943797145, + "learning_rate": 0.000804, + "loss": 6.4338, + "step": 268 + }, + { + "epoch": 0.00269, + "grad_norm": 1.0440114151727509, + "learning_rate": 0.0008070000000000001, + "loss": 6.4227, + "step": 269 + }, + { + "epoch": 0.0027, + "grad_norm": 1.170351969791303, + "learning_rate": 0.0008100000000000001, + "loss": 6.4362, + "step": 270 + }, + { + "epoch": 0.00271, + "grad_norm": 0.6832222674646221, + "learning_rate": 0.000813, + "loss": 6.4253, + "step": 271 + }, + { + "epoch": 0.00272, + "grad_norm": 0.617936452008115, + "learning_rate": 0.0008160000000000001, + "loss": 6.4063, + "step": 272 + }, + { + "epoch": 0.00273, + "grad_norm": 0.6750066600068259, + "learning_rate": 0.0008190000000000001, + "loss": 6.4008, + "step": 273 + }, + { + "epoch": 0.00274, + "grad_norm": 0.6315560378177079, + "learning_rate": 0.000822, + "loss": 6.379, + "step": 274 + }, + { + "epoch": 0.00275, + "grad_norm": 0.6774266517760174, + "learning_rate": 0.0008250000000000001, + "loss": 6.3852, + "step": 275 + }, + { + "epoch": 0.00276, + "grad_norm": 0.7272895822836024, + "learning_rate": 0.0008280000000000001, + "loss": 6.3794, + "step": 276 + }, + { + "epoch": 0.00277, + "grad_norm": 0.6752471037637485, + "learning_rate": 0.0008310000000000001, + "loss": 6.3735, + "step": 277 + }, + { + "epoch": 0.00278, + "grad_norm": 0.5678457826039285, + "learning_rate": 0.0008340000000000001, + "loss": 6.354, + "step": 278 + }, + { + "epoch": 0.00279, + "grad_norm": 0.4611700189072147, + "learning_rate": 0.0008370000000000001, + "loss": 6.3529, + "step": 279 + }, + { + "epoch": 0.0028, + "grad_norm": 0.343285643042232, + "learning_rate": 0.0008400000000000001, + "loss": 6.3329, + "step": 280 + }, + { + "epoch": 0.00281, + "grad_norm": 0.4519631747446028, + "learning_rate": 0.0008430000000000001, + "loss": 6.3253, + "step": 281 + }, + { + "epoch": 0.00282, + "grad_norm": 0.3255189118052276, + "learning_rate": 0.000846, + "loss": 6.3232, + "step": 282 + }, + { + "epoch": 0.00283, + "grad_norm": 0.4297016476682907, + "learning_rate": 0.0008489999999999999, + "loss": 6.3149, + "step": 283 + }, + { + "epoch": 0.00284, + "grad_norm": 0.37515936714697207, + "learning_rate": 0.0008519999999999999, + "loss": 6.3058, + "step": 284 + }, + { + "epoch": 0.00285, + "grad_norm": 0.3458870104505622, + "learning_rate": 0.000855, + "loss": 6.3075, + "step": 285 + }, + { + "epoch": 0.00286, + "grad_norm": 0.39943359732325345, + "learning_rate": 0.0008579999999999999, + "loss": 6.2857, + "step": 286 + }, + { + "epoch": 0.00287, + "grad_norm": 0.3631751365570726, + "learning_rate": 0.000861, + "loss": 6.2875, + "step": 287 + }, + { + "epoch": 0.00288, + "grad_norm": 0.552280032213235, + "learning_rate": 0.000864, + "loss": 6.2707, + "step": 288 + }, + { + "epoch": 0.00289, + "grad_norm": 0.9396362484724781, + "learning_rate": 0.0008669999999999999, + "loss": 6.2923, + "step": 289 + }, + { + "epoch": 0.0029, + "grad_norm": 1.3959756273937387, + "learning_rate": 0.00087, + "loss": 6.2945, + "step": 290 + }, + { + "epoch": 0.00291, + "grad_norm": 0.6791923155853262, + "learning_rate": 0.000873, + "loss": 6.2663, + "step": 291 + }, + { + "epoch": 0.00292, + "grad_norm": 0.9879732835250642, + "learning_rate": 0.0008759999999999999, + "loss": 6.2794, + "step": 292 + }, + { + "epoch": 0.00293, + "grad_norm": 1.0140384817226566, + "learning_rate": 0.000879, + "loss": 6.2652, + "step": 293 + }, + { + "epoch": 0.00294, + "grad_norm": 0.9889225551341856, + "learning_rate": 0.000882, + "loss": 6.2822, + "step": 294 + }, + { + "epoch": 0.00295, + "grad_norm": 1.0429477871581094, + "learning_rate": 0.0008849999999999999, + "loss": 6.2468, + "step": 295 + }, + { + "epoch": 0.00296, + "grad_norm": 1.1774929945794055, + "learning_rate": 0.000888, + "loss": 6.2705, + "step": 296 + }, + { + "epoch": 0.00297, + "grad_norm": 0.7364156162079134, + "learning_rate": 0.000891, + "loss": 6.2278, + "step": 297 + }, + { + "epoch": 0.00298, + "grad_norm": 0.8424638476384282, + "learning_rate": 0.0008939999999999999, + "loss": 6.2455, + "step": 298 + }, + { + "epoch": 0.00299, + "grad_norm": 0.8668489286879963, + "learning_rate": 0.000897, + "loss": 6.2225, + "step": 299 + }, + { + "epoch": 0.003, + "grad_norm": 0.9039057952602142, + "learning_rate": 0.0009, + "loss": 6.2236, + "step": 300 + }, + { + "epoch": 0.00301, + "grad_norm": 1.0344858724084711, + "learning_rate": 0.0009029999999999999, + "loss": 6.2222, + "step": 301 + }, + { + "epoch": 0.00302, + "grad_norm": 0.801399966246171, + "learning_rate": 0.000906, + "loss": 6.2007, + "step": 302 + }, + { + "epoch": 0.00303, + "grad_norm": 0.7276255151675343, + "learning_rate": 0.000909, + "loss": 6.2106, + "step": 303 + }, + { + "epoch": 0.00304, + "grad_norm": 0.8306162070729353, + "learning_rate": 0.000912, + "loss": 6.205, + "step": 304 + }, + { + "epoch": 0.00305, + "grad_norm": 0.7650178489304597, + "learning_rate": 0.000915, + "loss": 6.2045, + "step": 305 + }, + { + "epoch": 0.00306, + "grad_norm": 0.7024963687074245, + "learning_rate": 0.000918, + "loss": 6.1878, + "step": 306 + }, + { + "epoch": 0.00307, + "grad_norm": 0.5687961336654864, + "learning_rate": 0.000921, + "loss": 6.1778, + "step": 307 + }, + { + "epoch": 0.00308, + "grad_norm": 0.4515866134049927, + "learning_rate": 0.000924, + "loss": 6.1586, + "step": 308 + }, + { + "epoch": 0.00309, + "grad_norm": 0.5454081565882548, + "learning_rate": 0.000927, + "loss": 6.1763, + "step": 309 + }, + { + "epoch": 0.0031, + "grad_norm": 0.4033959865123679, + "learning_rate": 0.00093, + "loss": 6.1549, + "step": 310 + }, + { + "epoch": 0.00311, + "grad_norm": 0.5004494164305024, + "learning_rate": 0.000933, + "loss": 6.1478, + "step": 311 + }, + { + "epoch": 0.00312, + "grad_norm": 0.470361361901935, + "learning_rate": 0.000936, + "loss": 6.1275, + "step": 312 + }, + { + "epoch": 0.00313, + "grad_norm": 0.4973667290148138, + "learning_rate": 0.0009390000000000001, + "loss": 6.1399, + "step": 313 + }, + { + "epoch": 0.00314, + "grad_norm": 0.47398584855820086, + "learning_rate": 0.000942, + "loss": 6.1453, + "step": 314 + }, + { + "epoch": 0.00315, + "grad_norm": 0.39081338977861474, + "learning_rate": 0.000945, + "loss": 6.1206, + "step": 315 + }, + { + "epoch": 0.00316, + "grad_norm": 0.308452368547838, + "learning_rate": 0.0009480000000000001, + "loss": 6.0925, + "step": 316 + }, + { + "epoch": 0.00317, + "grad_norm": 0.33735261718955184, + "learning_rate": 0.000951, + "loss": 6.1112, + "step": 317 + }, + { + "epoch": 0.00318, + "grad_norm": 0.3843792856632324, + "learning_rate": 0.000954, + "loss": 6.1055, + "step": 318 + }, + { + "epoch": 0.00319, + "grad_norm": 0.45015697169720664, + "learning_rate": 0.0009570000000000001, + "loss": 6.0951, + "step": 319 + }, + { + "epoch": 0.0032, + "grad_norm": 0.6347836654615971, + "learning_rate": 0.00096, + "loss": 6.097, + "step": 320 + }, + { + "epoch": 0.00321, + "grad_norm": 0.9264517947757075, + "learning_rate": 0.000963, + "loss": 6.0955, + "step": 321 + }, + { + "epoch": 0.00322, + "grad_norm": 1.3311662164937155, + "learning_rate": 0.0009660000000000001, + "loss": 6.0984, + "step": 322 + }, + { + "epoch": 0.00323, + "grad_norm": 0.7192768486088142, + "learning_rate": 0.000969, + "loss": 6.081, + "step": 323 + }, + { + "epoch": 0.00324, + "grad_norm": 0.8815686637890671, + "learning_rate": 0.0009720000000000001, + "loss": 6.0921, + "step": 324 + }, + { + "epoch": 0.00325, + "grad_norm": 0.9399377971403509, + "learning_rate": 0.0009750000000000001, + "loss": 6.0559, + "step": 325 + }, + { + "epoch": 0.00326, + "grad_norm": 1.1161084705724094, + "learning_rate": 0.0009780000000000001, + "loss": 6.0866, + "step": 326 + }, + { + "epoch": 0.00327, + "grad_norm": 0.9798577225908143, + "learning_rate": 0.000981, + "loss": 6.08, + "step": 327 + }, + { + "epoch": 0.00328, + "grad_norm": 1.4594200076446016, + "learning_rate": 0.000984, + "loss": 6.0949, + "step": 328 + }, + { + "epoch": 0.00329, + "grad_norm": 0.7530414187502621, + "learning_rate": 0.000987, + "loss": 6.0592, + "step": 329 + }, + { + "epoch": 0.0033, + "grad_norm": 0.9983029089238832, + "learning_rate": 0.00099, + "loss": 6.0707, + "step": 330 + }, + { + "epoch": 0.00331, + "grad_norm": 0.835334238631822, + "learning_rate": 0.0009930000000000002, + "loss": 6.0609, + "step": 331 + }, + { + "epoch": 0.00332, + "grad_norm": 1.2633352311024129, + "learning_rate": 0.0009960000000000001, + "loss": 6.0632, + "step": 332 + }, + { + "epoch": 0.00333, + "grad_norm": 0.9415989578137678, + "learning_rate": 0.000999, + "loss": 6.0388, + "step": 333 + }, + { + "epoch": 0.00334, + "grad_norm": 0.752057040577036, + "learning_rate": 0.001002, + "loss": 6.0434, + "step": 334 + }, + { + "epoch": 0.00335, + "grad_norm": 0.6368712711427076, + "learning_rate": 0.001005, + "loss": 6.039, + "step": 335 + }, + { + "epoch": 0.00336, + "grad_norm": 0.5277321252477392, + "learning_rate": 0.001008, + "loss": 6.0255, + "step": 336 + }, + { + "epoch": 0.00337, + "grad_norm": 0.545640031716998, + "learning_rate": 0.0010110000000000002, + "loss": 6.0051, + "step": 337 + }, + { + "epoch": 0.00338, + "grad_norm": 0.5605817392374447, + "learning_rate": 0.0010140000000000001, + "loss": 6.0183, + "step": 338 + }, + { + "epoch": 0.00339, + "grad_norm": 0.4963166785854256, + "learning_rate": 0.0010170000000000001, + "loss": 6.0214, + "step": 339 + }, + { + "epoch": 0.0034, + "grad_norm": 0.496225629593103, + "learning_rate": 0.00102, + "loss": 5.9915, + "step": 340 + }, + { + "epoch": 0.00341, + "grad_norm": 0.45013704339915594, + "learning_rate": 0.001023, + "loss": 5.9719, + "step": 341 + }, + { + "epoch": 0.00342, + "grad_norm": 0.37899658405778774, + "learning_rate": 0.001026, + "loss": 5.9765, + "step": 342 + }, + { + "epoch": 0.00343, + "grad_norm": 0.4216072327604619, + "learning_rate": 0.0010290000000000002, + "loss": 5.9773, + "step": 343 + }, + { + "epoch": 0.00344, + "grad_norm": 0.4964437378733662, + "learning_rate": 0.001032, + "loss": 5.9678, + "step": 344 + }, + { + "epoch": 0.00345, + "grad_norm": 0.6031142308234815, + "learning_rate": 0.001035, + "loss": 5.9605, + "step": 345 + }, + { + "epoch": 0.00346, + "grad_norm": 0.6349778009023861, + "learning_rate": 0.0010379999999999999, + "loss": 5.9675, + "step": 346 + }, + { + "epoch": 0.00347, + "grad_norm": 0.6213366174245898, + "learning_rate": 0.001041, + "loss": 5.9393, + "step": 347 + }, + { + "epoch": 0.00348, + "grad_norm": 0.684857817555668, + "learning_rate": 0.001044, + "loss": 5.9499, + "step": 348 + }, + { + "epoch": 0.00349, + "grad_norm": 0.8421661368559449, + "learning_rate": 0.001047, + "loss": 5.9398, + "step": 349 + }, + { + "epoch": 0.0035, + "grad_norm": 0.8062802732253019, + "learning_rate": 0.00105, + "loss": 5.9568, + "step": 350 + }, + { + "epoch": 0.00351, + "grad_norm": 0.7669273491138234, + "learning_rate": 0.001053, + "loss": 5.9549, + "step": 351 + }, + { + "epoch": 0.00352, + "grad_norm": 1.0673384228730578, + "learning_rate": 0.0010559999999999999, + "loss": 5.9491, + "step": 352 + }, + { + "epoch": 0.00353, + "grad_norm": 1.313057737158452, + "learning_rate": 0.001059, + "loss": 5.9596, + "step": 353 + }, + { + "epoch": 0.00354, + "grad_norm": 0.8748293544965601, + "learning_rate": 0.001062, + "loss": 5.9228, + "step": 354 + }, + { + "epoch": 0.00355, + "grad_norm": 0.7617258072087358, + "learning_rate": 0.001065, + "loss": 5.9348, + "step": 355 + }, + { + "epoch": 0.00356, + "grad_norm": 0.9561923051718775, + "learning_rate": 0.001068, + "loss": 5.9306, + "step": 356 + }, + { + "epoch": 0.00357, + "grad_norm": 0.8631216371911231, + "learning_rate": 0.001071, + "loss": 5.9199, + "step": 357 + }, + { + "epoch": 0.00358, + "grad_norm": 0.9525953351992685, + "learning_rate": 0.001074, + "loss": 5.9184, + "step": 358 + }, + { + "epoch": 0.00359, + "grad_norm": 0.6858307664974747, + "learning_rate": 0.001077, + "loss": 5.8973, + "step": 359 + }, + { + "epoch": 0.0036, + "grad_norm": 0.7167037627907079, + "learning_rate": 0.00108, + "loss": 5.9065, + "step": 360 + }, + { + "epoch": 0.00361, + "grad_norm": 0.7969047134484166, + "learning_rate": 0.001083, + "loss": 5.8986, + "step": 361 + }, + { + "epoch": 0.00362, + "grad_norm": 0.8186204145394074, + "learning_rate": 0.001086, + "loss": 5.888, + "step": 362 + }, + { + "epoch": 0.00363, + "grad_norm": 0.7436182013517663, + "learning_rate": 0.001089, + "loss": 5.8766, + "step": 363 + }, + { + "epoch": 0.00364, + "grad_norm": 0.6658023261534547, + "learning_rate": 0.001092, + "loss": 5.8872, + "step": 364 + }, + { + "epoch": 0.00365, + "grad_norm": 0.7206668214896482, + "learning_rate": 0.001095, + "loss": 5.8872, + "step": 365 + }, + { + "epoch": 0.00366, + "grad_norm": 0.6345413922647961, + "learning_rate": 0.001098, + "loss": 5.8617, + "step": 366 + }, + { + "epoch": 0.00367, + "grad_norm": 0.6094924231218852, + "learning_rate": 0.001101, + "loss": 5.8618, + "step": 367 + }, + { + "epoch": 0.00368, + "grad_norm": 0.7732260654201254, + "learning_rate": 0.001104, + "loss": 5.8653, + "step": 368 + }, + { + "epoch": 0.00369, + "grad_norm": 0.9451132874875877, + "learning_rate": 0.001107, + "loss": 5.8779, + "step": 369 + }, + { + "epoch": 0.0037, + "grad_norm": 1.047387946938651, + "learning_rate": 0.00111, + "loss": 5.8603, + "step": 370 + }, + { + "epoch": 0.00371, + "grad_norm": 0.8709497379832931, + "learning_rate": 0.001113, + "loss": 5.8262, + "step": 371 + }, + { + "epoch": 0.00372, + "grad_norm": 0.6993621521276565, + "learning_rate": 0.001116, + "loss": 5.8434, + "step": 372 + }, + { + "epoch": 0.00373, + "grad_norm": 0.6889019297226876, + "learning_rate": 0.001119, + "loss": 5.8344, + "step": 373 + }, + { + "epoch": 0.00374, + "grad_norm": 0.8187808414499578, + "learning_rate": 0.001122, + "loss": 5.8424, + "step": 374 + }, + { + "epoch": 0.00375, + "grad_norm": 0.7395556739972736, + "learning_rate": 0.0011250000000000001, + "loss": 5.8404, + "step": 375 + }, + { + "epoch": 0.00376, + "grad_norm": 0.5027446756427529, + "learning_rate": 0.001128, + "loss": 5.8296, + "step": 376 + }, + { + "epoch": 0.00377, + "grad_norm": 0.45297103786338255, + "learning_rate": 0.001131, + "loss": 5.8239, + "step": 377 + }, + { + "epoch": 0.00378, + "grad_norm": 0.5150298222384522, + "learning_rate": 0.001134, + "loss": 5.82, + "step": 378 + }, + { + "epoch": 0.00379, + "grad_norm": 0.4216428009711753, + "learning_rate": 0.001137, + "loss": 5.8036, + "step": 379 + }, + { + "epoch": 0.0038, + "grad_norm": 0.43574801532624385, + "learning_rate": 0.00114, + "loss": 5.8234, + "step": 380 + }, + { + "epoch": 0.00381, + "grad_norm": 0.4737821231317218, + "learning_rate": 0.0011430000000000001, + "loss": 5.7908, + "step": 381 + }, + { + "epoch": 0.00382, + "grad_norm": 0.5236457098681065, + "learning_rate": 0.001146, + "loss": 5.7778, + "step": 382 + }, + { + "epoch": 0.00383, + "grad_norm": 0.5006340354259897, + "learning_rate": 0.001149, + "loss": 5.787, + "step": 383 + }, + { + "epoch": 0.00384, + "grad_norm": 0.5762330042414852, + "learning_rate": 0.001152, + "loss": 5.7915, + "step": 384 + }, + { + "epoch": 0.00385, + "grad_norm": 0.6623935085819848, + "learning_rate": 0.001155, + "loss": 5.7808, + "step": 385 + }, + { + "epoch": 0.00386, + "grad_norm": 0.6780931079980513, + "learning_rate": 0.001158, + "loss": 5.7864, + "step": 386 + }, + { + "epoch": 0.00387, + "grad_norm": 0.6283557380392781, + "learning_rate": 0.0011610000000000001, + "loss": 5.76, + "step": 387 + }, + { + "epoch": 0.00388, + "grad_norm": 0.8661947460887196, + "learning_rate": 0.001164, + "loss": 5.7642, + "step": 388 + }, + { + "epoch": 0.00389, + "grad_norm": 1.1877567285678448, + "learning_rate": 0.001167, + "loss": 5.7898, + "step": 389 + }, + { + "epoch": 0.0039, + "grad_norm": 0.5285139722189788, + "learning_rate": 0.00117, + "loss": 5.7491, + "step": 390 + }, + { + "epoch": 0.00391, + "grad_norm": 0.6430118692881639, + "learning_rate": 0.001173, + "loss": 5.7622, + "step": 391 + }, + { + "epoch": 0.00392, + "grad_norm": 0.6748100403552353, + "learning_rate": 0.001176, + "loss": 5.7713, + "step": 392 + }, + { + "epoch": 0.00393, + "grad_norm": 0.7344199345621275, + "learning_rate": 0.0011790000000000001, + "loss": 5.7448, + "step": 393 + }, + { + "epoch": 0.00394, + "grad_norm": 0.8611653440542995, + "learning_rate": 0.001182, + "loss": 5.7377, + "step": 394 + }, + { + "epoch": 0.00395, + "grad_norm": 1.0200403565527223, + "learning_rate": 0.001185, + "loss": 5.7297, + "step": 395 + }, + { + "epoch": 0.00396, + "grad_norm": 1.3219082304761296, + "learning_rate": 0.001188, + "loss": 5.7571, + "step": 396 + }, + { + "epoch": 0.00397, + "grad_norm": 1.0648841185793536, + "learning_rate": 0.001191, + "loss": 5.7452, + "step": 397 + }, + { + "epoch": 0.00398, + "grad_norm": 0.8818852137031271, + "learning_rate": 0.0011940000000000002, + "loss": 5.7476, + "step": 398 + }, + { + "epoch": 0.00399, + "grad_norm": 0.7229082271352473, + "learning_rate": 0.0011970000000000001, + "loss": 5.7455, + "step": 399 + }, + { + "epoch": 0.004, + "grad_norm": 0.7464785543636239, + "learning_rate": 0.0012000000000000001, + "loss": 5.7335, + "step": 400 + }, + { + "epoch": 0.00401, + "grad_norm": 0.788008045277313, + "learning_rate": 0.001203, + "loss": 5.7147, + "step": 401 + }, + { + "epoch": 0.00402, + "grad_norm": 0.8844811042429299, + "learning_rate": 0.001206, + "loss": 5.7255, + "step": 402 + }, + { + "epoch": 0.00403, + "grad_norm": 0.8786909754931423, + "learning_rate": 0.001209, + "loss": 5.7201, + "step": 403 + }, + { + "epoch": 0.00404, + "grad_norm": 0.9763167280044874, + "learning_rate": 0.0012120000000000002, + "loss": 5.7188, + "step": 404 + }, + { + "epoch": 0.00405, + "grad_norm": 0.9708490732732808, + "learning_rate": 0.0012150000000000002, + "loss": 5.7216, + "step": 405 + }, + { + "epoch": 0.00406, + "grad_norm": 0.7953769652219763, + "learning_rate": 0.0012180000000000001, + "loss": 5.6987, + "step": 406 + }, + { + "epoch": 0.00407, + "grad_norm": 0.7231086032909518, + "learning_rate": 0.0012209999999999999, + "loss": 5.6945, + "step": 407 + }, + { + "epoch": 0.00408, + "grad_norm": 0.6539281657127057, + "learning_rate": 0.001224, + "loss": 5.7108, + "step": 408 + }, + { + "epoch": 0.00409, + "grad_norm": 0.48725126694443294, + "learning_rate": 0.001227, + "loss": 5.6827, + "step": 409 + }, + { + "epoch": 0.0041, + "grad_norm": 0.5935989194477762, + "learning_rate": 0.00123, + "loss": 5.6771, + "step": 410 + }, + { + "epoch": 0.00411, + "grad_norm": 0.6524672694061662, + "learning_rate": 0.001233, + "loss": 5.6945, + "step": 411 + }, + { + "epoch": 0.00412, + "grad_norm": 0.5686252583506046, + "learning_rate": 0.001236, + "loss": 5.6796, + "step": 412 + }, + { + "epoch": 0.00413, + "grad_norm": 0.5206794027062916, + "learning_rate": 0.0012389999999999999, + "loss": 5.676, + "step": 413 + }, + { + "epoch": 0.00414, + "grad_norm": 0.5100681259570212, + "learning_rate": 0.001242, + "loss": 5.6561, + "step": 414 + }, + { + "epoch": 0.00415, + "grad_norm": 0.5323494958966052, + "learning_rate": 0.001245, + "loss": 5.6572, + "step": 415 + }, + { + "epoch": 0.00416, + "grad_norm": 0.593833682795931, + "learning_rate": 0.001248, + "loss": 5.6419, + "step": 416 + }, + { + "epoch": 0.00417, + "grad_norm": 0.5881502982050868, + "learning_rate": 0.001251, + "loss": 5.6711, + "step": 417 + }, + { + "epoch": 0.00418, + "grad_norm": 0.6571826832460801, + "learning_rate": 0.001254, + "loss": 5.6451, + "step": 418 + }, + { + "epoch": 0.00419, + "grad_norm": 0.8242389183582979, + "learning_rate": 0.0012569999999999999, + "loss": 5.6393, + "step": 419 + }, + { + "epoch": 0.0042, + "grad_norm": 0.7322278119135938, + "learning_rate": 0.00126, + "loss": 5.6531, + "step": 420 + }, + { + "epoch": 0.00421, + "grad_norm": 0.6080486142741918, + "learning_rate": 0.001263, + "loss": 5.6415, + "step": 421 + }, + { + "epoch": 0.00422, + "grad_norm": 0.6271805214461816, + "learning_rate": 0.001266, + "loss": 5.6327, + "step": 422 + }, + { + "epoch": 0.00423, + "grad_norm": 0.5934268893913589, + "learning_rate": 0.001269, + "loss": 5.6286, + "step": 423 + }, + { + "epoch": 0.00424, + "grad_norm": 0.6795296945522938, + "learning_rate": 0.001272, + "loss": 5.6299, + "step": 424 + }, + { + "epoch": 0.00425, + "grad_norm": 0.7845276448161947, + "learning_rate": 0.001275, + "loss": 5.6344, + "step": 425 + }, + { + "epoch": 0.00426, + "grad_norm": 0.9176562117983922, + "learning_rate": 0.001278, + "loss": 5.6402, + "step": 426 + }, + { + "epoch": 0.00427, + "grad_norm": 1.0306205830362438, + "learning_rate": 0.001281, + "loss": 5.6293, + "step": 427 + }, + { + "epoch": 0.00428, + "grad_norm": 1.0252381178312269, + "learning_rate": 0.001284, + "loss": 5.6086, + "step": 428 + }, + { + "epoch": 0.00429, + "grad_norm": 1.3332162612637855, + "learning_rate": 0.001287, + "loss": 5.6393, + "step": 429 + }, + { + "epoch": 0.0043, + "grad_norm": 0.8174291079939794, + "learning_rate": 0.00129, + "loss": 5.6202, + "step": 430 + }, + { + "epoch": 0.00431, + "grad_norm": 0.821028463418781, + "learning_rate": 0.001293, + "loss": 5.6102, + "step": 431 + }, + { + "epoch": 0.00432, + "grad_norm": 0.7475137161763143, + "learning_rate": 0.001296, + "loss": 5.6163, + "step": 432 + }, + { + "epoch": 0.00433, + "grad_norm": 0.7571870052992741, + "learning_rate": 0.001299, + "loss": 5.604, + "step": 433 + }, + { + "epoch": 0.00434, + "grad_norm": 1.0543560255015263, + "learning_rate": 0.001302, + "loss": 5.6184, + "step": 434 + }, + { + "epoch": 0.00435, + "grad_norm": 0.8758758304266553, + "learning_rate": 0.001305, + "loss": 5.6066, + "step": 435 + }, + { + "epoch": 0.00436, + "grad_norm": 0.9137062421440546, + "learning_rate": 0.001308, + "loss": 5.5859, + "step": 436 + }, + { + "epoch": 0.00437, + "grad_norm": 1.301736875083812, + "learning_rate": 0.001311, + "loss": 5.6173, + "step": 437 + }, + { + "epoch": 0.00438, + "grad_norm": 0.9356953917037294, + "learning_rate": 0.001314, + "loss": 5.5919, + "step": 438 + }, + { + "epoch": 0.00439, + "grad_norm": 0.8522821992819578, + "learning_rate": 0.001317, + "loss": 5.6158, + "step": 439 + }, + { + "epoch": 0.0044, + "grad_norm": 0.7182761753028103, + "learning_rate": 0.00132, + "loss": 5.5821, + "step": 440 + }, + { + "epoch": 0.00441, + "grad_norm": 0.6293266892726601, + "learning_rate": 0.001323, + "loss": 5.577, + "step": 441 + }, + { + "epoch": 0.00442, + "grad_norm": 0.8189921873128356, + "learning_rate": 0.0013260000000000001, + "loss": 5.5762, + "step": 442 + }, + { + "epoch": 0.00443, + "grad_norm": 1.0212422821054057, + "learning_rate": 0.001329, + "loss": 5.5904, + "step": 443 + }, + { + "epoch": 0.00444, + "grad_norm": 0.8951785018901781, + "learning_rate": 0.001332, + "loss": 5.5851, + "step": 444 + }, + { + "epoch": 0.00445, + "grad_norm": 0.7868769140150608, + "learning_rate": 0.001335, + "loss": 5.5661, + "step": 445 + }, + { + "epoch": 0.00446, + "grad_norm": 0.790102612629763, + "learning_rate": 0.001338, + "loss": 5.5718, + "step": 446 + }, + { + "epoch": 0.00447, + "grad_norm": 0.8396194874372788, + "learning_rate": 0.001341, + "loss": 5.5716, + "step": 447 + }, + { + "epoch": 0.00448, + "grad_norm": 0.9120841535821665, + "learning_rate": 0.0013440000000000001, + "loss": 5.5589, + "step": 448 + }, + { + "epoch": 0.00449, + "grad_norm": 0.8573073152890212, + "learning_rate": 0.001347, + "loss": 5.5603, + "step": 449 + }, + { + "epoch": 0.0045, + "grad_norm": 0.7420732830276576, + "learning_rate": 0.00135, + "loss": 5.5551, + "step": 450 + }, + { + "epoch": 0.00451, + "grad_norm": 0.7054051762730813, + "learning_rate": 0.001353, + "loss": 5.5451, + "step": 451 + }, + { + "epoch": 0.00452, + "grad_norm": 0.5383194985068459, + "learning_rate": 0.001356, + "loss": 5.5229, + "step": 452 + }, + { + "epoch": 0.00453, + "grad_norm": 0.5845224529530345, + "learning_rate": 0.001359, + "loss": 5.5144, + "step": 453 + }, + { + "epoch": 0.00454, + "grad_norm": 0.5197277965757966, + "learning_rate": 0.0013620000000000001, + "loss": 5.5269, + "step": 454 + }, + { + "epoch": 0.00455, + "grad_norm": 0.44901215490386587, + "learning_rate": 0.0013650000000000001, + "loss": 5.5227, + "step": 455 + }, + { + "epoch": 0.00456, + "grad_norm": 0.40823042735319937, + "learning_rate": 0.001368, + "loss": 5.5155, + "step": 456 + }, + { + "epoch": 0.00457, + "grad_norm": 0.4077054091161063, + "learning_rate": 0.001371, + "loss": 5.5045, + "step": 457 + }, + { + "epoch": 0.00458, + "grad_norm": 0.4051721280659754, + "learning_rate": 0.001374, + "loss": 5.5115, + "step": 458 + }, + { + "epoch": 0.00459, + "grad_norm": 0.3819222938644956, + "learning_rate": 0.0013770000000000002, + "loss": 5.4986, + "step": 459 + }, + { + "epoch": 0.0046, + "grad_norm": 0.4096791477469412, + "learning_rate": 0.0013800000000000002, + "loss": 5.4851, + "step": 460 + }, + { + "epoch": 0.00461, + "grad_norm": 0.44167018915391465, + "learning_rate": 0.0013830000000000001, + "loss": 5.4805, + "step": 461 + }, + { + "epoch": 0.00462, + "grad_norm": 0.49346826910496583, + "learning_rate": 0.001386, + "loss": 5.4928, + "step": 462 + }, + { + "epoch": 0.00463, + "grad_norm": 0.463489633996114, + "learning_rate": 0.001389, + "loss": 5.4652, + "step": 463 + }, + { + "epoch": 0.00464, + "grad_norm": 0.507205937400668, + "learning_rate": 0.001392, + "loss": 5.4859, + "step": 464 + }, + { + "epoch": 0.00465, + "grad_norm": 0.5318639728571777, + "learning_rate": 0.0013950000000000002, + "loss": 5.4572, + "step": 465 + }, + { + "epoch": 0.00466, + "grad_norm": 0.6489252257104292, + "learning_rate": 0.0013980000000000002, + "loss": 5.4792, + "step": 466 + }, + { + "epoch": 0.00467, + "grad_norm": 0.7957789059873086, + "learning_rate": 0.0014010000000000001, + "loss": 5.4727, + "step": 467 + }, + { + "epoch": 0.00468, + "grad_norm": 0.7621042481808248, + "learning_rate": 0.001404, + "loss": 5.462, + "step": 468 + }, + { + "epoch": 0.00469, + "grad_norm": 0.573822393217689, + "learning_rate": 0.001407, + "loss": 5.457, + "step": 469 + }, + { + "epoch": 0.0047, + "grad_norm": 0.7543746323219513, + "learning_rate": 0.00141, + "loss": 5.4709, + "step": 470 + }, + { + "epoch": 0.00471, + "grad_norm": 0.8641651762100534, + "learning_rate": 0.001413, + "loss": 5.4567, + "step": 471 + }, + { + "epoch": 0.00472, + "grad_norm": 0.9026674726231502, + "learning_rate": 0.001416, + "loss": 5.4699, + "step": 472 + }, + { + "epoch": 0.00473, + "grad_norm": 0.8212885101559565, + "learning_rate": 0.001419, + "loss": 5.4778, + "step": 473 + }, + { + "epoch": 0.00474, + "grad_norm": 0.8914030740906659, + "learning_rate": 0.0014219999999999999, + "loss": 5.461, + "step": 474 + }, + { + "epoch": 0.00475, + "grad_norm": 1.0570094425693455, + "learning_rate": 0.001425, + "loss": 5.4652, + "step": 475 + }, + { + "epoch": 0.00476, + "grad_norm": 0.9736444976311589, + "learning_rate": 0.001428, + "loss": 5.4875, + "step": 476 + }, + { + "epoch": 0.00477, + "grad_norm": 1.1550380737092787, + "learning_rate": 0.001431, + "loss": 5.4568, + "step": 477 + }, + { + "epoch": 0.00478, + "grad_norm": 0.848321570803796, + "learning_rate": 0.001434, + "loss": 5.4695, + "step": 478 + }, + { + "epoch": 0.00479, + "grad_norm": 0.9517827225501269, + "learning_rate": 0.001437, + "loss": 5.4501, + "step": 479 + }, + { + "epoch": 0.0048, + "grad_norm": 1.0883787540754652, + "learning_rate": 0.0014399999999999999, + "loss": 5.4562, + "step": 480 + }, + { + "epoch": 0.00481, + "grad_norm": 0.9422991164230814, + "learning_rate": 0.001443, + "loss": 5.4516, + "step": 481 + }, + { + "epoch": 0.00482, + "grad_norm": 0.9159499791385636, + "learning_rate": 0.001446, + "loss": 5.4273, + "step": 482 + }, + { + "epoch": 0.00483, + "grad_norm": 0.9688645055474768, + "learning_rate": 0.001449, + "loss": 5.44, + "step": 483 + }, + { + "epoch": 0.00484, + "grad_norm": 1.1114303023214132, + "learning_rate": 0.001452, + "loss": 5.4474, + "step": 484 + }, + { + "epoch": 0.00485, + "grad_norm": 0.9051569573634253, + "learning_rate": 0.001455, + "loss": 5.4468, + "step": 485 + }, + { + "epoch": 0.00486, + "grad_norm": 0.9247454458721566, + "learning_rate": 0.001458, + "loss": 5.4217, + "step": 486 + }, + { + "epoch": 0.00487, + "grad_norm": 0.852928162562673, + "learning_rate": 0.001461, + "loss": 5.4339, + "step": 487 + }, + { + "epoch": 0.00488, + "grad_norm": 0.8142513932978471, + "learning_rate": 0.001464, + "loss": 5.41, + "step": 488 + }, + { + "epoch": 0.00489, + "grad_norm": 0.9329231917883894, + "learning_rate": 0.001467, + "loss": 5.403, + "step": 489 + }, + { + "epoch": 0.0049, + "grad_norm": 0.9855383918059676, + "learning_rate": 0.00147, + "loss": 5.4217, + "step": 490 + }, + { + "epoch": 0.00491, + "grad_norm": 0.7649189111405409, + "learning_rate": 0.001473, + "loss": 5.4131, + "step": 491 + }, + { + "epoch": 0.00492, + "grad_norm": 0.7909135941762935, + "learning_rate": 0.001476, + "loss": 5.4088, + "step": 492 + }, + { + "epoch": 0.00493, + "grad_norm": 0.9895574066894314, + "learning_rate": 0.001479, + "loss": 5.412, + "step": 493 + }, + { + "epoch": 0.00494, + "grad_norm": 1.132991334341666, + "learning_rate": 0.001482, + "loss": 5.429, + "step": 494 + }, + { + "epoch": 0.00495, + "grad_norm": 0.761507737933228, + "learning_rate": 0.001485, + "loss": 5.4134, + "step": 495 + }, + { + "epoch": 0.00496, + "grad_norm": 0.9726159326361932, + "learning_rate": 0.001488, + "loss": 5.4067, + "step": 496 + }, + { + "epoch": 0.00497, + "grad_norm": 1.1482864163713484, + "learning_rate": 0.001491, + "loss": 5.3997, + "step": 497 + }, + { + "epoch": 0.00498, + "grad_norm": 0.8327332046897746, + "learning_rate": 0.001494, + "loss": 5.4147, + "step": 498 + }, + { + "epoch": 0.00499, + "grad_norm": 0.725916611519047, + "learning_rate": 0.001497, + "loss": 5.3792, + "step": 499 + }, + { + "epoch": 0.005, + "grad_norm": 0.6719782846245154, + "learning_rate": 0.0015, + "loss": 5.3842, + "step": 500 + }, + { + "epoch": 0.00501, + "grad_norm": 0.7463076928465905, + "learning_rate": 0.001503, + "loss": 5.3796, + "step": 501 + }, + { + "epoch": 0.00502, + "grad_norm": 0.8408104186601356, + "learning_rate": 0.001506, + "loss": 5.389, + "step": 502 + }, + { + "epoch": 0.00503, + "grad_norm": 0.860909611402231, + "learning_rate": 0.0015090000000000001, + "loss": 5.3926, + "step": 503 + }, + { + "epoch": 0.00504, + "grad_norm": 0.7026490274993983, + "learning_rate": 0.001512, + "loss": 5.3646, + "step": 504 + }, + { + "epoch": 0.00505, + "grad_norm": 0.7184807991697565, + "learning_rate": 0.001515, + "loss": 5.3547, + "step": 505 + }, + { + "epoch": 0.00506, + "grad_norm": 0.7839475253802514, + "learning_rate": 0.001518, + "loss": 5.3611, + "step": 506 + }, + { + "epoch": 0.00507, + "grad_norm": 0.7039499110993044, + "learning_rate": 0.001521, + "loss": 5.3522, + "step": 507 + }, + { + "epoch": 0.00508, + "grad_norm": 0.5587706859739108, + "learning_rate": 0.001524, + "loss": 5.3512, + "step": 508 + }, + { + "epoch": 0.00509, + "grad_norm": 0.4952941325228141, + "learning_rate": 0.0015270000000000001, + "loss": 5.326, + "step": 509 + }, + { + "epoch": 0.0051, + "grad_norm": 0.5131490664852795, + "learning_rate": 0.0015300000000000001, + "loss": 5.3428, + "step": 510 + }, + { + "epoch": 0.00511, + "grad_norm": 0.5539213487194597, + "learning_rate": 0.001533, + "loss": 5.3196, + "step": 511 + }, + { + "epoch": 0.00512, + "grad_norm": 0.5937876471208409, + "learning_rate": 0.001536, + "loss": 5.3158, + "step": 512 + }, + { + "epoch": 0.00513, + "grad_norm": 0.5441672327389838, + "learning_rate": 0.001539, + "loss": 5.327, + "step": 513 + }, + { + "epoch": 0.00514, + "grad_norm": 0.5371789671410057, + "learning_rate": 0.001542, + "loss": 5.3038, + "step": 514 + }, + { + "epoch": 0.00515, + "grad_norm": 0.5194765862771661, + "learning_rate": 0.0015450000000000001, + "loss": 5.3109, + "step": 515 + }, + { + "epoch": 0.00516, + "grad_norm": 0.5575198815834714, + "learning_rate": 0.0015480000000000001, + "loss": 5.3083, + "step": 516 + }, + { + "epoch": 0.00517, + "grad_norm": 0.5237583962443445, + "learning_rate": 0.001551, + "loss": 5.2962, + "step": 517 + }, + { + "epoch": 0.00518, + "grad_norm": 0.5237181167506534, + "learning_rate": 0.001554, + "loss": 5.2786, + "step": 518 + }, + { + "epoch": 0.00519, + "grad_norm": 0.6652146969315359, + "learning_rate": 0.001557, + "loss": 5.2847, + "step": 519 + }, + { + "epoch": 0.0052, + "grad_norm": 0.8817258231902962, + "learning_rate": 0.0015600000000000002, + "loss": 5.2966, + "step": 520 + }, + { + "epoch": 0.00521, + "grad_norm": 0.9047592596397305, + "learning_rate": 0.0015630000000000002, + "loss": 5.2732, + "step": 521 + }, + { + "epoch": 0.00522, + "grad_norm": 0.6139926424217688, + "learning_rate": 0.0015660000000000001, + "loss": 5.2701, + "step": 522 + }, + { + "epoch": 0.00523, + "grad_norm": 0.6292464658556638, + "learning_rate": 0.001569, + "loss": 5.2826, + "step": 523 + }, + { + "epoch": 0.00524, + "grad_norm": 0.752030715547053, + "learning_rate": 0.001572, + "loss": 5.2856, + "step": 524 + }, + { + "epoch": 0.00525, + "grad_norm": 0.929770536798091, + "learning_rate": 0.001575, + "loss": 5.287, + "step": 525 + }, + { + "epoch": 0.00526, + "grad_norm": 0.887370520628206, + "learning_rate": 0.0015780000000000002, + "loss": 5.2635, + "step": 526 + }, + { + "epoch": 0.00527, + "grad_norm": 0.7819104471156305, + "learning_rate": 0.0015810000000000002, + "loss": 5.2728, + "step": 527 + }, + { + "epoch": 0.00528, + "grad_norm": 0.9038037239389326, + "learning_rate": 0.0015840000000000001, + "loss": 5.2534, + "step": 528 + }, + { + "epoch": 0.00529, + "grad_norm": 0.8898068580069259, + "learning_rate": 0.001587, + "loss": 5.265, + "step": 529 + }, + { + "epoch": 0.0053, + "grad_norm": 1.0010848916209774, + "learning_rate": 0.00159, + "loss": 5.2764, + "step": 530 + }, + { + "epoch": 0.00531, + "grad_norm": 1.0010984765594055, + "learning_rate": 0.001593, + "loss": 5.2677, + "step": 531 + }, + { + "epoch": 0.00532, + "grad_norm": 1.026224264336229, + "learning_rate": 0.0015960000000000002, + "loss": 5.2779, + "step": 532 + }, + { + "epoch": 0.00533, + "grad_norm": 0.9759289805508353, + "learning_rate": 0.0015990000000000002, + "loss": 5.2678, + "step": 533 + }, + { + "epoch": 0.00534, + "grad_norm": 1.1376558518204782, + "learning_rate": 0.0016020000000000001, + "loss": 5.2612, + "step": 534 + }, + { + "epoch": 0.00535, + "grad_norm": 1.0517455067238486, + "learning_rate": 0.001605, + "loss": 5.2672, + "step": 535 + }, + { + "epoch": 0.00536, + "grad_norm": 0.9398243649272562, + "learning_rate": 0.001608, + "loss": 5.2627, + "step": 536 + }, + { + "epoch": 0.00537, + "grad_norm": 0.9512995727424398, + "learning_rate": 0.0016110000000000002, + "loss": 5.2558, + "step": 537 + }, + { + "epoch": 0.00538, + "grad_norm": 1.2746657535312511, + "learning_rate": 0.0016140000000000002, + "loss": 5.2451, + "step": 538 + }, + { + "epoch": 0.00539, + "grad_norm": 1.067518036012326, + "learning_rate": 0.0016170000000000002, + "loss": 5.2415, + "step": 539 + }, + { + "epoch": 0.0054, + "grad_norm": 1.1859628541053806, + "learning_rate": 0.0016200000000000001, + "loss": 5.2618, + "step": 540 + }, + { + "epoch": 0.00541, + "grad_norm": 0.8882936824028492, + "learning_rate": 0.001623, + "loss": 5.2308, + "step": 541 + }, + { + "epoch": 0.00542, + "grad_norm": 0.8517075205384302, + "learning_rate": 0.001626, + "loss": 5.2545, + "step": 542 + }, + { + "epoch": 0.00543, + "grad_norm": 0.8283552605034004, + "learning_rate": 0.0016290000000000002, + "loss": 5.2098, + "step": 543 + }, + { + "epoch": 0.00544, + "grad_norm": 0.9087829134911761, + "learning_rate": 0.0016320000000000002, + "loss": 5.2265, + "step": 544 + }, + { + "epoch": 0.00545, + "grad_norm": 0.8034144620978907, + "learning_rate": 0.0016350000000000002, + "loss": 5.2348, + "step": 545 + }, + { + "epoch": 0.00546, + "grad_norm": 0.7091235133563132, + "learning_rate": 0.0016380000000000001, + "loss": 5.2004, + "step": 546 + }, + { + "epoch": 0.00547, + "grad_norm": 0.6683331586466694, + "learning_rate": 0.001641, + "loss": 5.1915, + "step": 547 + }, + { + "epoch": 0.00548, + "grad_norm": 0.5441552662279447, + "learning_rate": 0.001644, + "loss": 5.1843, + "step": 548 + }, + { + "epoch": 0.00549, + "grad_norm": 0.5462993858197037, + "learning_rate": 0.0016470000000000002, + "loss": 5.1806, + "step": 549 + }, + { + "epoch": 0.0055, + "grad_norm": 0.5691406737163984, + "learning_rate": 0.0016500000000000002, + "loss": 5.1914, + "step": 550 + }, + { + "epoch": 0.00551, + "grad_norm": 0.57998143312047, + "learning_rate": 0.0016530000000000002, + "loss": 5.1789, + "step": 551 + }, + { + "epoch": 0.00552, + "grad_norm": 0.5286482487653069, + "learning_rate": 0.0016560000000000001, + "loss": 5.158, + "step": 552 + }, + { + "epoch": 0.00553, + "grad_norm": 0.48759014943874474, + "learning_rate": 0.001659, + "loss": 5.1465, + "step": 553 + }, + { + "epoch": 0.00554, + "grad_norm": 0.4792524526847805, + "learning_rate": 0.0016620000000000003, + "loss": 5.1537, + "step": 554 + }, + { + "epoch": 0.00555, + "grad_norm": 0.5054837577144806, + "learning_rate": 0.0016650000000000002, + "loss": 5.1496, + "step": 555 + }, + { + "epoch": 0.00556, + "grad_norm": 0.5148667639200912, + "learning_rate": 0.0016680000000000002, + "loss": 5.1288, + "step": 556 + }, + { + "epoch": 0.00557, + "grad_norm": 0.5122706189056161, + "learning_rate": 0.0016710000000000002, + "loss": 5.1153, + "step": 557 + }, + { + "epoch": 0.00558, + "grad_norm": 0.6046224327964763, + "learning_rate": 0.0016740000000000001, + "loss": 5.1358, + "step": 558 + }, + { + "epoch": 0.00559, + "grad_norm": 0.6851636176151574, + "learning_rate": 0.001677, + "loss": 5.1268, + "step": 559 + }, + { + "epoch": 0.0056, + "grad_norm": 0.6922547745331437, + "learning_rate": 0.0016800000000000003, + "loss": 5.096, + "step": 560 + }, + { + "epoch": 0.00561, + "grad_norm": 0.7411075733344746, + "learning_rate": 0.0016830000000000003, + "loss": 5.1176, + "step": 561 + }, + { + "epoch": 0.00562, + "grad_norm": 0.8981627852593407, + "learning_rate": 0.0016860000000000002, + "loss": 5.1206, + "step": 562 + }, + { + "epoch": 0.00563, + "grad_norm": 1.1260200882948381, + "learning_rate": 0.001689, + "loss": 5.1239, + "step": 563 + }, + { + "epoch": 0.00564, + "grad_norm": 1.1027210513289374, + "learning_rate": 0.001692, + "loss": 5.113, + "step": 564 + }, + { + "epoch": 0.00565, + "grad_norm": 0.890213024695838, + "learning_rate": 0.001695, + "loss": 5.1071, + "step": 565 + }, + { + "epoch": 0.00566, + "grad_norm": 0.8627979775394023, + "learning_rate": 0.0016979999999999999, + "loss": 5.0976, + "step": 566 + }, + { + "epoch": 0.00567, + "grad_norm": 0.9823945407746334, + "learning_rate": 0.0017009999999999998, + "loss": 5.101, + "step": 567 + }, + { + "epoch": 0.00568, + "grad_norm": 0.9715849000742567, + "learning_rate": 0.0017039999999999998, + "loss": 5.1139, + "step": 568 + }, + { + "epoch": 0.00569, + "grad_norm": 0.9989913016974431, + "learning_rate": 0.001707, + "loss": 5.0861, + "step": 569 + }, + { + "epoch": 0.0057, + "grad_norm": 1.0927877338999235, + "learning_rate": 0.00171, + "loss": 5.1262, + "step": 570 + }, + { + "epoch": 0.00571, + "grad_norm": 1.0328737042176641, + "learning_rate": 0.001713, + "loss": 5.1192, + "step": 571 + }, + { + "epoch": 0.00572, + "grad_norm": 1.3722388350456287, + "learning_rate": 0.0017159999999999999, + "loss": 5.1049, + "step": 572 + }, + { + "epoch": 0.00573, + "grad_norm": 0.9525979203379623, + "learning_rate": 0.0017189999999999998, + "loss": 5.1081, + "step": 573 + }, + { + "epoch": 0.00574, + "grad_norm": 1.0626128534442882, + "learning_rate": 0.001722, + "loss": 5.1048, + "step": 574 + }, + { + "epoch": 0.00575, + "grad_norm": 0.9331527599734185, + "learning_rate": 0.001725, + "loss": 5.074, + "step": 575 + }, + { + "epoch": 0.00576, + "grad_norm": 0.9277735357052385, + "learning_rate": 0.001728, + "loss": 5.0843, + "step": 576 + }, + { + "epoch": 0.00577, + "grad_norm": 0.9070967517565243, + "learning_rate": 0.001731, + "loss": 5.0908, + "step": 577 + }, + { + "epoch": 0.00578, + "grad_norm": 0.8451551366430134, + "learning_rate": 0.0017339999999999999, + "loss": 5.0704, + "step": 578 + }, + { + "epoch": 0.00579, + "grad_norm": 0.7590562073625285, + "learning_rate": 0.0017369999999999998, + "loss": 5.058, + "step": 579 + }, + { + "epoch": 0.0058, + "grad_norm": 0.6385326977156373, + "learning_rate": 0.00174, + "loss": 5.0662, + "step": 580 + }, + { + "epoch": 0.00581, + "grad_norm": 0.5982129206576257, + "learning_rate": 0.001743, + "loss": 5.0624, + "step": 581 + }, + { + "epoch": 0.00582, + "grad_norm": 0.7358321954275717, + "learning_rate": 0.001746, + "loss": 5.0371, + "step": 582 + }, + { + "epoch": 0.00583, + "grad_norm": 0.790704997209863, + "learning_rate": 0.001749, + "loss": 5.0641, + "step": 583 + }, + { + "epoch": 0.00584, + "grad_norm": 0.7351931257233056, + "learning_rate": 0.0017519999999999999, + "loss": 5.0549, + "step": 584 + }, + { + "epoch": 0.00585, + "grad_norm": 0.6167540927458872, + "learning_rate": 0.0017549999999999998, + "loss": 5.0362, + "step": 585 + }, + { + "epoch": 0.00586, + "grad_norm": 0.6143004439139178, + "learning_rate": 0.001758, + "loss": 5.0316, + "step": 586 + }, + { + "epoch": 0.00587, + "grad_norm": 0.619830997492515, + "learning_rate": 0.001761, + "loss": 5.018, + "step": 587 + }, + { + "epoch": 0.00588, + "grad_norm": 0.676754014852622, + "learning_rate": 0.001764, + "loss": 5.0113, + "step": 588 + }, + { + "epoch": 0.00589, + "grad_norm": 0.8152367183756798, + "learning_rate": 0.001767, + "loss": 5.0063, + "step": 589 + }, + { + "epoch": 0.0059, + "grad_norm": 0.8067730505459064, + "learning_rate": 0.0017699999999999999, + "loss": 5.0009, + "step": 590 + }, + { + "epoch": 0.00591, + "grad_norm": 0.7857115694134444, + "learning_rate": 0.001773, + "loss": 5.0014, + "step": 591 + }, + { + "epoch": 0.00592, + "grad_norm": 1.0305248086573016, + "learning_rate": 0.001776, + "loss": 5.0384, + "step": 592 + }, + { + "epoch": 0.00593, + "grad_norm": 1.1034322728795254, + "learning_rate": 0.001779, + "loss": 5.0387, + "step": 593 + }, + { + "epoch": 0.00594, + "grad_norm": 0.796782800390064, + "learning_rate": 0.001782, + "loss": 5.0138, + "step": 594 + }, + { + "epoch": 0.00595, + "grad_norm": 0.947951394101208, + "learning_rate": 0.001785, + "loss": 5.0288, + "step": 595 + }, + { + "epoch": 0.00596, + "grad_norm": 0.948001541672118, + "learning_rate": 0.0017879999999999999, + "loss": 5.0236, + "step": 596 + }, + { + "epoch": 0.00597, + "grad_norm": 0.8614311642183788, + "learning_rate": 0.001791, + "loss": 4.983, + "step": 597 + }, + { + "epoch": 0.00598, + "grad_norm": 0.7822502963641336, + "learning_rate": 0.001794, + "loss": 5.0032, + "step": 598 + }, + { + "epoch": 0.00599, + "grad_norm": 0.662502314672726, + "learning_rate": 0.001797, + "loss": 4.9824, + "step": 599 + }, + { + "epoch": 0.006, + "grad_norm": 0.612691797581944, + "learning_rate": 0.0018, + "loss": 4.9865, + "step": 600 + }, + { + "epoch": 0.00601, + "grad_norm": 0.6872791409552744, + "learning_rate": 0.001803, + "loss": 4.9768, + "step": 601 + }, + { + "epoch": 0.00602, + "grad_norm": 0.6805787880992017, + "learning_rate": 0.0018059999999999999, + "loss": 4.9611, + "step": 602 + }, + { + "epoch": 0.00603, + "grad_norm": 0.6726518476757681, + "learning_rate": 0.001809, + "loss": 4.9863, + "step": 603 + }, + { + "epoch": 0.00604, + "grad_norm": 0.6569683723636671, + "learning_rate": 0.001812, + "loss": 4.9594, + "step": 604 + }, + { + "epoch": 0.00605, + "grad_norm": 0.7635898698708745, + "learning_rate": 0.001815, + "loss": 4.9733, + "step": 605 + }, + { + "epoch": 0.00606, + "grad_norm": 0.7455554600622951, + "learning_rate": 0.001818, + "loss": 4.9743, + "step": 606 + }, + { + "epoch": 0.00607, + "grad_norm": 0.7493825131119356, + "learning_rate": 0.001821, + "loss": 4.968, + "step": 607 + }, + { + "epoch": 0.00608, + "grad_norm": 0.7957828614510537, + "learning_rate": 0.001824, + "loss": 4.9536, + "step": 608 + }, + { + "epoch": 0.00609, + "grad_norm": 0.8304899179729943, + "learning_rate": 0.001827, + "loss": 4.9564, + "step": 609 + }, + { + "epoch": 0.0061, + "grad_norm": 0.8109281284885322, + "learning_rate": 0.00183, + "loss": 4.9141, + "step": 610 + }, + { + "epoch": 0.00611, + "grad_norm": 0.7476158062529945, + "learning_rate": 0.001833, + "loss": 4.936, + "step": 611 + }, + { + "epoch": 0.00612, + "grad_norm": 0.7704555466663776, + "learning_rate": 0.001836, + "loss": 4.9262, + "step": 612 + }, + { + "epoch": 0.00613, + "grad_norm": 0.7179188189545482, + "learning_rate": 0.001839, + "loss": 4.9285, + "step": 613 + }, + { + "epoch": 0.00614, + "grad_norm": 0.6487628040476678, + "learning_rate": 0.001842, + "loss": 4.9483, + "step": 614 + }, + { + "epoch": 0.00615, + "grad_norm": 0.6346341426047015, + "learning_rate": 0.001845, + "loss": 4.9336, + "step": 615 + }, + { + "epoch": 0.00616, + "grad_norm": 0.6919482296512011, + "learning_rate": 0.001848, + "loss": 4.92, + "step": 616 + }, + { + "epoch": 0.00617, + "grad_norm": 0.8470324326487232, + "learning_rate": 0.001851, + "loss": 4.937, + "step": 617 + }, + { + "epoch": 0.00618, + "grad_norm": 0.9780451420741652, + "learning_rate": 0.001854, + "loss": 4.918, + "step": 618 + }, + { + "epoch": 0.00619, + "grad_norm": 1.0393940937639112, + "learning_rate": 0.001857, + "loss": 4.9393, + "step": 619 + }, + { + "epoch": 0.0062, + "grad_norm": 1.0864888594503201, + "learning_rate": 0.00186, + "loss": 4.9315, + "step": 620 + }, + { + "epoch": 0.00621, + "grad_norm": 1.1243573748169302, + "learning_rate": 0.001863, + "loss": 4.931, + "step": 621 + }, + { + "epoch": 0.00622, + "grad_norm": 1.2701399731469665, + "learning_rate": 0.001866, + "loss": 4.9502, + "step": 622 + }, + { + "epoch": 0.00623, + "grad_norm": 0.8273605340089976, + "learning_rate": 0.001869, + "loss": 4.9139, + "step": 623 + }, + { + "epoch": 0.00624, + "grad_norm": 0.9643311427939416, + "learning_rate": 0.001872, + "loss": 4.9253, + "step": 624 + }, + { + "epoch": 0.00625, + "grad_norm": 1.2310098040523034, + "learning_rate": 0.001875, + "loss": 4.9539, + "step": 625 + }, + { + "epoch": 0.00626, + "grad_norm": 0.8268848993203812, + "learning_rate": 0.0018780000000000001, + "loss": 4.9199, + "step": 626 + }, + { + "epoch": 0.00627, + "grad_norm": 0.9575639973355906, + "learning_rate": 0.001881, + "loss": 4.909, + "step": 627 + }, + { + "epoch": 0.00628, + "grad_norm": 0.9391450767019313, + "learning_rate": 0.001884, + "loss": 4.9487, + "step": 628 + }, + { + "epoch": 0.00629, + "grad_norm": 0.8387975784827085, + "learning_rate": 0.001887, + "loss": 4.8976, + "step": 629 + }, + { + "epoch": 0.0063, + "grad_norm": 0.7745062985579545, + "learning_rate": 0.00189, + "loss": 4.9062, + "step": 630 + }, + { + "epoch": 0.00631, + "grad_norm": 0.8213366097020988, + "learning_rate": 0.0018930000000000002, + "loss": 4.9006, + "step": 631 + }, + { + "epoch": 0.00632, + "grad_norm": 0.8670608316656828, + "learning_rate": 0.0018960000000000001, + "loss": 4.9226, + "step": 632 + }, + { + "epoch": 0.00633, + "grad_norm": 0.9183940724690223, + "learning_rate": 0.001899, + "loss": 4.9055, + "step": 633 + }, + { + "epoch": 0.00634, + "grad_norm": 0.8953336214105924, + "learning_rate": 0.001902, + "loss": 4.8931, + "step": 634 + }, + { + "epoch": 0.00635, + "grad_norm": 0.8869843611928877, + "learning_rate": 0.001905, + "loss": 4.8851, + "step": 635 + }, + { + "epoch": 0.00636, + "grad_norm": 1.014617544478078, + "learning_rate": 0.001908, + "loss": 4.911, + "step": 636 + }, + { + "epoch": 0.00637, + "grad_norm": 0.8681767489690893, + "learning_rate": 0.0019110000000000002, + "loss": 4.909, + "step": 637 + }, + { + "epoch": 0.00638, + "grad_norm": 0.9792692499883113, + "learning_rate": 0.0019140000000000001, + "loss": 4.9073, + "step": 638 + }, + { + "epoch": 0.00639, + "grad_norm": 0.8702384519183142, + "learning_rate": 0.001917, + "loss": 4.8724, + "step": 639 + }, + { + "epoch": 0.0064, + "grad_norm": 0.7656979903996737, + "learning_rate": 0.00192, + "loss": 4.8809, + "step": 640 + }, + { + "epoch": 0.00641, + "grad_norm": 0.7895438458212233, + "learning_rate": 0.001923, + "loss": 4.8762, + "step": 641 + }, + { + "epoch": 0.00642, + "grad_norm": 0.8281699495124627, + "learning_rate": 0.001926, + "loss": 4.8749, + "step": 642 + }, + { + "epoch": 0.00643, + "grad_norm": 0.7845360432833325, + "learning_rate": 0.0019290000000000002, + "loss": 4.8856, + "step": 643 + }, + { + "epoch": 0.00644, + "grad_norm": 0.6857666802048429, + "learning_rate": 0.0019320000000000001, + "loss": 4.8583, + "step": 644 + }, + { + "epoch": 0.00645, + "grad_norm": 0.5056941566313082, + "learning_rate": 0.001935, + "loss": 4.8522, + "step": 645 + }, + { + "epoch": 0.00646, + "grad_norm": 0.47964630717910517, + "learning_rate": 0.001938, + "loss": 4.8628, + "step": 646 + }, + { + "epoch": 0.00647, + "grad_norm": 0.519046406898008, + "learning_rate": 0.001941, + "loss": 4.8552, + "step": 647 + }, + { + "epoch": 0.00648, + "grad_norm": 0.4935185080219269, + "learning_rate": 0.0019440000000000002, + "loss": 4.8488, + "step": 648 + }, + { + "epoch": 0.00649, + "grad_norm": 0.5205599594018984, + "learning_rate": 0.0019470000000000002, + "loss": 4.8333, + "step": 649 + }, + { + "epoch": 0.0065, + "grad_norm": 0.45295352715192905, + "learning_rate": 0.0019500000000000001, + "loss": 4.8617, + "step": 650 + }, + { + "epoch": 0.00651, + "grad_norm": 0.4378817965282719, + "learning_rate": 0.001953, + "loss": 4.84, + "step": 651 + }, + { + "epoch": 0.00652, + "grad_norm": 0.48395025642257355, + "learning_rate": 0.0019560000000000003, + "loss": 4.8266, + "step": 652 + }, + { + "epoch": 0.00653, + "grad_norm": 0.5194845024092328, + "learning_rate": 0.0019590000000000002, + "loss": 4.8294, + "step": 653 + }, + { + "epoch": 0.00654, + "grad_norm": 0.5707068593609262, + "learning_rate": 0.001962, + "loss": 4.8026, + "step": 654 + }, + { + "epoch": 0.00655, + "grad_norm": 0.6725172209521785, + "learning_rate": 0.001965, + "loss": 4.834, + "step": 655 + }, + { + "epoch": 0.00656, + "grad_norm": 0.7729568432311581, + "learning_rate": 0.001968, + "loss": 4.8398, + "step": 656 + }, + { + "epoch": 0.00657, + "grad_norm": 0.9182317536069756, + "learning_rate": 0.001971, + "loss": 4.8175, + "step": 657 + }, + { + "epoch": 0.00658, + "grad_norm": 0.9241834469814327, + "learning_rate": 0.001974, + "loss": 4.8482, + "step": 658 + }, + { + "epoch": 0.00659, + "grad_norm": 0.7554289117841486, + "learning_rate": 0.001977, + "loss": 4.8199, + "step": 659 + }, + { + "epoch": 0.0066, + "grad_norm": 0.786222644002383, + "learning_rate": 0.00198, + "loss": 4.8336, + "step": 660 + }, + { + "epoch": 0.00661, + "grad_norm": 0.7486956414167568, + "learning_rate": 0.001983, + "loss": 4.8035, + "step": 661 + }, + { + "epoch": 0.00662, + "grad_norm": 0.8224936329580866, + "learning_rate": 0.0019860000000000004, + "loss": 4.8038, + "step": 662 + }, + { + "epoch": 0.00663, + "grad_norm": 1.0333891900276588, + "learning_rate": 0.0019890000000000003, + "loss": 4.8216, + "step": 663 + }, + { + "epoch": 0.00664, + "grad_norm": 1.0454184895583787, + "learning_rate": 0.0019920000000000003, + "loss": 4.826, + "step": 664 + }, + { + "epoch": 0.00665, + "grad_norm": 0.9684168839706867, + "learning_rate": 0.0019950000000000002, + "loss": 4.7952, + "step": 665 + }, + { + "epoch": 0.00666, + "grad_norm": 1.0353944208678731, + "learning_rate": 0.001998, + "loss": 4.8213, + "step": 666 + }, + { + "epoch": 0.00667, + "grad_norm": 0.9360230286599108, + "learning_rate": 0.002001, + "loss": 4.8166, + "step": 667 + }, + { + "epoch": 0.00668, + "grad_norm": 0.7256915823313501, + "learning_rate": 0.002004, + "loss": 4.83, + "step": 668 + }, + { + "epoch": 0.00669, + "grad_norm": 0.8394069732991961, + "learning_rate": 0.002007, + "loss": 4.8494, + "step": 669 + }, + { + "epoch": 0.0067, + "grad_norm": 1.077577977883947, + "learning_rate": 0.00201, + "loss": 4.8532, + "step": 670 + }, + { + "epoch": 0.00671, + "grad_norm": 0.944781077919628, + "learning_rate": 0.002013, + "loss": 4.8236, + "step": 671 + }, + { + "epoch": 0.00672, + "grad_norm": 0.7024086289087288, + "learning_rate": 0.002016, + "loss": 4.7966, + "step": 672 + }, + { + "epoch": 0.00673, + "grad_norm": 0.6931174948166909, + "learning_rate": 0.002019, + "loss": 4.809, + "step": 673 + }, + { + "epoch": 0.00674, + "grad_norm": 0.6108100930625565, + "learning_rate": 0.0020220000000000004, + "loss": 4.8005, + "step": 674 + }, + { + "epoch": 0.00675, + "grad_norm": 0.5498762136107717, + "learning_rate": 0.0020250000000000003, + "loss": 4.8171, + "step": 675 + }, + { + "epoch": 0.00676, + "grad_norm": 0.6244881777878202, + "learning_rate": 0.0020280000000000003, + "loss": 4.7951, + "step": 676 + }, + { + "epoch": 0.00677, + "grad_norm": 0.5815667573949113, + "learning_rate": 0.0020310000000000003, + "loss": 4.7913, + "step": 677 + }, + { + "epoch": 0.00678, + "grad_norm": 0.521491074875552, + "learning_rate": 0.0020340000000000002, + "loss": 4.7959, + "step": 678 + }, + { + "epoch": 0.00679, + "grad_norm": 0.5801192243287052, + "learning_rate": 0.002037, + "loss": 4.7864, + "step": 679 + }, + { + "epoch": 0.0068, + "grad_norm": 0.6398788453868361, + "learning_rate": 0.00204, + "loss": 4.7875, + "step": 680 + }, + { + "epoch": 0.00681, + "grad_norm": 0.6456567912241706, + "learning_rate": 0.002043, + "loss": 4.7807, + "step": 681 + }, + { + "epoch": 0.00682, + "grad_norm": 0.6889535888719373, + "learning_rate": 0.002046, + "loss": 4.7776, + "step": 682 + }, + { + "epoch": 0.00683, + "grad_norm": 0.68132819276425, + "learning_rate": 0.002049, + "loss": 4.7612, + "step": 683 + }, + { + "epoch": 0.00684, + "grad_norm": 0.6724100911918068, + "learning_rate": 0.002052, + "loss": 4.7814, + "step": 684 + }, + { + "epoch": 0.00685, + "grad_norm": 0.815928121483196, + "learning_rate": 0.0020550000000000004, + "loss": 4.7913, + "step": 685 + }, + { + "epoch": 0.00686, + "grad_norm": 0.7537150884203632, + "learning_rate": 0.0020580000000000004, + "loss": 4.7584, + "step": 686 + }, + { + "epoch": 0.00687, + "grad_norm": 0.7326770278926477, + "learning_rate": 0.0020610000000000003, + "loss": 4.7655, + "step": 687 + }, + { + "epoch": 0.00688, + "grad_norm": 0.7145131916015852, + "learning_rate": 0.002064, + "loss": 4.7631, + "step": 688 + }, + { + "epoch": 0.00689, + "grad_norm": 0.694828893634318, + "learning_rate": 0.002067, + "loss": 4.7685, + "step": 689 + }, + { + "epoch": 0.0069, + "grad_norm": 0.7402628986076095, + "learning_rate": 0.00207, + "loss": 4.7629, + "step": 690 + }, + { + "epoch": 0.00691, + "grad_norm": 0.7809038473743037, + "learning_rate": 0.0020729999999999998, + "loss": 4.734, + "step": 691 + }, + { + "epoch": 0.00692, + "grad_norm": 0.9321737974219415, + "learning_rate": 0.0020759999999999997, + "loss": 4.7623, + "step": 692 + }, + { + "epoch": 0.00693, + "grad_norm": 1.0940875671633443, + "learning_rate": 0.0020789999999999997, + "loss": 4.788, + "step": 693 + }, + { + "epoch": 0.00694, + "grad_norm": 0.9030072391970948, + "learning_rate": 0.002082, + "loss": 4.7582, + "step": 694 + }, + { + "epoch": 0.00695, + "grad_norm": 0.8857004379777869, + "learning_rate": 0.002085, + "loss": 4.7649, + "step": 695 + }, + { + "epoch": 0.00696, + "grad_norm": 0.8991843577664433, + "learning_rate": 0.002088, + "loss": 4.7712, + "step": 696 + }, + { + "epoch": 0.00697, + "grad_norm": 1.0094726318481275, + "learning_rate": 0.002091, + "loss": 4.7869, + "step": 697 + }, + { + "epoch": 0.00698, + "grad_norm": 1.09376996979269, + "learning_rate": 0.002094, + "loss": 4.7773, + "step": 698 + }, + { + "epoch": 0.00699, + "grad_norm": 0.8766228601427865, + "learning_rate": 0.002097, + "loss": 4.7548, + "step": 699 + }, + { + "epoch": 0.007, + "grad_norm": 0.8580408760742435, + "learning_rate": 0.0021, + "loss": 4.7811, + "step": 700 + }, + { + "epoch": 0.00701, + "grad_norm": 0.7398805390269545, + "learning_rate": 0.002103, + "loss": 4.7496, + "step": 701 + }, + { + "epoch": 0.00702, + "grad_norm": 0.8329034135969645, + "learning_rate": 0.002106, + "loss": 4.762, + "step": 702 + }, + { + "epoch": 0.00703, + "grad_norm": 0.8505784581176613, + "learning_rate": 0.0021089999999999998, + "loss": 4.7427, + "step": 703 + }, + { + "epoch": 0.00704, + "grad_norm": 0.8003313240374758, + "learning_rate": 0.0021119999999999997, + "loss": 4.7516, + "step": 704 + }, + { + "epoch": 0.00705, + "grad_norm": 0.7109629500256387, + "learning_rate": 0.002115, + "loss": 4.7635, + "step": 705 + }, + { + "epoch": 0.00706, + "grad_norm": 0.6289559048526784, + "learning_rate": 0.002118, + "loss": 4.7473, + "step": 706 + }, + { + "epoch": 0.00707, + "grad_norm": 0.5482382237711074, + "learning_rate": 0.002121, + "loss": 4.7193, + "step": 707 + }, + { + "epoch": 0.00708, + "grad_norm": 0.5598390325768088, + "learning_rate": 0.002124, + "loss": 4.7015, + "step": 708 + }, + { + "epoch": 0.00709, + "grad_norm": 0.5529551760214286, + "learning_rate": 0.002127, + "loss": 4.7331, + "step": 709 + }, + { + "epoch": 0.0071, + "grad_norm": 0.6671084870548735, + "learning_rate": 0.00213, + "loss": 4.7271, + "step": 710 + }, + { + "epoch": 0.00711, + "grad_norm": 1.1336515667563218, + "learning_rate": 0.002133, + "loss": 4.7225, + "step": 711 + }, + { + "epoch": 0.00712, + "grad_norm": 0.8878745395415352, + "learning_rate": 0.002136, + "loss": 4.761, + "step": 712 + }, + { + "epoch": 0.00713, + "grad_norm": 0.6767692274074788, + "learning_rate": 0.002139, + "loss": 4.7547, + "step": 713 + }, + { + "epoch": 0.00714, + "grad_norm": 0.7698487468076475, + "learning_rate": 0.002142, + "loss": 4.7544, + "step": 714 + }, + { + "epoch": 0.00715, + "grad_norm": 0.6189384208707256, + "learning_rate": 0.0021449999999999998, + "loss": 4.7077, + "step": 715 + }, + { + "epoch": 0.00716, + "grad_norm": 0.5968344999809513, + "learning_rate": 0.002148, + "loss": 4.7064, + "step": 716 + }, + { + "epoch": 0.00717, + "grad_norm": 0.5117629580972962, + "learning_rate": 0.002151, + "loss": 4.7111, + "step": 717 + }, + { + "epoch": 0.00718, + "grad_norm": 0.5586203984201494, + "learning_rate": 0.002154, + "loss": 4.7079, + "step": 718 + }, + { + "epoch": 0.00719, + "grad_norm": 0.5075479633130432, + "learning_rate": 0.002157, + "loss": 4.7121, + "step": 719 + }, + { + "epoch": 0.0072, + "grad_norm": 0.4840284295584723, + "learning_rate": 0.00216, + "loss": 4.7045, + "step": 720 + }, + { + "epoch": 0.00721, + "grad_norm": 0.4971167150526394, + "learning_rate": 0.002163, + "loss": 4.7056, + "step": 721 + }, + { + "epoch": 0.00722, + "grad_norm": 0.5347204422021664, + "learning_rate": 0.002166, + "loss": 4.6807, + "step": 722 + }, + { + "epoch": 0.00723, + "grad_norm": 0.5424206330484346, + "learning_rate": 0.002169, + "loss": 4.6942, + "step": 723 + }, + { + "epoch": 0.00724, + "grad_norm": 0.5219918511718452, + "learning_rate": 0.002172, + "loss": 4.6912, + "step": 724 + }, + { + "epoch": 0.00725, + "grad_norm": 0.4975444213381179, + "learning_rate": 0.002175, + "loss": 4.6834, + "step": 725 + }, + { + "epoch": 0.00726, + "grad_norm": 0.5840922524665152, + "learning_rate": 0.002178, + "loss": 4.7008, + "step": 726 + }, + { + "epoch": 0.00727, + "grad_norm": 0.7633698368434719, + "learning_rate": 0.0021809999999999998, + "loss": 4.6777, + "step": 727 + }, + { + "epoch": 0.00728, + "grad_norm": 0.9052599955462425, + "learning_rate": 0.002184, + "loss": 4.7084, + "step": 728 + }, + { + "epoch": 0.00729, + "grad_norm": 0.9052862109699649, + "learning_rate": 0.002187, + "loss": 4.7065, + "step": 729 + }, + { + "epoch": 0.0073, + "grad_norm": 0.9788098741089467, + "learning_rate": 0.00219, + "loss": 4.7163, + "step": 730 + }, + { + "epoch": 0.00731, + "grad_norm": 0.9218858714566311, + "learning_rate": 0.002193, + "loss": 4.7193, + "step": 731 + }, + { + "epoch": 0.00732, + "grad_norm": 1.1412145043722297, + "learning_rate": 0.002196, + "loss": 4.7513, + "step": 732 + }, + { + "epoch": 0.00733, + "grad_norm": 1.0851884555194036, + "learning_rate": 0.002199, + "loss": 4.7135, + "step": 733 + }, + { + "epoch": 0.00734, + "grad_norm": 0.9295060612046938, + "learning_rate": 0.002202, + "loss": 4.745, + "step": 734 + }, + { + "epoch": 0.00735, + "grad_norm": 1.1744848346988581, + "learning_rate": 0.002205, + "loss": 4.766, + "step": 735 + }, + { + "epoch": 0.00736, + "grad_norm": 0.8278931321658332, + "learning_rate": 0.002208, + "loss": 4.7413, + "step": 736 + }, + { + "epoch": 0.00737, + "grad_norm": 0.8170748474782394, + "learning_rate": 0.002211, + "loss": 4.701, + "step": 737 + }, + { + "epoch": 0.00738, + "grad_norm": 0.772640277783885, + "learning_rate": 0.002214, + "loss": 4.7083, + "step": 738 + }, + { + "epoch": 0.00739, + "grad_norm": 0.6328057525928463, + "learning_rate": 0.0022170000000000002, + "loss": 4.6886, + "step": 739 + }, + { + "epoch": 0.0074, + "grad_norm": 0.6456777453177256, + "learning_rate": 0.00222, + "loss": 4.7001, + "step": 740 + }, + { + "epoch": 0.00741, + "grad_norm": 0.5912580097221507, + "learning_rate": 0.002223, + "loss": 4.6652, + "step": 741 + }, + { + "epoch": 0.00742, + "grad_norm": 0.4622238730689768, + "learning_rate": 0.002226, + "loss": 4.6972, + "step": 742 + }, + { + "epoch": 0.00743, + "grad_norm": 0.4904966133460914, + "learning_rate": 0.002229, + "loss": 4.6585, + "step": 743 + }, + { + "epoch": 0.00744, + "grad_norm": 0.5611948954376664, + "learning_rate": 0.002232, + "loss": 4.6643, + "step": 744 + }, + { + "epoch": 0.00745, + "grad_norm": 0.6222946629538655, + "learning_rate": 0.002235, + "loss": 4.6777, + "step": 745 + }, + { + "epoch": 0.00746, + "grad_norm": 0.6678221674623648, + "learning_rate": 0.002238, + "loss": 4.6735, + "step": 746 + }, + { + "epoch": 0.00747, + "grad_norm": 0.7945455347777014, + "learning_rate": 0.002241, + "loss": 4.6637, + "step": 747 + }, + { + "epoch": 0.00748, + "grad_norm": 0.9027593299924651, + "learning_rate": 0.002244, + "loss": 4.683, + "step": 748 + }, + { + "epoch": 0.00749, + "grad_norm": 0.8224517051186598, + "learning_rate": 0.002247, + "loss": 4.6612, + "step": 749 + }, + { + "epoch": 0.0075, + "grad_norm": 0.6737194715747743, + "learning_rate": 0.0022500000000000003, + "loss": 4.6955, + "step": 750 + }, + { + "epoch": 0.00751, + "grad_norm": 0.7232822830562855, + "learning_rate": 0.0022530000000000002, + "loss": 4.6893, + "step": 751 + }, + { + "epoch": 0.00752, + "grad_norm": 0.6449225627821611, + "learning_rate": 0.002256, + "loss": 4.637, + "step": 752 + }, + { + "epoch": 0.00753, + "grad_norm": 0.6558342053271724, + "learning_rate": 0.002259, + "loss": 4.6782, + "step": 753 + }, + { + "epoch": 0.00754, + "grad_norm": 0.7056102529212358, + "learning_rate": 0.002262, + "loss": 4.6906, + "step": 754 + }, + { + "epoch": 0.00755, + "grad_norm": 0.9043185481447295, + "learning_rate": 0.002265, + "loss": 4.6606, + "step": 755 + }, + { + "epoch": 0.00756, + "grad_norm": 0.8770375990036015, + "learning_rate": 0.002268, + "loss": 4.6551, + "step": 756 + }, + { + "epoch": 0.00757, + "grad_norm": 0.784655485815756, + "learning_rate": 0.002271, + "loss": 4.6655, + "step": 757 + }, + { + "epoch": 0.00758, + "grad_norm": 0.6912079950422929, + "learning_rate": 0.002274, + "loss": 4.655, + "step": 758 + }, + { + "epoch": 0.00759, + "grad_norm": 0.6281371231708264, + "learning_rate": 0.002277, + "loss": 4.6629, + "step": 759 + }, + { + "epoch": 0.0076, + "grad_norm": 0.5992557531443571, + "learning_rate": 0.00228, + "loss": 4.6463, + "step": 760 + }, + { + "epoch": 0.00761, + "grad_norm": 0.6034804931159479, + "learning_rate": 0.002283, + "loss": 4.6146, + "step": 761 + }, + { + "epoch": 0.00762, + "grad_norm": 0.6060668832918386, + "learning_rate": 0.0022860000000000003, + "loss": 4.5953, + "step": 762 + }, + { + "epoch": 0.00763, + "grad_norm": 0.5816892837581542, + "learning_rate": 0.0022890000000000002, + "loss": 4.6459, + "step": 763 + }, + { + "epoch": 0.00764, + "grad_norm": 0.599786845424844, + "learning_rate": 0.002292, + "loss": 4.6554, + "step": 764 + }, + { + "epoch": 0.00765, + "grad_norm": 0.6640243443171511, + "learning_rate": 0.002295, + "loss": 4.6451, + "step": 765 + }, + { + "epoch": 0.00766, + "grad_norm": 0.6663176934335011, + "learning_rate": 0.002298, + "loss": 4.6438, + "step": 766 + }, + { + "epoch": 0.00767, + "grad_norm": 0.6384162747773746, + "learning_rate": 0.002301, + "loss": 4.6392, + "step": 767 + }, + { + "epoch": 0.00768, + "grad_norm": 0.8537658256313299, + "learning_rate": 0.002304, + "loss": 4.6457, + "step": 768 + }, + { + "epoch": 0.00769, + "grad_norm": 1.0098120726516953, + "learning_rate": 0.002307, + "loss": 4.6694, + "step": 769 + }, + { + "epoch": 0.0077, + "grad_norm": 0.7423003820485509, + "learning_rate": 0.00231, + "loss": 4.6223, + "step": 770 + }, + { + "epoch": 0.00771, + "grad_norm": 0.6691351211965346, + "learning_rate": 0.002313, + "loss": 4.6569, + "step": 771 + }, + { + "epoch": 0.00772, + "grad_norm": 0.8185609860643415, + "learning_rate": 0.002316, + "loss": 4.6531, + "step": 772 + }, + { + "epoch": 0.00773, + "grad_norm": 1.0039613061235502, + "learning_rate": 0.0023190000000000003, + "loss": 4.6664, + "step": 773 + }, + { + "epoch": 0.00774, + "grad_norm": 0.8500607234716588, + "learning_rate": 0.0023220000000000003, + "loss": 4.6845, + "step": 774 + }, + { + "epoch": 0.00775, + "grad_norm": 0.7057192127032131, + "learning_rate": 0.0023250000000000002, + "loss": 4.6688, + "step": 775 + }, + { + "epoch": 0.00776, + "grad_norm": 0.8527617729361273, + "learning_rate": 0.002328, + "loss": 4.6732, + "step": 776 + }, + { + "epoch": 0.00777, + "grad_norm": 0.7987783133918631, + "learning_rate": 0.002331, + "loss": 4.6469, + "step": 777 + }, + { + "epoch": 0.00778, + "grad_norm": 0.8221364860163118, + "learning_rate": 0.002334, + "loss": 4.6645, + "step": 778 + }, + { + "epoch": 0.00779, + "grad_norm": 0.8901185821524535, + "learning_rate": 0.002337, + "loss": 4.6243, + "step": 779 + }, + { + "epoch": 0.0078, + "grad_norm": 0.889956887452623, + "learning_rate": 0.00234, + "loss": 4.646, + "step": 780 + }, + { + "epoch": 0.00781, + "grad_norm": 0.763195026878423, + "learning_rate": 0.002343, + "loss": 4.6465, + "step": 781 + }, + { + "epoch": 0.00782, + "grad_norm": 0.6941360203492353, + "learning_rate": 0.002346, + "loss": 4.6378, + "step": 782 + }, + { + "epoch": 0.00783, + "grad_norm": 0.6504570243922513, + "learning_rate": 0.002349, + "loss": 4.6195, + "step": 783 + }, + { + "epoch": 0.00784, + "grad_norm": 0.5612286515783087, + "learning_rate": 0.002352, + "loss": 4.6293, + "step": 784 + }, + { + "epoch": 0.00785, + "grad_norm": 0.6491841258536649, + "learning_rate": 0.0023550000000000003, + "loss": 4.6258, + "step": 785 + }, + { + "epoch": 0.00786, + "grad_norm": 0.7618604140557245, + "learning_rate": 0.0023580000000000003, + "loss": 4.6353, + "step": 786 + }, + { + "epoch": 0.00787, + "grad_norm": 0.9289251225559676, + "learning_rate": 0.0023610000000000003, + "loss": 4.6478, + "step": 787 + }, + { + "epoch": 0.00788, + "grad_norm": 1.0160410825626025, + "learning_rate": 0.002364, + "loss": 4.6569, + "step": 788 + }, + { + "epoch": 0.00789, + "grad_norm": 0.9904632547103471, + "learning_rate": 0.002367, + "loss": 4.6841, + "step": 789 + }, + { + "epoch": 0.0079, + "grad_norm": 0.978549262142993, + "learning_rate": 0.00237, + "loss": 4.6525, + "step": 790 + }, + { + "epoch": 0.00791, + "grad_norm": 0.7626000247604355, + "learning_rate": 0.002373, + "loss": 4.6008, + "step": 791 + }, + { + "epoch": 0.00792, + "grad_norm": 0.7032433100729657, + "learning_rate": 0.002376, + "loss": 4.6361, + "step": 792 + }, + { + "epoch": 0.00793, + "grad_norm": 0.6164393933780811, + "learning_rate": 0.002379, + "loss": 4.6263, + "step": 793 + }, + { + "epoch": 0.00794, + "grad_norm": 0.6849838631141278, + "learning_rate": 0.002382, + "loss": 4.6402, + "step": 794 + }, + { + "epoch": 0.00795, + "grad_norm": 0.7541453624973187, + "learning_rate": 0.002385, + "loss": 4.6181, + "step": 795 + }, + { + "epoch": 0.00796, + "grad_norm": 0.7900314117496404, + "learning_rate": 0.0023880000000000004, + "loss": 4.6154, + "step": 796 + }, + { + "epoch": 0.00797, + "grad_norm": 0.8032302751261208, + "learning_rate": 0.0023910000000000003, + "loss": 4.6598, + "step": 797 + }, + { + "epoch": 0.00798, + "grad_norm": 0.7037506066431396, + "learning_rate": 0.0023940000000000003, + "loss": 4.6206, + "step": 798 + }, + { + "epoch": 0.00799, + "grad_norm": 0.6873143294249761, + "learning_rate": 0.0023970000000000003, + "loss": 4.6565, + "step": 799 + }, + { + "epoch": 0.008, + "grad_norm": 0.5611968184575689, + "learning_rate": 0.0024000000000000002, + "loss": 4.6002, + "step": 800 + }, + { + "epoch": 0.00801, + "grad_norm": 0.6431524925156966, + "learning_rate": 0.002403, + "loss": 4.6326, + "step": 801 + }, + { + "epoch": 0.00802, + "grad_norm": 0.664361851854149, + "learning_rate": 0.002406, + "loss": 4.624, + "step": 802 + }, + { + "epoch": 0.00803, + "grad_norm": 0.7385532765716779, + "learning_rate": 0.002409, + "loss": 4.6034, + "step": 803 + }, + { + "epoch": 0.00804, + "grad_norm": 0.6763586866376322, + "learning_rate": 0.002412, + "loss": 4.609, + "step": 804 + }, + { + "epoch": 0.00805, + "grad_norm": 0.4984516329675046, + "learning_rate": 0.002415, + "loss": 4.5958, + "step": 805 + }, + { + "epoch": 0.00806, + "grad_norm": 0.6422151194456196, + "learning_rate": 0.002418, + "loss": 4.6202, + "step": 806 + }, + { + "epoch": 0.00807, + "grad_norm": 0.6520347813599764, + "learning_rate": 0.0024210000000000004, + "loss": 4.6111, + "step": 807 + }, + { + "epoch": 0.00808, + "grad_norm": 0.6466117231395049, + "learning_rate": 0.0024240000000000004, + "loss": 4.5938, + "step": 808 + }, + { + "epoch": 0.00809, + "grad_norm": 0.7173763198404414, + "learning_rate": 0.0024270000000000003, + "loss": 4.5875, + "step": 809 + }, + { + "epoch": 0.0081, + "grad_norm": 0.7630708316428362, + "learning_rate": 0.0024300000000000003, + "loss": 4.6257, + "step": 810 + }, + { + "epoch": 0.00811, + "grad_norm": 0.6597258634486322, + "learning_rate": 0.0024330000000000003, + "loss": 4.5916, + "step": 811 + }, + { + "epoch": 0.00812, + "grad_norm": 0.5596135689444884, + "learning_rate": 0.0024360000000000002, + "loss": 4.5639, + "step": 812 + }, + { + "epoch": 0.00813, + "grad_norm": 0.5455291997717118, + "learning_rate": 0.0024389999999999998, + "loss": 4.6039, + "step": 813 + }, + { + "epoch": 0.00814, + "grad_norm": 0.643394229982758, + "learning_rate": 0.0024419999999999997, + "loss": 4.6125, + "step": 814 + }, + { + "epoch": 0.00815, + "grad_norm": 0.7154153020141445, + "learning_rate": 0.0024449999999999997, + "loss": 4.6204, + "step": 815 + }, + { + "epoch": 0.00816, + "grad_norm": 0.7061775094598466, + "learning_rate": 0.002448, + "loss": 4.6005, + "step": 816 + }, + { + "epoch": 0.00817, + "grad_norm": 0.7497949470997279, + "learning_rate": 0.002451, + "loss": 4.6054, + "step": 817 + }, + { + "epoch": 0.00818, + "grad_norm": 0.8303821582765404, + "learning_rate": 0.002454, + "loss": 4.6047, + "step": 818 + }, + { + "epoch": 0.00819, + "grad_norm": 0.7870772660210968, + "learning_rate": 0.002457, + "loss": 4.6129, + "step": 819 + }, + { + "epoch": 0.0082, + "grad_norm": 0.8045799279983024, + "learning_rate": 0.00246, + "loss": 4.5959, + "step": 820 + }, + { + "epoch": 0.00821, + "grad_norm": 0.6847814476141592, + "learning_rate": 0.002463, + "loss": 4.5753, + "step": 821 + }, + { + "epoch": 0.00822, + "grad_norm": 0.6767342952639601, + "learning_rate": 0.002466, + "loss": 4.5743, + "step": 822 + }, + { + "epoch": 0.00823, + "grad_norm": 0.7814972414925468, + "learning_rate": 0.002469, + "loss": 4.5892, + "step": 823 + }, + { + "epoch": 0.00824, + "grad_norm": 0.7970591329920511, + "learning_rate": 0.002472, + "loss": 4.6124, + "step": 824 + }, + { + "epoch": 0.00825, + "grad_norm": 0.7237059608391628, + "learning_rate": 0.0024749999999999998, + "loss": 4.6158, + "step": 825 + }, + { + "epoch": 0.00826, + "grad_norm": 0.8825063151832095, + "learning_rate": 0.0024779999999999997, + "loss": 4.5838, + "step": 826 + }, + { + "epoch": 0.00827, + "grad_norm": 0.895197450200386, + "learning_rate": 0.002481, + "loss": 4.6224, + "step": 827 + }, + { + "epoch": 0.00828, + "grad_norm": 0.7760924275661016, + "learning_rate": 0.002484, + "loss": 4.6087, + "step": 828 + }, + { + "epoch": 0.00829, + "grad_norm": 0.8835354563964355, + "learning_rate": 0.002487, + "loss": 4.6157, + "step": 829 + }, + { + "epoch": 0.0083, + "grad_norm": 0.8984013262238549, + "learning_rate": 0.00249, + "loss": 4.6232, + "step": 830 + }, + { + "epoch": 0.00831, + "grad_norm": 0.8653834946214679, + "learning_rate": 0.002493, + "loss": 4.6038, + "step": 831 + }, + { + "epoch": 0.00832, + "grad_norm": 0.8641563564291244, + "learning_rate": 0.002496, + "loss": 4.6089, + "step": 832 + }, + { + "epoch": 0.00833, + "grad_norm": 0.8849940351412556, + "learning_rate": 0.002499, + "loss": 4.6069, + "step": 833 + }, + { + "epoch": 0.00834, + "grad_norm": 0.9333438912141526, + "learning_rate": 0.002502, + "loss": 4.6165, + "step": 834 + }, + { + "epoch": 0.00835, + "grad_norm": 0.8853225829488065, + "learning_rate": 0.002505, + "loss": 4.6108, + "step": 835 + }, + { + "epoch": 0.00836, + "grad_norm": 0.9042165265638229, + "learning_rate": 0.002508, + "loss": 4.6433, + "step": 836 + }, + { + "epoch": 0.00837, + "grad_norm": 0.8034248166459536, + "learning_rate": 0.0025109999999999998, + "loss": 4.5745, + "step": 837 + }, + { + "epoch": 0.00838, + "grad_norm": 0.8086998635281932, + "learning_rate": 0.0025139999999999997, + "loss": 4.6114, + "step": 838 + }, + { + "epoch": 0.00839, + "grad_norm": 0.587375456181757, + "learning_rate": 0.002517, + "loss": 4.5954, + "step": 839 + }, + { + "epoch": 0.0084, + "grad_norm": 0.6561997865962053, + "learning_rate": 0.00252, + "loss": 4.5818, + "step": 840 + }, + { + "epoch": 0.00841, + "grad_norm": 0.6860624256233325, + "learning_rate": 0.002523, + "loss": 4.5813, + "step": 841 + }, + { + "epoch": 0.00842, + "grad_norm": 0.6384378914075728, + "learning_rate": 0.002526, + "loss": 4.5783, + "step": 842 + }, + { + "epoch": 0.00843, + "grad_norm": 0.5397302637540236, + "learning_rate": 0.002529, + "loss": 4.5728, + "step": 843 + }, + { + "epoch": 0.00844, + "grad_norm": 0.5753282450552862, + "learning_rate": 0.002532, + "loss": 4.5912, + "step": 844 + }, + { + "epoch": 0.00845, + "grad_norm": 0.605188801053789, + "learning_rate": 0.002535, + "loss": 4.5588, + "step": 845 + }, + { + "epoch": 0.00846, + "grad_norm": 0.523231421872222, + "learning_rate": 0.002538, + "loss": 4.5771, + "step": 846 + }, + { + "epoch": 0.00847, + "grad_norm": 0.4442194357732046, + "learning_rate": 0.002541, + "loss": 4.5539, + "step": 847 + }, + { + "epoch": 0.00848, + "grad_norm": 0.4340595323225716, + "learning_rate": 0.002544, + "loss": 4.5562, + "step": 848 + }, + { + "epoch": 0.00849, + "grad_norm": 0.40987686572717924, + "learning_rate": 0.002547, + "loss": 4.5471, + "step": 849 + }, + { + "epoch": 0.0085, + "grad_norm": 0.38823479672015115, + "learning_rate": 0.00255, + "loss": 4.5547, + "step": 850 + }, + { + "epoch": 0.00851, + "grad_norm": 0.42997075207568086, + "learning_rate": 0.002553, + "loss": 4.5733, + "step": 851 + }, + { + "epoch": 0.00852, + "grad_norm": 0.5192269912554525, + "learning_rate": 0.002556, + "loss": 4.5624, + "step": 852 + }, + { + "epoch": 0.00853, + "grad_norm": 0.6626364019198889, + "learning_rate": 0.002559, + "loss": 4.5465, + "step": 853 + }, + { + "epoch": 0.00854, + "grad_norm": 0.9475750496728129, + "learning_rate": 0.002562, + "loss": 4.5544, + "step": 854 + }, + { + "epoch": 0.00855, + "grad_norm": 0.9196861875673602, + "learning_rate": 0.002565, + "loss": 4.5554, + "step": 855 + }, + { + "epoch": 0.00856, + "grad_norm": 0.6972970172901616, + "learning_rate": 0.002568, + "loss": 4.5747, + "step": 856 + }, + { + "epoch": 0.00857, + "grad_norm": 0.7425487517589463, + "learning_rate": 0.002571, + "loss": 4.5717, + "step": 857 + }, + { + "epoch": 0.00858, + "grad_norm": 0.678911102843075, + "learning_rate": 0.002574, + "loss": 4.5576, + "step": 858 + }, + { + "epoch": 0.00859, + "grad_norm": 0.6680695283315139, + "learning_rate": 0.002577, + "loss": 4.5494, + "step": 859 + }, + { + "epoch": 0.0086, + "grad_norm": 0.7323389776740598, + "learning_rate": 0.00258, + "loss": 4.5797, + "step": 860 + }, + { + "epoch": 0.00861, + "grad_norm": 0.6564756555111457, + "learning_rate": 0.0025830000000000002, + "loss": 4.5928, + "step": 861 + }, + { + "epoch": 0.00862, + "grad_norm": 0.6458920384469554, + "learning_rate": 0.002586, + "loss": 4.5402, + "step": 862 + }, + { + "epoch": 0.00863, + "grad_norm": 0.56307833723133, + "learning_rate": 0.002589, + "loss": 4.5659, + "step": 863 + }, + { + "epoch": 0.00864, + "grad_norm": 0.6301527121328465, + "learning_rate": 0.002592, + "loss": 4.5169, + "step": 864 + }, + { + "epoch": 0.00865, + "grad_norm": 0.5741592962502369, + "learning_rate": 0.002595, + "loss": 4.5342, + "step": 865 + }, + { + "epoch": 0.00866, + "grad_norm": 0.6185631950272453, + "learning_rate": 0.002598, + "loss": 4.5624, + "step": 866 + }, + { + "epoch": 0.00867, + "grad_norm": 0.7226342011827854, + "learning_rate": 0.002601, + "loss": 4.5538, + "step": 867 + }, + { + "epoch": 0.00868, + "grad_norm": 0.6871063796143311, + "learning_rate": 0.002604, + "loss": 4.5354, + "step": 868 + }, + { + "epoch": 0.00869, + "grad_norm": 0.6146544871717051, + "learning_rate": 0.002607, + "loss": 4.5411, + "step": 869 + }, + { + "epoch": 0.0087, + "grad_norm": 0.5850681754527672, + "learning_rate": 0.00261, + "loss": 4.5279, + "step": 870 + }, + { + "epoch": 0.00871, + "grad_norm": 0.6619050810997609, + "learning_rate": 0.002613, + "loss": 4.5496, + "step": 871 + }, + { + "epoch": 0.00872, + "grad_norm": 0.6309510333250544, + "learning_rate": 0.002616, + "loss": 4.5312, + "step": 872 + }, + { + "epoch": 0.00873, + "grad_norm": 0.620458951387438, + "learning_rate": 0.0026190000000000002, + "loss": 4.5409, + "step": 873 + }, + { + "epoch": 0.00874, + "grad_norm": 0.7675582961358233, + "learning_rate": 0.002622, + "loss": 4.5679, + "step": 874 + }, + { + "epoch": 0.00875, + "grad_norm": 0.8413199287183839, + "learning_rate": 0.002625, + "loss": 4.5217, + "step": 875 + }, + { + "epoch": 0.00876, + "grad_norm": 0.6439305106538762, + "learning_rate": 0.002628, + "loss": 4.5435, + "step": 876 + }, + { + "epoch": 0.00877, + "grad_norm": 0.6155973146282673, + "learning_rate": 0.002631, + "loss": 4.5607, + "step": 877 + }, + { + "epoch": 0.00878, + "grad_norm": 0.6572664983016872, + "learning_rate": 0.002634, + "loss": 4.4959, + "step": 878 + }, + { + "epoch": 0.00879, + "grad_norm": 0.678308084092591, + "learning_rate": 0.002637, + "loss": 4.526, + "step": 879 + }, + { + "epoch": 0.0088, + "grad_norm": 0.682644138366955, + "learning_rate": 0.00264, + "loss": 4.5397, + "step": 880 + }, + { + "epoch": 0.00881, + "grad_norm": 0.6916960221367301, + "learning_rate": 0.002643, + "loss": 4.5424, + "step": 881 + }, + { + "epoch": 0.00882, + "grad_norm": 0.8625989079873627, + "learning_rate": 0.002646, + "loss": 4.5523, + "step": 882 + }, + { + "epoch": 0.00883, + "grad_norm": 0.9115056000231379, + "learning_rate": 0.002649, + "loss": 4.5538, + "step": 883 + }, + { + "epoch": 0.00884, + "grad_norm": 1.1386535317654105, + "learning_rate": 0.0026520000000000003, + "loss": 4.5528, + "step": 884 + }, + { + "epoch": 0.00885, + "grad_norm": 0.9401738564991132, + "learning_rate": 0.0026550000000000002, + "loss": 4.5478, + "step": 885 + }, + { + "epoch": 0.00886, + "grad_norm": 1.0261918984073035, + "learning_rate": 0.002658, + "loss": 4.5785, + "step": 886 + }, + { + "epoch": 0.00887, + "grad_norm": 1.2427377192417122, + "learning_rate": 0.002661, + "loss": 4.6, + "step": 887 + }, + { + "epoch": 0.00888, + "grad_norm": 1.061546131863703, + "learning_rate": 0.002664, + "loss": 4.5885, + "step": 888 + }, + { + "epoch": 0.00889, + "grad_norm": 0.907593068920597, + "learning_rate": 0.002667, + "loss": 4.535, + "step": 889 + }, + { + "epoch": 0.0089, + "grad_norm": 0.8129313603021157, + "learning_rate": 0.00267, + "loss": 4.5855, + "step": 890 + }, + { + "epoch": 0.00891, + "grad_norm": 0.8985699151758613, + "learning_rate": 0.002673, + "loss": 4.5762, + "step": 891 + }, + { + "epoch": 0.00892, + "grad_norm": 0.9328845387585828, + "learning_rate": 0.002676, + "loss": 4.5962, + "step": 892 + }, + { + "epoch": 0.00893, + "grad_norm": 0.9658499469408992, + "learning_rate": 0.002679, + "loss": 4.5743, + "step": 893 + }, + { + "epoch": 0.00894, + "grad_norm": 0.9262060176852961, + "learning_rate": 0.002682, + "loss": 4.5624, + "step": 894 + }, + { + "epoch": 0.00895, + "grad_norm": 1.0340643070320608, + "learning_rate": 0.0026850000000000003, + "loss": 4.5798, + "step": 895 + }, + { + "epoch": 0.00896, + "grad_norm": 0.961568519948311, + "learning_rate": 0.0026880000000000003, + "loss": 4.5724, + "step": 896 + }, + { + "epoch": 0.00897, + "grad_norm": 0.9522979094666822, + "learning_rate": 0.0026910000000000002, + "loss": 4.5901, + "step": 897 + }, + { + "epoch": 0.00898, + "grad_norm": 0.90558868323236, + "learning_rate": 0.002694, + "loss": 4.5802, + "step": 898 + }, + { + "epoch": 0.00899, + "grad_norm": 0.7752291080346148, + "learning_rate": 0.002697, + "loss": 4.5848, + "step": 899 + }, + { + "epoch": 0.009, + "grad_norm": 0.7198757797557334, + "learning_rate": 0.0027, + "loss": 4.5998, + "step": 900 + }, + { + "epoch": 0.00901, + "grad_norm": 0.7328950523365488, + "learning_rate": 0.002703, + "loss": 4.5552, + "step": 901 + }, + { + "epoch": 0.00902, + "grad_norm": 0.5974121750232518, + "learning_rate": 0.002706, + "loss": 4.558, + "step": 902 + }, + { + "epoch": 0.00903, + "grad_norm": 0.5753314236304967, + "learning_rate": 0.002709, + "loss": 4.5534, + "step": 903 + }, + { + "epoch": 0.00904, + "grad_norm": 0.6128643778394687, + "learning_rate": 0.002712, + "loss": 4.5329, + "step": 904 + }, + { + "epoch": 0.00905, + "grad_norm": 0.5691916638541509, + "learning_rate": 0.002715, + "loss": 4.5446, + "step": 905 + }, + { + "epoch": 0.00906, + "grad_norm": 0.5122473095411203, + "learning_rate": 0.002718, + "loss": 4.5538, + "step": 906 + }, + { + "epoch": 0.00907, + "grad_norm": 0.4479956748150508, + "learning_rate": 0.0027210000000000003, + "loss": 4.5603, + "step": 907 + }, + { + "epoch": 0.00908, + "grad_norm": 0.4387217017050332, + "learning_rate": 0.0027240000000000003, + "loss": 4.5304, + "step": 908 + }, + { + "epoch": 0.00909, + "grad_norm": 0.4358600325368994, + "learning_rate": 0.0027270000000000003, + "loss": 4.5454, + "step": 909 + }, + { + "epoch": 0.0091, + "grad_norm": 0.3849373702796595, + "learning_rate": 0.0027300000000000002, + "loss": 4.5477, + "step": 910 + }, + { + "epoch": 0.00911, + "grad_norm": 0.4574591880095763, + "learning_rate": 0.002733, + "loss": 4.5038, + "step": 911 + }, + { + "epoch": 0.00912, + "grad_norm": 0.6121326743360266, + "learning_rate": 0.002736, + "loss": 4.5131, + "step": 912 + }, + { + "epoch": 0.00913, + "grad_norm": 0.6845161679319078, + "learning_rate": 0.002739, + "loss": 4.513, + "step": 913 + }, + { + "epoch": 0.00914, + "grad_norm": 0.7333722528690678, + "learning_rate": 0.002742, + "loss": 4.5432, + "step": 914 + }, + { + "epoch": 0.00915, + "grad_norm": 0.6714442153045657, + "learning_rate": 0.002745, + "loss": 4.5048, + "step": 915 + }, + { + "epoch": 0.00916, + "grad_norm": 0.5861682792098877, + "learning_rate": 0.002748, + "loss": 4.4768, + "step": 916 + }, + { + "epoch": 0.00917, + "grad_norm": 0.7568015257717708, + "learning_rate": 0.002751, + "loss": 4.4985, + "step": 917 + }, + { + "epoch": 0.00918, + "grad_norm": 0.67845085346177, + "learning_rate": 0.0027540000000000004, + "loss": 4.5185, + "step": 918 + }, + { + "epoch": 0.00919, + "grad_norm": 0.6376249584521229, + "learning_rate": 0.0027570000000000003, + "loss": 4.4976, + "step": 919 + }, + { + "epoch": 0.0092, + "grad_norm": 0.5953295490629925, + "learning_rate": 0.0027600000000000003, + "loss": 4.5104, + "step": 920 + }, + { + "epoch": 0.00921, + "grad_norm": 0.5470113372090761, + "learning_rate": 0.0027630000000000003, + "loss": 4.4892, + "step": 921 + }, + { + "epoch": 0.00922, + "grad_norm": 0.5586775976924615, + "learning_rate": 0.0027660000000000002, + "loss": 4.5007, + "step": 922 + }, + { + "epoch": 0.00923, + "grad_norm": 0.5602976125001545, + "learning_rate": 0.002769, + "loss": 4.4767, + "step": 923 + }, + { + "epoch": 0.00924, + "grad_norm": 0.5512257756150963, + "learning_rate": 0.002772, + "loss": 4.4992, + "step": 924 + }, + { + "epoch": 0.00925, + "grad_norm": 0.6104014154605217, + "learning_rate": 0.002775, + "loss": 4.4804, + "step": 925 + }, + { + "epoch": 0.00926, + "grad_norm": 0.627882987139063, + "learning_rate": 0.002778, + "loss": 4.4994, + "step": 926 + }, + { + "epoch": 0.00927, + "grad_norm": 0.5637389395797978, + "learning_rate": 0.002781, + "loss": 4.4582, + "step": 927 + }, + { + "epoch": 0.00928, + "grad_norm": 0.5793012970445044, + "learning_rate": 0.002784, + "loss": 4.505, + "step": 928 + }, + { + "epoch": 0.00929, + "grad_norm": 0.5538600481661593, + "learning_rate": 0.0027870000000000004, + "loss": 4.4867, + "step": 929 + }, + { + "epoch": 0.0093, + "grad_norm": 0.4915384406563099, + "learning_rate": 0.0027900000000000004, + "loss": 4.4652, + "step": 930 + }, + { + "epoch": 0.00931, + "grad_norm": 0.5291025545236148, + "learning_rate": 0.0027930000000000003, + "loss": 4.4796, + "step": 931 + }, + { + "epoch": 0.00932, + "grad_norm": 0.5189072536981689, + "learning_rate": 0.0027960000000000003, + "loss": 4.4881, + "step": 932 + }, + { + "epoch": 0.00933, + "grad_norm": 0.4391692113712953, + "learning_rate": 0.0027990000000000003, + "loss": 4.4495, + "step": 933 + }, + { + "epoch": 0.00934, + "grad_norm": 0.5067956527040228, + "learning_rate": 0.0028020000000000002, + "loss": 4.4811, + "step": 934 + }, + { + "epoch": 0.00935, + "grad_norm": 0.5752082887405049, + "learning_rate": 0.002805, + "loss": 4.4893, + "step": 935 + }, + { + "epoch": 0.00936, + "grad_norm": 0.8597610157431208, + "learning_rate": 0.002808, + "loss": 4.5012, + "step": 936 + }, + { + "epoch": 0.00937, + "grad_norm": 1.0923006968336397, + "learning_rate": 0.002811, + "loss": 4.5462, + "step": 937 + }, + { + "epoch": 0.00938, + "grad_norm": 0.8211742236965874, + "learning_rate": 0.002814, + "loss": 4.5052, + "step": 938 + }, + { + "epoch": 0.00939, + "grad_norm": 0.699189062953733, + "learning_rate": 0.002817, + "loss": 4.4989, + "step": 939 + }, + { + "epoch": 0.0094, + "grad_norm": 0.6251265662758575, + "learning_rate": 0.00282, + "loss": 4.495, + "step": 940 + }, + { + "epoch": 0.00941, + "grad_norm": 0.8614570213641495, + "learning_rate": 0.002823, + "loss": 4.4823, + "step": 941 + }, + { + "epoch": 0.00942, + "grad_norm": 0.9710835908546839, + "learning_rate": 0.002826, + "loss": 4.5282, + "step": 942 + }, + { + "epoch": 0.00943, + "grad_norm": 0.8255156288963836, + "learning_rate": 0.002829, + "loss": 4.5175, + "step": 943 + }, + { + "epoch": 0.00944, + "grad_norm": 0.8529303551550831, + "learning_rate": 0.002832, + "loss": 4.5233, + "step": 944 + }, + { + "epoch": 0.00945, + "grad_norm": 0.8805636588256177, + "learning_rate": 0.002835, + "loss": 4.5307, + "step": 945 + }, + { + "epoch": 0.00946, + "grad_norm": 0.8815394682869104, + "learning_rate": 0.002838, + "loss": 4.545, + "step": 946 + }, + { + "epoch": 0.00947, + "grad_norm": 0.9792537736888025, + "learning_rate": 0.0028409999999999998, + "loss": 4.5259, + "step": 947 + }, + { + "epoch": 0.00948, + "grad_norm": 1.16480627354709, + "learning_rate": 0.0028439999999999997, + "loss": 4.5258, + "step": 948 + }, + { + "epoch": 0.00949, + "grad_norm": 0.9311558570406092, + "learning_rate": 0.002847, + "loss": 4.5678, + "step": 949 + }, + { + "epoch": 0.0095, + "grad_norm": 0.9151500308490546, + "learning_rate": 0.00285, + "loss": 4.5566, + "step": 950 + }, + { + "epoch": 0.00951, + "grad_norm": 0.8788038193040383, + "learning_rate": 0.002853, + "loss": 4.539, + "step": 951 + }, + { + "epoch": 0.00952, + "grad_norm": 0.6883343465241535, + "learning_rate": 0.002856, + "loss": 4.5492, + "step": 952 + }, + { + "epoch": 0.00953, + "grad_norm": 0.7428316161957196, + "learning_rate": 0.002859, + "loss": 4.4912, + "step": 953 + }, + { + "epoch": 0.00954, + "grad_norm": 0.6589075049765231, + "learning_rate": 0.002862, + "loss": 4.5478, + "step": 954 + }, + { + "epoch": 0.00955, + "grad_norm": 0.6374906159512967, + "learning_rate": 0.002865, + "loss": 4.5305, + "step": 955 + }, + { + "epoch": 0.00956, + "grad_norm": 0.7037593758453927, + "learning_rate": 0.002868, + "loss": 4.5123, + "step": 956 + }, + { + "epoch": 0.00957, + "grad_norm": 0.7575837270625414, + "learning_rate": 0.002871, + "loss": 4.5364, + "step": 957 + }, + { + "epoch": 0.00958, + "grad_norm": 0.8136537306721928, + "learning_rate": 0.002874, + "loss": 4.5221, + "step": 958 + }, + { + "epoch": 0.00959, + "grad_norm": 0.7529303499273822, + "learning_rate": 0.002877, + "loss": 4.5134, + "step": 959 + }, + { + "epoch": 0.0096, + "grad_norm": 0.7723176518474248, + "learning_rate": 0.0028799999999999997, + "loss": 4.5128, + "step": 960 + }, + { + "epoch": 0.00961, + "grad_norm": 0.6826701565045171, + "learning_rate": 0.002883, + "loss": 4.4802, + "step": 961 + }, + { + "epoch": 0.00962, + "grad_norm": 0.6022941634910601, + "learning_rate": 0.002886, + "loss": 4.5086, + "step": 962 + }, + { + "epoch": 0.00963, + "grad_norm": 0.5423609760386227, + "learning_rate": 0.002889, + "loss": 4.5025, + "step": 963 + }, + { + "epoch": 0.00964, + "grad_norm": 0.5430502092899187, + "learning_rate": 0.002892, + "loss": 4.493, + "step": 964 + }, + { + "epoch": 0.00965, + "grad_norm": 0.5046734063237629, + "learning_rate": 0.002895, + "loss": 4.4643, + "step": 965 + }, + { + "epoch": 0.00966, + "grad_norm": 0.4625107724617308, + "learning_rate": 0.002898, + "loss": 4.4758, + "step": 966 + }, + { + "epoch": 0.00967, + "grad_norm": 0.4403845150464526, + "learning_rate": 0.002901, + "loss": 4.5117, + "step": 967 + }, + { + "epoch": 0.00968, + "grad_norm": 0.40739936008519134, + "learning_rate": 0.002904, + "loss": 4.4717, + "step": 968 + }, + { + "epoch": 0.00969, + "grad_norm": 0.41620271754109195, + "learning_rate": 0.002907, + "loss": 4.5059, + "step": 969 + }, + { + "epoch": 0.0097, + "grad_norm": 0.3884080596123629, + "learning_rate": 0.00291, + "loss": 4.4561, + "step": 970 + }, + { + "epoch": 0.00971, + "grad_norm": 0.40963591049792675, + "learning_rate": 0.002913, + "loss": 4.4624, + "step": 971 + }, + { + "epoch": 0.00972, + "grad_norm": 0.427860611994679, + "learning_rate": 0.002916, + "loss": 4.4693, + "step": 972 + }, + { + "epoch": 0.00973, + "grad_norm": 0.6266830722962007, + "learning_rate": 0.002919, + "loss": 4.4689, + "step": 973 + }, + { + "epoch": 0.00974, + "grad_norm": 0.7795241673951117, + "learning_rate": 0.002922, + "loss": 4.4734, + "step": 974 + }, + { + "epoch": 0.00975, + "grad_norm": 0.7392088002205494, + "learning_rate": 0.002925, + "loss": 4.4915, + "step": 975 + }, + { + "epoch": 0.00976, + "grad_norm": 0.6090919745791438, + "learning_rate": 0.002928, + "loss": 4.4688, + "step": 976 + }, + { + "epoch": 0.00977, + "grad_norm": 0.6721667952819499, + "learning_rate": 0.002931, + "loss": 4.5114, + "step": 977 + }, + { + "epoch": 0.00978, + "grad_norm": 0.7157395676490952, + "learning_rate": 0.002934, + "loss": 4.4828, + "step": 978 + }, + { + "epoch": 0.00979, + "grad_norm": 0.6673604586868628, + "learning_rate": 0.002937, + "loss": 4.4724, + "step": 979 + }, + { + "epoch": 0.0098, + "grad_norm": 0.6241827666303977, + "learning_rate": 0.00294, + "loss": 4.4641, + "step": 980 + }, + { + "epoch": 0.00981, + "grad_norm": 0.6477735028588306, + "learning_rate": 0.002943, + "loss": 4.479, + "step": 981 + }, + { + "epoch": 0.00982, + "grad_norm": 0.5903229373466107, + "learning_rate": 0.002946, + "loss": 4.4717, + "step": 982 + }, + { + "epoch": 0.00983, + "grad_norm": 0.5367081021491611, + "learning_rate": 0.0029490000000000002, + "loss": 4.4684, + "step": 983 + }, + { + "epoch": 0.00984, + "grad_norm": 0.55855952891459, + "learning_rate": 0.002952, + "loss": 4.455, + "step": 984 + }, + { + "epoch": 0.00985, + "grad_norm": 0.6566800838357816, + "learning_rate": 0.002955, + "loss": 4.4789, + "step": 985 + }, + { + "epoch": 0.00986, + "grad_norm": 0.7602203626787503, + "learning_rate": 0.002958, + "loss": 4.4461, + "step": 986 + }, + { + "epoch": 0.00987, + "grad_norm": 0.7515002206199092, + "learning_rate": 0.002961, + "loss": 4.4656, + "step": 987 + }, + { + "epoch": 0.00988, + "grad_norm": 0.7510028045825011, + "learning_rate": 0.002964, + "loss": 4.497, + "step": 988 + }, + { + "epoch": 0.00989, + "grad_norm": 0.6686524266681442, + "learning_rate": 0.002967, + "loss": 4.438, + "step": 989 + }, + { + "epoch": 0.0099, + "grad_norm": 0.7110181433203292, + "learning_rate": 0.00297, + "loss": 4.4759, + "step": 990 + }, + { + "epoch": 0.00991, + "grad_norm": 0.9487700940206629, + "learning_rate": 0.002973, + "loss": 4.4909, + "step": 991 + }, + { + "epoch": 0.00992, + "grad_norm": 1.0128664486589338, + "learning_rate": 0.002976, + "loss": 4.4936, + "step": 992 + }, + { + "epoch": 0.00993, + "grad_norm": 0.776315556008045, + "learning_rate": 0.002979, + "loss": 4.4927, + "step": 993 + }, + { + "epoch": 0.00994, + "grad_norm": 0.6954881784010936, + "learning_rate": 0.002982, + "loss": 4.4894, + "step": 994 + }, + { + "epoch": 0.00995, + "grad_norm": 0.8343170542469196, + "learning_rate": 0.0029850000000000002, + "loss": 4.4966, + "step": 995 + }, + { + "epoch": 0.00996, + "grad_norm": 0.8064256733311401, + "learning_rate": 0.002988, + "loss": 4.4974, + "step": 996 + }, + { + "epoch": 0.00997, + "grad_norm": 0.7600837252115415, + "learning_rate": 0.002991, + "loss": 4.4859, + "step": 997 + }, + { + "epoch": 0.00998, + "grad_norm": 0.7413851219310601, + "learning_rate": 0.002994, + "loss": 4.4878, + "step": 998 + }, + { + "epoch": 0.00999, + "grad_norm": 0.7789142980889687, + "learning_rate": 0.002997, + "loss": 4.4774, + "step": 999 + }, + { + "epoch": 0.01, + "grad_norm": 0.5973093879700552, + "learning_rate": 0.003, + "loss": 4.4903, + "step": 1000 + }, + { + "epoch": 0.01001, + "grad_norm": 0.5861183186896064, + "learning_rate": 0.003, + "loss": 4.4874, + "step": 1001 + }, + { + "epoch": 0.01002, + "grad_norm": 0.6212400302116947, + "learning_rate": 0.003, + "loss": 4.4557, + "step": 1002 + }, + { + "epoch": 0.01003, + "grad_norm": 0.6083357500633214, + "learning_rate": 0.003, + "loss": 4.4779, + "step": 1003 + }, + { + "epoch": 0.01004, + "grad_norm": 0.5830404519486425, + "learning_rate": 0.003, + "loss": 4.4653, + "step": 1004 + }, + { + "epoch": 0.01005, + "grad_norm": 0.7477664572876674, + "learning_rate": 0.003, + "loss": 4.4465, + "step": 1005 + }, + { + "epoch": 0.01006, + "grad_norm": 0.7841384646860249, + "learning_rate": 0.003, + "loss": 4.449, + "step": 1006 + }, + { + "epoch": 0.01007, + "grad_norm": 0.7235135066769743, + "learning_rate": 0.003, + "loss": 4.4927, + "step": 1007 + }, + { + "epoch": 0.01008, + "grad_norm": 0.6474532468386683, + "learning_rate": 0.003, + "loss": 4.4993, + "step": 1008 + }, + { + "epoch": 0.01009, + "grad_norm": 0.5980408215414467, + "learning_rate": 0.003, + "loss": 4.4651, + "step": 1009 + }, + { + "epoch": 0.0101, + "grad_norm": 0.5537282779479755, + "learning_rate": 0.003, + "loss": 4.4693, + "step": 1010 + }, + { + "epoch": 0.01011, + "grad_norm": 0.5708819324924145, + "learning_rate": 0.003, + "loss": 4.4785, + "step": 1011 + }, + { + "epoch": 0.01012, + "grad_norm": 0.4974883261450977, + "learning_rate": 0.003, + "loss": 4.4576, + "step": 1012 + }, + { + "epoch": 0.01013, + "grad_norm": 0.4665247551568978, + "learning_rate": 0.003, + "loss": 4.4798, + "step": 1013 + }, + { + "epoch": 0.01014, + "grad_norm": 0.42165803422689113, + "learning_rate": 0.003, + "loss": 4.4182, + "step": 1014 + }, + { + "epoch": 0.01015, + "grad_norm": 0.4067134475964037, + "learning_rate": 0.003, + "loss": 4.4608, + "step": 1015 + }, + { + "epoch": 0.01016, + "grad_norm": 0.5123147411688123, + "learning_rate": 0.003, + "loss": 4.4714, + "step": 1016 + }, + { + "epoch": 0.01017, + "grad_norm": 0.6275035755668664, + "learning_rate": 0.003, + "loss": 4.4159, + "step": 1017 + }, + { + "epoch": 0.01018, + "grad_norm": 0.8581692725167155, + "learning_rate": 0.003, + "loss": 4.4817, + "step": 1018 + }, + { + "epoch": 0.01019, + "grad_norm": 0.8393947155453974, + "learning_rate": 0.003, + "loss": 4.4755, + "step": 1019 + }, + { + "epoch": 0.0102, + "grad_norm": 0.6933428051970444, + "learning_rate": 0.003, + "loss": 4.4592, + "step": 1020 + }, + { + "epoch": 0.01021, + "grad_norm": 0.7990025577528214, + "learning_rate": 0.003, + "loss": 4.4512, + "step": 1021 + }, + { + "epoch": 0.01022, + "grad_norm": 0.8110623623420127, + "learning_rate": 0.003, + "loss": 4.4643, + "step": 1022 + }, + { + "epoch": 0.01023, + "grad_norm": 0.7434736091722897, + "learning_rate": 0.003, + "loss": 4.4609, + "step": 1023 + }, + { + "epoch": 0.01024, + "grad_norm": 0.7213203816141717, + "learning_rate": 0.003, + "loss": 4.417, + "step": 1024 + }, + { + "epoch": 0.01025, + "grad_norm": 0.7010589428906292, + "learning_rate": 0.003, + "loss": 4.4518, + "step": 1025 + }, + { + "epoch": 0.01026, + "grad_norm": 0.6280712957780686, + "learning_rate": 0.003, + "loss": 4.44, + "step": 1026 + }, + { + "epoch": 0.01027, + "grad_norm": 0.5410145302803399, + "learning_rate": 0.003, + "loss": 4.4579, + "step": 1027 + }, + { + "epoch": 0.01028, + "grad_norm": 0.4828232064846382, + "learning_rate": 0.003, + "loss": 4.4353, + "step": 1028 + }, + { + "epoch": 0.01029, + "grad_norm": 0.4739043118914256, + "learning_rate": 0.003, + "loss": 4.412, + "step": 1029 + }, + { + "epoch": 0.0103, + "grad_norm": 0.45225551629540334, + "learning_rate": 0.003, + "loss": 4.4408, + "step": 1030 + }, + { + "epoch": 0.01031, + "grad_norm": 0.40901639056794215, + "learning_rate": 0.003, + "loss": 4.4521, + "step": 1031 + }, + { + "epoch": 0.01032, + "grad_norm": 0.4266157306452721, + "learning_rate": 0.003, + "loss": 4.4241, + "step": 1032 + }, + { + "epoch": 0.01033, + "grad_norm": 0.476318572377966, + "learning_rate": 0.003, + "loss": 4.4623, + "step": 1033 + }, + { + "epoch": 0.01034, + "grad_norm": 0.4976367752405493, + "learning_rate": 0.003, + "loss": 4.4286, + "step": 1034 + }, + { + "epoch": 0.01035, + "grad_norm": 0.48706616162522565, + "learning_rate": 0.003, + "loss": 4.427, + "step": 1035 + }, + { + "epoch": 0.01036, + "grad_norm": 0.5540613528135478, + "learning_rate": 0.003, + "loss": 4.4185, + "step": 1036 + }, + { + "epoch": 0.01037, + "grad_norm": 0.5585932570414466, + "learning_rate": 0.003, + "loss": 4.4458, + "step": 1037 + }, + { + "epoch": 0.01038, + "grad_norm": 0.5014561937587317, + "learning_rate": 0.003, + "loss": 4.44, + "step": 1038 + }, + { + "epoch": 0.01039, + "grad_norm": 0.5070723079813054, + "learning_rate": 0.003, + "loss": 4.4296, + "step": 1039 + }, + { + "epoch": 0.0104, + "grad_norm": 0.5609823517547058, + "learning_rate": 0.003, + "loss": 4.4462, + "step": 1040 + }, + { + "epoch": 0.01041, + "grad_norm": 0.664341991841828, + "learning_rate": 0.003, + "loss": 4.4332, + "step": 1041 + }, + { + "epoch": 0.01042, + "grad_norm": 0.631896992045698, + "learning_rate": 0.003, + "loss": 4.458, + "step": 1042 + }, + { + "epoch": 0.01043, + "grad_norm": 0.8655430264592057, + "learning_rate": 0.003, + "loss": 4.4103, + "step": 1043 + }, + { + "epoch": 0.01044, + "grad_norm": 1.0843670364695495, + "learning_rate": 0.003, + "loss": 4.4702, + "step": 1044 + }, + { + "epoch": 0.01045, + "grad_norm": 0.9544830822145469, + "learning_rate": 0.003, + "loss": 4.4986, + "step": 1045 + }, + { + "epoch": 0.01046, + "grad_norm": 0.9579295900139632, + "learning_rate": 0.003, + "loss": 4.4718, + "step": 1046 + }, + { + "epoch": 0.01047, + "grad_norm": 0.9275123969574217, + "learning_rate": 0.003, + "loss": 4.4903, + "step": 1047 + }, + { + "epoch": 0.01048, + "grad_norm": 0.8417384302831535, + "learning_rate": 0.003, + "loss": 4.4694, + "step": 1048 + }, + { + "epoch": 0.01049, + "grad_norm": 0.7711730353324007, + "learning_rate": 0.003, + "loss": 4.449, + "step": 1049 + }, + { + "epoch": 0.0105, + "grad_norm": 0.747892326408584, + "learning_rate": 0.003, + "loss": 4.4932, + "step": 1050 + }, + { + "epoch": 0.01051, + "grad_norm": 0.8100969209243005, + "learning_rate": 0.003, + "loss": 4.4724, + "step": 1051 + }, + { + "epoch": 0.01052, + "grad_norm": 0.7211696952483453, + "learning_rate": 0.003, + "loss": 4.4859, + "step": 1052 + }, + { + "epoch": 0.01053, + "grad_norm": 0.761937042029771, + "learning_rate": 0.003, + "loss": 4.4907, + "step": 1053 + }, + { + "epoch": 0.01054, + "grad_norm": 0.791018085034848, + "learning_rate": 0.003, + "loss": 4.4849, + "step": 1054 + }, + { + "epoch": 0.01055, + "grad_norm": 0.76513637315221, + "learning_rate": 0.003, + "loss": 4.464, + "step": 1055 + }, + { + "epoch": 0.01056, + "grad_norm": 0.844999763652046, + "learning_rate": 0.003, + "loss": 4.4686, + "step": 1056 + }, + { + "epoch": 0.01057, + "grad_norm": 0.7706046759321704, + "learning_rate": 0.003, + "loss": 4.476, + "step": 1057 + }, + { + "epoch": 0.01058, + "grad_norm": 0.6536852085679028, + "learning_rate": 0.003, + "loss": 4.4607, + "step": 1058 + }, + { + "epoch": 0.01059, + "grad_norm": 0.7239862733740466, + "learning_rate": 0.003, + "loss": 4.4656, + "step": 1059 + }, + { + "epoch": 0.0106, + "grad_norm": 0.8167528488457484, + "learning_rate": 0.003, + "loss": 4.4597, + "step": 1060 + }, + { + "epoch": 0.01061, + "grad_norm": 0.6791440431932436, + "learning_rate": 0.003, + "loss": 4.4551, + "step": 1061 + }, + { + "epoch": 0.01062, + "grad_norm": 0.7024673414145854, + "learning_rate": 0.003, + "loss": 4.4488, + "step": 1062 + }, + { + "epoch": 0.01063, + "grad_norm": 0.5950427080817663, + "learning_rate": 0.003, + "loss": 4.4376, + "step": 1063 + }, + { + "epoch": 0.01064, + "grad_norm": 0.4616517916448735, + "learning_rate": 0.003, + "loss": 4.4487, + "step": 1064 + }, + { + "epoch": 0.01065, + "grad_norm": 0.43744796242075984, + "learning_rate": 0.003, + "loss": 4.4449, + "step": 1065 + }, + { + "epoch": 0.01066, + "grad_norm": 0.3954250450266796, + "learning_rate": 0.003, + "loss": 4.4441, + "step": 1066 + }, + { + "epoch": 0.01067, + "grad_norm": 0.44911980407136426, + "learning_rate": 0.003, + "loss": 4.4535, + "step": 1067 + }, + { + "epoch": 0.01068, + "grad_norm": 0.6121110106038351, + "learning_rate": 0.003, + "loss": 4.4381, + "step": 1068 + }, + { + "epoch": 0.01069, + "grad_norm": 0.8069569048435549, + "learning_rate": 0.003, + "loss": 4.4677, + "step": 1069 + }, + { + "epoch": 0.0107, + "grad_norm": 0.9567755776549173, + "learning_rate": 0.003, + "loss": 4.5257, + "step": 1070 + }, + { + "epoch": 0.01071, + "grad_norm": 0.7710306613088521, + "learning_rate": 0.003, + "loss": 4.4413, + "step": 1071 + }, + { + "epoch": 0.01072, + "grad_norm": 0.6442072811444085, + "learning_rate": 0.003, + "loss": 4.4742, + "step": 1072 + }, + { + "epoch": 0.01073, + "grad_norm": 0.6426873609153096, + "learning_rate": 0.003, + "loss": 4.464, + "step": 1073 + }, + { + "epoch": 0.01074, + "grad_norm": 0.5296718591337678, + "learning_rate": 0.003, + "loss": 4.4244, + "step": 1074 + }, + { + "epoch": 0.01075, + "grad_norm": 0.4598820284284391, + "learning_rate": 0.003, + "loss": 4.4589, + "step": 1075 + }, + { + "epoch": 0.01076, + "grad_norm": 0.4671476178414801, + "learning_rate": 0.003, + "loss": 4.4514, + "step": 1076 + }, + { + "epoch": 0.01077, + "grad_norm": 0.5378166305898745, + "learning_rate": 0.003, + "loss": 4.4483, + "step": 1077 + }, + { + "epoch": 0.01078, + "grad_norm": 0.4897385837475064, + "learning_rate": 0.003, + "loss": 4.3993, + "step": 1078 + }, + { + "epoch": 0.01079, + "grad_norm": 0.40845103276318817, + "learning_rate": 0.003, + "loss": 4.4297, + "step": 1079 + }, + { + "epoch": 0.0108, + "grad_norm": 0.36771124599267696, + "learning_rate": 0.003, + "loss": 4.4028, + "step": 1080 + }, + { + "epoch": 0.01081, + "grad_norm": 0.4088680521129332, + "learning_rate": 0.003, + "loss": 4.41, + "step": 1081 + }, + { + "epoch": 0.01082, + "grad_norm": 0.41501548740150246, + "learning_rate": 0.003, + "loss": 4.4279, + "step": 1082 + }, + { + "epoch": 0.01083, + "grad_norm": 0.3846754433134812, + "learning_rate": 0.003, + "loss": 4.4131, + "step": 1083 + }, + { + "epoch": 0.01084, + "grad_norm": 0.3759434314813712, + "learning_rate": 0.003, + "loss": 4.4174, + "step": 1084 + }, + { + "epoch": 0.01085, + "grad_norm": 0.38199974093520966, + "learning_rate": 0.003, + "loss": 4.4048, + "step": 1085 + }, + { + "epoch": 0.01086, + "grad_norm": 0.417068769159195, + "learning_rate": 0.003, + "loss": 4.3806, + "step": 1086 + }, + { + "epoch": 0.01087, + "grad_norm": 0.4600653555422747, + "learning_rate": 0.003, + "loss": 4.3971, + "step": 1087 + }, + { + "epoch": 0.01088, + "grad_norm": 0.46093681942570613, + "learning_rate": 0.003, + "loss": 4.4326, + "step": 1088 + }, + { + "epoch": 0.01089, + "grad_norm": 0.3917178547896987, + "learning_rate": 0.003, + "loss": 4.3982, + "step": 1089 + }, + { + "epoch": 0.0109, + "grad_norm": 0.40807888361788897, + "learning_rate": 0.003, + "loss": 4.4241, + "step": 1090 + }, + { + "epoch": 0.01091, + "grad_norm": 0.48077039467346944, + "learning_rate": 0.003, + "loss": 4.3944, + "step": 1091 + }, + { + "epoch": 0.01092, + "grad_norm": 0.5412841894896883, + "learning_rate": 0.003, + "loss": 4.3919, + "step": 1092 + }, + { + "epoch": 0.01093, + "grad_norm": 0.6050990959157666, + "learning_rate": 0.003, + "loss": 4.4164, + "step": 1093 + }, + { + "epoch": 0.01094, + "grad_norm": 0.7349015923043832, + "learning_rate": 0.003, + "loss": 4.4107, + "step": 1094 + }, + { + "epoch": 0.01095, + "grad_norm": 0.8311905656610761, + "learning_rate": 0.003, + "loss": 4.4503, + "step": 1095 + }, + { + "epoch": 0.01096, + "grad_norm": 0.7514639212089553, + "learning_rate": 0.003, + "loss": 4.4205, + "step": 1096 + }, + { + "epoch": 0.01097, + "grad_norm": 0.7811392154893482, + "learning_rate": 0.003, + "loss": 4.4143, + "step": 1097 + }, + { + "epoch": 0.01098, + "grad_norm": 0.784974458960817, + "learning_rate": 0.003, + "loss": 4.4358, + "step": 1098 + }, + { + "epoch": 0.01099, + "grad_norm": 0.8366574206737059, + "learning_rate": 0.003, + "loss": 4.4105, + "step": 1099 + }, + { + "epoch": 0.011, + "grad_norm": 0.8308538935944765, + "learning_rate": 0.003, + "loss": 4.4197, + "step": 1100 + }, + { + "epoch": 0.01101, + "grad_norm": 0.7744395626168694, + "learning_rate": 0.003, + "loss": 4.4297, + "step": 1101 + }, + { + "epoch": 0.01102, + "grad_norm": 0.790296389339303, + "learning_rate": 0.003, + "loss": 4.4394, + "step": 1102 + }, + { + "epoch": 0.01103, + "grad_norm": 0.888052985760201, + "learning_rate": 0.003, + "loss": 4.4459, + "step": 1103 + }, + { + "epoch": 0.01104, + "grad_norm": 0.7149547126246749, + "learning_rate": 0.003, + "loss": 4.429, + "step": 1104 + }, + { + "epoch": 0.01105, + "grad_norm": 0.6372550706190871, + "learning_rate": 0.003, + "loss": 4.4469, + "step": 1105 + }, + { + "epoch": 0.01106, + "grad_norm": 0.6914402582964255, + "learning_rate": 0.003, + "loss": 4.4208, + "step": 1106 + }, + { + "epoch": 0.01107, + "grad_norm": 0.6242116385034774, + "learning_rate": 0.003, + "loss": 4.4302, + "step": 1107 + }, + { + "epoch": 0.01108, + "grad_norm": 0.5046709947552265, + "learning_rate": 0.003, + "loss": 4.3916, + "step": 1108 + }, + { + "epoch": 0.01109, + "grad_norm": 0.44296219090623196, + "learning_rate": 0.003, + "loss": 4.4076, + "step": 1109 + }, + { + "epoch": 0.0111, + "grad_norm": 0.42911746175989035, + "learning_rate": 0.003, + "loss": 4.4253, + "step": 1110 + }, + { + "epoch": 0.01111, + "grad_norm": 0.4399471864730289, + "learning_rate": 0.003, + "loss": 4.401, + "step": 1111 + }, + { + "epoch": 0.01112, + "grad_norm": 0.4608476260695391, + "learning_rate": 0.003, + "loss": 4.4197, + "step": 1112 + }, + { + "epoch": 0.01113, + "grad_norm": 0.48813138086989816, + "learning_rate": 0.003, + "loss": 4.388, + "step": 1113 + }, + { + "epoch": 0.01114, + "grad_norm": 0.5565547407761932, + "learning_rate": 0.003, + "loss": 4.4024, + "step": 1114 + }, + { + "epoch": 0.01115, + "grad_norm": 0.6052109602996041, + "learning_rate": 0.003, + "loss": 4.4297, + "step": 1115 + }, + { + "epoch": 0.01116, + "grad_norm": 0.6641979249916196, + "learning_rate": 0.003, + "loss": 4.4207, + "step": 1116 + }, + { + "epoch": 0.01117, + "grad_norm": 0.6083566793907378, + "learning_rate": 0.003, + "loss": 4.4056, + "step": 1117 + }, + { + "epoch": 0.01118, + "grad_norm": 0.5238007850801012, + "learning_rate": 0.003, + "loss": 4.4035, + "step": 1118 + }, + { + "epoch": 0.01119, + "grad_norm": 0.5793839332710249, + "learning_rate": 0.003, + "loss": 4.4112, + "step": 1119 + }, + { + "epoch": 0.0112, + "grad_norm": 0.5969906941980586, + "learning_rate": 0.003, + "loss": 4.3958, + "step": 1120 + }, + { + "epoch": 0.01121, + "grad_norm": 0.6030781006503133, + "learning_rate": 0.003, + "loss": 4.4075, + "step": 1121 + }, + { + "epoch": 0.01122, + "grad_norm": 0.5551009392133712, + "learning_rate": 0.003, + "loss": 4.382, + "step": 1122 + }, + { + "epoch": 0.01123, + "grad_norm": 0.5109047416253846, + "learning_rate": 0.003, + "loss": 4.3689, + "step": 1123 + }, + { + "epoch": 0.01124, + "grad_norm": 0.5407283441136803, + "learning_rate": 0.003, + "loss": 4.4164, + "step": 1124 + }, + { + "epoch": 0.01125, + "grad_norm": 0.538491909735447, + "learning_rate": 0.003, + "loss": 4.3877, + "step": 1125 + }, + { + "epoch": 0.01126, + "grad_norm": 0.4826610539173529, + "learning_rate": 0.003, + "loss": 4.392, + "step": 1126 + }, + { + "epoch": 0.01127, + "grad_norm": 0.41526400192632873, + "learning_rate": 0.003, + "loss": 4.421, + "step": 1127 + }, + { + "epoch": 0.01128, + "grad_norm": 0.4898630750070376, + "learning_rate": 0.003, + "loss": 4.4001, + "step": 1128 + }, + { + "epoch": 0.01129, + "grad_norm": 0.6900647874251847, + "learning_rate": 0.003, + "loss": 4.3988, + "step": 1129 + }, + { + "epoch": 0.0113, + "grad_norm": 0.6715183668037129, + "learning_rate": 0.003, + "loss": 4.3803, + "step": 1130 + }, + { + "epoch": 0.01131, + "grad_norm": 0.5369210919725328, + "learning_rate": 0.003, + "loss": 4.3895, + "step": 1131 + }, + { + "epoch": 0.01132, + "grad_norm": 0.7090208352393726, + "learning_rate": 0.003, + "loss": 4.4216, + "step": 1132 + }, + { + "epoch": 0.01133, + "grad_norm": 0.8026720884346313, + "learning_rate": 0.003, + "loss": 4.3951, + "step": 1133 + }, + { + "epoch": 0.01134, + "grad_norm": 0.7235729334025307, + "learning_rate": 0.003, + "loss": 4.4035, + "step": 1134 + }, + { + "epoch": 0.01135, + "grad_norm": 0.5789287370870784, + "learning_rate": 0.003, + "loss": 4.4255, + "step": 1135 + }, + { + "epoch": 0.01136, + "grad_norm": 0.5872111107371502, + "learning_rate": 0.003, + "loss": 4.4037, + "step": 1136 + }, + { + "epoch": 0.01137, + "grad_norm": 0.5973670624507018, + "learning_rate": 0.003, + "loss": 4.4236, + "step": 1137 + }, + { + "epoch": 0.01138, + "grad_norm": 0.5814805523631724, + "learning_rate": 0.003, + "loss": 4.3808, + "step": 1138 + }, + { + "epoch": 0.01139, + "grad_norm": 0.5749927487336652, + "learning_rate": 0.003, + "loss": 4.389, + "step": 1139 + }, + { + "epoch": 0.0114, + "grad_norm": 0.5177155580469588, + "learning_rate": 0.003, + "loss": 4.3725, + "step": 1140 + }, + { + "epoch": 0.01141, + "grad_norm": 0.5098128826528755, + "learning_rate": 0.003, + "loss": 4.3956, + "step": 1141 + }, + { + "epoch": 0.01142, + "grad_norm": 0.5805509550413942, + "learning_rate": 0.003, + "loss": 4.3764, + "step": 1142 + }, + { + "epoch": 0.01143, + "grad_norm": 0.5868263165469564, + "learning_rate": 0.003, + "loss": 4.3885, + "step": 1143 + }, + { + "epoch": 0.01144, + "grad_norm": 0.6024984248815738, + "learning_rate": 0.003, + "loss": 4.3595, + "step": 1144 + }, + { + "epoch": 0.01145, + "grad_norm": 0.6011422862217553, + "learning_rate": 0.003, + "loss": 4.396, + "step": 1145 + }, + { + "epoch": 0.01146, + "grad_norm": 0.6291389752386858, + "learning_rate": 0.003, + "loss": 4.3845, + "step": 1146 + }, + { + "epoch": 0.01147, + "grad_norm": 0.5858383738783732, + "learning_rate": 0.003, + "loss": 4.3734, + "step": 1147 + }, + { + "epoch": 0.01148, + "grad_norm": 0.5690631266524724, + "learning_rate": 0.003, + "loss": 4.3732, + "step": 1148 + }, + { + "epoch": 0.01149, + "grad_norm": 0.47603449917616164, + "learning_rate": 0.003, + "loss": 4.3879, + "step": 1149 + }, + { + "epoch": 0.0115, + "grad_norm": 0.5321591521167512, + "learning_rate": 0.003, + "loss": 4.3815, + "step": 1150 + }, + { + "epoch": 0.01151, + "grad_norm": 0.5902114027403508, + "learning_rate": 0.003, + "loss": 4.3804, + "step": 1151 + }, + { + "epoch": 0.01152, + "grad_norm": 0.720836087411827, + "learning_rate": 0.003, + "loss": 4.3994, + "step": 1152 + }, + { + "epoch": 0.01153, + "grad_norm": 0.858014811127527, + "learning_rate": 0.003, + "loss": 4.4233, + "step": 1153 + }, + { + "epoch": 0.01154, + "grad_norm": 0.8276250061884273, + "learning_rate": 0.003, + "loss": 4.4035, + "step": 1154 + }, + { + "epoch": 0.01155, + "grad_norm": 0.9316719148571102, + "learning_rate": 0.003, + "loss": 4.4003, + "step": 1155 + }, + { + "epoch": 0.01156, + "grad_norm": 0.9492720138035956, + "learning_rate": 0.003, + "loss": 4.4137, + "step": 1156 + }, + { + "epoch": 0.01157, + "grad_norm": 1.1562300430530366, + "learning_rate": 0.003, + "loss": 4.471, + "step": 1157 + }, + { + "epoch": 0.01158, + "grad_norm": 0.8082778562361257, + "learning_rate": 0.003, + "loss": 4.4212, + "step": 1158 + }, + { + "epoch": 0.01159, + "grad_norm": 0.8296228637182822, + "learning_rate": 0.003, + "loss": 4.4254, + "step": 1159 + }, + { + "epoch": 0.0116, + "grad_norm": 0.9729894065836546, + "learning_rate": 0.003, + "loss": 4.4435, + "step": 1160 + }, + { + "epoch": 0.01161, + "grad_norm": 0.986208959900233, + "learning_rate": 0.003, + "loss": 4.4396, + "step": 1161 + }, + { + "epoch": 0.01162, + "grad_norm": 0.9917874948238292, + "learning_rate": 0.003, + "loss": 4.4412, + "step": 1162 + }, + { + "epoch": 0.01163, + "grad_norm": 1.022093528985829, + "learning_rate": 0.003, + "loss": 4.4772, + "step": 1163 + }, + { + "epoch": 0.01164, + "grad_norm": 0.8274710934079806, + "learning_rate": 0.003, + "loss": 4.4588, + "step": 1164 + }, + { + "epoch": 0.01165, + "grad_norm": 0.8544376883862705, + "learning_rate": 0.003, + "loss": 4.4518, + "step": 1165 + }, + { + "epoch": 0.01166, + "grad_norm": 0.9826443233479919, + "learning_rate": 0.003, + "loss": 4.4596, + "step": 1166 + }, + { + "epoch": 0.01167, + "grad_norm": 0.8638656323292462, + "learning_rate": 0.003, + "loss": 4.427, + "step": 1167 + }, + { + "epoch": 0.01168, + "grad_norm": 0.7965515530800152, + "learning_rate": 0.003, + "loss": 4.4285, + "step": 1168 + }, + { + "epoch": 0.01169, + "grad_norm": 0.621289788302785, + "learning_rate": 0.003, + "loss": 4.4631, + "step": 1169 + }, + { + "epoch": 0.0117, + "grad_norm": 0.5251543844281131, + "learning_rate": 0.003, + "loss": 4.441, + "step": 1170 + }, + { + "epoch": 0.01171, + "grad_norm": 0.6074391771886244, + "learning_rate": 0.003, + "loss": 4.423, + "step": 1171 + }, + { + "epoch": 0.01172, + "grad_norm": 0.5839489329030457, + "learning_rate": 0.003, + "loss": 4.4084, + "step": 1172 + }, + { + "epoch": 0.01173, + "grad_norm": 0.561004783856832, + "learning_rate": 0.003, + "loss": 4.4404, + "step": 1173 + }, + { + "epoch": 0.01174, + "grad_norm": 0.5821159441803843, + "learning_rate": 0.003, + "loss": 4.4125, + "step": 1174 + }, + { + "epoch": 0.01175, + "grad_norm": 0.5941266163389948, + "learning_rate": 0.003, + "loss": 4.3949, + "step": 1175 + }, + { + "epoch": 0.01176, + "grad_norm": 0.6502157975630524, + "learning_rate": 0.003, + "loss": 4.432, + "step": 1176 + }, + { + "epoch": 0.01177, + "grad_norm": 0.8335241720178251, + "learning_rate": 0.003, + "loss": 4.4041, + "step": 1177 + }, + { + "epoch": 0.01178, + "grad_norm": 1.0000155108324966, + "learning_rate": 0.003, + "loss": 4.4462, + "step": 1178 + }, + { + "epoch": 0.01179, + "grad_norm": 0.9223184266441969, + "learning_rate": 0.003, + "loss": 4.4309, + "step": 1179 + }, + { + "epoch": 0.0118, + "grad_norm": 0.6248971968290589, + "learning_rate": 0.003, + "loss": 4.4722, + "step": 1180 + }, + { + "epoch": 0.01181, + "grad_norm": 0.6327502868796897, + "learning_rate": 0.003, + "loss": 4.4606, + "step": 1181 + }, + { + "epoch": 0.01182, + "grad_norm": 0.6032057741631047, + "learning_rate": 0.003, + "loss": 4.4179, + "step": 1182 + }, + { + "epoch": 0.01183, + "grad_norm": 0.6177419756766389, + "learning_rate": 0.003, + "loss": 4.4068, + "step": 1183 + }, + { + "epoch": 0.01184, + "grad_norm": 0.4603977151305031, + "learning_rate": 0.003, + "loss": 4.4388, + "step": 1184 + }, + { + "epoch": 0.01185, + "grad_norm": 0.39775931274104387, + "learning_rate": 0.003, + "loss": 4.4039, + "step": 1185 + }, + { + "epoch": 0.01186, + "grad_norm": 0.43457323577233337, + "learning_rate": 0.003, + "loss": 4.4263, + "step": 1186 + }, + { + "epoch": 0.01187, + "grad_norm": 0.3445355147627072, + "learning_rate": 0.003, + "loss": 4.3841, + "step": 1187 + }, + { + "epoch": 0.01188, + "grad_norm": 0.31570813535234143, + "learning_rate": 0.003, + "loss": 4.3952, + "step": 1188 + }, + { + "epoch": 0.01189, + "grad_norm": 0.3234807855930626, + "learning_rate": 0.003, + "loss": 4.3874, + "step": 1189 + }, + { + "epoch": 0.0119, + "grad_norm": 0.3007248052516205, + "learning_rate": 0.003, + "loss": 4.3694, + "step": 1190 + }, + { + "epoch": 0.01191, + "grad_norm": 0.289692411083719, + "learning_rate": 0.003, + "loss": 4.3476, + "step": 1191 + }, + { + "epoch": 0.01192, + "grad_norm": 0.2772752758842908, + "learning_rate": 0.003, + "loss": 4.3779, + "step": 1192 + }, + { + "epoch": 0.01193, + "grad_norm": 0.2637988935747097, + "learning_rate": 0.003, + "loss": 4.3787, + "step": 1193 + }, + { + "epoch": 0.01194, + "grad_norm": 0.2480894935347364, + "learning_rate": 0.003, + "loss": 4.3791, + "step": 1194 + }, + { + "epoch": 0.01195, + "grad_norm": 0.2744828686479978, + "learning_rate": 0.003, + "loss": 4.3558, + "step": 1195 + }, + { + "epoch": 0.01196, + "grad_norm": 0.2694264980010376, + "learning_rate": 0.003, + "loss": 4.3608, + "step": 1196 + }, + { + "epoch": 0.01197, + "grad_norm": 0.2911995649935099, + "learning_rate": 0.003, + "loss": 4.3243, + "step": 1197 + }, + { + "epoch": 0.01198, + "grad_norm": 0.3049686361769555, + "learning_rate": 0.003, + "loss": 4.3978, + "step": 1198 + }, + { + "epoch": 0.01199, + "grad_norm": 0.4091366608598886, + "learning_rate": 0.003, + "loss": 4.3744, + "step": 1199 + }, + { + "epoch": 0.012, + "grad_norm": 0.5260409537460498, + "learning_rate": 0.003, + "loss": 4.3673, + "step": 1200 + }, + { + "epoch": 0.01201, + "grad_norm": 0.6358041499118771, + "learning_rate": 0.003, + "loss": 4.4018, + "step": 1201 + }, + { + "epoch": 0.01202, + "grad_norm": 0.5923034972204151, + "learning_rate": 0.003, + "loss": 4.3715, + "step": 1202 + }, + { + "epoch": 0.01203, + "grad_norm": 0.49351792103508907, + "learning_rate": 0.003, + "loss": 4.3737, + "step": 1203 + }, + { + "epoch": 0.01204, + "grad_norm": 0.6405227463764683, + "learning_rate": 0.003, + "loss": 4.3772, + "step": 1204 + }, + { + "epoch": 0.01205, + "grad_norm": 0.7035300072883678, + "learning_rate": 0.003, + "loss": 4.3759, + "step": 1205 + }, + { + "epoch": 0.01206, + "grad_norm": 0.6623021094722211, + "learning_rate": 0.003, + "loss": 4.4176, + "step": 1206 + }, + { + "epoch": 0.01207, + "grad_norm": 0.6906429556810797, + "learning_rate": 0.003, + "loss": 4.3398, + "step": 1207 + }, + { + "epoch": 0.01208, + "grad_norm": 0.7416830674542866, + "learning_rate": 0.003, + "loss": 4.4101, + "step": 1208 + }, + { + "epoch": 0.01209, + "grad_norm": 0.7431220077552587, + "learning_rate": 0.003, + "loss": 4.3922, + "step": 1209 + }, + { + "epoch": 0.0121, + "grad_norm": 0.7626206113161277, + "learning_rate": 0.003, + "loss": 4.397, + "step": 1210 + }, + { + "epoch": 0.01211, + "grad_norm": 0.7273957054715338, + "learning_rate": 0.003, + "loss": 4.381, + "step": 1211 + }, + { + "epoch": 0.01212, + "grad_norm": 0.5840029568104971, + "learning_rate": 0.003, + "loss": 4.4002, + "step": 1212 + }, + { + "epoch": 0.01213, + "grad_norm": 0.5534819591366041, + "learning_rate": 0.003, + "loss": 4.3747, + "step": 1213 + }, + { + "epoch": 0.01214, + "grad_norm": 0.5479838122682379, + "learning_rate": 0.003, + "loss": 4.3596, + "step": 1214 + }, + { + "epoch": 0.01215, + "grad_norm": 0.6541197102272803, + "learning_rate": 0.003, + "loss": 4.3736, + "step": 1215 + }, + { + "epoch": 0.01216, + "grad_norm": 0.6998905075239406, + "learning_rate": 0.003, + "loss": 4.369, + "step": 1216 + }, + { + "epoch": 0.01217, + "grad_norm": 0.6272982074936522, + "learning_rate": 0.003, + "loss": 4.3725, + "step": 1217 + }, + { + "epoch": 0.01218, + "grad_norm": 0.6037550089398132, + "learning_rate": 0.003, + "loss": 4.3687, + "step": 1218 + }, + { + "epoch": 0.01219, + "grad_norm": 0.5872809556780119, + "learning_rate": 0.003, + "loss": 4.3682, + "step": 1219 + }, + { + "epoch": 0.0122, + "grad_norm": 0.45035452365127926, + "learning_rate": 0.003, + "loss": 4.3537, + "step": 1220 + }, + { + "epoch": 0.01221, + "grad_norm": 0.41160354363285073, + "learning_rate": 0.003, + "loss": 4.3666, + "step": 1221 + }, + { + "epoch": 0.01222, + "grad_norm": 0.5075830142490442, + "learning_rate": 0.003, + "loss": 4.3511, + "step": 1222 + }, + { + "epoch": 0.01223, + "grad_norm": 0.5662251248767171, + "learning_rate": 0.003, + "loss": 4.3541, + "step": 1223 + }, + { + "epoch": 0.01224, + "grad_norm": 0.690794501870828, + "learning_rate": 0.003, + "loss": 4.3444, + "step": 1224 + }, + { + "epoch": 0.01225, + "grad_norm": 0.675909010637242, + "learning_rate": 0.003, + "loss": 4.3698, + "step": 1225 + }, + { + "epoch": 0.01226, + "grad_norm": 0.6653173755559965, + "learning_rate": 0.003, + "loss": 4.36, + "step": 1226 + }, + { + "epoch": 0.01227, + "grad_norm": 0.8124014201072856, + "learning_rate": 0.003, + "loss": 4.4022, + "step": 1227 + }, + { + "epoch": 0.01228, + "grad_norm": 0.8601590167217709, + "learning_rate": 0.003, + "loss": 4.3869, + "step": 1228 + }, + { + "epoch": 0.01229, + "grad_norm": 0.7816229118417856, + "learning_rate": 0.003, + "loss": 4.3791, + "step": 1229 + }, + { + "epoch": 0.0123, + "grad_norm": 0.7167415795127342, + "learning_rate": 0.003, + "loss": 4.4104, + "step": 1230 + }, + { + "epoch": 0.01231, + "grad_norm": 0.7227237760755019, + "learning_rate": 0.003, + "loss": 4.3789, + "step": 1231 + }, + { + "epoch": 0.01232, + "grad_norm": 0.6338303976227229, + "learning_rate": 0.003, + "loss": 4.3516, + "step": 1232 + }, + { + "epoch": 0.01233, + "grad_norm": 0.5121760575897172, + "learning_rate": 0.003, + "loss": 4.3548, + "step": 1233 + }, + { + "epoch": 0.01234, + "grad_norm": 0.562103643632394, + "learning_rate": 0.003, + "loss": 4.3735, + "step": 1234 + }, + { + "epoch": 0.01235, + "grad_norm": 0.5094370853466236, + "learning_rate": 0.003, + "loss": 4.3603, + "step": 1235 + }, + { + "epoch": 0.01236, + "grad_norm": 0.42756368516441545, + "learning_rate": 0.003, + "loss": 4.3622, + "step": 1236 + }, + { + "epoch": 0.01237, + "grad_norm": 0.41715621646232526, + "learning_rate": 0.003, + "loss": 4.3707, + "step": 1237 + }, + { + "epoch": 0.01238, + "grad_norm": 0.41900290536826335, + "learning_rate": 0.003, + "loss": 4.3664, + "step": 1238 + }, + { + "epoch": 0.01239, + "grad_norm": 0.4057196989115332, + "learning_rate": 0.003, + "loss": 4.3485, + "step": 1239 + }, + { + "epoch": 0.0124, + "grad_norm": 0.41828480066605506, + "learning_rate": 0.003, + "loss": 4.331, + "step": 1240 + }, + { + "epoch": 0.01241, + "grad_norm": 0.4114484855240935, + "learning_rate": 0.003, + "loss": 4.3417, + "step": 1241 + }, + { + "epoch": 0.01242, + "grad_norm": 0.4335820493661237, + "learning_rate": 0.003, + "loss": 4.348, + "step": 1242 + }, + { + "epoch": 0.01243, + "grad_norm": 0.5022671727550078, + "learning_rate": 0.003, + "loss": 4.3316, + "step": 1243 + }, + { + "epoch": 0.01244, + "grad_norm": 0.5275944783156026, + "learning_rate": 0.003, + "loss": 4.3445, + "step": 1244 + }, + { + "epoch": 0.01245, + "grad_norm": 0.5485058077885037, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1245 + }, + { + "epoch": 0.01246, + "grad_norm": 0.5308550603908434, + "learning_rate": 0.003, + "loss": 4.312, + "step": 1246 + }, + { + "epoch": 0.01247, + "grad_norm": 0.5603455205230277, + "learning_rate": 0.003, + "loss": 4.373, + "step": 1247 + }, + { + "epoch": 0.01248, + "grad_norm": 0.6616614283266387, + "learning_rate": 0.003, + "loss": 4.3622, + "step": 1248 + }, + { + "epoch": 0.01249, + "grad_norm": 0.6950989703790047, + "learning_rate": 0.003, + "loss": 4.3579, + "step": 1249 + }, + { + "epoch": 0.0125, + "grad_norm": 0.6807657788060711, + "learning_rate": 0.003, + "loss": 4.3266, + "step": 1250 + }, + { + "epoch": 0.01251, + "grad_norm": 0.7044512298181651, + "learning_rate": 0.003, + "loss": 4.3715, + "step": 1251 + }, + { + "epoch": 0.01252, + "grad_norm": 0.7187283948109732, + "learning_rate": 0.003, + "loss": 4.3491, + "step": 1252 + }, + { + "epoch": 0.01253, + "grad_norm": 0.6168851566630172, + "learning_rate": 0.003, + "loss": 4.3238, + "step": 1253 + }, + { + "epoch": 0.01254, + "grad_norm": 0.5482483685536774, + "learning_rate": 0.003, + "loss": 4.3671, + "step": 1254 + }, + { + "epoch": 0.01255, + "grad_norm": 0.5158716427473324, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1255 + }, + { + "epoch": 0.01256, + "grad_norm": 0.46371154610018284, + "learning_rate": 0.003, + "loss": 4.3421, + "step": 1256 + }, + { + "epoch": 0.01257, + "grad_norm": 0.4261160552146568, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1257 + }, + { + "epoch": 0.01258, + "grad_norm": 0.507565461332445, + "learning_rate": 0.003, + "loss": 4.3375, + "step": 1258 + }, + { + "epoch": 0.01259, + "grad_norm": 0.7023530299208391, + "learning_rate": 0.003, + "loss": 4.3461, + "step": 1259 + }, + { + "epoch": 0.0126, + "grad_norm": 0.8862638702569765, + "learning_rate": 0.003, + "loss": 4.3945, + "step": 1260 + }, + { + "epoch": 0.01261, + "grad_norm": 0.92528404873397, + "learning_rate": 0.003, + "loss": 4.3876, + "step": 1261 + }, + { + "epoch": 0.01262, + "grad_norm": 0.9810348621626278, + "learning_rate": 0.003, + "loss": 4.3924, + "step": 1262 + }, + { + "epoch": 0.01263, + "grad_norm": 0.7811680454044853, + "learning_rate": 0.003, + "loss": 4.3835, + "step": 1263 + }, + { + "epoch": 0.01264, + "grad_norm": 0.6458262464633842, + "learning_rate": 0.003, + "loss": 4.365, + "step": 1264 + }, + { + "epoch": 0.01265, + "grad_norm": 0.6726321620447105, + "learning_rate": 0.003, + "loss": 4.3777, + "step": 1265 + }, + { + "epoch": 0.01266, + "grad_norm": 0.7462208985516781, + "learning_rate": 0.003, + "loss": 4.3808, + "step": 1266 + }, + { + "epoch": 0.01267, + "grad_norm": 0.8334608532491081, + "learning_rate": 0.003, + "loss": 4.4053, + "step": 1267 + }, + { + "epoch": 0.01268, + "grad_norm": 0.722253192550455, + "learning_rate": 0.003, + "loss": 4.3914, + "step": 1268 + }, + { + "epoch": 0.01269, + "grad_norm": 0.5434607483157148, + "learning_rate": 0.003, + "loss": 4.3468, + "step": 1269 + }, + { + "epoch": 0.0127, + "grad_norm": 0.4389640950286333, + "learning_rate": 0.003, + "loss": 4.3757, + "step": 1270 + }, + { + "epoch": 0.01271, + "grad_norm": 0.4493097096002282, + "learning_rate": 0.003, + "loss": 4.357, + "step": 1271 + }, + { + "epoch": 0.01272, + "grad_norm": 0.4602878891635523, + "learning_rate": 0.003, + "loss": 4.3487, + "step": 1272 + }, + { + "epoch": 0.01273, + "grad_norm": 0.4816441128313513, + "learning_rate": 0.003, + "loss": 4.3755, + "step": 1273 + }, + { + "epoch": 0.01274, + "grad_norm": 0.5275730605295648, + "learning_rate": 0.003, + "loss": 4.3705, + "step": 1274 + }, + { + "epoch": 0.01275, + "grad_norm": 0.5285253035410874, + "learning_rate": 0.003, + "loss": 4.3384, + "step": 1275 + }, + { + "epoch": 0.01276, + "grad_norm": 0.42608260843601453, + "learning_rate": 0.003, + "loss": 4.3398, + "step": 1276 + }, + { + "epoch": 0.01277, + "grad_norm": 0.394113210363875, + "learning_rate": 0.003, + "loss": 4.3444, + "step": 1277 + }, + { + "epoch": 0.01278, + "grad_norm": 0.485427057016395, + "learning_rate": 0.003, + "loss": 4.3534, + "step": 1278 + }, + { + "epoch": 0.01279, + "grad_norm": 0.4912898619415999, + "learning_rate": 0.003, + "loss": 4.3362, + "step": 1279 + }, + { + "epoch": 0.0128, + "grad_norm": 0.4970947767829747, + "learning_rate": 0.003, + "loss": 4.3331, + "step": 1280 + }, + { + "epoch": 0.01281, + "grad_norm": 0.45661885236926897, + "learning_rate": 0.003, + "loss": 4.3167, + "step": 1281 + }, + { + "epoch": 0.01282, + "grad_norm": 0.3777414491844073, + "learning_rate": 0.003, + "loss": 4.3562, + "step": 1282 + }, + { + "epoch": 0.01283, + "grad_norm": 0.4194385613191725, + "learning_rate": 0.003, + "loss": 4.3264, + "step": 1283 + }, + { + "epoch": 0.01284, + "grad_norm": 0.455784660804564, + "learning_rate": 0.003, + "loss": 4.328, + "step": 1284 + }, + { + "epoch": 0.01285, + "grad_norm": 0.5164075234519105, + "learning_rate": 0.003, + "loss": 4.3319, + "step": 1285 + }, + { + "epoch": 0.01286, + "grad_norm": 0.6148100521816782, + "learning_rate": 0.003, + "loss": 4.3568, + "step": 1286 + }, + { + "epoch": 0.01287, + "grad_norm": 0.8435250802638579, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1287 + }, + { + "epoch": 0.01288, + "grad_norm": 0.8948772567277533, + "learning_rate": 0.003, + "loss": 4.3616, + "step": 1288 + }, + { + "epoch": 0.01289, + "grad_norm": 0.5768385697936498, + "learning_rate": 0.003, + "loss": 4.3412, + "step": 1289 + }, + { + "epoch": 0.0129, + "grad_norm": 0.48744560346189664, + "learning_rate": 0.003, + "loss": 4.3621, + "step": 1290 + }, + { + "epoch": 0.01291, + "grad_norm": 0.5888239794628021, + "learning_rate": 0.003, + "loss": 4.3458, + "step": 1291 + }, + { + "epoch": 0.01292, + "grad_norm": 0.5667412086864034, + "learning_rate": 0.003, + "loss": 4.3773, + "step": 1292 + }, + { + "epoch": 0.01293, + "grad_norm": 0.6066487822204776, + "learning_rate": 0.003, + "loss": 4.3648, + "step": 1293 + }, + { + "epoch": 0.01294, + "grad_norm": 0.7065627866796411, + "learning_rate": 0.003, + "loss": 4.3371, + "step": 1294 + }, + { + "epoch": 0.01295, + "grad_norm": 0.7389325253179186, + "learning_rate": 0.003, + "loss": 4.3679, + "step": 1295 + }, + { + "epoch": 0.01296, + "grad_norm": 0.7660662423328507, + "learning_rate": 0.003, + "loss": 4.3399, + "step": 1296 + }, + { + "epoch": 0.01297, + "grad_norm": 0.7847737463247922, + "learning_rate": 0.003, + "loss": 4.3573, + "step": 1297 + }, + { + "epoch": 0.01298, + "grad_norm": 0.8150967050528192, + "learning_rate": 0.003, + "loss": 4.3384, + "step": 1298 + }, + { + "epoch": 0.01299, + "grad_norm": 0.7237792681987708, + "learning_rate": 0.003, + "loss": 4.3533, + "step": 1299 + }, + { + "epoch": 0.013, + "grad_norm": 0.6452098996944176, + "learning_rate": 0.003, + "loss": 4.3355, + "step": 1300 + }, + { + "epoch": 0.01301, + "grad_norm": 0.6428524108659182, + "learning_rate": 0.003, + "loss": 4.3495, + "step": 1301 + }, + { + "epoch": 0.01302, + "grad_norm": 0.7475054826389519, + "learning_rate": 0.003, + "loss": 4.4013, + "step": 1302 + }, + { + "epoch": 0.01303, + "grad_norm": 0.9085755519721904, + "learning_rate": 0.003, + "loss": 4.3576, + "step": 1303 + }, + { + "epoch": 0.01304, + "grad_norm": 0.8565196075884696, + "learning_rate": 0.003, + "loss": 4.3934, + "step": 1304 + }, + { + "epoch": 0.01305, + "grad_norm": 0.8916073679840805, + "learning_rate": 0.003, + "loss": 4.3789, + "step": 1305 + }, + { + "epoch": 0.01306, + "grad_norm": 0.7910970693035262, + "learning_rate": 0.003, + "loss": 4.3681, + "step": 1306 + }, + { + "epoch": 0.01307, + "grad_norm": 0.7398667792205602, + "learning_rate": 0.003, + "loss": 4.3875, + "step": 1307 + }, + { + "epoch": 0.01308, + "grad_norm": 0.6800690310649735, + "learning_rate": 0.003, + "loss": 4.3597, + "step": 1308 + }, + { + "epoch": 0.01309, + "grad_norm": 0.610129543183007, + "learning_rate": 0.003, + "loss": 4.3978, + "step": 1309 + }, + { + "epoch": 0.0131, + "grad_norm": 0.6342445693889617, + "learning_rate": 0.003, + "loss": 4.3435, + "step": 1310 + }, + { + "epoch": 0.01311, + "grad_norm": 0.5612187898652411, + "learning_rate": 0.003, + "loss": 4.3787, + "step": 1311 + }, + { + "epoch": 0.01312, + "grad_norm": 0.6173335576080043, + "learning_rate": 0.003, + "loss": 4.3472, + "step": 1312 + }, + { + "epoch": 0.01313, + "grad_norm": 0.5958431337675327, + "learning_rate": 0.003, + "loss": 4.3387, + "step": 1313 + }, + { + "epoch": 0.01314, + "grad_norm": 0.6624135563680444, + "learning_rate": 0.003, + "loss": 4.3367, + "step": 1314 + }, + { + "epoch": 0.01315, + "grad_norm": 0.6490907209314275, + "learning_rate": 0.003, + "loss": 4.368, + "step": 1315 + }, + { + "epoch": 0.01316, + "grad_norm": 0.5500794253188088, + "learning_rate": 0.003, + "loss": 4.3457, + "step": 1316 + }, + { + "epoch": 0.01317, + "grad_norm": 0.6275958010748339, + "learning_rate": 0.003, + "loss": 4.3746, + "step": 1317 + }, + { + "epoch": 0.01318, + "grad_norm": 0.5648473326010665, + "learning_rate": 0.003, + "loss": 4.3513, + "step": 1318 + }, + { + "epoch": 0.01319, + "grad_norm": 0.45510587840319355, + "learning_rate": 0.003, + "loss": 4.3528, + "step": 1319 + }, + { + "epoch": 0.0132, + "grad_norm": 0.44264387716031117, + "learning_rate": 0.003, + "loss": 4.3472, + "step": 1320 + }, + { + "epoch": 0.01321, + "grad_norm": 0.40376640077570203, + "learning_rate": 0.003, + "loss": 4.322, + "step": 1321 + }, + { + "epoch": 0.01322, + "grad_norm": 0.3877836192326668, + "learning_rate": 0.003, + "loss": 4.3502, + "step": 1322 + }, + { + "epoch": 0.01323, + "grad_norm": 0.3937183906523392, + "learning_rate": 0.003, + "loss": 4.341, + "step": 1323 + }, + { + "epoch": 0.01324, + "grad_norm": 0.4100022537274817, + "learning_rate": 0.003, + "loss": 4.3255, + "step": 1324 + }, + { + "epoch": 0.01325, + "grad_norm": 0.46279809510243225, + "learning_rate": 0.003, + "loss": 4.3315, + "step": 1325 + }, + { + "epoch": 0.01326, + "grad_norm": 0.4767280607611473, + "learning_rate": 0.003, + "loss": 4.3173, + "step": 1326 + }, + { + "epoch": 0.01327, + "grad_norm": 0.5123257836661763, + "learning_rate": 0.003, + "loss": 4.3366, + "step": 1327 + }, + { + "epoch": 0.01328, + "grad_norm": 0.5434476080816805, + "learning_rate": 0.003, + "loss": 4.3286, + "step": 1328 + }, + { + "epoch": 0.01329, + "grad_norm": 0.48979784037018337, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1329 + }, + { + "epoch": 0.0133, + "grad_norm": 0.42591921766495494, + "learning_rate": 0.003, + "loss": 4.3315, + "step": 1330 + }, + { + "epoch": 0.01331, + "grad_norm": 0.48229010910172454, + "learning_rate": 0.003, + "loss": 4.318, + "step": 1331 + }, + { + "epoch": 0.01332, + "grad_norm": 0.514388044498456, + "learning_rate": 0.003, + "loss": 4.3386, + "step": 1332 + }, + { + "epoch": 0.01333, + "grad_norm": 0.5443479600217087, + "learning_rate": 0.003, + "loss": 4.3145, + "step": 1333 + }, + { + "epoch": 0.01334, + "grad_norm": 0.5537269490906184, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1334 + }, + { + "epoch": 0.01335, + "grad_norm": 0.5158317522709531, + "learning_rate": 0.003, + "loss": 4.3253, + "step": 1335 + }, + { + "epoch": 0.01336, + "grad_norm": 0.5108167456714484, + "learning_rate": 0.003, + "loss": 4.3223, + "step": 1336 + }, + { + "epoch": 0.01337, + "grad_norm": 0.6228714832065689, + "learning_rate": 0.003, + "loss": 4.3418, + "step": 1337 + }, + { + "epoch": 0.01338, + "grad_norm": 0.7769527340926119, + "learning_rate": 0.003, + "loss": 4.3169, + "step": 1338 + }, + { + "epoch": 0.01339, + "grad_norm": 1.0163715747866278, + "learning_rate": 0.003, + "loss": 4.3527, + "step": 1339 + }, + { + "epoch": 0.0134, + "grad_norm": 0.8875484689453775, + "learning_rate": 0.003, + "loss": 4.3448, + "step": 1340 + }, + { + "epoch": 0.01341, + "grad_norm": 0.7593448848803411, + "learning_rate": 0.003, + "loss": 4.3557, + "step": 1341 + }, + { + "epoch": 0.01342, + "grad_norm": 0.7038826124517171, + "learning_rate": 0.003, + "loss": 4.3272, + "step": 1342 + }, + { + "epoch": 0.01343, + "grad_norm": 0.6570836329978591, + "learning_rate": 0.003, + "loss": 4.3287, + "step": 1343 + }, + { + "epoch": 0.01344, + "grad_norm": 0.7840667137315036, + "learning_rate": 0.003, + "loss": 4.3338, + "step": 1344 + }, + { + "epoch": 0.01345, + "grad_norm": 0.8057220270427362, + "learning_rate": 0.003, + "loss": 4.3808, + "step": 1345 + }, + { + "epoch": 0.01346, + "grad_norm": 0.7768442613998496, + "learning_rate": 0.003, + "loss": 4.3555, + "step": 1346 + }, + { + "epoch": 0.01347, + "grad_norm": 0.7491448291061306, + "learning_rate": 0.003, + "loss": 4.3463, + "step": 1347 + }, + { + "epoch": 0.01348, + "grad_norm": 0.7233252261262297, + "learning_rate": 0.003, + "loss": 4.3476, + "step": 1348 + }, + { + "epoch": 0.01349, + "grad_norm": 0.7147253457500066, + "learning_rate": 0.003, + "loss": 4.3662, + "step": 1349 + }, + { + "epoch": 0.0135, + "grad_norm": 0.8413435804466313, + "learning_rate": 0.003, + "loss": 4.3673, + "step": 1350 + }, + { + "epoch": 0.01351, + "grad_norm": 0.9169233695908002, + "learning_rate": 0.003, + "loss": 4.406, + "step": 1351 + }, + { + "epoch": 0.01352, + "grad_norm": 0.7136998484165762, + "learning_rate": 0.003, + "loss": 4.371, + "step": 1352 + }, + { + "epoch": 0.01353, + "grad_norm": 0.7015996798304995, + "learning_rate": 0.003, + "loss": 4.3575, + "step": 1353 + }, + { + "epoch": 0.01354, + "grad_norm": 0.7349461226651943, + "learning_rate": 0.003, + "loss": 4.3427, + "step": 1354 + }, + { + "epoch": 0.01355, + "grad_norm": 0.7908101359678147, + "learning_rate": 0.003, + "loss": 4.3615, + "step": 1355 + }, + { + "epoch": 0.01356, + "grad_norm": 0.6826667144170498, + "learning_rate": 0.003, + "loss": 4.3532, + "step": 1356 + }, + { + "epoch": 0.01357, + "grad_norm": 0.5539967821197357, + "learning_rate": 0.003, + "loss": 4.3308, + "step": 1357 + }, + { + "epoch": 0.01358, + "grad_norm": 0.6087032954279707, + "learning_rate": 0.003, + "loss": 4.3295, + "step": 1358 + }, + { + "epoch": 0.01359, + "grad_norm": 0.7945356614677406, + "learning_rate": 0.003, + "loss": 4.3753, + "step": 1359 + }, + { + "epoch": 0.0136, + "grad_norm": 0.7371837851861581, + "learning_rate": 0.003, + "loss": 4.3479, + "step": 1360 + }, + { + "epoch": 0.01361, + "grad_norm": 0.5619004170613455, + "learning_rate": 0.003, + "loss": 4.3527, + "step": 1361 + }, + { + "epoch": 0.01362, + "grad_norm": 0.5285167943277829, + "learning_rate": 0.003, + "loss": 4.3621, + "step": 1362 + }, + { + "epoch": 0.01363, + "grad_norm": 0.5076817025344985, + "learning_rate": 0.003, + "loss": 4.3625, + "step": 1363 + }, + { + "epoch": 0.01364, + "grad_norm": 0.45959981115038123, + "learning_rate": 0.003, + "loss": 4.3487, + "step": 1364 + }, + { + "epoch": 0.01365, + "grad_norm": 0.4790054557157507, + "learning_rate": 0.003, + "loss": 4.3663, + "step": 1365 + }, + { + "epoch": 0.01366, + "grad_norm": 0.47982525149920063, + "learning_rate": 0.003, + "loss": 4.3186, + "step": 1366 + }, + { + "epoch": 0.01367, + "grad_norm": 0.4158841510813793, + "learning_rate": 0.003, + "loss": 4.3634, + "step": 1367 + }, + { + "epoch": 0.01368, + "grad_norm": 0.40516006301845847, + "learning_rate": 0.003, + "loss": 4.3337, + "step": 1368 + }, + { + "epoch": 0.01369, + "grad_norm": 0.4130828938355455, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1369 + }, + { + "epoch": 0.0137, + "grad_norm": 0.4124457908402306, + "learning_rate": 0.003, + "loss": 4.3486, + "step": 1370 + }, + { + "epoch": 0.01371, + "grad_norm": 0.3816075969714427, + "learning_rate": 0.003, + "loss": 4.3376, + "step": 1371 + }, + { + "epoch": 0.01372, + "grad_norm": 0.360968745189086, + "learning_rate": 0.003, + "loss": 4.2843, + "step": 1372 + }, + { + "epoch": 0.01373, + "grad_norm": 0.3546042076579899, + "learning_rate": 0.003, + "loss": 4.3232, + "step": 1373 + }, + { + "epoch": 0.01374, + "grad_norm": 0.35618421664945743, + "learning_rate": 0.003, + "loss": 4.3287, + "step": 1374 + }, + { + "epoch": 0.01375, + "grad_norm": 0.3353008720630168, + "learning_rate": 0.003, + "loss": 4.2881, + "step": 1375 + }, + { + "epoch": 0.01376, + "grad_norm": 0.3199771170841034, + "learning_rate": 0.003, + "loss": 4.3127, + "step": 1376 + }, + { + "epoch": 0.01377, + "grad_norm": 0.3779659083446569, + "learning_rate": 0.003, + "loss": 4.3119, + "step": 1377 + }, + { + "epoch": 0.01378, + "grad_norm": 0.5745139158210405, + "learning_rate": 0.003, + "loss": 4.3459, + "step": 1378 + }, + { + "epoch": 0.01379, + "grad_norm": 0.8747330770544977, + "learning_rate": 0.003, + "loss": 4.3327, + "step": 1379 + }, + { + "epoch": 0.0138, + "grad_norm": 1.1013712635151454, + "learning_rate": 0.003, + "loss": 4.3554, + "step": 1380 + }, + { + "epoch": 0.01381, + "grad_norm": 0.6702950145107445, + "learning_rate": 0.003, + "loss": 4.3457, + "step": 1381 + }, + { + "epoch": 0.01382, + "grad_norm": 0.674972596014113, + "learning_rate": 0.003, + "loss": 4.3437, + "step": 1382 + }, + { + "epoch": 0.01383, + "grad_norm": 0.6997537624642522, + "learning_rate": 0.003, + "loss": 4.3369, + "step": 1383 + }, + { + "epoch": 0.01384, + "grad_norm": 0.6469176723359975, + "learning_rate": 0.003, + "loss": 4.3638, + "step": 1384 + }, + { + "epoch": 0.01385, + "grad_norm": 0.5530980119483301, + "learning_rate": 0.003, + "loss": 4.3456, + "step": 1385 + }, + { + "epoch": 0.01386, + "grad_norm": 0.6135404549370512, + "learning_rate": 0.003, + "loss": 4.3261, + "step": 1386 + }, + { + "epoch": 0.01387, + "grad_norm": 0.552332815368935, + "learning_rate": 0.003, + "loss": 4.3116, + "step": 1387 + }, + { + "epoch": 0.01388, + "grad_norm": 0.49081872927843356, + "learning_rate": 0.003, + "loss": 4.3469, + "step": 1388 + }, + { + "epoch": 0.01389, + "grad_norm": 0.495957727461612, + "learning_rate": 0.003, + "loss": 4.3269, + "step": 1389 + }, + { + "epoch": 0.0139, + "grad_norm": 0.4896458918623629, + "learning_rate": 0.003, + "loss": 4.3267, + "step": 1390 + }, + { + "epoch": 0.01391, + "grad_norm": 0.49070092673030247, + "learning_rate": 0.003, + "loss": 4.3285, + "step": 1391 + }, + { + "epoch": 0.01392, + "grad_norm": 0.475140973661053, + "learning_rate": 0.003, + "loss": 4.3381, + "step": 1392 + }, + { + "epoch": 0.01393, + "grad_norm": 0.45720050629487835, + "learning_rate": 0.003, + "loss": 4.3042, + "step": 1393 + }, + { + "epoch": 0.01394, + "grad_norm": 0.5119120302085031, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1394 + }, + { + "epoch": 0.01395, + "grad_norm": 0.6190916799083034, + "learning_rate": 0.003, + "loss": 4.3279, + "step": 1395 + }, + { + "epoch": 0.01396, + "grad_norm": 0.6508817551577758, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1396 + }, + { + "epoch": 0.01397, + "grad_norm": 0.6075070397922983, + "learning_rate": 0.003, + "loss": 4.3359, + "step": 1397 + }, + { + "epoch": 0.01398, + "grad_norm": 0.6196238264215417, + "learning_rate": 0.003, + "loss": 4.3165, + "step": 1398 + }, + { + "epoch": 0.01399, + "grad_norm": 0.6249218455396569, + "learning_rate": 0.003, + "loss": 4.3396, + "step": 1399 + }, + { + "epoch": 0.014, + "grad_norm": 0.7148437815408427, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1400 + }, + { + "epoch": 0.01401, + "grad_norm": 0.6723818092453638, + "learning_rate": 0.003, + "loss": 4.3228, + "step": 1401 + }, + { + "epoch": 0.01402, + "grad_norm": 0.7455518751558462, + "learning_rate": 0.003, + "loss": 4.3392, + "step": 1402 + }, + { + "epoch": 0.01403, + "grad_norm": 0.6611806665670097, + "learning_rate": 0.003, + "loss": 4.3114, + "step": 1403 + }, + { + "epoch": 0.01404, + "grad_norm": 0.6150645931793568, + "learning_rate": 0.003, + "loss": 4.3339, + "step": 1404 + }, + { + "epoch": 0.01405, + "grad_norm": 0.591029520159785, + "learning_rate": 0.003, + "loss": 4.3493, + "step": 1405 + }, + { + "epoch": 0.01406, + "grad_norm": 0.6587405723247819, + "learning_rate": 0.003, + "loss": 4.3328, + "step": 1406 + }, + { + "epoch": 0.01407, + "grad_norm": 0.7112635318635802, + "learning_rate": 0.003, + "loss": 4.3508, + "step": 1407 + }, + { + "epoch": 0.01408, + "grad_norm": 0.7760050777299251, + "learning_rate": 0.003, + "loss": 4.3163, + "step": 1408 + }, + { + "epoch": 0.01409, + "grad_norm": 0.7313606609426694, + "learning_rate": 0.003, + "loss": 4.3489, + "step": 1409 + }, + { + "epoch": 0.0141, + "grad_norm": 0.7213404812022338, + "learning_rate": 0.003, + "loss": 4.326, + "step": 1410 + }, + { + "epoch": 0.01411, + "grad_norm": 0.6616036396606474, + "learning_rate": 0.003, + "loss": 4.3016, + "step": 1411 + }, + { + "epoch": 0.01412, + "grad_norm": 0.6008069011437791, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1412 + }, + { + "epoch": 0.01413, + "grad_norm": 0.7020614194036953, + "learning_rate": 0.003, + "loss": 4.3356, + "step": 1413 + }, + { + "epoch": 0.01414, + "grad_norm": 0.6966655821964342, + "learning_rate": 0.003, + "loss": 4.322, + "step": 1414 + }, + { + "epoch": 0.01415, + "grad_norm": 0.7878825992848615, + "learning_rate": 0.003, + "loss": 4.3182, + "step": 1415 + }, + { + "epoch": 0.01416, + "grad_norm": 0.9133447782042373, + "learning_rate": 0.003, + "loss": 4.3606, + "step": 1416 + }, + { + "epoch": 0.01417, + "grad_norm": 1.0833856715197905, + "learning_rate": 0.003, + "loss": 4.3655, + "step": 1417 + }, + { + "epoch": 0.01418, + "grad_norm": 0.8776183907401212, + "learning_rate": 0.003, + "loss": 4.3779, + "step": 1418 + }, + { + "epoch": 0.01419, + "grad_norm": 0.8716355229415133, + "learning_rate": 0.003, + "loss": 4.3745, + "step": 1419 + }, + { + "epoch": 0.0142, + "grad_norm": 0.6900477550364122, + "learning_rate": 0.003, + "loss": 4.3678, + "step": 1420 + }, + { + "epoch": 0.01421, + "grad_norm": 0.6518139426774641, + "learning_rate": 0.003, + "loss": 4.3347, + "step": 1421 + }, + { + "epoch": 0.01422, + "grad_norm": 0.6549157122857109, + "learning_rate": 0.003, + "loss": 4.3382, + "step": 1422 + }, + { + "epoch": 0.01423, + "grad_norm": 0.6092791265519953, + "learning_rate": 0.003, + "loss": 4.3402, + "step": 1423 + }, + { + "epoch": 0.01424, + "grad_norm": 0.5943123649374407, + "learning_rate": 0.003, + "loss": 4.3537, + "step": 1424 + }, + { + "epoch": 0.01425, + "grad_norm": 0.5086164834561862, + "learning_rate": 0.003, + "loss": 4.3528, + "step": 1425 + }, + { + "epoch": 0.01426, + "grad_norm": 0.45109246822140664, + "learning_rate": 0.003, + "loss": 4.3278, + "step": 1426 + }, + { + "epoch": 0.01427, + "grad_norm": 0.43801724179324136, + "learning_rate": 0.003, + "loss": 4.3209, + "step": 1427 + }, + { + "epoch": 0.01428, + "grad_norm": 0.5018488931475374, + "learning_rate": 0.003, + "loss": 4.3159, + "step": 1428 + }, + { + "epoch": 0.01429, + "grad_norm": 0.47066094753828186, + "learning_rate": 0.003, + "loss": 4.3234, + "step": 1429 + }, + { + "epoch": 0.0143, + "grad_norm": 0.4403238802934058, + "learning_rate": 0.003, + "loss": 4.3319, + "step": 1430 + }, + { + "epoch": 0.01431, + "grad_norm": 0.44039117208650785, + "learning_rate": 0.003, + "loss": 4.3183, + "step": 1431 + }, + { + "epoch": 0.01432, + "grad_norm": 0.477019489570791, + "learning_rate": 0.003, + "loss": 4.3074, + "step": 1432 + }, + { + "epoch": 0.01433, + "grad_norm": 0.4264410276196829, + "learning_rate": 0.003, + "loss": 4.3017, + "step": 1433 + }, + { + "epoch": 0.01434, + "grad_norm": 0.3944694636285646, + "learning_rate": 0.003, + "loss": 4.2866, + "step": 1434 + }, + { + "epoch": 0.01435, + "grad_norm": 0.3924802801708403, + "learning_rate": 0.003, + "loss": 4.2942, + "step": 1435 + }, + { + "epoch": 0.01436, + "grad_norm": 0.4172733802501552, + "learning_rate": 0.003, + "loss": 4.3128, + "step": 1436 + }, + { + "epoch": 0.01437, + "grad_norm": 0.3851734831243834, + "learning_rate": 0.003, + "loss": 4.3237, + "step": 1437 + }, + { + "epoch": 0.01438, + "grad_norm": 0.35404107026104975, + "learning_rate": 0.003, + "loss": 4.3163, + "step": 1438 + }, + { + "epoch": 0.01439, + "grad_norm": 0.3935862503777362, + "learning_rate": 0.003, + "loss": 4.2798, + "step": 1439 + }, + { + "epoch": 0.0144, + "grad_norm": 0.48537387134962556, + "learning_rate": 0.003, + "loss": 4.2941, + "step": 1440 + }, + { + "epoch": 0.01441, + "grad_norm": 0.6205869970599557, + "learning_rate": 0.003, + "loss": 4.3081, + "step": 1441 + }, + { + "epoch": 0.01442, + "grad_norm": 0.733625820231546, + "learning_rate": 0.003, + "loss": 4.2954, + "step": 1442 + }, + { + "epoch": 0.01443, + "grad_norm": 0.6539626450528481, + "learning_rate": 0.003, + "loss": 4.3597, + "step": 1443 + }, + { + "epoch": 0.01444, + "grad_norm": 0.4715932337839517, + "learning_rate": 0.003, + "loss": 4.3144, + "step": 1444 + }, + { + "epoch": 0.01445, + "grad_norm": 0.5734177290330968, + "learning_rate": 0.003, + "loss": 4.2799, + "step": 1445 + }, + { + "epoch": 0.01446, + "grad_norm": 0.6345599077429956, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1446 + }, + { + "epoch": 0.01447, + "grad_norm": 0.7159211854064349, + "learning_rate": 0.003, + "loss": 4.3124, + "step": 1447 + }, + { + "epoch": 0.01448, + "grad_norm": 0.847805509409058, + "learning_rate": 0.003, + "loss": 4.3346, + "step": 1448 + }, + { + "epoch": 0.01449, + "grad_norm": 0.8334367502615841, + "learning_rate": 0.003, + "loss": 4.3316, + "step": 1449 + }, + { + "epoch": 0.0145, + "grad_norm": 0.8387485961034247, + "learning_rate": 0.003, + "loss": 4.3521, + "step": 1450 + }, + { + "epoch": 0.01451, + "grad_norm": 0.7069735242725316, + "learning_rate": 0.003, + "loss": 4.3254, + "step": 1451 + }, + { + "epoch": 0.01452, + "grad_norm": 0.7128306367496979, + "learning_rate": 0.003, + "loss": 4.3315, + "step": 1452 + }, + { + "epoch": 0.01453, + "grad_norm": 0.7607386324594154, + "learning_rate": 0.003, + "loss": 4.3465, + "step": 1453 + }, + { + "epoch": 0.01454, + "grad_norm": 0.8531443855411419, + "learning_rate": 0.003, + "loss": 4.3152, + "step": 1454 + }, + { + "epoch": 0.01455, + "grad_norm": 0.9072653937920705, + "learning_rate": 0.003, + "loss": 4.3602, + "step": 1455 + }, + { + "epoch": 0.01456, + "grad_norm": 0.9678032867174141, + "learning_rate": 0.003, + "loss": 4.3448, + "step": 1456 + }, + { + "epoch": 0.01457, + "grad_norm": 0.8263837469314969, + "learning_rate": 0.003, + "loss": 4.3358, + "step": 1457 + }, + { + "epoch": 0.01458, + "grad_norm": 0.7586740850669937, + "learning_rate": 0.003, + "loss": 4.3274, + "step": 1458 + }, + { + "epoch": 0.01459, + "grad_norm": 0.7104303096772606, + "learning_rate": 0.003, + "loss": 4.3506, + "step": 1459 + }, + { + "epoch": 0.0146, + "grad_norm": 0.7105635489785329, + "learning_rate": 0.003, + "loss": 4.358, + "step": 1460 + }, + { + "epoch": 0.01461, + "grad_norm": 0.6051123991598398, + "learning_rate": 0.003, + "loss": 4.3255, + "step": 1461 + }, + { + "epoch": 0.01462, + "grad_norm": 0.49818465497872577, + "learning_rate": 0.003, + "loss": 4.3277, + "step": 1462 + }, + { + "epoch": 0.01463, + "grad_norm": 0.4705387398623177, + "learning_rate": 0.003, + "loss": 4.3234, + "step": 1463 + }, + { + "epoch": 0.01464, + "grad_norm": 0.43420725887330847, + "learning_rate": 0.003, + "loss": 4.2947, + "step": 1464 + }, + { + "epoch": 0.01465, + "grad_norm": 0.416394344592158, + "learning_rate": 0.003, + "loss": 4.3376, + "step": 1465 + }, + { + "epoch": 0.01466, + "grad_norm": 0.46274333085308, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1466 + }, + { + "epoch": 0.01467, + "grad_norm": 0.412473133729069, + "learning_rate": 0.003, + "loss": 4.339, + "step": 1467 + }, + { + "epoch": 0.01468, + "grad_norm": 0.4311218055740771, + "learning_rate": 0.003, + "loss": 4.3146, + "step": 1468 + }, + { + "epoch": 0.01469, + "grad_norm": 0.4074958191436831, + "learning_rate": 0.003, + "loss": 4.312, + "step": 1469 + }, + { + "epoch": 0.0147, + "grad_norm": 0.43399248954304653, + "learning_rate": 0.003, + "loss": 4.3033, + "step": 1470 + }, + { + "epoch": 0.01471, + "grad_norm": 0.41361534167776814, + "learning_rate": 0.003, + "loss": 4.3141, + "step": 1471 + }, + { + "epoch": 0.01472, + "grad_norm": 0.4291933422709367, + "learning_rate": 0.003, + "loss": 4.2989, + "step": 1472 + }, + { + "epoch": 0.01473, + "grad_norm": 0.4534239206252403, + "learning_rate": 0.003, + "loss": 4.3122, + "step": 1473 + }, + { + "epoch": 0.01474, + "grad_norm": 0.5901119500480289, + "learning_rate": 0.003, + "loss": 4.3004, + "step": 1474 + }, + { + "epoch": 0.01475, + "grad_norm": 0.893255157155039, + "learning_rate": 0.003, + "loss": 4.3326, + "step": 1475 + }, + { + "epoch": 0.01476, + "grad_norm": 1.4183061783370785, + "learning_rate": 0.003, + "loss": 4.3574, + "step": 1476 + }, + { + "epoch": 0.01477, + "grad_norm": 0.6731343966041683, + "learning_rate": 0.003, + "loss": 4.3345, + "step": 1477 + }, + { + "epoch": 0.01478, + "grad_norm": 0.6759295392700941, + "learning_rate": 0.003, + "loss": 4.3246, + "step": 1478 + }, + { + "epoch": 0.01479, + "grad_norm": 0.5808036667947172, + "learning_rate": 0.003, + "loss": 4.3486, + "step": 1479 + }, + { + "epoch": 0.0148, + "grad_norm": 0.5015217636867718, + "learning_rate": 0.003, + "loss": 4.3142, + "step": 1480 + }, + { + "epoch": 0.01481, + "grad_norm": 0.4567088802488701, + "learning_rate": 0.003, + "loss": 4.3125, + "step": 1481 + }, + { + "epoch": 0.01482, + "grad_norm": 0.4183399036149268, + "learning_rate": 0.003, + "loss": 4.2956, + "step": 1482 + }, + { + "epoch": 0.01483, + "grad_norm": 0.4450691330955631, + "learning_rate": 0.003, + "loss": 4.2937, + "step": 1483 + }, + { + "epoch": 0.01484, + "grad_norm": 0.3765600896203587, + "learning_rate": 0.003, + "loss": 4.3398, + "step": 1484 + }, + { + "epoch": 0.01485, + "grad_norm": 0.3992681409647294, + "learning_rate": 0.003, + "loss": 4.3029, + "step": 1485 + }, + { + "epoch": 0.01486, + "grad_norm": 0.3969999829277036, + "learning_rate": 0.003, + "loss": 4.2984, + "step": 1486 + }, + { + "epoch": 0.01487, + "grad_norm": 0.3308465500623018, + "learning_rate": 0.003, + "loss": 4.2984, + "step": 1487 + }, + { + "epoch": 0.01488, + "grad_norm": 0.2814090123328206, + "learning_rate": 0.003, + "loss": 4.3052, + "step": 1488 + }, + { + "epoch": 0.01489, + "grad_norm": 0.30005415710473554, + "learning_rate": 0.003, + "loss": 4.2986, + "step": 1489 + }, + { + "epoch": 0.0149, + "grad_norm": 0.32421623348009687, + "learning_rate": 0.003, + "loss": 4.2912, + "step": 1490 + }, + { + "epoch": 0.01491, + "grad_norm": 0.35115426496092417, + "learning_rate": 0.003, + "loss": 4.3043, + "step": 1491 + }, + { + "epoch": 0.01492, + "grad_norm": 0.37018314484045245, + "learning_rate": 0.003, + "loss": 4.2686, + "step": 1492 + }, + { + "epoch": 0.01493, + "grad_norm": 0.4075858968364619, + "learning_rate": 0.003, + "loss": 4.3015, + "step": 1493 + }, + { + "epoch": 0.01494, + "grad_norm": 0.4912830051602376, + "learning_rate": 0.003, + "loss": 4.2818, + "step": 1494 + }, + { + "epoch": 0.01495, + "grad_norm": 0.5565886808339906, + "learning_rate": 0.003, + "loss": 4.2955, + "step": 1495 + }, + { + "epoch": 0.01496, + "grad_norm": 0.6104515927182645, + "learning_rate": 0.003, + "loss": 4.2731, + "step": 1496 + }, + { + "epoch": 0.01497, + "grad_norm": 0.6049970308529968, + "learning_rate": 0.003, + "loss": 4.2492, + "step": 1497 + }, + { + "epoch": 0.01498, + "grad_norm": 0.5392848754302815, + "learning_rate": 0.003, + "loss": 4.304, + "step": 1498 + }, + { + "epoch": 0.01499, + "grad_norm": 0.544325685587062, + "learning_rate": 0.003, + "loss": 4.3252, + "step": 1499 + }, + { + "epoch": 0.015, + "grad_norm": 0.5744687962127912, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1500 + }, + { + "epoch": 0.01501, + "grad_norm": 0.6650575718497268, + "learning_rate": 0.003, + "loss": 4.2836, + "step": 1501 + }, + { + "epoch": 0.01502, + "grad_norm": 0.762494815469482, + "learning_rate": 0.003, + "loss": 4.3344, + "step": 1502 + }, + { + "epoch": 0.01503, + "grad_norm": 0.8042887596944258, + "learning_rate": 0.003, + "loss": 4.3203, + "step": 1503 + }, + { + "epoch": 0.01504, + "grad_norm": 0.8888233604685658, + "learning_rate": 0.003, + "loss": 4.3247, + "step": 1504 + }, + { + "epoch": 0.01505, + "grad_norm": 0.9445171252403785, + "learning_rate": 0.003, + "loss": 4.3231, + "step": 1505 + }, + { + "epoch": 0.01506, + "grad_norm": 1.0444284185336012, + "learning_rate": 0.003, + "loss": 4.3641, + "step": 1506 + }, + { + "epoch": 0.01507, + "grad_norm": 1.1261023785455877, + "learning_rate": 0.003, + "loss": 4.3931, + "step": 1507 + }, + { + "epoch": 0.01508, + "grad_norm": 1.234913329930944, + "learning_rate": 0.003, + "loss": 4.3763, + "step": 1508 + }, + { + "epoch": 0.01509, + "grad_norm": 0.8497166266198067, + "learning_rate": 0.003, + "loss": 4.3805, + "step": 1509 + }, + { + "epoch": 0.0151, + "grad_norm": 0.8376711335758371, + "learning_rate": 0.003, + "loss": 4.3542, + "step": 1510 + }, + { + "epoch": 0.01511, + "grad_norm": 0.724530376603094, + "learning_rate": 0.003, + "loss": 4.3387, + "step": 1511 + }, + { + "epoch": 0.01512, + "grad_norm": 0.6831424127546084, + "learning_rate": 0.003, + "loss": 4.3429, + "step": 1512 + }, + { + "epoch": 0.01513, + "grad_norm": 0.5603527644059662, + "learning_rate": 0.003, + "loss": 4.3405, + "step": 1513 + }, + { + "epoch": 0.01514, + "grad_norm": 0.5930589812891277, + "learning_rate": 0.003, + "loss": 4.3465, + "step": 1514 + }, + { + "epoch": 0.01515, + "grad_norm": 0.7232891849880219, + "learning_rate": 0.003, + "loss": 4.339, + "step": 1515 + }, + { + "epoch": 0.01516, + "grad_norm": 0.7759860995114962, + "learning_rate": 0.003, + "loss": 4.3508, + "step": 1516 + }, + { + "epoch": 0.01517, + "grad_norm": 0.6404044980036384, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1517 + }, + { + "epoch": 0.01518, + "grad_norm": 0.640364805854496, + "learning_rate": 0.003, + "loss": 4.3263, + "step": 1518 + }, + { + "epoch": 0.01519, + "grad_norm": 0.7392318945229942, + "learning_rate": 0.003, + "loss": 4.3661, + "step": 1519 + }, + { + "epoch": 0.0152, + "grad_norm": 0.9450167317883038, + "learning_rate": 0.003, + "loss": 4.3284, + "step": 1520 + }, + { + "epoch": 0.01521, + "grad_norm": 0.9107576371118652, + "learning_rate": 0.003, + "loss": 4.3495, + "step": 1521 + }, + { + "epoch": 0.01522, + "grad_norm": 0.7612821914439016, + "learning_rate": 0.003, + "loss": 4.3394, + "step": 1522 + }, + { + "epoch": 0.01523, + "grad_norm": 0.6613113244902528, + "learning_rate": 0.003, + "loss": 4.3287, + "step": 1523 + }, + { + "epoch": 0.01524, + "grad_norm": 0.5975485724244859, + "learning_rate": 0.003, + "loss": 4.3558, + "step": 1524 + }, + { + "epoch": 0.01525, + "grad_norm": 0.49953522198986544, + "learning_rate": 0.003, + "loss": 4.349, + "step": 1525 + }, + { + "epoch": 0.01526, + "grad_norm": 0.5020786027972391, + "learning_rate": 0.003, + "loss": 4.339, + "step": 1526 + }, + { + "epoch": 0.01527, + "grad_norm": 0.4272044641011338, + "learning_rate": 0.003, + "loss": 4.301, + "step": 1527 + }, + { + "epoch": 0.01528, + "grad_norm": 0.38333031076180046, + "learning_rate": 0.003, + "loss": 4.3172, + "step": 1528 + }, + { + "epoch": 0.01529, + "grad_norm": 0.37710300633649013, + "learning_rate": 0.003, + "loss": 4.3078, + "step": 1529 + }, + { + "epoch": 0.0153, + "grad_norm": 0.3539206796857484, + "learning_rate": 0.003, + "loss": 4.3136, + "step": 1530 + }, + { + "epoch": 0.01531, + "grad_norm": 0.31579499408524375, + "learning_rate": 0.003, + "loss": 4.3044, + "step": 1531 + }, + { + "epoch": 0.01532, + "grad_norm": 0.28740282558130736, + "learning_rate": 0.003, + "loss": 4.2793, + "step": 1532 + }, + { + "epoch": 0.01533, + "grad_norm": 0.32298125914434667, + "learning_rate": 0.003, + "loss": 4.3053, + "step": 1533 + }, + { + "epoch": 0.01534, + "grad_norm": 0.34152544816126584, + "learning_rate": 0.003, + "loss": 4.302, + "step": 1534 + }, + { + "epoch": 0.01535, + "grad_norm": 0.40345568055827374, + "learning_rate": 0.003, + "loss": 4.3133, + "step": 1535 + }, + { + "epoch": 0.01536, + "grad_norm": 0.47716276358652043, + "learning_rate": 0.003, + "loss": 4.2948, + "step": 1536 + }, + { + "epoch": 0.01537, + "grad_norm": 0.5673450395929736, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1537 + }, + { + "epoch": 0.01538, + "grad_norm": 0.5662986079723249, + "learning_rate": 0.003, + "loss": 4.2688, + "step": 1538 + }, + { + "epoch": 0.01539, + "grad_norm": 0.5250656872774493, + "learning_rate": 0.003, + "loss": 4.2993, + "step": 1539 + }, + { + "epoch": 0.0154, + "grad_norm": 0.4759070450942551, + "learning_rate": 0.003, + "loss": 4.3011, + "step": 1540 + }, + { + "epoch": 0.01541, + "grad_norm": 0.525208601732405, + "learning_rate": 0.003, + "loss": 4.2841, + "step": 1541 + }, + { + "epoch": 0.01542, + "grad_norm": 0.6195424842603703, + "learning_rate": 0.003, + "loss": 4.3158, + "step": 1542 + }, + { + "epoch": 0.01543, + "grad_norm": 0.7066044834019704, + "learning_rate": 0.003, + "loss": 4.3015, + "step": 1543 + }, + { + "epoch": 0.01544, + "grad_norm": 0.6094492242185264, + "learning_rate": 0.003, + "loss": 4.2886, + "step": 1544 + }, + { + "epoch": 0.01545, + "grad_norm": 0.6542596992859665, + "learning_rate": 0.003, + "loss": 4.3554, + "step": 1545 + }, + { + "epoch": 0.01546, + "grad_norm": 0.7470834345800471, + "learning_rate": 0.003, + "loss": 4.3329, + "step": 1546 + }, + { + "epoch": 0.01547, + "grad_norm": 0.7200211733418024, + "learning_rate": 0.003, + "loss": 4.3307, + "step": 1547 + }, + { + "epoch": 0.01548, + "grad_norm": 0.6295943931814879, + "learning_rate": 0.003, + "loss": 4.2934, + "step": 1548 + }, + { + "epoch": 0.01549, + "grad_norm": 0.54945766010764, + "learning_rate": 0.003, + "loss": 4.2766, + "step": 1549 + }, + { + "epoch": 0.0155, + "grad_norm": 0.5950909013773433, + "learning_rate": 0.003, + "loss": 4.297, + "step": 1550 + }, + { + "epoch": 0.01551, + "grad_norm": 0.6224936273193388, + "learning_rate": 0.003, + "loss": 4.2776, + "step": 1551 + }, + { + "epoch": 0.01552, + "grad_norm": 0.6624794177801069, + "learning_rate": 0.003, + "loss": 4.2999, + "step": 1552 + }, + { + "epoch": 0.01553, + "grad_norm": 0.7863145855853837, + "learning_rate": 0.003, + "loss": 4.335, + "step": 1553 + }, + { + "epoch": 0.01554, + "grad_norm": 0.7964168301162362, + "learning_rate": 0.003, + "loss": 4.3213, + "step": 1554 + }, + { + "epoch": 0.01555, + "grad_norm": 0.6686376199306979, + "learning_rate": 0.003, + "loss": 4.292, + "step": 1555 + }, + { + "epoch": 0.01556, + "grad_norm": 0.6587445991969533, + "learning_rate": 0.003, + "loss": 4.327, + "step": 1556 + }, + { + "epoch": 0.01557, + "grad_norm": 0.8168668532513341, + "learning_rate": 0.003, + "loss": 4.3101, + "step": 1557 + }, + { + "epoch": 0.01558, + "grad_norm": 0.6341929215834538, + "learning_rate": 0.003, + "loss": 4.2928, + "step": 1558 + }, + { + "epoch": 0.01559, + "grad_norm": 0.5639732491475488, + "learning_rate": 0.003, + "loss": 4.3265, + "step": 1559 + }, + { + "epoch": 0.0156, + "grad_norm": 0.6344606998016725, + "learning_rate": 0.003, + "loss": 4.3189, + "step": 1560 + }, + { + "epoch": 0.01561, + "grad_norm": 0.6020928854417142, + "learning_rate": 0.003, + "loss": 4.301, + "step": 1561 + }, + { + "epoch": 0.01562, + "grad_norm": 0.6447473041378132, + "learning_rate": 0.003, + "loss": 4.2957, + "step": 1562 + }, + { + "epoch": 0.01563, + "grad_norm": 0.637527615370107, + "learning_rate": 0.003, + "loss": 4.3012, + "step": 1563 + }, + { + "epoch": 0.01564, + "grad_norm": 0.6065116381126623, + "learning_rate": 0.003, + "loss": 4.3275, + "step": 1564 + }, + { + "epoch": 0.01565, + "grad_norm": 0.5530540273129289, + "learning_rate": 0.003, + "loss": 4.3008, + "step": 1565 + }, + { + "epoch": 0.01566, + "grad_norm": 0.523767221886022, + "learning_rate": 0.003, + "loss": 4.3185, + "step": 1566 + }, + { + "epoch": 0.01567, + "grad_norm": 0.44666292802850066, + "learning_rate": 0.003, + "loss": 4.3073, + "step": 1567 + }, + { + "epoch": 0.01568, + "grad_norm": 0.4825649782380437, + "learning_rate": 0.003, + "loss": 4.2912, + "step": 1568 + }, + { + "epoch": 0.01569, + "grad_norm": 0.4768074953920202, + "learning_rate": 0.003, + "loss": 4.2946, + "step": 1569 + }, + { + "epoch": 0.0157, + "grad_norm": 0.4756489341732161, + "learning_rate": 0.003, + "loss": 4.3112, + "step": 1570 + }, + { + "epoch": 0.01571, + "grad_norm": 0.5079994555205132, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1571 + }, + { + "epoch": 0.01572, + "grad_norm": 0.5289839099051613, + "learning_rate": 0.003, + "loss": 4.3054, + "step": 1572 + }, + { + "epoch": 0.01573, + "grad_norm": 0.5222081857718813, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1573 + }, + { + "epoch": 0.01574, + "grad_norm": 0.6451289760332761, + "learning_rate": 0.003, + "loss": 4.3076, + "step": 1574 + }, + { + "epoch": 0.01575, + "grad_norm": 0.7877974193239016, + "learning_rate": 0.003, + "loss": 4.2813, + "step": 1575 + }, + { + "epoch": 0.01576, + "grad_norm": 0.8607804040917739, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1576 + }, + { + "epoch": 0.01577, + "grad_norm": 0.8142185085462708, + "learning_rate": 0.003, + "loss": 4.2736, + "step": 1577 + }, + { + "epoch": 0.01578, + "grad_norm": 0.7720878068888671, + "learning_rate": 0.003, + "loss": 4.3026, + "step": 1578 + }, + { + "epoch": 0.01579, + "grad_norm": 0.7192241887142894, + "learning_rate": 0.003, + "loss": 4.2996, + "step": 1579 + }, + { + "epoch": 0.0158, + "grad_norm": 0.606159564012465, + "learning_rate": 0.003, + "loss": 4.3013, + "step": 1580 + }, + { + "epoch": 0.01581, + "grad_norm": 0.4880751313224698, + "learning_rate": 0.003, + "loss": 4.3071, + "step": 1581 + }, + { + "epoch": 0.01582, + "grad_norm": 0.5304647614551631, + "learning_rate": 0.003, + "loss": 4.2891, + "step": 1582 + }, + { + "epoch": 0.01583, + "grad_norm": 0.46382514166554145, + "learning_rate": 0.003, + "loss": 4.2732, + "step": 1583 + }, + { + "epoch": 0.01584, + "grad_norm": 0.47578841605412814, + "learning_rate": 0.003, + "loss": 4.3048, + "step": 1584 + }, + { + "epoch": 0.01585, + "grad_norm": 0.4985689869447637, + "learning_rate": 0.003, + "loss": 4.2972, + "step": 1585 + }, + { + "epoch": 0.01586, + "grad_norm": 0.4793261849715971, + "learning_rate": 0.003, + "loss": 4.2718, + "step": 1586 + }, + { + "epoch": 0.01587, + "grad_norm": 0.4867805801070396, + "learning_rate": 0.003, + "loss": 4.288, + "step": 1587 + }, + { + "epoch": 0.01588, + "grad_norm": 0.49512882436423494, + "learning_rate": 0.003, + "loss": 4.3139, + "step": 1588 + }, + { + "epoch": 0.01589, + "grad_norm": 0.5289005946834403, + "learning_rate": 0.003, + "loss": 4.306, + "step": 1589 + }, + { + "epoch": 0.0159, + "grad_norm": 0.6096773535511837, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1590 + }, + { + "epoch": 0.01591, + "grad_norm": 0.6229667469755624, + "learning_rate": 0.003, + "loss": 4.3102, + "step": 1591 + }, + { + "epoch": 0.01592, + "grad_norm": 0.4936084502041978, + "learning_rate": 0.003, + "loss": 4.2967, + "step": 1592 + }, + { + "epoch": 0.01593, + "grad_norm": 0.42446523481816184, + "learning_rate": 0.003, + "loss": 4.2946, + "step": 1593 + }, + { + "epoch": 0.01594, + "grad_norm": 0.40613386955274233, + "learning_rate": 0.003, + "loss": 4.2814, + "step": 1594 + }, + { + "epoch": 0.01595, + "grad_norm": 0.4303780684833319, + "learning_rate": 0.003, + "loss": 4.308, + "step": 1595 + }, + { + "epoch": 0.01596, + "grad_norm": 0.4038673300031157, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1596 + }, + { + "epoch": 0.01597, + "grad_norm": 0.42824275182777244, + "learning_rate": 0.003, + "loss": 4.2611, + "step": 1597 + }, + { + "epoch": 0.01598, + "grad_norm": 0.4949313860176736, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1598 + }, + { + "epoch": 0.01599, + "grad_norm": 0.5005123390361799, + "learning_rate": 0.003, + "loss": 4.2803, + "step": 1599 + }, + { + "epoch": 0.016, + "grad_norm": 0.5929838548199676, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 1600 + }, + { + "epoch": 0.01601, + "grad_norm": 0.8301504230585989, + "learning_rate": 0.003, + "loss": 4.315, + "step": 1601 + }, + { + "epoch": 0.01602, + "grad_norm": 0.9853750506753596, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1602 + }, + { + "epoch": 0.01603, + "grad_norm": 1.003124318602637, + "learning_rate": 0.003, + "loss": 4.3258, + "step": 1603 + }, + { + "epoch": 0.01604, + "grad_norm": 0.912578535311941, + "learning_rate": 0.003, + "loss": 4.327, + "step": 1604 + }, + { + "epoch": 0.01605, + "grad_norm": 0.8301067621469277, + "learning_rate": 0.003, + "loss": 4.3002, + "step": 1605 + }, + { + "epoch": 0.01606, + "grad_norm": 0.7646428413230717, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1606 + }, + { + "epoch": 0.01607, + "grad_norm": 0.8210614746463125, + "learning_rate": 0.003, + "loss": 4.3396, + "step": 1607 + }, + { + "epoch": 0.01608, + "grad_norm": 0.8162978830688478, + "learning_rate": 0.003, + "loss": 4.3084, + "step": 1608 + }, + { + "epoch": 0.01609, + "grad_norm": 0.7940470937374723, + "learning_rate": 0.003, + "loss": 4.3225, + "step": 1609 + }, + { + "epoch": 0.0161, + "grad_norm": 0.7532673019192272, + "learning_rate": 0.003, + "loss": 4.2882, + "step": 1610 + }, + { + "epoch": 0.01611, + "grad_norm": 0.77086180161133, + "learning_rate": 0.003, + "loss": 4.3117, + "step": 1611 + }, + { + "epoch": 0.01612, + "grad_norm": 0.6754537663457288, + "learning_rate": 0.003, + "loss": 4.2961, + "step": 1612 + }, + { + "epoch": 0.01613, + "grad_norm": 0.599175162495873, + "learning_rate": 0.003, + "loss": 4.3386, + "step": 1613 + }, + { + "epoch": 0.01614, + "grad_norm": 0.5411062109149707, + "learning_rate": 0.003, + "loss": 4.3566, + "step": 1614 + }, + { + "epoch": 0.01615, + "grad_norm": 0.512685672879493, + "learning_rate": 0.003, + "loss": 4.3203, + "step": 1615 + }, + { + "epoch": 0.01616, + "grad_norm": 0.49714926242353863, + "learning_rate": 0.003, + "loss": 4.3009, + "step": 1616 + }, + { + "epoch": 0.01617, + "grad_norm": 0.4964011411644138, + "learning_rate": 0.003, + "loss": 4.2889, + "step": 1617 + }, + { + "epoch": 0.01618, + "grad_norm": 0.5061608293930215, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1618 + }, + { + "epoch": 0.01619, + "grad_norm": 0.5494277838056353, + "learning_rate": 0.003, + "loss": 4.2811, + "step": 1619 + }, + { + "epoch": 0.0162, + "grad_norm": 0.6308404917903175, + "learning_rate": 0.003, + "loss": 4.2928, + "step": 1620 + }, + { + "epoch": 0.01621, + "grad_norm": 0.731307523044738, + "learning_rate": 0.003, + "loss": 4.2937, + "step": 1621 + }, + { + "epoch": 0.01622, + "grad_norm": 0.9046402251963843, + "learning_rate": 0.003, + "loss": 4.3346, + "step": 1622 + }, + { + "epoch": 0.01623, + "grad_norm": 0.7735409772436379, + "learning_rate": 0.003, + "loss": 4.3147, + "step": 1623 + }, + { + "epoch": 0.01624, + "grad_norm": 0.672645275853148, + "learning_rate": 0.003, + "loss": 4.3229, + "step": 1624 + }, + { + "epoch": 0.01625, + "grad_norm": 0.6719787877984026, + "learning_rate": 0.003, + "loss": 4.2983, + "step": 1625 + }, + { + "epoch": 0.01626, + "grad_norm": 0.6708647627350693, + "learning_rate": 0.003, + "loss": 4.3168, + "step": 1626 + }, + { + "epoch": 0.01627, + "grad_norm": 0.649063452794678, + "learning_rate": 0.003, + "loss": 4.2886, + "step": 1627 + }, + { + "epoch": 0.01628, + "grad_norm": 0.6161891488716968, + "learning_rate": 0.003, + "loss": 4.2901, + "step": 1628 + }, + { + "epoch": 0.01629, + "grad_norm": 0.5565589154413988, + "learning_rate": 0.003, + "loss": 4.2903, + "step": 1629 + }, + { + "epoch": 0.0163, + "grad_norm": 0.5030071975984001, + "learning_rate": 0.003, + "loss": 4.2755, + "step": 1630 + }, + { + "epoch": 0.01631, + "grad_norm": 0.48590911922639535, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1631 + }, + { + "epoch": 0.01632, + "grad_norm": 0.5725399795731132, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1632 + }, + { + "epoch": 0.01633, + "grad_norm": 0.7370492556479813, + "learning_rate": 0.003, + "loss": 4.2846, + "step": 1633 + }, + { + "epoch": 0.01634, + "grad_norm": 0.8641758267856038, + "learning_rate": 0.003, + "loss": 4.3131, + "step": 1634 + }, + { + "epoch": 0.01635, + "grad_norm": 0.7344051689167792, + "learning_rate": 0.003, + "loss": 4.2907, + "step": 1635 + }, + { + "epoch": 0.01636, + "grad_norm": 0.5215072843138687, + "learning_rate": 0.003, + "loss": 4.2982, + "step": 1636 + }, + { + "epoch": 0.01637, + "grad_norm": 0.4749653803767528, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1637 + }, + { + "epoch": 0.01638, + "grad_norm": 0.4605003028840984, + "learning_rate": 0.003, + "loss": 4.2856, + "step": 1638 + }, + { + "epoch": 0.01639, + "grad_norm": 0.42552013547580053, + "learning_rate": 0.003, + "loss": 4.3073, + "step": 1639 + }, + { + "epoch": 0.0164, + "grad_norm": 0.4220228883023831, + "learning_rate": 0.003, + "loss": 4.2853, + "step": 1640 + }, + { + "epoch": 0.01641, + "grad_norm": 0.48264096364388565, + "learning_rate": 0.003, + "loss": 4.2601, + "step": 1641 + }, + { + "epoch": 0.01642, + "grad_norm": 0.5522814057439278, + "learning_rate": 0.003, + "loss": 4.3056, + "step": 1642 + }, + { + "epoch": 0.01643, + "grad_norm": 0.5448290909435038, + "learning_rate": 0.003, + "loss": 4.2899, + "step": 1643 + }, + { + "epoch": 0.01644, + "grad_norm": 0.4693252715517831, + "learning_rate": 0.003, + "loss": 4.3048, + "step": 1644 + }, + { + "epoch": 0.01645, + "grad_norm": 0.38397009629291196, + "learning_rate": 0.003, + "loss": 4.2478, + "step": 1645 + }, + { + "epoch": 0.01646, + "grad_norm": 0.47479208422272845, + "learning_rate": 0.003, + "loss": 4.2841, + "step": 1646 + }, + { + "epoch": 0.01647, + "grad_norm": 0.5095257415638307, + "learning_rate": 0.003, + "loss": 4.2814, + "step": 1647 + }, + { + "epoch": 0.01648, + "grad_norm": 0.5223808020316233, + "learning_rate": 0.003, + "loss": 4.2591, + "step": 1648 + }, + { + "epoch": 0.01649, + "grad_norm": 0.4909186651871344, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1649 + }, + { + "epoch": 0.0165, + "grad_norm": 0.4445354228997141, + "learning_rate": 0.003, + "loss": 4.2687, + "step": 1650 + }, + { + "epoch": 0.01651, + "grad_norm": 0.47462201592081604, + "learning_rate": 0.003, + "loss": 4.2796, + "step": 1651 + }, + { + "epoch": 0.01652, + "grad_norm": 0.5712842642113961, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1652 + }, + { + "epoch": 0.01653, + "grad_norm": 0.6633586551393839, + "learning_rate": 0.003, + "loss": 4.2872, + "step": 1653 + }, + { + "epoch": 0.01654, + "grad_norm": 0.6882337845469141, + "learning_rate": 0.003, + "loss": 4.2797, + "step": 1654 + }, + { + "epoch": 0.01655, + "grad_norm": 0.5883903260155589, + "learning_rate": 0.003, + "loss": 4.2855, + "step": 1655 + }, + { + "epoch": 0.01656, + "grad_norm": 0.692175129507891, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1656 + }, + { + "epoch": 0.01657, + "grad_norm": 0.7208024346780962, + "learning_rate": 0.003, + "loss": 4.285, + "step": 1657 + }, + { + "epoch": 0.01658, + "grad_norm": 0.7405524865262961, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 1658 + }, + { + "epoch": 0.01659, + "grad_norm": 0.6893828101785304, + "learning_rate": 0.003, + "loss": 4.2781, + "step": 1659 + }, + { + "epoch": 0.0166, + "grad_norm": 0.6598372867680238, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1660 + }, + { + "epoch": 0.01661, + "grad_norm": 0.47529948561931934, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 1661 + }, + { + "epoch": 0.01662, + "grad_norm": 0.46023265898975174, + "learning_rate": 0.003, + "loss": 4.2646, + "step": 1662 + }, + { + "epoch": 0.01663, + "grad_norm": 0.4507377258313933, + "learning_rate": 0.003, + "loss": 4.2842, + "step": 1663 + }, + { + "epoch": 0.01664, + "grad_norm": 0.4200561869900329, + "learning_rate": 0.003, + "loss": 4.2999, + "step": 1664 + }, + { + "epoch": 0.01665, + "grad_norm": 0.4244798072607477, + "learning_rate": 0.003, + "loss": 4.2616, + "step": 1665 + }, + { + "epoch": 0.01666, + "grad_norm": 0.5427314965460833, + "learning_rate": 0.003, + "loss": 4.2661, + "step": 1666 + }, + { + "epoch": 0.01667, + "grad_norm": 0.5950166710073447, + "learning_rate": 0.003, + "loss": 4.2869, + "step": 1667 + }, + { + "epoch": 0.01668, + "grad_norm": 0.7033617268618768, + "learning_rate": 0.003, + "loss": 4.2937, + "step": 1668 + }, + { + "epoch": 0.01669, + "grad_norm": 0.7337795007439258, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1669 + }, + { + "epoch": 0.0167, + "grad_norm": 0.7228924020186139, + "learning_rate": 0.003, + "loss": 4.2759, + "step": 1670 + }, + { + "epoch": 0.01671, + "grad_norm": 0.9293602473006644, + "learning_rate": 0.003, + "loss": 4.2785, + "step": 1671 + }, + { + "epoch": 0.01672, + "grad_norm": 1.029746366393395, + "learning_rate": 0.003, + "loss": 4.3064, + "step": 1672 + }, + { + "epoch": 0.01673, + "grad_norm": 0.9727787844526897, + "learning_rate": 0.003, + "loss": 4.3102, + "step": 1673 + }, + { + "epoch": 0.01674, + "grad_norm": 0.9554074542733949, + "learning_rate": 0.003, + "loss": 4.3312, + "step": 1674 + }, + { + "epoch": 0.01675, + "grad_norm": 0.9706031573401521, + "learning_rate": 0.003, + "loss": 4.3194, + "step": 1675 + }, + { + "epoch": 0.01676, + "grad_norm": 0.9822377528346978, + "learning_rate": 0.003, + "loss": 4.3319, + "step": 1676 + }, + { + "epoch": 0.01677, + "grad_norm": 0.9324715324675055, + "learning_rate": 0.003, + "loss": 4.335, + "step": 1677 + }, + { + "epoch": 0.01678, + "grad_norm": 0.8555988190467146, + "learning_rate": 0.003, + "loss": 4.3399, + "step": 1678 + }, + { + "epoch": 0.01679, + "grad_norm": 0.9906267827738694, + "learning_rate": 0.003, + "loss": 4.3344, + "step": 1679 + }, + { + "epoch": 0.0168, + "grad_norm": 0.9961431022000035, + "learning_rate": 0.003, + "loss": 4.3626, + "step": 1680 + }, + { + "epoch": 0.01681, + "grad_norm": 0.8880584596730197, + "learning_rate": 0.003, + "loss": 4.362, + "step": 1681 + }, + { + "epoch": 0.01682, + "grad_norm": 0.864035612954157, + "learning_rate": 0.003, + "loss": 4.3519, + "step": 1682 + }, + { + "epoch": 0.01683, + "grad_norm": 0.830453292981266, + "learning_rate": 0.003, + "loss": 4.3671, + "step": 1683 + }, + { + "epoch": 0.01684, + "grad_norm": 0.8911963046678297, + "learning_rate": 0.003, + "loss": 4.3454, + "step": 1684 + }, + { + "epoch": 0.01685, + "grad_norm": 0.8195919434333439, + "learning_rate": 0.003, + "loss": 4.3594, + "step": 1685 + }, + { + "epoch": 0.01686, + "grad_norm": 0.8629708344208772, + "learning_rate": 0.003, + "loss": 4.3459, + "step": 1686 + }, + { + "epoch": 0.01687, + "grad_norm": 0.7564331005412712, + "learning_rate": 0.003, + "loss": 4.3417, + "step": 1687 + }, + { + "epoch": 0.01688, + "grad_norm": 0.5556779425751791, + "learning_rate": 0.003, + "loss": 4.3178, + "step": 1688 + }, + { + "epoch": 0.01689, + "grad_norm": 0.48730261859047036, + "learning_rate": 0.003, + "loss": 4.3272, + "step": 1689 + }, + { + "epoch": 0.0169, + "grad_norm": 0.44730185947220713, + "learning_rate": 0.003, + "loss": 4.3108, + "step": 1690 + }, + { + "epoch": 0.01691, + "grad_norm": 0.4517455858191217, + "learning_rate": 0.003, + "loss": 4.3185, + "step": 1691 + }, + { + "epoch": 0.01692, + "grad_norm": 0.3859623506260394, + "learning_rate": 0.003, + "loss": 4.3134, + "step": 1692 + }, + { + "epoch": 0.01693, + "grad_norm": 0.4280684753336336, + "learning_rate": 0.003, + "loss": 4.3374, + "step": 1693 + }, + { + "epoch": 0.01694, + "grad_norm": 0.430827914877835, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1694 + }, + { + "epoch": 0.01695, + "grad_norm": 0.43375392594165474, + "learning_rate": 0.003, + "loss": 4.3125, + "step": 1695 + }, + { + "epoch": 0.01696, + "grad_norm": 0.38718080857219184, + "learning_rate": 0.003, + "loss": 4.2992, + "step": 1696 + }, + { + "epoch": 0.01697, + "grad_norm": 0.30363972361127406, + "learning_rate": 0.003, + "loss": 4.272, + "step": 1697 + }, + { + "epoch": 0.01698, + "grad_norm": 0.31558190433180666, + "learning_rate": 0.003, + "loss": 4.2652, + "step": 1698 + }, + { + "epoch": 0.01699, + "grad_norm": 0.3101672448572644, + "learning_rate": 0.003, + "loss": 4.2732, + "step": 1699 + }, + { + "epoch": 0.017, + "grad_norm": 0.28571559628804677, + "learning_rate": 0.003, + "loss": 4.2885, + "step": 1700 + }, + { + "epoch": 0.01701, + "grad_norm": 0.31584250090208377, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1701 + }, + { + "epoch": 0.01702, + "grad_norm": 0.36960398708639597, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1702 + }, + { + "epoch": 0.01703, + "grad_norm": 0.4412523037970288, + "learning_rate": 0.003, + "loss": 4.277, + "step": 1703 + }, + { + "epoch": 0.01704, + "grad_norm": 0.6716654424542439, + "learning_rate": 0.003, + "loss": 4.2707, + "step": 1704 + }, + { + "epoch": 0.01705, + "grad_norm": 1.1072937714524183, + "learning_rate": 0.003, + "loss": 4.2932, + "step": 1705 + }, + { + "epoch": 0.01706, + "grad_norm": 0.8287243662947943, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1706 + }, + { + "epoch": 0.01707, + "grad_norm": 0.6387856390419926, + "learning_rate": 0.003, + "loss": 4.3079, + "step": 1707 + }, + { + "epoch": 0.01708, + "grad_norm": 0.7665689860690651, + "learning_rate": 0.003, + "loss": 4.2925, + "step": 1708 + }, + { + "epoch": 0.01709, + "grad_norm": 0.6842592484083714, + "learning_rate": 0.003, + "loss": 4.3099, + "step": 1709 + }, + { + "epoch": 0.0171, + "grad_norm": 0.5640128407150151, + "learning_rate": 0.003, + "loss": 4.2554, + "step": 1710 + }, + { + "epoch": 0.01711, + "grad_norm": 0.5706999470034969, + "learning_rate": 0.003, + "loss": 4.2824, + "step": 1711 + }, + { + "epoch": 0.01712, + "grad_norm": 0.4795902281641074, + "learning_rate": 0.003, + "loss": 4.2884, + "step": 1712 + }, + { + "epoch": 0.01713, + "grad_norm": 0.425849272235493, + "learning_rate": 0.003, + "loss": 4.2609, + "step": 1713 + }, + { + "epoch": 0.01714, + "grad_norm": 0.38297971865347935, + "learning_rate": 0.003, + "loss": 4.2951, + "step": 1714 + }, + { + "epoch": 0.01715, + "grad_norm": 0.3866137138438359, + "learning_rate": 0.003, + "loss": 4.3021, + "step": 1715 + }, + { + "epoch": 0.01716, + "grad_norm": 0.35314982927049987, + "learning_rate": 0.003, + "loss": 4.2748, + "step": 1716 + }, + { + "epoch": 0.01717, + "grad_norm": 0.31552154478040295, + "learning_rate": 0.003, + "loss": 4.3044, + "step": 1717 + }, + { + "epoch": 0.01718, + "grad_norm": 0.30638930868708425, + "learning_rate": 0.003, + "loss": 4.2709, + "step": 1718 + }, + { + "epoch": 0.01719, + "grad_norm": 0.3127248594754382, + "learning_rate": 0.003, + "loss": 4.2905, + "step": 1719 + }, + { + "epoch": 0.0172, + "grad_norm": 0.33132313581376, + "learning_rate": 0.003, + "loss": 4.2482, + "step": 1720 + }, + { + "epoch": 0.01721, + "grad_norm": 0.3400058323027828, + "learning_rate": 0.003, + "loss": 4.2635, + "step": 1721 + }, + { + "epoch": 0.01722, + "grad_norm": 0.3468025060877176, + "learning_rate": 0.003, + "loss": 4.2559, + "step": 1722 + }, + { + "epoch": 0.01723, + "grad_norm": 0.35695313771503173, + "learning_rate": 0.003, + "loss": 4.2636, + "step": 1723 + }, + { + "epoch": 0.01724, + "grad_norm": 0.34758213513255676, + "learning_rate": 0.003, + "loss": 4.2648, + "step": 1724 + }, + { + "epoch": 0.01725, + "grad_norm": 0.40361756493758844, + "learning_rate": 0.003, + "loss": 4.2508, + "step": 1725 + }, + { + "epoch": 0.01726, + "grad_norm": 0.5053798954640991, + "learning_rate": 0.003, + "loss": 4.245, + "step": 1726 + }, + { + "epoch": 0.01727, + "grad_norm": 0.6257283402319211, + "learning_rate": 0.003, + "loss": 4.2558, + "step": 1727 + }, + { + "epoch": 0.01728, + "grad_norm": 0.5898179215522085, + "learning_rate": 0.003, + "loss": 4.2591, + "step": 1728 + }, + { + "epoch": 0.01729, + "grad_norm": 0.5313524252553607, + "learning_rate": 0.003, + "loss": 4.2643, + "step": 1729 + }, + { + "epoch": 0.0173, + "grad_norm": 0.6103904325056944, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 1730 + }, + { + "epoch": 0.01731, + "grad_norm": 0.498773172096663, + "learning_rate": 0.003, + "loss": 4.2592, + "step": 1731 + }, + { + "epoch": 0.01732, + "grad_norm": 0.6373537928005123, + "learning_rate": 0.003, + "loss": 4.2725, + "step": 1732 + }, + { + "epoch": 0.01733, + "grad_norm": 0.7241473038492131, + "learning_rate": 0.003, + "loss": 4.257, + "step": 1733 + }, + { + "epoch": 0.01734, + "grad_norm": 0.8759884069698398, + "learning_rate": 0.003, + "loss": 4.2776, + "step": 1734 + }, + { + "epoch": 0.01735, + "grad_norm": 0.8320108473030087, + "learning_rate": 0.003, + "loss": 4.2551, + "step": 1735 + }, + { + "epoch": 0.01736, + "grad_norm": 0.6879588721711866, + "learning_rate": 0.003, + "loss": 4.2513, + "step": 1736 + }, + { + "epoch": 0.01737, + "grad_norm": 0.6180790162276626, + "learning_rate": 0.003, + "loss": 4.2782, + "step": 1737 + }, + { + "epoch": 0.01738, + "grad_norm": 0.6570007295297866, + "learning_rate": 0.003, + "loss": 4.2689, + "step": 1738 + }, + { + "epoch": 0.01739, + "grad_norm": 0.7234192244460489, + "learning_rate": 0.003, + "loss": 4.288, + "step": 1739 + }, + { + "epoch": 0.0174, + "grad_norm": 0.646718379715682, + "learning_rate": 0.003, + "loss": 4.2832, + "step": 1740 + }, + { + "epoch": 0.01741, + "grad_norm": 0.7667295546998241, + "learning_rate": 0.003, + "loss": 4.2989, + "step": 1741 + }, + { + "epoch": 0.01742, + "grad_norm": 0.6671673777635381, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1742 + }, + { + "epoch": 0.01743, + "grad_norm": 0.6506146600686363, + "learning_rate": 0.003, + "loss": 4.2938, + "step": 1743 + }, + { + "epoch": 0.01744, + "grad_norm": 0.6573560282994302, + "learning_rate": 0.003, + "loss": 4.2796, + "step": 1744 + }, + { + "epoch": 0.01745, + "grad_norm": 0.640684434992135, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 1745 + }, + { + "epoch": 0.01746, + "grad_norm": 0.657793650803233, + "learning_rate": 0.003, + "loss": 4.2671, + "step": 1746 + }, + { + "epoch": 0.01747, + "grad_norm": 0.7709605353943576, + "learning_rate": 0.003, + "loss": 4.2717, + "step": 1747 + }, + { + "epoch": 0.01748, + "grad_norm": 0.8278007400279545, + "learning_rate": 0.003, + "loss": 4.2678, + "step": 1748 + }, + { + "epoch": 0.01749, + "grad_norm": 0.7466841943172032, + "learning_rate": 0.003, + "loss": 4.2633, + "step": 1749 + }, + { + "epoch": 0.0175, + "grad_norm": 0.763703983777172, + "learning_rate": 0.003, + "loss": 4.2821, + "step": 1750 + }, + { + "epoch": 0.01751, + "grad_norm": 0.7380887125320881, + "learning_rate": 0.003, + "loss": 4.2741, + "step": 1751 + }, + { + "epoch": 0.01752, + "grad_norm": 0.6415685853882744, + "learning_rate": 0.003, + "loss": 4.314, + "step": 1752 + }, + { + "epoch": 0.01753, + "grad_norm": 0.6180937585023022, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 1753 + }, + { + "epoch": 0.01754, + "grad_norm": 0.6258159102980642, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1754 + }, + { + "epoch": 0.01755, + "grad_norm": 0.5258294344568889, + "learning_rate": 0.003, + "loss": 4.2275, + "step": 1755 + }, + { + "epoch": 0.01756, + "grad_norm": 0.4224710169236117, + "learning_rate": 0.003, + "loss": 4.3014, + "step": 1756 + }, + { + "epoch": 0.01757, + "grad_norm": 0.4609603622580685, + "learning_rate": 0.003, + "loss": 4.2799, + "step": 1757 + }, + { + "epoch": 0.01758, + "grad_norm": 0.5088646105086777, + "learning_rate": 0.003, + "loss": 4.2854, + "step": 1758 + }, + { + "epoch": 0.01759, + "grad_norm": 0.6116749982391423, + "learning_rate": 0.003, + "loss": 4.2559, + "step": 1759 + }, + { + "epoch": 0.0176, + "grad_norm": 0.6421141472355831, + "learning_rate": 0.003, + "loss": 4.2809, + "step": 1760 + }, + { + "epoch": 0.01761, + "grad_norm": 0.683100042759788, + "learning_rate": 0.003, + "loss": 4.2664, + "step": 1761 + }, + { + "epoch": 0.01762, + "grad_norm": 0.6507618810640869, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1762 + }, + { + "epoch": 0.01763, + "grad_norm": 0.6268432713329561, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1763 + }, + { + "epoch": 0.01764, + "grad_norm": 0.5913417648906798, + "learning_rate": 0.003, + "loss": 4.2433, + "step": 1764 + }, + { + "epoch": 0.01765, + "grad_norm": 0.6076223877479996, + "learning_rate": 0.003, + "loss": 4.2712, + "step": 1765 + }, + { + "epoch": 0.01766, + "grad_norm": 0.7266026161134381, + "learning_rate": 0.003, + "loss": 4.2888, + "step": 1766 + }, + { + "epoch": 0.01767, + "grad_norm": 0.7022487568636244, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 1767 + }, + { + "epoch": 0.01768, + "grad_norm": 0.5934615837163861, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1768 + }, + { + "epoch": 0.01769, + "grad_norm": 0.4361100704507323, + "learning_rate": 0.003, + "loss": 4.2511, + "step": 1769 + }, + { + "epoch": 0.0177, + "grad_norm": 0.48611120007740843, + "learning_rate": 0.003, + "loss": 4.2382, + "step": 1770 + }, + { + "epoch": 0.01771, + "grad_norm": 0.45731585839886957, + "learning_rate": 0.003, + "loss": 4.266, + "step": 1771 + }, + { + "epoch": 0.01772, + "grad_norm": 0.4581498365355611, + "learning_rate": 0.003, + "loss": 4.2403, + "step": 1772 + }, + { + "epoch": 0.01773, + "grad_norm": 0.3835918309901609, + "learning_rate": 0.003, + "loss": 4.2284, + "step": 1773 + }, + { + "epoch": 0.01774, + "grad_norm": 0.4161002686293629, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1774 + }, + { + "epoch": 0.01775, + "grad_norm": 0.48314560217147945, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 1775 + }, + { + "epoch": 0.01776, + "grad_norm": 0.550180805967614, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 1776 + }, + { + "epoch": 0.01777, + "grad_norm": 0.570528755486745, + "learning_rate": 0.003, + "loss": 4.2588, + "step": 1777 + }, + { + "epoch": 0.01778, + "grad_norm": 0.6830277922868914, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1778 + }, + { + "epoch": 0.01779, + "grad_norm": 0.8262794189445563, + "learning_rate": 0.003, + "loss": 4.281, + "step": 1779 + }, + { + "epoch": 0.0178, + "grad_norm": 0.85849663326723, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1780 + }, + { + "epoch": 0.01781, + "grad_norm": 0.7027313657081399, + "learning_rate": 0.003, + "loss": 4.2685, + "step": 1781 + }, + { + "epoch": 0.01782, + "grad_norm": 0.5620584115129139, + "learning_rate": 0.003, + "loss": 4.2878, + "step": 1782 + }, + { + "epoch": 0.01783, + "grad_norm": 0.5096347371759039, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1783 + }, + { + "epoch": 0.01784, + "grad_norm": 0.5297569056185621, + "learning_rate": 0.003, + "loss": 4.2633, + "step": 1784 + }, + { + "epoch": 0.01785, + "grad_norm": 0.6034694109958456, + "learning_rate": 0.003, + "loss": 4.2267, + "step": 1785 + }, + { + "epoch": 0.01786, + "grad_norm": 0.6309039198553906, + "learning_rate": 0.003, + "loss": 4.2565, + "step": 1786 + }, + { + "epoch": 0.01787, + "grad_norm": 0.6879917544738386, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 1787 + }, + { + "epoch": 0.01788, + "grad_norm": 0.6818187529366616, + "learning_rate": 0.003, + "loss": 4.2789, + "step": 1788 + }, + { + "epoch": 0.01789, + "grad_norm": 0.5930877660332807, + "learning_rate": 0.003, + "loss": 4.2445, + "step": 1789 + }, + { + "epoch": 0.0179, + "grad_norm": 0.5104382089041954, + "learning_rate": 0.003, + "loss": 4.2605, + "step": 1790 + }, + { + "epoch": 0.01791, + "grad_norm": 0.48343390667073793, + "learning_rate": 0.003, + "loss": 4.2706, + "step": 1791 + }, + { + "epoch": 0.01792, + "grad_norm": 0.5502398127361193, + "learning_rate": 0.003, + "loss": 4.2742, + "step": 1792 + }, + { + "epoch": 0.01793, + "grad_norm": 0.514623976410524, + "learning_rate": 0.003, + "loss": 4.235, + "step": 1793 + }, + { + "epoch": 0.01794, + "grad_norm": 0.5328860529823664, + "learning_rate": 0.003, + "loss": 4.2517, + "step": 1794 + }, + { + "epoch": 0.01795, + "grad_norm": 0.5324866930035196, + "learning_rate": 0.003, + "loss": 4.2511, + "step": 1795 + }, + { + "epoch": 0.01796, + "grad_norm": 0.4431327306726926, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1796 + }, + { + "epoch": 0.01797, + "grad_norm": 0.4163239838007479, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 1797 + }, + { + "epoch": 0.01798, + "grad_norm": 0.4680057225647005, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 1798 + }, + { + "epoch": 0.01799, + "grad_norm": 0.5257929933308816, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 1799 + }, + { + "epoch": 0.018, + "grad_norm": 0.6818533479906342, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1800 + }, + { + "epoch": 0.01801, + "grad_norm": 0.8347036722912681, + "learning_rate": 0.003, + "loss": 4.2531, + "step": 1801 + }, + { + "epoch": 0.01802, + "grad_norm": 0.8911873404291202, + "learning_rate": 0.003, + "loss": 4.2851, + "step": 1802 + }, + { + "epoch": 0.01803, + "grad_norm": 0.8674521299479792, + "learning_rate": 0.003, + "loss": 4.2905, + "step": 1803 + }, + { + "epoch": 0.01804, + "grad_norm": 0.8602615225653094, + "learning_rate": 0.003, + "loss": 4.2682, + "step": 1804 + }, + { + "epoch": 0.01805, + "grad_norm": 0.8301928044098125, + "learning_rate": 0.003, + "loss": 4.2953, + "step": 1805 + }, + { + "epoch": 0.01806, + "grad_norm": 0.7981063012004259, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1806 + }, + { + "epoch": 0.01807, + "grad_norm": 0.8614656914400735, + "learning_rate": 0.003, + "loss": 4.2917, + "step": 1807 + }, + { + "epoch": 0.01808, + "grad_norm": 0.7448619096800406, + "learning_rate": 0.003, + "loss": 4.2867, + "step": 1808 + }, + { + "epoch": 0.01809, + "grad_norm": 0.6456900848105672, + "learning_rate": 0.003, + "loss": 4.2963, + "step": 1809 + }, + { + "epoch": 0.0181, + "grad_norm": 0.5841374638529157, + "learning_rate": 0.003, + "loss": 4.2907, + "step": 1810 + }, + { + "epoch": 0.01811, + "grad_norm": 0.5584017956424331, + "learning_rate": 0.003, + "loss": 4.2718, + "step": 1811 + }, + { + "epoch": 0.01812, + "grad_norm": 0.5780683839817958, + "learning_rate": 0.003, + "loss": 4.2546, + "step": 1812 + }, + { + "epoch": 0.01813, + "grad_norm": 0.5210766747257082, + "learning_rate": 0.003, + "loss": 4.298, + "step": 1813 + }, + { + "epoch": 0.01814, + "grad_norm": 0.5061687855504291, + "learning_rate": 0.003, + "loss": 4.2617, + "step": 1814 + }, + { + "epoch": 0.01815, + "grad_norm": 0.5646752857396716, + "learning_rate": 0.003, + "loss": 4.2497, + "step": 1815 + }, + { + "epoch": 0.01816, + "grad_norm": 0.6329676204068563, + "learning_rate": 0.003, + "loss": 4.2779, + "step": 1816 + }, + { + "epoch": 0.01817, + "grad_norm": 0.576517926530056, + "learning_rate": 0.003, + "loss": 4.2664, + "step": 1817 + }, + { + "epoch": 0.01818, + "grad_norm": 0.4950065288657199, + "learning_rate": 0.003, + "loss": 4.2615, + "step": 1818 + }, + { + "epoch": 0.01819, + "grad_norm": 0.5161367831332595, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 1819 + }, + { + "epoch": 0.0182, + "grad_norm": 0.5814618735674658, + "learning_rate": 0.003, + "loss": 4.2663, + "step": 1820 + }, + { + "epoch": 0.01821, + "grad_norm": 0.556938739734867, + "learning_rate": 0.003, + "loss": 4.251, + "step": 1821 + }, + { + "epoch": 0.01822, + "grad_norm": 0.5708870221750882, + "learning_rate": 0.003, + "loss": 4.2636, + "step": 1822 + }, + { + "epoch": 0.01823, + "grad_norm": 0.5398260260215829, + "learning_rate": 0.003, + "loss": 4.2764, + "step": 1823 + }, + { + "epoch": 0.01824, + "grad_norm": 0.5454416785979428, + "learning_rate": 0.003, + "loss": 4.2571, + "step": 1824 + }, + { + "epoch": 0.01825, + "grad_norm": 0.6656252297688491, + "learning_rate": 0.003, + "loss": 4.2668, + "step": 1825 + }, + { + "epoch": 0.01826, + "grad_norm": 0.9269512928572629, + "learning_rate": 0.003, + "loss": 4.289, + "step": 1826 + }, + { + "epoch": 0.01827, + "grad_norm": 1.076217745090645, + "learning_rate": 0.003, + "loss": 4.2819, + "step": 1827 + }, + { + "epoch": 0.01828, + "grad_norm": 0.650029561792675, + "learning_rate": 0.003, + "loss": 4.2717, + "step": 1828 + }, + { + "epoch": 0.01829, + "grad_norm": 0.5962218186152778, + "learning_rate": 0.003, + "loss": 4.2866, + "step": 1829 + }, + { + "epoch": 0.0183, + "grad_norm": 0.6672932732108263, + "learning_rate": 0.003, + "loss": 4.2811, + "step": 1830 + }, + { + "epoch": 0.01831, + "grad_norm": 0.6051971108876394, + "learning_rate": 0.003, + "loss": 4.2699, + "step": 1831 + }, + { + "epoch": 0.01832, + "grad_norm": 0.5439007581448924, + "learning_rate": 0.003, + "loss": 4.2598, + "step": 1832 + }, + { + "epoch": 0.01833, + "grad_norm": 0.55902144104146, + "learning_rate": 0.003, + "loss": 4.2938, + "step": 1833 + }, + { + "epoch": 0.01834, + "grad_norm": 0.5728509198658349, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1834 + }, + { + "epoch": 0.01835, + "grad_norm": 0.5340971547409412, + "learning_rate": 0.003, + "loss": 4.2588, + "step": 1835 + }, + { + "epoch": 0.01836, + "grad_norm": 0.47778636042203493, + "learning_rate": 0.003, + "loss": 4.2504, + "step": 1836 + }, + { + "epoch": 0.01837, + "grad_norm": 0.4158033301420617, + "learning_rate": 0.003, + "loss": 4.2504, + "step": 1837 + }, + { + "epoch": 0.01838, + "grad_norm": 0.5117944575544118, + "learning_rate": 0.003, + "loss": 4.2549, + "step": 1838 + }, + { + "epoch": 0.01839, + "grad_norm": 0.6067550068103876, + "learning_rate": 0.003, + "loss": 4.2791, + "step": 1839 + }, + { + "epoch": 0.0184, + "grad_norm": 0.6159872555117356, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 1840 + }, + { + "epoch": 0.01841, + "grad_norm": 0.5609724495757069, + "learning_rate": 0.003, + "loss": 4.2137, + "step": 1841 + }, + { + "epoch": 0.01842, + "grad_norm": 0.5051094597061737, + "learning_rate": 0.003, + "loss": 4.2465, + "step": 1842 + }, + { + "epoch": 0.01843, + "grad_norm": 0.5100819064264481, + "learning_rate": 0.003, + "loss": 4.228, + "step": 1843 + }, + { + "epoch": 0.01844, + "grad_norm": 0.5582900536993392, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 1844 + }, + { + "epoch": 0.01845, + "grad_norm": 0.6111866156562342, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 1845 + }, + { + "epoch": 0.01846, + "grad_norm": 0.6619002433974149, + "learning_rate": 0.003, + "loss": 4.2608, + "step": 1846 + }, + { + "epoch": 0.01847, + "grad_norm": 1.0200537693834115, + "learning_rate": 0.003, + "loss": 4.2703, + "step": 1847 + }, + { + "epoch": 0.01848, + "grad_norm": 1.0513050095404686, + "learning_rate": 0.003, + "loss": 4.3125, + "step": 1848 + }, + { + "epoch": 0.01849, + "grad_norm": 0.8927900194204563, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1849 + }, + { + "epoch": 0.0185, + "grad_norm": 0.9502466432757984, + "learning_rate": 0.003, + "loss": 4.316, + "step": 1850 + }, + { + "epoch": 0.01851, + "grad_norm": 0.9099370930692962, + "learning_rate": 0.003, + "loss": 4.2922, + "step": 1851 + }, + { + "epoch": 0.01852, + "grad_norm": 0.7734254610813497, + "learning_rate": 0.003, + "loss": 4.293, + "step": 1852 + }, + { + "epoch": 0.01853, + "grad_norm": 0.7300466041466253, + "learning_rate": 0.003, + "loss": 4.2883, + "step": 1853 + }, + { + "epoch": 0.01854, + "grad_norm": 0.7064556599969221, + "learning_rate": 0.003, + "loss": 4.319, + "step": 1854 + }, + { + "epoch": 0.01855, + "grad_norm": 0.5941104488816943, + "learning_rate": 0.003, + "loss": 4.2971, + "step": 1855 + }, + { + "epoch": 0.01856, + "grad_norm": 0.5639252296572004, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 1856 + }, + { + "epoch": 0.01857, + "grad_norm": 0.5497157641923732, + "learning_rate": 0.003, + "loss": 4.261, + "step": 1857 + }, + { + "epoch": 0.01858, + "grad_norm": 0.5999940363832145, + "learning_rate": 0.003, + "loss": 4.2675, + "step": 1858 + }, + { + "epoch": 0.01859, + "grad_norm": 0.6524431060775852, + "learning_rate": 0.003, + "loss": 4.2673, + "step": 1859 + }, + { + "epoch": 0.0186, + "grad_norm": 0.5786076880789475, + "learning_rate": 0.003, + "loss": 4.2435, + "step": 1860 + }, + { + "epoch": 0.01861, + "grad_norm": 0.5737732368480609, + "learning_rate": 0.003, + "loss": 4.2866, + "step": 1861 + }, + { + "epoch": 0.01862, + "grad_norm": 0.6266977494361998, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1862 + }, + { + "epoch": 0.01863, + "grad_norm": 0.5490443064279302, + "learning_rate": 0.003, + "loss": 4.2575, + "step": 1863 + }, + { + "epoch": 0.01864, + "grad_norm": 0.4951546054862347, + "learning_rate": 0.003, + "loss": 4.2885, + "step": 1864 + }, + { + "epoch": 0.01865, + "grad_norm": 0.4870782718763887, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 1865 + }, + { + "epoch": 0.01866, + "grad_norm": 0.46735035399196706, + "learning_rate": 0.003, + "loss": 4.2603, + "step": 1866 + }, + { + "epoch": 0.01867, + "grad_norm": 0.4716752041768468, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 1867 + }, + { + "epoch": 0.01868, + "grad_norm": 0.44705830429737475, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 1868 + }, + { + "epoch": 0.01869, + "grad_norm": 0.4991544420243406, + "learning_rate": 0.003, + "loss": 4.2625, + "step": 1869 + }, + { + "epoch": 0.0187, + "grad_norm": 0.4942905986952634, + "learning_rate": 0.003, + "loss": 4.2547, + "step": 1870 + }, + { + "epoch": 0.01871, + "grad_norm": 0.5121508182199742, + "learning_rate": 0.003, + "loss": 4.2287, + "step": 1871 + }, + { + "epoch": 0.01872, + "grad_norm": 0.5132249094210571, + "learning_rate": 0.003, + "loss": 4.2747, + "step": 1872 + }, + { + "epoch": 0.01873, + "grad_norm": 0.45833884531988217, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 1873 + }, + { + "epoch": 0.01874, + "grad_norm": 0.4075899028314144, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1874 + }, + { + "epoch": 0.01875, + "grad_norm": 0.4250746194260601, + "learning_rate": 0.003, + "loss": 4.2517, + "step": 1875 + }, + { + "epoch": 0.01876, + "grad_norm": 0.4231054620615194, + "learning_rate": 0.003, + "loss": 4.2577, + "step": 1876 + }, + { + "epoch": 0.01877, + "grad_norm": 0.39967365013750716, + "learning_rate": 0.003, + "loss": 4.2584, + "step": 1877 + }, + { + "epoch": 0.01878, + "grad_norm": 0.4152490558819513, + "learning_rate": 0.003, + "loss": 4.2427, + "step": 1878 + }, + { + "epoch": 0.01879, + "grad_norm": 0.42538683613055095, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 1879 + }, + { + "epoch": 0.0188, + "grad_norm": 0.44463665765346116, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 1880 + }, + { + "epoch": 0.01881, + "grad_norm": 0.4351959985696958, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 1881 + }, + { + "epoch": 0.01882, + "grad_norm": 0.5024559892307489, + "learning_rate": 0.003, + "loss": 4.2542, + "step": 1882 + }, + { + "epoch": 0.01883, + "grad_norm": 0.6063994875754019, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 1883 + }, + { + "epoch": 0.01884, + "grad_norm": 0.7335676135390363, + "learning_rate": 0.003, + "loss": 4.2381, + "step": 1884 + }, + { + "epoch": 0.01885, + "grad_norm": 0.9057741844151627, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 1885 + }, + { + "epoch": 0.01886, + "grad_norm": 1.0977669981210454, + "learning_rate": 0.003, + "loss": 4.2863, + "step": 1886 + }, + { + "epoch": 0.01887, + "grad_norm": 0.6506964260225092, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1887 + }, + { + "epoch": 0.01888, + "grad_norm": 0.7018734791941831, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1888 + }, + { + "epoch": 0.01889, + "grad_norm": 0.6726610800126339, + "learning_rate": 0.003, + "loss": 4.2655, + "step": 1889 + }, + { + "epoch": 0.0189, + "grad_norm": 0.7428766906603239, + "learning_rate": 0.003, + "loss": 4.2512, + "step": 1890 + }, + { + "epoch": 0.01891, + "grad_norm": 0.684421659513949, + "learning_rate": 0.003, + "loss": 4.2969, + "step": 1891 + }, + { + "epoch": 0.01892, + "grad_norm": 0.6909566019627704, + "learning_rate": 0.003, + "loss": 4.2744, + "step": 1892 + }, + { + "epoch": 0.01893, + "grad_norm": 0.6520197616765968, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1893 + }, + { + "epoch": 0.01894, + "grad_norm": 0.6739359420697347, + "learning_rate": 0.003, + "loss": 4.2591, + "step": 1894 + }, + { + "epoch": 0.01895, + "grad_norm": 0.697458296569538, + "learning_rate": 0.003, + "loss": 4.2498, + "step": 1895 + }, + { + "epoch": 0.01896, + "grad_norm": 0.5936568619706635, + "learning_rate": 0.003, + "loss": 4.2688, + "step": 1896 + }, + { + "epoch": 0.01897, + "grad_norm": 0.5494055764115975, + "learning_rate": 0.003, + "loss": 4.2647, + "step": 1897 + }, + { + "epoch": 0.01898, + "grad_norm": 0.5676106882870927, + "learning_rate": 0.003, + "loss": 4.2339, + "step": 1898 + }, + { + "epoch": 0.01899, + "grad_norm": 0.6101424369862174, + "learning_rate": 0.003, + "loss": 4.2482, + "step": 1899 + }, + { + "epoch": 0.019, + "grad_norm": 0.6545357483725045, + "learning_rate": 0.003, + "loss": 4.2443, + "step": 1900 + }, + { + "epoch": 0.01901, + "grad_norm": 0.6221145886969454, + "learning_rate": 0.003, + "loss": 4.29, + "step": 1901 + }, + { + "epoch": 0.01902, + "grad_norm": 0.47771631002661874, + "learning_rate": 0.003, + "loss": 4.2605, + "step": 1902 + }, + { + "epoch": 0.01903, + "grad_norm": 0.503444994520635, + "learning_rate": 0.003, + "loss": 4.215, + "step": 1903 + }, + { + "epoch": 0.01904, + "grad_norm": 0.5071441864936951, + "learning_rate": 0.003, + "loss": 4.2711, + "step": 1904 + }, + { + "epoch": 0.01905, + "grad_norm": 0.5754503645956631, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1905 + }, + { + "epoch": 0.01906, + "grad_norm": 0.6654408282834325, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1906 + }, + { + "epoch": 0.01907, + "grad_norm": 0.7906613760805888, + "learning_rate": 0.003, + "loss": 4.2662, + "step": 1907 + }, + { + "epoch": 0.01908, + "grad_norm": 0.8850433480932498, + "learning_rate": 0.003, + "loss": 4.2906, + "step": 1908 + }, + { + "epoch": 0.01909, + "grad_norm": 0.806456676126049, + "learning_rate": 0.003, + "loss": 4.2573, + "step": 1909 + }, + { + "epoch": 0.0191, + "grad_norm": 0.5544806973924623, + "learning_rate": 0.003, + "loss": 4.256, + "step": 1910 + }, + { + "epoch": 0.01911, + "grad_norm": 0.6152192065162949, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 1911 + }, + { + "epoch": 0.01912, + "grad_norm": 0.5774523882886098, + "learning_rate": 0.003, + "loss": 4.2416, + "step": 1912 + }, + { + "epoch": 0.01913, + "grad_norm": 0.5526940825090116, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 1913 + }, + { + "epoch": 0.01914, + "grad_norm": 0.5202691625530954, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 1914 + }, + { + "epoch": 0.01915, + "grad_norm": 0.4828726686790076, + "learning_rate": 0.003, + "loss": 4.257, + "step": 1915 + }, + { + "epoch": 0.01916, + "grad_norm": 0.4684476049653884, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 1916 + }, + { + "epoch": 0.01917, + "grad_norm": 0.47660708071333646, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1917 + }, + { + "epoch": 0.01918, + "grad_norm": 0.4979902159298128, + "learning_rate": 0.003, + "loss": 4.2404, + "step": 1918 + }, + { + "epoch": 0.01919, + "grad_norm": 0.4677465287869337, + "learning_rate": 0.003, + "loss": 4.2412, + "step": 1919 + }, + { + "epoch": 0.0192, + "grad_norm": 0.42528920687498767, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 1920 + }, + { + "epoch": 0.01921, + "grad_norm": 0.4717244787464016, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 1921 + }, + { + "epoch": 0.01922, + "grad_norm": 0.4420806610942072, + "learning_rate": 0.003, + "loss": 4.2448, + "step": 1922 + }, + { + "epoch": 0.01923, + "grad_norm": 0.42058067854047654, + "learning_rate": 0.003, + "loss": 4.2491, + "step": 1923 + }, + { + "epoch": 0.01924, + "grad_norm": 0.3467314160541135, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1924 + }, + { + "epoch": 0.01925, + "grad_norm": 0.3293466925660612, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 1925 + }, + { + "epoch": 0.01926, + "grad_norm": 0.41102145828379955, + "learning_rate": 0.003, + "loss": 4.2416, + "step": 1926 + }, + { + "epoch": 0.01927, + "grad_norm": 0.720656359870687, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1927 + }, + { + "epoch": 0.01928, + "grad_norm": 1.1844212503280394, + "learning_rate": 0.003, + "loss": 4.249, + "step": 1928 + }, + { + "epoch": 0.01929, + "grad_norm": 0.9144474201054079, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 1929 + }, + { + "epoch": 0.0193, + "grad_norm": 0.8894910892570072, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 1930 + }, + { + "epoch": 0.01931, + "grad_norm": 1.0125767880809944, + "learning_rate": 0.003, + "loss": 4.2809, + "step": 1931 + }, + { + "epoch": 0.01932, + "grad_norm": 0.6957841268471371, + "learning_rate": 0.003, + "loss": 4.2689, + "step": 1932 + }, + { + "epoch": 0.01933, + "grad_norm": 0.967902005641644, + "learning_rate": 0.003, + "loss": 4.2719, + "step": 1933 + }, + { + "epoch": 0.01934, + "grad_norm": 0.9288023682490801, + "learning_rate": 0.003, + "loss": 4.2872, + "step": 1934 + }, + { + "epoch": 0.01935, + "grad_norm": 0.9414019337954325, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1935 + }, + { + "epoch": 0.01936, + "grad_norm": 0.9694221827159286, + "learning_rate": 0.003, + "loss": 4.2962, + "step": 1936 + }, + { + "epoch": 0.01937, + "grad_norm": 1.001938596118208, + "learning_rate": 0.003, + "loss": 4.3121, + "step": 1937 + }, + { + "epoch": 0.01938, + "grad_norm": 1.0140385438058088, + "learning_rate": 0.003, + "loss": 4.3241, + "step": 1938 + }, + { + "epoch": 0.01939, + "grad_norm": 1.2235452961348985, + "learning_rate": 0.003, + "loss": 4.3646, + "step": 1939 + }, + { + "epoch": 0.0194, + "grad_norm": 0.8846695796998231, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1940 + }, + { + "epoch": 0.01941, + "grad_norm": 0.8286594640280003, + "learning_rate": 0.003, + "loss": 4.3066, + "step": 1941 + }, + { + "epoch": 0.01942, + "grad_norm": 0.7123881249011045, + "learning_rate": 0.003, + "loss": 4.3226, + "step": 1942 + }, + { + "epoch": 0.01943, + "grad_norm": 0.6089638608494183, + "learning_rate": 0.003, + "loss": 4.288, + "step": 1943 + }, + { + "epoch": 0.01944, + "grad_norm": 0.5712540744847615, + "learning_rate": 0.003, + "loss": 4.3258, + "step": 1944 + }, + { + "epoch": 0.01945, + "grad_norm": 0.5597808479773261, + "learning_rate": 0.003, + "loss": 4.2885, + "step": 1945 + }, + { + "epoch": 0.01946, + "grad_norm": 0.5954089244059145, + "learning_rate": 0.003, + "loss": 4.2798, + "step": 1946 + }, + { + "epoch": 0.01947, + "grad_norm": 0.5594072113001303, + "learning_rate": 0.003, + "loss": 4.2869, + "step": 1947 + }, + { + "epoch": 0.01948, + "grad_norm": 0.47470802243769694, + "learning_rate": 0.003, + "loss": 4.3012, + "step": 1948 + }, + { + "epoch": 0.01949, + "grad_norm": 0.5123315729723936, + "learning_rate": 0.003, + "loss": 4.2904, + "step": 1949 + }, + { + "epoch": 0.0195, + "grad_norm": 0.5068647256374985, + "learning_rate": 0.003, + "loss": 4.2768, + "step": 1950 + }, + { + "epoch": 0.01951, + "grad_norm": 0.4942238546854525, + "learning_rate": 0.003, + "loss": 4.2687, + "step": 1951 + }, + { + "epoch": 0.01952, + "grad_norm": 0.4526189118950263, + "learning_rate": 0.003, + "loss": 4.271, + "step": 1952 + }, + { + "epoch": 0.01953, + "grad_norm": 0.4370221049383596, + "learning_rate": 0.003, + "loss": 4.275, + "step": 1953 + }, + { + "epoch": 0.01954, + "grad_norm": 0.46324265238758683, + "learning_rate": 0.003, + "loss": 4.2733, + "step": 1954 + }, + { + "epoch": 0.01955, + "grad_norm": 0.5071666518202474, + "learning_rate": 0.003, + "loss": 4.2654, + "step": 1955 + }, + { + "epoch": 0.01956, + "grad_norm": 0.6181690576282608, + "learning_rate": 0.003, + "loss": 4.2624, + "step": 1956 + }, + { + "epoch": 0.01957, + "grad_norm": 0.5796028080408675, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 1957 + }, + { + "epoch": 0.01958, + "grad_norm": 0.41315504837285055, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 1958 + }, + { + "epoch": 0.01959, + "grad_norm": 0.3341067438397588, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1959 + }, + { + "epoch": 0.0196, + "grad_norm": 0.3891099796372788, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1960 + }, + { + "epoch": 0.01961, + "grad_norm": 0.43000625586768587, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1961 + }, + { + "epoch": 0.01962, + "grad_norm": 0.43925709381263844, + "learning_rate": 0.003, + "loss": 4.232, + "step": 1962 + }, + { + "epoch": 0.01963, + "grad_norm": 0.42981173506186865, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1963 + }, + { + "epoch": 0.01964, + "grad_norm": 0.4097996230629943, + "learning_rate": 0.003, + "loss": 4.2465, + "step": 1964 + }, + { + "epoch": 0.01965, + "grad_norm": 0.3678222300164623, + "learning_rate": 0.003, + "loss": 4.2354, + "step": 1965 + }, + { + "epoch": 0.01966, + "grad_norm": 0.3826622250643601, + "learning_rate": 0.003, + "loss": 4.2372, + "step": 1966 + }, + { + "epoch": 0.01967, + "grad_norm": 0.40760983332927614, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 1967 + }, + { + "epoch": 0.01968, + "grad_norm": 0.47242767064200136, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 1968 + }, + { + "epoch": 0.01969, + "grad_norm": 0.588742886193704, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 1969 + }, + { + "epoch": 0.0197, + "grad_norm": 0.7240833755896967, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 1970 + }, + { + "epoch": 0.01971, + "grad_norm": 0.7681447765560055, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1971 + }, + { + "epoch": 0.01972, + "grad_norm": 0.6596049230522953, + "learning_rate": 0.003, + "loss": 4.2329, + "step": 1972 + }, + { + "epoch": 0.01973, + "grad_norm": 0.5971490272128623, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 1973 + }, + { + "epoch": 0.01974, + "grad_norm": 0.5788210723943882, + "learning_rate": 0.003, + "loss": 4.2527, + "step": 1974 + }, + { + "epoch": 0.01975, + "grad_norm": 0.5716517631988168, + "learning_rate": 0.003, + "loss": 4.2434, + "step": 1975 + }, + { + "epoch": 0.01976, + "grad_norm": 0.5714647695261786, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 1976 + }, + { + "epoch": 0.01977, + "grad_norm": 0.4672772924879757, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 1977 + }, + { + "epoch": 0.01978, + "grad_norm": 0.4919465044055173, + "learning_rate": 0.003, + "loss": 4.2624, + "step": 1978 + }, + { + "epoch": 0.01979, + "grad_norm": 0.5361240818371313, + "learning_rate": 0.003, + "loss": 4.25, + "step": 1979 + }, + { + "epoch": 0.0198, + "grad_norm": 0.6323418711078611, + "learning_rate": 0.003, + "loss": 4.2334, + "step": 1980 + }, + { + "epoch": 0.01981, + "grad_norm": 0.6568538387899918, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 1981 + }, + { + "epoch": 0.01982, + "grad_norm": 0.5202437330212393, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 1982 + }, + { + "epoch": 0.01983, + "grad_norm": 0.50161503607477, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 1983 + }, + { + "epoch": 0.01984, + "grad_norm": 0.6088039267450153, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 1984 + }, + { + "epoch": 0.01985, + "grad_norm": 0.8039703319398841, + "learning_rate": 0.003, + "loss": 4.247, + "step": 1985 + }, + { + "epoch": 0.01986, + "grad_norm": 1.0279100079228762, + "learning_rate": 0.003, + "loss": 4.2896, + "step": 1986 + }, + { + "epoch": 0.01987, + "grad_norm": 0.8414880380409095, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 1987 + }, + { + "epoch": 0.01988, + "grad_norm": 0.8182097035427517, + "learning_rate": 0.003, + "loss": 4.2565, + "step": 1988 + }, + { + "epoch": 0.01989, + "grad_norm": 0.7584478804000142, + "learning_rate": 0.003, + "loss": 4.2488, + "step": 1989 + }, + { + "epoch": 0.0199, + "grad_norm": 0.6298670596846694, + "learning_rate": 0.003, + "loss": 4.2542, + "step": 1990 + }, + { + "epoch": 0.01991, + "grad_norm": 0.5918495849768016, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 1991 + }, + { + "epoch": 0.01992, + "grad_norm": 0.6341092964743213, + "learning_rate": 0.003, + "loss": 4.274, + "step": 1992 + }, + { + "epoch": 0.01993, + "grad_norm": 0.690814419492087, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 1993 + }, + { + "epoch": 0.01994, + "grad_norm": 0.738327319887457, + "learning_rate": 0.003, + "loss": 4.2804, + "step": 1994 + }, + { + "epoch": 0.01995, + "grad_norm": 0.9314757846548688, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1995 + }, + { + "epoch": 0.01996, + "grad_norm": 0.9849485592924706, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 1996 + }, + { + "epoch": 0.01997, + "grad_norm": 0.8581365774503161, + "learning_rate": 0.003, + "loss": 4.2889, + "step": 1997 + }, + { + "epoch": 0.01998, + "grad_norm": 0.824906277718389, + "learning_rate": 0.003, + "loss": 4.2717, + "step": 1998 + }, + { + "epoch": 0.01999, + "grad_norm": 0.8642378679098298, + "learning_rate": 0.003, + "loss": 4.274, + "step": 1999 + }, + { + "epoch": 0.02, + "grad_norm": 0.7257405855967545, + "learning_rate": 0.003, + "loss": 4.25, + "step": 2000 + }, + { + "epoch": 0.02001, + "grad_norm": 0.7164159344618267, + "learning_rate": 0.003, + "loss": 4.2892, + "step": 2001 + }, + { + "epoch": 0.02002, + "grad_norm": 0.7067957991027757, + "learning_rate": 0.003, + "loss": 4.2811, + "step": 2002 + }, + { + "epoch": 0.02003, + "grad_norm": 0.7144493226672485, + "learning_rate": 0.003, + "loss": 4.2802, + "step": 2003 + }, + { + "epoch": 0.02004, + "grad_norm": 0.8392479210866416, + "learning_rate": 0.003, + "loss": 4.2749, + "step": 2004 + }, + { + "epoch": 0.02005, + "grad_norm": 0.7896031315305061, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 2005 + }, + { + "epoch": 0.02006, + "grad_norm": 0.7812166363163485, + "learning_rate": 0.003, + "loss": 4.2791, + "step": 2006 + }, + { + "epoch": 0.02007, + "grad_norm": 0.720616320355632, + "learning_rate": 0.003, + "loss": 4.2906, + "step": 2007 + }, + { + "epoch": 0.02008, + "grad_norm": 0.7275037377798674, + "learning_rate": 0.003, + "loss": 4.2831, + "step": 2008 + }, + { + "epoch": 0.02009, + "grad_norm": 0.8040540578216816, + "learning_rate": 0.003, + "loss": 4.2757, + "step": 2009 + }, + { + "epoch": 0.0201, + "grad_norm": 0.8028591997627386, + "learning_rate": 0.003, + "loss": 4.2928, + "step": 2010 + }, + { + "epoch": 0.02011, + "grad_norm": 0.7203630829267829, + "learning_rate": 0.003, + "loss": 4.2738, + "step": 2011 + }, + { + "epoch": 0.02012, + "grad_norm": 0.6305312780193182, + "learning_rate": 0.003, + "loss": 4.2707, + "step": 2012 + }, + { + "epoch": 0.02013, + "grad_norm": 0.4989290466950007, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 2013 + }, + { + "epoch": 0.02014, + "grad_norm": 0.5235253503408636, + "learning_rate": 0.003, + "loss": 4.2549, + "step": 2014 + }, + { + "epoch": 0.02015, + "grad_norm": 0.49148432915624135, + "learning_rate": 0.003, + "loss": 4.2456, + "step": 2015 + }, + { + "epoch": 0.02016, + "grad_norm": 0.5171965972543717, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 2016 + }, + { + "epoch": 0.02017, + "grad_norm": 0.48975410612827497, + "learning_rate": 0.003, + "loss": 4.2428, + "step": 2017 + }, + { + "epoch": 0.02018, + "grad_norm": 0.532740553989744, + "learning_rate": 0.003, + "loss": 4.2458, + "step": 2018 + }, + { + "epoch": 0.02019, + "grad_norm": 0.48750173027142363, + "learning_rate": 0.003, + "loss": 4.2504, + "step": 2019 + }, + { + "epoch": 0.0202, + "grad_norm": 0.4089724033791691, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 2020 + }, + { + "epoch": 0.02021, + "grad_norm": 0.4207487504513156, + "learning_rate": 0.003, + "loss": 4.265, + "step": 2021 + }, + { + "epoch": 0.02022, + "grad_norm": 0.4057461757528402, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 2022 + }, + { + "epoch": 0.02023, + "grad_norm": 0.3580229208354821, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2023 + }, + { + "epoch": 0.02024, + "grad_norm": 0.3567498277820771, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2024 + }, + { + "epoch": 0.02025, + "grad_norm": 0.43295354825834576, + "learning_rate": 0.003, + "loss": 4.2663, + "step": 2025 + }, + { + "epoch": 0.02026, + "grad_norm": 0.4300502575103401, + "learning_rate": 0.003, + "loss": 4.2452, + "step": 2026 + }, + { + "epoch": 0.02027, + "grad_norm": 0.5507801143322275, + "learning_rate": 0.003, + "loss": 4.2429, + "step": 2027 + }, + { + "epoch": 0.02028, + "grad_norm": 0.6603955599139062, + "learning_rate": 0.003, + "loss": 4.2319, + "step": 2028 + }, + { + "epoch": 0.02029, + "grad_norm": 0.7462999560296353, + "learning_rate": 0.003, + "loss": 4.2391, + "step": 2029 + }, + { + "epoch": 0.0203, + "grad_norm": 0.6312368220072189, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 2030 + }, + { + "epoch": 0.02031, + "grad_norm": 0.5002880523213, + "learning_rate": 0.003, + "loss": 4.2308, + "step": 2031 + }, + { + "epoch": 0.02032, + "grad_norm": 0.7101695088498386, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 2032 + }, + { + "epoch": 0.02033, + "grad_norm": 0.7026230897172011, + "learning_rate": 0.003, + "loss": 4.2596, + "step": 2033 + }, + { + "epoch": 0.02034, + "grad_norm": 0.4672671459141678, + "learning_rate": 0.003, + "loss": 4.2396, + "step": 2034 + }, + { + "epoch": 0.02035, + "grad_norm": 0.5167562590378834, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2035 + }, + { + "epoch": 0.02036, + "grad_norm": 0.5204324719963792, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 2036 + }, + { + "epoch": 0.02037, + "grad_norm": 0.44171452211516693, + "learning_rate": 0.003, + "loss": 4.2379, + "step": 2037 + }, + { + "epoch": 0.02038, + "grad_norm": 0.443929190404615, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 2038 + }, + { + "epoch": 0.02039, + "grad_norm": 0.48540831581824334, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 2039 + }, + { + "epoch": 0.0204, + "grad_norm": 0.5410457621233177, + "learning_rate": 0.003, + "loss": 4.236, + "step": 2040 + }, + { + "epoch": 0.02041, + "grad_norm": 0.5544368623280378, + "learning_rate": 0.003, + "loss": 4.1991, + "step": 2041 + }, + { + "epoch": 0.02042, + "grad_norm": 0.5894517852974022, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 2042 + }, + { + "epoch": 0.02043, + "grad_norm": 0.6318863235101739, + "learning_rate": 0.003, + "loss": 4.2488, + "step": 2043 + }, + { + "epoch": 0.02044, + "grad_norm": 0.6044374925441985, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 2044 + }, + { + "epoch": 0.02045, + "grad_norm": 0.5638648488316749, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 2045 + }, + { + "epoch": 0.02046, + "grad_norm": 0.514843677649273, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2046 + }, + { + "epoch": 0.02047, + "grad_norm": 0.4850628116632027, + "learning_rate": 0.003, + "loss": 4.2466, + "step": 2047 + }, + { + "epoch": 0.02048, + "grad_norm": 0.5067500543661038, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 2048 + }, + { + "epoch": 0.02049, + "grad_norm": 0.5005632559023494, + "learning_rate": 0.003, + "loss": 4.2517, + "step": 2049 + }, + { + "epoch": 0.0205, + "grad_norm": 0.5338842758485439, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 2050 + }, + { + "epoch": 0.02051, + "grad_norm": 0.5673366342029961, + "learning_rate": 0.003, + "loss": 4.2056, + "step": 2051 + }, + { + "epoch": 0.02052, + "grad_norm": 0.6208513233991159, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 2052 + }, + { + "epoch": 0.02053, + "grad_norm": 0.7925288751630323, + "learning_rate": 0.003, + "loss": 4.2162, + "step": 2053 + }, + { + "epoch": 0.02054, + "grad_norm": 0.9689478763760375, + "learning_rate": 0.003, + "loss": 4.2703, + "step": 2054 + }, + { + "epoch": 0.02055, + "grad_norm": 0.9174532624746199, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 2055 + }, + { + "epoch": 0.02056, + "grad_norm": 0.7759474219779441, + "learning_rate": 0.003, + "loss": 4.2289, + "step": 2056 + }, + { + "epoch": 0.02057, + "grad_norm": 0.6855719086766776, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 2057 + }, + { + "epoch": 0.02058, + "grad_norm": 0.664344253711328, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 2058 + }, + { + "epoch": 0.02059, + "grad_norm": 0.7126024175545171, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 2059 + }, + { + "epoch": 0.0206, + "grad_norm": 0.8304297596763678, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 2060 + }, + { + "epoch": 0.02061, + "grad_norm": 0.8055338565488124, + "learning_rate": 0.003, + "loss": 4.2502, + "step": 2061 + }, + { + "epoch": 0.02062, + "grad_norm": 0.7656045427907143, + "learning_rate": 0.003, + "loss": 4.2474, + "step": 2062 + }, + { + "epoch": 0.02063, + "grad_norm": 0.7423249019964357, + "learning_rate": 0.003, + "loss": 4.2458, + "step": 2063 + }, + { + "epoch": 0.02064, + "grad_norm": 0.7236715566519402, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 2064 + }, + { + "epoch": 0.02065, + "grad_norm": 0.7531415944255598, + "learning_rate": 0.003, + "loss": 4.2369, + "step": 2065 + }, + { + "epoch": 0.02066, + "grad_norm": 0.6867969137582566, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 2066 + }, + { + "epoch": 0.02067, + "grad_norm": 0.5938985343647097, + "learning_rate": 0.003, + "loss": 4.245, + "step": 2067 + }, + { + "epoch": 0.02068, + "grad_norm": 0.4863087277096077, + "learning_rate": 0.003, + "loss": 4.2209, + "step": 2068 + }, + { + "epoch": 0.02069, + "grad_norm": 0.4635605849081945, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 2069 + }, + { + "epoch": 0.0207, + "grad_norm": 0.5457323036329618, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 2070 + }, + { + "epoch": 0.02071, + "grad_norm": 0.5432685692905308, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 2071 + }, + { + "epoch": 0.02072, + "grad_norm": 0.5685190700350947, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 2072 + }, + { + "epoch": 0.02073, + "grad_norm": 0.535137018993922, + "learning_rate": 0.003, + "loss": 4.2291, + "step": 2073 + }, + { + "epoch": 0.02074, + "grad_norm": 0.5467639426283901, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 2074 + }, + { + "epoch": 0.02075, + "grad_norm": 0.5451184946505037, + "learning_rate": 0.003, + "loss": 4.2644, + "step": 2075 + }, + { + "epoch": 0.02076, + "grad_norm": 0.4600936162495081, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 2076 + }, + { + "epoch": 0.02077, + "grad_norm": 0.465650503497137, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2077 + }, + { + "epoch": 0.02078, + "grad_norm": 0.5411332311026555, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 2078 + }, + { + "epoch": 0.02079, + "grad_norm": 0.5765467214387223, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 2079 + }, + { + "epoch": 0.0208, + "grad_norm": 0.5738914259466256, + "learning_rate": 0.003, + "loss": 4.2137, + "step": 2080 + }, + { + "epoch": 0.02081, + "grad_norm": 0.5810191862607845, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 2081 + }, + { + "epoch": 0.02082, + "grad_norm": 0.5527464904961061, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 2082 + }, + { + "epoch": 0.02083, + "grad_norm": 0.5331974469414995, + "learning_rate": 0.003, + "loss": 4.2351, + "step": 2083 + }, + { + "epoch": 0.02084, + "grad_norm": 0.5880130787094308, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 2084 + }, + { + "epoch": 0.02085, + "grad_norm": 0.6630695440600266, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 2085 + }, + { + "epoch": 0.02086, + "grad_norm": 0.6095683840332357, + "learning_rate": 0.003, + "loss": 4.2148, + "step": 2086 + }, + { + "epoch": 0.02087, + "grad_norm": 0.6289181823836659, + "learning_rate": 0.003, + "loss": 4.214, + "step": 2087 + }, + { + "epoch": 0.02088, + "grad_norm": 0.6691073473732546, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 2088 + }, + { + "epoch": 0.02089, + "grad_norm": 0.757943428461077, + "learning_rate": 0.003, + "loss": 4.2252, + "step": 2089 + }, + { + "epoch": 0.0209, + "grad_norm": 0.6942072774820578, + "learning_rate": 0.003, + "loss": 4.2464, + "step": 2090 + }, + { + "epoch": 0.02091, + "grad_norm": 0.5265412592049787, + "learning_rate": 0.003, + "loss": 4.224, + "step": 2091 + }, + { + "epoch": 0.02092, + "grad_norm": 0.6385734634683451, + "learning_rate": 0.003, + "loss": 4.2259, + "step": 2092 + }, + { + "epoch": 0.02093, + "grad_norm": 0.7053440248117775, + "learning_rate": 0.003, + "loss": 4.1968, + "step": 2093 + }, + { + "epoch": 0.02094, + "grad_norm": 0.696978044878893, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2094 + }, + { + "epoch": 0.02095, + "grad_norm": 0.6272978652521859, + "learning_rate": 0.003, + "loss": 4.2496, + "step": 2095 + }, + { + "epoch": 0.02096, + "grad_norm": 0.643764393951801, + "learning_rate": 0.003, + "loss": 4.2395, + "step": 2096 + }, + { + "epoch": 0.02097, + "grad_norm": 0.6586650506136257, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 2097 + }, + { + "epoch": 0.02098, + "grad_norm": 0.5374026111379894, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 2098 + }, + { + "epoch": 0.02099, + "grad_norm": 0.463394138236571, + "learning_rate": 0.003, + "loss": 4.2426, + "step": 2099 + }, + { + "epoch": 0.021, + "grad_norm": 0.43247948811237297, + "learning_rate": 0.003, + "loss": 4.25, + "step": 2100 + }, + { + "epoch": 0.02101, + "grad_norm": 0.4029126824700864, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 2101 + }, + { + "epoch": 0.02102, + "grad_norm": 0.4076539308302002, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 2102 + }, + { + "epoch": 0.02103, + "grad_norm": 0.48572239313264476, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2103 + }, + { + "epoch": 0.02104, + "grad_norm": 0.6011024498535701, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2104 + }, + { + "epoch": 0.02105, + "grad_norm": 0.7452127300170479, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 2105 + }, + { + "epoch": 0.02106, + "grad_norm": 0.8682450979969631, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 2106 + }, + { + "epoch": 0.02107, + "grad_norm": 0.8835439612331999, + "learning_rate": 0.003, + "loss": 4.2496, + "step": 2107 + }, + { + "epoch": 0.02108, + "grad_norm": 0.7854943005617361, + "learning_rate": 0.003, + "loss": 4.226, + "step": 2108 + }, + { + "epoch": 0.02109, + "grad_norm": 0.7880515723278634, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 2109 + }, + { + "epoch": 0.0211, + "grad_norm": 0.7374506931044654, + "learning_rate": 0.003, + "loss": 4.234, + "step": 2110 + }, + { + "epoch": 0.02111, + "grad_norm": 0.8090504896236194, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 2111 + }, + { + "epoch": 0.02112, + "grad_norm": 0.7918025257999607, + "learning_rate": 0.003, + "loss": 4.2488, + "step": 2112 + }, + { + "epoch": 0.02113, + "grad_norm": 0.8131758042516601, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 2113 + }, + { + "epoch": 0.02114, + "grad_norm": 0.7527860897593042, + "learning_rate": 0.003, + "loss": 4.2365, + "step": 2114 + }, + { + "epoch": 0.02115, + "grad_norm": 0.7359409139929027, + "learning_rate": 0.003, + "loss": 4.231, + "step": 2115 + }, + { + "epoch": 0.02116, + "grad_norm": 0.6808675729987482, + "learning_rate": 0.003, + "loss": 4.2544, + "step": 2116 + }, + { + "epoch": 0.02117, + "grad_norm": 0.5755888828726318, + "learning_rate": 0.003, + "loss": 4.2431, + "step": 2117 + }, + { + "epoch": 0.02118, + "grad_norm": 0.5921868695179046, + "learning_rate": 0.003, + "loss": 4.2397, + "step": 2118 + }, + { + "epoch": 0.02119, + "grad_norm": 0.5455603705053529, + "learning_rate": 0.003, + "loss": 4.2459, + "step": 2119 + }, + { + "epoch": 0.0212, + "grad_norm": 0.627867188843003, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 2120 + }, + { + "epoch": 0.02121, + "grad_norm": 0.611945450027123, + "learning_rate": 0.003, + "loss": 4.2171, + "step": 2121 + }, + { + "epoch": 0.02122, + "grad_norm": 0.568436827457235, + "learning_rate": 0.003, + "loss": 4.226, + "step": 2122 + }, + { + "epoch": 0.02123, + "grad_norm": 0.5350506674860982, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2123 + }, + { + "epoch": 0.02124, + "grad_norm": 0.5325059896638067, + "learning_rate": 0.003, + "loss": 4.2254, + "step": 2124 + }, + { + "epoch": 0.02125, + "grad_norm": 0.55145580695339, + "learning_rate": 0.003, + "loss": 4.2384, + "step": 2125 + }, + { + "epoch": 0.02126, + "grad_norm": 0.6190295521827949, + "learning_rate": 0.003, + "loss": 4.2377, + "step": 2126 + }, + { + "epoch": 0.02127, + "grad_norm": 0.5815814384311068, + "learning_rate": 0.003, + "loss": 4.244, + "step": 2127 + }, + { + "epoch": 0.02128, + "grad_norm": 0.624207657173836, + "learning_rate": 0.003, + "loss": 4.2362, + "step": 2128 + }, + { + "epoch": 0.02129, + "grad_norm": 0.637928608874603, + "learning_rate": 0.003, + "loss": 4.2364, + "step": 2129 + }, + { + "epoch": 0.0213, + "grad_norm": 0.5372631019114236, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 2130 + }, + { + "epoch": 0.02131, + "grad_norm": 0.5117120142897393, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2131 + }, + { + "epoch": 0.02132, + "grad_norm": 0.5272503917214815, + "learning_rate": 0.003, + "loss": 4.2429, + "step": 2132 + }, + { + "epoch": 0.02133, + "grad_norm": 0.4922684782001162, + "learning_rate": 0.003, + "loss": 4.2337, + "step": 2133 + }, + { + "epoch": 0.02134, + "grad_norm": 0.4711594178288037, + "learning_rate": 0.003, + "loss": 4.233, + "step": 2134 + }, + { + "epoch": 0.02135, + "grad_norm": 0.5354798117430423, + "learning_rate": 0.003, + "loss": 4.2327, + "step": 2135 + }, + { + "epoch": 0.02136, + "grad_norm": 0.5972846763679789, + "learning_rate": 0.003, + "loss": 4.2059, + "step": 2136 + }, + { + "epoch": 0.02137, + "grad_norm": 0.6875217013113815, + "learning_rate": 0.003, + "loss": 4.2468, + "step": 2137 + }, + { + "epoch": 0.02138, + "grad_norm": 0.9636076155805002, + "learning_rate": 0.003, + "loss": 4.273, + "step": 2138 + }, + { + "epoch": 0.02139, + "grad_norm": 1.148302508200364, + "learning_rate": 0.003, + "loss": 4.2769, + "step": 2139 + }, + { + "epoch": 0.0214, + "grad_norm": 0.7025550992151242, + "learning_rate": 0.003, + "loss": 4.2515, + "step": 2140 + }, + { + "epoch": 0.02141, + "grad_norm": 0.7381063691718428, + "learning_rate": 0.003, + "loss": 4.2408, + "step": 2141 + }, + { + "epoch": 0.02142, + "grad_norm": 0.6164956740364308, + "learning_rate": 0.003, + "loss": 4.2223, + "step": 2142 + }, + { + "epoch": 0.02143, + "grad_norm": 0.6846147414300063, + "learning_rate": 0.003, + "loss": 4.2414, + "step": 2143 + }, + { + "epoch": 0.02144, + "grad_norm": 0.7043565504624957, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 2144 + }, + { + "epoch": 0.02145, + "grad_norm": 0.6306040325927637, + "learning_rate": 0.003, + "loss": 4.2262, + "step": 2145 + }, + { + "epoch": 0.02146, + "grad_norm": 0.5782503991731893, + "learning_rate": 0.003, + "loss": 4.2372, + "step": 2146 + }, + { + "epoch": 0.02147, + "grad_norm": 0.5354606170184357, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 2147 + }, + { + "epoch": 0.02148, + "grad_norm": 0.4707102689450845, + "learning_rate": 0.003, + "loss": 4.2629, + "step": 2148 + }, + { + "epoch": 0.02149, + "grad_norm": 0.5098358773254507, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 2149 + }, + { + "epoch": 0.0215, + "grad_norm": 0.5420555838700287, + "learning_rate": 0.003, + "loss": 4.2354, + "step": 2150 + }, + { + "epoch": 0.02151, + "grad_norm": 0.5090710234825387, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 2151 + }, + { + "epoch": 0.02152, + "grad_norm": 0.5210838533104809, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 2152 + }, + { + "epoch": 0.02153, + "grad_norm": 0.4582952267704145, + "learning_rate": 0.003, + "loss": 4.2212, + "step": 2153 + }, + { + "epoch": 0.02154, + "grad_norm": 0.5288143265950994, + "learning_rate": 0.003, + "loss": 4.214, + "step": 2154 + }, + { + "epoch": 0.02155, + "grad_norm": 0.568233364630795, + "learning_rate": 0.003, + "loss": 4.2465, + "step": 2155 + }, + { + "epoch": 0.02156, + "grad_norm": 0.6709041798606031, + "learning_rate": 0.003, + "loss": 4.2356, + "step": 2156 + }, + { + "epoch": 0.02157, + "grad_norm": 0.837609334642183, + "learning_rate": 0.003, + "loss": 4.2105, + "step": 2157 + }, + { + "epoch": 0.02158, + "grad_norm": 0.9025483102003374, + "learning_rate": 0.003, + "loss": 4.2406, + "step": 2158 + }, + { + "epoch": 0.02159, + "grad_norm": 0.6757453813747746, + "learning_rate": 0.003, + "loss": 4.2475, + "step": 2159 + }, + { + "epoch": 0.0216, + "grad_norm": 0.6195925236969745, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 2160 + }, + { + "epoch": 0.02161, + "grad_norm": 0.6541705747885649, + "learning_rate": 0.003, + "loss": 4.2346, + "step": 2161 + }, + { + "epoch": 0.02162, + "grad_norm": 0.713506704868693, + "learning_rate": 0.003, + "loss": 4.2445, + "step": 2162 + }, + { + "epoch": 0.02163, + "grad_norm": 0.7618153376621027, + "learning_rate": 0.003, + "loss": 4.2619, + "step": 2163 + }, + { + "epoch": 0.02164, + "grad_norm": 0.7193438764197988, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 2164 + }, + { + "epoch": 0.02165, + "grad_norm": 0.7614872081882262, + "learning_rate": 0.003, + "loss": 4.2185, + "step": 2165 + }, + { + "epoch": 0.02166, + "grad_norm": 0.7583750488187075, + "learning_rate": 0.003, + "loss": 4.2616, + "step": 2166 + }, + { + "epoch": 0.02167, + "grad_norm": 0.5886983647176827, + "learning_rate": 0.003, + "loss": 4.2327, + "step": 2167 + }, + { + "epoch": 0.02168, + "grad_norm": 0.5673249719944714, + "learning_rate": 0.003, + "loss": 4.229, + "step": 2168 + }, + { + "epoch": 0.02169, + "grad_norm": 0.539350720737554, + "learning_rate": 0.003, + "loss": 4.2253, + "step": 2169 + }, + { + "epoch": 0.0217, + "grad_norm": 0.505175388578022, + "learning_rate": 0.003, + "loss": 4.2316, + "step": 2170 + }, + { + "epoch": 0.02171, + "grad_norm": 0.523933575801358, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 2171 + }, + { + "epoch": 0.02172, + "grad_norm": 0.4755293785333703, + "learning_rate": 0.003, + "loss": 4.2128, + "step": 2172 + }, + { + "epoch": 0.02173, + "grad_norm": 0.5360790560792152, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 2173 + }, + { + "epoch": 0.02174, + "grad_norm": 0.5613507175665455, + "learning_rate": 0.003, + "loss": 4.216, + "step": 2174 + }, + { + "epoch": 0.02175, + "grad_norm": 0.7273309266731635, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 2175 + }, + { + "epoch": 0.02176, + "grad_norm": 0.8055224057820037, + "learning_rate": 0.003, + "loss": 4.2357, + "step": 2176 + }, + { + "epoch": 0.02177, + "grad_norm": 0.7934658981354097, + "learning_rate": 0.003, + "loss": 4.2405, + "step": 2177 + }, + { + "epoch": 0.02178, + "grad_norm": 0.680464890242631, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 2178 + }, + { + "epoch": 0.02179, + "grad_norm": 0.6249756747331623, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 2179 + }, + { + "epoch": 0.0218, + "grad_norm": 0.6384620261389609, + "learning_rate": 0.003, + "loss": 4.245, + "step": 2180 + }, + { + "epoch": 0.02181, + "grad_norm": 0.6338413006913513, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 2181 + }, + { + "epoch": 0.02182, + "grad_norm": 0.602201817766222, + "learning_rate": 0.003, + "loss": 4.2311, + "step": 2182 + }, + { + "epoch": 0.02183, + "grad_norm": 0.6159529942047347, + "learning_rate": 0.003, + "loss": 4.2415, + "step": 2183 + }, + { + "epoch": 0.02184, + "grad_norm": 0.654749250757439, + "learning_rate": 0.003, + "loss": 4.2162, + "step": 2184 + }, + { + "epoch": 0.02185, + "grad_norm": 0.798418167230075, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 2185 + }, + { + "epoch": 0.02186, + "grad_norm": 0.9118929371333987, + "learning_rate": 0.003, + "loss": 4.2486, + "step": 2186 + }, + { + "epoch": 0.02187, + "grad_norm": 0.8282661348302262, + "learning_rate": 0.003, + "loss": 4.2454, + "step": 2187 + }, + { + "epoch": 0.02188, + "grad_norm": 0.6832425004401769, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 2188 + }, + { + "epoch": 0.02189, + "grad_norm": 0.6446063652482951, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 2189 + }, + { + "epoch": 0.0219, + "grad_norm": 0.5887684645304796, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 2190 + }, + { + "epoch": 0.02191, + "grad_norm": 0.5977971628739048, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 2191 + }, + { + "epoch": 0.02192, + "grad_norm": 0.4949212999134859, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 2192 + }, + { + "epoch": 0.02193, + "grad_norm": 0.5003189031862624, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 2193 + }, + { + "epoch": 0.02194, + "grad_norm": 0.48226309241618065, + "learning_rate": 0.003, + "loss": 4.2678, + "step": 2194 + }, + { + "epoch": 0.02195, + "grad_norm": 0.544272590519486, + "learning_rate": 0.003, + "loss": 4.2332, + "step": 2195 + }, + { + "epoch": 0.02196, + "grad_norm": 0.6071662200973025, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 2196 + }, + { + "epoch": 0.02197, + "grad_norm": 0.7366010094026804, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 2197 + }, + { + "epoch": 0.02198, + "grad_norm": 0.7102801228364903, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 2198 + }, + { + "epoch": 0.02199, + "grad_norm": 0.6007866089006326, + "learning_rate": 0.003, + "loss": 4.194, + "step": 2199 + }, + { + "epoch": 0.022, + "grad_norm": 0.5670114788301772, + "learning_rate": 0.003, + "loss": 4.2194, + "step": 2200 + }, + { + "epoch": 0.02201, + "grad_norm": 0.4763785124586338, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 2201 + }, + { + "epoch": 0.02202, + "grad_norm": 0.42903876892090725, + "learning_rate": 0.003, + "loss": 4.2163, + "step": 2202 + }, + { + "epoch": 0.02203, + "grad_norm": 0.4156329769757173, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 2203 + }, + { + "epoch": 0.02204, + "grad_norm": 0.4681197264286167, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 2204 + }, + { + "epoch": 0.02205, + "grad_norm": 0.5059629089964722, + "learning_rate": 0.003, + "loss": 4.2388, + "step": 2205 + }, + { + "epoch": 0.02206, + "grad_norm": 0.6309485655238869, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 2206 + }, + { + "epoch": 0.02207, + "grad_norm": 0.6824022074223641, + "learning_rate": 0.003, + "loss": 4.2243, + "step": 2207 + }, + { + "epoch": 0.02208, + "grad_norm": 0.6950242305361404, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 2208 + }, + { + "epoch": 0.02209, + "grad_norm": 0.8264786459594594, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 2209 + }, + { + "epoch": 0.0221, + "grad_norm": 0.8486387875780823, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 2210 + }, + { + "epoch": 0.02211, + "grad_norm": 0.7761571300506404, + "learning_rate": 0.003, + "loss": 4.2269, + "step": 2211 + }, + { + "epoch": 0.02212, + "grad_norm": 0.6379539901749441, + "learning_rate": 0.003, + "loss": 4.2233, + "step": 2212 + }, + { + "epoch": 0.02213, + "grad_norm": 0.6089238962453585, + "learning_rate": 0.003, + "loss": 4.2296, + "step": 2213 + }, + { + "epoch": 0.02214, + "grad_norm": 0.5796568814987876, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 2214 + }, + { + "epoch": 0.02215, + "grad_norm": 0.5660272621079528, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 2215 + }, + { + "epoch": 0.02216, + "grad_norm": 0.5885926435515716, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 2216 + }, + { + "epoch": 0.02217, + "grad_norm": 0.5475660813804676, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 2217 + }, + { + "epoch": 0.02218, + "grad_norm": 0.5260715025238017, + "learning_rate": 0.003, + "loss": 4.2304, + "step": 2218 + }, + { + "epoch": 0.02219, + "grad_norm": 0.5716509549906074, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 2219 + }, + { + "epoch": 0.0222, + "grad_norm": 0.5917539957492576, + "learning_rate": 0.003, + "loss": 4.2195, + "step": 2220 + }, + { + "epoch": 0.02221, + "grad_norm": 0.6044205892693139, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 2221 + }, + { + "epoch": 0.02222, + "grad_norm": 0.567760157124041, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 2222 + }, + { + "epoch": 0.02223, + "grad_norm": 0.577201891352313, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 2223 + }, + { + "epoch": 0.02224, + "grad_norm": 0.5589539299230697, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 2224 + }, + { + "epoch": 0.02225, + "grad_norm": 0.578872867820463, + "learning_rate": 0.003, + "loss": 4.206, + "step": 2225 + }, + { + "epoch": 0.02226, + "grad_norm": 0.6018889202924489, + "learning_rate": 0.003, + "loss": 4.2223, + "step": 2226 + }, + { + "epoch": 0.02227, + "grad_norm": 0.6659111341486127, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 2227 + }, + { + "epoch": 0.02228, + "grad_norm": 0.6307543898445819, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 2228 + }, + { + "epoch": 0.02229, + "grad_norm": 0.5993087464100946, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2229 + }, + { + "epoch": 0.0223, + "grad_norm": 0.6286376964291409, + "learning_rate": 0.003, + "loss": 4.226, + "step": 2230 + }, + { + "epoch": 0.02231, + "grad_norm": 0.872328678246551, + "learning_rate": 0.003, + "loss": 4.2394, + "step": 2231 + }, + { + "epoch": 0.02232, + "grad_norm": 1.0461803323605487, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 2232 + }, + { + "epoch": 0.02233, + "grad_norm": 0.9329588220097066, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 2233 + }, + { + "epoch": 0.02234, + "grad_norm": 0.6760145633983702, + "learning_rate": 0.003, + "loss": 4.2457, + "step": 2234 + }, + { + "epoch": 0.02235, + "grad_norm": 0.7832849984021282, + "learning_rate": 0.003, + "loss": 4.281, + "step": 2235 + }, + { + "epoch": 0.02236, + "grad_norm": 0.6196461184379554, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 2236 + }, + { + "epoch": 0.02237, + "grad_norm": 0.5630794415245316, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 2237 + }, + { + "epoch": 0.02238, + "grad_norm": 0.4338854754415404, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2238 + }, + { + "epoch": 0.02239, + "grad_norm": 0.43304313838731295, + "learning_rate": 0.003, + "loss": 4.2118, + "step": 2239 + }, + { + "epoch": 0.0224, + "grad_norm": 0.3849146263035166, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 2240 + }, + { + "epoch": 0.02241, + "grad_norm": 0.3700093297043444, + "learning_rate": 0.003, + "loss": 4.221, + "step": 2241 + }, + { + "epoch": 0.02242, + "grad_norm": 0.3483834690743904, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 2242 + }, + { + "epoch": 0.02243, + "grad_norm": 0.38069058438024445, + "learning_rate": 0.003, + "loss": 4.2335, + "step": 2243 + }, + { + "epoch": 0.02244, + "grad_norm": 0.4503598461041079, + "learning_rate": 0.003, + "loss": 4.21, + "step": 2244 + }, + { + "epoch": 0.02245, + "grad_norm": 0.5173040984777282, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2245 + }, + { + "epoch": 0.02246, + "grad_norm": 0.5743673128739117, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 2246 + }, + { + "epoch": 0.02247, + "grad_norm": 0.6564841670845484, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 2247 + }, + { + "epoch": 0.02248, + "grad_norm": 0.7271803573068377, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 2248 + }, + { + "epoch": 0.02249, + "grad_norm": 0.6357921170928627, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2249 + }, + { + "epoch": 0.0225, + "grad_norm": 0.5333359806181992, + "learning_rate": 0.003, + "loss": 4.2185, + "step": 2250 + }, + { + "epoch": 0.02251, + "grad_norm": 0.5481515163618663, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 2251 + }, + { + "epoch": 0.02252, + "grad_norm": 0.5849718573683582, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 2252 + }, + { + "epoch": 0.02253, + "grad_norm": 0.692221940613648, + "learning_rate": 0.003, + "loss": 4.2188, + "step": 2253 + }, + { + "epoch": 0.02254, + "grad_norm": 0.8350743358894537, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 2254 + }, + { + "epoch": 0.02255, + "grad_norm": 0.9700169658063796, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2255 + }, + { + "epoch": 0.02256, + "grad_norm": 1.1849964912746291, + "learning_rate": 0.003, + "loss": 4.2446, + "step": 2256 + }, + { + "epoch": 0.02257, + "grad_norm": 0.8019987212669306, + "learning_rate": 0.003, + "loss": 4.2432, + "step": 2257 + }, + { + "epoch": 0.02258, + "grad_norm": 0.7178719029324007, + "learning_rate": 0.003, + "loss": 4.2513, + "step": 2258 + }, + { + "epoch": 0.02259, + "grad_norm": 0.8182127992878601, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 2259 + }, + { + "epoch": 0.0226, + "grad_norm": 0.7294977278785716, + "learning_rate": 0.003, + "loss": 4.2695, + "step": 2260 + }, + { + "epoch": 0.02261, + "grad_norm": 0.6869467420486108, + "learning_rate": 0.003, + "loss": 4.2439, + "step": 2261 + }, + { + "epoch": 0.02262, + "grad_norm": 0.7258556366490507, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 2262 + }, + { + "epoch": 0.02263, + "grad_norm": 0.6949107967496511, + "learning_rate": 0.003, + "loss": 4.2554, + "step": 2263 + }, + { + "epoch": 0.02264, + "grad_norm": 0.6840843296451264, + "learning_rate": 0.003, + "loss": 4.2468, + "step": 2264 + }, + { + "epoch": 0.02265, + "grad_norm": 0.6101088207573817, + "learning_rate": 0.003, + "loss": 4.2012, + "step": 2265 + }, + { + "epoch": 0.02266, + "grad_norm": 0.6149463166497737, + "learning_rate": 0.003, + "loss": 4.2182, + "step": 2266 + }, + { + "epoch": 0.02267, + "grad_norm": 0.566032425336374, + "learning_rate": 0.003, + "loss": 4.2427, + "step": 2267 + }, + { + "epoch": 0.02268, + "grad_norm": 0.5712592856836044, + "learning_rate": 0.003, + "loss": 4.2156, + "step": 2268 + }, + { + "epoch": 0.02269, + "grad_norm": 0.5122859378101323, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 2269 + }, + { + "epoch": 0.0227, + "grad_norm": 0.4651490853981594, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 2270 + }, + { + "epoch": 0.02271, + "grad_norm": 0.39776686394339583, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 2271 + }, + { + "epoch": 0.02272, + "grad_norm": 0.4302685557347805, + "learning_rate": 0.003, + "loss": 4.2274, + "step": 2272 + }, + { + "epoch": 0.02273, + "grad_norm": 0.467593917531494, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2273 + }, + { + "epoch": 0.02274, + "grad_norm": 0.632433536625551, + "learning_rate": 0.003, + "loss": 4.2141, + "step": 2274 + }, + { + "epoch": 0.02275, + "grad_norm": 0.8580139662336462, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 2275 + }, + { + "epoch": 0.02276, + "grad_norm": 0.9516059646785548, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2276 + }, + { + "epoch": 0.02277, + "grad_norm": 0.7436294288419117, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 2277 + }, + { + "epoch": 0.02278, + "grad_norm": 0.6418671737236206, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2278 + }, + { + "epoch": 0.02279, + "grad_norm": 0.7656816707126055, + "learning_rate": 0.003, + "loss": 4.2435, + "step": 2279 + }, + { + "epoch": 0.0228, + "grad_norm": 0.6177672920210129, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 2280 + }, + { + "epoch": 0.02281, + "grad_norm": 0.5895210519635931, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 2281 + }, + { + "epoch": 0.02282, + "grad_norm": 0.5868753488307538, + "learning_rate": 0.003, + "loss": 4.2369, + "step": 2282 + }, + { + "epoch": 0.02283, + "grad_norm": 0.6052691283316047, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 2283 + }, + { + "epoch": 0.02284, + "grad_norm": 0.5832141161476754, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 2284 + }, + { + "epoch": 0.02285, + "grad_norm": 0.5796401382132979, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2285 + }, + { + "epoch": 0.02286, + "grad_norm": 0.5270223366823183, + "learning_rate": 0.003, + "loss": 4.2443, + "step": 2286 + }, + { + "epoch": 0.02287, + "grad_norm": 0.4022883813311557, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 2287 + }, + { + "epoch": 0.02288, + "grad_norm": 0.4249187639605795, + "learning_rate": 0.003, + "loss": 4.2409, + "step": 2288 + }, + { + "epoch": 0.02289, + "grad_norm": 0.39026571046663244, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 2289 + }, + { + "epoch": 0.0229, + "grad_norm": 0.4142451551190753, + "learning_rate": 0.003, + "loss": 4.2147, + "step": 2290 + }, + { + "epoch": 0.02291, + "grad_norm": 0.48765765895038643, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 2291 + }, + { + "epoch": 0.02292, + "grad_norm": 0.5671990063966691, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 2292 + }, + { + "epoch": 0.02293, + "grad_norm": 0.5656764518255547, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2293 + }, + { + "epoch": 0.02294, + "grad_norm": 0.5275578505406908, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 2294 + }, + { + "epoch": 0.02295, + "grad_norm": 0.48221866990126555, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 2295 + }, + { + "epoch": 0.02296, + "grad_norm": 0.5188277549634542, + "learning_rate": 0.003, + "loss": 4.225, + "step": 2296 + }, + { + "epoch": 0.02297, + "grad_norm": 0.6417395019786397, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2297 + }, + { + "epoch": 0.02298, + "grad_norm": 0.7094858197213939, + "learning_rate": 0.003, + "loss": 4.2151, + "step": 2298 + }, + { + "epoch": 0.02299, + "grad_norm": 0.7715781893149263, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2299 + }, + { + "epoch": 0.023, + "grad_norm": 0.8183493893838668, + "learning_rate": 0.003, + "loss": 4.2495, + "step": 2300 + }, + { + "epoch": 0.02301, + "grad_norm": 0.7727005595462422, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2301 + }, + { + "epoch": 0.02302, + "grad_norm": 0.789250865027278, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 2302 + }, + { + "epoch": 0.02303, + "grad_norm": 0.8561833032989024, + "learning_rate": 0.003, + "loss": 4.2229, + "step": 2303 + }, + { + "epoch": 0.02304, + "grad_norm": 0.7983476489527453, + "learning_rate": 0.003, + "loss": 4.2526, + "step": 2304 + }, + { + "epoch": 0.02305, + "grad_norm": 0.792691144527778, + "learning_rate": 0.003, + "loss": 4.2307, + "step": 2305 + }, + { + "epoch": 0.02306, + "grad_norm": 0.8333614803540381, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2306 + }, + { + "epoch": 0.02307, + "grad_norm": 0.8709733257410114, + "learning_rate": 0.003, + "loss": 4.248, + "step": 2307 + }, + { + "epoch": 0.02308, + "grad_norm": 0.7840461913162227, + "learning_rate": 0.003, + "loss": 4.2376, + "step": 2308 + }, + { + "epoch": 0.02309, + "grad_norm": 0.7160902424813279, + "learning_rate": 0.003, + "loss": 4.263, + "step": 2309 + }, + { + "epoch": 0.0231, + "grad_norm": 0.7757918757353481, + "learning_rate": 0.003, + "loss": 4.2432, + "step": 2310 + }, + { + "epoch": 0.02311, + "grad_norm": 0.8687865089693275, + "learning_rate": 0.003, + "loss": 4.2573, + "step": 2311 + }, + { + "epoch": 0.02312, + "grad_norm": 0.7846302126542865, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 2312 + }, + { + "epoch": 0.02313, + "grad_norm": 0.7297339955221905, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2313 + }, + { + "epoch": 0.02314, + "grad_norm": 0.6956047853880688, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 2314 + }, + { + "epoch": 0.02315, + "grad_norm": 0.6937368414181051, + "learning_rate": 0.003, + "loss": 4.2413, + "step": 2315 + }, + { + "epoch": 0.02316, + "grad_norm": 0.6709916211621512, + "learning_rate": 0.003, + "loss": 4.2522, + "step": 2316 + }, + { + "epoch": 0.02317, + "grad_norm": 0.6213877076356003, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 2317 + }, + { + "epoch": 0.02318, + "grad_norm": 0.6088693827583501, + "learning_rate": 0.003, + "loss": 4.2184, + "step": 2318 + }, + { + "epoch": 0.02319, + "grad_norm": 0.5989733797969562, + "learning_rate": 0.003, + "loss": 4.241, + "step": 2319 + }, + { + "epoch": 0.0232, + "grad_norm": 0.55615386704494, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 2320 + }, + { + "epoch": 0.02321, + "grad_norm": 0.5520477296154882, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2321 + }, + { + "epoch": 0.02322, + "grad_norm": 0.5223023663129024, + "learning_rate": 0.003, + "loss": 4.2284, + "step": 2322 + }, + { + "epoch": 0.02323, + "grad_norm": 0.5716320780098701, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 2323 + }, + { + "epoch": 0.02324, + "grad_norm": 0.6956733472532048, + "learning_rate": 0.003, + "loss": 4.208, + "step": 2324 + }, + { + "epoch": 0.02325, + "grad_norm": 0.6924168616855679, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 2325 + }, + { + "epoch": 0.02326, + "grad_norm": 0.6921696722241616, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 2326 + }, + { + "epoch": 0.02327, + "grad_norm": 0.6351653773148184, + "learning_rate": 0.003, + "loss": 4.2319, + "step": 2327 + }, + { + "epoch": 0.02328, + "grad_norm": 0.5655372914501792, + "learning_rate": 0.003, + "loss": 4.2261, + "step": 2328 + }, + { + "epoch": 0.02329, + "grad_norm": 0.6342553615471473, + "learning_rate": 0.003, + "loss": 4.2194, + "step": 2329 + }, + { + "epoch": 0.0233, + "grad_norm": 0.7154577148277532, + "learning_rate": 0.003, + "loss": 4.2152, + "step": 2330 + }, + { + "epoch": 0.02331, + "grad_norm": 0.73894366097317, + "learning_rate": 0.003, + "loss": 4.2587, + "step": 2331 + }, + { + "epoch": 0.02332, + "grad_norm": 0.7851778924150568, + "learning_rate": 0.003, + "loss": 4.2242, + "step": 2332 + }, + { + "epoch": 0.02333, + "grad_norm": 0.7287502366417434, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 2333 + }, + { + "epoch": 0.02334, + "grad_norm": 0.69094548866664, + "learning_rate": 0.003, + "loss": 4.197, + "step": 2334 + }, + { + "epoch": 0.02335, + "grad_norm": 0.6431810869476228, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 2335 + }, + { + "epoch": 0.02336, + "grad_norm": 0.6507673975963663, + "learning_rate": 0.003, + "loss": 4.2276, + "step": 2336 + }, + { + "epoch": 0.02337, + "grad_norm": 0.7087163474933083, + "learning_rate": 0.003, + "loss": 4.2296, + "step": 2337 + }, + { + "epoch": 0.02338, + "grad_norm": 0.6659743716890278, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 2338 + }, + { + "epoch": 0.02339, + "grad_norm": 0.5625472175847084, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 2339 + }, + { + "epoch": 0.0234, + "grad_norm": 0.536202676220793, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 2340 + }, + { + "epoch": 0.02341, + "grad_norm": 0.4892866338203838, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2341 + }, + { + "epoch": 0.02342, + "grad_norm": 0.4796875190367091, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2342 + }, + { + "epoch": 0.02343, + "grad_norm": 0.43636850083944123, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 2343 + }, + { + "epoch": 0.02344, + "grad_norm": 0.4075473684449614, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 2344 + }, + { + "epoch": 0.02345, + "grad_norm": 0.4126961162680678, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2345 + }, + { + "epoch": 0.02346, + "grad_norm": 0.4484646299481725, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 2346 + }, + { + "epoch": 0.02347, + "grad_norm": 0.4884767738301177, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 2347 + }, + { + "epoch": 0.02348, + "grad_norm": 0.4734903472149947, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2348 + }, + { + "epoch": 0.02349, + "grad_norm": 0.4515057387841677, + "learning_rate": 0.003, + "loss": 4.2087, + "step": 2349 + }, + { + "epoch": 0.0235, + "grad_norm": 0.43769490231507474, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2350 + }, + { + "epoch": 0.02351, + "grad_norm": 0.3875892706722824, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2351 + }, + { + "epoch": 0.02352, + "grad_norm": 0.42012415359844946, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 2352 + }, + { + "epoch": 0.02353, + "grad_norm": 0.5800085038244271, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 2353 + }, + { + "epoch": 0.02354, + "grad_norm": 0.9008192298721557, + "learning_rate": 0.003, + "loss": 4.2263, + "step": 2354 + }, + { + "epoch": 0.02355, + "grad_norm": 1.174355012660842, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 2355 + }, + { + "epoch": 0.02356, + "grad_norm": 0.6757502362960481, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 2356 + }, + { + "epoch": 0.02357, + "grad_norm": 0.7784416987559685, + "learning_rate": 0.003, + "loss": 4.2185, + "step": 2357 + }, + { + "epoch": 0.02358, + "grad_norm": 0.8207171439845993, + "learning_rate": 0.003, + "loss": 4.2385, + "step": 2358 + }, + { + "epoch": 0.02359, + "grad_norm": 0.6974291909709516, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 2359 + }, + { + "epoch": 0.0236, + "grad_norm": 0.6185472104370391, + "learning_rate": 0.003, + "loss": 4.2471, + "step": 2360 + }, + { + "epoch": 0.02361, + "grad_norm": 0.5773489611025617, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 2361 + }, + { + "epoch": 0.02362, + "grad_norm": 0.5560165828585443, + "learning_rate": 0.003, + "loss": 4.217, + "step": 2362 + }, + { + "epoch": 0.02363, + "grad_norm": 0.5892532979179639, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2363 + }, + { + "epoch": 0.02364, + "grad_norm": 0.6748161886878429, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 2364 + }, + { + "epoch": 0.02365, + "grad_norm": 0.6598699309472811, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 2365 + }, + { + "epoch": 0.02366, + "grad_norm": 0.6235317744814385, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2366 + }, + { + "epoch": 0.02367, + "grad_norm": 0.6532173738874076, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 2367 + }, + { + "epoch": 0.02368, + "grad_norm": 0.6882158454966492, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2368 + }, + { + "epoch": 0.02369, + "grad_norm": 0.7139133860202842, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 2369 + }, + { + "epoch": 0.0237, + "grad_norm": 0.6776525799328785, + "learning_rate": 0.003, + "loss": 4.2355, + "step": 2370 + }, + { + "epoch": 0.02371, + "grad_norm": 0.7002833727021253, + "learning_rate": 0.003, + "loss": 4.2219, + "step": 2371 + }, + { + "epoch": 0.02372, + "grad_norm": 0.7223141017308237, + "learning_rate": 0.003, + "loss": 4.2127, + "step": 2372 + }, + { + "epoch": 0.02373, + "grad_norm": 0.7004144412962903, + "learning_rate": 0.003, + "loss": 4.2003, + "step": 2373 + }, + { + "epoch": 0.02374, + "grad_norm": 0.6806743664027574, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 2374 + }, + { + "epoch": 0.02375, + "grad_norm": 0.6705132166888766, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2375 + }, + { + "epoch": 0.02376, + "grad_norm": 0.5005122224104306, + "learning_rate": 0.003, + "loss": 4.2267, + "step": 2376 + }, + { + "epoch": 0.02377, + "grad_norm": 0.4877832394535443, + "learning_rate": 0.003, + "loss": 4.194, + "step": 2377 + }, + { + "epoch": 0.02378, + "grad_norm": 0.5211637717940496, + "learning_rate": 0.003, + "loss": 4.203, + "step": 2378 + }, + { + "epoch": 0.02379, + "grad_norm": 0.6384313580109909, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 2379 + }, + { + "epoch": 0.0238, + "grad_norm": 0.7131026886322172, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 2380 + }, + { + "epoch": 0.02381, + "grad_norm": 0.6937779462786806, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2381 + }, + { + "epoch": 0.02382, + "grad_norm": 0.6337775600473754, + "learning_rate": 0.003, + "loss": 4.2401, + "step": 2382 + }, + { + "epoch": 0.02383, + "grad_norm": 0.5646754547957066, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 2383 + }, + { + "epoch": 0.02384, + "grad_norm": 0.5471514877116583, + "learning_rate": 0.003, + "loss": 4.1863, + "step": 2384 + }, + { + "epoch": 0.02385, + "grad_norm": 0.5791601386868904, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 2385 + }, + { + "epoch": 0.02386, + "grad_norm": 0.5976046883190522, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2386 + }, + { + "epoch": 0.02387, + "grad_norm": 0.6655482933999028, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 2387 + }, + { + "epoch": 0.02388, + "grad_norm": 0.7334421734543064, + "learning_rate": 0.003, + "loss": 4.2504, + "step": 2388 + }, + { + "epoch": 0.02389, + "grad_norm": 0.7475937838490452, + "learning_rate": 0.003, + "loss": 4.2143, + "step": 2389 + }, + { + "epoch": 0.0239, + "grad_norm": 0.7091495046860523, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2390 + }, + { + "epoch": 0.02391, + "grad_norm": 0.6788986600308993, + "learning_rate": 0.003, + "loss": 4.2281, + "step": 2391 + }, + { + "epoch": 0.02392, + "grad_norm": 0.6156974740652746, + "learning_rate": 0.003, + "loss": 4.2308, + "step": 2392 + }, + { + "epoch": 0.02393, + "grad_norm": 0.5941793885935496, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 2393 + }, + { + "epoch": 0.02394, + "grad_norm": 0.5863170046009283, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2394 + }, + { + "epoch": 0.02395, + "grad_norm": 0.7380191806299783, + "learning_rate": 0.003, + "loss": 4.2191, + "step": 2395 + }, + { + "epoch": 0.02396, + "grad_norm": 0.8712169644578591, + "learning_rate": 0.003, + "loss": 4.2425, + "step": 2396 + }, + { + "epoch": 0.02397, + "grad_norm": 0.8536583911176343, + "learning_rate": 0.003, + "loss": 4.2305, + "step": 2397 + }, + { + "epoch": 0.02398, + "grad_norm": 0.7082106435478536, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 2398 + }, + { + "epoch": 0.02399, + "grad_norm": 0.6784314246730624, + "learning_rate": 0.003, + "loss": 4.2124, + "step": 2399 + }, + { + "epoch": 0.024, + "grad_norm": 0.7409520225246221, + "learning_rate": 0.003, + "loss": 4.2253, + "step": 2400 + }, + { + "epoch": 0.02401, + "grad_norm": 0.7471077758325675, + "learning_rate": 0.003, + "loss": 4.2007, + "step": 2401 + }, + { + "epoch": 0.02402, + "grad_norm": 0.7167664651670812, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 2402 + }, + { + "epoch": 0.02403, + "grad_norm": 0.7133428609804063, + "learning_rate": 0.003, + "loss": 4.2011, + "step": 2403 + }, + { + "epoch": 0.02404, + "grad_norm": 0.6914213864059761, + "learning_rate": 0.003, + "loss": 4.2426, + "step": 2404 + }, + { + "epoch": 0.02405, + "grad_norm": 0.6223491766257139, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 2405 + }, + { + "epoch": 0.02406, + "grad_norm": 0.6402462299592815, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 2406 + }, + { + "epoch": 0.02407, + "grad_norm": 0.5600794609103489, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 2407 + }, + { + "epoch": 0.02408, + "grad_norm": 0.4916021474517548, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 2408 + }, + { + "epoch": 0.02409, + "grad_norm": 0.46461844041782707, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 2409 + }, + { + "epoch": 0.0241, + "grad_norm": 0.5180903118600274, + "learning_rate": 0.003, + "loss": 4.2239, + "step": 2410 + }, + { + "epoch": 0.02411, + "grad_norm": 0.6289211485922821, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2411 + }, + { + "epoch": 0.02412, + "grad_norm": 0.625292309840426, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 2412 + }, + { + "epoch": 0.02413, + "grad_norm": 0.5316162511007585, + "learning_rate": 0.003, + "loss": 4.1942, + "step": 2413 + }, + { + "epoch": 0.02414, + "grad_norm": 0.569080239332035, + "learning_rate": 0.003, + "loss": 4.2105, + "step": 2414 + }, + { + "epoch": 0.02415, + "grad_norm": 0.6830500278720256, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2415 + }, + { + "epoch": 0.02416, + "grad_norm": 0.8230987152330054, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 2416 + }, + { + "epoch": 0.02417, + "grad_norm": 0.9992750696467811, + "learning_rate": 0.003, + "loss": 4.241, + "step": 2417 + }, + { + "epoch": 0.02418, + "grad_norm": 0.7172338223207045, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 2418 + }, + { + "epoch": 0.02419, + "grad_norm": 0.6066239591362272, + "learning_rate": 0.003, + "loss": 4.2197, + "step": 2419 + }, + { + "epoch": 0.0242, + "grad_norm": 0.6615008268071535, + "learning_rate": 0.003, + "loss": 4.2339, + "step": 2420 + }, + { + "epoch": 0.02421, + "grad_norm": 0.601398245236051, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 2421 + }, + { + "epoch": 0.02422, + "grad_norm": 0.5164536803360787, + "learning_rate": 0.003, + "loss": 4.2444, + "step": 2422 + }, + { + "epoch": 0.02423, + "grad_norm": 0.47425853655922334, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 2423 + }, + { + "epoch": 0.02424, + "grad_norm": 0.45557836325278056, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 2424 + }, + { + "epoch": 0.02425, + "grad_norm": 0.37992471673130124, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 2425 + }, + { + "epoch": 0.02426, + "grad_norm": 0.4388918700897719, + "learning_rate": 0.003, + "loss": 4.2146, + "step": 2426 + }, + { + "epoch": 0.02427, + "grad_norm": 0.47592644536510775, + "learning_rate": 0.003, + "loss": 4.2395, + "step": 2427 + }, + { + "epoch": 0.02428, + "grad_norm": 0.41079892436338244, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2428 + }, + { + "epoch": 0.02429, + "grad_norm": 0.48266596555203656, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 2429 + }, + { + "epoch": 0.0243, + "grad_norm": 0.58630499435022, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 2430 + }, + { + "epoch": 0.02431, + "grad_norm": 0.6362348679805433, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2431 + }, + { + "epoch": 0.02432, + "grad_norm": 0.6409895194011197, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 2432 + }, + { + "epoch": 0.02433, + "grad_norm": 0.7045349857959661, + "learning_rate": 0.003, + "loss": 4.239, + "step": 2433 + }, + { + "epoch": 0.02434, + "grad_norm": 0.6987807589267505, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2434 + }, + { + "epoch": 0.02435, + "grad_norm": 0.5878135110602212, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 2435 + }, + { + "epoch": 0.02436, + "grad_norm": 0.5898177674736178, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 2436 + }, + { + "epoch": 0.02437, + "grad_norm": 0.5929083773979196, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 2437 + }, + { + "epoch": 0.02438, + "grad_norm": 0.6434362451996457, + "learning_rate": 0.003, + "loss": 4.2142, + "step": 2438 + }, + { + "epoch": 0.02439, + "grad_norm": 0.6611917410629563, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2439 + }, + { + "epoch": 0.0244, + "grad_norm": 0.7743759672777403, + "learning_rate": 0.003, + "loss": 4.2424, + "step": 2440 + }, + { + "epoch": 0.02441, + "grad_norm": 0.7462269728851785, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2441 + }, + { + "epoch": 0.02442, + "grad_norm": 0.7608157833129767, + "learning_rate": 0.003, + "loss": 4.2431, + "step": 2442 + }, + { + "epoch": 0.02443, + "grad_norm": 0.79170341480151, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2443 + }, + { + "epoch": 0.02444, + "grad_norm": 0.7939192302282532, + "learning_rate": 0.003, + "loss": 4.2161, + "step": 2444 + }, + { + "epoch": 0.02445, + "grad_norm": 0.7598676209330011, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 2445 + }, + { + "epoch": 0.02446, + "grad_norm": 0.657657458880643, + "learning_rate": 0.003, + "loss": 4.2177, + "step": 2446 + }, + { + "epoch": 0.02447, + "grad_norm": 0.636985404351659, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 2447 + }, + { + "epoch": 0.02448, + "grad_norm": 0.6540065013234154, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 2448 + }, + { + "epoch": 0.02449, + "grad_norm": 0.6430617212308102, + "learning_rate": 0.003, + "loss": 4.218, + "step": 2449 + }, + { + "epoch": 0.0245, + "grad_norm": 0.6859449218709958, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 2450 + }, + { + "epoch": 0.02451, + "grad_norm": 0.6585394183553759, + "learning_rate": 0.003, + "loss": 4.203, + "step": 2451 + }, + { + "epoch": 0.02452, + "grad_norm": 0.5944042022079538, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2452 + }, + { + "epoch": 0.02453, + "grad_norm": 0.6245280633667185, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2453 + }, + { + "epoch": 0.02454, + "grad_norm": 0.6285302652257769, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2454 + }, + { + "epoch": 0.02455, + "grad_norm": 0.630536065900029, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 2455 + }, + { + "epoch": 0.02456, + "grad_norm": 0.6392454947407117, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 2456 + }, + { + "epoch": 0.02457, + "grad_norm": 0.732191714374005, + "learning_rate": 0.003, + "loss": 4.2215, + "step": 2457 + }, + { + "epoch": 0.02458, + "grad_norm": 0.8168261785432145, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 2458 + }, + { + "epoch": 0.02459, + "grad_norm": 0.8118895866121905, + "learning_rate": 0.003, + "loss": 4.2313, + "step": 2459 + }, + { + "epoch": 0.0246, + "grad_norm": 0.7597126257704254, + "learning_rate": 0.003, + "loss": 4.2347, + "step": 2460 + }, + { + "epoch": 0.02461, + "grad_norm": 0.7096048570719959, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2461 + }, + { + "epoch": 0.02462, + "grad_norm": 0.6412516044054245, + "learning_rate": 0.003, + "loss": 4.2333, + "step": 2462 + }, + { + "epoch": 0.02463, + "grad_norm": 0.6158305280803775, + "learning_rate": 0.003, + "loss": 4.2316, + "step": 2463 + }, + { + "epoch": 0.02464, + "grad_norm": 0.5837936979418801, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 2464 + }, + { + "epoch": 0.02465, + "grad_norm": 0.5652391806413423, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 2465 + }, + { + "epoch": 0.02466, + "grad_norm": 0.5314994263522052, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 2466 + }, + { + "epoch": 0.02467, + "grad_norm": 0.5492980443942803, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 2467 + }, + { + "epoch": 0.02468, + "grad_norm": 0.5476949286484541, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 2468 + }, + { + "epoch": 0.02469, + "grad_norm": 0.5739182126857109, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 2469 + }, + { + "epoch": 0.0247, + "grad_norm": 0.6455625601186472, + "learning_rate": 0.003, + "loss": 4.2003, + "step": 2470 + }, + { + "epoch": 0.02471, + "grad_norm": 0.7500801989374446, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 2471 + }, + { + "epoch": 0.02472, + "grad_norm": 0.8015350177741498, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 2472 + }, + { + "epoch": 0.02473, + "grad_norm": 0.7700030742386493, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 2473 + }, + { + "epoch": 0.02474, + "grad_norm": 0.6591968552189625, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 2474 + }, + { + "epoch": 0.02475, + "grad_norm": 0.5913092934763802, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2475 + }, + { + "epoch": 0.02476, + "grad_norm": 0.6828419322627938, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2476 + }, + { + "epoch": 0.02477, + "grad_norm": 0.7429757949041498, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 2477 + }, + { + "epoch": 0.02478, + "grad_norm": 0.7533856527859732, + "learning_rate": 0.003, + "loss": 4.2029, + "step": 2478 + }, + { + "epoch": 0.02479, + "grad_norm": 0.6271813598380395, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 2479 + }, + { + "epoch": 0.0248, + "grad_norm": 0.7048163325838352, + "learning_rate": 0.003, + "loss": 4.2341, + "step": 2480 + }, + { + "epoch": 0.02481, + "grad_norm": 0.7059455660386844, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2481 + }, + { + "epoch": 0.02482, + "grad_norm": 0.6095979307747832, + "learning_rate": 0.003, + "loss": 4.2232, + "step": 2482 + }, + { + "epoch": 0.02483, + "grad_norm": 0.589007344890445, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2483 + }, + { + "epoch": 0.02484, + "grad_norm": 0.5795798273023001, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 2484 + }, + { + "epoch": 0.02485, + "grad_norm": 0.6558357821242041, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 2485 + }, + { + "epoch": 0.02486, + "grad_norm": 0.6694443853300054, + "learning_rate": 0.003, + "loss": 4.2506, + "step": 2486 + }, + { + "epoch": 0.02487, + "grad_norm": 0.7548583562826651, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 2487 + }, + { + "epoch": 0.02488, + "grad_norm": 0.6086868759932925, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 2488 + }, + { + "epoch": 0.02489, + "grad_norm": 0.5543068870038478, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 2489 + }, + { + "epoch": 0.0249, + "grad_norm": 0.4695404599404173, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 2490 + }, + { + "epoch": 0.02491, + "grad_norm": 0.5098980691467636, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 2491 + }, + { + "epoch": 0.02492, + "grad_norm": 0.5029414147729624, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2492 + }, + { + "epoch": 0.02493, + "grad_norm": 0.5644399016032551, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2493 + }, + { + "epoch": 0.02494, + "grad_norm": 0.6367872115761151, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2494 + }, + { + "epoch": 0.02495, + "grad_norm": 0.6767787025658096, + "learning_rate": 0.003, + "loss": 4.2079, + "step": 2495 + }, + { + "epoch": 0.02496, + "grad_norm": 0.696192347480972, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 2496 + }, + { + "epoch": 0.02497, + "grad_norm": 0.6551341280875559, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 2497 + }, + { + "epoch": 0.02498, + "grad_norm": 0.6087269213019051, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2498 + }, + { + "epoch": 0.02499, + "grad_norm": 0.6686291413156635, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2499 + }, + { + "epoch": 0.025, + "grad_norm": 0.6959767014574844, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 2500 + }, + { + "epoch": 0.02501, + "grad_norm": 0.6984509411978452, + "learning_rate": 0.003, + "loss": 4.2249, + "step": 2501 + }, + { + "epoch": 0.02502, + "grad_norm": 0.6296748558389454, + "learning_rate": 0.003, + "loss": 4.204, + "step": 2502 + }, + { + "epoch": 0.02503, + "grad_norm": 0.7588933964884115, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 2503 + }, + { + "epoch": 0.02504, + "grad_norm": 0.7608780929496015, + "learning_rate": 0.003, + "loss": 4.2023, + "step": 2504 + }, + { + "epoch": 0.02505, + "grad_norm": 0.7626036518863434, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 2505 + }, + { + "epoch": 0.02506, + "grad_norm": 0.8721628417499054, + "learning_rate": 0.003, + "loss": 4.2222, + "step": 2506 + }, + { + "epoch": 0.02507, + "grad_norm": 0.77178561857626, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 2507 + }, + { + "epoch": 0.02508, + "grad_norm": 0.728729150203243, + "learning_rate": 0.003, + "loss": 4.2073, + "step": 2508 + }, + { + "epoch": 0.02509, + "grad_norm": 0.6476865147041562, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 2509 + }, + { + "epoch": 0.0251, + "grad_norm": 0.6920620401373876, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 2510 + }, + { + "epoch": 0.02511, + "grad_norm": 0.7129030622439039, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 2511 + }, + { + "epoch": 0.02512, + "grad_norm": 0.7903195113153619, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 2512 + }, + { + "epoch": 0.02513, + "grad_norm": 0.7730609450390953, + "learning_rate": 0.003, + "loss": 4.2181, + "step": 2513 + }, + { + "epoch": 0.02514, + "grad_norm": 0.656826259281822, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 2514 + }, + { + "epoch": 0.02515, + "grad_norm": 0.6482134555064485, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2515 + }, + { + "epoch": 0.02516, + "grad_norm": 0.5967395863996815, + "learning_rate": 0.003, + "loss": 4.205, + "step": 2516 + }, + { + "epoch": 0.02517, + "grad_norm": 0.5007565780061344, + "learning_rate": 0.003, + "loss": 4.2015, + "step": 2517 + }, + { + "epoch": 0.02518, + "grad_norm": 0.5217525966473281, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 2518 + }, + { + "epoch": 0.02519, + "grad_norm": 0.4784613905987089, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 2519 + }, + { + "epoch": 0.0252, + "grad_norm": 0.4743563932590768, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2520 + }, + { + "epoch": 0.02521, + "grad_norm": 0.487720505759532, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 2521 + }, + { + "epoch": 0.02522, + "grad_norm": 0.5322961687406275, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 2522 + }, + { + "epoch": 0.02523, + "grad_norm": 0.5314708118979667, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 2523 + }, + { + "epoch": 0.02524, + "grad_norm": 0.565643482510681, + "learning_rate": 0.003, + "loss": 4.2052, + "step": 2524 + }, + { + "epoch": 0.02525, + "grad_norm": 0.5980154086501233, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 2525 + }, + { + "epoch": 0.02526, + "grad_norm": 0.668292503382125, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 2526 + }, + { + "epoch": 0.02527, + "grad_norm": 0.8972899326232674, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2527 + }, + { + "epoch": 0.02528, + "grad_norm": 1.071209575597214, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 2528 + }, + { + "epoch": 0.02529, + "grad_norm": 0.7116312192642624, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 2529 + }, + { + "epoch": 0.0253, + "grad_norm": 0.6832964815647001, + "learning_rate": 0.003, + "loss": 4.2324, + "step": 2530 + }, + { + "epoch": 0.02531, + "grad_norm": 0.625455406806653, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 2531 + }, + { + "epoch": 0.02532, + "grad_norm": 0.5481673077994964, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 2532 + }, + { + "epoch": 0.02533, + "grad_norm": 0.5897257835444514, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2533 + }, + { + "epoch": 0.02534, + "grad_norm": 0.5146548286556276, + "learning_rate": 0.003, + "loss": 4.2083, + "step": 2534 + }, + { + "epoch": 0.02535, + "grad_norm": 0.5177859418119969, + "learning_rate": 0.003, + "loss": 4.2022, + "step": 2535 + }, + { + "epoch": 0.02536, + "grad_norm": 0.5298155181993902, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 2536 + }, + { + "epoch": 0.02537, + "grad_norm": 0.4935521410250174, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2537 + }, + { + "epoch": 0.02538, + "grad_norm": 0.46468573263998625, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 2538 + }, + { + "epoch": 0.02539, + "grad_norm": 0.517571911001853, + "learning_rate": 0.003, + "loss": 4.2155, + "step": 2539 + }, + { + "epoch": 0.0254, + "grad_norm": 0.5342760221554671, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 2540 + }, + { + "epoch": 0.02541, + "grad_norm": 0.54814214731398, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 2541 + }, + { + "epoch": 0.02542, + "grad_norm": 0.6094140763884728, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2542 + }, + { + "epoch": 0.02543, + "grad_norm": 0.7209045415511117, + "learning_rate": 0.003, + "loss": 4.2097, + "step": 2543 + }, + { + "epoch": 0.02544, + "grad_norm": 0.8126054331647753, + "learning_rate": 0.003, + "loss": 4.2042, + "step": 2544 + }, + { + "epoch": 0.02545, + "grad_norm": 0.7645479610935821, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 2545 + }, + { + "epoch": 0.02546, + "grad_norm": 0.6438407031724117, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2546 + }, + { + "epoch": 0.02547, + "grad_norm": 0.6119848106712312, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 2547 + }, + { + "epoch": 0.02548, + "grad_norm": 0.5927448337585458, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 2548 + }, + { + "epoch": 0.02549, + "grad_norm": 0.8336977020901954, + "learning_rate": 0.003, + "loss": 4.2269, + "step": 2549 + }, + { + "epoch": 0.0255, + "grad_norm": 1.0008812022371014, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2550 + }, + { + "epoch": 0.02551, + "grad_norm": 1.047497616869583, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 2551 + }, + { + "epoch": 0.02552, + "grad_norm": 0.9125531296176879, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 2552 + }, + { + "epoch": 0.02553, + "grad_norm": 1.0094140787791404, + "learning_rate": 0.003, + "loss": 4.2917, + "step": 2553 + }, + { + "epoch": 0.02554, + "grad_norm": 0.9401220726777249, + "learning_rate": 0.003, + "loss": 4.2404, + "step": 2554 + }, + { + "epoch": 0.02555, + "grad_norm": 0.8810312726451306, + "learning_rate": 0.003, + "loss": 4.233, + "step": 2555 + }, + { + "epoch": 0.02556, + "grad_norm": 0.8398872404556826, + "learning_rate": 0.003, + "loss": 4.242, + "step": 2556 + }, + { + "epoch": 0.02557, + "grad_norm": 0.8674958106494866, + "learning_rate": 0.003, + "loss": 4.251, + "step": 2557 + }, + { + "epoch": 0.02558, + "grad_norm": 0.9359611753863145, + "learning_rate": 0.003, + "loss": 4.2763, + "step": 2558 + }, + { + "epoch": 0.02559, + "grad_norm": 0.919728435180346, + "learning_rate": 0.003, + "loss": 4.2634, + "step": 2559 + }, + { + "epoch": 0.0256, + "grad_norm": 0.9704240601933475, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 2560 + }, + { + "epoch": 0.02561, + "grad_norm": 0.9645281717104208, + "learning_rate": 0.003, + "loss": 4.2533, + "step": 2561 + }, + { + "epoch": 0.02562, + "grad_norm": 0.9553791325261367, + "learning_rate": 0.003, + "loss": 4.2754, + "step": 2562 + }, + { + "epoch": 0.02563, + "grad_norm": 0.7877299682148988, + "learning_rate": 0.003, + "loss": 4.2515, + "step": 2563 + }, + { + "epoch": 0.02564, + "grad_norm": 0.6960170044487899, + "learning_rate": 0.003, + "loss": 4.264, + "step": 2564 + }, + { + "epoch": 0.02565, + "grad_norm": 0.5853819163555847, + "learning_rate": 0.003, + "loss": 4.2346, + "step": 2565 + }, + { + "epoch": 0.02566, + "grad_norm": 0.56684551943134, + "learning_rate": 0.003, + "loss": 4.2543, + "step": 2566 + }, + { + "epoch": 0.02567, + "grad_norm": 0.5518127543434307, + "learning_rate": 0.003, + "loss": 4.251, + "step": 2567 + }, + { + "epoch": 0.02568, + "grad_norm": 0.5196560828829251, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 2568 + }, + { + "epoch": 0.02569, + "grad_norm": 0.595356322869423, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 2569 + }, + { + "epoch": 0.0257, + "grad_norm": 0.7758881983100933, + "learning_rate": 0.003, + "loss": 4.2302, + "step": 2570 + }, + { + "epoch": 0.02571, + "grad_norm": 0.9784633915798607, + "learning_rate": 0.003, + "loss": 4.2591, + "step": 2571 + }, + { + "epoch": 0.02572, + "grad_norm": 0.9890948659534822, + "learning_rate": 0.003, + "loss": 4.2413, + "step": 2572 + }, + { + "epoch": 0.02573, + "grad_norm": 0.7618440411400482, + "learning_rate": 0.003, + "loss": 4.2415, + "step": 2573 + }, + { + "epoch": 0.02574, + "grad_norm": 0.5829959121875294, + "learning_rate": 0.003, + "loss": 4.2662, + "step": 2574 + }, + { + "epoch": 0.02575, + "grad_norm": 0.5754964401101073, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2575 + }, + { + "epoch": 0.02576, + "grad_norm": 0.5023263337601672, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2576 + }, + { + "epoch": 0.02577, + "grad_norm": 0.5094113404385299, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 2577 + }, + { + "epoch": 0.02578, + "grad_norm": 0.45458875378333236, + "learning_rate": 0.003, + "loss": 4.216, + "step": 2578 + }, + { + "epoch": 0.02579, + "grad_norm": 0.376766997653558, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2579 + }, + { + "epoch": 0.0258, + "grad_norm": 0.3615714006582123, + "learning_rate": 0.003, + "loss": 4.2143, + "step": 2580 + }, + { + "epoch": 0.02581, + "grad_norm": 0.34139864457507546, + "learning_rate": 0.003, + "loss": 4.2314, + "step": 2581 + }, + { + "epoch": 0.02582, + "grad_norm": 0.31484861485882354, + "learning_rate": 0.003, + "loss": 4.225, + "step": 2582 + }, + { + "epoch": 0.02583, + "grad_norm": 0.27908396977569033, + "learning_rate": 0.003, + "loss": 4.1944, + "step": 2583 + }, + { + "epoch": 0.02584, + "grad_norm": 0.28343226190675325, + "learning_rate": 0.003, + "loss": 4.2059, + "step": 2584 + }, + { + "epoch": 0.02585, + "grad_norm": 0.31066810836348263, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2585 + }, + { + "epoch": 0.02586, + "grad_norm": 0.29167242806584853, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2586 + }, + { + "epoch": 0.02587, + "grad_norm": 0.3309461372612561, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 2587 + }, + { + "epoch": 0.02588, + "grad_norm": 0.3812190342327403, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 2588 + }, + { + "epoch": 0.02589, + "grad_norm": 0.48608400124636386, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2589 + }, + { + "epoch": 0.0259, + "grad_norm": 0.6708650835599492, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 2590 + }, + { + "epoch": 0.02591, + "grad_norm": 0.7815243168293263, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2591 + }, + { + "epoch": 0.02592, + "grad_norm": 0.6070733396030551, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2592 + }, + { + "epoch": 0.02593, + "grad_norm": 0.36755746327659883, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2593 + }, + { + "epoch": 0.02594, + "grad_norm": 0.5461931648813682, + "learning_rate": 0.003, + "loss": 4.2113, + "step": 2594 + }, + { + "epoch": 0.02595, + "grad_norm": 0.5548085387163001, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2595 + }, + { + "epoch": 0.02596, + "grad_norm": 0.6279120258670815, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 2596 + }, + { + "epoch": 0.02597, + "grad_norm": 0.6833879500904073, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 2597 + }, + { + "epoch": 0.02598, + "grad_norm": 0.7079000881755875, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 2598 + }, + { + "epoch": 0.02599, + "grad_norm": 0.6366436124583713, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2599 + }, + { + "epoch": 0.026, + "grad_norm": 0.5901740968221895, + "learning_rate": 0.003, + "loss": 4.211, + "step": 2600 + }, + { + "epoch": 0.02601, + "grad_norm": 0.5956938888246321, + "learning_rate": 0.003, + "loss": 4.2151, + "step": 2601 + }, + { + "epoch": 0.02602, + "grad_norm": 0.5493627512392831, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 2602 + }, + { + "epoch": 0.02603, + "grad_norm": 0.44261728447277765, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2603 + }, + { + "epoch": 0.02604, + "grad_norm": 0.47710609408550947, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 2604 + }, + { + "epoch": 0.02605, + "grad_norm": 0.4918626291452229, + "learning_rate": 0.003, + "loss": 4.1854, + "step": 2605 + }, + { + "epoch": 0.02606, + "grad_norm": 0.46913897415532124, + "learning_rate": 0.003, + "loss": 4.204, + "step": 2606 + }, + { + "epoch": 0.02607, + "grad_norm": 0.48033538034271883, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2607 + }, + { + "epoch": 0.02608, + "grad_norm": 0.5430081805575081, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 2608 + }, + { + "epoch": 0.02609, + "grad_norm": 0.7745086701981942, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 2609 + }, + { + "epoch": 0.0261, + "grad_norm": 0.9079228266743159, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2610 + }, + { + "epoch": 0.02611, + "grad_norm": 0.8915671065001584, + "learning_rate": 0.003, + "loss": 4.183, + "step": 2611 + }, + { + "epoch": 0.02612, + "grad_norm": 0.8970335727218054, + "learning_rate": 0.003, + "loss": 4.2272, + "step": 2612 + }, + { + "epoch": 0.02613, + "grad_norm": 0.7184841949590459, + "learning_rate": 0.003, + "loss": 4.215, + "step": 2613 + }, + { + "epoch": 0.02614, + "grad_norm": 0.6361221793660993, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2614 + }, + { + "epoch": 0.02615, + "grad_norm": 0.6008095298009088, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2615 + }, + { + "epoch": 0.02616, + "grad_norm": 0.5894242905465984, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2616 + }, + { + "epoch": 0.02617, + "grad_norm": 0.6436011542740302, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 2617 + }, + { + "epoch": 0.02618, + "grad_norm": 0.7283160674464684, + "learning_rate": 0.003, + "loss": 4.2126, + "step": 2618 + }, + { + "epoch": 0.02619, + "grad_norm": 0.7948866659668812, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 2619 + }, + { + "epoch": 0.0262, + "grad_norm": 0.7483085908476613, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2620 + }, + { + "epoch": 0.02621, + "grad_norm": 0.6674824981271792, + "learning_rate": 0.003, + "loss": 4.217, + "step": 2621 + }, + { + "epoch": 0.02622, + "grad_norm": 0.6038002286952583, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2622 + }, + { + "epoch": 0.02623, + "grad_norm": 0.6014737867483161, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 2623 + }, + { + "epoch": 0.02624, + "grad_norm": 0.5903601826725474, + "learning_rate": 0.003, + "loss": 4.2493, + "step": 2624 + }, + { + "epoch": 0.02625, + "grad_norm": 0.6662689883238633, + "learning_rate": 0.003, + "loss": 4.2487, + "step": 2625 + }, + { + "epoch": 0.02626, + "grad_norm": 0.6773686120638865, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2626 + }, + { + "epoch": 0.02627, + "grad_norm": 0.6189409649870486, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 2627 + }, + { + "epoch": 0.02628, + "grad_norm": 0.6566341590011362, + "learning_rate": 0.003, + "loss": 4.2282, + "step": 2628 + }, + { + "epoch": 0.02629, + "grad_norm": 0.7499843369890263, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2629 + }, + { + "epoch": 0.0263, + "grad_norm": 0.7612666434746392, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 2630 + }, + { + "epoch": 0.02631, + "grad_norm": 0.6535861726328742, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 2631 + }, + { + "epoch": 0.02632, + "grad_norm": 0.6394295181872582, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2632 + }, + { + "epoch": 0.02633, + "grad_norm": 0.6261623421196629, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 2633 + }, + { + "epoch": 0.02634, + "grad_norm": 0.574244319715974, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2634 + }, + { + "epoch": 0.02635, + "grad_norm": 0.4907934619440069, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2635 + }, + { + "epoch": 0.02636, + "grad_norm": 0.47213288352020044, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2636 + }, + { + "epoch": 0.02637, + "grad_norm": 0.4457276996768415, + "learning_rate": 0.003, + "loss": 4.2168, + "step": 2637 + }, + { + "epoch": 0.02638, + "grad_norm": 0.45345762795109046, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2638 + }, + { + "epoch": 0.02639, + "grad_norm": 0.49627655516291913, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 2639 + }, + { + "epoch": 0.0264, + "grad_norm": 0.5785078427721256, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2640 + }, + { + "epoch": 0.02641, + "grad_norm": 0.7855264178492048, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2641 + }, + { + "epoch": 0.02642, + "grad_norm": 0.9564797199743369, + "learning_rate": 0.003, + "loss": 4.2106, + "step": 2642 + }, + { + "epoch": 0.02643, + "grad_norm": 0.9009180286527748, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 2643 + }, + { + "epoch": 0.02644, + "grad_norm": 0.6806094030785705, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 2644 + }, + { + "epoch": 0.02645, + "grad_norm": 0.6950198239199242, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 2645 + }, + { + "epoch": 0.02646, + "grad_norm": 0.6983266391512944, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2646 + }, + { + "epoch": 0.02647, + "grad_norm": 0.6601491893781873, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 2647 + }, + { + "epoch": 0.02648, + "grad_norm": 0.6355139962402977, + "learning_rate": 0.003, + "loss": 4.2257, + "step": 2648 + }, + { + "epoch": 0.02649, + "grad_norm": 0.5820480039604559, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2649 + }, + { + "epoch": 0.0265, + "grad_norm": 0.5775761818558494, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2650 + }, + { + "epoch": 0.02651, + "grad_norm": 0.5277676161314399, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2651 + }, + { + "epoch": 0.02652, + "grad_norm": 0.5297857391988389, + "learning_rate": 0.003, + "loss": 4.2161, + "step": 2652 + }, + { + "epoch": 0.02653, + "grad_norm": 0.5897074985274656, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 2653 + }, + { + "epoch": 0.02654, + "grad_norm": 0.6546317550221565, + "learning_rate": 0.003, + "loss": 4.1856, + "step": 2654 + }, + { + "epoch": 0.02655, + "grad_norm": 0.6571917314744374, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 2655 + }, + { + "epoch": 0.02656, + "grad_norm": 0.6300742703920978, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 2656 + }, + { + "epoch": 0.02657, + "grad_norm": 0.6437986671993883, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2657 + }, + { + "epoch": 0.02658, + "grad_norm": 0.6540624785623048, + "learning_rate": 0.003, + "loss": 4.204, + "step": 2658 + }, + { + "epoch": 0.02659, + "grad_norm": 0.6833039431896774, + "learning_rate": 0.003, + "loss": 4.2004, + "step": 2659 + }, + { + "epoch": 0.0266, + "grad_norm": 0.6560030968945707, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2660 + }, + { + "epoch": 0.02661, + "grad_norm": 0.6045077056358651, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2661 + }, + { + "epoch": 0.02662, + "grad_norm": 0.5982743265464419, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 2662 + }, + { + "epoch": 0.02663, + "grad_norm": 0.6142468679426575, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2663 + }, + { + "epoch": 0.02664, + "grad_norm": 0.5662553451652035, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 2664 + }, + { + "epoch": 0.02665, + "grad_norm": 0.5761075041061567, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 2665 + }, + { + "epoch": 0.02666, + "grad_norm": 0.6870114427260798, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 2666 + }, + { + "epoch": 0.02667, + "grad_norm": 0.8311453279031553, + "learning_rate": 0.003, + "loss": 4.2051, + "step": 2667 + }, + { + "epoch": 0.02668, + "grad_norm": 0.9940869003642662, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 2668 + }, + { + "epoch": 0.02669, + "grad_norm": 0.9365809099626523, + "learning_rate": 0.003, + "loss": 4.2092, + "step": 2669 + }, + { + "epoch": 0.0267, + "grad_norm": 0.7814988387932911, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 2670 + }, + { + "epoch": 0.02671, + "grad_norm": 0.7536334880750852, + "learning_rate": 0.003, + "loss": 4.2232, + "step": 2671 + }, + { + "epoch": 0.02672, + "grad_norm": 0.6570195531624667, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2672 + }, + { + "epoch": 0.02673, + "grad_norm": 0.5512466402663184, + "learning_rate": 0.003, + "loss": 4.203, + "step": 2673 + }, + { + "epoch": 0.02674, + "grad_norm": 0.61723011024151, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2674 + }, + { + "epoch": 0.02675, + "grad_norm": 0.6694888220820355, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2675 + }, + { + "epoch": 0.02676, + "grad_norm": 0.684903042237445, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2676 + }, + { + "epoch": 0.02677, + "grad_norm": 0.6916594775917959, + "learning_rate": 0.003, + "loss": 4.2254, + "step": 2677 + }, + { + "epoch": 0.02678, + "grad_norm": 0.5888995948139836, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 2678 + }, + { + "epoch": 0.02679, + "grad_norm": 0.46001488738668683, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 2679 + }, + { + "epoch": 0.0268, + "grad_norm": 0.48559488053080857, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 2680 + }, + { + "epoch": 0.02681, + "grad_norm": 0.5328577853550642, + "learning_rate": 0.003, + "loss": 4.2162, + "step": 2681 + }, + { + "epoch": 0.02682, + "grad_norm": 0.5478209027693539, + "learning_rate": 0.003, + "loss": 4.1942, + "step": 2682 + }, + { + "epoch": 0.02683, + "grad_norm": 0.5231582728945893, + "learning_rate": 0.003, + "loss": 4.214, + "step": 2683 + }, + { + "epoch": 0.02684, + "grad_norm": 0.5004578011633336, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2684 + }, + { + "epoch": 0.02685, + "grad_norm": 0.5339999052577957, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 2685 + }, + { + "epoch": 0.02686, + "grad_norm": 0.5808031106775001, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2686 + }, + { + "epoch": 0.02687, + "grad_norm": 0.6474970996641412, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2687 + }, + { + "epoch": 0.02688, + "grad_norm": 0.6071304394655341, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 2688 + }, + { + "epoch": 0.02689, + "grad_norm": 0.6303081168202617, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 2689 + }, + { + "epoch": 0.0269, + "grad_norm": 0.6875418523901222, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 2690 + }, + { + "epoch": 0.02691, + "grad_norm": 0.8385284997095487, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 2691 + }, + { + "epoch": 0.02692, + "grad_norm": 0.9480846054518673, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2692 + }, + { + "epoch": 0.02693, + "grad_norm": 0.8071009083907058, + "learning_rate": 0.003, + "loss": 4.2239, + "step": 2693 + }, + { + "epoch": 0.02694, + "grad_norm": 0.726325622879689, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2694 + }, + { + "epoch": 0.02695, + "grad_norm": 0.6357501756622079, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 2695 + }, + { + "epoch": 0.02696, + "grad_norm": 0.5914031503189017, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2696 + }, + { + "epoch": 0.02697, + "grad_norm": 0.6523831217047013, + "learning_rate": 0.003, + "loss": 4.2118, + "step": 2697 + }, + { + "epoch": 0.02698, + "grad_norm": 0.6028229475417466, + "learning_rate": 0.003, + "loss": 4.1939, + "step": 2698 + }, + { + "epoch": 0.02699, + "grad_norm": 0.5476388167406291, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 2699 + }, + { + "epoch": 0.027, + "grad_norm": 0.4742225811072004, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2700 + }, + { + "epoch": 0.02701, + "grad_norm": 0.4202221652460036, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 2701 + }, + { + "epoch": 0.02702, + "grad_norm": 0.43365355703010106, + "learning_rate": 0.003, + "loss": 4.2144, + "step": 2702 + }, + { + "epoch": 0.02703, + "grad_norm": 0.4838782470885569, + "learning_rate": 0.003, + "loss": 4.2107, + "step": 2703 + }, + { + "epoch": 0.02704, + "grad_norm": 0.5383020066295563, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 2704 + }, + { + "epoch": 0.02705, + "grad_norm": 0.6989516882897651, + "learning_rate": 0.003, + "loss": 4.2137, + "step": 2705 + }, + { + "epoch": 0.02706, + "grad_norm": 0.7486144178237021, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2706 + }, + { + "epoch": 0.02707, + "grad_norm": 0.7126817427692109, + "learning_rate": 0.003, + "loss": 4.175, + "step": 2707 + }, + { + "epoch": 0.02708, + "grad_norm": 0.6105439188174862, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2708 + }, + { + "epoch": 0.02709, + "grad_norm": 0.5700397686001112, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2709 + }, + { + "epoch": 0.0271, + "grad_norm": 0.6208775729436938, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2710 + }, + { + "epoch": 0.02711, + "grad_norm": 0.6014094255252589, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 2711 + }, + { + "epoch": 0.02712, + "grad_norm": 0.5261948655268764, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2712 + }, + { + "epoch": 0.02713, + "grad_norm": 0.5550123847377345, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 2713 + }, + { + "epoch": 0.02714, + "grad_norm": 0.56736840403282, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2714 + }, + { + "epoch": 0.02715, + "grad_norm": 0.6628029062916901, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2715 + }, + { + "epoch": 0.02716, + "grad_norm": 0.768979209194684, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2716 + }, + { + "epoch": 0.02717, + "grad_norm": 0.7482884136468829, + "learning_rate": 0.003, + "loss": 4.193, + "step": 2717 + }, + { + "epoch": 0.02718, + "grad_norm": 0.8846409333789726, + "learning_rate": 0.003, + "loss": 4.219, + "step": 2718 + }, + { + "epoch": 0.02719, + "grad_norm": 0.9429345703060683, + "learning_rate": 0.003, + "loss": 4.2045, + "step": 2719 + }, + { + "epoch": 0.0272, + "grad_norm": 1.002947468833307, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2720 + }, + { + "epoch": 0.02721, + "grad_norm": 1.0695835852166422, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 2721 + }, + { + "epoch": 0.02722, + "grad_norm": 1.1056646889932085, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 2722 + }, + { + "epoch": 0.02723, + "grad_norm": 0.8267701861491207, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 2723 + }, + { + "epoch": 0.02724, + "grad_norm": 0.7367716189571131, + "learning_rate": 0.003, + "loss": 4.2292, + "step": 2724 + }, + { + "epoch": 0.02725, + "grad_norm": 0.6044896083683174, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 2725 + }, + { + "epoch": 0.02726, + "grad_norm": 0.6956030827952617, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2726 + }, + { + "epoch": 0.02727, + "grad_norm": 0.7192323257722564, + "learning_rate": 0.003, + "loss": 4.2197, + "step": 2727 + }, + { + "epoch": 0.02728, + "grad_norm": 0.7783208693346495, + "learning_rate": 0.003, + "loss": 4.2088, + "step": 2728 + }, + { + "epoch": 0.02729, + "grad_norm": 0.8318571752999154, + "learning_rate": 0.003, + "loss": 4.2427, + "step": 2729 + }, + { + "epoch": 0.0273, + "grad_norm": 0.6686152556603595, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 2730 + }, + { + "epoch": 0.02731, + "grad_norm": 0.6832226230920126, + "learning_rate": 0.003, + "loss": 4.211, + "step": 2731 + }, + { + "epoch": 0.02732, + "grad_norm": 0.6331150968875209, + "learning_rate": 0.003, + "loss": 4.1987, + "step": 2732 + }, + { + "epoch": 0.02733, + "grad_norm": 0.5278467341667037, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2733 + }, + { + "epoch": 0.02734, + "grad_norm": 0.6138764744804129, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 2734 + }, + { + "epoch": 0.02735, + "grad_norm": 0.561739809597342, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 2735 + }, + { + "epoch": 0.02736, + "grad_norm": 0.5908346453733819, + "learning_rate": 0.003, + "loss": 4.1928, + "step": 2736 + }, + { + "epoch": 0.02737, + "grad_norm": 0.6511959635659452, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 2737 + }, + { + "epoch": 0.02738, + "grad_norm": 0.758698109151286, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 2738 + }, + { + "epoch": 0.02739, + "grad_norm": 0.6956474809291019, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 2739 + }, + { + "epoch": 0.0274, + "grad_norm": 0.5058919640608095, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2740 + }, + { + "epoch": 0.02741, + "grad_norm": 0.5563232333292056, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2741 + }, + { + "epoch": 0.02742, + "grad_norm": 0.5987655354363043, + "learning_rate": 0.003, + "loss": 4.1885, + "step": 2742 + }, + { + "epoch": 0.02743, + "grad_norm": 0.5914696396737046, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 2743 + }, + { + "epoch": 0.02744, + "grad_norm": 0.5630038713206789, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2744 + }, + { + "epoch": 0.02745, + "grad_norm": 0.587542894365965, + "learning_rate": 0.003, + "loss": 4.2027, + "step": 2745 + }, + { + "epoch": 0.02746, + "grad_norm": 0.5934042861360144, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2746 + }, + { + "epoch": 0.02747, + "grad_norm": 0.6474787518036716, + "learning_rate": 0.003, + "loss": 4.212, + "step": 2747 + }, + { + "epoch": 0.02748, + "grad_norm": 0.8113625213803506, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 2748 + }, + { + "epoch": 0.02749, + "grad_norm": 0.8229694069124209, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2749 + }, + { + "epoch": 0.0275, + "grad_norm": 0.6793478072709392, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 2750 + }, + { + "epoch": 0.02751, + "grad_norm": 0.5656831593340759, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 2751 + }, + { + "epoch": 0.02752, + "grad_norm": 0.5267924527455705, + "learning_rate": 0.003, + "loss": 4.2017, + "step": 2752 + }, + { + "epoch": 0.02753, + "grad_norm": 0.502928202841188, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2753 + }, + { + "epoch": 0.02754, + "grad_norm": 0.5044833881071361, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2754 + }, + { + "epoch": 0.02755, + "grad_norm": 0.5267594677995416, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 2755 + }, + { + "epoch": 0.02756, + "grad_norm": 0.456921415827541, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 2756 + }, + { + "epoch": 0.02757, + "grad_norm": 0.3836455619810196, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 2757 + }, + { + "epoch": 0.02758, + "grad_norm": 0.34559498426529056, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 2758 + }, + { + "epoch": 0.02759, + "grad_norm": 0.34478709094038296, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2759 + }, + { + "epoch": 0.0276, + "grad_norm": 0.4284232121155303, + "learning_rate": 0.003, + "loss": 4.1802, + "step": 2760 + }, + { + "epoch": 0.02761, + "grad_norm": 0.5071377792443562, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2761 + }, + { + "epoch": 0.02762, + "grad_norm": 0.6171334423886827, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 2762 + }, + { + "epoch": 0.02763, + "grad_norm": 0.7563311828472695, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2763 + }, + { + "epoch": 0.02764, + "grad_norm": 0.8185996809603724, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 2764 + }, + { + "epoch": 0.02765, + "grad_norm": 0.7325003643978839, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 2765 + }, + { + "epoch": 0.02766, + "grad_norm": 0.6977141939643235, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2766 + }, + { + "epoch": 0.02767, + "grad_norm": 0.7178457591084105, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 2767 + }, + { + "epoch": 0.02768, + "grad_norm": 0.6603543944468468, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 2768 + }, + { + "epoch": 0.02769, + "grad_norm": 0.5908808640128104, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 2769 + }, + { + "epoch": 0.0277, + "grad_norm": 0.5584733753440203, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 2770 + }, + { + "epoch": 0.02771, + "grad_norm": 0.5513412132170109, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2771 + }, + { + "epoch": 0.02772, + "grad_norm": 0.601007172067716, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 2772 + }, + { + "epoch": 0.02773, + "grad_norm": 0.6222859555759426, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 2773 + }, + { + "epoch": 0.02774, + "grad_norm": 0.6298455082538394, + "learning_rate": 0.003, + "loss": 4.2142, + "step": 2774 + }, + { + "epoch": 0.02775, + "grad_norm": 0.6384186594957891, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 2775 + }, + { + "epoch": 0.02776, + "grad_norm": 0.7183697808219275, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2776 + }, + { + "epoch": 0.02777, + "grad_norm": 0.6932591793310324, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 2777 + }, + { + "epoch": 0.02778, + "grad_norm": 0.6667113422946584, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 2778 + }, + { + "epoch": 0.02779, + "grad_norm": 0.7407795606572514, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 2779 + }, + { + "epoch": 0.0278, + "grad_norm": 0.7295794698337943, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2780 + }, + { + "epoch": 0.02781, + "grad_norm": 0.6893914165082023, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2781 + }, + { + "epoch": 0.02782, + "grad_norm": 0.6620017500188085, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2782 + }, + { + "epoch": 0.02783, + "grad_norm": 0.6131468932443346, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 2783 + }, + { + "epoch": 0.02784, + "grad_norm": 0.5579371518365059, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2784 + }, + { + "epoch": 0.02785, + "grad_norm": 0.49723848657236286, + "learning_rate": 0.003, + "loss": 4.192, + "step": 2785 + }, + { + "epoch": 0.02786, + "grad_norm": 0.5125002182675953, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 2786 + }, + { + "epoch": 0.02787, + "grad_norm": 0.4914565242864978, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 2787 + }, + { + "epoch": 0.02788, + "grad_norm": 0.526484975930702, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2788 + }, + { + "epoch": 0.02789, + "grad_norm": 0.5321807051700945, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2789 + }, + { + "epoch": 0.0279, + "grad_norm": 0.5840495108099473, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2790 + }, + { + "epoch": 0.02791, + "grad_norm": 0.9202391536169947, + "learning_rate": 0.003, + "loss": 4.177, + "step": 2791 + }, + { + "epoch": 0.02792, + "grad_norm": 1.2841478556987087, + "learning_rate": 0.003, + "loss": 4.2354, + "step": 2792 + }, + { + "epoch": 0.02793, + "grad_norm": 0.6986972433822171, + "learning_rate": 0.003, + "loss": 4.2026, + "step": 2793 + }, + { + "epoch": 0.02794, + "grad_norm": 0.7400669972097395, + "learning_rate": 0.003, + "loss": 4.2072, + "step": 2794 + }, + { + "epoch": 0.02795, + "grad_norm": 0.7145362265408897, + "learning_rate": 0.003, + "loss": 4.2139, + "step": 2795 + }, + { + "epoch": 0.02796, + "grad_norm": 0.7481134935341638, + "learning_rate": 0.003, + "loss": 4.2118, + "step": 2796 + }, + { + "epoch": 0.02797, + "grad_norm": 0.8256336591556148, + "learning_rate": 0.003, + "loss": 4.2281, + "step": 2797 + }, + { + "epoch": 0.02798, + "grad_norm": 0.7297049193325166, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 2798 + }, + { + "epoch": 0.02799, + "grad_norm": 0.7233108511585719, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 2799 + }, + { + "epoch": 0.028, + "grad_norm": 0.6743448137124989, + "learning_rate": 0.003, + "loss": 4.2137, + "step": 2800 + }, + { + "epoch": 0.02801, + "grad_norm": 0.5867417428513181, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2801 + }, + { + "epoch": 0.02802, + "grad_norm": 0.557083005494101, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 2802 + }, + { + "epoch": 0.02803, + "grad_norm": 0.5239629190819096, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2803 + }, + { + "epoch": 0.02804, + "grad_norm": 0.6122062528440474, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 2804 + }, + { + "epoch": 0.02805, + "grad_norm": 0.6123404104585692, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2805 + }, + { + "epoch": 0.02806, + "grad_norm": 0.5813003741795452, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2806 + }, + { + "epoch": 0.02807, + "grad_norm": 0.555658293190141, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2807 + }, + { + "epoch": 0.02808, + "grad_norm": 0.4856792434285324, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 2808 + }, + { + "epoch": 0.02809, + "grad_norm": 0.468227745280838, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 2809 + }, + { + "epoch": 0.0281, + "grad_norm": 0.5953915884658988, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 2810 + }, + { + "epoch": 0.02811, + "grad_norm": 0.9133221862875325, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 2811 + }, + { + "epoch": 0.02812, + "grad_norm": 1.275925834959494, + "learning_rate": 0.003, + "loss": 4.2304, + "step": 2812 + }, + { + "epoch": 0.02813, + "grad_norm": 0.8483651283891129, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2813 + }, + { + "epoch": 0.02814, + "grad_norm": 0.7605354432635538, + "learning_rate": 0.003, + "loss": 4.202, + "step": 2814 + }, + { + "epoch": 0.02815, + "grad_norm": 0.7567043943015741, + "learning_rate": 0.003, + "loss": 4.2385, + "step": 2815 + }, + { + "epoch": 0.02816, + "grad_norm": 0.7448943831936015, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2816 + }, + { + "epoch": 0.02817, + "grad_norm": 0.7400121216625585, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2817 + }, + { + "epoch": 0.02818, + "grad_norm": 0.7612379755026737, + "learning_rate": 0.003, + "loss": 4.1838, + "step": 2818 + }, + { + "epoch": 0.02819, + "grad_norm": 0.8916358831666468, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 2819 + }, + { + "epoch": 0.0282, + "grad_norm": 0.9386399567045983, + "learning_rate": 0.003, + "loss": 4.2191, + "step": 2820 + }, + { + "epoch": 0.02821, + "grad_norm": 0.7968339375438817, + "learning_rate": 0.003, + "loss": 4.2369, + "step": 2821 + }, + { + "epoch": 0.02822, + "grad_norm": 0.6707514048211725, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2822 + }, + { + "epoch": 0.02823, + "grad_norm": 0.5937548176700269, + "learning_rate": 0.003, + "loss": 4.2229, + "step": 2823 + }, + { + "epoch": 0.02824, + "grad_norm": 0.6160995840345511, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 2824 + }, + { + "epoch": 0.02825, + "grad_norm": 0.5897186283971302, + "learning_rate": 0.003, + "loss": 4.2084, + "step": 2825 + }, + { + "epoch": 0.02826, + "grad_norm": 0.5674030456447229, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 2826 + }, + { + "epoch": 0.02827, + "grad_norm": 0.5509162115210677, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 2827 + }, + { + "epoch": 0.02828, + "grad_norm": 0.5443211688206793, + "learning_rate": 0.003, + "loss": 4.208, + "step": 2828 + }, + { + "epoch": 0.02829, + "grad_norm": 0.5032331740089974, + "learning_rate": 0.003, + "loss": 4.227, + "step": 2829 + }, + { + "epoch": 0.0283, + "grad_norm": 0.5225637902550168, + "learning_rate": 0.003, + "loss": 4.1889, + "step": 2830 + }, + { + "epoch": 0.02831, + "grad_norm": 0.5311323075927029, + "learning_rate": 0.003, + "loss": 4.1907, + "step": 2831 + }, + { + "epoch": 0.02832, + "grad_norm": 0.5001952573238827, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2832 + }, + { + "epoch": 0.02833, + "grad_norm": 0.466662882413672, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 2833 + }, + { + "epoch": 0.02834, + "grad_norm": 0.4772870203488476, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 2834 + }, + { + "epoch": 0.02835, + "grad_norm": 0.5535871706800527, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2835 + }, + { + "epoch": 0.02836, + "grad_norm": 0.7239867170041147, + "learning_rate": 0.003, + "loss": 4.2004, + "step": 2836 + }, + { + "epoch": 0.02837, + "grad_norm": 0.7597045607127227, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2837 + }, + { + "epoch": 0.02838, + "grad_norm": 0.7252587205241631, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 2838 + }, + { + "epoch": 0.02839, + "grad_norm": 0.8402252149672428, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2839 + }, + { + "epoch": 0.0284, + "grad_norm": 0.8241726365065422, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 2840 + }, + { + "epoch": 0.02841, + "grad_norm": 0.8096094003742889, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 2841 + }, + { + "epoch": 0.02842, + "grad_norm": 0.7156086401419698, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2842 + }, + { + "epoch": 0.02843, + "grad_norm": 0.6040106819890795, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 2843 + }, + { + "epoch": 0.02844, + "grad_norm": 0.5060645820894467, + "learning_rate": 0.003, + "loss": 4.194, + "step": 2844 + }, + { + "epoch": 0.02845, + "grad_norm": 0.5041938926420962, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2845 + }, + { + "epoch": 0.02846, + "grad_norm": 0.5025147774215044, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 2846 + }, + { + "epoch": 0.02847, + "grad_norm": 0.5625355648833802, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 2847 + }, + { + "epoch": 0.02848, + "grad_norm": 0.6048155911479377, + "learning_rate": 0.003, + "loss": 4.1823, + "step": 2848 + }, + { + "epoch": 0.02849, + "grad_norm": 0.593616598386133, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 2849 + }, + { + "epoch": 0.0285, + "grad_norm": 0.5421469421547838, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2850 + }, + { + "epoch": 0.02851, + "grad_norm": 0.5145785466987678, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 2851 + }, + { + "epoch": 0.02852, + "grad_norm": 0.5364202493290657, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 2852 + }, + { + "epoch": 0.02853, + "grad_norm": 0.5092167845332397, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 2853 + }, + { + "epoch": 0.02854, + "grad_norm": 0.505959030155662, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 2854 + }, + { + "epoch": 0.02855, + "grad_norm": 0.48727747027775575, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 2855 + }, + { + "epoch": 0.02856, + "grad_norm": 0.5114663663985488, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2856 + }, + { + "epoch": 0.02857, + "grad_norm": 0.6029175042652142, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2857 + }, + { + "epoch": 0.02858, + "grad_norm": 0.811258279718228, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2858 + }, + { + "epoch": 0.02859, + "grad_norm": 1.000140796502162, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2859 + }, + { + "epoch": 0.0286, + "grad_norm": 1.2449496760884864, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 2860 + }, + { + "epoch": 0.02861, + "grad_norm": 0.7665309921680512, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2861 + }, + { + "epoch": 0.02862, + "grad_norm": 0.7010737782221613, + "learning_rate": 0.003, + "loss": 4.1959, + "step": 2862 + }, + { + "epoch": 0.02863, + "grad_norm": 0.6652925807236452, + "learning_rate": 0.003, + "loss": 4.216, + "step": 2863 + }, + { + "epoch": 0.02864, + "grad_norm": 0.6660638892873599, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2864 + }, + { + "epoch": 0.02865, + "grad_norm": 0.7030991193559302, + "learning_rate": 0.003, + "loss": 4.2231, + "step": 2865 + }, + { + "epoch": 0.02866, + "grad_norm": 0.6597702895413415, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 2866 + }, + { + "epoch": 0.02867, + "grad_norm": 0.582862490658125, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 2867 + }, + { + "epoch": 0.02868, + "grad_norm": 0.5208311845539698, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 2868 + }, + { + "epoch": 0.02869, + "grad_norm": 0.5269361881811672, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2869 + }, + { + "epoch": 0.0287, + "grad_norm": 0.49588966178431665, + "learning_rate": 0.003, + "loss": 4.1982, + "step": 2870 + }, + { + "epoch": 0.02871, + "grad_norm": 0.4427797615796878, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2871 + }, + { + "epoch": 0.02872, + "grad_norm": 0.46354755259563285, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2872 + }, + { + "epoch": 0.02873, + "grad_norm": 0.5003897365609394, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2873 + }, + { + "epoch": 0.02874, + "grad_norm": 0.5095674852487324, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 2874 + }, + { + "epoch": 0.02875, + "grad_norm": 0.5767552596221609, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 2875 + }, + { + "epoch": 0.02876, + "grad_norm": 0.7553686836410436, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2876 + }, + { + "epoch": 0.02877, + "grad_norm": 0.927267465496802, + "learning_rate": 0.003, + "loss": 4.181, + "step": 2877 + }, + { + "epoch": 0.02878, + "grad_norm": 0.8275286861154829, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2878 + }, + { + "epoch": 0.02879, + "grad_norm": 0.6254369579498115, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2879 + }, + { + "epoch": 0.0288, + "grad_norm": 0.6608202030691483, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2880 + }, + { + "epoch": 0.02881, + "grad_norm": 0.6264026993054118, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2881 + }, + { + "epoch": 0.02882, + "grad_norm": 0.6456049764969926, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 2882 + }, + { + "epoch": 0.02883, + "grad_norm": 0.631140439713875, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2883 + }, + { + "epoch": 0.02884, + "grad_norm": 0.6167258778841849, + "learning_rate": 0.003, + "loss": 4.2212, + "step": 2884 + }, + { + "epoch": 0.02885, + "grad_norm": 0.6237538943365489, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 2885 + }, + { + "epoch": 0.02886, + "grad_norm": 0.577915341916395, + "learning_rate": 0.003, + "loss": 4.1922, + "step": 2886 + }, + { + "epoch": 0.02887, + "grad_norm": 0.5869716498143316, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 2887 + }, + { + "epoch": 0.02888, + "grad_norm": 0.6367834629014674, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2888 + }, + { + "epoch": 0.02889, + "grad_norm": 0.6424828882281004, + "learning_rate": 0.003, + "loss": 4.167, + "step": 2889 + }, + { + "epoch": 0.0289, + "grad_norm": 0.7456946160039144, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2890 + }, + { + "epoch": 0.02891, + "grad_norm": 0.8611032962542791, + "learning_rate": 0.003, + "loss": 4.2268, + "step": 2891 + }, + { + "epoch": 0.02892, + "grad_norm": 1.0298003752084743, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 2892 + }, + { + "epoch": 0.02893, + "grad_norm": 1.0054991154141122, + "learning_rate": 0.003, + "loss": 4.1976, + "step": 2893 + }, + { + "epoch": 0.02894, + "grad_norm": 0.8964526887398983, + "learning_rate": 0.003, + "loss": 4.2137, + "step": 2894 + }, + { + "epoch": 0.02895, + "grad_norm": 0.786662499637371, + "learning_rate": 0.003, + "loss": 4.2494, + "step": 2895 + }, + { + "epoch": 0.02896, + "grad_norm": 0.8924645045032006, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 2896 + }, + { + "epoch": 0.02897, + "grad_norm": 0.9444560121857192, + "learning_rate": 0.003, + "loss": 4.1979, + "step": 2897 + }, + { + "epoch": 0.02898, + "grad_norm": 0.9041547595653124, + "learning_rate": 0.003, + "loss": 4.2475, + "step": 2898 + }, + { + "epoch": 0.02899, + "grad_norm": 0.8479058583034529, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 2899 + }, + { + "epoch": 0.029, + "grad_norm": 0.8224886244859558, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2900 + }, + { + "epoch": 0.02901, + "grad_norm": 0.8698134304184088, + "learning_rate": 0.003, + "loss": 4.2375, + "step": 2901 + }, + { + "epoch": 0.02902, + "grad_norm": 0.8486827254456236, + "learning_rate": 0.003, + "loss": 4.2524, + "step": 2902 + }, + { + "epoch": 0.02903, + "grad_norm": 0.7415128492690074, + "learning_rate": 0.003, + "loss": 4.2059, + "step": 2903 + }, + { + "epoch": 0.02904, + "grad_norm": 0.7083318227885205, + "learning_rate": 0.003, + "loss": 4.2241, + "step": 2904 + }, + { + "epoch": 0.02905, + "grad_norm": 0.7648353967865369, + "learning_rate": 0.003, + "loss": 4.2383, + "step": 2905 + }, + { + "epoch": 0.02906, + "grad_norm": 0.7593644185333667, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 2906 + }, + { + "epoch": 0.02907, + "grad_norm": 0.7189838848454702, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 2907 + }, + { + "epoch": 0.02908, + "grad_norm": 0.7101023301849051, + "learning_rate": 0.003, + "loss": 4.2105, + "step": 2908 + }, + { + "epoch": 0.02909, + "grad_norm": 0.6273709094950615, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 2909 + }, + { + "epoch": 0.0291, + "grad_norm": 0.5304669195124173, + "learning_rate": 0.003, + "loss": 4.2082, + "step": 2910 + }, + { + "epoch": 0.02911, + "grad_norm": 0.509577660549234, + "learning_rate": 0.003, + "loss": 4.2021, + "step": 2911 + }, + { + "epoch": 0.02912, + "grad_norm": 0.4255206741887875, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 2912 + }, + { + "epoch": 0.02913, + "grad_norm": 0.4561469328827923, + "learning_rate": 0.003, + "loss": 4.1998, + "step": 2913 + }, + { + "epoch": 0.02914, + "grad_norm": 0.42117739519724395, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2914 + }, + { + "epoch": 0.02915, + "grad_norm": 0.4011573145594706, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 2915 + }, + { + "epoch": 0.02916, + "grad_norm": 0.42970374324262517, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 2916 + }, + { + "epoch": 0.02917, + "grad_norm": 0.33514332187348916, + "learning_rate": 0.003, + "loss": 4.2087, + "step": 2917 + }, + { + "epoch": 0.02918, + "grad_norm": 0.34222241717451907, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2918 + }, + { + "epoch": 0.02919, + "grad_norm": 0.36583286994162584, + "learning_rate": 0.003, + "loss": 4.2045, + "step": 2919 + }, + { + "epoch": 0.0292, + "grad_norm": 0.4446133039291897, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2920 + }, + { + "epoch": 0.02921, + "grad_norm": 0.6109713288069767, + "learning_rate": 0.003, + "loss": 4.1869, + "step": 2921 + }, + { + "epoch": 0.02922, + "grad_norm": 0.9167859634758244, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 2922 + }, + { + "epoch": 0.02923, + "grad_norm": 1.009808447907654, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 2923 + }, + { + "epoch": 0.02924, + "grad_norm": 0.8524826722350543, + "learning_rate": 0.003, + "loss": 4.2159, + "step": 2924 + }, + { + "epoch": 0.02925, + "grad_norm": 0.8032284343679224, + "learning_rate": 0.003, + "loss": 4.2199, + "step": 2925 + }, + { + "epoch": 0.02926, + "grad_norm": 0.749692494912656, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2926 + }, + { + "epoch": 0.02927, + "grad_norm": 0.7088429283617566, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 2927 + }, + { + "epoch": 0.02928, + "grad_norm": 0.6042074975753539, + "learning_rate": 0.003, + "loss": 4.2151, + "step": 2928 + }, + { + "epoch": 0.02929, + "grad_norm": 0.563152730036225, + "learning_rate": 0.003, + "loss": 4.2, + "step": 2929 + }, + { + "epoch": 0.0293, + "grad_norm": 0.44275110512836524, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2930 + }, + { + "epoch": 0.02931, + "grad_norm": 0.4579944440403804, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2931 + }, + { + "epoch": 0.02932, + "grad_norm": 0.4515235240773822, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 2932 + }, + { + "epoch": 0.02933, + "grad_norm": 0.46401883936201904, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2933 + }, + { + "epoch": 0.02934, + "grad_norm": 0.4813031783733538, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 2934 + }, + { + "epoch": 0.02935, + "grad_norm": 0.4967851984899721, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2935 + }, + { + "epoch": 0.02936, + "grad_norm": 0.5320015979752634, + "learning_rate": 0.003, + "loss": 4.189, + "step": 2936 + }, + { + "epoch": 0.02937, + "grad_norm": 0.5208542547284235, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 2937 + }, + { + "epoch": 0.02938, + "grad_norm": 0.4869610046256959, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 2938 + }, + { + "epoch": 0.02939, + "grad_norm": 0.43982195879012664, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 2939 + }, + { + "epoch": 0.0294, + "grad_norm": 0.43165943569371323, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 2940 + }, + { + "epoch": 0.02941, + "grad_norm": 0.42317517860985626, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2941 + }, + { + "epoch": 0.02942, + "grad_norm": 0.46023364048218307, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2942 + }, + { + "epoch": 0.02943, + "grad_norm": 0.5385981811058449, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2943 + }, + { + "epoch": 0.02944, + "grad_norm": 0.7649280046904252, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2944 + }, + { + "epoch": 0.02945, + "grad_norm": 1.1539036490870935, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 2945 + }, + { + "epoch": 0.02946, + "grad_norm": 0.9395984374200569, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2946 + }, + { + "epoch": 0.02947, + "grad_norm": 0.8080599605170256, + "learning_rate": 0.003, + "loss": 4.224, + "step": 2947 + }, + { + "epoch": 0.02948, + "grad_norm": 0.7434332174406536, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 2948 + }, + { + "epoch": 0.02949, + "grad_norm": 0.7928578212224346, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 2949 + }, + { + "epoch": 0.0295, + "grad_norm": 0.8333301095683224, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 2950 + }, + { + "epoch": 0.02951, + "grad_norm": 0.9482608719829599, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 2951 + }, + { + "epoch": 0.02952, + "grad_norm": 0.8328172653762101, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2952 + }, + { + "epoch": 0.02953, + "grad_norm": 0.8348300580576069, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 2953 + }, + { + "epoch": 0.02954, + "grad_norm": 0.8191965718315765, + "learning_rate": 0.003, + "loss": 4.2499, + "step": 2954 + }, + { + "epoch": 0.02955, + "grad_norm": 0.8592579474886709, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 2955 + }, + { + "epoch": 0.02956, + "grad_norm": 0.768153733637731, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 2956 + }, + { + "epoch": 0.02957, + "grad_norm": 0.619447070815721, + "learning_rate": 0.003, + "loss": 4.2157, + "step": 2957 + }, + { + "epoch": 0.02958, + "grad_norm": 0.6117541520949576, + "learning_rate": 0.003, + "loss": 4.2284, + "step": 2958 + }, + { + "epoch": 0.02959, + "grad_norm": 0.5765518020385798, + "learning_rate": 0.003, + "loss": 4.2146, + "step": 2959 + }, + { + "epoch": 0.0296, + "grad_norm": 0.5472369901466232, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 2960 + }, + { + "epoch": 0.02961, + "grad_norm": 0.5697675686509917, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2961 + }, + { + "epoch": 0.02962, + "grad_norm": 0.5322939405778847, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 2962 + }, + { + "epoch": 0.02963, + "grad_norm": 0.4995062550237162, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 2963 + }, + { + "epoch": 0.02964, + "grad_norm": 0.48408735290928806, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 2964 + }, + { + "epoch": 0.02965, + "grad_norm": 0.49796931779139914, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 2965 + }, + { + "epoch": 0.02966, + "grad_norm": 0.49309681218010465, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 2966 + }, + { + "epoch": 0.02967, + "grad_norm": 0.46885663161103275, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 2967 + }, + { + "epoch": 0.02968, + "grad_norm": 0.4307711779753192, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2968 + }, + { + "epoch": 0.02969, + "grad_norm": 0.38466337051238414, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 2969 + }, + { + "epoch": 0.0297, + "grad_norm": 0.46985591464325377, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 2970 + }, + { + "epoch": 0.02971, + "grad_norm": 0.6319423638686542, + "learning_rate": 0.003, + "loss": 4.186, + "step": 2971 + }, + { + "epoch": 0.02972, + "grad_norm": 0.8754437886538992, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 2972 + }, + { + "epoch": 0.02973, + "grad_norm": 1.020941685436833, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 2973 + }, + { + "epoch": 0.02974, + "grad_norm": 0.6824462354167666, + "learning_rate": 0.003, + "loss": 4.193, + "step": 2974 + }, + { + "epoch": 0.02975, + "grad_norm": 0.5456820041365307, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2975 + }, + { + "epoch": 0.02976, + "grad_norm": 0.7254587394041584, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 2976 + }, + { + "epoch": 0.02977, + "grad_norm": 0.6896012196641018, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2977 + }, + { + "epoch": 0.02978, + "grad_norm": 0.6446956558129, + "learning_rate": 0.003, + "loss": 4.1956, + "step": 2978 + }, + { + "epoch": 0.02979, + "grad_norm": 0.5613291446841128, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 2979 + }, + { + "epoch": 0.0298, + "grad_norm": 0.578998122498601, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2980 + }, + { + "epoch": 0.02981, + "grad_norm": 0.6195176572378017, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2981 + }, + { + "epoch": 0.02982, + "grad_norm": 0.6240562576963727, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 2982 + }, + { + "epoch": 0.02983, + "grad_norm": 0.5968266153844792, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2983 + }, + { + "epoch": 0.02984, + "grad_norm": 0.5306795307358292, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2984 + }, + { + "epoch": 0.02985, + "grad_norm": 0.5463323846035709, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 2985 + }, + { + "epoch": 0.02986, + "grad_norm": 0.5421781401610759, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 2986 + }, + { + "epoch": 0.02987, + "grad_norm": 0.5293611287733503, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 2987 + }, + { + "epoch": 0.02988, + "grad_norm": 0.6408997854011281, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 2988 + }, + { + "epoch": 0.02989, + "grad_norm": 0.7426268122761913, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 2989 + }, + { + "epoch": 0.0299, + "grad_norm": 0.7556899506663288, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 2990 + }, + { + "epoch": 0.02991, + "grad_norm": 0.6415270525884926, + "learning_rate": 0.003, + "loss": 4.1985, + "step": 2991 + }, + { + "epoch": 0.02992, + "grad_norm": 0.6368634341681172, + "learning_rate": 0.003, + "loss": 4.1951, + "step": 2992 + }, + { + "epoch": 0.02993, + "grad_norm": 0.6410538934602643, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2993 + }, + { + "epoch": 0.02994, + "grad_norm": 0.6332716830387952, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 2994 + }, + { + "epoch": 0.02995, + "grad_norm": 0.7865491161110946, + "learning_rate": 0.003, + "loss": 4.2175, + "step": 2995 + }, + { + "epoch": 0.02996, + "grad_norm": 0.9230963051423904, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2996 + }, + { + "epoch": 0.02997, + "grad_norm": 1.1769719664750078, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 2997 + }, + { + "epoch": 0.02998, + "grad_norm": 0.965735521679244, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 2998 + }, + { + "epoch": 0.02999, + "grad_norm": 0.9025433565090871, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 2999 + }, + { + "epoch": 0.03, + "grad_norm": 0.9697259711524333, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 3000 + }, + { + "epoch": 0.03001, + "grad_norm": 0.9254213834179852, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 3001 + }, + { + "epoch": 0.03002, + "grad_norm": 0.8893198223091714, + "learning_rate": 0.003, + "loss": 4.225, + "step": 3002 + }, + { + "epoch": 0.03003, + "grad_norm": 0.7192875473487559, + "learning_rate": 0.003, + "loss": 4.197, + "step": 3003 + }, + { + "epoch": 0.03004, + "grad_norm": 0.7886569353807684, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 3004 + }, + { + "epoch": 0.03005, + "grad_norm": 0.8011423250434415, + "learning_rate": 0.003, + "loss": 4.2294, + "step": 3005 + }, + { + "epoch": 0.03006, + "grad_norm": 0.7535404952002084, + "learning_rate": 0.003, + "loss": 4.2201, + "step": 3006 + }, + { + "epoch": 0.03007, + "grad_norm": 0.7314424724357438, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 3007 + }, + { + "epoch": 0.03008, + "grad_norm": 0.6096876125818365, + "learning_rate": 0.003, + "loss": 4.215, + "step": 3008 + }, + { + "epoch": 0.03009, + "grad_norm": 0.5575234963619043, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 3009 + }, + { + "epoch": 0.0301, + "grad_norm": 0.5598489300006476, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 3010 + }, + { + "epoch": 0.03011, + "grad_norm": 0.7310511801710661, + "learning_rate": 0.003, + "loss": 4.1885, + "step": 3011 + }, + { + "epoch": 0.03012, + "grad_norm": 0.8536268304907085, + "learning_rate": 0.003, + "loss": 4.2002, + "step": 3012 + }, + { + "epoch": 0.03013, + "grad_norm": 0.9271684945658873, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 3013 + }, + { + "epoch": 0.03014, + "grad_norm": 0.748204257953396, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 3014 + }, + { + "epoch": 0.03015, + "grad_norm": 0.7294374875175993, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 3015 + }, + { + "epoch": 0.03016, + "grad_norm": 0.8462639497588901, + "learning_rate": 0.003, + "loss": 4.204, + "step": 3016 + }, + { + "epoch": 0.03017, + "grad_norm": 0.6499456234415616, + "learning_rate": 0.003, + "loss": 4.189, + "step": 3017 + }, + { + "epoch": 0.03018, + "grad_norm": 0.5875121164816828, + "learning_rate": 0.003, + "loss": 4.1824, + "step": 3018 + }, + { + "epoch": 0.03019, + "grad_norm": 0.47508344745103015, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 3019 + }, + { + "epoch": 0.0302, + "grad_norm": 0.5142366251911704, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 3020 + }, + { + "epoch": 0.03021, + "grad_norm": 0.5290255986656376, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 3021 + }, + { + "epoch": 0.03022, + "grad_norm": 0.5261392358925515, + "learning_rate": 0.003, + "loss": 4.2026, + "step": 3022 + }, + { + "epoch": 0.03023, + "grad_norm": 0.5585956938981086, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 3023 + }, + { + "epoch": 0.03024, + "grad_norm": 0.6114817871124547, + "learning_rate": 0.003, + "loss": 4.2096, + "step": 3024 + }, + { + "epoch": 0.03025, + "grad_norm": 0.6824857482246931, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 3025 + }, + { + "epoch": 0.03026, + "grad_norm": 0.7984707065148595, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 3026 + }, + { + "epoch": 0.03027, + "grad_norm": 0.7174251636707903, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 3027 + }, + { + "epoch": 0.03028, + "grad_norm": 0.7106059576910744, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3028 + }, + { + "epoch": 0.03029, + "grad_norm": 0.6122719077257656, + "learning_rate": 0.003, + "loss": 4.1927, + "step": 3029 + }, + { + "epoch": 0.0303, + "grad_norm": 0.5305954769529888, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3030 + }, + { + "epoch": 0.03031, + "grad_norm": 0.4997654568254282, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3031 + }, + { + "epoch": 0.03032, + "grad_norm": 0.4912614975804802, + "learning_rate": 0.003, + "loss": 4.178, + "step": 3032 + }, + { + "epoch": 0.03033, + "grad_norm": 0.4920559211009805, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3033 + }, + { + "epoch": 0.03034, + "grad_norm": 0.5605701380054549, + "learning_rate": 0.003, + "loss": 4.1747, + "step": 3034 + }, + { + "epoch": 0.03035, + "grad_norm": 0.6077583896543486, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 3035 + }, + { + "epoch": 0.03036, + "grad_norm": 0.7202855019986794, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 3036 + }, + { + "epoch": 0.03037, + "grad_norm": 0.6575890137997805, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 3037 + }, + { + "epoch": 0.03038, + "grad_norm": 0.46568815919813755, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 3038 + }, + { + "epoch": 0.03039, + "grad_norm": 0.49641007446862917, + "learning_rate": 0.003, + "loss": 4.166, + "step": 3039 + }, + { + "epoch": 0.0304, + "grad_norm": 0.559012034381986, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 3040 + }, + { + "epoch": 0.03041, + "grad_norm": 0.5533954561872954, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 3041 + }, + { + "epoch": 0.03042, + "grad_norm": 0.47365032322165673, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 3042 + }, + { + "epoch": 0.03043, + "grad_norm": 0.4920139603559505, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 3043 + }, + { + "epoch": 0.03044, + "grad_norm": 0.48267895485751067, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 3044 + }, + { + "epoch": 0.03045, + "grad_norm": 0.40892249484891385, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 3045 + }, + { + "epoch": 0.03046, + "grad_norm": 0.5179956198309833, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 3046 + }, + { + "epoch": 0.03047, + "grad_norm": 0.6458691485842255, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3047 + }, + { + "epoch": 0.03048, + "grad_norm": 0.805749623290173, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3048 + }, + { + "epoch": 0.03049, + "grad_norm": 0.8776696886409071, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 3049 + }, + { + "epoch": 0.0305, + "grad_norm": 0.8822405330269477, + "learning_rate": 0.003, + "loss": 4.1854, + "step": 3050 + }, + { + "epoch": 0.03051, + "grad_norm": 0.9277197792519675, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 3051 + }, + { + "epoch": 0.03052, + "grad_norm": 0.9287607223949175, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 3052 + }, + { + "epoch": 0.03053, + "grad_norm": 0.8621618117255709, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 3053 + }, + { + "epoch": 0.03054, + "grad_norm": 0.9192475292904252, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 3054 + }, + { + "epoch": 0.03055, + "grad_norm": 0.9250509957208142, + "learning_rate": 0.003, + "loss": 4.2062, + "step": 3055 + }, + { + "epoch": 0.03056, + "grad_norm": 0.9134243567876251, + "learning_rate": 0.003, + "loss": 4.217, + "step": 3056 + }, + { + "epoch": 0.03057, + "grad_norm": 1.0483835943699353, + "learning_rate": 0.003, + "loss": 4.2258, + "step": 3057 + }, + { + "epoch": 0.03058, + "grad_norm": 1.0909797627134183, + "learning_rate": 0.003, + "loss": 4.2072, + "step": 3058 + }, + { + "epoch": 0.03059, + "grad_norm": 1.0929367898362812, + "learning_rate": 0.003, + "loss": 4.2194, + "step": 3059 + }, + { + "epoch": 0.0306, + "grad_norm": 0.9499317872652826, + "learning_rate": 0.003, + "loss": 4.2104, + "step": 3060 + }, + { + "epoch": 0.03061, + "grad_norm": 0.7530643446613368, + "learning_rate": 0.003, + "loss": 4.227, + "step": 3061 + }, + { + "epoch": 0.03062, + "grad_norm": 0.7822441045416462, + "learning_rate": 0.003, + "loss": 4.25, + "step": 3062 + }, + { + "epoch": 0.03063, + "grad_norm": 0.7991934606488392, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 3063 + }, + { + "epoch": 0.03064, + "grad_norm": 0.7614903963432643, + "learning_rate": 0.003, + "loss": 4.2217, + "step": 3064 + }, + { + "epoch": 0.03065, + "grad_norm": 0.6047999858157871, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3065 + }, + { + "epoch": 0.03066, + "grad_norm": 0.6254938393897825, + "learning_rate": 0.003, + "loss": 4.231, + "step": 3066 + }, + { + "epoch": 0.03067, + "grad_norm": 0.5333801571885867, + "learning_rate": 0.003, + "loss": 4.2275, + "step": 3067 + }, + { + "epoch": 0.03068, + "grad_norm": 0.47159284158763254, + "learning_rate": 0.003, + "loss": 4.1942, + "step": 3068 + }, + { + "epoch": 0.03069, + "grad_norm": 0.4095985602684375, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 3069 + }, + { + "epoch": 0.0307, + "grad_norm": 0.4380075531913516, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 3070 + }, + { + "epoch": 0.03071, + "grad_norm": 0.520842979900822, + "learning_rate": 0.003, + "loss": 4.181, + "step": 3071 + }, + { + "epoch": 0.03072, + "grad_norm": 0.6276268407397474, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 3072 + }, + { + "epoch": 0.03073, + "grad_norm": 0.7317184572201801, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 3073 + }, + { + "epoch": 0.03074, + "grad_norm": 0.7701795105904193, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 3074 + }, + { + "epoch": 0.03075, + "grad_norm": 0.6957977102957552, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 3075 + }, + { + "epoch": 0.03076, + "grad_norm": 0.6144911834647919, + "learning_rate": 0.003, + "loss": 4.2033, + "step": 3076 + }, + { + "epoch": 0.03077, + "grad_norm": 0.692008332614251, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 3077 + }, + { + "epoch": 0.03078, + "grad_norm": 0.6946737312829024, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 3078 + }, + { + "epoch": 0.03079, + "grad_norm": 0.7552084198206422, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3079 + }, + { + "epoch": 0.0308, + "grad_norm": 0.7025348126305976, + "learning_rate": 0.003, + "loss": 4.203, + "step": 3080 + }, + { + "epoch": 0.03081, + "grad_norm": 0.5947816851076059, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 3081 + }, + { + "epoch": 0.03082, + "grad_norm": 0.6508903175127416, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3082 + }, + { + "epoch": 0.03083, + "grad_norm": 0.8364672171654759, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3083 + }, + { + "epoch": 0.03084, + "grad_norm": 0.7270205647692827, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 3084 + }, + { + "epoch": 0.03085, + "grad_norm": 0.5251957082188036, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 3085 + }, + { + "epoch": 0.03086, + "grad_norm": 0.4928794972094713, + "learning_rate": 0.003, + "loss": 4.2012, + "step": 3086 + }, + { + "epoch": 0.03087, + "grad_norm": 0.4574857190743631, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 3087 + }, + { + "epoch": 0.03088, + "grad_norm": 0.48967300167720096, + "learning_rate": 0.003, + "loss": 4.19, + "step": 3088 + }, + { + "epoch": 0.03089, + "grad_norm": 0.5020327161233649, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3089 + }, + { + "epoch": 0.0309, + "grad_norm": 0.5318564623713369, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 3090 + }, + { + "epoch": 0.03091, + "grad_norm": 0.5321720354076895, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 3091 + }, + { + "epoch": 0.03092, + "grad_norm": 0.45641360749571647, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 3092 + }, + { + "epoch": 0.03093, + "grad_norm": 0.40991320622723604, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3093 + }, + { + "epoch": 0.03094, + "grad_norm": 0.4792246990497778, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 3094 + }, + { + "epoch": 0.03095, + "grad_norm": 0.5559515338310668, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 3095 + }, + { + "epoch": 0.03096, + "grad_norm": 0.7462971856723287, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 3096 + }, + { + "epoch": 0.03097, + "grad_norm": 0.9002817455009458, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 3097 + }, + { + "epoch": 0.03098, + "grad_norm": 0.8351003992845788, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 3098 + }, + { + "epoch": 0.03099, + "grad_norm": 0.5935585890716526, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 3099 + }, + { + "epoch": 0.031, + "grad_norm": 0.6353446062973269, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3100 + }, + { + "epoch": 0.03101, + "grad_norm": 0.7592462024173295, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 3101 + }, + { + "epoch": 0.03102, + "grad_norm": 0.8288461234391342, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3102 + }, + { + "epoch": 0.03103, + "grad_norm": 0.8186091723896449, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3103 + }, + { + "epoch": 0.03104, + "grad_norm": 0.730911001668181, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 3104 + }, + { + "epoch": 0.03105, + "grad_norm": 0.6266459445253275, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 3105 + }, + { + "epoch": 0.03106, + "grad_norm": 0.6545747432475193, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3106 + }, + { + "epoch": 0.03107, + "grad_norm": 0.5231672698914397, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 3107 + }, + { + "epoch": 0.03108, + "grad_norm": 0.5734245951904136, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 3108 + }, + { + "epoch": 0.03109, + "grad_norm": 0.5288754437005536, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 3109 + }, + { + "epoch": 0.0311, + "grad_norm": 0.5320691683351324, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 3110 + }, + { + "epoch": 0.03111, + "grad_norm": 0.5370642425053924, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 3111 + }, + { + "epoch": 0.03112, + "grad_norm": 0.49890497665091543, + "learning_rate": 0.003, + "loss": 4.178, + "step": 3112 + }, + { + "epoch": 0.03113, + "grad_norm": 0.4336869241892021, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 3113 + }, + { + "epoch": 0.03114, + "grad_norm": 0.45513880225763054, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3114 + }, + { + "epoch": 0.03115, + "grad_norm": 0.5541487651615646, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 3115 + }, + { + "epoch": 0.03116, + "grad_norm": 0.6085416142601979, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3116 + }, + { + "epoch": 0.03117, + "grad_norm": 0.6289436131363786, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 3117 + }, + { + "epoch": 0.03118, + "grad_norm": 0.7648349803443671, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3118 + }, + { + "epoch": 0.03119, + "grad_norm": 0.9905397386097399, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 3119 + }, + { + "epoch": 0.0312, + "grad_norm": 1.002027828062209, + "learning_rate": 0.003, + "loss": 4.2085, + "step": 3120 + }, + { + "epoch": 0.03121, + "grad_norm": 0.9945235603188975, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 3121 + }, + { + "epoch": 0.03122, + "grad_norm": 0.9749277139432609, + "learning_rate": 0.003, + "loss": 4.2158, + "step": 3122 + }, + { + "epoch": 0.03123, + "grad_norm": 0.9176347284109458, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 3123 + }, + { + "epoch": 0.03124, + "grad_norm": 0.8222379595933585, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3124 + }, + { + "epoch": 0.03125, + "grad_norm": 0.8029594745648053, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 3125 + }, + { + "epoch": 0.03126, + "grad_norm": 0.7973924670105094, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 3126 + }, + { + "epoch": 0.03127, + "grad_norm": 0.8183339868293253, + "learning_rate": 0.003, + "loss": 4.2012, + "step": 3127 + }, + { + "epoch": 0.03128, + "grad_norm": 0.8564648965676082, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 3128 + }, + { + "epoch": 0.03129, + "grad_norm": 0.6856535056536724, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 3129 + }, + { + "epoch": 0.0313, + "grad_norm": 0.6641053792145878, + "learning_rate": 0.003, + "loss": 4.2086, + "step": 3130 + }, + { + "epoch": 0.03131, + "grad_norm": 0.5479229508159585, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 3131 + }, + { + "epoch": 0.03132, + "grad_norm": 0.5311614836536821, + "learning_rate": 0.003, + "loss": 4.2024, + "step": 3132 + }, + { + "epoch": 0.03133, + "grad_norm": 0.5105588102398402, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 3133 + }, + { + "epoch": 0.03134, + "grad_norm": 0.5009888970015204, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 3134 + }, + { + "epoch": 0.03135, + "grad_norm": 0.4297510332483796, + "learning_rate": 0.003, + "loss": 4.15, + "step": 3135 + }, + { + "epoch": 0.03136, + "grad_norm": 0.4354586739325799, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 3136 + }, + { + "epoch": 0.03137, + "grad_norm": 0.5083671646646589, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3137 + }, + { + "epoch": 0.03138, + "grad_norm": 0.6199429454274028, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 3138 + }, + { + "epoch": 0.03139, + "grad_norm": 0.7271551302730402, + "learning_rate": 0.003, + "loss": 4.179, + "step": 3139 + }, + { + "epoch": 0.0314, + "grad_norm": 0.7358854576257201, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 3140 + }, + { + "epoch": 0.03141, + "grad_norm": 0.6326154627482451, + "learning_rate": 0.003, + "loss": 4.208, + "step": 3141 + }, + { + "epoch": 0.03142, + "grad_norm": 0.5588364977660311, + "learning_rate": 0.003, + "loss": 4.183, + "step": 3142 + }, + { + "epoch": 0.03143, + "grad_norm": 0.6311591083308525, + "learning_rate": 0.003, + "loss": 4.1728, + "step": 3143 + }, + { + "epoch": 0.03144, + "grad_norm": 0.5392546760146013, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3144 + }, + { + "epoch": 0.03145, + "grad_norm": 0.4557888424936214, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 3145 + }, + { + "epoch": 0.03146, + "grad_norm": 0.4216370024975091, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 3146 + }, + { + "epoch": 0.03147, + "grad_norm": 0.4295618903624941, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3147 + }, + { + "epoch": 0.03148, + "grad_norm": 0.4097670324789773, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 3148 + }, + { + "epoch": 0.03149, + "grad_norm": 0.4802416636246548, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 3149 + }, + { + "epoch": 0.0315, + "grad_norm": 0.46441368907739394, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 3150 + }, + { + "epoch": 0.03151, + "grad_norm": 0.44059093730228843, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3151 + }, + { + "epoch": 0.03152, + "grad_norm": 0.4172894456146658, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3152 + }, + { + "epoch": 0.03153, + "grad_norm": 0.42959521977879306, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 3153 + }, + { + "epoch": 0.03154, + "grad_norm": 0.43398804524875734, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3154 + }, + { + "epoch": 0.03155, + "grad_norm": 0.4562270589856803, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3155 + }, + { + "epoch": 0.03156, + "grad_norm": 0.5126486371492089, + "learning_rate": 0.003, + "loss": 4.142, + "step": 3156 + }, + { + "epoch": 0.03157, + "grad_norm": 0.6526844959193616, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 3157 + }, + { + "epoch": 0.03158, + "grad_norm": 0.9480089169547891, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 3158 + }, + { + "epoch": 0.03159, + "grad_norm": 1.3561889811332835, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 3159 + }, + { + "epoch": 0.0316, + "grad_norm": 0.5982846825437411, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 3160 + }, + { + "epoch": 0.03161, + "grad_norm": 0.8732729212340313, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 3161 + }, + { + "epoch": 0.03162, + "grad_norm": 0.886847324325859, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 3162 + }, + { + "epoch": 0.03163, + "grad_norm": 0.99270442281753, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 3163 + }, + { + "epoch": 0.03164, + "grad_norm": 0.9630909842912381, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 3164 + }, + { + "epoch": 0.03165, + "grad_norm": 0.7916297793185595, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 3165 + }, + { + "epoch": 0.03166, + "grad_norm": 0.6926990051674062, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 3166 + }, + { + "epoch": 0.03167, + "grad_norm": 0.681279202527116, + "learning_rate": 0.003, + "loss": 4.1779, + "step": 3167 + }, + { + "epoch": 0.03168, + "grad_norm": 0.6889533564228932, + "learning_rate": 0.003, + "loss": 4.2042, + "step": 3168 + }, + { + "epoch": 0.03169, + "grad_norm": 0.8396344102094035, + "learning_rate": 0.003, + "loss": 4.2217, + "step": 3169 + }, + { + "epoch": 0.0317, + "grad_norm": 0.9428653644577029, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 3170 + }, + { + "epoch": 0.03171, + "grad_norm": 0.691282163039264, + "learning_rate": 0.003, + "loss": 4.2143, + "step": 3171 + }, + { + "epoch": 0.03172, + "grad_norm": 0.665795792414971, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 3172 + }, + { + "epoch": 0.03173, + "grad_norm": 0.6751127918552098, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3173 + }, + { + "epoch": 0.03174, + "grad_norm": 0.6066754986654972, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 3174 + }, + { + "epoch": 0.03175, + "grad_norm": 0.5400765945883416, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 3175 + }, + { + "epoch": 0.03176, + "grad_norm": 0.5542770572580099, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3176 + }, + { + "epoch": 0.03177, + "grad_norm": 0.5967671493496074, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 3177 + }, + { + "epoch": 0.03178, + "grad_norm": 0.6105572005396745, + "learning_rate": 0.003, + "loss": 4.222, + "step": 3178 + }, + { + "epoch": 0.03179, + "grad_norm": 0.6783275770388425, + "learning_rate": 0.003, + "loss": 4.186, + "step": 3179 + }, + { + "epoch": 0.0318, + "grad_norm": 0.7073200859467578, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3180 + }, + { + "epoch": 0.03181, + "grad_norm": 0.6783381680497546, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 3181 + }, + { + "epoch": 0.03182, + "grad_norm": 0.7230568807336094, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 3182 + }, + { + "epoch": 0.03183, + "grad_norm": 0.873578952331532, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 3183 + }, + { + "epoch": 0.03184, + "grad_norm": 0.9905329983068216, + "learning_rate": 0.003, + "loss": 4.1817, + "step": 3184 + }, + { + "epoch": 0.03185, + "grad_norm": 1.2552538351706124, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 3185 + }, + { + "epoch": 0.03186, + "grad_norm": 0.6881365341663627, + "learning_rate": 0.003, + "loss": 4.1977, + "step": 3186 + }, + { + "epoch": 0.03187, + "grad_norm": 0.6603777519922737, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3187 + }, + { + "epoch": 0.03188, + "grad_norm": 0.6114536474568962, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 3188 + }, + { + "epoch": 0.03189, + "grad_norm": 0.7652970454563929, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 3189 + }, + { + "epoch": 0.0319, + "grad_norm": 0.7756244045621472, + "learning_rate": 0.003, + "loss": 4.1785, + "step": 3190 + }, + { + "epoch": 0.03191, + "grad_norm": 0.7586188953077919, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 3191 + }, + { + "epoch": 0.03192, + "grad_norm": 0.7248898112293884, + "learning_rate": 0.003, + "loss": 4.186, + "step": 3192 + }, + { + "epoch": 0.03193, + "grad_norm": 0.6819049982170161, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 3193 + }, + { + "epoch": 0.03194, + "grad_norm": 0.6384912160022395, + "learning_rate": 0.003, + "loss": 4.208, + "step": 3194 + }, + { + "epoch": 0.03195, + "grad_norm": 0.6207563594114237, + "learning_rate": 0.003, + "loss": 4.172, + "step": 3195 + }, + { + "epoch": 0.03196, + "grad_norm": 0.5716998964191464, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 3196 + }, + { + "epoch": 0.03197, + "grad_norm": 0.5492769888070425, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 3197 + }, + { + "epoch": 0.03198, + "grad_norm": 0.5349886855188608, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 3198 + }, + { + "epoch": 0.03199, + "grad_norm": 0.4722551144791865, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3199 + }, + { + "epoch": 0.032, + "grad_norm": 0.5241213036464709, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 3200 + }, + { + "epoch": 0.03201, + "grad_norm": 0.5556232177618735, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3201 + }, + { + "epoch": 0.03202, + "grad_norm": 0.5951717317344422, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 3202 + }, + { + "epoch": 0.03203, + "grad_norm": 0.5874723863194432, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 3203 + }, + { + "epoch": 0.03204, + "grad_norm": 0.5662458323617429, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 3204 + }, + { + "epoch": 0.03205, + "grad_norm": 0.55236769595308, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 3205 + }, + { + "epoch": 0.03206, + "grad_norm": 0.5082397875263351, + "learning_rate": 0.003, + "loss": 4.16, + "step": 3206 + }, + { + "epoch": 0.03207, + "grad_norm": 0.4452387164637836, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 3207 + }, + { + "epoch": 0.03208, + "grad_norm": 0.48293819644672603, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 3208 + }, + { + "epoch": 0.03209, + "grad_norm": 0.4498162894819178, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 3209 + }, + { + "epoch": 0.0321, + "grad_norm": 0.5708600401920843, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3210 + }, + { + "epoch": 0.03211, + "grad_norm": 0.6897782500954599, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 3211 + }, + { + "epoch": 0.03212, + "grad_norm": 0.8577311234564894, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3212 + }, + { + "epoch": 0.03213, + "grad_norm": 0.7479816094202391, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3213 + }, + { + "epoch": 0.03214, + "grad_norm": 0.5575837603075381, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 3214 + }, + { + "epoch": 0.03215, + "grad_norm": 0.619113370679725, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 3215 + }, + { + "epoch": 0.03216, + "grad_norm": 0.6857581625178918, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 3216 + }, + { + "epoch": 0.03217, + "grad_norm": 0.6339316692312089, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 3217 + }, + { + "epoch": 0.03218, + "grad_norm": 0.5361569097117066, + "learning_rate": 0.003, + "loss": 4.147, + "step": 3218 + }, + { + "epoch": 0.03219, + "grad_norm": 0.5146188024497156, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 3219 + }, + { + "epoch": 0.0322, + "grad_norm": 0.5580750782179962, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 3220 + }, + { + "epoch": 0.03221, + "grad_norm": 0.5127806683543551, + "learning_rate": 0.003, + "loss": 4.171, + "step": 3221 + }, + { + "epoch": 0.03222, + "grad_norm": 0.5372160864602477, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3222 + }, + { + "epoch": 0.03223, + "grad_norm": 0.5923510679532299, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3223 + }, + { + "epoch": 0.03224, + "grad_norm": 0.6861592854273826, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3224 + }, + { + "epoch": 0.03225, + "grad_norm": 0.7582363764178276, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3225 + }, + { + "epoch": 0.03226, + "grad_norm": 0.8207513579274669, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 3226 + }, + { + "epoch": 0.03227, + "grad_norm": 0.7077066475989446, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3227 + }, + { + "epoch": 0.03228, + "grad_norm": 0.6232151026364198, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 3228 + }, + { + "epoch": 0.03229, + "grad_norm": 0.6228354375566014, + "learning_rate": 0.003, + "loss": 4.1653, + "step": 3229 + }, + { + "epoch": 0.0323, + "grad_norm": 0.6011100180442941, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3230 + }, + { + "epoch": 0.03231, + "grad_norm": 0.6278235480602589, + "learning_rate": 0.003, + "loss": 4.1733, + "step": 3231 + }, + { + "epoch": 0.03232, + "grad_norm": 0.5999968897563468, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 3232 + }, + { + "epoch": 0.03233, + "grad_norm": 0.5467043906444932, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 3233 + }, + { + "epoch": 0.03234, + "grad_norm": 0.616637901979197, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 3234 + }, + { + "epoch": 0.03235, + "grad_norm": 0.7932914435950059, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 3235 + }, + { + "epoch": 0.03236, + "grad_norm": 0.9734805326028607, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 3236 + }, + { + "epoch": 0.03237, + "grad_norm": 0.9035113763708699, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3237 + }, + { + "epoch": 0.03238, + "grad_norm": 0.8124977646963972, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 3238 + }, + { + "epoch": 0.03239, + "grad_norm": 0.7169306328435582, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 3239 + }, + { + "epoch": 0.0324, + "grad_norm": 0.7507848511489641, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 3240 + }, + { + "epoch": 0.03241, + "grad_norm": 0.871055197405415, + "learning_rate": 0.003, + "loss": 4.1945, + "step": 3241 + }, + { + "epoch": 0.03242, + "grad_norm": 0.9726257649238742, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 3242 + }, + { + "epoch": 0.03243, + "grad_norm": 0.8829139791894937, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 3243 + }, + { + "epoch": 0.03244, + "grad_norm": 0.9226257984088151, + "learning_rate": 0.003, + "loss": 4.1964, + "step": 3244 + }, + { + "epoch": 0.03245, + "grad_norm": 1.0602033368371822, + "learning_rate": 0.003, + "loss": 4.2203, + "step": 3245 + }, + { + "epoch": 0.03246, + "grad_norm": 0.8832790843056474, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 3246 + }, + { + "epoch": 0.03247, + "grad_norm": 0.8993256456137283, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 3247 + }, + { + "epoch": 0.03248, + "grad_norm": 0.9121935412385299, + "learning_rate": 0.003, + "loss": 4.2361, + "step": 3248 + }, + { + "epoch": 0.03249, + "grad_norm": 0.8613725910612215, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3249 + }, + { + "epoch": 0.0325, + "grad_norm": 0.8167292628339079, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 3250 + }, + { + "epoch": 0.03251, + "grad_norm": 0.7863361659710474, + "learning_rate": 0.003, + "loss": 4.2274, + "step": 3251 + }, + { + "epoch": 0.03252, + "grad_norm": 0.7493875681769112, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3252 + }, + { + "epoch": 0.03253, + "grad_norm": 0.7679048561396902, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 3253 + }, + { + "epoch": 0.03254, + "grad_norm": 0.7654672625645944, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 3254 + }, + { + "epoch": 0.03255, + "grad_norm": 0.7464808737181244, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 3255 + }, + { + "epoch": 0.03256, + "grad_norm": 0.6526585542637006, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 3256 + }, + { + "epoch": 0.03257, + "grad_norm": 0.6001937061564249, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 3257 + }, + { + "epoch": 0.03258, + "grad_norm": 0.5472158831270236, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 3258 + }, + { + "epoch": 0.03259, + "grad_norm": 0.4785164815267161, + "learning_rate": 0.003, + "loss": 4.1806, + "step": 3259 + }, + { + "epoch": 0.0326, + "grad_norm": 0.38861603840814574, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 3260 + }, + { + "epoch": 0.03261, + "grad_norm": 0.41654558334327835, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 3261 + }, + { + "epoch": 0.03262, + "grad_norm": 0.41147900352780037, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 3262 + }, + { + "epoch": 0.03263, + "grad_norm": 0.4340386920329771, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 3263 + }, + { + "epoch": 0.03264, + "grad_norm": 0.47093951437519954, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 3264 + }, + { + "epoch": 0.03265, + "grad_norm": 0.5195090423674352, + "learning_rate": 0.003, + "loss": 4.184, + "step": 3265 + }, + { + "epoch": 0.03266, + "grad_norm": 0.5846656667385064, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3266 + }, + { + "epoch": 0.03267, + "grad_norm": 0.6621919529113969, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 3267 + }, + { + "epoch": 0.03268, + "grad_norm": 0.6864096228211959, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3268 + }, + { + "epoch": 0.03269, + "grad_norm": 0.5308237484050747, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3269 + }, + { + "epoch": 0.0327, + "grad_norm": 0.44850448788276254, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 3270 + }, + { + "epoch": 0.03271, + "grad_norm": 0.4693563167350624, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 3271 + }, + { + "epoch": 0.03272, + "grad_norm": 0.5323713025056388, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 3272 + }, + { + "epoch": 0.03273, + "grad_norm": 0.6692459000105278, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 3273 + }, + { + "epoch": 0.03274, + "grad_norm": 0.7924729751268105, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3274 + }, + { + "epoch": 0.03275, + "grad_norm": 0.8832852537291822, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 3275 + }, + { + "epoch": 0.03276, + "grad_norm": 0.9440382055401524, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3276 + }, + { + "epoch": 0.03277, + "grad_norm": 0.9157723514845009, + "learning_rate": 0.003, + "loss": 4.181, + "step": 3277 + }, + { + "epoch": 0.03278, + "grad_norm": 0.7757959986300896, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 3278 + }, + { + "epoch": 0.03279, + "grad_norm": 0.7052884741699046, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3279 + }, + { + "epoch": 0.0328, + "grad_norm": 0.6748931568117461, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3280 + }, + { + "epoch": 0.03281, + "grad_norm": 0.5834976642550973, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 3281 + }, + { + "epoch": 0.03282, + "grad_norm": 0.5830462450862388, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 3282 + }, + { + "epoch": 0.03283, + "grad_norm": 0.6046371763595715, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3283 + }, + { + "epoch": 0.03284, + "grad_norm": 0.498692543866341, + "learning_rate": 0.003, + "loss": 4.188, + "step": 3284 + }, + { + "epoch": 0.03285, + "grad_norm": 0.5300083365332573, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 3285 + }, + { + "epoch": 0.03286, + "grad_norm": 0.5302483285107112, + "learning_rate": 0.003, + "loss": 4.173, + "step": 3286 + }, + { + "epoch": 0.03287, + "grad_norm": 0.5280212789702404, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3287 + }, + { + "epoch": 0.03288, + "grad_norm": 0.49328823308562514, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3288 + }, + { + "epoch": 0.03289, + "grad_norm": 0.5955758847359914, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 3289 + }, + { + "epoch": 0.0329, + "grad_norm": 0.7318966219724384, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 3290 + }, + { + "epoch": 0.03291, + "grad_norm": 0.8969015331966617, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 3291 + }, + { + "epoch": 0.03292, + "grad_norm": 0.8171747463806796, + "learning_rate": 0.003, + "loss": 4.15, + "step": 3292 + }, + { + "epoch": 0.03293, + "grad_norm": 0.6454312253277896, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3293 + }, + { + "epoch": 0.03294, + "grad_norm": 0.5476436730454269, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 3294 + }, + { + "epoch": 0.03295, + "grad_norm": 0.6343237379540597, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3295 + }, + { + "epoch": 0.03296, + "grad_norm": 0.6111699411729721, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3296 + }, + { + "epoch": 0.03297, + "grad_norm": 0.5298030355791016, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3297 + }, + { + "epoch": 0.03298, + "grad_norm": 0.4836663999032712, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3298 + }, + { + "epoch": 0.03299, + "grad_norm": 0.4701496219943231, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 3299 + }, + { + "epoch": 0.033, + "grad_norm": 0.4623026228792837, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3300 + }, + { + "epoch": 0.03301, + "grad_norm": 0.43500908939154875, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 3301 + }, + { + "epoch": 0.03302, + "grad_norm": 0.4287201358461857, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3302 + }, + { + "epoch": 0.03303, + "grad_norm": 0.46641640405345786, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 3303 + }, + { + "epoch": 0.03304, + "grad_norm": 0.588941013591916, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 3304 + }, + { + "epoch": 0.03305, + "grad_norm": 0.7079568978470296, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3305 + }, + { + "epoch": 0.03306, + "grad_norm": 0.829985756720969, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 3306 + }, + { + "epoch": 0.03307, + "grad_norm": 0.8462386242609186, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 3307 + }, + { + "epoch": 0.03308, + "grad_norm": 0.8293495815638529, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3308 + }, + { + "epoch": 0.03309, + "grad_norm": 0.8456212329927598, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3309 + }, + { + "epoch": 0.0331, + "grad_norm": 0.8565848835132709, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3310 + }, + { + "epoch": 0.03311, + "grad_norm": 0.9394898981807235, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 3311 + }, + { + "epoch": 0.03312, + "grad_norm": 1.0401594073450542, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 3312 + }, + { + "epoch": 0.03313, + "grad_norm": 0.8840161283334239, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 3313 + }, + { + "epoch": 0.03314, + "grad_norm": 0.8560525559559692, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 3314 + }, + { + "epoch": 0.03315, + "grad_norm": 0.7954589166824607, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 3315 + }, + { + "epoch": 0.03316, + "grad_norm": 0.7407877779230415, + "learning_rate": 0.003, + "loss": 4.2018, + "step": 3316 + }, + { + "epoch": 0.03317, + "grad_norm": 0.7593343697168953, + "learning_rate": 0.003, + "loss": 4.1938, + "step": 3317 + }, + { + "epoch": 0.03318, + "grad_norm": 0.7588615866933531, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 3318 + }, + { + "epoch": 0.03319, + "grad_norm": 0.6406451388828956, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 3319 + }, + { + "epoch": 0.0332, + "grad_norm": 0.6857052599814105, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 3320 + }, + { + "epoch": 0.03321, + "grad_norm": 0.6857879313241626, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 3321 + }, + { + "epoch": 0.03322, + "grad_norm": 0.7686016779789973, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3322 + }, + { + "epoch": 0.03323, + "grad_norm": 0.8269613306774392, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 3323 + }, + { + "epoch": 0.03324, + "grad_norm": 0.9528423775782789, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3324 + }, + { + "epoch": 0.03325, + "grad_norm": 1.0794820717536844, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 3325 + }, + { + "epoch": 0.03326, + "grad_norm": 0.8445799040478842, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3326 + }, + { + "epoch": 0.03327, + "grad_norm": 0.6629463030927865, + "learning_rate": 0.003, + "loss": 4.1918, + "step": 3327 + }, + { + "epoch": 0.03328, + "grad_norm": 0.6487519835718909, + "learning_rate": 0.003, + "loss": 4.211, + "step": 3328 + }, + { + "epoch": 0.03329, + "grad_norm": 0.6298798087889773, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 3329 + }, + { + "epoch": 0.0333, + "grad_norm": 0.6333191541087022, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 3330 + }, + { + "epoch": 0.03331, + "grad_norm": 0.7837692756055505, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 3331 + }, + { + "epoch": 0.03332, + "grad_norm": 0.9187938715930933, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 3332 + }, + { + "epoch": 0.03333, + "grad_norm": 0.7284609287843756, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 3333 + }, + { + "epoch": 0.03334, + "grad_norm": 0.5950364415684378, + "learning_rate": 0.003, + "loss": 4.182, + "step": 3334 + }, + { + "epoch": 0.03335, + "grad_norm": 0.5204758750774462, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 3335 + }, + { + "epoch": 0.03336, + "grad_norm": 0.5061539602167332, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 3336 + }, + { + "epoch": 0.03337, + "grad_norm": 0.42913500260913456, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 3337 + }, + { + "epoch": 0.03338, + "grad_norm": 0.4415664602358343, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 3338 + }, + { + "epoch": 0.03339, + "grad_norm": 0.4739540929710416, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3339 + }, + { + "epoch": 0.0334, + "grad_norm": 0.4177466598922738, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3340 + }, + { + "epoch": 0.03341, + "grad_norm": 0.4450986126917478, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 3341 + }, + { + "epoch": 0.03342, + "grad_norm": 0.4857573129812489, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 3342 + }, + { + "epoch": 0.03343, + "grad_norm": 0.6478603105495, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3343 + }, + { + "epoch": 0.03344, + "grad_norm": 0.7654176175893398, + "learning_rate": 0.003, + "loss": 4.1891, + "step": 3344 + }, + { + "epoch": 0.03345, + "grad_norm": 0.708719161890371, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 3345 + }, + { + "epoch": 0.03346, + "grad_norm": 0.5692838217453585, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 3346 + }, + { + "epoch": 0.03347, + "grad_norm": 0.5899959723847616, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 3347 + }, + { + "epoch": 0.03348, + "grad_norm": 0.6486978345976652, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3348 + }, + { + "epoch": 0.03349, + "grad_norm": 0.5576273675617567, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 3349 + }, + { + "epoch": 0.0335, + "grad_norm": 0.537004248184913, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 3350 + }, + { + "epoch": 0.03351, + "grad_norm": 0.5567929131713881, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3351 + }, + { + "epoch": 0.03352, + "grad_norm": 0.6714613195190436, + "learning_rate": 0.003, + "loss": 4.171, + "step": 3352 + }, + { + "epoch": 0.03353, + "grad_norm": 0.7428392470438228, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 3353 + }, + { + "epoch": 0.03354, + "grad_norm": 0.709523442230735, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3354 + }, + { + "epoch": 0.03355, + "grad_norm": 0.5743763408378283, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 3355 + }, + { + "epoch": 0.03356, + "grad_norm": 0.5372193296353462, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 3356 + }, + { + "epoch": 0.03357, + "grad_norm": 0.6152109328121239, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3357 + }, + { + "epoch": 0.03358, + "grad_norm": 0.693520738126006, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 3358 + }, + { + "epoch": 0.03359, + "grad_norm": 0.6800317331746373, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3359 + }, + { + "epoch": 0.0336, + "grad_norm": 0.6462929043552973, + "learning_rate": 0.003, + "loss": 4.179, + "step": 3360 + }, + { + "epoch": 0.03361, + "grad_norm": 0.6413289136062978, + "learning_rate": 0.003, + "loss": 4.1788, + "step": 3361 + }, + { + "epoch": 0.03362, + "grad_norm": 0.6955328156949359, + "learning_rate": 0.003, + "loss": 4.1767, + "step": 3362 + }, + { + "epoch": 0.03363, + "grad_norm": 0.6235123487755582, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3363 + }, + { + "epoch": 0.03364, + "grad_norm": 0.602598837086171, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 3364 + }, + { + "epoch": 0.03365, + "grad_norm": 0.7356364448082628, + "learning_rate": 0.003, + "loss": 4.144, + "step": 3365 + }, + { + "epoch": 0.03366, + "grad_norm": 0.8244656072335483, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3366 + }, + { + "epoch": 0.03367, + "grad_norm": 0.7703618126132621, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 3367 + }, + { + "epoch": 0.03368, + "grad_norm": 0.6946923139107304, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 3368 + }, + { + "epoch": 0.03369, + "grad_norm": 0.7107703988620854, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3369 + }, + { + "epoch": 0.0337, + "grad_norm": 0.6762732803354303, + "learning_rate": 0.003, + "loss": 4.1834, + "step": 3370 + }, + { + "epoch": 0.03371, + "grad_norm": 0.6545021306919707, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3371 + }, + { + "epoch": 0.03372, + "grad_norm": 0.6538700683175451, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3372 + }, + { + "epoch": 0.03373, + "grad_norm": 0.601220154679101, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3373 + }, + { + "epoch": 0.03374, + "grad_norm": 0.5930617819105674, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 3374 + }, + { + "epoch": 0.03375, + "grad_norm": 0.5268626720580842, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3375 + }, + { + "epoch": 0.03376, + "grad_norm": 0.5347783123792315, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3376 + }, + { + "epoch": 0.03377, + "grad_norm": 0.5216348207352511, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 3377 + }, + { + "epoch": 0.03378, + "grad_norm": 0.6194694687851028, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3378 + }, + { + "epoch": 0.03379, + "grad_norm": 0.6549669142309517, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3379 + }, + { + "epoch": 0.0338, + "grad_norm": 0.6860555079944507, + "learning_rate": 0.003, + "loss": 4.179, + "step": 3380 + }, + { + "epoch": 0.03381, + "grad_norm": 0.7654947303377718, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 3381 + }, + { + "epoch": 0.03382, + "grad_norm": 0.8842593404054855, + "learning_rate": 0.003, + "loss": 4.177, + "step": 3382 + }, + { + "epoch": 0.03383, + "grad_norm": 1.2792086110225658, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 3383 + }, + { + "epoch": 0.03384, + "grad_norm": 0.9270742666385557, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 3384 + }, + { + "epoch": 0.03385, + "grad_norm": 0.7871637830554715, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 3385 + }, + { + "epoch": 0.03386, + "grad_norm": 0.8239773523698274, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 3386 + }, + { + "epoch": 0.03387, + "grad_norm": 0.73835170938435, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 3387 + }, + { + "epoch": 0.03388, + "grad_norm": 0.6929973646655294, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 3388 + }, + { + "epoch": 0.03389, + "grad_norm": 0.7227682380021057, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 3389 + }, + { + "epoch": 0.0339, + "grad_norm": 0.6143798247462846, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 3390 + }, + { + "epoch": 0.03391, + "grad_norm": 0.6734624195180631, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3391 + }, + { + "epoch": 0.03392, + "grad_norm": 0.6594794409943178, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3392 + }, + { + "epoch": 0.03393, + "grad_norm": 0.5959392487968519, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 3393 + }, + { + "epoch": 0.03394, + "grad_norm": 0.5142376929909485, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 3394 + }, + { + "epoch": 0.03395, + "grad_norm": 0.48066315678667376, + "learning_rate": 0.003, + "loss": 4.1776, + "step": 3395 + }, + { + "epoch": 0.03396, + "grad_norm": 0.4969135823877688, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 3396 + }, + { + "epoch": 0.03397, + "grad_norm": 0.4793333077073278, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3397 + }, + { + "epoch": 0.03398, + "grad_norm": 0.5618846612168108, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 3398 + }, + { + "epoch": 0.03399, + "grad_norm": 0.598645896556551, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 3399 + }, + { + "epoch": 0.034, + "grad_norm": 0.5909466070203917, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3400 + }, + { + "epoch": 0.03401, + "grad_norm": 0.5544973914148591, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 3401 + }, + { + "epoch": 0.03402, + "grad_norm": 0.5799704954663493, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 3402 + }, + { + "epoch": 0.03403, + "grad_norm": 0.6679893543628809, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 3403 + }, + { + "epoch": 0.03404, + "grad_norm": 0.7464230169510123, + "learning_rate": 0.003, + "loss": 4.1854, + "step": 3404 + }, + { + "epoch": 0.03405, + "grad_norm": 0.7894447498122126, + "learning_rate": 0.003, + "loss": 4.168, + "step": 3405 + }, + { + "epoch": 0.03406, + "grad_norm": 0.7021432861635766, + "learning_rate": 0.003, + "loss": 4.152, + "step": 3406 + }, + { + "epoch": 0.03407, + "grad_norm": 0.5878306021412367, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 3407 + }, + { + "epoch": 0.03408, + "grad_norm": 0.5913015706762682, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 3408 + }, + { + "epoch": 0.03409, + "grad_norm": 0.6312028321034342, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 3409 + }, + { + "epoch": 0.0341, + "grad_norm": 0.7898832429117774, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3410 + }, + { + "epoch": 0.03411, + "grad_norm": 0.7714877743902544, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 3411 + }, + { + "epoch": 0.03412, + "grad_norm": 0.7513990558104037, + "learning_rate": 0.003, + "loss": 4.19, + "step": 3412 + }, + { + "epoch": 0.03413, + "grad_norm": 0.7461988498435763, + "learning_rate": 0.003, + "loss": 4.1994, + "step": 3413 + }, + { + "epoch": 0.03414, + "grad_norm": 0.7030092576304908, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 3414 + }, + { + "epoch": 0.03415, + "grad_norm": 0.7996163589913181, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 3415 + }, + { + "epoch": 0.03416, + "grad_norm": 0.9173276805644571, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 3416 + }, + { + "epoch": 0.03417, + "grad_norm": 1.1614777308174031, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 3417 + }, + { + "epoch": 0.03418, + "grad_norm": 1.0497658143246895, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3418 + }, + { + "epoch": 0.03419, + "grad_norm": 0.8865365686182854, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 3419 + }, + { + "epoch": 0.0342, + "grad_norm": 0.9104853531188979, + "learning_rate": 0.003, + "loss": 4.2051, + "step": 3420 + }, + { + "epoch": 0.03421, + "grad_norm": 0.8728209224738288, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 3421 + }, + { + "epoch": 0.03422, + "grad_norm": 0.8151487908159633, + "learning_rate": 0.003, + "loss": 4.2121, + "step": 3422 + }, + { + "epoch": 0.03423, + "grad_norm": 0.7760561564619395, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 3423 + }, + { + "epoch": 0.03424, + "grad_norm": 0.7902274791310419, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 3424 + }, + { + "epoch": 0.03425, + "grad_norm": 0.7840677839560138, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3425 + }, + { + "epoch": 0.03426, + "grad_norm": 0.7200842223622085, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 3426 + }, + { + "epoch": 0.03427, + "grad_norm": 0.6907951008112484, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3427 + }, + { + "epoch": 0.03428, + "grad_norm": 0.6706145072117141, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 3428 + }, + { + "epoch": 0.03429, + "grad_norm": 0.5775498037003282, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3429 + }, + { + "epoch": 0.0343, + "grad_norm": 0.5280875253924556, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3430 + }, + { + "epoch": 0.03431, + "grad_norm": 0.5133378996898088, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3431 + }, + { + "epoch": 0.03432, + "grad_norm": 0.5402448334590585, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3432 + }, + { + "epoch": 0.03433, + "grad_norm": 0.5171387300004384, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 3433 + }, + { + "epoch": 0.03434, + "grad_norm": 0.5654372536507138, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 3434 + }, + { + "epoch": 0.03435, + "grad_norm": 0.7195487294501742, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3435 + }, + { + "epoch": 0.03436, + "grad_norm": 0.9836128063160017, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3436 + }, + { + "epoch": 0.03437, + "grad_norm": 1.000819383038808, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 3437 + }, + { + "epoch": 0.03438, + "grad_norm": 0.7833210304414927, + "learning_rate": 0.003, + "loss": 4.1883, + "step": 3438 + }, + { + "epoch": 0.03439, + "grad_norm": 0.7155950508827565, + "learning_rate": 0.003, + "loss": 4.1864, + "step": 3439 + }, + { + "epoch": 0.0344, + "grad_norm": 0.7239664621860699, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3440 + }, + { + "epoch": 0.03441, + "grad_norm": 0.6825035000283939, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 3441 + }, + { + "epoch": 0.03442, + "grad_norm": 0.6099118062197426, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3442 + }, + { + "epoch": 0.03443, + "grad_norm": 0.6931617154820389, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 3443 + }, + { + "epoch": 0.03444, + "grad_norm": 0.6606283519642352, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 3444 + }, + { + "epoch": 0.03445, + "grad_norm": 0.5738857319007685, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 3445 + }, + { + "epoch": 0.03446, + "grad_norm": 0.6247920672226246, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 3446 + }, + { + "epoch": 0.03447, + "grad_norm": 0.6141551380791748, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 3447 + }, + { + "epoch": 0.03448, + "grad_norm": 0.5719627595458622, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 3448 + }, + { + "epoch": 0.03449, + "grad_norm": 0.5136603871940236, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 3449 + }, + { + "epoch": 0.0345, + "grad_norm": 0.4556692204413723, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 3450 + }, + { + "epoch": 0.03451, + "grad_norm": 0.43785806673883965, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 3451 + }, + { + "epoch": 0.03452, + "grad_norm": 0.4135943023058251, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3452 + }, + { + "epoch": 0.03453, + "grad_norm": 0.41685265470403127, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 3453 + }, + { + "epoch": 0.03454, + "grad_norm": 0.4541366407006024, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 3454 + }, + { + "epoch": 0.03455, + "grad_norm": 0.4820485430446041, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3455 + }, + { + "epoch": 0.03456, + "grad_norm": 0.6070175235924494, + "learning_rate": 0.003, + "loss": 4.15, + "step": 3456 + }, + { + "epoch": 0.03457, + "grad_norm": 0.9155671860985577, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 3457 + }, + { + "epoch": 0.03458, + "grad_norm": 1.0743568532560335, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 3458 + }, + { + "epoch": 0.03459, + "grad_norm": 0.6911554531998789, + "learning_rate": 0.003, + "loss": 4.1933, + "step": 3459 + }, + { + "epoch": 0.0346, + "grad_norm": 0.7160032080825989, + "learning_rate": 0.003, + "loss": 4.1932, + "step": 3460 + }, + { + "epoch": 0.03461, + "grad_norm": 0.7205146036072163, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 3461 + }, + { + "epoch": 0.03462, + "grad_norm": 0.6026232163742171, + "learning_rate": 0.003, + "loss": 4.158, + "step": 3462 + }, + { + "epoch": 0.03463, + "grad_norm": 0.6794227213204947, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 3463 + }, + { + "epoch": 0.03464, + "grad_norm": 0.6456036736035401, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 3464 + }, + { + "epoch": 0.03465, + "grad_norm": 0.7130971772716332, + "learning_rate": 0.003, + "loss": 4.1705, + "step": 3465 + }, + { + "epoch": 0.03466, + "grad_norm": 0.6881411959714172, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 3466 + }, + { + "epoch": 0.03467, + "grad_norm": 0.5261069185968785, + "learning_rate": 0.003, + "loss": 4.147, + "step": 3467 + }, + { + "epoch": 0.03468, + "grad_norm": 0.6133848295254514, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3468 + }, + { + "epoch": 0.03469, + "grad_norm": 0.5746409041252338, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3469 + }, + { + "epoch": 0.0347, + "grad_norm": 0.509910411545196, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 3470 + }, + { + "epoch": 0.03471, + "grad_norm": 0.5549185257359567, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 3471 + }, + { + "epoch": 0.03472, + "grad_norm": 0.5919654155995967, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 3472 + }, + { + "epoch": 0.03473, + "grad_norm": 0.6922934316382823, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3473 + }, + { + "epoch": 0.03474, + "grad_norm": 0.664824640285904, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 3474 + }, + { + "epoch": 0.03475, + "grad_norm": 0.690088727777654, + "learning_rate": 0.003, + "loss": 4.157, + "step": 3475 + }, + { + "epoch": 0.03476, + "grad_norm": 0.6776855408026875, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 3476 + }, + { + "epoch": 0.03477, + "grad_norm": 0.7001953445833428, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3477 + }, + { + "epoch": 0.03478, + "grad_norm": 0.7620329811614004, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 3478 + }, + { + "epoch": 0.03479, + "grad_norm": 0.8981370425674162, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 3479 + }, + { + "epoch": 0.0348, + "grad_norm": 0.8250170448327679, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 3480 + }, + { + "epoch": 0.03481, + "grad_norm": 0.6849653124584689, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 3481 + }, + { + "epoch": 0.03482, + "grad_norm": 0.6955143755202231, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3482 + }, + { + "epoch": 0.03483, + "grad_norm": 0.8314937740860378, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 3483 + }, + { + "epoch": 0.03484, + "grad_norm": 0.8591376385119152, + "learning_rate": 0.003, + "loss": 4.205, + "step": 3484 + }, + { + "epoch": 0.03485, + "grad_norm": 0.908594692668108, + "learning_rate": 0.003, + "loss": 4.1994, + "step": 3485 + }, + { + "epoch": 0.03486, + "grad_norm": 0.9172194797930587, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3486 + }, + { + "epoch": 0.03487, + "grad_norm": 0.8921568436215902, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 3487 + }, + { + "epoch": 0.03488, + "grad_norm": 0.7601873754760123, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 3488 + }, + { + "epoch": 0.03489, + "grad_norm": 0.7787148687779002, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 3489 + }, + { + "epoch": 0.0349, + "grad_norm": 0.7649701226102734, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3490 + }, + { + "epoch": 0.03491, + "grad_norm": 0.5441805689029632, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 3491 + }, + { + "epoch": 0.03492, + "grad_norm": 0.5662780979613814, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 3492 + }, + { + "epoch": 0.03493, + "grad_norm": 0.5236480373981434, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3493 + }, + { + "epoch": 0.03494, + "grad_norm": 0.5399778596820843, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 3494 + }, + { + "epoch": 0.03495, + "grad_norm": 0.5540768413769496, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3495 + }, + { + "epoch": 0.03496, + "grad_norm": 0.5933746135701535, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3496 + }, + { + "epoch": 0.03497, + "grad_norm": 0.5852489998748386, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3497 + }, + { + "epoch": 0.03498, + "grad_norm": 0.6599448025327912, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 3498 + }, + { + "epoch": 0.03499, + "grad_norm": 0.7172393903833442, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 3499 + }, + { + "epoch": 0.035, + "grad_norm": 0.6405157363259029, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 3500 + }, + { + "epoch": 0.03501, + "grad_norm": 0.5654687699259198, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3501 + }, + { + "epoch": 0.03502, + "grad_norm": 0.618591430417172, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 3502 + }, + { + "epoch": 0.03503, + "grad_norm": 0.6943996447233877, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3503 + }, + { + "epoch": 0.03504, + "grad_norm": 0.6621001515018434, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3504 + }, + { + "epoch": 0.03505, + "grad_norm": 0.597184679883689, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3505 + }, + { + "epoch": 0.03506, + "grad_norm": 0.5659263992573133, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3506 + }, + { + "epoch": 0.03507, + "grad_norm": 0.5952738202818753, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 3507 + }, + { + "epoch": 0.03508, + "grad_norm": 0.6319710224532459, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3508 + }, + { + "epoch": 0.03509, + "grad_norm": 0.7003635970154289, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 3509 + }, + { + "epoch": 0.0351, + "grad_norm": 0.7544632618493929, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 3510 + }, + { + "epoch": 0.03511, + "grad_norm": 0.8665700214441323, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 3511 + }, + { + "epoch": 0.03512, + "grad_norm": 0.8564897161621264, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 3512 + }, + { + "epoch": 0.03513, + "grad_norm": 0.9197981119921014, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 3513 + }, + { + "epoch": 0.03514, + "grad_norm": 0.8492203049153929, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3514 + }, + { + "epoch": 0.03515, + "grad_norm": 0.8901010567034684, + "learning_rate": 0.003, + "loss": 4.1863, + "step": 3515 + }, + { + "epoch": 0.03516, + "grad_norm": 0.9283878054698729, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 3516 + }, + { + "epoch": 0.03517, + "grad_norm": 0.8804785262543259, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 3517 + }, + { + "epoch": 0.03518, + "grad_norm": 0.8785403040867484, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3518 + }, + { + "epoch": 0.03519, + "grad_norm": 0.8144080502479469, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 3519 + }, + { + "epoch": 0.0352, + "grad_norm": 0.7871723040057216, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 3520 + }, + { + "epoch": 0.03521, + "grad_norm": 0.7368932966297566, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3521 + }, + { + "epoch": 0.03522, + "grad_norm": 0.7353338197865396, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 3522 + }, + { + "epoch": 0.03523, + "grad_norm": 0.80714477911056, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 3523 + }, + { + "epoch": 0.03524, + "grad_norm": 0.8183318273894814, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 3524 + }, + { + "epoch": 0.03525, + "grad_norm": 0.7083620836915047, + "learning_rate": 0.003, + "loss": 4.1538, + "step": 3525 + }, + { + "epoch": 0.03526, + "grad_norm": 0.6358752706280991, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 3526 + }, + { + "epoch": 0.03527, + "grad_norm": 0.5882454641700064, + "learning_rate": 0.003, + "loss": 4.1866, + "step": 3527 + }, + { + "epoch": 0.03528, + "grad_norm": 0.5418834609571274, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 3528 + }, + { + "epoch": 0.03529, + "grad_norm": 0.4938060985233679, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 3529 + }, + { + "epoch": 0.0353, + "grad_norm": 0.47222674272043563, + "learning_rate": 0.003, + "loss": 4.159, + "step": 3530 + }, + { + "epoch": 0.03531, + "grad_norm": 0.4494800604692719, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 3531 + }, + { + "epoch": 0.03532, + "grad_norm": 0.46025033508749047, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 3532 + }, + { + "epoch": 0.03533, + "grad_norm": 0.5287514582627705, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3533 + }, + { + "epoch": 0.03534, + "grad_norm": 0.6330388929499089, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 3534 + }, + { + "epoch": 0.03535, + "grad_norm": 0.6032059148264879, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3535 + }, + { + "epoch": 0.03536, + "grad_norm": 0.5769078485745863, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3536 + }, + { + "epoch": 0.03537, + "grad_norm": 0.7065282154758016, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3537 + }, + { + "epoch": 0.03538, + "grad_norm": 0.8671025316072222, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 3538 + }, + { + "epoch": 0.03539, + "grad_norm": 1.0158613655750308, + "learning_rate": 0.003, + "loss": 4.153, + "step": 3539 + }, + { + "epoch": 0.0354, + "grad_norm": 0.8533960614991879, + "learning_rate": 0.003, + "loss": 4.1884, + "step": 3540 + }, + { + "epoch": 0.03541, + "grad_norm": 0.6659419541207255, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 3541 + }, + { + "epoch": 0.03542, + "grad_norm": 0.6774420684048681, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3542 + }, + { + "epoch": 0.03543, + "grad_norm": 0.7402573585555693, + "learning_rate": 0.003, + "loss": 4.1676, + "step": 3543 + }, + { + "epoch": 0.03544, + "grad_norm": 0.6226751107190542, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3544 + }, + { + "epoch": 0.03545, + "grad_norm": 0.5532949073684433, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 3545 + }, + { + "epoch": 0.03546, + "grad_norm": 0.521811868333153, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3546 + }, + { + "epoch": 0.03547, + "grad_norm": 0.5542126267573285, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 3547 + }, + { + "epoch": 0.03548, + "grad_norm": 0.5294342446885085, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 3548 + }, + { + "epoch": 0.03549, + "grad_norm": 0.46845119607442226, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 3549 + }, + { + "epoch": 0.0355, + "grad_norm": 0.4173574369972876, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 3550 + }, + { + "epoch": 0.03551, + "grad_norm": 0.5597688720149216, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3551 + }, + { + "epoch": 0.03552, + "grad_norm": 0.684279124843178, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3552 + }, + { + "epoch": 0.03553, + "grad_norm": 0.7906598471557775, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 3553 + }, + { + "epoch": 0.03554, + "grad_norm": 0.8477558409372302, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 3554 + }, + { + "epoch": 0.03555, + "grad_norm": 0.9130757811703537, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3555 + }, + { + "epoch": 0.03556, + "grad_norm": 1.0233508386779784, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3556 + }, + { + "epoch": 0.03557, + "grad_norm": 0.9577495816046034, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 3557 + }, + { + "epoch": 0.03558, + "grad_norm": 0.6971645674236855, + "learning_rate": 0.003, + "loss": 4.1907, + "step": 3558 + }, + { + "epoch": 0.03559, + "grad_norm": 0.615773531816, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3559 + }, + { + "epoch": 0.0356, + "grad_norm": 0.7565427067348032, + "learning_rate": 0.003, + "loss": 4.179, + "step": 3560 + }, + { + "epoch": 0.03561, + "grad_norm": 0.7654646543935898, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 3561 + }, + { + "epoch": 0.03562, + "grad_norm": 0.7171787225466916, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3562 + }, + { + "epoch": 0.03563, + "grad_norm": 0.7202637052044578, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 3563 + }, + { + "epoch": 0.03564, + "grad_norm": 0.7813914185176697, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3564 + }, + { + "epoch": 0.03565, + "grad_norm": 0.9681384729900618, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 3565 + }, + { + "epoch": 0.03566, + "grad_norm": 0.9815276969782409, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 3566 + }, + { + "epoch": 0.03567, + "grad_norm": 0.8545575756116207, + "learning_rate": 0.003, + "loss": 4.1799, + "step": 3567 + }, + { + "epoch": 0.03568, + "grad_norm": 0.8579697593867708, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3568 + }, + { + "epoch": 0.03569, + "grad_norm": 0.9295006292416876, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 3569 + }, + { + "epoch": 0.0357, + "grad_norm": 0.7824959494204987, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 3570 + }, + { + "epoch": 0.03571, + "grad_norm": 0.715056502621953, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 3571 + }, + { + "epoch": 0.03572, + "grad_norm": 0.7939274654308445, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 3572 + }, + { + "epoch": 0.03573, + "grad_norm": 0.8990540619234125, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 3573 + }, + { + "epoch": 0.03574, + "grad_norm": 0.8924167007734459, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 3574 + }, + { + "epoch": 0.03575, + "grad_norm": 0.9633999021454877, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 3575 + }, + { + "epoch": 0.03576, + "grad_norm": 0.8337435293260566, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 3576 + }, + { + "epoch": 0.03577, + "grad_norm": 0.7990789046598609, + "learning_rate": 0.003, + "loss": 4.1996, + "step": 3577 + }, + { + "epoch": 0.03578, + "grad_norm": 0.740247280194808, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 3578 + }, + { + "epoch": 0.03579, + "grad_norm": 0.736535666256918, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 3579 + }, + { + "epoch": 0.0358, + "grad_norm": 0.7123148002792392, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 3580 + }, + { + "epoch": 0.03581, + "grad_norm": 0.6711995328778474, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 3581 + }, + { + "epoch": 0.03582, + "grad_norm": 0.6405574599918185, + "learning_rate": 0.003, + "loss": 4.166, + "step": 3582 + }, + { + "epoch": 0.03583, + "grad_norm": 0.6115022182938825, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 3583 + }, + { + "epoch": 0.03584, + "grad_norm": 0.5561430180920142, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 3584 + }, + { + "epoch": 0.03585, + "grad_norm": 0.5150442741371403, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 3585 + }, + { + "epoch": 0.03586, + "grad_norm": 0.5324475951916865, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 3586 + }, + { + "epoch": 0.03587, + "grad_norm": 0.5865627283177408, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 3587 + }, + { + "epoch": 0.03588, + "grad_norm": 0.6404975493032687, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 3588 + }, + { + "epoch": 0.03589, + "grad_norm": 0.6198477170373378, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 3589 + }, + { + "epoch": 0.0359, + "grad_norm": 0.5436091514667869, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 3590 + }, + { + "epoch": 0.03591, + "grad_norm": 0.5153318776417137, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 3591 + }, + { + "epoch": 0.03592, + "grad_norm": 0.4789482382988687, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3592 + }, + { + "epoch": 0.03593, + "grad_norm": 0.4332591255641176, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3593 + }, + { + "epoch": 0.03594, + "grad_norm": 0.46006005882861767, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 3594 + }, + { + "epoch": 0.03595, + "grad_norm": 0.39445716900409733, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 3595 + }, + { + "epoch": 0.03596, + "grad_norm": 0.40621407877255555, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3596 + }, + { + "epoch": 0.03597, + "grad_norm": 0.4360669588971265, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 3597 + }, + { + "epoch": 0.03598, + "grad_norm": 0.5144534194928903, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 3598 + }, + { + "epoch": 0.03599, + "grad_norm": 0.7248031737778203, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3599 + }, + { + "epoch": 0.036, + "grad_norm": 1.0800621032582747, + "learning_rate": 0.003, + "loss": 4.1892, + "step": 3600 + }, + { + "epoch": 0.03601, + "grad_norm": 0.9508478687848839, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3601 + }, + { + "epoch": 0.03602, + "grad_norm": 0.6514600121121226, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3602 + }, + { + "epoch": 0.03603, + "grad_norm": 0.6278285308162193, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3603 + }, + { + "epoch": 0.03604, + "grad_norm": 0.7950592611481373, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3604 + }, + { + "epoch": 0.03605, + "grad_norm": 0.7151902116041485, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3605 + }, + { + "epoch": 0.03606, + "grad_norm": 0.7086662453671662, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 3606 + }, + { + "epoch": 0.03607, + "grad_norm": 0.7175966521512804, + "learning_rate": 0.003, + "loss": 4.147, + "step": 3607 + }, + { + "epoch": 0.03608, + "grad_norm": 0.587764679138397, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 3608 + }, + { + "epoch": 0.03609, + "grad_norm": 0.5305480603629071, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3609 + }, + { + "epoch": 0.0361, + "grad_norm": 0.5522309324109884, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3610 + }, + { + "epoch": 0.03611, + "grad_norm": 0.5828315338943078, + "learning_rate": 0.003, + "loss": 4.125, + "step": 3611 + }, + { + "epoch": 0.03612, + "grad_norm": 0.6300633453025702, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 3612 + }, + { + "epoch": 0.03613, + "grad_norm": 0.617268364161966, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 3613 + }, + { + "epoch": 0.03614, + "grad_norm": 0.6309109479392245, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 3614 + }, + { + "epoch": 0.03615, + "grad_norm": 0.570782892253401, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 3615 + }, + { + "epoch": 0.03616, + "grad_norm": 0.5541049972487797, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3616 + }, + { + "epoch": 0.03617, + "grad_norm": 0.5951507982053302, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3617 + }, + { + "epoch": 0.03618, + "grad_norm": 0.7159646588024825, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3618 + }, + { + "epoch": 0.03619, + "grad_norm": 0.8176290428711182, + "learning_rate": 0.003, + "loss": 4.158, + "step": 3619 + }, + { + "epoch": 0.0362, + "grad_norm": 0.920159490234485, + "learning_rate": 0.003, + "loss": 4.173, + "step": 3620 + }, + { + "epoch": 0.03621, + "grad_norm": 0.9049277558974237, + "learning_rate": 0.003, + "loss": 4.173, + "step": 3621 + }, + { + "epoch": 0.03622, + "grad_norm": 0.9134672000988132, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 3622 + }, + { + "epoch": 0.03623, + "grad_norm": 0.8632217222934753, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 3623 + }, + { + "epoch": 0.03624, + "grad_norm": 0.7748608966158578, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3624 + }, + { + "epoch": 0.03625, + "grad_norm": 0.6549902845991534, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 3625 + }, + { + "epoch": 0.03626, + "grad_norm": 0.6025866731625074, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3626 + }, + { + "epoch": 0.03627, + "grad_norm": 0.5718452823869131, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 3627 + }, + { + "epoch": 0.03628, + "grad_norm": 0.5380938313028744, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3628 + }, + { + "epoch": 0.03629, + "grad_norm": 0.5051723060670195, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 3629 + }, + { + "epoch": 0.0363, + "grad_norm": 0.5404489026019382, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 3630 + }, + { + "epoch": 0.03631, + "grad_norm": 0.6007935793023943, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3631 + }, + { + "epoch": 0.03632, + "grad_norm": 0.620386838070519, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 3632 + }, + { + "epoch": 0.03633, + "grad_norm": 0.7326091251339881, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3633 + }, + { + "epoch": 0.03634, + "grad_norm": 0.9520374375008709, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 3634 + }, + { + "epoch": 0.03635, + "grad_norm": 0.9435644392052507, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 3635 + }, + { + "epoch": 0.03636, + "grad_norm": 0.8034652176528725, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 3636 + }, + { + "epoch": 0.03637, + "grad_norm": 0.8215010352273984, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 3637 + }, + { + "epoch": 0.03638, + "grad_norm": 0.7729628364890975, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 3638 + }, + { + "epoch": 0.03639, + "grad_norm": 0.7939041200619633, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 3639 + }, + { + "epoch": 0.0364, + "grad_norm": 0.7624978886682702, + "learning_rate": 0.003, + "loss": 4.1911, + "step": 3640 + }, + { + "epoch": 0.03641, + "grad_norm": 0.6972108651120591, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 3641 + }, + { + "epoch": 0.03642, + "grad_norm": 0.686374864648811, + "learning_rate": 0.003, + "loss": 4.192, + "step": 3642 + }, + { + "epoch": 0.03643, + "grad_norm": 0.6994446209399746, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 3643 + }, + { + "epoch": 0.03644, + "grad_norm": 0.7789189969606384, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 3644 + }, + { + "epoch": 0.03645, + "grad_norm": 0.7984886139835612, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3645 + }, + { + "epoch": 0.03646, + "grad_norm": 0.6181112574802159, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 3646 + }, + { + "epoch": 0.03647, + "grad_norm": 0.6061654420727669, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3647 + }, + { + "epoch": 0.03648, + "grad_norm": 0.6696338465160281, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 3648 + }, + { + "epoch": 0.03649, + "grad_norm": 0.7331012822671169, + "learning_rate": 0.003, + "loss": 4.1808, + "step": 3649 + }, + { + "epoch": 0.0365, + "grad_norm": 0.7037525814274137, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3650 + }, + { + "epoch": 0.03651, + "grad_norm": 0.692119661540781, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 3651 + }, + { + "epoch": 0.03652, + "grad_norm": 0.6604395418921041, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 3652 + }, + { + "epoch": 0.03653, + "grad_norm": 0.6346433017565251, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 3653 + }, + { + "epoch": 0.03654, + "grad_norm": 0.6759345560485471, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3654 + }, + { + "epoch": 0.03655, + "grad_norm": 0.7374777892156636, + "learning_rate": 0.003, + "loss": 4.1719, + "step": 3655 + }, + { + "epoch": 0.03656, + "grad_norm": 0.8895381978397107, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 3656 + }, + { + "epoch": 0.03657, + "grad_norm": 1.0700577195843572, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 3657 + }, + { + "epoch": 0.03658, + "grad_norm": 0.8922825561167493, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 3658 + }, + { + "epoch": 0.03659, + "grad_norm": 0.754621627580339, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 3659 + }, + { + "epoch": 0.0366, + "grad_norm": 0.8002619699354713, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 3660 + }, + { + "epoch": 0.03661, + "grad_norm": 0.7711130288745789, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 3661 + }, + { + "epoch": 0.03662, + "grad_norm": 0.7607029556463581, + "learning_rate": 0.003, + "loss": 4.162, + "step": 3662 + }, + { + "epoch": 0.03663, + "grad_norm": 0.7440287835939471, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 3663 + }, + { + "epoch": 0.03664, + "grad_norm": 0.563780432136041, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 3664 + }, + { + "epoch": 0.03665, + "grad_norm": 0.4357399894939763, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 3665 + }, + { + "epoch": 0.03666, + "grad_norm": 0.45515608243379135, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 3666 + }, + { + "epoch": 0.03667, + "grad_norm": 0.4706495224710868, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 3667 + }, + { + "epoch": 0.03668, + "grad_norm": 0.4175208514628951, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 3668 + }, + { + "epoch": 0.03669, + "grad_norm": 0.3785566363290307, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3669 + }, + { + "epoch": 0.0367, + "grad_norm": 0.3437728850428848, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 3670 + }, + { + "epoch": 0.03671, + "grad_norm": 0.3820679778361713, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 3671 + }, + { + "epoch": 0.03672, + "grad_norm": 0.44348210068257454, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 3672 + }, + { + "epoch": 0.03673, + "grad_norm": 0.5490971493740642, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 3673 + }, + { + "epoch": 0.03674, + "grad_norm": 0.6223490419700332, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 3674 + }, + { + "epoch": 0.03675, + "grad_norm": 0.6342035142985044, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 3675 + }, + { + "epoch": 0.03676, + "grad_norm": 0.631553220636773, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 3676 + }, + { + "epoch": 0.03677, + "grad_norm": 0.8208223460663321, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3677 + }, + { + "epoch": 0.03678, + "grad_norm": 1.057395404894455, + "learning_rate": 0.003, + "loss": 4.1876, + "step": 3678 + }, + { + "epoch": 0.03679, + "grad_norm": 1.059489837092536, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 3679 + }, + { + "epoch": 0.0368, + "grad_norm": 0.9648807335139974, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3680 + }, + { + "epoch": 0.03681, + "grad_norm": 0.9082948485919852, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 3681 + }, + { + "epoch": 0.03682, + "grad_norm": 0.7608012026223391, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 3682 + }, + { + "epoch": 0.03683, + "grad_norm": 0.7012632382999905, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 3683 + }, + { + "epoch": 0.03684, + "grad_norm": 0.7244234234962599, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3684 + }, + { + "epoch": 0.03685, + "grad_norm": 0.6796055439571254, + "learning_rate": 0.003, + "loss": 4.153, + "step": 3685 + }, + { + "epoch": 0.03686, + "grad_norm": 0.6397461073626532, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 3686 + }, + { + "epoch": 0.03687, + "grad_norm": 0.6837549545590554, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 3687 + }, + { + "epoch": 0.03688, + "grad_norm": 0.612454393388502, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 3688 + }, + { + "epoch": 0.03689, + "grad_norm": 0.6808423570337984, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3689 + }, + { + "epoch": 0.0369, + "grad_norm": 0.6866647703326533, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 3690 + }, + { + "epoch": 0.03691, + "grad_norm": 0.7296388488733391, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3691 + }, + { + "epoch": 0.03692, + "grad_norm": 0.8733542369985744, + "learning_rate": 0.003, + "loss": 4.2056, + "step": 3692 + }, + { + "epoch": 0.03693, + "grad_norm": 0.9120312999635036, + "learning_rate": 0.003, + "loss": 4.1787, + "step": 3693 + }, + { + "epoch": 0.03694, + "grad_norm": 0.7928228320045024, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 3694 + }, + { + "epoch": 0.03695, + "grad_norm": 0.7574597489784519, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3695 + }, + { + "epoch": 0.03696, + "grad_norm": 0.7264595127076895, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 3696 + }, + { + "epoch": 0.03697, + "grad_norm": 0.797502557212387, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 3697 + }, + { + "epoch": 0.03698, + "grad_norm": 0.8741564039842601, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 3698 + }, + { + "epoch": 0.03699, + "grad_norm": 0.8855946379470947, + "learning_rate": 0.003, + "loss": 4.1856, + "step": 3699 + }, + { + "epoch": 0.037, + "grad_norm": 0.8586025114326333, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 3700 + }, + { + "epoch": 0.03701, + "grad_norm": 0.8930614814476187, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 3701 + }, + { + "epoch": 0.03702, + "grad_norm": 0.8802482281802025, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 3702 + }, + { + "epoch": 0.03703, + "grad_norm": 0.7549215853060027, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 3703 + }, + { + "epoch": 0.03704, + "grad_norm": 0.6295591556831198, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 3704 + }, + { + "epoch": 0.03705, + "grad_norm": 0.5580114287621604, + "learning_rate": 0.003, + "loss": 4.1901, + "step": 3705 + }, + { + "epoch": 0.03706, + "grad_norm": 0.493748778061454, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3706 + }, + { + "epoch": 0.03707, + "grad_norm": 0.501411513908697, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 3707 + }, + { + "epoch": 0.03708, + "grad_norm": 0.46976615957109447, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3708 + }, + { + "epoch": 0.03709, + "grad_norm": 0.4882450257172406, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 3709 + }, + { + "epoch": 0.0371, + "grad_norm": 0.5609237339392169, + "learning_rate": 0.003, + "loss": 4.1796, + "step": 3710 + }, + { + "epoch": 0.03711, + "grad_norm": 0.577744009339436, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3711 + }, + { + "epoch": 0.03712, + "grad_norm": 0.5617641082054192, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3712 + }, + { + "epoch": 0.03713, + "grad_norm": 0.6129339279774879, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 3713 + }, + { + "epoch": 0.03714, + "grad_norm": 0.761778934710984, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 3714 + }, + { + "epoch": 0.03715, + "grad_norm": 0.936702433405622, + "learning_rate": 0.003, + "loss": 4.1692, + "step": 3715 + }, + { + "epoch": 0.03716, + "grad_norm": 1.0004110813808647, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 3716 + }, + { + "epoch": 0.03717, + "grad_norm": 0.9246847467387567, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 3717 + }, + { + "epoch": 0.03718, + "grad_norm": 0.8130004478592973, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3718 + }, + { + "epoch": 0.03719, + "grad_norm": 0.8110626454843745, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 3719 + }, + { + "epoch": 0.0372, + "grad_norm": 0.6478944328491517, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 3720 + }, + { + "epoch": 0.03721, + "grad_norm": 0.6036054997551548, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 3721 + }, + { + "epoch": 0.03722, + "grad_norm": 0.47934112751684016, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3722 + }, + { + "epoch": 0.03723, + "grad_norm": 0.5465870050613278, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 3723 + }, + { + "epoch": 0.03724, + "grad_norm": 0.5794138576951311, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3724 + }, + { + "epoch": 0.03725, + "grad_norm": 0.5728374193198114, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 3725 + }, + { + "epoch": 0.03726, + "grad_norm": 0.5999625053130218, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3726 + }, + { + "epoch": 0.03727, + "grad_norm": 0.6205981266168359, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 3727 + }, + { + "epoch": 0.03728, + "grad_norm": 0.6360363945769689, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 3728 + }, + { + "epoch": 0.03729, + "grad_norm": 0.5842605492528835, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 3729 + }, + { + "epoch": 0.0373, + "grad_norm": 0.5328025691146133, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3730 + }, + { + "epoch": 0.03731, + "grad_norm": 0.53824919717479, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 3731 + }, + { + "epoch": 0.03732, + "grad_norm": 0.6816149389539496, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 3732 + }, + { + "epoch": 0.03733, + "grad_norm": 0.7649227254747324, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 3733 + }, + { + "epoch": 0.03734, + "grad_norm": 0.723858182349595, + "learning_rate": 0.003, + "loss": 4.147, + "step": 3734 + }, + { + "epoch": 0.03735, + "grad_norm": 0.7221215059978716, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3735 + }, + { + "epoch": 0.03736, + "grad_norm": 0.7400835788486658, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 3736 + }, + { + "epoch": 0.03737, + "grad_norm": 0.6969381819145593, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 3737 + }, + { + "epoch": 0.03738, + "grad_norm": 0.6127048568575166, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3738 + }, + { + "epoch": 0.03739, + "grad_norm": 0.6440493642792959, + "learning_rate": 0.003, + "loss": 4.139, + "step": 3739 + }, + { + "epoch": 0.0374, + "grad_norm": 0.72519391026854, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3740 + }, + { + "epoch": 0.03741, + "grad_norm": 0.7715264840802699, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 3741 + }, + { + "epoch": 0.03742, + "grad_norm": 0.6234960786729838, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 3742 + }, + { + "epoch": 0.03743, + "grad_norm": 0.529614255393143, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 3743 + }, + { + "epoch": 0.03744, + "grad_norm": 0.5395949119845926, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 3744 + }, + { + "epoch": 0.03745, + "grad_norm": 0.5717733035088027, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 3745 + }, + { + "epoch": 0.03746, + "grad_norm": 0.7148827950766982, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3746 + }, + { + "epoch": 0.03747, + "grad_norm": 0.9619906254460838, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 3747 + }, + { + "epoch": 0.03748, + "grad_norm": 0.9454260095382823, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 3748 + }, + { + "epoch": 0.03749, + "grad_norm": 0.8505540045662432, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 3749 + }, + { + "epoch": 0.0375, + "grad_norm": 0.8001368316453729, + "learning_rate": 0.003, + "loss": 4.1825, + "step": 3750 + }, + { + "epoch": 0.03751, + "grad_norm": 0.6787039450042418, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3751 + }, + { + "epoch": 0.03752, + "grad_norm": 0.6675510373317245, + "learning_rate": 0.003, + "loss": 4.1573, + "step": 3752 + }, + { + "epoch": 0.03753, + "grad_norm": 0.659564153036602, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 3753 + }, + { + "epoch": 0.03754, + "grad_norm": 0.5879492269067329, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 3754 + }, + { + "epoch": 0.03755, + "grad_norm": 0.5251202407537541, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3755 + }, + { + "epoch": 0.03756, + "grad_norm": 0.47935868138717536, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 3756 + }, + { + "epoch": 0.03757, + "grad_norm": 0.5841333632592701, + "learning_rate": 0.003, + "loss": 4.16, + "step": 3757 + }, + { + "epoch": 0.03758, + "grad_norm": 0.7521988739330352, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3758 + }, + { + "epoch": 0.03759, + "grad_norm": 0.8503529306111015, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 3759 + }, + { + "epoch": 0.0376, + "grad_norm": 0.8653069459339167, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 3760 + }, + { + "epoch": 0.03761, + "grad_norm": 0.8195705554984235, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3761 + }, + { + "epoch": 0.03762, + "grad_norm": 0.7776430458932468, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 3762 + }, + { + "epoch": 0.03763, + "grad_norm": 0.8090187425891137, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 3763 + }, + { + "epoch": 0.03764, + "grad_norm": 0.7299629395412605, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 3764 + }, + { + "epoch": 0.03765, + "grad_norm": 0.7220941046527883, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 3765 + }, + { + "epoch": 0.03766, + "grad_norm": 0.7938574946684543, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3766 + }, + { + "epoch": 0.03767, + "grad_norm": 0.848179339225032, + "learning_rate": 0.003, + "loss": 4.1885, + "step": 3767 + }, + { + "epoch": 0.03768, + "grad_norm": 0.8273555270559375, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 3768 + }, + { + "epoch": 0.03769, + "grad_norm": 0.7189795085544652, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 3769 + }, + { + "epoch": 0.0377, + "grad_norm": 0.6877419998361242, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 3770 + }, + { + "epoch": 0.03771, + "grad_norm": 0.7347886642206836, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 3771 + }, + { + "epoch": 0.03772, + "grad_norm": 0.7369268379376811, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3772 + }, + { + "epoch": 0.03773, + "grad_norm": 0.7648153892095982, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 3773 + }, + { + "epoch": 0.03774, + "grad_norm": 0.689330468097358, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 3774 + }, + { + "epoch": 0.03775, + "grad_norm": 0.7165358486039451, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 3775 + }, + { + "epoch": 0.03776, + "grad_norm": 0.7489974209205834, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 3776 + }, + { + "epoch": 0.03777, + "grad_norm": 0.8249125931936281, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 3777 + }, + { + "epoch": 0.03778, + "grad_norm": 0.7978997683857884, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 3778 + }, + { + "epoch": 0.03779, + "grad_norm": 0.7763812864712815, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 3779 + }, + { + "epoch": 0.0378, + "grad_norm": 0.6854145264424516, + "learning_rate": 0.003, + "loss": 4.163, + "step": 3780 + }, + { + "epoch": 0.03781, + "grad_norm": 0.7017290642180594, + "learning_rate": 0.003, + "loss": 4.1924, + "step": 3781 + }, + { + "epoch": 0.03782, + "grad_norm": 0.7499872472851988, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3782 + }, + { + "epoch": 0.03783, + "grad_norm": 0.7066696741848768, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 3783 + }, + { + "epoch": 0.03784, + "grad_norm": 0.6293823118533233, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 3784 + }, + { + "epoch": 0.03785, + "grad_norm": 0.5600002308125822, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3785 + }, + { + "epoch": 0.03786, + "grad_norm": 0.4985251419936004, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3786 + }, + { + "epoch": 0.03787, + "grad_norm": 0.4682470683406906, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3787 + }, + { + "epoch": 0.03788, + "grad_norm": 0.507866310001821, + "learning_rate": 0.003, + "loss": 4.1774, + "step": 3788 + }, + { + "epoch": 0.03789, + "grad_norm": 0.5463073884456529, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 3789 + }, + { + "epoch": 0.0379, + "grad_norm": 0.6172159531979761, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3790 + }, + { + "epoch": 0.03791, + "grad_norm": 0.7402780853790922, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 3791 + }, + { + "epoch": 0.03792, + "grad_norm": 0.9096314434511164, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 3792 + }, + { + "epoch": 0.03793, + "grad_norm": 0.8009156787875767, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 3793 + }, + { + "epoch": 0.03794, + "grad_norm": 0.6007769119418681, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 3794 + }, + { + "epoch": 0.03795, + "grad_norm": 0.6248284046940696, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3795 + }, + { + "epoch": 0.03796, + "grad_norm": 0.6313675517193291, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3796 + }, + { + "epoch": 0.03797, + "grad_norm": 0.6260974457476847, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3797 + }, + { + "epoch": 0.03798, + "grad_norm": 0.6329676643978727, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 3798 + }, + { + "epoch": 0.03799, + "grad_norm": 0.6413791763016119, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 3799 + }, + { + "epoch": 0.038, + "grad_norm": 0.5746316704038654, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 3800 + }, + { + "epoch": 0.03801, + "grad_norm": 0.5630475643545394, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3801 + }, + { + "epoch": 0.03802, + "grad_norm": 0.5962762680426152, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 3802 + }, + { + "epoch": 0.03803, + "grad_norm": 0.7181192675393114, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 3803 + }, + { + "epoch": 0.03804, + "grad_norm": 0.7862836747313122, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3804 + }, + { + "epoch": 0.03805, + "grad_norm": 0.8710510112491459, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3805 + }, + { + "epoch": 0.03806, + "grad_norm": 0.8521659658182485, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 3806 + }, + { + "epoch": 0.03807, + "grad_norm": 0.7446144627159563, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 3807 + }, + { + "epoch": 0.03808, + "grad_norm": 0.7136381828235824, + "learning_rate": 0.003, + "loss": 4.167, + "step": 3808 + }, + { + "epoch": 0.03809, + "grad_norm": 0.6617555150784667, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3809 + }, + { + "epoch": 0.0381, + "grad_norm": 0.5589720929569325, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3810 + }, + { + "epoch": 0.03811, + "grad_norm": 0.513976956124665, + "learning_rate": 0.003, + "loss": 4.1583, + "step": 3811 + }, + { + "epoch": 0.03812, + "grad_norm": 0.5341053812083173, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 3812 + }, + { + "epoch": 0.03813, + "grad_norm": 0.5180551827098935, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3813 + }, + { + "epoch": 0.03814, + "grad_norm": 0.5664522386981432, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3814 + }, + { + "epoch": 0.03815, + "grad_norm": 0.7610361605531731, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 3815 + }, + { + "epoch": 0.03816, + "grad_norm": 0.9587327701936716, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 3816 + }, + { + "epoch": 0.03817, + "grad_norm": 0.935563437309735, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3817 + }, + { + "epoch": 0.03818, + "grad_norm": 1.0060426088546508, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 3818 + }, + { + "epoch": 0.03819, + "grad_norm": 0.9721837543200701, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3819 + }, + { + "epoch": 0.0382, + "grad_norm": 1.1304312225449777, + "learning_rate": 0.003, + "loss": 4.2256, + "step": 3820 + }, + { + "epoch": 0.03821, + "grad_norm": 1.0177206332843154, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 3821 + }, + { + "epoch": 0.03822, + "grad_norm": 1.031220537341774, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 3822 + }, + { + "epoch": 0.03823, + "grad_norm": 1.0388061376025717, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 3823 + }, + { + "epoch": 0.03824, + "grad_norm": 0.9677858739885562, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 3824 + }, + { + "epoch": 0.03825, + "grad_norm": 0.937987879509307, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 3825 + }, + { + "epoch": 0.03826, + "grad_norm": 0.8230886831468718, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 3826 + }, + { + "epoch": 0.03827, + "grad_norm": 0.6935375745622523, + "learning_rate": 0.003, + "loss": 4.2068, + "step": 3827 + }, + { + "epoch": 0.03828, + "grad_norm": 0.6856721015265079, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 3828 + }, + { + "epoch": 0.03829, + "grad_norm": 0.6394378471082416, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 3829 + }, + { + "epoch": 0.0383, + "grad_norm": 0.6570235063756633, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 3830 + }, + { + "epoch": 0.03831, + "grad_norm": 0.6505218073754993, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 3831 + }, + { + "epoch": 0.03832, + "grad_norm": 0.5403168994861741, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 3832 + }, + { + "epoch": 0.03833, + "grad_norm": 0.42095483529999317, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 3833 + }, + { + "epoch": 0.03834, + "grad_norm": 0.40483366089412304, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 3834 + }, + { + "epoch": 0.03835, + "grad_norm": 0.4630812031805051, + "learning_rate": 0.003, + "loss": 4.172, + "step": 3835 + }, + { + "epoch": 0.03836, + "grad_norm": 0.4841208007987568, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 3836 + }, + { + "epoch": 0.03837, + "grad_norm": 0.536907199381203, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 3837 + }, + { + "epoch": 0.03838, + "grad_norm": 0.6117335045256618, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3838 + }, + { + "epoch": 0.03839, + "grad_norm": 0.6780085079827656, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 3839 + }, + { + "epoch": 0.0384, + "grad_norm": 0.697829286285506, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3840 + }, + { + "epoch": 0.03841, + "grad_norm": 0.8027464809531286, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 3841 + }, + { + "epoch": 0.03842, + "grad_norm": 0.8177694533436255, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 3842 + }, + { + "epoch": 0.03843, + "grad_norm": 0.6989170877767756, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 3843 + }, + { + "epoch": 0.03844, + "grad_norm": 0.6383070406944257, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 3844 + }, + { + "epoch": 0.03845, + "grad_norm": 0.6201943361574275, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 3845 + }, + { + "epoch": 0.03846, + "grad_norm": 0.6699207633477972, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 3846 + }, + { + "epoch": 0.03847, + "grad_norm": 0.5790976587711703, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 3847 + }, + { + "epoch": 0.03848, + "grad_norm": 0.5911075701092868, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 3848 + }, + { + "epoch": 0.03849, + "grad_norm": 0.6969380862395563, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 3849 + }, + { + "epoch": 0.0385, + "grad_norm": 0.7972901632396154, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3850 + }, + { + "epoch": 0.03851, + "grad_norm": 0.8999501436569268, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 3851 + }, + { + "epoch": 0.03852, + "grad_norm": 0.9359674147442539, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 3852 + }, + { + "epoch": 0.03853, + "grad_norm": 0.7466141042924882, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 3853 + }, + { + "epoch": 0.03854, + "grad_norm": 0.5969895365980001, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3854 + }, + { + "epoch": 0.03855, + "grad_norm": 0.5270288768727529, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 3855 + }, + { + "epoch": 0.03856, + "grad_norm": 0.5338061114857245, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 3856 + }, + { + "epoch": 0.03857, + "grad_norm": 0.5294295191115548, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 3857 + }, + { + "epoch": 0.03858, + "grad_norm": 0.5523522268399486, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 3858 + }, + { + "epoch": 0.03859, + "grad_norm": 0.6152187330517869, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3859 + }, + { + "epoch": 0.0386, + "grad_norm": 0.6117532516023865, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3860 + }, + { + "epoch": 0.03861, + "grad_norm": 0.5524498434352159, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 3861 + }, + { + "epoch": 0.03862, + "grad_norm": 0.5236517072626808, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 3862 + }, + { + "epoch": 0.03863, + "grad_norm": 0.6248759021675342, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 3863 + }, + { + "epoch": 0.03864, + "grad_norm": 0.679464652102799, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3864 + }, + { + "epoch": 0.03865, + "grad_norm": 0.6453393413543939, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 3865 + }, + { + "epoch": 0.03866, + "grad_norm": 0.6101015046125351, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 3866 + }, + { + "epoch": 0.03867, + "grad_norm": 0.5167462630710293, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3867 + }, + { + "epoch": 0.03868, + "grad_norm": 0.501446932546633, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 3868 + }, + { + "epoch": 0.03869, + "grad_norm": 0.4709811100238584, + "learning_rate": 0.003, + "loss": 4.132, + "step": 3869 + }, + { + "epoch": 0.0387, + "grad_norm": 0.4875337874718785, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3870 + }, + { + "epoch": 0.03871, + "grad_norm": 0.5681317313929587, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3871 + }, + { + "epoch": 0.03872, + "grad_norm": 0.6978195894520031, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 3872 + }, + { + "epoch": 0.03873, + "grad_norm": 0.9026562899987328, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 3873 + }, + { + "epoch": 0.03874, + "grad_norm": 1.2643110286573773, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 3874 + }, + { + "epoch": 0.03875, + "grad_norm": 0.8522074037692399, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 3875 + }, + { + "epoch": 0.03876, + "grad_norm": 0.7221351700519376, + "learning_rate": 0.003, + "loss": 4.1817, + "step": 3876 + }, + { + "epoch": 0.03877, + "grad_norm": 0.6790217877458674, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 3877 + }, + { + "epoch": 0.03878, + "grad_norm": 0.6442745541470235, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 3878 + }, + { + "epoch": 0.03879, + "grad_norm": 0.5773572680715529, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 3879 + }, + { + "epoch": 0.0388, + "grad_norm": 0.6630378193548411, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 3880 + }, + { + "epoch": 0.03881, + "grad_norm": 0.6679169810032245, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 3881 + }, + { + "epoch": 0.03882, + "grad_norm": 0.7019796842116967, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 3882 + }, + { + "epoch": 0.03883, + "grad_norm": 0.6656806453760354, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 3883 + }, + { + "epoch": 0.03884, + "grad_norm": 0.7319329418461972, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 3884 + }, + { + "epoch": 0.03885, + "grad_norm": 0.9033013251391556, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 3885 + }, + { + "epoch": 0.03886, + "grad_norm": 1.0950277457821216, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 3886 + }, + { + "epoch": 0.03887, + "grad_norm": 0.7794992948946918, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 3887 + }, + { + "epoch": 0.03888, + "grad_norm": 0.6776556880519633, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 3888 + }, + { + "epoch": 0.03889, + "grad_norm": 0.6740634088352366, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 3889 + }, + { + "epoch": 0.0389, + "grad_norm": 0.6800323734503214, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 3890 + }, + { + "epoch": 0.03891, + "grad_norm": 0.6548487525930513, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3891 + }, + { + "epoch": 0.03892, + "grad_norm": 0.6949410167297108, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3892 + }, + { + "epoch": 0.03893, + "grad_norm": 0.71600442881932, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 3893 + }, + { + "epoch": 0.03894, + "grad_norm": 0.6445397470488516, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 3894 + }, + { + "epoch": 0.03895, + "grad_norm": 0.5602823203406133, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 3895 + }, + { + "epoch": 0.03896, + "grad_norm": 0.5429794704774388, + "learning_rate": 0.003, + "loss": 4.151, + "step": 3896 + }, + { + "epoch": 0.03897, + "grad_norm": 0.5936531994862636, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3897 + }, + { + "epoch": 0.03898, + "grad_norm": 0.6856826306823404, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3898 + }, + { + "epoch": 0.03899, + "grad_norm": 0.7274726655026192, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 3899 + }, + { + "epoch": 0.039, + "grad_norm": 0.8037659881555415, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 3900 + }, + { + "epoch": 0.03901, + "grad_norm": 0.8316671306971751, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 3901 + }, + { + "epoch": 0.03902, + "grad_norm": 0.8611987009716957, + "learning_rate": 0.003, + "loss": 4.145, + "step": 3902 + }, + { + "epoch": 0.03903, + "grad_norm": 0.8344071211818236, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 3903 + }, + { + "epoch": 0.03904, + "grad_norm": 0.7880927476353365, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 3904 + }, + { + "epoch": 0.03905, + "grad_norm": 0.778724013223126, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 3905 + }, + { + "epoch": 0.03906, + "grad_norm": 0.7715175718675301, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 3906 + }, + { + "epoch": 0.03907, + "grad_norm": 0.7877862833527057, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 3907 + }, + { + "epoch": 0.03908, + "grad_norm": 0.5993067619626092, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 3908 + }, + { + "epoch": 0.03909, + "grad_norm": 0.5889198426858406, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 3909 + }, + { + "epoch": 0.0391, + "grad_norm": 0.5523722340939393, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 3910 + }, + { + "epoch": 0.03911, + "grad_norm": 0.5127903685980498, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 3911 + }, + { + "epoch": 0.03912, + "grad_norm": 0.4766305366831587, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 3912 + }, + { + "epoch": 0.03913, + "grad_norm": 0.5086608596079191, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 3913 + }, + { + "epoch": 0.03914, + "grad_norm": 0.5215056781723886, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 3914 + }, + { + "epoch": 0.03915, + "grad_norm": 0.5669876961115498, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3915 + }, + { + "epoch": 0.03916, + "grad_norm": 0.5924249172540584, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 3916 + }, + { + "epoch": 0.03917, + "grad_norm": 0.6024923400756873, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3917 + }, + { + "epoch": 0.03918, + "grad_norm": 0.7101951940455182, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 3918 + }, + { + "epoch": 0.03919, + "grad_norm": 0.8265326178992785, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3919 + }, + { + "epoch": 0.0392, + "grad_norm": 1.076222106498144, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 3920 + }, + { + "epoch": 0.03921, + "grad_norm": 0.9195914439109347, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 3921 + }, + { + "epoch": 0.03922, + "grad_norm": 0.8170632240304666, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 3922 + }, + { + "epoch": 0.03923, + "grad_norm": 0.8296100524137636, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 3923 + }, + { + "epoch": 0.03924, + "grad_norm": 0.8264396450686449, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3924 + }, + { + "epoch": 0.03925, + "grad_norm": 0.8983164449879958, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 3925 + }, + { + "epoch": 0.03926, + "grad_norm": 0.8746883993886628, + "learning_rate": 0.003, + "loss": 4.1589, + "step": 3926 + }, + { + "epoch": 0.03927, + "grad_norm": 0.8997462460768232, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3927 + }, + { + "epoch": 0.03928, + "grad_norm": 0.8314963835673442, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 3928 + }, + { + "epoch": 0.03929, + "grad_norm": 0.7199296155552709, + "learning_rate": 0.003, + "loss": 4.191, + "step": 3929 + }, + { + "epoch": 0.0393, + "grad_norm": 0.6856199549997152, + "learning_rate": 0.003, + "loss": 4.175, + "step": 3930 + }, + { + "epoch": 0.03931, + "grad_norm": 0.6802907106791054, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 3931 + }, + { + "epoch": 0.03932, + "grad_norm": 0.7056782066262727, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 3932 + }, + { + "epoch": 0.03933, + "grad_norm": 0.84224228414277, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 3933 + }, + { + "epoch": 0.03934, + "grad_norm": 0.9417246071568263, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 3934 + }, + { + "epoch": 0.03935, + "grad_norm": 0.8605169810225617, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 3935 + }, + { + "epoch": 0.03936, + "grad_norm": 0.8668689820297688, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 3936 + }, + { + "epoch": 0.03937, + "grad_norm": 0.7302292343482224, + "learning_rate": 0.003, + "loss": 4.1638, + "step": 3937 + }, + { + "epoch": 0.03938, + "grad_norm": 0.7852118199812202, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 3938 + }, + { + "epoch": 0.03939, + "grad_norm": 0.7372234256187263, + "learning_rate": 0.003, + "loss": 4.144, + "step": 3939 + }, + { + "epoch": 0.0394, + "grad_norm": 0.7135965416181232, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 3940 + }, + { + "epoch": 0.03941, + "grad_norm": 0.598564224938674, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 3941 + }, + { + "epoch": 0.03942, + "grad_norm": 0.5266569159659855, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 3942 + }, + { + "epoch": 0.03943, + "grad_norm": 0.5409156386273862, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 3943 + }, + { + "epoch": 0.03944, + "grad_norm": 0.604265732288973, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 3944 + }, + { + "epoch": 0.03945, + "grad_norm": 0.5831656289599461, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3945 + }, + { + "epoch": 0.03946, + "grad_norm": 0.5502377902249579, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 3946 + }, + { + "epoch": 0.03947, + "grad_norm": 0.48330198086471926, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 3947 + }, + { + "epoch": 0.03948, + "grad_norm": 0.45335380074512316, + "learning_rate": 0.003, + "loss": 4.18, + "step": 3948 + }, + { + "epoch": 0.03949, + "grad_norm": 0.4365298076816215, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 3949 + }, + { + "epoch": 0.0395, + "grad_norm": 0.4330234892790934, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 3950 + }, + { + "epoch": 0.03951, + "grad_norm": 0.4549033232539336, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 3951 + }, + { + "epoch": 0.03952, + "grad_norm": 0.5088568380699875, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 3952 + }, + { + "epoch": 0.03953, + "grad_norm": 0.611258865255683, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3953 + }, + { + "epoch": 0.03954, + "grad_norm": 0.6882585765157053, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 3954 + }, + { + "epoch": 0.03955, + "grad_norm": 0.7656685556557281, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 3955 + }, + { + "epoch": 0.03956, + "grad_norm": 0.8113542097479195, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 3956 + }, + { + "epoch": 0.03957, + "grad_norm": 0.8632342447207374, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 3957 + }, + { + "epoch": 0.03958, + "grad_norm": 1.0300805032856515, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 3958 + }, + { + "epoch": 0.03959, + "grad_norm": 1.4201099525549408, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 3959 + }, + { + "epoch": 0.0396, + "grad_norm": 0.8108000226279894, + "learning_rate": 0.003, + "loss": 4.1741, + "step": 3960 + }, + { + "epoch": 0.03961, + "grad_norm": 0.8230830778410857, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 3961 + }, + { + "epoch": 0.03962, + "grad_norm": 0.8784449509664598, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 3962 + }, + { + "epoch": 0.03963, + "grad_norm": 0.8465039181944972, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 3963 + }, + { + "epoch": 0.03964, + "grad_norm": 0.8689576663713054, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 3964 + }, + { + "epoch": 0.03965, + "grad_norm": 0.8789965519139235, + "learning_rate": 0.003, + "loss": 4.1722, + "step": 3965 + }, + { + "epoch": 0.03966, + "grad_norm": 0.7504804588349563, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 3966 + }, + { + "epoch": 0.03967, + "grad_norm": 0.7521340457571327, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 3967 + }, + { + "epoch": 0.03968, + "grad_norm": 0.6566639153668963, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 3968 + }, + { + "epoch": 0.03969, + "grad_norm": 0.6193517682452371, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 3969 + }, + { + "epoch": 0.0397, + "grad_norm": 0.618307384827639, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 3970 + }, + { + "epoch": 0.03971, + "grad_norm": 0.5807398202018539, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 3971 + }, + { + "epoch": 0.03972, + "grad_norm": 0.5003114971259446, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 3972 + }, + { + "epoch": 0.03973, + "grad_norm": 0.5672839085406842, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 3973 + }, + { + "epoch": 0.03974, + "grad_norm": 0.6148540153542106, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 3974 + }, + { + "epoch": 0.03975, + "grad_norm": 0.6299273778345961, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 3975 + }, + { + "epoch": 0.03976, + "grad_norm": 0.7124442131113389, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 3976 + }, + { + "epoch": 0.03977, + "grad_norm": 0.8025542381027738, + "learning_rate": 0.003, + "loss": 4.178, + "step": 3977 + }, + { + "epoch": 0.03978, + "grad_norm": 0.8588749028025513, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 3978 + }, + { + "epoch": 0.03979, + "grad_norm": 0.8573131443233668, + "learning_rate": 0.003, + "loss": 4.169, + "step": 3979 + }, + { + "epoch": 0.0398, + "grad_norm": 0.99280645435225, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 3980 + }, + { + "epoch": 0.03981, + "grad_norm": 1.1460073770386754, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 3981 + }, + { + "epoch": 0.03982, + "grad_norm": 0.9873321030261758, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3982 + }, + { + "epoch": 0.03983, + "grad_norm": 0.8416549998631022, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 3983 + }, + { + "epoch": 0.03984, + "grad_norm": 0.7402406551683491, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 3984 + }, + { + "epoch": 0.03985, + "grad_norm": 0.7030279267262828, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 3985 + }, + { + "epoch": 0.03986, + "grad_norm": 0.6894694286447128, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 3986 + }, + { + "epoch": 0.03987, + "grad_norm": 0.7314973515669282, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 3987 + }, + { + "epoch": 0.03988, + "grad_norm": 0.8590107763209545, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 3988 + }, + { + "epoch": 0.03989, + "grad_norm": 1.074747879568787, + "learning_rate": 0.003, + "loss": 4.1798, + "step": 3989 + }, + { + "epoch": 0.0399, + "grad_norm": 0.8434989382900636, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 3990 + }, + { + "epoch": 0.03991, + "grad_norm": 0.7206979857545632, + "learning_rate": 0.003, + "loss": 4.1859, + "step": 3991 + }, + { + "epoch": 0.03992, + "grad_norm": 0.8037268570780255, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 3992 + }, + { + "epoch": 0.03993, + "grad_norm": 0.7620905271249203, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 3993 + }, + { + "epoch": 0.03994, + "grad_norm": 0.738025311420094, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 3994 + }, + { + "epoch": 0.03995, + "grad_norm": 0.6172466009583261, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 3995 + }, + { + "epoch": 0.03996, + "grad_norm": 0.452402581761887, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 3996 + }, + { + "epoch": 0.03997, + "grad_norm": 0.48368404250480046, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 3997 + }, + { + "epoch": 0.03998, + "grad_norm": 0.47207135976156783, + "learning_rate": 0.003, + "loss": 4.17, + "step": 3998 + }, + { + "epoch": 0.03999, + "grad_norm": 0.502083976437197, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 3999 + }, + { + "epoch": 0.04, + "grad_norm": 0.4504866573185321, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 4000 + }, + { + "epoch": 0.04001, + "grad_norm": 0.37734509945831174, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4001 + }, + { + "epoch": 0.04002, + "grad_norm": 0.4339763243340729, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 4002 + }, + { + "epoch": 0.04003, + "grad_norm": 0.5726849542049914, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 4003 + }, + { + "epoch": 0.04004, + "grad_norm": 0.71841237008688, + "learning_rate": 0.003, + "loss": 4.167, + "step": 4004 + }, + { + "epoch": 0.04005, + "grad_norm": 0.9930905212295545, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 4005 + }, + { + "epoch": 0.04006, + "grad_norm": 0.9991984295961828, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4006 + }, + { + "epoch": 0.04007, + "grad_norm": 0.7547073099000725, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 4007 + }, + { + "epoch": 0.04008, + "grad_norm": 0.8454739420896483, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4008 + }, + { + "epoch": 0.04009, + "grad_norm": 0.9067851236139133, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 4009 + }, + { + "epoch": 0.0401, + "grad_norm": 0.8062732196382164, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 4010 + }, + { + "epoch": 0.04011, + "grad_norm": 0.6938007348622179, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 4011 + }, + { + "epoch": 0.04012, + "grad_norm": 0.7757979776927771, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 4012 + }, + { + "epoch": 0.04013, + "grad_norm": 0.7089158793355179, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 4013 + }, + { + "epoch": 0.04014, + "grad_norm": 0.6111952836402943, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 4014 + }, + { + "epoch": 0.04015, + "grad_norm": 0.5804891668774357, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4015 + }, + { + "epoch": 0.04016, + "grad_norm": 0.5604585117010126, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 4016 + }, + { + "epoch": 0.04017, + "grad_norm": 0.5830753494163682, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 4017 + }, + { + "epoch": 0.04018, + "grad_norm": 0.52884054520163, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 4018 + }, + { + "epoch": 0.04019, + "grad_norm": 0.5435284400458248, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 4019 + }, + { + "epoch": 0.0402, + "grad_norm": 0.46054232620433116, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4020 + }, + { + "epoch": 0.04021, + "grad_norm": 0.5140849815317029, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 4021 + }, + { + "epoch": 0.04022, + "grad_norm": 0.6311629918528258, + "learning_rate": 0.003, + "loss": 4.13, + "step": 4022 + }, + { + "epoch": 0.04023, + "grad_norm": 0.7137869361747575, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 4023 + }, + { + "epoch": 0.04024, + "grad_norm": 0.738420470053228, + "learning_rate": 0.003, + "loss": 4.124, + "step": 4024 + }, + { + "epoch": 0.04025, + "grad_norm": 0.6521094783928104, + "learning_rate": 0.003, + "loss": 4.13, + "step": 4025 + }, + { + "epoch": 0.04026, + "grad_norm": 0.5809109807556073, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 4026 + }, + { + "epoch": 0.04027, + "grad_norm": 0.6439840472927373, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 4027 + }, + { + "epoch": 0.04028, + "grad_norm": 0.6879545278017376, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 4028 + }, + { + "epoch": 0.04029, + "grad_norm": 0.7939138849392092, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 4029 + }, + { + "epoch": 0.0403, + "grad_norm": 0.7119398099333982, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 4030 + }, + { + "epoch": 0.04031, + "grad_norm": 0.6966756671526879, + "learning_rate": 0.003, + "loss": 4.139, + "step": 4031 + }, + { + "epoch": 0.04032, + "grad_norm": 0.6596032753240082, + "learning_rate": 0.003, + "loss": 4.1647, + "step": 4032 + }, + { + "epoch": 0.04033, + "grad_norm": 0.6274155855378644, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 4033 + }, + { + "epoch": 0.04034, + "grad_norm": 0.6095955743702651, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 4034 + }, + { + "epoch": 0.04035, + "grad_norm": 0.555350198090057, + "learning_rate": 0.003, + "loss": 4.148, + "step": 4035 + }, + { + "epoch": 0.04036, + "grad_norm": 0.5197769667331041, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 4036 + }, + { + "epoch": 0.04037, + "grad_norm": 0.5661546926366932, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 4037 + }, + { + "epoch": 0.04038, + "grad_norm": 0.7699720543941366, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 4038 + }, + { + "epoch": 0.04039, + "grad_norm": 1.210094587806928, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4039 + }, + { + "epoch": 0.0404, + "grad_norm": 0.868075390136909, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 4040 + }, + { + "epoch": 0.04041, + "grad_norm": 0.7416181269518949, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 4041 + }, + { + "epoch": 0.04042, + "grad_norm": 0.6648042947771692, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 4042 + }, + { + "epoch": 0.04043, + "grad_norm": 0.6609860437382262, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 4043 + }, + { + "epoch": 0.04044, + "grad_norm": 0.7121985413545872, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 4044 + }, + { + "epoch": 0.04045, + "grad_norm": 0.7259672418788464, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4045 + }, + { + "epoch": 0.04046, + "grad_norm": 0.6676245217285697, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 4046 + }, + { + "epoch": 0.04047, + "grad_norm": 0.6576773790552027, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 4047 + }, + { + "epoch": 0.04048, + "grad_norm": 0.7504199241468649, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 4048 + }, + { + "epoch": 0.04049, + "grad_norm": 0.6934067855311765, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 4049 + }, + { + "epoch": 0.0405, + "grad_norm": 0.6792487502012533, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 4050 + }, + { + "epoch": 0.04051, + "grad_norm": 0.6504046520169164, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 4051 + }, + { + "epoch": 0.04052, + "grad_norm": 0.5049122912594802, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 4052 + }, + { + "epoch": 0.04053, + "grad_norm": 0.5305812046390571, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4053 + }, + { + "epoch": 0.04054, + "grad_norm": 0.595657788891312, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 4054 + }, + { + "epoch": 0.04055, + "grad_norm": 0.8174417297511951, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 4055 + }, + { + "epoch": 0.04056, + "grad_norm": 1.2164744618417813, + "learning_rate": 0.003, + "loss": 4.166, + "step": 4056 + }, + { + "epoch": 0.04057, + "grad_norm": 0.7248345125453411, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 4057 + }, + { + "epoch": 0.04058, + "grad_norm": 0.6128740383918565, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 4058 + }, + { + "epoch": 0.04059, + "grad_norm": 0.7656551573931919, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 4059 + }, + { + "epoch": 0.0406, + "grad_norm": 0.7644654016064842, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 4060 + }, + { + "epoch": 0.04061, + "grad_norm": 0.7754515611223848, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4061 + }, + { + "epoch": 0.04062, + "grad_norm": 0.8608274358738601, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 4062 + }, + { + "epoch": 0.04063, + "grad_norm": 0.9067607362608292, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4063 + }, + { + "epoch": 0.04064, + "grad_norm": 1.0529705240815344, + "learning_rate": 0.003, + "loss": 4.1734, + "step": 4064 + }, + { + "epoch": 0.04065, + "grad_norm": 1.0102927624010265, + "learning_rate": 0.003, + "loss": 4.215, + "step": 4065 + }, + { + "epoch": 0.04066, + "grad_norm": 0.9478175395069827, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 4066 + }, + { + "epoch": 0.04067, + "grad_norm": 0.8098599255455219, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 4067 + }, + { + "epoch": 0.04068, + "grad_norm": 0.724209064256899, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 4068 + }, + { + "epoch": 0.04069, + "grad_norm": 0.6809092940169613, + "learning_rate": 0.003, + "loss": 4.168, + "step": 4069 + }, + { + "epoch": 0.0407, + "grad_norm": 0.8132916778007174, + "learning_rate": 0.003, + "loss": 4.174, + "step": 4070 + }, + { + "epoch": 0.04071, + "grad_norm": 1.041265005432495, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 4071 + }, + { + "epoch": 0.04072, + "grad_norm": 1.0457271292201864, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 4072 + }, + { + "epoch": 0.04073, + "grad_norm": 1.0651363452233251, + "learning_rate": 0.003, + "loss": 4.167, + "step": 4073 + }, + { + "epoch": 0.04074, + "grad_norm": 0.7177727168488233, + "learning_rate": 0.003, + "loss": 4.1844, + "step": 4074 + }, + { + "epoch": 0.04075, + "grad_norm": 0.6005805584962535, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 4075 + }, + { + "epoch": 0.04076, + "grad_norm": 0.5914694996561694, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 4076 + }, + { + "epoch": 0.04077, + "grad_norm": 0.6911502018139879, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 4077 + }, + { + "epoch": 0.04078, + "grad_norm": 0.7912939945570531, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 4078 + }, + { + "epoch": 0.04079, + "grad_norm": 0.8570836696530718, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 4079 + }, + { + "epoch": 0.0408, + "grad_norm": 0.7585471415380697, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 4080 + }, + { + "epoch": 0.04081, + "grad_norm": 0.7520683150191194, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4081 + }, + { + "epoch": 0.04082, + "grad_norm": 0.6537470491867149, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 4082 + }, + { + "epoch": 0.04083, + "grad_norm": 0.6723574711008242, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 4083 + }, + { + "epoch": 0.04084, + "grad_norm": 0.6506573739916557, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 4084 + }, + { + "epoch": 0.04085, + "grad_norm": 0.7215838325103069, + "learning_rate": 0.003, + "loss": 4.1658, + "step": 4085 + }, + { + "epoch": 0.04086, + "grad_norm": 0.7304080342645308, + "learning_rate": 0.003, + "loss": 4.1831, + "step": 4086 + }, + { + "epoch": 0.04087, + "grad_norm": 0.736672307736761, + "learning_rate": 0.003, + "loss": 4.164, + "step": 4087 + }, + { + "epoch": 0.04088, + "grad_norm": 0.7327696397562437, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4088 + }, + { + "epoch": 0.04089, + "grad_norm": 0.737838253705245, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 4089 + }, + { + "epoch": 0.0409, + "grad_norm": 0.6618217098522069, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 4090 + }, + { + "epoch": 0.04091, + "grad_norm": 0.6534380155154119, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 4091 + }, + { + "epoch": 0.04092, + "grad_norm": 0.6174715229139612, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 4092 + }, + { + "epoch": 0.04093, + "grad_norm": 0.6189167119848286, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 4093 + }, + { + "epoch": 0.04094, + "grad_norm": 0.6641703981374822, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 4094 + }, + { + "epoch": 0.04095, + "grad_norm": 0.6395123879631477, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 4095 + }, + { + "epoch": 0.04096, + "grad_norm": 0.6349994704357759, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4096 + }, + { + "epoch": 0.04097, + "grad_norm": 0.6463455807108017, + "learning_rate": 0.003, + "loss": 4.141, + "step": 4097 + }, + { + "epoch": 0.04098, + "grad_norm": 0.6932323673060099, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4098 + }, + { + "epoch": 0.04099, + "grad_norm": 0.8807834378700455, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 4099 + }, + { + "epoch": 0.041, + "grad_norm": 1.010325661821192, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 4100 + }, + { + "epoch": 0.04101, + "grad_norm": 0.8888939381607318, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4101 + }, + { + "epoch": 0.04102, + "grad_norm": 0.7433180771066499, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 4102 + }, + { + "epoch": 0.04103, + "grad_norm": 0.6923500471319527, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 4103 + }, + { + "epoch": 0.04104, + "grad_norm": 0.6523846830610261, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 4104 + }, + { + "epoch": 0.04105, + "grad_norm": 0.6124945853172781, + "learning_rate": 0.003, + "loss": 4.12, + "step": 4105 + }, + { + "epoch": 0.04106, + "grad_norm": 0.6180742314657086, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 4106 + }, + { + "epoch": 0.04107, + "grad_norm": 0.7675574080126237, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 4107 + }, + { + "epoch": 0.04108, + "grad_norm": 0.7972279438911031, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 4108 + }, + { + "epoch": 0.04109, + "grad_norm": 0.7338110856400634, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4109 + }, + { + "epoch": 0.0411, + "grad_norm": 0.5620305965378466, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4110 + }, + { + "epoch": 0.04111, + "grad_norm": 0.5981690023054039, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4111 + }, + { + "epoch": 0.04112, + "grad_norm": 0.5672857968992558, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 4112 + }, + { + "epoch": 0.04113, + "grad_norm": 0.6312145325378008, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 4113 + }, + { + "epoch": 0.04114, + "grad_norm": 0.6035265736834228, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4114 + }, + { + "epoch": 0.04115, + "grad_norm": 0.5962258124728763, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 4115 + }, + { + "epoch": 0.04116, + "grad_norm": 0.5290947291389287, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 4116 + }, + { + "epoch": 0.04117, + "grad_norm": 0.475360712634243, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 4117 + }, + { + "epoch": 0.04118, + "grad_norm": 0.48139980415634037, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 4118 + }, + { + "epoch": 0.04119, + "grad_norm": 0.4768773023811471, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4119 + }, + { + "epoch": 0.0412, + "grad_norm": 0.4715918252478233, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 4120 + }, + { + "epoch": 0.04121, + "grad_norm": 0.4631389077981398, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 4121 + }, + { + "epoch": 0.04122, + "grad_norm": 0.5695292166099397, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4122 + }, + { + "epoch": 0.04123, + "grad_norm": 0.6344096447219111, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 4123 + }, + { + "epoch": 0.04124, + "grad_norm": 0.7628425133873309, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 4124 + }, + { + "epoch": 0.04125, + "grad_norm": 1.0834853187583475, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 4125 + }, + { + "epoch": 0.04126, + "grad_norm": 1.108357219795881, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 4126 + }, + { + "epoch": 0.04127, + "grad_norm": 0.7334164794838763, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 4127 + }, + { + "epoch": 0.04128, + "grad_norm": 0.7315856737953997, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 4128 + }, + { + "epoch": 0.04129, + "grad_norm": 0.67750855822388, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 4129 + }, + { + "epoch": 0.0413, + "grad_norm": 0.6477790148509263, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4130 + }, + { + "epoch": 0.04131, + "grad_norm": 0.6976553932039135, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 4131 + }, + { + "epoch": 0.04132, + "grad_norm": 0.7575720136801286, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 4132 + }, + { + "epoch": 0.04133, + "grad_norm": 0.7834309464101223, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 4133 + }, + { + "epoch": 0.04134, + "grad_norm": 0.8769385664533508, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 4134 + }, + { + "epoch": 0.04135, + "grad_norm": 0.9288719850360465, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4135 + }, + { + "epoch": 0.04136, + "grad_norm": 0.8613590746245083, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 4136 + }, + { + "epoch": 0.04137, + "grad_norm": 0.7914311956529176, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 4137 + }, + { + "epoch": 0.04138, + "grad_norm": 0.8323094727140243, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 4138 + }, + { + "epoch": 0.04139, + "grad_norm": 0.7498292112429189, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 4139 + }, + { + "epoch": 0.0414, + "grad_norm": 0.7470547424534141, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 4140 + }, + { + "epoch": 0.04141, + "grad_norm": 0.7086395261007737, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 4141 + }, + { + "epoch": 0.04142, + "grad_norm": 0.7475027864042157, + "learning_rate": 0.003, + "loss": 4.166, + "step": 4142 + }, + { + "epoch": 0.04143, + "grad_norm": 0.7096941472876405, + "learning_rate": 0.003, + "loss": 4.151, + "step": 4143 + }, + { + "epoch": 0.04144, + "grad_norm": 0.6807921123663281, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 4144 + }, + { + "epoch": 0.04145, + "grad_norm": 0.6209916627342668, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 4145 + }, + { + "epoch": 0.04146, + "grad_norm": 0.6603935131450228, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4146 + }, + { + "epoch": 0.04147, + "grad_norm": 0.766665555184346, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 4147 + }, + { + "epoch": 0.04148, + "grad_norm": 0.8232977544301274, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 4148 + }, + { + "epoch": 0.04149, + "grad_norm": 0.9384903955246353, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4149 + }, + { + "epoch": 0.0415, + "grad_norm": 0.8526053994314199, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 4150 + }, + { + "epoch": 0.04151, + "grad_norm": 0.703764014227338, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 4151 + }, + { + "epoch": 0.04152, + "grad_norm": 0.5883825477373993, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4152 + }, + { + "epoch": 0.04153, + "grad_norm": 0.5178464710197032, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 4153 + }, + { + "epoch": 0.04154, + "grad_norm": 0.4801954114695538, + "learning_rate": 0.003, + "loss": 4.152, + "step": 4154 + }, + { + "epoch": 0.04155, + "grad_norm": 0.4883851788041572, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4155 + }, + { + "epoch": 0.04156, + "grad_norm": 0.5305154970213135, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4156 + }, + { + "epoch": 0.04157, + "grad_norm": 0.5969580348221766, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4157 + }, + { + "epoch": 0.04158, + "grad_norm": 0.604419085621543, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 4158 + }, + { + "epoch": 0.04159, + "grad_norm": 0.5952666616236362, + "learning_rate": 0.003, + "loss": 4.136, + "step": 4159 + }, + { + "epoch": 0.0416, + "grad_norm": 0.6015748305688609, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4160 + }, + { + "epoch": 0.04161, + "grad_norm": 0.6509334448144796, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 4161 + }, + { + "epoch": 0.04162, + "grad_norm": 0.8275491529133575, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 4162 + }, + { + "epoch": 0.04163, + "grad_norm": 0.9190060913193079, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 4163 + }, + { + "epoch": 0.04164, + "grad_norm": 1.0804296755917129, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 4164 + }, + { + "epoch": 0.04165, + "grad_norm": 0.8918375998135468, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4165 + }, + { + "epoch": 0.04166, + "grad_norm": 0.6444987754007603, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4166 + }, + { + "epoch": 0.04167, + "grad_norm": 0.5987188953434583, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 4167 + }, + { + "epoch": 0.04168, + "grad_norm": 0.6931415065499815, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 4168 + }, + { + "epoch": 0.04169, + "grad_norm": 0.7939284298013517, + "learning_rate": 0.003, + "loss": 4.1706, + "step": 4169 + }, + { + "epoch": 0.0417, + "grad_norm": 0.8929923791952147, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 4170 + }, + { + "epoch": 0.04171, + "grad_norm": 0.8713733021965931, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 4171 + }, + { + "epoch": 0.04172, + "grad_norm": 0.741654776741016, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 4172 + }, + { + "epoch": 0.04173, + "grad_norm": 0.6869389765405233, + "learning_rate": 0.003, + "loss": 4.169, + "step": 4173 + }, + { + "epoch": 0.04174, + "grad_norm": 0.7104276340717113, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 4174 + }, + { + "epoch": 0.04175, + "grad_norm": 0.7652108394533633, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 4175 + }, + { + "epoch": 0.04176, + "grad_norm": 0.6595942295349072, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 4176 + }, + { + "epoch": 0.04177, + "grad_norm": 0.6050614217333806, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4177 + }, + { + "epoch": 0.04178, + "grad_norm": 0.6649317208659106, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 4178 + }, + { + "epoch": 0.04179, + "grad_norm": 0.7497130033497713, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4179 + }, + { + "epoch": 0.0418, + "grad_norm": 0.8135721504617128, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4180 + }, + { + "epoch": 0.04181, + "grad_norm": 0.8382943348914027, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4181 + }, + { + "epoch": 0.04182, + "grad_norm": 0.7581608595797077, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 4182 + }, + { + "epoch": 0.04183, + "grad_norm": 0.7367386623848073, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4183 + }, + { + "epoch": 0.04184, + "grad_norm": 0.8223299343686243, + "learning_rate": 0.003, + "loss": 4.2008, + "step": 4184 + }, + { + "epoch": 0.04185, + "grad_norm": 0.8253315995103551, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 4185 + }, + { + "epoch": 0.04186, + "grad_norm": 1.021967899357659, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 4186 + }, + { + "epoch": 0.04187, + "grad_norm": 1.1194144390778462, + "learning_rate": 0.003, + "loss": 4.196, + "step": 4187 + }, + { + "epoch": 0.04188, + "grad_norm": 0.8663014575987595, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4188 + }, + { + "epoch": 0.04189, + "grad_norm": 0.7730217042128824, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4189 + }, + { + "epoch": 0.0419, + "grad_norm": 0.718425867412397, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 4190 + }, + { + "epoch": 0.04191, + "grad_norm": 0.634536414286615, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 4191 + }, + { + "epoch": 0.04192, + "grad_norm": 0.6074939161177713, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 4192 + }, + { + "epoch": 0.04193, + "grad_norm": 0.5595906619172795, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 4193 + }, + { + "epoch": 0.04194, + "grad_norm": 0.591321468676565, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4194 + }, + { + "epoch": 0.04195, + "grad_norm": 0.6331925592194563, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 4195 + }, + { + "epoch": 0.04196, + "grad_norm": 0.66777685730357, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 4196 + }, + { + "epoch": 0.04197, + "grad_norm": 0.6581215996860487, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 4197 + }, + { + "epoch": 0.04198, + "grad_norm": 0.5775968736275588, + "learning_rate": 0.003, + "loss": 4.128, + "step": 4198 + }, + { + "epoch": 0.04199, + "grad_norm": 0.5115173192319249, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4199 + }, + { + "epoch": 0.042, + "grad_norm": 0.4907326781591073, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 4200 + }, + { + "epoch": 0.04201, + "grad_norm": 0.4340051119862648, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4201 + }, + { + "epoch": 0.04202, + "grad_norm": 0.4270978633731861, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 4202 + }, + { + "epoch": 0.04203, + "grad_norm": 0.5415535220350182, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 4203 + }, + { + "epoch": 0.04204, + "grad_norm": 0.6653136591047163, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 4204 + }, + { + "epoch": 0.04205, + "grad_norm": 0.9015195834615578, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 4205 + }, + { + "epoch": 0.04206, + "grad_norm": 0.9511163331115945, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 4206 + }, + { + "epoch": 0.04207, + "grad_norm": 0.8183175873095769, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 4207 + }, + { + "epoch": 0.04208, + "grad_norm": 0.7983775516444301, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4208 + }, + { + "epoch": 0.04209, + "grad_norm": 0.9111818683711085, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 4209 + }, + { + "epoch": 0.0421, + "grad_norm": 0.8760572861926779, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 4210 + }, + { + "epoch": 0.04211, + "grad_norm": 0.8385383847029609, + "learning_rate": 0.003, + "loss": 4.158, + "step": 4211 + }, + { + "epoch": 0.04212, + "grad_norm": 0.8523572548315141, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 4212 + }, + { + "epoch": 0.04213, + "grad_norm": 0.87521517308783, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4213 + }, + { + "epoch": 0.04214, + "grad_norm": 0.8765783386936129, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 4214 + }, + { + "epoch": 0.04215, + "grad_norm": 0.797769330073817, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 4215 + }, + { + "epoch": 0.04216, + "grad_norm": 0.7247822776339875, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4216 + }, + { + "epoch": 0.04217, + "grad_norm": 0.7261327052126046, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 4217 + }, + { + "epoch": 0.04218, + "grad_norm": 0.7319178674334204, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 4218 + }, + { + "epoch": 0.04219, + "grad_norm": 0.7694832016892729, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4219 + }, + { + "epoch": 0.0422, + "grad_norm": 0.757141250932244, + "learning_rate": 0.003, + "loss": 4.17, + "step": 4220 + }, + { + "epoch": 0.04221, + "grad_norm": 0.7290165220800948, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4221 + }, + { + "epoch": 0.04222, + "grad_norm": 0.6346702624845955, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 4222 + }, + { + "epoch": 0.04223, + "grad_norm": 0.583770301965496, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 4223 + }, + { + "epoch": 0.04224, + "grad_norm": 0.5350148270680584, + "learning_rate": 0.003, + "loss": 4.17, + "step": 4224 + }, + { + "epoch": 0.04225, + "grad_norm": 0.5316639979135299, + "learning_rate": 0.003, + "loss": 4.154, + "step": 4225 + }, + { + "epoch": 0.04226, + "grad_norm": 0.5763902675717133, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 4226 + }, + { + "epoch": 0.04227, + "grad_norm": 0.6468360436042053, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4227 + }, + { + "epoch": 0.04228, + "grad_norm": 0.7289273143868911, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4228 + }, + { + "epoch": 0.04229, + "grad_norm": 0.759894491275707, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 4229 + }, + { + "epoch": 0.0423, + "grad_norm": 0.7516690067272278, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 4230 + }, + { + "epoch": 0.04231, + "grad_norm": 0.6070127843457722, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 4231 + }, + { + "epoch": 0.04232, + "grad_norm": 0.6987145523188838, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4232 + }, + { + "epoch": 0.04233, + "grad_norm": 0.9287852741458625, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 4233 + }, + { + "epoch": 0.04234, + "grad_norm": 0.9918383708905326, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 4234 + }, + { + "epoch": 0.04235, + "grad_norm": 0.9367701682265075, + "learning_rate": 0.003, + "loss": 4.134, + "step": 4235 + }, + { + "epoch": 0.04236, + "grad_norm": 0.8703490177121154, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 4236 + }, + { + "epoch": 0.04237, + "grad_norm": 0.8156234434712465, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4237 + }, + { + "epoch": 0.04238, + "grad_norm": 0.6835933384976534, + "learning_rate": 0.003, + "loss": 4.175, + "step": 4238 + }, + { + "epoch": 0.04239, + "grad_norm": 0.6391989756025438, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4239 + }, + { + "epoch": 0.0424, + "grad_norm": 0.6447631067586589, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 4240 + }, + { + "epoch": 0.04241, + "grad_norm": 0.6715155999130873, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 4241 + }, + { + "epoch": 0.04242, + "grad_norm": 0.6273748467552226, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4242 + }, + { + "epoch": 0.04243, + "grad_norm": 0.6467254126434867, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 4243 + }, + { + "epoch": 0.04244, + "grad_norm": 0.6955126933633631, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 4244 + }, + { + "epoch": 0.04245, + "grad_norm": 0.681078617510812, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 4245 + }, + { + "epoch": 0.04246, + "grad_norm": 0.69651922194907, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 4246 + }, + { + "epoch": 0.04247, + "grad_norm": 0.7659042143848519, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4247 + }, + { + "epoch": 0.04248, + "grad_norm": 0.8045443211607587, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 4248 + }, + { + "epoch": 0.04249, + "grad_norm": 0.7807725187988103, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 4249 + }, + { + "epoch": 0.0425, + "grad_norm": 0.797629821833922, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 4250 + }, + { + "epoch": 0.04251, + "grad_norm": 0.8773737541427695, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 4251 + }, + { + "epoch": 0.04252, + "grad_norm": 0.8764481339534494, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 4252 + }, + { + "epoch": 0.04253, + "grad_norm": 0.771295115347259, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4253 + }, + { + "epoch": 0.04254, + "grad_norm": 0.6812529052304258, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4254 + }, + { + "epoch": 0.04255, + "grad_norm": 0.6528883858102312, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 4255 + }, + { + "epoch": 0.04256, + "grad_norm": 0.6360597468005796, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 4256 + }, + { + "epoch": 0.04257, + "grad_norm": 0.6045194570542003, + "learning_rate": 0.003, + "loss": 4.152, + "step": 4257 + }, + { + "epoch": 0.04258, + "grad_norm": 0.6237209269525407, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 4258 + }, + { + "epoch": 0.04259, + "grad_norm": 0.6534065774304211, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4259 + }, + { + "epoch": 0.0426, + "grad_norm": 0.7228160196304746, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4260 + }, + { + "epoch": 0.04261, + "grad_norm": 0.7225504963671482, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 4261 + }, + { + "epoch": 0.04262, + "grad_norm": 0.6030677337778022, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 4262 + }, + { + "epoch": 0.04263, + "grad_norm": 0.5576016271704423, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 4263 + }, + { + "epoch": 0.04264, + "grad_norm": 0.4622794122851678, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4264 + }, + { + "epoch": 0.04265, + "grad_norm": 0.4498062625739869, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 4265 + }, + { + "epoch": 0.04266, + "grad_norm": 0.4663874509947672, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 4266 + }, + { + "epoch": 0.04267, + "grad_norm": 0.5819381640130065, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4267 + }, + { + "epoch": 0.04268, + "grad_norm": 0.7240543464190408, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 4268 + }, + { + "epoch": 0.04269, + "grad_norm": 0.7869636361035491, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4269 + }, + { + "epoch": 0.0427, + "grad_norm": 0.7316564623442252, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 4270 + }, + { + "epoch": 0.04271, + "grad_norm": 0.6962340655904857, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 4271 + }, + { + "epoch": 0.04272, + "grad_norm": 0.7468024354946826, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 4272 + }, + { + "epoch": 0.04273, + "grad_norm": 0.8315685932461937, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 4273 + }, + { + "epoch": 0.04274, + "grad_norm": 0.7971355124213763, + "learning_rate": 0.003, + "loss": 4.144, + "step": 4274 + }, + { + "epoch": 0.04275, + "grad_norm": 0.7711074822549306, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4275 + }, + { + "epoch": 0.04276, + "grad_norm": 0.8849517437250696, + "learning_rate": 0.003, + "loss": 4.176, + "step": 4276 + }, + { + "epoch": 0.04277, + "grad_norm": 1.058466724741216, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 4277 + }, + { + "epoch": 0.04278, + "grad_norm": 0.9806877039749324, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 4278 + }, + { + "epoch": 0.04279, + "grad_norm": 1.1506116450442327, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4279 + }, + { + "epoch": 0.0428, + "grad_norm": 1.0111616301127737, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 4280 + }, + { + "epoch": 0.04281, + "grad_norm": 1.1240409097888207, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 4281 + }, + { + "epoch": 0.04282, + "grad_norm": 0.9707964829866726, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 4282 + }, + { + "epoch": 0.04283, + "grad_norm": 0.8952580683408602, + "learning_rate": 0.003, + "loss": 4.2067, + "step": 4283 + }, + { + "epoch": 0.04284, + "grad_norm": 0.8037590999213429, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4284 + }, + { + "epoch": 0.04285, + "grad_norm": 0.7180846206853434, + "learning_rate": 0.003, + "loss": 4.1689, + "step": 4285 + }, + { + "epoch": 0.04286, + "grad_norm": 0.6600031955008437, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 4286 + }, + { + "epoch": 0.04287, + "grad_norm": 0.624643362166525, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 4287 + }, + { + "epoch": 0.04288, + "grad_norm": 0.7314395975343404, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4288 + }, + { + "epoch": 0.04289, + "grad_norm": 0.7552436651554467, + "learning_rate": 0.003, + "loss": 4.149, + "step": 4289 + }, + { + "epoch": 0.0429, + "grad_norm": 0.7444307165904086, + "learning_rate": 0.003, + "loss": 4.136, + "step": 4290 + }, + { + "epoch": 0.04291, + "grad_norm": 0.6792525045452005, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 4291 + }, + { + "epoch": 0.04292, + "grad_norm": 0.6366634706551982, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 4292 + }, + { + "epoch": 0.04293, + "grad_norm": 0.5177321350541024, + "learning_rate": 0.003, + "loss": 4.1696, + "step": 4293 + }, + { + "epoch": 0.04294, + "grad_norm": 0.5199647077622895, + "learning_rate": 0.003, + "loss": 4.1474, + "step": 4294 + }, + { + "epoch": 0.04295, + "grad_norm": 0.533363434433449, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 4295 + }, + { + "epoch": 0.04296, + "grad_norm": 0.4910620913342609, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 4296 + }, + { + "epoch": 0.04297, + "grad_norm": 0.4528594955305011, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 4297 + }, + { + "epoch": 0.04298, + "grad_norm": 0.5508851288276945, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4298 + }, + { + "epoch": 0.04299, + "grad_norm": 0.6926359286804635, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4299 + }, + { + "epoch": 0.043, + "grad_norm": 0.776299147583224, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4300 + }, + { + "epoch": 0.04301, + "grad_norm": 0.7543517253530487, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4301 + }, + { + "epoch": 0.04302, + "grad_norm": 0.7190955364079332, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4302 + }, + { + "epoch": 0.04303, + "grad_norm": 0.7273212030383343, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 4303 + }, + { + "epoch": 0.04304, + "grad_norm": 0.8042177427504058, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4304 + }, + { + "epoch": 0.04305, + "grad_norm": 0.889845680686584, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 4305 + }, + { + "epoch": 0.04306, + "grad_norm": 0.8019734206580067, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 4306 + }, + { + "epoch": 0.04307, + "grad_norm": 0.7599014266097025, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 4307 + }, + { + "epoch": 0.04308, + "grad_norm": 0.6928021109826253, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4308 + }, + { + "epoch": 0.04309, + "grad_norm": 0.6623360644735573, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4309 + }, + { + "epoch": 0.0431, + "grad_norm": 0.643977742791814, + "learning_rate": 0.003, + "loss": 4.1542, + "step": 4310 + }, + { + "epoch": 0.04311, + "grad_norm": 0.6752224622600257, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4311 + }, + { + "epoch": 0.04312, + "grad_norm": 0.685683319654323, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 4312 + }, + { + "epoch": 0.04313, + "grad_norm": 0.6828332575779398, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 4313 + }, + { + "epoch": 0.04314, + "grad_norm": 0.7207034799508291, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 4314 + }, + { + "epoch": 0.04315, + "grad_norm": 0.6920628089309946, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 4315 + }, + { + "epoch": 0.04316, + "grad_norm": 0.6697894292684422, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4316 + }, + { + "epoch": 0.04317, + "grad_norm": 0.6529885706954907, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4317 + }, + { + "epoch": 0.04318, + "grad_norm": 0.6780113889358966, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 4318 + }, + { + "epoch": 0.04319, + "grad_norm": 0.6696493184651962, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4319 + }, + { + "epoch": 0.0432, + "grad_norm": 0.6379331339678891, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 4320 + }, + { + "epoch": 0.04321, + "grad_norm": 0.713977041740171, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4321 + }, + { + "epoch": 0.04322, + "grad_norm": 0.9061136199776986, + "learning_rate": 0.003, + "loss": 4.1716, + "step": 4322 + }, + { + "epoch": 0.04323, + "grad_norm": 1.0087052417142888, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4323 + }, + { + "epoch": 0.04324, + "grad_norm": 0.9066972139100388, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 4324 + }, + { + "epoch": 0.04325, + "grad_norm": 0.829210403933647, + "learning_rate": 0.003, + "loss": 4.1709, + "step": 4325 + }, + { + "epoch": 0.04326, + "grad_norm": 0.766217873896832, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 4326 + }, + { + "epoch": 0.04327, + "grad_norm": 0.66336507203185, + "learning_rate": 0.003, + "loss": 4.125, + "step": 4327 + }, + { + "epoch": 0.04328, + "grad_norm": 0.6687205522581232, + "learning_rate": 0.003, + "loss": 4.143, + "step": 4328 + }, + { + "epoch": 0.04329, + "grad_norm": 0.6982687394228022, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 4329 + }, + { + "epoch": 0.0433, + "grad_norm": 0.8226633234518121, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 4330 + }, + { + "epoch": 0.04331, + "grad_norm": 0.7328602014834829, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 4331 + }, + { + "epoch": 0.04332, + "grad_norm": 0.7041041791657667, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 4332 + }, + { + "epoch": 0.04333, + "grad_norm": 0.7435632839058932, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 4333 + }, + { + "epoch": 0.04334, + "grad_norm": 0.8773446956801049, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 4334 + }, + { + "epoch": 0.04335, + "grad_norm": 0.8446265063365133, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 4335 + }, + { + "epoch": 0.04336, + "grad_norm": 0.7068555070364958, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 4336 + }, + { + "epoch": 0.04337, + "grad_norm": 0.7529497673494582, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4337 + }, + { + "epoch": 0.04338, + "grad_norm": 0.8822495743753999, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 4338 + }, + { + "epoch": 0.04339, + "grad_norm": 0.8362014809442191, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4339 + }, + { + "epoch": 0.0434, + "grad_norm": 0.8580345532471818, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 4340 + }, + { + "epoch": 0.04341, + "grad_norm": 0.8559919642238094, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4341 + }, + { + "epoch": 0.04342, + "grad_norm": 0.9879150305605277, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 4342 + }, + { + "epoch": 0.04343, + "grad_norm": 1.0224948991525673, + "learning_rate": 0.003, + "loss": 4.1953, + "step": 4343 + }, + { + "epoch": 0.04344, + "grad_norm": 0.8797661268558509, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 4344 + }, + { + "epoch": 0.04345, + "grad_norm": 0.7071974800056289, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 4345 + }, + { + "epoch": 0.04346, + "grad_norm": 0.7231453989893986, + "learning_rate": 0.003, + "loss": 4.166, + "step": 4346 + }, + { + "epoch": 0.04347, + "grad_norm": 0.7413522380132354, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 4347 + }, + { + "epoch": 0.04348, + "grad_norm": 0.861089019789159, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4348 + }, + { + "epoch": 0.04349, + "grad_norm": 0.8929649634389639, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 4349 + }, + { + "epoch": 0.0435, + "grad_norm": 0.8422594079232042, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 4350 + }, + { + "epoch": 0.04351, + "grad_norm": 0.6766722178647709, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4351 + }, + { + "epoch": 0.04352, + "grad_norm": 0.6065148592107631, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 4352 + }, + { + "epoch": 0.04353, + "grad_norm": 0.650665279490602, + "learning_rate": 0.003, + "loss": 4.139, + "step": 4353 + }, + { + "epoch": 0.04354, + "grad_norm": 0.6886947880787938, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 4354 + }, + { + "epoch": 0.04355, + "grad_norm": 0.671956570402557, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 4355 + }, + { + "epoch": 0.04356, + "grad_norm": 0.6051867008044065, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 4356 + }, + { + "epoch": 0.04357, + "grad_norm": 0.565831993323353, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 4357 + }, + { + "epoch": 0.04358, + "grad_norm": 0.6223890122466902, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 4358 + }, + { + "epoch": 0.04359, + "grad_norm": 0.5817182346909322, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 4359 + }, + { + "epoch": 0.0436, + "grad_norm": 0.5416327077206258, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 4360 + }, + { + "epoch": 0.04361, + "grad_norm": 0.5828369400063332, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4361 + }, + { + "epoch": 0.04362, + "grad_norm": 0.6659513919032584, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 4362 + }, + { + "epoch": 0.04363, + "grad_norm": 0.6690881007982171, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4363 + }, + { + "epoch": 0.04364, + "grad_norm": 0.6338233307172799, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 4364 + }, + { + "epoch": 0.04365, + "grad_norm": 0.6477441011384779, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 4365 + }, + { + "epoch": 0.04366, + "grad_norm": 0.7001290336621013, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 4366 + }, + { + "epoch": 0.04367, + "grad_norm": 0.7750910475232337, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 4367 + }, + { + "epoch": 0.04368, + "grad_norm": 0.653538990880571, + "learning_rate": 0.003, + "loss": 4.135, + "step": 4368 + }, + { + "epoch": 0.04369, + "grad_norm": 0.5454509649351917, + "learning_rate": 0.003, + "loss": 4.124, + "step": 4369 + }, + { + "epoch": 0.0437, + "grad_norm": 0.587005709751278, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 4370 + }, + { + "epoch": 0.04371, + "grad_norm": 0.6982746587986959, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 4371 + }, + { + "epoch": 0.04372, + "grad_norm": 0.7399237590808896, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 4372 + }, + { + "epoch": 0.04373, + "grad_norm": 0.7016341160191462, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 4373 + }, + { + "epoch": 0.04374, + "grad_norm": 0.7430013647684137, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 4374 + }, + { + "epoch": 0.04375, + "grad_norm": 0.7446061707027156, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 4375 + }, + { + "epoch": 0.04376, + "grad_norm": 0.7746810507611076, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 4376 + }, + { + "epoch": 0.04377, + "grad_norm": 0.9323940396007363, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 4377 + }, + { + "epoch": 0.04378, + "grad_norm": 1.139045470645915, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 4378 + }, + { + "epoch": 0.04379, + "grad_norm": 0.9525647154578588, + "learning_rate": 0.003, + "loss": 4.1848, + "step": 4379 + }, + { + "epoch": 0.0438, + "grad_norm": 0.8243260327364462, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 4380 + }, + { + "epoch": 0.04381, + "grad_norm": 0.6831625467049471, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 4381 + }, + { + "epoch": 0.04382, + "grad_norm": 0.612690019661821, + "learning_rate": 0.003, + "loss": 4.148, + "step": 4382 + }, + { + "epoch": 0.04383, + "grad_norm": 0.540088011425151, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 4383 + }, + { + "epoch": 0.04384, + "grad_norm": 0.48067565504109466, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4384 + }, + { + "epoch": 0.04385, + "grad_norm": 0.4940838283566669, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 4385 + }, + { + "epoch": 0.04386, + "grad_norm": 0.47625894049094, + "learning_rate": 0.003, + "loss": 4.134, + "step": 4386 + }, + { + "epoch": 0.04387, + "grad_norm": 0.48055396677108986, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 4387 + }, + { + "epoch": 0.04388, + "grad_norm": 0.4652346965489437, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 4388 + }, + { + "epoch": 0.04389, + "grad_norm": 0.6054008250326102, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 4389 + }, + { + "epoch": 0.0439, + "grad_norm": 0.8224358615264853, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 4390 + }, + { + "epoch": 0.04391, + "grad_norm": 0.9124868764503055, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4391 + }, + { + "epoch": 0.04392, + "grad_norm": 0.9806682632795687, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 4392 + }, + { + "epoch": 0.04393, + "grad_norm": 0.9973385060246203, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 4393 + }, + { + "epoch": 0.04394, + "grad_norm": 0.9949589217331475, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 4394 + }, + { + "epoch": 0.04395, + "grad_norm": 0.9709607301881599, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 4395 + }, + { + "epoch": 0.04396, + "grad_norm": 0.9645527804872946, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4396 + }, + { + "epoch": 0.04397, + "grad_norm": 0.9816350655638709, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4397 + }, + { + "epoch": 0.04398, + "grad_norm": 1.0455773944250841, + "learning_rate": 0.003, + "loss": 4.1626, + "step": 4398 + }, + { + "epoch": 0.04399, + "grad_norm": 0.9119841908639416, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 4399 + }, + { + "epoch": 0.044, + "grad_norm": 0.9349641099143394, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 4400 + }, + { + "epoch": 0.04401, + "grad_norm": 1.012943202195574, + "learning_rate": 0.003, + "loss": 4.1876, + "step": 4401 + }, + { + "epoch": 0.04402, + "grad_norm": 0.9301815653214077, + "learning_rate": 0.003, + "loss": 4.1887, + "step": 4402 + }, + { + "epoch": 0.04403, + "grad_norm": 0.8751459284087095, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4403 + }, + { + "epoch": 0.04404, + "grad_norm": 0.8722167037599222, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 4404 + }, + { + "epoch": 0.04405, + "grad_norm": 0.8424088244833199, + "learning_rate": 0.003, + "loss": 4.176, + "step": 4405 + }, + { + "epoch": 0.04406, + "grad_norm": 0.8029456116125556, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 4406 + }, + { + "epoch": 0.04407, + "grad_norm": 0.7600665803647753, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 4407 + }, + { + "epoch": 0.04408, + "grad_norm": 0.7143442872514836, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 4408 + }, + { + "epoch": 0.04409, + "grad_norm": 0.7965192349032346, + "learning_rate": 0.003, + "loss": 4.178, + "step": 4409 + }, + { + "epoch": 0.0441, + "grad_norm": 0.8792702370392786, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 4410 + }, + { + "epoch": 0.04411, + "grad_norm": 1.0243973709917655, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4411 + }, + { + "epoch": 0.04412, + "grad_norm": 0.8145959259792708, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4412 + }, + { + "epoch": 0.04413, + "grad_norm": 0.57578344484875, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4413 + }, + { + "epoch": 0.04414, + "grad_norm": 0.5763746372907456, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 4414 + }, + { + "epoch": 0.04415, + "grad_norm": 0.4849406384890711, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 4415 + }, + { + "epoch": 0.04416, + "grad_norm": 0.5645793903941705, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 4416 + }, + { + "epoch": 0.04417, + "grad_norm": 0.561088827608562, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4417 + }, + { + "epoch": 0.04418, + "grad_norm": 0.5718569607650594, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 4418 + }, + { + "epoch": 0.04419, + "grad_norm": 0.502350939750905, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4419 + }, + { + "epoch": 0.0442, + "grad_norm": 0.45571038809991676, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 4420 + }, + { + "epoch": 0.04421, + "grad_norm": 0.42375826081335205, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 4421 + }, + { + "epoch": 0.04422, + "grad_norm": 0.42008823401459217, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 4422 + }, + { + "epoch": 0.04423, + "grad_norm": 0.46008487233587403, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 4423 + }, + { + "epoch": 0.04424, + "grad_norm": 0.5469701885775844, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 4424 + }, + { + "epoch": 0.04425, + "grad_norm": 0.6192720902784608, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 4425 + }, + { + "epoch": 0.04426, + "grad_norm": 0.6752662810050423, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 4426 + }, + { + "epoch": 0.04427, + "grad_norm": 0.8141717149045535, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4427 + }, + { + "epoch": 0.04428, + "grad_norm": 1.1138698669039535, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 4428 + }, + { + "epoch": 0.04429, + "grad_norm": 1.1125457214220589, + "learning_rate": 0.003, + "loss": 4.1893, + "step": 4429 + }, + { + "epoch": 0.0443, + "grad_norm": 0.7372479542562949, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4430 + }, + { + "epoch": 0.04431, + "grad_norm": 0.7590662782504862, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 4431 + }, + { + "epoch": 0.04432, + "grad_norm": 0.7851056797042649, + "learning_rate": 0.003, + "loss": 4.169, + "step": 4432 + }, + { + "epoch": 0.04433, + "grad_norm": 0.6565615710180471, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 4433 + }, + { + "epoch": 0.04434, + "grad_norm": 0.5918951930492853, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 4434 + }, + { + "epoch": 0.04435, + "grad_norm": 0.5712466192614544, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 4435 + }, + { + "epoch": 0.04436, + "grad_norm": 0.6540737676285047, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 4436 + }, + { + "epoch": 0.04437, + "grad_norm": 0.6512769646169722, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 4437 + }, + { + "epoch": 0.04438, + "grad_norm": 0.6313619524940465, + "learning_rate": 0.003, + "loss": 4.181, + "step": 4438 + }, + { + "epoch": 0.04439, + "grad_norm": 0.752951040748533, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 4439 + }, + { + "epoch": 0.0444, + "grad_norm": 0.7444628766487162, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 4440 + }, + { + "epoch": 0.04441, + "grad_norm": 0.7171166547645912, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 4441 + }, + { + "epoch": 0.04442, + "grad_norm": 0.7004680425501666, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 4442 + }, + { + "epoch": 0.04443, + "grad_norm": 0.699693309924618, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4443 + }, + { + "epoch": 0.04444, + "grad_norm": 0.605352884343741, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4444 + }, + { + "epoch": 0.04445, + "grad_norm": 0.6910385491365711, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 4445 + }, + { + "epoch": 0.04446, + "grad_norm": 0.6750353497967526, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4446 + }, + { + "epoch": 0.04447, + "grad_norm": 0.5704187456661106, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 4447 + }, + { + "epoch": 0.04448, + "grad_norm": 0.5831627691900045, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 4448 + }, + { + "epoch": 0.04449, + "grad_norm": 0.7383838513937211, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4449 + }, + { + "epoch": 0.0445, + "grad_norm": 0.9275667982571733, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4450 + }, + { + "epoch": 0.04451, + "grad_norm": 1.1278443077545208, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 4451 + }, + { + "epoch": 0.04452, + "grad_norm": 0.7200631704178471, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 4452 + }, + { + "epoch": 0.04453, + "grad_norm": 0.7319095347479152, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 4453 + }, + { + "epoch": 0.04454, + "grad_norm": 0.828360672866726, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4454 + }, + { + "epoch": 0.04455, + "grad_norm": 0.8149109397250537, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 4455 + }, + { + "epoch": 0.04456, + "grad_norm": 0.7949979530915787, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 4456 + }, + { + "epoch": 0.04457, + "grad_norm": 0.7815335465443298, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4457 + }, + { + "epoch": 0.04458, + "grad_norm": 0.6667090111498273, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4458 + }, + { + "epoch": 0.04459, + "grad_norm": 0.6478410064314194, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 4459 + }, + { + "epoch": 0.0446, + "grad_norm": 0.692693711021832, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 4460 + }, + { + "epoch": 0.04461, + "grad_norm": 0.7661349867430685, + "learning_rate": 0.003, + "loss": 4.1671, + "step": 4461 + }, + { + "epoch": 0.04462, + "grad_norm": 0.9456365590075492, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 4462 + }, + { + "epoch": 0.04463, + "grad_norm": 0.9219959129201606, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4463 + }, + { + "epoch": 0.04464, + "grad_norm": 0.8148900268740181, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 4464 + }, + { + "epoch": 0.04465, + "grad_norm": 0.8174238600511685, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 4465 + }, + { + "epoch": 0.04466, + "grad_norm": 0.6910175481107252, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4466 + }, + { + "epoch": 0.04467, + "grad_norm": 0.8503966818321036, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 4467 + }, + { + "epoch": 0.04468, + "grad_norm": 0.8010356993406599, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4468 + }, + { + "epoch": 0.04469, + "grad_norm": 0.7388705798425241, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 4469 + }, + { + "epoch": 0.0447, + "grad_norm": 0.776572618481807, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 4470 + }, + { + "epoch": 0.04471, + "grad_norm": 0.8915085141537737, + "learning_rate": 0.003, + "loss": 4.185, + "step": 4471 + }, + { + "epoch": 0.04472, + "grad_norm": 0.9200136447240024, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 4472 + }, + { + "epoch": 0.04473, + "grad_norm": 0.8156139485901094, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 4473 + }, + { + "epoch": 0.04474, + "grad_norm": 0.7239763646047743, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4474 + }, + { + "epoch": 0.04475, + "grad_norm": 0.6835215989709738, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 4475 + }, + { + "epoch": 0.04476, + "grad_norm": 0.6497813876295598, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 4476 + }, + { + "epoch": 0.04477, + "grad_norm": 0.647414838457496, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 4477 + }, + { + "epoch": 0.04478, + "grad_norm": 0.559137538294316, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 4478 + }, + { + "epoch": 0.04479, + "grad_norm": 0.5859013586768095, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 4479 + }, + { + "epoch": 0.0448, + "grad_norm": 0.5435317461744096, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4480 + }, + { + "epoch": 0.04481, + "grad_norm": 0.6011946337549432, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 4481 + }, + { + "epoch": 0.04482, + "grad_norm": 0.7806327420943877, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 4482 + }, + { + "epoch": 0.04483, + "grad_norm": 1.0587276903826628, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4483 + }, + { + "epoch": 0.04484, + "grad_norm": 0.9760427349795051, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 4484 + }, + { + "epoch": 0.04485, + "grad_norm": 0.7488828747485865, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 4485 + }, + { + "epoch": 0.04486, + "grad_norm": 0.8084521903070014, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 4486 + }, + { + "epoch": 0.04487, + "grad_norm": 0.9117696333780938, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 4487 + }, + { + "epoch": 0.04488, + "grad_norm": 0.8996969646129346, + "learning_rate": 0.003, + "loss": 4.163, + "step": 4488 + }, + { + "epoch": 0.04489, + "grad_norm": 0.8641616901342983, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4489 + }, + { + "epoch": 0.0449, + "grad_norm": 0.7280388583966139, + "learning_rate": 0.003, + "loss": 4.1572, + "step": 4490 + }, + { + "epoch": 0.04491, + "grad_norm": 0.6733842898509934, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 4491 + }, + { + "epoch": 0.04492, + "grad_norm": 0.6313608339838067, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 4492 + }, + { + "epoch": 0.04493, + "grad_norm": 0.6192244819984872, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 4493 + }, + { + "epoch": 0.04494, + "grad_norm": 0.5488813855675564, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 4494 + }, + { + "epoch": 0.04495, + "grad_norm": 0.4779659159856146, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 4495 + }, + { + "epoch": 0.04496, + "grad_norm": 0.5060747754385665, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 4496 + }, + { + "epoch": 0.04497, + "grad_norm": 0.5808549525919214, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 4497 + }, + { + "epoch": 0.04498, + "grad_norm": 0.6694153212993217, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 4498 + }, + { + "epoch": 0.04499, + "grad_norm": 0.8090682417638635, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 4499 + }, + { + "epoch": 0.045, + "grad_norm": 0.9223841526186204, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4500 + }, + { + "epoch": 0.04501, + "grad_norm": 0.8349758444195646, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 4501 + }, + { + "epoch": 0.04502, + "grad_norm": 0.7254525437975357, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 4502 + }, + { + "epoch": 0.04503, + "grad_norm": 0.649056521260754, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 4503 + }, + { + "epoch": 0.04504, + "grad_norm": 0.713115140118379, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 4504 + }, + { + "epoch": 0.04505, + "grad_norm": 0.6117909603348726, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 4505 + }, + { + "epoch": 0.04506, + "grad_norm": 0.6145202957029064, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 4506 + }, + { + "epoch": 0.04507, + "grad_norm": 0.5247412068715119, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4507 + }, + { + "epoch": 0.04508, + "grad_norm": 0.5806017179220667, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 4508 + }, + { + "epoch": 0.04509, + "grad_norm": 0.5825330876210861, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 4509 + }, + { + "epoch": 0.0451, + "grad_norm": 0.7871833251091375, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 4510 + }, + { + "epoch": 0.04511, + "grad_norm": 0.8750417888525687, + "learning_rate": 0.003, + "loss": 4.147, + "step": 4511 + }, + { + "epoch": 0.04512, + "grad_norm": 0.8564995340864695, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4512 + }, + { + "epoch": 0.04513, + "grad_norm": 0.7843001776115731, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4513 + }, + { + "epoch": 0.04514, + "grad_norm": 0.6796657929720622, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 4514 + }, + { + "epoch": 0.04515, + "grad_norm": 0.6410320420606614, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 4515 + }, + { + "epoch": 0.04516, + "grad_norm": 0.6710348380128827, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 4516 + }, + { + "epoch": 0.04517, + "grad_norm": 0.8318718390934561, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 4517 + }, + { + "epoch": 0.04518, + "grad_norm": 0.895874624608404, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 4518 + }, + { + "epoch": 0.04519, + "grad_norm": 0.9529303615004293, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4519 + }, + { + "epoch": 0.0452, + "grad_norm": 1.00692610224058, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4520 + }, + { + "epoch": 0.04521, + "grad_norm": 1.0366622726930343, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 4521 + }, + { + "epoch": 0.04522, + "grad_norm": 1.0505298007829553, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 4522 + }, + { + "epoch": 0.04523, + "grad_norm": 1.0189524405337644, + "learning_rate": 0.003, + "loss": 4.158, + "step": 4523 + }, + { + "epoch": 0.04524, + "grad_norm": 1.030832446298102, + "learning_rate": 0.003, + "loss": 4.138, + "step": 4524 + }, + { + "epoch": 0.04525, + "grad_norm": 1.0054730825905025, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4525 + }, + { + "epoch": 0.04526, + "grad_norm": 0.9357043959907038, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 4526 + }, + { + "epoch": 0.04527, + "grad_norm": 0.880967833594335, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 4527 + }, + { + "epoch": 0.04528, + "grad_norm": 0.95913766957737, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4528 + }, + { + "epoch": 0.04529, + "grad_norm": 0.7989651014070877, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4529 + }, + { + "epoch": 0.0453, + "grad_norm": 0.7380961288194754, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4530 + }, + { + "epoch": 0.04531, + "grad_norm": 0.7085687602336856, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 4531 + }, + { + "epoch": 0.04532, + "grad_norm": 0.6574885202110717, + "learning_rate": 0.003, + "loss": 4.16, + "step": 4532 + }, + { + "epoch": 0.04533, + "grad_norm": 0.6519010842998008, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 4533 + }, + { + "epoch": 0.04534, + "grad_norm": 0.6093622128244879, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 4534 + }, + { + "epoch": 0.04535, + "grad_norm": 0.5837584644000219, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 4535 + }, + { + "epoch": 0.04536, + "grad_norm": 0.6897731390503988, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 4536 + }, + { + "epoch": 0.04537, + "grad_norm": 0.6949801331902556, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 4537 + }, + { + "epoch": 0.04538, + "grad_norm": 0.6776975735023881, + "learning_rate": 0.003, + "loss": 4.15, + "step": 4538 + }, + { + "epoch": 0.04539, + "grad_norm": 0.678568300988415, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4539 + }, + { + "epoch": 0.0454, + "grad_norm": 0.6603018556737656, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 4540 + }, + { + "epoch": 0.04541, + "grad_norm": 0.6133835370391497, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 4541 + }, + { + "epoch": 0.04542, + "grad_norm": 0.635535341053592, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 4542 + }, + { + "epoch": 0.04543, + "grad_norm": 0.668577889488378, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 4543 + }, + { + "epoch": 0.04544, + "grad_norm": 0.7427455997801305, + "learning_rate": 0.003, + "loss": 4.128, + "step": 4544 + }, + { + "epoch": 0.04545, + "grad_norm": 0.8639923142163211, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 4545 + }, + { + "epoch": 0.04546, + "grad_norm": 0.6895806692659665, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4546 + }, + { + "epoch": 0.04547, + "grad_norm": 0.49153824435946125, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 4547 + }, + { + "epoch": 0.04548, + "grad_norm": 0.5516451722786672, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 4548 + }, + { + "epoch": 0.04549, + "grad_norm": 0.6029789439713806, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4549 + }, + { + "epoch": 0.0455, + "grad_norm": 0.66622853988702, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 4550 + }, + { + "epoch": 0.04551, + "grad_norm": 0.7238008534924777, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 4551 + }, + { + "epoch": 0.04552, + "grad_norm": 0.7824263722798367, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 4552 + }, + { + "epoch": 0.04553, + "grad_norm": 0.7388703999593268, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 4553 + }, + { + "epoch": 0.04554, + "grad_norm": 0.6913086022347059, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4554 + }, + { + "epoch": 0.04555, + "grad_norm": 0.6442867508319275, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 4555 + }, + { + "epoch": 0.04556, + "grad_norm": 0.7238849740450599, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 4556 + }, + { + "epoch": 0.04557, + "grad_norm": 0.7976327742193281, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 4557 + }, + { + "epoch": 0.04558, + "grad_norm": 0.8032287746582655, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 4558 + }, + { + "epoch": 0.04559, + "grad_norm": 0.784841139322348, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 4559 + }, + { + "epoch": 0.0456, + "grad_norm": 0.8095586358578389, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4560 + }, + { + "epoch": 0.04561, + "grad_norm": 0.6748297662204855, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 4561 + }, + { + "epoch": 0.04562, + "grad_norm": 0.6987048066021282, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 4562 + }, + { + "epoch": 0.04563, + "grad_norm": 0.772254646470346, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 4563 + }, + { + "epoch": 0.04564, + "grad_norm": 0.8035846762926936, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 4564 + }, + { + "epoch": 0.04565, + "grad_norm": 0.7913541854291436, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 4565 + }, + { + "epoch": 0.04566, + "grad_norm": 0.7821636491006997, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4566 + }, + { + "epoch": 0.04567, + "grad_norm": 0.6863252350497664, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4567 + }, + { + "epoch": 0.04568, + "grad_norm": 0.6211213056274318, + "learning_rate": 0.003, + "loss": 4.112, + "step": 4568 + }, + { + "epoch": 0.04569, + "grad_norm": 0.7600485305292163, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 4569 + }, + { + "epoch": 0.0457, + "grad_norm": 0.7935297140012192, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 4570 + }, + { + "epoch": 0.04571, + "grad_norm": 0.8707627695603387, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 4571 + }, + { + "epoch": 0.04572, + "grad_norm": 0.9504850878398086, + "learning_rate": 0.003, + "loss": 4.146, + "step": 4572 + }, + { + "epoch": 0.04573, + "grad_norm": 0.9772978151032115, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 4573 + }, + { + "epoch": 0.04574, + "grad_norm": 0.8349860867562747, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4574 + }, + { + "epoch": 0.04575, + "grad_norm": 0.7275515356936376, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 4575 + }, + { + "epoch": 0.04576, + "grad_norm": 0.7453152360629928, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4576 + }, + { + "epoch": 0.04577, + "grad_norm": 0.731228336021725, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 4577 + }, + { + "epoch": 0.04578, + "grad_norm": 0.7565651611671647, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 4578 + }, + { + "epoch": 0.04579, + "grad_norm": 0.9198271859494617, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4579 + }, + { + "epoch": 0.0458, + "grad_norm": 1.0574643718931676, + "learning_rate": 0.003, + "loss": 4.168, + "step": 4580 + }, + { + "epoch": 0.04581, + "grad_norm": 0.9388725861321914, + "learning_rate": 0.003, + "loss": 4.1802, + "step": 4581 + }, + { + "epoch": 0.04582, + "grad_norm": 0.7599199575014808, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 4582 + }, + { + "epoch": 0.04583, + "grad_norm": 0.6683292234471547, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 4583 + }, + { + "epoch": 0.04584, + "grad_norm": 0.7512487692406998, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 4584 + }, + { + "epoch": 0.04585, + "grad_norm": 0.7081640921869536, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 4585 + }, + { + "epoch": 0.04586, + "grad_norm": 0.7168079051261622, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4586 + }, + { + "epoch": 0.04587, + "grad_norm": 0.814216089662895, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 4587 + }, + { + "epoch": 0.04588, + "grad_norm": 0.8322807065043963, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 4588 + }, + { + "epoch": 0.04589, + "grad_norm": 0.8173670039992074, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 4589 + }, + { + "epoch": 0.0459, + "grad_norm": 0.7002687692328883, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4590 + }, + { + "epoch": 0.04591, + "grad_norm": 0.667313250694922, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 4591 + }, + { + "epoch": 0.04592, + "grad_norm": 0.7721193873262023, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 4592 + }, + { + "epoch": 0.04593, + "grad_norm": 0.719951932975662, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4593 + }, + { + "epoch": 0.04594, + "grad_norm": 0.624806723533564, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 4594 + }, + { + "epoch": 0.04595, + "grad_norm": 0.5756945318000921, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 4595 + }, + { + "epoch": 0.04596, + "grad_norm": 0.5657179032543712, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 4596 + }, + { + "epoch": 0.04597, + "grad_norm": 0.6256040629924149, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 4597 + }, + { + "epoch": 0.04598, + "grad_norm": 0.6013110035097162, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 4598 + }, + { + "epoch": 0.04599, + "grad_norm": 0.7296067063170582, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 4599 + }, + { + "epoch": 0.046, + "grad_norm": 0.9209598285204096, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4600 + }, + { + "epoch": 0.04601, + "grad_norm": 1.0160633328517232, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 4601 + }, + { + "epoch": 0.04602, + "grad_norm": 0.8169968744441843, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4602 + }, + { + "epoch": 0.04603, + "grad_norm": 0.659664339286615, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4603 + }, + { + "epoch": 0.04604, + "grad_norm": 0.6592904260797935, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4604 + }, + { + "epoch": 0.04605, + "grad_norm": 0.6689967757886678, + "learning_rate": 0.003, + "loss": 4.159, + "step": 4605 + }, + { + "epoch": 0.04606, + "grad_norm": 0.6839466982971314, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 4606 + }, + { + "epoch": 0.04607, + "grad_norm": 0.6080788219356916, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 4607 + }, + { + "epoch": 0.04608, + "grad_norm": 0.6229335033564855, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 4608 + }, + { + "epoch": 0.04609, + "grad_norm": 0.6110178676005829, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 4609 + }, + { + "epoch": 0.0461, + "grad_norm": 0.6009057478520481, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 4610 + }, + { + "epoch": 0.04611, + "grad_norm": 0.667830986428792, + "learning_rate": 0.003, + "loss": 4.1615, + "step": 4611 + }, + { + "epoch": 0.04612, + "grad_norm": 0.8308783400085262, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4612 + }, + { + "epoch": 0.04613, + "grad_norm": 0.9122587432173699, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 4613 + }, + { + "epoch": 0.04614, + "grad_norm": 1.0201054294863492, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 4614 + }, + { + "epoch": 0.04615, + "grad_norm": 0.7968952064740754, + "learning_rate": 0.003, + "loss": 4.145, + "step": 4615 + }, + { + "epoch": 0.04616, + "grad_norm": 0.7040440287863108, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 4616 + }, + { + "epoch": 0.04617, + "grad_norm": 0.650743607835788, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 4617 + }, + { + "epoch": 0.04618, + "grad_norm": 0.6876364900587607, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4618 + }, + { + "epoch": 0.04619, + "grad_norm": 0.674403145056888, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 4619 + }, + { + "epoch": 0.0462, + "grad_norm": 0.6278438951559039, + "learning_rate": 0.003, + "loss": 4.158, + "step": 4620 + }, + { + "epoch": 0.04621, + "grad_norm": 0.5882186858354869, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 4621 + }, + { + "epoch": 0.04622, + "grad_norm": 0.6093666817375198, + "learning_rate": 0.003, + "loss": 4.155, + "step": 4622 + }, + { + "epoch": 0.04623, + "grad_norm": 0.5680931823986678, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 4623 + }, + { + "epoch": 0.04624, + "grad_norm": 0.6168812041909036, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 4624 + }, + { + "epoch": 0.04625, + "grad_norm": 0.6071577011256134, + "learning_rate": 0.003, + "loss": 4.136, + "step": 4625 + }, + { + "epoch": 0.04626, + "grad_norm": 0.5293202352744564, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 4626 + }, + { + "epoch": 0.04627, + "grad_norm": 0.5193967115713368, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 4627 + }, + { + "epoch": 0.04628, + "grad_norm": 0.6344196546552844, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 4628 + }, + { + "epoch": 0.04629, + "grad_norm": 0.6660491057377373, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4629 + }, + { + "epoch": 0.0463, + "grad_norm": 0.6585553354499009, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 4630 + }, + { + "epoch": 0.04631, + "grad_norm": 0.6065008696674108, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 4631 + }, + { + "epoch": 0.04632, + "grad_norm": 0.5310090972041924, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 4632 + }, + { + "epoch": 0.04633, + "grad_norm": 0.6357968891584513, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 4633 + }, + { + "epoch": 0.04634, + "grad_norm": 0.941578980575187, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4634 + }, + { + "epoch": 0.04635, + "grad_norm": 1.3572160948425378, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 4635 + }, + { + "epoch": 0.04636, + "grad_norm": 0.6013442724378576, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 4636 + }, + { + "epoch": 0.04637, + "grad_norm": 0.6555645652247263, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 4637 + }, + { + "epoch": 0.04638, + "grad_norm": 0.7328409108901823, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 4638 + }, + { + "epoch": 0.04639, + "grad_norm": 0.6867315466218923, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 4639 + }, + { + "epoch": 0.0464, + "grad_norm": 0.78928501351717, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 4640 + }, + { + "epoch": 0.04641, + "grad_norm": 0.8455497454514354, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 4641 + }, + { + "epoch": 0.04642, + "grad_norm": 0.8465515736905618, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 4642 + }, + { + "epoch": 0.04643, + "grad_norm": 0.7201267511516922, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 4643 + }, + { + "epoch": 0.04644, + "grad_norm": 0.7720786991059095, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 4644 + }, + { + "epoch": 0.04645, + "grad_norm": 0.7658630542557853, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4645 + }, + { + "epoch": 0.04646, + "grad_norm": 0.8677913304110078, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 4646 + }, + { + "epoch": 0.04647, + "grad_norm": 0.8914361615284006, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4647 + }, + { + "epoch": 0.04648, + "grad_norm": 0.8603752804558079, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 4648 + }, + { + "epoch": 0.04649, + "grad_norm": 0.8183443337243556, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 4649 + }, + { + "epoch": 0.0465, + "grad_norm": 0.6592577702353686, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4650 + }, + { + "epoch": 0.04651, + "grad_norm": 0.6033831169642737, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 4651 + }, + { + "epoch": 0.04652, + "grad_norm": 0.6694785473055216, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 4652 + }, + { + "epoch": 0.04653, + "grad_norm": 0.6797455964877727, + "learning_rate": 0.003, + "loss": 4.143, + "step": 4653 + }, + { + "epoch": 0.04654, + "grad_norm": 0.7400041353958778, + "learning_rate": 0.003, + "loss": 4.139, + "step": 4654 + }, + { + "epoch": 0.04655, + "grad_norm": 0.884024246484862, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 4655 + }, + { + "epoch": 0.04656, + "grad_norm": 0.9724046681595542, + "learning_rate": 0.003, + "loss": 4.127, + "step": 4656 + }, + { + "epoch": 0.04657, + "grad_norm": 0.9321801110747265, + "learning_rate": 0.003, + "loss": 4.1641, + "step": 4657 + }, + { + "epoch": 0.04658, + "grad_norm": 0.7977247035986766, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 4658 + }, + { + "epoch": 0.04659, + "grad_norm": 0.7837947068767307, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4659 + }, + { + "epoch": 0.0466, + "grad_norm": 0.8489578033894326, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 4660 + }, + { + "epoch": 0.04661, + "grad_norm": 0.9070979393786719, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4661 + }, + { + "epoch": 0.04662, + "grad_norm": 0.9271278255074006, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 4662 + }, + { + "epoch": 0.04663, + "grad_norm": 0.9626508931970053, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 4663 + }, + { + "epoch": 0.04664, + "grad_norm": 1.112988241634731, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4664 + }, + { + "epoch": 0.04665, + "grad_norm": 0.8921340954370226, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 4665 + }, + { + "epoch": 0.04666, + "grad_norm": 0.7837443287711118, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4666 + }, + { + "epoch": 0.04667, + "grad_norm": 0.7219249423509705, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 4667 + }, + { + "epoch": 0.04668, + "grad_norm": 0.7307342741354349, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 4668 + }, + { + "epoch": 0.04669, + "grad_norm": 0.7401471208519357, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 4669 + }, + { + "epoch": 0.0467, + "grad_norm": 0.7077495687150961, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 4670 + }, + { + "epoch": 0.04671, + "grad_norm": 0.7086785987292082, + "learning_rate": 0.003, + "loss": 4.136, + "step": 4671 + }, + { + "epoch": 0.04672, + "grad_norm": 0.7271653898870789, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 4672 + }, + { + "epoch": 0.04673, + "grad_norm": 0.7152687743613511, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4673 + }, + { + "epoch": 0.04674, + "grad_norm": 0.8387550898694738, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 4674 + }, + { + "epoch": 0.04675, + "grad_norm": 0.9264998756521228, + "learning_rate": 0.003, + "loss": 4.156, + "step": 4675 + }, + { + "epoch": 0.04676, + "grad_norm": 0.9760003719176258, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4676 + }, + { + "epoch": 0.04677, + "grad_norm": 0.9605322775687504, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 4677 + }, + { + "epoch": 0.04678, + "grad_norm": 0.760672310236161, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 4678 + }, + { + "epoch": 0.04679, + "grad_norm": 0.6288734540073927, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 4679 + }, + { + "epoch": 0.0468, + "grad_norm": 0.5340420649165649, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 4680 + }, + { + "epoch": 0.04681, + "grad_norm": 0.5383218654638344, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 4681 + }, + { + "epoch": 0.04682, + "grad_norm": 0.5921170754815444, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 4682 + }, + { + "epoch": 0.04683, + "grad_norm": 0.5986112675201656, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4683 + }, + { + "epoch": 0.04684, + "grad_norm": 0.6515018805380663, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 4684 + }, + { + "epoch": 0.04685, + "grad_norm": 0.6225025584760357, + "learning_rate": 0.003, + "loss": 4.125, + "step": 4685 + }, + { + "epoch": 0.04686, + "grad_norm": 0.5748567786413258, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 4686 + }, + { + "epoch": 0.04687, + "grad_norm": 0.597224880471858, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 4687 + }, + { + "epoch": 0.04688, + "grad_norm": 0.7033215388348866, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4688 + }, + { + "epoch": 0.04689, + "grad_norm": 0.8369783564473339, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 4689 + }, + { + "epoch": 0.0469, + "grad_norm": 0.8637378056248713, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4690 + }, + { + "epoch": 0.04691, + "grad_norm": 0.9863994238127417, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 4691 + }, + { + "epoch": 0.04692, + "grad_norm": 1.1588055062272775, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 4692 + }, + { + "epoch": 0.04693, + "grad_norm": 1.0154258289459517, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 4693 + }, + { + "epoch": 0.04694, + "grad_norm": 0.8542851528225356, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4694 + }, + { + "epoch": 0.04695, + "grad_norm": 0.7936283873091949, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 4695 + }, + { + "epoch": 0.04696, + "grad_norm": 0.8706455027386264, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 4696 + }, + { + "epoch": 0.04697, + "grad_norm": 0.888025520306328, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 4697 + }, + { + "epoch": 0.04698, + "grad_norm": 0.8024507884497375, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 4698 + }, + { + "epoch": 0.04699, + "grad_norm": 0.6716482593675086, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4699 + }, + { + "epoch": 0.047, + "grad_norm": 0.6875782963509716, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4700 + }, + { + "epoch": 0.04701, + "grad_norm": 0.6233215143134673, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 4701 + }, + { + "epoch": 0.04702, + "grad_norm": 0.5392566756700929, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 4702 + }, + { + "epoch": 0.04703, + "grad_norm": 0.5917569266603563, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 4703 + }, + { + "epoch": 0.04704, + "grad_norm": 0.6011918134456015, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 4704 + }, + { + "epoch": 0.04705, + "grad_norm": 0.6280725769151366, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 4705 + }, + { + "epoch": 0.04706, + "grad_norm": 0.6696710280228617, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4706 + }, + { + "epoch": 0.04707, + "grad_norm": 0.5908607329438127, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 4707 + }, + { + "epoch": 0.04708, + "grad_norm": 0.5153382206986533, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4708 + }, + { + "epoch": 0.04709, + "grad_norm": 0.513601992478386, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 4709 + }, + { + "epoch": 0.0471, + "grad_norm": 0.6490521053147635, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 4710 + }, + { + "epoch": 0.04711, + "grad_norm": 0.7794691577627103, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 4711 + }, + { + "epoch": 0.04712, + "grad_norm": 0.7688497592240825, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 4712 + }, + { + "epoch": 0.04713, + "grad_norm": 0.8439539645333411, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4713 + }, + { + "epoch": 0.04714, + "grad_norm": 0.9146059884037103, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 4714 + }, + { + "epoch": 0.04715, + "grad_norm": 0.8646027137635436, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 4715 + }, + { + "epoch": 0.04716, + "grad_norm": 0.7274583804499275, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4716 + }, + { + "epoch": 0.04717, + "grad_norm": 0.683321783957918, + "learning_rate": 0.003, + "loss": 4.144, + "step": 4717 + }, + { + "epoch": 0.04718, + "grad_norm": 0.7716759413978014, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4718 + }, + { + "epoch": 0.04719, + "grad_norm": 0.79629647999161, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 4719 + }, + { + "epoch": 0.0472, + "grad_norm": 0.7050760628166665, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 4720 + }, + { + "epoch": 0.04721, + "grad_norm": 0.6252610323631939, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 4721 + }, + { + "epoch": 0.04722, + "grad_norm": 0.6123095727231639, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 4722 + }, + { + "epoch": 0.04723, + "grad_norm": 0.6308650251833371, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 4723 + }, + { + "epoch": 0.04724, + "grad_norm": 0.6995282704649718, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 4724 + }, + { + "epoch": 0.04725, + "grad_norm": 0.822004537328782, + "learning_rate": 0.003, + "loss": 4.141, + "step": 4725 + }, + { + "epoch": 0.04726, + "grad_norm": 0.9108754250302765, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 4726 + }, + { + "epoch": 0.04727, + "grad_norm": 0.9359931741571359, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 4727 + }, + { + "epoch": 0.04728, + "grad_norm": 0.8319271568838779, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 4728 + }, + { + "epoch": 0.04729, + "grad_norm": 0.7272073389712985, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 4729 + }, + { + "epoch": 0.0473, + "grad_norm": 0.6527677653973615, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 4730 + }, + { + "epoch": 0.04731, + "grad_norm": 0.5930657999039424, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 4731 + }, + { + "epoch": 0.04732, + "grad_norm": 0.6549784460880107, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 4732 + }, + { + "epoch": 0.04733, + "grad_norm": 0.8327578035736742, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4733 + }, + { + "epoch": 0.04734, + "grad_norm": 1.104638767629771, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 4734 + }, + { + "epoch": 0.04735, + "grad_norm": 0.9388734582276314, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 4735 + }, + { + "epoch": 0.04736, + "grad_norm": 0.7892905763980204, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 4736 + }, + { + "epoch": 0.04737, + "grad_norm": 0.8500450443730917, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 4737 + }, + { + "epoch": 0.04738, + "grad_norm": 0.9289248020340674, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 4738 + }, + { + "epoch": 0.04739, + "grad_norm": 0.955981339694237, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 4739 + }, + { + "epoch": 0.0474, + "grad_norm": 0.9041807476389423, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 4740 + }, + { + "epoch": 0.04741, + "grad_norm": 0.8492001903774913, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 4741 + }, + { + "epoch": 0.04742, + "grad_norm": 0.7853348835474101, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 4742 + }, + { + "epoch": 0.04743, + "grad_norm": 0.7795682667464989, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 4743 + }, + { + "epoch": 0.04744, + "grad_norm": 0.7267419747954841, + "learning_rate": 0.003, + "loss": 4.119, + "step": 4744 + }, + { + "epoch": 0.04745, + "grad_norm": 0.7262044399015473, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4745 + }, + { + "epoch": 0.04746, + "grad_norm": 0.7299893869109523, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4746 + }, + { + "epoch": 0.04747, + "grad_norm": 0.806239055464935, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 4747 + }, + { + "epoch": 0.04748, + "grad_norm": 0.6789998274727187, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 4748 + }, + { + "epoch": 0.04749, + "grad_norm": 0.54858705364636, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 4749 + }, + { + "epoch": 0.0475, + "grad_norm": 0.5060470689311919, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4750 + }, + { + "epoch": 0.04751, + "grad_norm": 0.5080665649296767, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 4751 + }, + { + "epoch": 0.04752, + "grad_norm": 0.5597535217525249, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 4752 + }, + { + "epoch": 0.04753, + "grad_norm": 0.5815761039735459, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4753 + }, + { + "epoch": 0.04754, + "grad_norm": 0.6895819778653248, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 4754 + }, + { + "epoch": 0.04755, + "grad_norm": 0.9251374027849657, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 4755 + }, + { + "epoch": 0.04756, + "grad_norm": 1.1438681733923501, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 4756 + }, + { + "epoch": 0.04757, + "grad_norm": 0.7788109201469021, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 4757 + }, + { + "epoch": 0.04758, + "grad_norm": 0.6624956821204955, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 4758 + }, + { + "epoch": 0.04759, + "grad_norm": 0.7037195304943054, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 4759 + }, + { + "epoch": 0.0476, + "grad_norm": 0.7411798494302422, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 4760 + }, + { + "epoch": 0.04761, + "grad_norm": 0.7054700847471357, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 4761 + }, + { + "epoch": 0.04762, + "grad_norm": 0.6226720435093628, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 4762 + }, + { + "epoch": 0.04763, + "grad_norm": 0.6886529154194588, + "learning_rate": 0.003, + "loss": 4.122, + "step": 4763 + }, + { + "epoch": 0.04764, + "grad_norm": 0.6992577861087502, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 4764 + }, + { + "epoch": 0.04765, + "grad_norm": 0.6613811777977939, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 4765 + }, + { + "epoch": 0.04766, + "grad_norm": 0.5851808207831782, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 4766 + }, + { + "epoch": 0.04767, + "grad_norm": 0.6679898130461849, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 4767 + }, + { + "epoch": 0.04768, + "grad_norm": 0.7821493077256183, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 4768 + }, + { + "epoch": 0.04769, + "grad_norm": 0.8355601404081668, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 4769 + }, + { + "epoch": 0.0477, + "grad_norm": 0.8479665704653694, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 4770 + }, + { + "epoch": 0.04771, + "grad_norm": 0.8866254860935738, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 4771 + }, + { + "epoch": 0.04772, + "grad_norm": 0.915742333919775, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 4772 + }, + { + "epoch": 0.04773, + "grad_norm": 0.754051301920501, + "learning_rate": 0.003, + "loss": 4.157, + "step": 4773 + }, + { + "epoch": 0.04774, + "grad_norm": 0.7232939078296818, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 4774 + }, + { + "epoch": 0.04775, + "grad_norm": 0.8169270534707214, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 4775 + }, + { + "epoch": 0.04776, + "grad_norm": 0.8573605280685423, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4776 + }, + { + "epoch": 0.04777, + "grad_norm": 0.9120618578003061, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4777 + }, + { + "epoch": 0.04778, + "grad_norm": 0.9392445159690339, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4778 + }, + { + "epoch": 0.04779, + "grad_norm": 0.8671768686651342, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4779 + }, + { + "epoch": 0.0478, + "grad_norm": 0.8025049205614178, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 4780 + }, + { + "epoch": 0.04781, + "grad_norm": 0.7254367126713761, + "learning_rate": 0.003, + "loss": 4.125, + "step": 4781 + }, + { + "epoch": 0.04782, + "grad_norm": 0.6859662827550447, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4782 + }, + { + "epoch": 0.04783, + "grad_norm": 0.6967231284903036, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 4783 + }, + { + "epoch": 0.04784, + "grad_norm": 0.7980703901010124, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 4784 + }, + { + "epoch": 0.04785, + "grad_norm": 0.8254970324649596, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 4785 + }, + { + "epoch": 0.04786, + "grad_norm": 0.8777922852113546, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4786 + }, + { + "epoch": 0.04787, + "grad_norm": 0.7491389893707523, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 4787 + }, + { + "epoch": 0.04788, + "grad_norm": 0.6503832601570363, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 4788 + }, + { + "epoch": 0.04789, + "grad_norm": 0.7031951667657491, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 4789 + }, + { + "epoch": 0.0479, + "grad_norm": 0.678065080300084, + "learning_rate": 0.003, + "loss": 4.131, + "step": 4790 + }, + { + "epoch": 0.04791, + "grad_norm": 0.523098323506981, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 4791 + }, + { + "epoch": 0.04792, + "grad_norm": 0.5260803142583377, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 4792 + }, + { + "epoch": 0.04793, + "grad_norm": 0.5215649768374346, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 4793 + }, + { + "epoch": 0.04794, + "grad_norm": 0.46968713617443264, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 4794 + }, + { + "epoch": 0.04795, + "grad_norm": 0.5247439754384676, + "learning_rate": 0.003, + "loss": 4.129, + "step": 4795 + }, + { + "epoch": 0.04796, + "grad_norm": 0.6933674582411209, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4796 + }, + { + "epoch": 0.04797, + "grad_norm": 0.9807791236957965, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 4797 + }, + { + "epoch": 0.04798, + "grad_norm": 1.1741198222289444, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 4798 + }, + { + "epoch": 0.04799, + "grad_norm": 0.6525704641811733, + "learning_rate": 0.003, + "loss": 4.116, + "step": 4799 + }, + { + "epoch": 0.048, + "grad_norm": 0.8089064094414999, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4800 + }, + { + "epoch": 0.04801, + "grad_norm": 0.870503514300495, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 4801 + }, + { + "epoch": 0.04802, + "grad_norm": 0.6635652157598942, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4802 + }, + { + "epoch": 0.04803, + "grad_norm": 0.6200303139785031, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4803 + }, + { + "epoch": 0.04804, + "grad_norm": 0.762754226201963, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 4804 + }, + { + "epoch": 0.04805, + "grad_norm": 0.7861473887229704, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 4805 + }, + { + "epoch": 0.04806, + "grad_norm": 0.7696022521328222, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 4806 + }, + { + "epoch": 0.04807, + "grad_norm": 0.7570918480886352, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4807 + }, + { + "epoch": 0.04808, + "grad_norm": 0.8705313373191277, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 4808 + }, + { + "epoch": 0.04809, + "grad_norm": 0.8255463817975629, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 4809 + }, + { + "epoch": 0.0481, + "grad_norm": 0.8537469900625964, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4810 + }, + { + "epoch": 0.04811, + "grad_norm": 0.8483723617783365, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4811 + }, + { + "epoch": 0.04812, + "grad_norm": 0.7645031119405934, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 4812 + }, + { + "epoch": 0.04813, + "grad_norm": 0.8283116512854986, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 4813 + }, + { + "epoch": 0.04814, + "grad_norm": 0.8641578230891004, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 4814 + }, + { + "epoch": 0.04815, + "grad_norm": 0.8919893073764681, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 4815 + }, + { + "epoch": 0.04816, + "grad_norm": 0.8648426828228101, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 4816 + }, + { + "epoch": 0.04817, + "grad_norm": 0.9477238526245142, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 4817 + }, + { + "epoch": 0.04818, + "grad_norm": 0.9252788072968092, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 4818 + }, + { + "epoch": 0.04819, + "grad_norm": 0.8797819035450328, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4819 + }, + { + "epoch": 0.0482, + "grad_norm": 0.9035562726868283, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 4820 + }, + { + "epoch": 0.04821, + "grad_norm": 0.858921817091334, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 4821 + }, + { + "epoch": 0.04822, + "grad_norm": 0.6583705277011013, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 4822 + }, + { + "epoch": 0.04823, + "grad_norm": 0.5779379743608074, + "learning_rate": 0.003, + "loss": 4.128, + "step": 4823 + }, + { + "epoch": 0.04824, + "grad_norm": 0.5505773694357159, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4824 + }, + { + "epoch": 0.04825, + "grad_norm": 0.5013668074041548, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4825 + }, + { + "epoch": 0.04826, + "grad_norm": 0.5240704697199119, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 4826 + }, + { + "epoch": 0.04827, + "grad_norm": 0.533629158460227, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 4827 + }, + { + "epoch": 0.04828, + "grad_norm": 0.5992841702695002, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 4828 + }, + { + "epoch": 0.04829, + "grad_norm": 0.6683602415515053, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 4829 + }, + { + "epoch": 0.0483, + "grad_norm": 0.7131645514243471, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 4830 + }, + { + "epoch": 0.04831, + "grad_norm": 0.7918054904389603, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4831 + }, + { + "epoch": 0.04832, + "grad_norm": 0.8874223110243422, + "learning_rate": 0.003, + "loss": 4.128, + "step": 4832 + }, + { + "epoch": 0.04833, + "grad_norm": 0.9718339603744532, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 4833 + }, + { + "epoch": 0.04834, + "grad_norm": 0.9758671382016597, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 4834 + }, + { + "epoch": 0.04835, + "grad_norm": 0.9485480338404471, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 4835 + }, + { + "epoch": 0.04836, + "grad_norm": 1.0708614312540052, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4836 + }, + { + "epoch": 0.04837, + "grad_norm": 1.005625524837446, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 4837 + }, + { + "epoch": 0.04838, + "grad_norm": 1.142957704221482, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 4838 + }, + { + "epoch": 0.04839, + "grad_norm": 0.8876209726930161, + "learning_rate": 0.003, + "loss": 4.1597, + "step": 4839 + }, + { + "epoch": 0.0484, + "grad_norm": 1.0000157052113117, + "learning_rate": 0.003, + "loss": 4.1726, + "step": 4840 + }, + { + "epoch": 0.04841, + "grad_norm": 0.8947816300534793, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 4841 + }, + { + "epoch": 0.04842, + "grad_norm": 1.0306766066880264, + "learning_rate": 0.003, + "loss": 4.1906, + "step": 4842 + }, + { + "epoch": 0.04843, + "grad_norm": 1.1670348326264735, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 4843 + }, + { + "epoch": 0.04844, + "grad_norm": 0.8379009760125189, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 4844 + }, + { + "epoch": 0.04845, + "grad_norm": 0.9239064010183352, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 4845 + }, + { + "epoch": 0.04846, + "grad_norm": 0.9084764134640881, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 4846 + }, + { + "epoch": 0.04847, + "grad_norm": 0.7704730780996573, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 4847 + }, + { + "epoch": 0.04848, + "grad_norm": 0.8040182191975439, + "learning_rate": 0.003, + "loss": 4.186, + "step": 4848 + }, + { + "epoch": 0.04849, + "grad_norm": 0.9074519148966281, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 4849 + }, + { + "epoch": 0.0485, + "grad_norm": 1.1771056378102565, + "learning_rate": 0.003, + "loss": 4.1773, + "step": 4850 + }, + { + "epoch": 0.04851, + "grad_norm": 0.9654993158924701, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 4851 + }, + { + "epoch": 0.04852, + "grad_norm": 0.7654220286623955, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4852 + }, + { + "epoch": 0.04853, + "grad_norm": 0.7291058591496458, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 4853 + }, + { + "epoch": 0.04854, + "grad_norm": 0.6997930550604129, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 4854 + }, + { + "epoch": 0.04855, + "grad_norm": 0.8234663642795009, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 4855 + }, + { + "epoch": 0.04856, + "grad_norm": 0.7798398187689213, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 4856 + }, + { + "epoch": 0.04857, + "grad_norm": 0.6411382707024734, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 4857 + }, + { + "epoch": 0.04858, + "grad_norm": 0.6914930361794667, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 4858 + }, + { + "epoch": 0.04859, + "grad_norm": 0.742724632667153, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 4859 + }, + { + "epoch": 0.0486, + "grad_norm": 0.6805430980474091, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 4860 + }, + { + "epoch": 0.04861, + "grad_norm": 0.538868342282314, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4861 + }, + { + "epoch": 0.04862, + "grad_norm": 0.4415725147391537, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 4862 + }, + { + "epoch": 0.04863, + "grad_norm": 0.46517280056665583, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 4863 + }, + { + "epoch": 0.04864, + "grad_norm": 0.44851841399681913, + "learning_rate": 0.003, + "loss": 4.13, + "step": 4864 + }, + { + "epoch": 0.04865, + "grad_norm": 0.4755135869431016, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 4865 + }, + { + "epoch": 0.04866, + "grad_norm": 0.449801980415272, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 4866 + }, + { + "epoch": 0.04867, + "grad_norm": 0.41976113144986854, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 4867 + }, + { + "epoch": 0.04868, + "grad_norm": 0.41274576265380564, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4868 + }, + { + "epoch": 0.04869, + "grad_norm": 0.4120235509377836, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 4869 + }, + { + "epoch": 0.0487, + "grad_norm": 0.45999567656300233, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 4870 + }, + { + "epoch": 0.04871, + "grad_norm": 0.5762534303017752, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 4871 + }, + { + "epoch": 0.04872, + "grad_norm": 0.6778749879598042, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 4872 + }, + { + "epoch": 0.04873, + "grad_norm": 0.7636826849696239, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 4873 + }, + { + "epoch": 0.04874, + "grad_norm": 0.7169955577952278, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4874 + }, + { + "epoch": 0.04875, + "grad_norm": 0.5747097191102749, + "learning_rate": 0.003, + "loss": 4.118, + "step": 4875 + }, + { + "epoch": 0.04876, + "grad_norm": 0.5097483392386485, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 4876 + }, + { + "epoch": 0.04877, + "grad_norm": 0.5291097317099629, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 4877 + }, + { + "epoch": 0.04878, + "grad_norm": 0.5314236989459756, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4878 + }, + { + "epoch": 0.04879, + "grad_norm": 0.5936650310790298, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4879 + }, + { + "epoch": 0.0488, + "grad_norm": 0.7538105145753773, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 4880 + }, + { + "epoch": 0.04881, + "grad_norm": 0.9419961284154782, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 4881 + }, + { + "epoch": 0.04882, + "grad_norm": 1.125784565290687, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 4882 + }, + { + "epoch": 0.04883, + "grad_norm": 0.7937742064689157, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 4883 + }, + { + "epoch": 0.04884, + "grad_norm": 0.7276138590729601, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 4884 + }, + { + "epoch": 0.04885, + "grad_norm": 0.7635121030741637, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4885 + }, + { + "epoch": 0.04886, + "grad_norm": 0.8525933214692812, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 4886 + }, + { + "epoch": 0.04887, + "grad_norm": 0.8979433474373131, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 4887 + }, + { + "epoch": 0.04888, + "grad_norm": 0.9257981316717354, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 4888 + }, + { + "epoch": 0.04889, + "grad_norm": 0.8410005218082275, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 4889 + }, + { + "epoch": 0.0489, + "grad_norm": 0.8463344038585867, + "learning_rate": 0.003, + "loss": 4.14, + "step": 4890 + }, + { + "epoch": 0.04891, + "grad_norm": 0.7647404294541085, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 4891 + }, + { + "epoch": 0.04892, + "grad_norm": 0.7679081300937539, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4892 + }, + { + "epoch": 0.04893, + "grad_norm": 0.8575058455749437, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 4893 + }, + { + "epoch": 0.04894, + "grad_norm": 0.8645404495029662, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 4894 + }, + { + "epoch": 0.04895, + "grad_norm": 0.8334125833446672, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 4895 + }, + { + "epoch": 0.04896, + "grad_norm": 0.7851358889893844, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 4896 + }, + { + "epoch": 0.04897, + "grad_norm": 0.7536500770250976, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 4897 + }, + { + "epoch": 0.04898, + "grad_norm": 0.7125975228161553, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 4898 + }, + { + "epoch": 0.04899, + "grad_norm": 0.6723493590903667, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 4899 + }, + { + "epoch": 0.049, + "grad_norm": 0.5994822751068893, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 4900 + }, + { + "epoch": 0.04901, + "grad_norm": 0.562371614276483, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 4901 + }, + { + "epoch": 0.04902, + "grad_norm": 0.6664145016393446, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 4902 + }, + { + "epoch": 0.04903, + "grad_norm": 0.6848367937927413, + "learning_rate": 0.003, + "loss": 4.147, + "step": 4903 + }, + { + "epoch": 0.04904, + "grad_norm": 0.6584184589527915, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4904 + }, + { + "epoch": 0.04905, + "grad_norm": 0.6238214939939397, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4905 + }, + { + "epoch": 0.04906, + "grad_norm": 0.6975587494661175, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 4906 + }, + { + "epoch": 0.04907, + "grad_norm": 0.8941666785365164, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 4907 + }, + { + "epoch": 0.04908, + "grad_norm": 1.081340571949139, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 4908 + }, + { + "epoch": 0.04909, + "grad_norm": 0.8257080561116473, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 4909 + }, + { + "epoch": 0.0491, + "grad_norm": 0.6982931611612875, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4910 + }, + { + "epoch": 0.04911, + "grad_norm": 0.8003274963133196, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 4911 + }, + { + "epoch": 0.04912, + "grad_norm": 0.8407576542296316, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 4912 + }, + { + "epoch": 0.04913, + "grad_norm": 0.9633535310499528, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 4913 + }, + { + "epoch": 0.04914, + "grad_norm": 0.9685795549695555, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 4914 + }, + { + "epoch": 0.04915, + "grad_norm": 0.8915068602810806, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 4915 + }, + { + "epoch": 0.04916, + "grad_norm": 0.7915988128949956, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 4916 + }, + { + "epoch": 0.04917, + "grad_norm": 0.7634607572088334, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4917 + }, + { + "epoch": 0.04918, + "grad_norm": 0.7786335105527836, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 4918 + }, + { + "epoch": 0.04919, + "grad_norm": 0.8773870158083756, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 4919 + }, + { + "epoch": 0.0492, + "grad_norm": 0.7845152066567127, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 4920 + }, + { + "epoch": 0.04921, + "grad_norm": 0.8250729750583956, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 4921 + }, + { + "epoch": 0.04922, + "grad_norm": 0.8194188230604832, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 4922 + }, + { + "epoch": 0.04923, + "grad_norm": 0.8678442267593626, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 4923 + }, + { + "epoch": 0.04924, + "grad_norm": 0.9322111328758901, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 4924 + }, + { + "epoch": 0.04925, + "grad_norm": 0.883656356593312, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 4925 + }, + { + "epoch": 0.04926, + "grad_norm": 0.7834220179620971, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 4926 + }, + { + "epoch": 0.04927, + "grad_norm": 0.648822263811632, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 4927 + }, + { + "epoch": 0.04928, + "grad_norm": 0.6606153717082371, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 4928 + }, + { + "epoch": 0.04929, + "grad_norm": 0.6448549657492914, + "learning_rate": 0.003, + "loss": 4.132, + "step": 4929 + }, + { + "epoch": 0.0493, + "grad_norm": 0.6537446324241543, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4930 + }, + { + "epoch": 0.04931, + "grad_norm": 0.6782203397244866, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 4931 + }, + { + "epoch": 0.04932, + "grad_norm": 0.7378598466407571, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 4932 + }, + { + "epoch": 0.04933, + "grad_norm": 0.7395809589970013, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 4933 + }, + { + "epoch": 0.04934, + "grad_norm": 0.7658041842655043, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 4934 + }, + { + "epoch": 0.04935, + "grad_norm": 0.9019517516492989, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 4935 + }, + { + "epoch": 0.04936, + "grad_norm": 1.1441200899996462, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 4936 + }, + { + "epoch": 0.04937, + "grad_norm": 0.8788107449320854, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4937 + }, + { + "epoch": 0.04938, + "grad_norm": 0.8007894624899063, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 4938 + }, + { + "epoch": 0.04939, + "grad_norm": 0.789503524635859, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 4939 + }, + { + "epoch": 0.0494, + "grad_norm": 0.787654716377685, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 4940 + }, + { + "epoch": 0.04941, + "grad_norm": 0.6688134872012756, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 4941 + }, + { + "epoch": 0.04942, + "grad_norm": 0.5978385156801125, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4942 + }, + { + "epoch": 0.04943, + "grad_norm": 0.6124698730462602, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 4943 + }, + { + "epoch": 0.04944, + "grad_norm": 0.5591922471734772, + "learning_rate": 0.003, + "loss": 4.154, + "step": 4944 + }, + { + "epoch": 0.04945, + "grad_norm": 0.4939591571209979, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 4945 + }, + { + "epoch": 0.04946, + "grad_norm": 0.49333816406820896, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 4946 + }, + { + "epoch": 0.04947, + "grad_norm": 0.5127970104844136, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 4947 + }, + { + "epoch": 0.04948, + "grad_norm": 0.5804249788539954, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 4948 + }, + { + "epoch": 0.04949, + "grad_norm": 0.6106490006275131, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 4949 + }, + { + "epoch": 0.0495, + "grad_norm": 0.7326284913442733, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4950 + }, + { + "epoch": 0.04951, + "grad_norm": 0.8806738854393903, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 4951 + }, + { + "epoch": 0.04952, + "grad_norm": 0.8907989929621909, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 4952 + }, + { + "epoch": 0.04953, + "grad_norm": 0.6504370122838272, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 4953 + }, + { + "epoch": 0.04954, + "grad_norm": 0.5399865076183303, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 4954 + }, + { + "epoch": 0.04955, + "grad_norm": 0.6426116557249149, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 4955 + }, + { + "epoch": 0.04956, + "grad_norm": 0.6565119107623256, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 4956 + }, + { + "epoch": 0.04957, + "grad_norm": 0.6722789026797903, + "learning_rate": 0.003, + "loss": 4.106, + "step": 4957 + }, + { + "epoch": 0.04958, + "grad_norm": 0.5890068452063758, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 4958 + }, + { + "epoch": 0.04959, + "grad_norm": 0.5878861479683556, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 4959 + }, + { + "epoch": 0.0496, + "grad_norm": 0.6509946110169393, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4960 + }, + { + "epoch": 0.04961, + "grad_norm": 0.6156850509647105, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 4961 + }, + { + "epoch": 0.04962, + "grad_norm": 0.6497590683615507, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 4962 + }, + { + "epoch": 0.04963, + "grad_norm": 0.6412215380917496, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 4963 + }, + { + "epoch": 0.04964, + "grad_norm": 0.6061448543738891, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 4964 + }, + { + "epoch": 0.04965, + "grad_norm": 0.6123953168115248, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 4965 + }, + { + "epoch": 0.04966, + "grad_norm": 0.8037505368816966, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 4966 + }, + { + "epoch": 0.04967, + "grad_norm": 0.9549949754561539, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 4967 + }, + { + "epoch": 0.04968, + "grad_norm": 1.1281595822055885, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 4968 + }, + { + "epoch": 0.04969, + "grad_norm": 0.8652927562240527, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 4969 + }, + { + "epoch": 0.0497, + "grad_norm": 0.7881031556981761, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 4970 + }, + { + "epoch": 0.04971, + "grad_norm": 0.8417863445996929, + "learning_rate": 0.003, + "loss": 4.153, + "step": 4971 + }, + { + "epoch": 0.04972, + "grad_norm": 0.8493800636053089, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 4972 + }, + { + "epoch": 0.04973, + "grad_norm": 0.9639644128846697, + "learning_rate": 0.003, + "loss": 4.1826, + "step": 4973 + }, + { + "epoch": 0.04974, + "grad_norm": 1.0979045911236074, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 4974 + }, + { + "epoch": 0.04975, + "grad_norm": 1.1499829958321686, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 4975 + }, + { + "epoch": 0.04976, + "grad_norm": 0.896257641249768, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 4976 + }, + { + "epoch": 0.04977, + "grad_norm": 0.8176607909447694, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 4977 + }, + { + "epoch": 0.04978, + "grad_norm": 0.8167430782012225, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 4978 + }, + { + "epoch": 0.04979, + "grad_norm": 0.8298370431983914, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 4979 + }, + { + "epoch": 0.0498, + "grad_norm": 0.8918442192162982, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 4980 + }, + { + "epoch": 0.04981, + "grad_norm": 0.8954883580668984, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 4981 + }, + { + "epoch": 0.04982, + "grad_norm": 0.8872182858380784, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 4982 + }, + { + "epoch": 0.04983, + "grad_norm": 0.7800382368496617, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 4983 + }, + { + "epoch": 0.04984, + "grad_norm": 0.7656679202701089, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4984 + }, + { + "epoch": 0.04985, + "grad_norm": 0.8692356760179402, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 4985 + }, + { + "epoch": 0.04986, + "grad_norm": 0.8658640091365618, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 4986 + }, + { + "epoch": 0.04987, + "grad_norm": 0.9418717422739423, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 4987 + }, + { + "epoch": 0.04988, + "grad_norm": 0.9193482062490026, + "learning_rate": 0.003, + "loss": 4.1628, + "step": 4988 + }, + { + "epoch": 0.04989, + "grad_norm": 0.8335538614629532, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 4989 + }, + { + "epoch": 0.0499, + "grad_norm": 0.7458856658289911, + "learning_rate": 0.003, + "loss": 4.136, + "step": 4990 + }, + { + "epoch": 0.04991, + "grad_norm": 0.6549431595291397, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 4991 + }, + { + "epoch": 0.04992, + "grad_norm": 0.6708966834866974, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4992 + }, + { + "epoch": 0.04993, + "grad_norm": 0.6367637070694386, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 4993 + }, + { + "epoch": 0.04994, + "grad_norm": 0.6415299490019786, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 4994 + }, + { + "epoch": 0.04995, + "grad_norm": 0.7355718950683062, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 4995 + }, + { + "epoch": 0.04996, + "grad_norm": 0.8795955713743572, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 4996 + }, + { + "epoch": 0.04997, + "grad_norm": 1.1069671993764165, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 4997 + }, + { + "epoch": 0.04998, + "grad_norm": 0.9565748099055139, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 4998 + }, + { + "epoch": 0.04999, + "grad_norm": 0.8361592261029797, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 4999 + }, + { + "epoch": 0.05, + "grad_norm": 0.9200109034528712, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 5000 + }, + { + "epoch": 0.05001, + "grad_norm": 0.9617359528015215, + "learning_rate": 0.003, + "loss": 4.172, + "step": 5001 + }, + { + "epoch": 0.05002, + "grad_norm": 0.8052353535299214, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5002 + }, + { + "epoch": 0.05003, + "grad_norm": 0.7285922794371577, + "learning_rate": 0.003, + "loss": 4.152, + "step": 5003 + }, + { + "epoch": 0.05004, + "grad_norm": 0.6735195112013481, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5004 + }, + { + "epoch": 0.05005, + "grad_norm": 0.6049497214440789, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 5005 + }, + { + "epoch": 0.05006, + "grad_norm": 0.6066071944555775, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5006 + }, + { + "epoch": 0.05007, + "grad_norm": 0.558024137943532, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 5007 + }, + { + "epoch": 0.05008, + "grad_norm": 0.5395388306678914, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 5008 + }, + { + "epoch": 0.05009, + "grad_norm": 0.5556098595038315, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5009 + }, + { + "epoch": 0.0501, + "grad_norm": 0.6653438296094251, + "learning_rate": 0.003, + "loss": 4.133, + "step": 5010 + }, + { + "epoch": 0.05011, + "grad_norm": 0.77449202414198, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 5011 + }, + { + "epoch": 0.05012, + "grad_norm": 0.7658330704952865, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5012 + }, + { + "epoch": 0.05013, + "grad_norm": 0.9103652169172048, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 5013 + }, + { + "epoch": 0.05014, + "grad_norm": 1.0759287643443909, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 5014 + }, + { + "epoch": 0.05015, + "grad_norm": 0.9691643650194389, + "learning_rate": 0.003, + "loss": 4.1534, + "step": 5015 + }, + { + "epoch": 0.05016, + "grad_norm": 0.9073305570987539, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 5016 + }, + { + "epoch": 0.05017, + "grad_norm": 0.8073149334354928, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 5017 + }, + { + "epoch": 0.05018, + "grad_norm": 0.7017737986112383, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 5018 + }, + { + "epoch": 0.05019, + "grad_norm": 0.6777526168401347, + "learning_rate": 0.003, + "loss": 4.157, + "step": 5019 + }, + { + "epoch": 0.0502, + "grad_norm": 0.7096093854788539, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 5020 + }, + { + "epoch": 0.05021, + "grad_norm": 0.6174117502687014, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5021 + }, + { + "epoch": 0.05022, + "grad_norm": 0.6289052553001543, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 5022 + }, + { + "epoch": 0.05023, + "grad_norm": 0.6439915077819784, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 5023 + }, + { + "epoch": 0.05024, + "grad_norm": 0.6395721391541763, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 5024 + }, + { + "epoch": 0.05025, + "grad_norm": 0.6213156408060325, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5025 + }, + { + "epoch": 0.05026, + "grad_norm": 0.6019908233874888, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 5026 + }, + { + "epoch": 0.05027, + "grad_norm": 0.7041501406597449, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5027 + }, + { + "epoch": 0.05028, + "grad_norm": 0.8719881236969269, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5028 + }, + { + "epoch": 0.05029, + "grad_norm": 0.9033846221678056, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 5029 + }, + { + "epoch": 0.0503, + "grad_norm": 0.8943098923163394, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5030 + }, + { + "epoch": 0.05031, + "grad_norm": 0.8280905971408706, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5031 + }, + { + "epoch": 0.05032, + "grad_norm": 0.8573742108993896, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 5032 + }, + { + "epoch": 0.05033, + "grad_norm": 0.8521216692297204, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5033 + }, + { + "epoch": 0.05034, + "grad_norm": 0.8336742938862856, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5034 + }, + { + "epoch": 0.05035, + "grad_norm": 0.7155132757115704, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 5035 + }, + { + "epoch": 0.05036, + "grad_norm": 0.7077255442757248, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 5036 + }, + { + "epoch": 0.05037, + "grad_norm": 0.6598592587060891, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 5037 + }, + { + "epoch": 0.05038, + "grad_norm": 0.6985786753992844, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 5038 + }, + { + "epoch": 0.05039, + "grad_norm": 0.7840225900778709, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 5039 + }, + { + "epoch": 0.0504, + "grad_norm": 0.8187206569784481, + "learning_rate": 0.003, + "loss": 4.139, + "step": 5040 + }, + { + "epoch": 0.05041, + "grad_norm": 0.8117416189384354, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 5041 + }, + { + "epoch": 0.05042, + "grad_norm": 0.701867915130458, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5042 + }, + { + "epoch": 0.05043, + "grad_norm": 0.6612953236447207, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 5043 + }, + { + "epoch": 0.05044, + "grad_norm": 0.6173987969701078, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 5044 + }, + { + "epoch": 0.05045, + "grad_norm": 0.5892878279295536, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 5045 + }, + { + "epoch": 0.05046, + "grad_norm": 0.6842586991400519, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 5046 + }, + { + "epoch": 0.05047, + "grad_norm": 0.7460792551860184, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 5047 + }, + { + "epoch": 0.05048, + "grad_norm": 0.7111178192864371, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 5048 + }, + { + "epoch": 0.05049, + "grad_norm": 0.646577206489084, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5049 + }, + { + "epoch": 0.0505, + "grad_norm": 0.7133657320392227, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 5050 + }, + { + "epoch": 0.05051, + "grad_norm": 0.8879083064557955, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 5051 + }, + { + "epoch": 0.05052, + "grad_norm": 1.108020136402945, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 5052 + }, + { + "epoch": 0.05053, + "grad_norm": 1.124376607679852, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 5053 + }, + { + "epoch": 0.05054, + "grad_norm": 0.8613572741956879, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 5054 + }, + { + "epoch": 0.05055, + "grad_norm": 0.9007503555201047, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 5055 + }, + { + "epoch": 0.05056, + "grad_norm": 0.9020326994115863, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5056 + }, + { + "epoch": 0.05057, + "grad_norm": 0.9194629842951593, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 5057 + }, + { + "epoch": 0.05058, + "grad_norm": 0.8058592428411899, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 5058 + }, + { + "epoch": 0.05059, + "grad_norm": 0.705098811845147, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 5059 + }, + { + "epoch": 0.0506, + "grad_norm": 0.6716192520354848, + "learning_rate": 0.003, + "loss": 4.1509, + "step": 5060 + }, + { + "epoch": 0.05061, + "grad_norm": 0.7122976459366105, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 5061 + }, + { + "epoch": 0.05062, + "grad_norm": 0.8458397806323464, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5062 + }, + { + "epoch": 0.05063, + "grad_norm": 0.8509332439877665, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 5063 + }, + { + "epoch": 0.05064, + "grad_norm": 0.8628499996665681, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 5064 + }, + { + "epoch": 0.05065, + "grad_norm": 0.7651349441929993, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5065 + }, + { + "epoch": 0.05066, + "grad_norm": 0.6441373019881442, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 5066 + }, + { + "epoch": 0.05067, + "grad_norm": 0.6162135983622039, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5067 + }, + { + "epoch": 0.05068, + "grad_norm": 0.5839874533063438, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 5068 + }, + { + "epoch": 0.05069, + "grad_norm": 0.602540955663345, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 5069 + }, + { + "epoch": 0.0507, + "grad_norm": 0.6829292685811817, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 5070 + }, + { + "epoch": 0.05071, + "grad_norm": 0.9040694995935689, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 5071 + }, + { + "epoch": 0.05072, + "grad_norm": 1.1363622378094762, + "learning_rate": 0.003, + "loss": 4.126, + "step": 5072 + }, + { + "epoch": 0.05073, + "grad_norm": 0.6884937706217313, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 5073 + }, + { + "epoch": 0.05074, + "grad_norm": 0.6473518308046033, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 5074 + }, + { + "epoch": 0.05075, + "grad_norm": 0.7765044046650167, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5075 + }, + { + "epoch": 0.05076, + "grad_norm": 0.8329788470025239, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 5076 + }, + { + "epoch": 0.05077, + "grad_norm": 0.799881313047519, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 5077 + }, + { + "epoch": 0.05078, + "grad_norm": 0.7389949572586678, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 5078 + }, + { + "epoch": 0.05079, + "grad_norm": 0.7330687471577605, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 5079 + }, + { + "epoch": 0.0508, + "grad_norm": 0.7072208630775181, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 5080 + }, + { + "epoch": 0.05081, + "grad_norm": 0.698914398304387, + "learning_rate": 0.003, + "loss": 4.135, + "step": 5081 + }, + { + "epoch": 0.05082, + "grad_norm": 0.6278523534587991, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 5082 + }, + { + "epoch": 0.05083, + "grad_norm": 0.5773478170755377, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 5083 + }, + { + "epoch": 0.05084, + "grad_norm": 0.4995734535669917, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 5084 + }, + { + "epoch": 0.05085, + "grad_norm": 0.503192190879057, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 5085 + }, + { + "epoch": 0.05086, + "grad_norm": 0.4849750033335867, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 5086 + }, + { + "epoch": 0.05087, + "grad_norm": 0.5401918808221647, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 5087 + }, + { + "epoch": 0.05088, + "grad_norm": 0.5431450812753981, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5088 + }, + { + "epoch": 0.05089, + "grad_norm": 0.5955708352853926, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5089 + }, + { + "epoch": 0.0509, + "grad_norm": 0.6944142097249647, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5090 + }, + { + "epoch": 0.05091, + "grad_norm": 0.7093523328553718, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 5091 + }, + { + "epoch": 0.05092, + "grad_norm": 0.711564968018302, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 5092 + }, + { + "epoch": 0.05093, + "grad_norm": 0.7293388159366392, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 5093 + }, + { + "epoch": 0.05094, + "grad_norm": 0.7789366794435865, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 5094 + }, + { + "epoch": 0.05095, + "grad_norm": 0.8943150130903914, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 5095 + }, + { + "epoch": 0.05096, + "grad_norm": 1.2235914801717511, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5096 + }, + { + "epoch": 0.05097, + "grad_norm": 0.8627777574244729, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5097 + }, + { + "epoch": 0.05098, + "grad_norm": 0.7848585017002957, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5098 + }, + { + "epoch": 0.05099, + "grad_norm": 0.7684212373548842, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 5099 + }, + { + "epoch": 0.051, + "grad_norm": 0.7191076738646075, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 5100 + }, + { + "epoch": 0.05101, + "grad_norm": 0.7450408749289968, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 5101 + }, + { + "epoch": 0.05102, + "grad_norm": 0.7788766932633848, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 5102 + }, + { + "epoch": 0.05103, + "grad_norm": 0.7398214547540392, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 5103 + }, + { + "epoch": 0.05104, + "grad_norm": 0.6558233797113744, + "learning_rate": 0.003, + "loss": 4.112, + "step": 5104 + }, + { + "epoch": 0.05105, + "grad_norm": 0.6381200092203351, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 5105 + }, + { + "epoch": 0.05106, + "grad_norm": 0.6828637852428295, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 5106 + }, + { + "epoch": 0.05107, + "grad_norm": 0.8597716654991152, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5107 + }, + { + "epoch": 0.05108, + "grad_norm": 1.2317482679596312, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 5108 + }, + { + "epoch": 0.05109, + "grad_norm": 0.8665302105620079, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5109 + }, + { + "epoch": 0.0511, + "grad_norm": 0.7969460956461537, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5110 + }, + { + "epoch": 0.05111, + "grad_norm": 0.8007840700796149, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 5111 + }, + { + "epoch": 0.05112, + "grad_norm": 0.7681571978199957, + "learning_rate": 0.003, + "loss": 4.14, + "step": 5112 + }, + { + "epoch": 0.05113, + "grad_norm": 0.6903861079813164, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 5113 + }, + { + "epoch": 0.05114, + "grad_norm": 0.7692348540414872, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 5114 + }, + { + "epoch": 0.05115, + "grad_norm": 0.7538155044071809, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 5115 + }, + { + "epoch": 0.05116, + "grad_norm": 0.7449617507426948, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 5116 + }, + { + "epoch": 0.05117, + "grad_norm": 0.8510379134985516, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 5117 + }, + { + "epoch": 0.05118, + "grad_norm": 1.005098903517839, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 5118 + }, + { + "epoch": 0.05119, + "grad_norm": 1.0115097346662045, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5119 + }, + { + "epoch": 0.0512, + "grad_norm": 0.9811018804142781, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 5120 + }, + { + "epoch": 0.05121, + "grad_norm": 0.9389084532957305, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5121 + }, + { + "epoch": 0.05122, + "grad_norm": 0.9691145521379205, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 5122 + }, + { + "epoch": 0.05123, + "grad_norm": 1.0049713917715095, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 5123 + }, + { + "epoch": 0.05124, + "grad_norm": 1.0419381303567035, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 5124 + }, + { + "epoch": 0.05125, + "grad_norm": 1.0446044411819475, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 5125 + }, + { + "epoch": 0.05126, + "grad_norm": 0.9019262378070061, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5126 + }, + { + "epoch": 0.05127, + "grad_norm": 0.8105987689505292, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 5127 + }, + { + "epoch": 0.05128, + "grad_norm": 0.7905970287718124, + "learning_rate": 0.003, + "loss": 4.1648, + "step": 5128 + }, + { + "epoch": 0.05129, + "grad_norm": 0.7417838905848333, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5129 + }, + { + "epoch": 0.0513, + "grad_norm": 0.7900814171820785, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 5130 + }, + { + "epoch": 0.05131, + "grad_norm": 0.9133150183860361, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 5131 + }, + { + "epoch": 0.05132, + "grad_norm": 1.095265924476809, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 5132 + }, + { + "epoch": 0.05133, + "grad_norm": 0.9471945091609965, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 5133 + }, + { + "epoch": 0.05134, + "grad_norm": 0.9244905428462661, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 5134 + }, + { + "epoch": 0.05135, + "grad_norm": 0.8334274421733712, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5135 + }, + { + "epoch": 0.05136, + "grad_norm": 0.709427559475487, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 5136 + }, + { + "epoch": 0.05137, + "grad_norm": 0.6442219169362268, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 5137 + }, + { + "epoch": 0.05138, + "grad_norm": 0.6467988163207677, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5138 + }, + { + "epoch": 0.05139, + "grad_norm": 0.733478794767403, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5139 + }, + { + "epoch": 0.0514, + "grad_norm": 0.7297546736273143, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 5140 + }, + { + "epoch": 0.05141, + "grad_norm": 0.6341213802128426, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 5141 + }, + { + "epoch": 0.05142, + "grad_norm": 0.5776293822275287, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 5142 + }, + { + "epoch": 0.05143, + "grad_norm": 0.5773363793450792, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5143 + }, + { + "epoch": 0.05144, + "grad_norm": 0.5526645311406644, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 5144 + }, + { + "epoch": 0.05145, + "grad_norm": 0.4890054978586333, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 5145 + }, + { + "epoch": 0.05146, + "grad_norm": 0.4190024922610593, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 5146 + }, + { + "epoch": 0.05147, + "grad_norm": 0.44043856065420567, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 5147 + }, + { + "epoch": 0.05148, + "grad_norm": 0.45582477546885664, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 5148 + }, + { + "epoch": 0.05149, + "grad_norm": 0.5300055959379171, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5149 + }, + { + "epoch": 0.0515, + "grad_norm": 0.5337764008665007, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5150 + }, + { + "epoch": 0.05151, + "grad_norm": 0.4777932939886135, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5151 + }, + { + "epoch": 0.05152, + "grad_norm": 0.4822263308806615, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 5152 + }, + { + "epoch": 0.05153, + "grad_norm": 0.5781114808382272, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 5153 + }, + { + "epoch": 0.05154, + "grad_norm": 0.7109250842959455, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 5154 + }, + { + "epoch": 0.05155, + "grad_norm": 0.9418503203747088, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 5155 + }, + { + "epoch": 0.05156, + "grad_norm": 1.2647418011213238, + "learning_rate": 0.003, + "loss": 4.1714, + "step": 5156 + }, + { + "epoch": 0.05157, + "grad_norm": 0.7308469531008109, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 5157 + }, + { + "epoch": 0.05158, + "grad_norm": 0.7875229543444077, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5158 + }, + { + "epoch": 0.05159, + "grad_norm": 0.8140638394753378, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 5159 + }, + { + "epoch": 0.0516, + "grad_norm": 0.8244642815731201, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5160 + }, + { + "epoch": 0.05161, + "grad_norm": 0.8540456002198428, + "learning_rate": 0.003, + "loss": 4.089, + "step": 5161 + }, + { + "epoch": 0.05162, + "grad_norm": 0.8746074474591011, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5162 + }, + { + "epoch": 0.05163, + "grad_norm": 0.8737342817314073, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 5163 + }, + { + "epoch": 0.05164, + "grad_norm": 1.0627737585477899, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5164 + }, + { + "epoch": 0.05165, + "grad_norm": 0.9694679199622456, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 5165 + }, + { + "epoch": 0.05166, + "grad_norm": 0.9344677027950545, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 5166 + }, + { + "epoch": 0.05167, + "grad_norm": 0.7891486782935601, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 5167 + }, + { + "epoch": 0.05168, + "grad_norm": 0.7729700436459497, + "learning_rate": 0.003, + "loss": 4.1763, + "step": 5168 + }, + { + "epoch": 0.05169, + "grad_norm": 0.7768744996829114, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 5169 + }, + { + "epoch": 0.0517, + "grad_norm": 0.9172514868358831, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 5170 + }, + { + "epoch": 0.05171, + "grad_norm": 0.987191990102733, + "learning_rate": 0.003, + "loss": 4.14, + "step": 5171 + }, + { + "epoch": 0.05172, + "grad_norm": 1.1074558271172996, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 5172 + }, + { + "epoch": 0.05173, + "grad_norm": 0.9112090797136994, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 5173 + }, + { + "epoch": 0.05174, + "grad_norm": 0.8508573888938128, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 5174 + }, + { + "epoch": 0.05175, + "grad_norm": 0.9226457815827247, + "learning_rate": 0.003, + "loss": 4.168, + "step": 5175 + }, + { + "epoch": 0.05176, + "grad_norm": 0.9401989711162229, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5176 + }, + { + "epoch": 0.05177, + "grad_norm": 1.1169814516735364, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5177 + }, + { + "epoch": 0.05178, + "grad_norm": 0.8757415219342446, + "learning_rate": 0.003, + "loss": 4.125, + "step": 5178 + }, + { + "epoch": 0.05179, + "grad_norm": 0.8953548399429396, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 5179 + }, + { + "epoch": 0.0518, + "grad_norm": 0.8270061517274871, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 5180 + }, + { + "epoch": 0.05181, + "grad_norm": 0.7996635692475278, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 5181 + }, + { + "epoch": 0.05182, + "grad_norm": 0.788819919508774, + "learning_rate": 0.003, + "loss": 4.1624, + "step": 5182 + }, + { + "epoch": 0.05183, + "grad_norm": 0.7520155289113659, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5183 + }, + { + "epoch": 0.05184, + "grad_norm": 0.7236732746268649, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 5184 + }, + { + "epoch": 0.05185, + "grad_norm": 0.638115349783052, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 5185 + }, + { + "epoch": 0.05186, + "grad_norm": 0.5723057160378794, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 5186 + }, + { + "epoch": 0.05187, + "grad_norm": 0.6462132298250771, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 5187 + }, + { + "epoch": 0.05188, + "grad_norm": 0.7054920468123554, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 5188 + }, + { + "epoch": 0.05189, + "grad_norm": 0.8179909749537473, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5189 + }, + { + "epoch": 0.0519, + "grad_norm": 0.9666616633136137, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 5190 + }, + { + "epoch": 0.05191, + "grad_norm": 0.8948010215701201, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 5191 + }, + { + "epoch": 0.05192, + "grad_norm": 0.7272065263730562, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 5192 + }, + { + "epoch": 0.05193, + "grad_norm": 0.7326420198310921, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5193 + }, + { + "epoch": 0.05194, + "grad_norm": 0.7400163904092779, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 5194 + }, + { + "epoch": 0.05195, + "grad_norm": 0.6229710454294611, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 5195 + }, + { + "epoch": 0.05196, + "grad_norm": 0.5732082487024617, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5196 + }, + { + "epoch": 0.05197, + "grad_norm": 0.5437048621681219, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5197 + }, + { + "epoch": 0.05198, + "grad_norm": 0.5106768440457569, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 5198 + }, + { + "epoch": 0.05199, + "grad_norm": 0.46246539222179384, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 5199 + }, + { + "epoch": 0.052, + "grad_norm": 0.48176692281015543, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 5200 + }, + { + "epoch": 0.05201, + "grad_norm": 0.4828183154586405, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 5201 + }, + { + "epoch": 0.05202, + "grad_norm": 0.5257398649757856, + "learning_rate": 0.003, + "loss": 4.103, + "step": 5202 + }, + { + "epoch": 0.05203, + "grad_norm": 0.5163132111756201, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 5203 + }, + { + "epoch": 0.05204, + "grad_norm": 0.47004258100640817, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5204 + }, + { + "epoch": 0.05205, + "grad_norm": 0.5080321067769532, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5205 + }, + { + "epoch": 0.05206, + "grad_norm": 0.6713429412709093, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 5206 + }, + { + "epoch": 0.05207, + "grad_norm": 0.9187866373124713, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 5207 + }, + { + "epoch": 0.05208, + "grad_norm": 1.2677728854398744, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5208 + }, + { + "epoch": 0.05209, + "grad_norm": 0.630859673605965, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 5209 + }, + { + "epoch": 0.0521, + "grad_norm": 0.7350336486110594, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 5210 + }, + { + "epoch": 0.05211, + "grad_norm": 0.8202119917312024, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 5211 + }, + { + "epoch": 0.05212, + "grad_norm": 0.6430476198919388, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 5212 + }, + { + "epoch": 0.05213, + "grad_norm": 0.6006713399050833, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5213 + }, + { + "epoch": 0.05214, + "grad_norm": 0.6412056440427675, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5214 + }, + { + "epoch": 0.05215, + "grad_norm": 0.7406745452916796, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5215 + }, + { + "epoch": 0.05216, + "grad_norm": 0.7338642035681033, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5216 + }, + { + "epoch": 0.05217, + "grad_norm": 0.7548188627814898, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 5217 + }, + { + "epoch": 0.05218, + "grad_norm": 0.7790433320975734, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 5218 + }, + { + "epoch": 0.05219, + "grad_norm": 0.921132206986461, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 5219 + }, + { + "epoch": 0.0522, + "grad_norm": 1.0167153212263549, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 5220 + }, + { + "epoch": 0.05221, + "grad_norm": 1.0121309018306264, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5221 + }, + { + "epoch": 0.05222, + "grad_norm": 0.907628952065032, + "learning_rate": 0.003, + "loss": 4.137, + "step": 5222 + }, + { + "epoch": 0.05223, + "grad_norm": 0.919506193257904, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 5223 + }, + { + "epoch": 0.05224, + "grad_norm": 1.1506026821508868, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 5224 + }, + { + "epoch": 0.05225, + "grad_norm": 0.899823660266816, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 5225 + }, + { + "epoch": 0.05226, + "grad_norm": 0.8527078083046193, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 5226 + }, + { + "epoch": 0.05227, + "grad_norm": 0.7079246341060081, + "learning_rate": 0.003, + "loss": 4.165, + "step": 5227 + }, + { + "epoch": 0.05228, + "grad_norm": 0.8044831433064553, + "learning_rate": 0.003, + "loss": 4.145, + "step": 5228 + }, + { + "epoch": 0.05229, + "grad_norm": 0.8720881208580172, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 5229 + }, + { + "epoch": 0.0523, + "grad_norm": 0.8795412540846314, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 5230 + }, + { + "epoch": 0.05231, + "grad_norm": 0.7917348425475053, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 5231 + }, + { + "epoch": 0.05232, + "grad_norm": 0.752556657478349, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 5232 + }, + { + "epoch": 0.05233, + "grad_norm": 0.6869072276302872, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5233 + }, + { + "epoch": 0.05234, + "grad_norm": 0.7324907875751323, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 5234 + }, + { + "epoch": 0.05235, + "grad_norm": 0.721046302164716, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 5235 + }, + { + "epoch": 0.05236, + "grad_norm": 0.575714162268126, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 5236 + }, + { + "epoch": 0.05237, + "grad_norm": 0.516817288215746, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 5237 + }, + { + "epoch": 0.05238, + "grad_norm": 0.6049328985205596, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5238 + }, + { + "epoch": 0.05239, + "grad_norm": 0.645411189054774, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 5239 + }, + { + "epoch": 0.0524, + "grad_norm": 0.7516501855794593, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5240 + }, + { + "epoch": 0.05241, + "grad_norm": 0.8580499313970857, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5241 + }, + { + "epoch": 0.05242, + "grad_norm": 0.8683371010488704, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 5242 + }, + { + "epoch": 0.05243, + "grad_norm": 1.0371281140802466, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5243 + }, + { + "epoch": 0.05244, + "grad_norm": 1.0230759917025969, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5244 + }, + { + "epoch": 0.05245, + "grad_norm": 0.8780400373348358, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 5245 + }, + { + "epoch": 0.05246, + "grad_norm": 0.8613363719642216, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5246 + }, + { + "epoch": 0.05247, + "grad_norm": 0.7710466126666472, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 5247 + }, + { + "epoch": 0.05248, + "grad_norm": 0.7486554824013956, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5248 + }, + { + "epoch": 0.05249, + "grad_norm": 0.7704795977105405, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 5249 + }, + { + "epoch": 0.0525, + "grad_norm": 0.7893461765166555, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 5250 + }, + { + "epoch": 0.05251, + "grad_norm": 0.7489202376303653, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 5251 + }, + { + "epoch": 0.05252, + "grad_norm": 0.6609520253229195, + "learning_rate": 0.003, + "loss": 4.148, + "step": 5252 + }, + { + "epoch": 0.05253, + "grad_norm": 0.719719165362111, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 5253 + }, + { + "epoch": 0.05254, + "grad_norm": 0.6856836855692663, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5254 + }, + { + "epoch": 0.05255, + "grad_norm": 0.6331973623044866, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 5255 + }, + { + "epoch": 0.05256, + "grad_norm": 0.6209166464347697, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5256 + }, + { + "epoch": 0.05257, + "grad_norm": 0.6703957390397629, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 5257 + }, + { + "epoch": 0.05258, + "grad_norm": 0.8005691943001225, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5258 + }, + { + "epoch": 0.05259, + "grad_norm": 1.178008398076746, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5259 + }, + { + "epoch": 0.0526, + "grad_norm": 0.9009691862241965, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 5260 + }, + { + "epoch": 0.05261, + "grad_norm": 0.636909529940799, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 5261 + }, + { + "epoch": 0.05262, + "grad_norm": 0.7583190817879178, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 5262 + }, + { + "epoch": 0.05263, + "grad_norm": 0.7940859742603288, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 5263 + }, + { + "epoch": 0.05264, + "grad_norm": 0.7387297524848515, + "learning_rate": 0.003, + "loss": 4.148, + "step": 5264 + }, + { + "epoch": 0.05265, + "grad_norm": 0.6381909927316334, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 5265 + }, + { + "epoch": 0.05266, + "grad_norm": 0.6211413466674444, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 5266 + }, + { + "epoch": 0.05267, + "grad_norm": 0.6304792137219847, + "learning_rate": 0.003, + "loss": 4.142, + "step": 5267 + }, + { + "epoch": 0.05268, + "grad_norm": 0.7489765141892385, + "learning_rate": 0.003, + "loss": 4.111, + "step": 5268 + }, + { + "epoch": 0.05269, + "grad_norm": 0.7861495121827866, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 5269 + }, + { + "epoch": 0.0527, + "grad_norm": 0.6974038599114732, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 5270 + }, + { + "epoch": 0.05271, + "grad_norm": 0.664530980238474, + "learning_rate": 0.003, + "loss": 4.085, + "step": 5271 + }, + { + "epoch": 0.05272, + "grad_norm": 0.7520422410793401, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5272 + }, + { + "epoch": 0.05273, + "grad_norm": 0.7793196930797887, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 5273 + }, + { + "epoch": 0.05274, + "grad_norm": 0.754818417244099, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 5274 + }, + { + "epoch": 0.05275, + "grad_norm": 0.6242567287914901, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5275 + }, + { + "epoch": 0.05276, + "grad_norm": 0.6899528691289879, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 5276 + }, + { + "epoch": 0.05277, + "grad_norm": 0.7574326268248618, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 5277 + }, + { + "epoch": 0.05278, + "grad_norm": 0.8507468220046873, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5278 + }, + { + "epoch": 0.05279, + "grad_norm": 1.0353055112068175, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 5279 + }, + { + "epoch": 0.0528, + "grad_norm": 0.9033296737516481, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 5280 + }, + { + "epoch": 0.05281, + "grad_norm": 0.7963839547721746, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 5281 + }, + { + "epoch": 0.05282, + "grad_norm": 0.7877787577000847, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 5282 + }, + { + "epoch": 0.05283, + "grad_norm": 0.7846974861191656, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5283 + }, + { + "epoch": 0.05284, + "grad_norm": 0.7458279107143788, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 5284 + }, + { + "epoch": 0.05285, + "grad_norm": 0.7229333553728315, + "learning_rate": 0.003, + "loss": 4.089, + "step": 5285 + }, + { + "epoch": 0.05286, + "grad_norm": 0.7184374854077129, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 5286 + }, + { + "epoch": 0.05287, + "grad_norm": 0.8297178227719375, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5287 + }, + { + "epoch": 0.05288, + "grad_norm": 0.7621123902466413, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5288 + }, + { + "epoch": 0.05289, + "grad_norm": 0.6339905585597592, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5289 + }, + { + "epoch": 0.0529, + "grad_norm": 0.6327911322808994, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5290 + }, + { + "epoch": 0.05291, + "grad_norm": 0.6026020302477917, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 5291 + }, + { + "epoch": 0.05292, + "grad_norm": 0.6437340105854593, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 5292 + }, + { + "epoch": 0.05293, + "grad_norm": 0.6314109214308004, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5293 + }, + { + "epoch": 0.05294, + "grad_norm": 0.695308432779079, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 5294 + }, + { + "epoch": 0.05295, + "grad_norm": 0.7787212860512488, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5295 + }, + { + "epoch": 0.05296, + "grad_norm": 0.8622976689690274, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 5296 + }, + { + "epoch": 0.05297, + "grad_norm": 1.0166098716112735, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 5297 + }, + { + "epoch": 0.05298, + "grad_norm": 1.0869357162009616, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 5298 + }, + { + "epoch": 0.05299, + "grad_norm": 0.885010326633525, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 5299 + }, + { + "epoch": 0.053, + "grad_norm": 0.9615808194178252, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 5300 + }, + { + "epoch": 0.05301, + "grad_norm": 0.938352188140003, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 5301 + }, + { + "epoch": 0.05302, + "grad_norm": 1.0040646683623418, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 5302 + }, + { + "epoch": 0.05303, + "grad_norm": 0.8200261070086993, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 5303 + }, + { + "epoch": 0.05304, + "grad_norm": 0.8306683481132952, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 5304 + }, + { + "epoch": 0.05305, + "grad_norm": 0.835451021401934, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5305 + }, + { + "epoch": 0.05306, + "grad_norm": 0.999299771292447, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 5306 + }, + { + "epoch": 0.05307, + "grad_norm": 1.1575629558742315, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 5307 + }, + { + "epoch": 0.05308, + "grad_norm": 0.8821432895719332, + "learning_rate": 0.003, + "loss": 4.1606, + "step": 5308 + }, + { + "epoch": 0.05309, + "grad_norm": 0.7478897292069714, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 5309 + }, + { + "epoch": 0.0531, + "grad_norm": 0.6757983881370974, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 5310 + }, + { + "epoch": 0.05311, + "grad_norm": 0.734265220677749, + "learning_rate": 0.003, + "loss": 4.165, + "step": 5311 + }, + { + "epoch": 0.05312, + "grad_norm": 0.7403158021925053, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5312 + }, + { + "epoch": 0.05313, + "grad_norm": 0.7001445193969814, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 5313 + }, + { + "epoch": 0.05314, + "grad_norm": 0.6690291331175298, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 5314 + }, + { + "epoch": 0.05315, + "grad_norm": 0.692987823884034, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5315 + }, + { + "epoch": 0.05316, + "grad_norm": 0.6767422015060799, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 5316 + }, + { + "epoch": 0.05317, + "grad_norm": 0.6288130150337862, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 5317 + }, + { + "epoch": 0.05318, + "grad_norm": 0.6126535331811985, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 5318 + }, + { + "epoch": 0.05319, + "grad_norm": 0.6921182533464885, + "learning_rate": 0.003, + "loss": 4.107, + "step": 5319 + }, + { + "epoch": 0.0532, + "grad_norm": 0.8654384133463597, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 5320 + }, + { + "epoch": 0.05321, + "grad_norm": 1.33574080453646, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5321 + }, + { + "epoch": 0.05322, + "grad_norm": 0.7268715200875471, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 5322 + }, + { + "epoch": 0.05323, + "grad_norm": 0.8318376363678193, + "learning_rate": 0.003, + "loss": 4.1707, + "step": 5323 + }, + { + "epoch": 0.05324, + "grad_norm": 0.9201296394107168, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 5324 + }, + { + "epoch": 0.05325, + "grad_norm": 0.9283907989830872, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 5325 + }, + { + "epoch": 0.05326, + "grad_norm": 0.9049955824604045, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 5326 + }, + { + "epoch": 0.05327, + "grad_norm": 0.8645534067837618, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 5327 + }, + { + "epoch": 0.05328, + "grad_norm": 0.7530167384055072, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 5328 + }, + { + "epoch": 0.05329, + "grad_norm": 0.7611151412670385, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5329 + }, + { + "epoch": 0.0533, + "grad_norm": 0.6706563948860642, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 5330 + }, + { + "epoch": 0.05331, + "grad_norm": 0.7287872383125151, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 5331 + }, + { + "epoch": 0.05332, + "grad_norm": 0.6990347485461705, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5332 + }, + { + "epoch": 0.05333, + "grad_norm": 0.7153797786309389, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 5333 + }, + { + "epoch": 0.05334, + "grad_norm": 0.7135824620276469, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 5334 + }, + { + "epoch": 0.05335, + "grad_norm": 0.7329026729817923, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 5335 + }, + { + "epoch": 0.05336, + "grad_norm": 0.7489264419035738, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 5336 + }, + { + "epoch": 0.05337, + "grad_norm": 0.7598169235752914, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 5337 + }, + { + "epoch": 0.05338, + "grad_norm": 0.7398262818382618, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 5338 + }, + { + "epoch": 0.05339, + "grad_norm": 0.7503534133517454, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 5339 + }, + { + "epoch": 0.0534, + "grad_norm": 0.729974918134158, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 5340 + }, + { + "epoch": 0.05341, + "grad_norm": 0.6588879077709251, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 5341 + }, + { + "epoch": 0.05342, + "grad_norm": 0.7419685883521477, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 5342 + }, + { + "epoch": 0.05343, + "grad_norm": 0.7435550390936541, + "learning_rate": 0.003, + "loss": 4.134, + "step": 5343 + }, + { + "epoch": 0.05344, + "grad_norm": 0.7863852626600986, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 5344 + }, + { + "epoch": 0.05345, + "grad_norm": 0.7422447513853418, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 5345 + }, + { + "epoch": 0.05346, + "grad_norm": 0.6629198259774158, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 5346 + }, + { + "epoch": 0.05347, + "grad_norm": 0.6664361566776935, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5347 + }, + { + "epoch": 0.05348, + "grad_norm": 0.7259069460111056, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 5348 + }, + { + "epoch": 0.05349, + "grad_norm": 0.8960329268103234, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 5349 + }, + { + "epoch": 0.0535, + "grad_norm": 1.2972339587908424, + "learning_rate": 0.003, + "loss": 4.127, + "step": 5350 + }, + { + "epoch": 0.05351, + "grad_norm": 0.8742838225263962, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 5351 + }, + { + "epoch": 0.05352, + "grad_norm": 0.7189234033594293, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5352 + }, + { + "epoch": 0.05353, + "grad_norm": 0.8249722357305902, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5353 + }, + { + "epoch": 0.05354, + "grad_norm": 0.9093171506347467, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5354 + }, + { + "epoch": 0.05355, + "grad_norm": 0.9642691428952453, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5355 + }, + { + "epoch": 0.05356, + "grad_norm": 1.0666591729077117, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 5356 + }, + { + "epoch": 0.05357, + "grad_norm": 1.1481513734933346, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 5357 + }, + { + "epoch": 0.05358, + "grad_norm": 0.8272666359467848, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5358 + }, + { + "epoch": 0.05359, + "grad_norm": 0.7172771084344366, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5359 + }, + { + "epoch": 0.0536, + "grad_norm": 0.6651106652374006, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 5360 + }, + { + "epoch": 0.05361, + "grad_norm": 0.7099023505414537, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5361 + }, + { + "epoch": 0.05362, + "grad_norm": 0.6814935799785445, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5362 + }, + { + "epoch": 0.05363, + "grad_norm": 0.7116843146393923, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 5363 + }, + { + "epoch": 0.05364, + "grad_norm": 0.8468291440238759, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5364 + }, + { + "epoch": 0.05365, + "grad_norm": 0.9232590055754448, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 5365 + }, + { + "epoch": 0.05366, + "grad_norm": 0.9207798580705393, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5366 + }, + { + "epoch": 0.05367, + "grad_norm": 0.903400216800342, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 5367 + }, + { + "epoch": 0.05368, + "grad_norm": 0.8164520073516316, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 5368 + }, + { + "epoch": 0.05369, + "grad_norm": 0.7772994418518063, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 5369 + }, + { + "epoch": 0.0537, + "grad_norm": 0.880991172920623, + "learning_rate": 0.003, + "loss": 4.1351, + "step": 5370 + }, + { + "epoch": 0.05371, + "grad_norm": 0.8578371500269605, + "learning_rate": 0.003, + "loss": 4.152, + "step": 5371 + }, + { + "epoch": 0.05372, + "grad_norm": 0.8581893520081385, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5372 + }, + { + "epoch": 0.05373, + "grad_norm": 0.981605807782507, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 5373 + }, + { + "epoch": 0.05374, + "grad_norm": 1.0221056617839879, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 5374 + }, + { + "epoch": 0.05375, + "grad_norm": 0.8349259297399914, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5375 + }, + { + "epoch": 0.05376, + "grad_norm": 0.6337577803423684, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5376 + }, + { + "epoch": 0.05377, + "grad_norm": 0.6889832338024495, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5377 + }, + { + "epoch": 0.05378, + "grad_norm": 0.7812861326156888, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 5378 + }, + { + "epoch": 0.05379, + "grad_norm": 0.8471373099098546, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 5379 + }, + { + "epoch": 0.0538, + "grad_norm": 0.8235164446325877, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 5380 + }, + { + "epoch": 0.05381, + "grad_norm": 0.7191229707868476, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 5381 + }, + { + "epoch": 0.05382, + "grad_norm": 0.6499844207771936, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 5382 + }, + { + "epoch": 0.05383, + "grad_norm": 0.6556993226066228, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 5383 + }, + { + "epoch": 0.05384, + "grad_norm": 0.6563361402381379, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 5384 + }, + { + "epoch": 0.05385, + "grad_norm": 0.7347199963990706, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 5385 + }, + { + "epoch": 0.05386, + "grad_norm": 0.7386855978909174, + "learning_rate": 0.003, + "loss": 4.123, + "step": 5386 + }, + { + "epoch": 0.05387, + "grad_norm": 0.6950422799141344, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 5387 + }, + { + "epoch": 0.05388, + "grad_norm": 0.5813469350124647, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 5388 + }, + { + "epoch": 0.05389, + "grad_norm": 0.5663689684535403, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 5389 + }, + { + "epoch": 0.0539, + "grad_norm": 0.5465150881145437, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 5390 + }, + { + "epoch": 0.05391, + "grad_norm": 0.6133981789080953, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 5391 + }, + { + "epoch": 0.05392, + "grad_norm": 0.6736582173461655, + "learning_rate": 0.003, + "loss": 4.108, + "step": 5392 + }, + { + "epoch": 0.05393, + "grad_norm": 0.7301761943101791, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 5393 + }, + { + "epoch": 0.05394, + "grad_norm": 0.8165543355372945, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 5394 + }, + { + "epoch": 0.05395, + "grad_norm": 0.9718965866957342, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 5395 + }, + { + "epoch": 0.05396, + "grad_norm": 1.187963468121871, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 5396 + }, + { + "epoch": 0.05397, + "grad_norm": 0.8173506639861493, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5397 + }, + { + "epoch": 0.05398, + "grad_norm": 0.8358938356109366, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5398 + }, + { + "epoch": 0.05399, + "grad_norm": 1.0290513982103986, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 5399 + }, + { + "epoch": 0.054, + "grad_norm": 1.080858286517883, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 5400 + }, + { + "epoch": 0.05401, + "grad_norm": 0.9077257276299444, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 5401 + }, + { + "epoch": 0.05402, + "grad_norm": 1.0072208959923346, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 5402 + }, + { + "epoch": 0.05403, + "grad_norm": 1.018927241348422, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 5403 + }, + { + "epoch": 0.05404, + "grad_norm": 0.869754384415822, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 5404 + }, + { + "epoch": 0.05405, + "grad_norm": 0.8797782462066771, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 5405 + }, + { + "epoch": 0.05406, + "grad_norm": 0.9001689861982061, + "learning_rate": 0.003, + "loss": 4.122, + "step": 5406 + }, + { + "epoch": 0.05407, + "grad_norm": 0.8502046414935157, + "learning_rate": 0.003, + "loss": 4.136, + "step": 5407 + }, + { + "epoch": 0.05408, + "grad_norm": 0.7664799629174679, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 5408 + }, + { + "epoch": 0.05409, + "grad_norm": 0.6908041565692945, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5409 + }, + { + "epoch": 0.0541, + "grad_norm": 0.6610468895485156, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5410 + }, + { + "epoch": 0.05411, + "grad_norm": 0.7119400907145296, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 5411 + }, + { + "epoch": 0.05412, + "grad_norm": 0.663756122102936, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 5412 + }, + { + "epoch": 0.05413, + "grad_norm": 0.6934560877059226, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 5413 + }, + { + "epoch": 0.05414, + "grad_norm": 0.7346678872710226, + "learning_rate": 0.003, + "loss": 4.1, + "step": 5414 + }, + { + "epoch": 0.05415, + "grad_norm": 0.8484600669855001, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 5415 + }, + { + "epoch": 0.05416, + "grad_norm": 1.0151357887958885, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 5416 + }, + { + "epoch": 0.05417, + "grad_norm": 1.0367007805029973, + "learning_rate": 0.003, + "loss": 4.133, + "step": 5417 + }, + { + "epoch": 0.05418, + "grad_norm": 0.8644340293592367, + "learning_rate": 0.003, + "loss": 4.119, + "step": 5418 + }, + { + "epoch": 0.05419, + "grad_norm": 0.7954788907717502, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5419 + }, + { + "epoch": 0.0542, + "grad_norm": 0.6744403315171975, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5420 + }, + { + "epoch": 0.05421, + "grad_norm": 0.6512161506363956, + "learning_rate": 0.003, + "loss": 4.11, + "step": 5421 + }, + { + "epoch": 0.05422, + "grad_norm": 0.6012962342683661, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 5422 + }, + { + "epoch": 0.05423, + "grad_norm": 0.5672068462884734, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 5423 + }, + { + "epoch": 0.05424, + "grad_norm": 0.5505478147926042, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 5424 + }, + { + "epoch": 0.05425, + "grad_norm": 0.5656768972103682, + "learning_rate": 0.003, + "loss": 4.117, + "step": 5425 + }, + { + "epoch": 0.05426, + "grad_norm": 0.6139416554795246, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 5426 + }, + { + "epoch": 0.05427, + "grad_norm": 0.7235910686302464, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 5427 + }, + { + "epoch": 0.05428, + "grad_norm": 0.9324553754288041, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 5428 + }, + { + "epoch": 0.05429, + "grad_norm": 1.240086702831149, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 5429 + }, + { + "epoch": 0.0543, + "grad_norm": 0.6358810736408764, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 5430 + }, + { + "epoch": 0.05431, + "grad_norm": 0.6446879509210102, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5431 + }, + { + "epoch": 0.05432, + "grad_norm": 0.7579426004521167, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 5432 + }, + { + "epoch": 0.05433, + "grad_norm": 0.7180614229941668, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5433 + }, + { + "epoch": 0.05434, + "grad_norm": 0.7525090079920899, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 5434 + }, + { + "epoch": 0.05435, + "grad_norm": 0.7330233888201235, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 5435 + }, + { + "epoch": 0.05436, + "grad_norm": 0.6353200125329312, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5436 + }, + { + "epoch": 0.05437, + "grad_norm": 0.5773159842467662, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5437 + }, + { + "epoch": 0.05438, + "grad_norm": 0.5441102130695965, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 5438 + }, + { + "epoch": 0.05439, + "grad_norm": 0.5835331287048283, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 5439 + }, + { + "epoch": 0.0544, + "grad_norm": 0.60640408988456, + "learning_rate": 0.003, + "loss": 4.1, + "step": 5440 + }, + { + "epoch": 0.05441, + "grad_norm": 0.6241847646948565, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5441 + }, + { + "epoch": 0.05442, + "grad_norm": 0.6464314803977715, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 5442 + }, + { + "epoch": 0.05443, + "grad_norm": 0.7145704106013125, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 5443 + }, + { + "epoch": 0.05444, + "grad_norm": 0.9202361132826254, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 5444 + }, + { + "epoch": 0.05445, + "grad_norm": 1.0938082156986777, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 5445 + }, + { + "epoch": 0.05446, + "grad_norm": 0.938114872647745, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 5446 + }, + { + "epoch": 0.05447, + "grad_norm": 0.7770727627970194, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5447 + }, + { + "epoch": 0.05448, + "grad_norm": 0.9145478575365268, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 5448 + }, + { + "epoch": 0.05449, + "grad_norm": 1.2469972170687962, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 5449 + }, + { + "epoch": 0.0545, + "grad_norm": 0.9093855660849631, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 5450 + }, + { + "epoch": 0.05451, + "grad_norm": 0.8857139983218779, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 5451 + }, + { + "epoch": 0.05452, + "grad_norm": 0.8900426362781307, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 5452 + }, + { + "epoch": 0.05453, + "grad_norm": 0.8375394525311214, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 5453 + }, + { + "epoch": 0.05454, + "grad_norm": 0.8515818685266928, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 5454 + }, + { + "epoch": 0.05455, + "grad_norm": 0.8260003014827662, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 5455 + }, + { + "epoch": 0.05456, + "grad_norm": 0.7287425178770975, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5456 + }, + { + "epoch": 0.05457, + "grad_norm": 0.7579041524211054, + "learning_rate": 0.003, + "loss": 4.152, + "step": 5457 + }, + { + "epoch": 0.05458, + "grad_norm": 0.8277850916477079, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 5458 + }, + { + "epoch": 0.05459, + "grad_norm": 0.9693114941982469, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 5459 + }, + { + "epoch": 0.0546, + "grad_norm": 0.9765459017034847, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 5460 + }, + { + "epoch": 0.05461, + "grad_norm": 0.791633262589639, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 5461 + }, + { + "epoch": 0.05462, + "grad_norm": 0.8490495779181783, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 5462 + }, + { + "epoch": 0.05463, + "grad_norm": 0.9783106702125903, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5463 + }, + { + "epoch": 0.05464, + "grad_norm": 1.0312992529126999, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 5464 + }, + { + "epoch": 0.05465, + "grad_norm": 1.0898859521438415, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 5465 + }, + { + "epoch": 0.05466, + "grad_norm": 0.9011074511913453, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 5466 + }, + { + "epoch": 0.05467, + "grad_norm": 0.7944273383073479, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 5467 + }, + { + "epoch": 0.05468, + "grad_norm": 0.7557868646316354, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 5468 + }, + { + "epoch": 0.05469, + "grad_norm": 0.7157700961434887, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 5469 + }, + { + "epoch": 0.0547, + "grad_norm": 0.768865679961248, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5470 + }, + { + "epoch": 0.05471, + "grad_norm": 0.7253564868288621, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5471 + }, + { + "epoch": 0.05472, + "grad_norm": 0.6484270659754396, + "learning_rate": 0.003, + "loss": 4.117, + "step": 5472 + }, + { + "epoch": 0.05473, + "grad_norm": 0.6149159552987197, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 5473 + }, + { + "epoch": 0.05474, + "grad_norm": 0.6394638714309846, + "learning_rate": 0.003, + "loss": 4.109, + "step": 5474 + }, + { + "epoch": 0.05475, + "grad_norm": 0.7019738393808744, + "learning_rate": 0.003, + "loss": 4.1679, + "step": 5475 + }, + { + "epoch": 0.05476, + "grad_norm": 0.7528679854178247, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5476 + }, + { + "epoch": 0.05477, + "grad_norm": 0.8641692686702116, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 5477 + }, + { + "epoch": 0.05478, + "grad_norm": 0.9329968095336455, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5478 + }, + { + "epoch": 0.05479, + "grad_norm": 0.8530617715262148, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5479 + }, + { + "epoch": 0.0548, + "grad_norm": 0.6851831025467184, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5480 + }, + { + "epoch": 0.05481, + "grad_norm": 0.6300236876868771, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 5481 + }, + { + "epoch": 0.05482, + "grad_norm": 0.6745238952273943, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5482 + }, + { + "epoch": 0.05483, + "grad_norm": 0.6653313106296262, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 5483 + }, + { + "epoch": 0.05484, + "grad_norm": 0.7093577864302207, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 5484 + }, + { + "epoch": 0.05485, + "grad_norm": 0.7814477121541693, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 5485 + }, + { + "epoch": 0.05486, + "grad_norm": 0.8390976033521745, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 5486 + }, + { + "epoch": 0.05487, + "grad_norm": 0.761138743567861, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5487 + }, + { + "epoch": 0.05488, + "grad_norm": 0.7296061446756823, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 5488 + }, + { + "epoch": 0.05489, + "grad_norm": 0.7768186211594449, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 5489 + }, + { + "epoch": 0.0549, + "grad_norm": 0.6724947950772555, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 5490 + }, + { + "epoch": 0.05491, + "grad_norm": 0.6345273171856477, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 5491 + }, + { + "epoch": 0.05492, + "grad_norm": 0.6187943049842857, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 5492 + }, + { + "epoch": 0.05493, + "grad_norm": 0.736687820563599, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 5493 + }, + { + "epoch": 0.05494, + "grad_norm": 0.7510302372633021, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 5494 + }, + { + "epoch": 0.05495, + "grad_norm": 0.8712240659486696, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 5495 + }, + { + "epoch": 0.05496, + "grad_norm": 1.0341625719582008, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 5496 + }, + { + "epoch": 0.05497, + "grad_norm": 1.1451338915999956, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 5497 + }, + { + "epoch": 0.05498, + "grad_norm": 0.8431680462491885, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5498 + }, + { + "epoch": 0.05499, + "grad_norm": 0.816700771440371, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5499 + }, + { + "epoch": 0.055, + "grad_norm": 0.799679407985695, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 5500 + }, + { + "epoch": 0.05501, + "grad_norm": 0.9037921014823048, + "learning_rate": 0.003, + "loss": 4.123, + "step": 5501 + }, + { + "epoch": 0.05502, + "grad_norm": 1.0586237652595547, + "learning_rate": 0.003, + "loss": 4.12, + "step": 5502 + }, + { + "epoch": 0.05503, + "grad_norm": 1.031551443780613, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 5503 + }, + { + "epoch": 0.05504, + "grad_norm": 0.9058140828678377, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 5504 + }, + { + "epoch": 0.05505, + "grad_norm": 0.8558402590801321, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 5505 + }, + { + "epoch": 0.05506, + "grad_norm": 0.9182865122687004, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 5506 + }, + { + "epoch": 0.05507, + "grad_norm": 0.7933354119287489, + "learning_rate": 0.003, + "loss": 4.158, + "step": 5507 + }, + { + "epoch": 0.05508, + "grad_norm": 0.7272589705323323, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 5508 + }, + { + "epoch": 0.05509, + "grad_norm": 0.7161156271163123, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5509 + }, + { + "epoch": 0.0551, + "grad_norm": 0.6839604168952217, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 5510 + }, + { + "epoch": 0.05511, + "grad_norm": 0.6139416847259401, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 5511 + }, + { + "epoch": 0.05512, + "grad_norm": 0.5621071546906728, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 5512 + }, + { + "epoch": 0.05513, + "grad_norm": 0.6096185049093434, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5513 + }, + { + "epoch": 0.05514, + "grad_norm": 0.6336178682323431, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5514 + }, + { + "epoch": 0.05515, + "grad_norm": 0.7031400606321546, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 5515 + }, + { + "epoch": 0.05516, + "grad_norm": 0.7447410189527172, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 5516 + }, + { + "epoch": 0.05517, + "grad_norm": 0.7841931699582834, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 5517 + }, + { + "epoch": 0.05518, + "grad_norm": 0.8917972840062397, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5518 + }, + { + "epoch": 0.05519, + "grad_norm": 1.0113040291838757, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 5519 + }, + { + "epoch": 0.0552, + "grad_norm": 1.0623735541880837, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 5520 + }, + { + "epoch": 0.05521, + "grad_norm": 0.7945011952243602, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 5521 + }, + { + "epoch": 0.05522, + "grad_norm": 0.6236986309747409, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 5522 + }, + { + "epoch": 0.05523, + "grad_norm": 0.6915416985072932, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5523 + }, + { + "epoch": 0.05524, + "grad_norm": 0.7772328845162221, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 5524 + }, + { + "epoch": 0.05525, + "grad_norm": 0.8883412809335927, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5525 + }, + { + "epoch": 0.05526, + "grad_norm": 0.8086300513135104, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 5526 + }, + { + "epoch": 0.05527, + "grad_norm": 0.6933818876338046, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5527 + }, + { + "epoch": 0.05528, + "grad_norm": 0.7255857040307583, + "learning_rate": 0.003, + "loss": 4.125, + "step": 5528 + }, + { + "epoch": 0.05529, + "grad_norm": 0.7262305634108036, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 5529 + }, + { + "epoch": 0.0553, + "grad_norm": 0.6799075981942326, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5530 + }, + { + "epoch": 0.05531, + "grad_norm": 0.7288072066854816, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5531 + }, + { + "epoch": 0.05532, + "grad_norm": 0.7212131851938903, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 5532 + }, + { + "epoch": 0.05533, + "grad_norm": 0.7813370825497535, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 5533 + }, + { + "epoch": 0.05534, + "grad_norm": 0.7625719940297025, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 5534 + }, + { + "epoch": 0.05535, + "grad_norm": 0.8190303715949169, + "learning_rate": 0.003, + "loss": 4.139, + "step": 5535 + }, + { + "epoch": 0.05536, + "grad_norm": 0.8250338882829615, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 5536 + }, + { + "epoch": 0.05537, + "grad_norm": 0.8788954790980155, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 5537 + }, + { + "epoch": 0.05538, + "grad_norm": 0.9012442797992367, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5538 + }, + { + "epoch": 0.05539, + "grad_norm": 0.8122983939229457, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 5539 + }, + { + "epoch": 0.0554, + "grad_norm": 0.7598578272137851, + "learning_rate": 0.003, + "loss": 4.164, + "step": 5540 + }, + { + "epoch": 0.05541, + "grad_norm": 0.7025045153629659, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 5541 + }, + { + "epoch": 0.05542, + "grad_norm": 0.7249716315464921, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 5542 + }, + { + "epoch": 0.05543, + "grad_norm": 0.7653470879890454, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 5543 + }, + { + "epoch": 0.05544, + "grad_norm": 0.9152516311386858, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 5544 + }, + { + "epoch": 0.05545, + "grad_norm": 0.9264085795715772, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 5545 + }, + { + "epoch": 0.05546, + "grad_norm": 0.75980006084045, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 5546 + }, + { + "epoch": 0.05547, + "grad_norm": 0.7568411552111042, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 5547 + }, + { + "epoch": 0.05548, + "grad_norm": 0.8058342997945618, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5548 + }, + { + "epoch": 0.05549, + "grad_norm": 0.7052563470892956, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 5549 + }, + { + "epoch": 0.0555, + "grad_norm": 0.6538707744806174, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 5550 + }, + { + "epoch": 0.05551, + "grad_norm": 0.6754492446230133, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5551 + }, + { + "epoch": 0.05552, + "grad_norm": 0.7000361531636052, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 5552 + }, + { + "epoch": 0.05553, + "grad_norm": 0.7471036467384345, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5553 + }, + { + "epoch": 0.05554, + "grad_norm": 0.6519148066664208, + "learning_rate": 0.003, + "loss": 4.107, + "step": 5554 + }, + { + "epoch": 0.05555, + "grad_norm": 0.6843431614210626, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 5555 + }, + { + "epoch": 0.05556, + "grad_norm": 0.6534346041116017, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 5556 + }, + { + "epoch": 0.05557, + "grad_norm": 0.816868082356806, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5557 + }, + { + "epoch": 0.05558, + "grad_norm": 1.0929732507789733, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 5558 + }, + { + "epoch": 0.05559, + "grad_norm": 0.8960362084356669, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 5559 + }, + { + "epoch": 0.0556, + "grad_norm": 0.6973978644009409, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 5560 + }, + { + "epoch": 0.05561, + "grad_norm": 0.7611043817319628, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 5561 + }, + { + "epoch": 0.05562, + "grad_norm": 0.8608099588467866, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5562 + }, + { + "epoch": 0.05563, + "grad_norm": 0.9082804694980892, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 5563 + }, + { + "epoch": 0.05564, + "grad_norm": 0.9906908375188082, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5564 + }, + { + "epoch": 0.05565, + "grad_norm": 1.0732377297521, + "learning_rate": 0.003, + "loss": 4.1639, + "step": 5565 + }, + { + "epoch": 0.05566, + "grad_norm": 1.0637503448630574, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 5566 + }, + { + "epoch": 0.05567, + "grad_norm": 1.0859660134329547, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 5567 + }, + { + "epoch": 0.05568, + "grad_norm": 1.0189336160886215, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 5568 + }, + { + "epoch": 0.05569, + "grad_norm": 1.004138283903561, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5569 + }, + { + "epoch": 0.0557, + "grad_norm": 1.190958221299472, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 5570 + }, + { + "epoch": 0.05571, + "grad_norm": 1.010721856861485, + "learning_rate": 0.003, + "loss": 4.155, + "step": 5571 + }, + { + "epoch": 0.05572, + "grad_norm": 0.9197968716418343, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 5572 + }, + { + "epoch": 0.05573, + "grad_norm": 0.9807472494013146, + "learning_rate": 0.003, + "loss": 4.1767, + "step": 5573 + }, + { + "epoch": 0.05574, + "grad_norm": 0.9948127166172148, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 5574 + }, + { + "epoch": 0.05575, + "grad_norm": 0.971313458379033, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 5575 + }, + { + "epoch": 0.05576, + "grad_norm": 1.0277422588399054, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 5576 + }, + { + "epoch": 0.05577, + "grad_norm": 0.9995940449974644, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 5577 + }, + { + "epoch": 0.05578, + "grad_norm": 1.0859216843814685, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 5578 + }, + { + "epoch": 0.05579, + "grad_norm": 0.9574713713729739, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 5579 + }, + { + "epoch": 0.0558, + "grad_norm": 0.8737247794468992, + "learning_rate": 0.003, + "loss": 4.1549, + "step": 5580 + }, + { + "epoch": 0.05581, + "grad_norm": 0.7539148254380033, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 5581 + }, + { + "epoch": 0.05582, + "grad_norm": 0.7731731504235131, + "learning_rate": 0.003, + "loss": 4.174, + "step": 5582 + }, + { + "epoch": 0.05583, + "grad_norm": 0.786183465585757, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 5583 + }, + { + "epoch": 0.05584, + "grad_norm": 0.8712656140654648, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 5584 + }, + { + "epoch": 0.05585, + "grad_norm": 0.8461555674443455, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 5585 + }, + { + "epoch": 0.05586, + "grad_norm": 0.6370593274231787, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 5586 + }, + { + "epoch": 0.05587, + "grad_norm": 0.6466924043657243, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 5587 + }, + { + "epoch": 0.05588, + "grad_norm": 0.6622345802465658, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5588 + }, + { + "epoch": 0.05589, + "grad_norm": 0.5913876457137929, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5589 + }, + { + "epoch": 0.0559, + "grad_norm": 0.5270714516413212, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5590 + }, + { + "epoch": 0.05591, + "grad_norm": 0.5304582733694935, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 5591 + }, + { + "epoch": 0.05592, + "grad_norm": 0.5313660136101308, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 5592 + }, + { + "epoch": 0.05593, + "grad_norm": 0.5134736329805021, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5593 + }, + { + "epoch": 0.05594, + "grad_norm": 0.4831292187313499, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 5594 + }, + { + "epoch": 0.05595, + "grad_norm": 0.47641374196121894, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 5595 + }, + { + "epoch": 0.05596, + "grad_norm": 0.5262150385564768, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 5596 + }, + { + "epoch": 0.05597, + "grad_norm": 0.5970056001301876, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 5597 + }, + { + "epoch": 0.05598, + "grad_norm": 0.5741914372250703, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 5598 + }, + { + "epoch": 0.05599, + "grad_norm": 0.6464864543941242, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 5599 + }, + { + "epoch": 0.056, + "grad_norm": 0.9195309880740932, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 5600 + }, + { + "epoch": 0.05601, + "grad_norm": 1.2480023251412764, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 5601 + }, + { + "epoch": 0.05602, + "grad_norm": 0.7822871448594372, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 5602 + }, + { + "epoch": 0.05603, + "grad_norm": 0.7289334276521395, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 5603 + }, + { + "epoch": 0.05604, + "grad_norm": 0.679359088573528, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 5604 + }, + { + "epoch": 0.05605, + "grad_norm": 0.7513465211377279, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5605 + }, + { + "epoch": 0.05606, + "grad_norm": 0.7553929167731772, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 5606 + }, + { + "epoch": 0.05607, + "grad_norm": 0.7847903163107046, + "learning_rate": 0.003, + "loss": 4.1462, + "step": 5607 + }, + { + "epoch": 0.05608, + "grad_norm": 0.7298569984980766, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5608 + }, + { + "epoch": 0.05609, + "grad_norm": 0.8968015359645949, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 5609 + }, + { + "epoch": 0.0561, + "grad_norm": 0.911039007619847, + "learning_rate": 0.003, + "loss": 4.158, + "step": 5610 + }, + { + "epoch": 0.05611, + "grad_norm": 0.6357740462584217, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 5611 + }, + { + "epoch": 0.05612, + "grad_norm": 0.6371928488100147, + "learning_rate": 0.003, + "loss": 4.109, + "step": 5612 + }, + { + "epoch": 0.05613, + "grad_norm": 0.7120154136241356, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 5613 + }, + { + "epoch": 0.05614, + "grad_norm": 0.6531842861897988, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5614 + }, + { + "epoch": 0.05615, + "grad_norm": 0.723116589410996, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 5615 + }, + { + "epoch": 0.05616, + "grad_norm": 0.6683266115411998, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 5616 + }, + { + "epoch": 0.05617, + "grad_norm": 0.5821372860770843, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 5617 + }, + { + "epoch": 0.05618, + "grad_norm": 0.5360283472097199, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 5618 + }, + { + "epoch": 0.05619, + "grad_norm": 0.5939904129474893, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 5619 + }, + { + "epoch": 0.0562, + "grad_norm": 0.667235826498158, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 5620 + }, + { + "epoch": 0.05621, + "grad_norm": 0.7593944756058572, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 5621 + }, + { + "epoch": 0.05622, + "grad_norm": 0.7560267363483694, + "learning_rate": 0.003, + "loss": 4.108, + "step": 5622 + }, + { + "epoch": 0.05623, + "grad_norm": 0.9633744968283545, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 5623 + }, + { + "epoch": 0.05624, + "grad_norm": 1.3159764636382347, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 5624 + }, + { + "epoch": 0.05625, + "grad_norm": 0.8598542406723987, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5625 + }, + { + "epoch": 0.05626, + "grad_norm": 0.8928264883638428, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 5626 + }, + { + "epoch": 0.05627, + "grad_norm": 0.9035581162578888, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 5627 + }, + { + "epoch": 0.05628, + "grad_norm": 0.9326763323161109, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5628 + }, + { + "epoch": 0.05629, + "grad_norm": 0.9858524239497275, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 5629 + }, + { + "epoch": 0.0563, + "grad_norm": 1.0278927211295703, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5630 + }, + { + "epoch": 0.05631, + "grad_norm": 0.8158010640041221, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5631 + }, + { + "epoch": 0.05632, + "grad_norm": 0.7904116451209343, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 5632 + }, + { + "epoch": 0.05633, + "grad_norm": 0.6674336130552363, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5633 + }, + { + "epoch": 0.05634, + "grad_norm": 0.7131647955202249, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 5634 + }, + { + "epoch": 0.05635, + "grad_norm": 0.7762681609027913, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5635 + }, + { + "epoch": 0.05636, + "grad_norm": 0.8021027171917042, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5636 + }, + { + "epoch": 0.05637, + "grad_norm": 0.925350555541478, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 5637 + }, + { + "epoch": 0.05638, + "grad_norm": 0.9727914008655373, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 5638 + }, + { + "epoch": 0.05639, + "grad_norm": 0.7920727200781233, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5639 + }, + { + "epoch": 0.0564, + "grad_norm": 0.6724027259224868, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5640 + }, + { + "epoch": 0.05641, + "grad_norm": 0.6404204234893655, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 5641 + }, + { + "epoch": 0.05642, + "grad_norm": 0.638929472798256, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 5642 + }, + { + "epoch": 0.05643, + "grad_norm": 0.6279646959643067, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5643 + }, + { + "epoch": 0.05644, + "grad_norm": 0.6050520516714298, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 5644 + }, + { + "epoch": 0.05645, + "grad_norm": 0.7210392134789482, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 5645 + }, + { + "epoch": 0.05646, + "grad_norm": 0.9082430985310681, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5646 + }, + { + "epoch": 0.05647, + "grad_norm": 1.0629749955090315, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 5647 + }, + { + "epoch": 0.05648, + "grad_norm": 0.9337407114146831, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 5648 + }, + { + "epoch": 0.05649, + "grad_norm": 0.7908235143189736, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5649 + }, + { + "epoch": 0.0565, + "grad_norm": 0.6719463167631939, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 5650 + }, + { + "epoch": 0.05651, + "grad_norm": 0.6366536023365831, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 5651 + }, + { + "epoch": 0.05652, + "grad_norm": 0.6620856715240755, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 5652 + }, + { + "epoch": 0.05653, + "grad_norm": 0.6375701410325152, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5653 + }, + { + "epoch": 0.05654, + "grad_norm": 0.7756105771774413, + "learning_rate": 0.003, + "loss": 4.14, + "step": 5654 + }, + { + "epoch": 0.05655, + "grad_norm": 0.8746926551877258, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 5655 + }, + { + "epoch": 0.05656, + "grad_norm": 0.824346564649059, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 5656 + }, + { + "epoch": 0.05657, + "grad_norm": 0.9029529758933031, + "learning_rate": 0.003, + "loss": 4.131, + "step": 5657 + }, + { + "epoch": 0.05658, + "grad_norm": 0.9351306065482489, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5658 + }, + { + "epoch": 0.05659, + "grad_norm": 0.8596436492121928, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5659 + }, + { + "epoch": 0.0566, + "grad_norm": 0.9772647763222754, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 5660 + }, + { + "epoch": 0.05661, + "grad_norm": 1.009709800322154, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 5661 + }, + { + "epoch": 0.05662, + "grad_norm": 0.8700878215220617, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 5662 + }, + { + "epoch": 0.05663, + "grad_norm": 0.7314812319835777, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5663 + }, + { + "epoch": 0.05664, + "grad_norm": 0.7218667255886607, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 5664 + }, + { + "epoch": 0.05665, + "grad_norm": 0.7545885907055334, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5665 + }, + { + "epoch": 0.05666, + "grad_norm": 0.750379441945731, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 5666 + }, + { + "epoch": 0.05667, + "grad_norm": 0.7649794680001714, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 5667 + }, + { + "epoch": 0.05668, + "grad_norm": 0.7398186277471477, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 5668 + }, + { + "epoch": 0.05669, + "grad_norm": 0.7190311775637344, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5669 + }, + { + "epoch": 0.0567, + "grad_norm": 0.7385720576479917, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 5670 + }, + { + "epoch": 0.05671, + "grad_norm": 0.7190124419045905, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 5671 + }, + { + "epoch": 0.05672, + "grad_norm": 0.6906608031924452, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 5672 + }, + { + "epoch": 0.05673, + "grad_norm": 0.7440346855522176, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 5673 + }, + { + "epoch": 0.05674, + "grad_norm": 0.8058800416308596, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5674 + }, + { + "epoch": 0.05675, + "grad_norm": 0.9524033317076179, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 5675 + }, + { + "epoch": 0.05676, + "grad_norm": 1.0466364934113919, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 5676 + }, + { + "epoch": 0.05677, + "grad_norm": 1.1290737878900445, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 5677 + }, + { + "epoch": 0.05678, + "grad_norm": 0.8056761897248448, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5678 + }, + { + "epoch": 0.05679, + "grad_norm": 0.8416448974640103, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 5679 + }, + { + "epoch": 0.0568, + "grad_norm": 0.8654292108286987, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 5680 + }, + { + "epoch": 0.05681, + "grad_norm": 0.8669869662711412, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 5681 + }, + { + "epoch": 0.05682, + "grad_norm": 0.7948737245049556, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5682 + }, + { + "epoch": 0.05683, + "grad_norm": 0.7377380328343832, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5683 + }, + { + "epoch": 0.05684, + "grad_norm": 0.8063644138611359, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 5684 + }, + { + "epoch": 0.05685, + "grad_norm": 0.8527467662490903, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 5685 + }, + { + "epoch": 0.05686, + "grad_norm": 0.727039984777113, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 5686 + }, + { + "epoch": 0.05687, + "grad_norm": 0.6543891605340604, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5687 + }, + { + "epoch": 0.05688, + "grad_norm": 0.6732327830992805, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 5688 + }, + { + "epoch": 0.05689, + "grad_norm": 0.6988792787125268, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 5689 + }, + { + "epoch": 0.0569, + "grad_norm": 0.7481389272798462, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5690 + }, + { + "epoch": 0.05691, + "grad_norm": 0.8277047142206917, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 5691 + }, + { + "epoch": 0.05692, + "grad_norm": 0.8689994157613129, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5692 + }, + { + "epoch": 0.05693, + "grad_norm": 0.8286352092123737, + "learning_rate": 0.003, + "loss": 4.127, + "step": 5693 + }, + { + "epoch": 0.05694, + "grad_norm": 0.8815887145121191, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5694 + }, + { + "epoch": 0.05695, + "grad_norm": 0.8249748591708475, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5695 + }, + { + "epoch": 0.05696, + "grad_norm": 0.8653183020137561, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 5696 + }, + { + "epoch": 0.05697, + "grad_norm": 0.8835391111395854, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 5697 + }, + { + "epoch": 0.05698, + "grad_norm": 0.8593431463479471, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 5698 + }, + { + "epoch": 0.05699, + "grad_norm": 0.8457692338051964, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 5699 + }, + { + "epoch": 0.057, + "grad_norm": 0.7057109559383719, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5700 + }, + { + "epoch": 0.05701, + "grad_norm": 0.7143700073271365, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 5701 + }, + { + "epoch": 0.05702, + "grad_norm": 0.7926557745297437, + "learning_rate": 0.003, + "loss": 4.112, + "step": 5702 + }, + { + "epoch": 0.05703, + "grad_norm": 0.788953414454626, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 5703 + }, + { + "epoch": 0.05704, + "grad_norm": 0.8987044741769469, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5704 + }, + { + "epoch": 0.05705, + "grad_norm": 0.9475911671053671, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 5705 + }, + { + "epoch": 0.05706, + "grad_norm": 0.9351503494399143, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 5706 + }, + { + "epoch": 0.05707, + "grad_norm": 0.908791879595692, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 5707 + }, + { + "epoch": 0.05708, + "grad_norm": 0.9702984464047565, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 5708 + }, + { + "epoch": 0.05709, + "grad_norm": 1.02607623723065, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5709 + }, + { + "epoch": 0.0571, + "grad_norm": 0.8383478585082147, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 5710 + }, + { + "epoch": 0.05711, + "grad_norm": 0.8409807556043147, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 5711 + }, + { + "epoch": 0.05712, + "grad_norm": 0.8246373489572517, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 5712 + }, + { + "epoch": 0.05713, + "grad_norm": 0.750695304399167, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 5713 + }, + { + "epoch": 0.05714, + "grad_norm": 0.6256785887976535, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 5714 + }, + { + "epoch": 0.05715, + "grad_norm": 0.6405112496698065, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 5715 + }, + { + "epoch": 0.05716, + "grad_norm": 0.7045480948725663, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 5716 + }, + { + "epoch": 0.05717, + "grad_norm": 0.7251180513932567, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 5717 + }, + { + "epoch": 0.05718, + "grad_norm": 0.7144875502995738, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5718 + }, + { + "epoch": 0.05719, + "grad_norm": 0.7691901405002558, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 5719 + }, + { + "epoch": 0.0572, + "grad_norm": 0.733838672256027, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5720 + }, + { + "epoch": 0.05721, + "grad_norm": 0.7007983767497351, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 5721 + }, + { + "epoch": 0.05722, + "grad_norm": 0.6786326167442739, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5722 + }, + { + "epoch": 0.05723, + "grad_norm": 0.7298075380572556, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 5723 + }, + { + "epoch": 0.05724, + "grad_norm": 0.8701417935987675, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 5724 + }, + { + "epoch": 0.05725, + "grad_norm": 0.9317116533192469, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5725 + }, + { + "epoch": 0.05726, + "grad_norm": 1.2423360551944322, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 5726 + }, + { + "epoch": 0.05727, + "grad_norm": 0.9566322040091452, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 5727 + }, + { + "epoch": 0.05728, + "grad_norm": 0.8630235197631045, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 5728 + }, + { + "epoch": 0.05729, + "grad_norm": 0.6367381629677203, + "learning_rate": 0.003, + "loss": 4.118, + "step": 5729 + }, + { + "epoch": 0.0573, + "grad_norm": 0.6307754247308351, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 5730 + }, + { + "epoch": 0.05731, + "grad_norm": 0.6637990852153008, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 5731 + }, + { + "epoch": 0.05732, + "grad_norm": 0.7152262181674999, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 5732 + }, + { + "epoch": 0.05733, + "grad_norm": 0.7583908888024418, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 5733 + }, + { + "epoch": 0.05734, + "grad_norm": 0.7979697396059119, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 5734 + }, + { + "epoch": 0.05735, + "grad_norm": 0.6659788683114847, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 5735 + }, + { + "epoch": 0.05736, + "grad_norm": 0.6718526951634197, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 5736 + }, + { + "epoch": 0.05737, + "grad_norm": 0.5932737131818991, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 5737 + }, + { + "epoch": 0.05738, + "grad_norm": 0.5552311466351124, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 5738 + }, + { + "epoch": 0.05739, + "grad_norm": 0.5192428824433908, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5739 + }, + { + "epoch": 0.0574, + "grad_norm": 0.5521978223598561, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5740 + }, + { + "epoch": 0.05741, + "grad_norm": 0.6827764477275866, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5741 + }, + { + "epoch": 0.05742, + "grad_norm": 0.9661334292481325, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 5742 + }, + { + "epoch": 0.05743, + "grad_norm": 1.3808942777208146, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5743 + }, + { + "epoch": 0.05744, + "grad_norm": 0.6634000683140943, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 5744 + }, + { + "epoch": 0.05745, + "grad_norm": 0.8051574353421863, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 5745 + }, + { + "epoch": 0.05746, + "grad_norm": 0.8451000982869576, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5746 + }, + { + "epoch": 0.05747, + "grad_norm": 0.7005027290316844, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 5747 + }, + { + "epoch": 0.05748, + "grad_norm": 0.7417025932554691, + "learning_rate": 0.003, + "loss": 4.1526, + "step": 5748 + }, + { + "epoch": 0.05749, + "grad_norm": 0.7770045474271049, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5749 + }, + { + "epoch": 0.0575, + "grad_norm": 0.8496769014517581, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5750 + }, + { + "epoch": 0.05751, + "grad_norm": 0.9049063166808049, + "learning_rate": 0.003, + "loss": 4.144, + "step": 5751 + }, + { + "epoch": 0.05752, + "grad_norm": 1.0297317331778117, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 5752 + }, + { + "epoch": 0.05753, + "grad_norm": 0.9151964474023491, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 5753 + }, + { + "epoch": 0.05754, + "grad_norm": 0.8031670531802124, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 5754 + }, + { + "epoch": 0.05755, + "grad_norm": 1.0112539909545384, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 5755 + }, + { + "epoch": 0.05756, + "grad_norm": 1.1823863497130285, + "learning_rate": 0.003, + "loss": 4.1551, + "step": 5756 + }, + { + "epoch": 0.05757, + "grad_norm": 0.82463840960989, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 5757 + }, + { + "epoch": 0.05758, + "grad_norm": 0.8370537741924009, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 5758 + }, + { + "epoch": 0.05759, + "grad_norm": 0.7690512085008506, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5759 + }, + { + "epoch": 0.0576, + "grad_norm": 0.8543370684675727, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 5760 + }, + { + "epoch": 0.05761, + "grad_norm": 0.8953621498467661, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 5761 + }, + { + "epoch": 0.05762, + "grad_norm": 0.8489500157996731, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 5762 + }, + { + "epoch": 0.05763, + "grad_norm": 0.7805963155950815, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 5763 + }, + { + "epoch": 0.05764, + "grad_norm": 0.7082991317232272, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 5764 + }, + { + "epoch": 0.05765, + "grad_norm": 0.7124286749136263, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5765 + }, + { + "epoch": 0.05766, + "grad_norm": 0.6760677428953438, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5766 + }, + { + "epoch": 0.05767, + "grad_norm": 0.8489631258318778, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 5767 + }, + { + "epoch": 0.05768, + "grad_norm": 0.9362514658122035, + "learning_rate": 0.003, + "loss": 4.139, + "step": 5768 + }, + { + "epoch": 0.05769, + "grad_norm": 0.8865918581298181, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 5769 + }, + { + "epoch": 0.0577, + "grad_norm": 0.7665977076446682, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 5770 + }, + { + "epoch": 0.05771, + "grad_norm": 0.692978560597086, + "learning_rate": 0.003, + "loss": 4.129, + "step": 5771 + }, + { + "epoch": 0.05772, + "grad_norm": 0.5670375020911395, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 5772 + }, + { + "epoch": 0.05773, + "grad_norm": 0.5994073546273513, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5773 + }, + { + "epoch": 0.05774, + "grad_norm": 0.6697442229469948, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 5774 + }, + { + "epoch": 0.05775, + "grad_norm": 0.7260764971579977, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5775 + }, + { + "epoch": 0.05776, + "grad_norm": 0.827091034088083, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 5776 + }, + { + "epoch": 0.05777, + "grad_norm": 0.85821320032967, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5777 + }, + { + "epoch": 0.05778, + "grad_norm": 0.75564632841601, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5778 + }, + { + "epoch": 0.05779, + "grad_norm": 0.758336533791715, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 5779 + }, + { + "epoch": 0.0578, + "grad_norm": 0.8762931646419935, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 5780 + }, + { + "epoch": 0.05781, + "grad_norm": 0.8498076917904277, + "learning_rate": 0.003, + "loss": 4.106, + "step": 5781 + }, + { + "epoch": 0.05782, + "grad_norm": 0.9973539387616522, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 5782 + }, + { + "epoch": 0.05783, + "grad_norm": 1.1963721721828537, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5783 + }, + { + "epoch": 0.05784, + "grad_norm": 0.8859060956969492, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5784 + }, + { + "epoch": 0.05785, + "grad_norm": 0.8870211693131135, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 5785 + }, + { + "epoch": 0.05786, + "grad_norm": 0.8028500867550125, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 5786 + }, + { + "epoch": 0.05787, + "grad_norm": 0.9175527403883458, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5787 + }, + { + "epoch": 0.05788, + "grad_norm": 0.9094044732533373, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 5788 + }, + { + "epoch": 0.05789, + "grad_norm": 0.9198435831448731, + "learning_rate": 0.003, + "loss": 4.131, + "step": 5789 + }, + { + "epoch": 0.0579, + "grad_norm": 0.8903452476156254, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 5790 + }, + { + "epoch": 0.05791, + "grad_norm": 0.7995965533827011, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 5791 + }, + { + "epoch": 0.05792, + "grad_norm": 0.7374601576105686, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 5792 + }, + { + "epoch": 0.05793, + "grad_norm": 0.8694986815433677, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 5793 + }, + { + "epoch": 0.05794, + "grad_norm": 1.0476632248094724, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 5794 + }, + { + "epoch": 0.05795, + "grad_norm": 1.1869938993812021, + "learning_rate": 0.003, + "loss": 4.171, + "step": 5795 + }, + { + "epoch": 0.05796, + "grad_norm": 0.8360088466789314, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5796 + }, + { + "epoch": 0.05797, + "grad_norm": 0.8084090857296047, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5797 + }, + { + "epoch": 0.05798, + "grad_norm": 0.7637314143049846, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 5798 + }, + { + "epoch": 0.05799, + "grad_norm": 0.8732093010633248, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 5799 + }, + { + "epoch": 0.058, + "grad_norm": 0.7122155596601744, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 5800 + }, + { + "epoch": 0.05801, + "grad_norm": 0.6818953681284188, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 5801 + }, + { + "epoch": 0.05802, + "grad_norm": 0.7665388126779251, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5802 + }, + { + "epoch": 0.05803, + "grad_norm": 0.8847770862686954, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 5803 + }, + { + "epoch": 0.05804, + "grad_norm": 0.8330623324317175, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 5804 + }, + { + "epoch": 0.05805, + "grad_norm": 0.667711721351076, + "learning_rate": 0.003, + "loss": 4.091, + "step": 5805 + }, + { + "epoch": 0.05806, + "grad_norm": 0.7138990216020714, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 5806 + }, + { + "epoch": 0.05807, + "grad_norm": 0.6844079777026275, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 5807 + }, + { + "epoch": 0.05808, + "grad_norm": 0.7329662618338171, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 5808 + }, + { + "epoch": 0.05809, + "grad_norm": 0.7127193982748581, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 5809 + }, + { + "epoch": 0.0581, + "grad_norm": 0.7849473665098132, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 5810 + }, + { + "epoch": 0.05811, + "grad_norm": 0.7772533700540266, + "learning_rate": 0.003, + "loss": 4.111, + "step": 5811 + }, + { + "epoch": 0.05812, + "grad_norm": 0.8535932209668625, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 5812 + }, + { + "epoch": 0.05813, + "grad_norm": 1.050675078146715, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 5813 + }, + { + "epoch": 0.05814, + "grad_norm": 0.9859964212140094, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 5814 + }, + { + "epoch": 0.05815, + "grad_norm": 0.8071168638710576, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5815 + }, + { + "epoch": 0.05816, + "grad_norm": 0.744822186311271, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 5816 + }, + { + "epoch": 0.05817, + "grad_norm": 0.859723113139702, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5817 + }, + { + "epoch": 0.05818, + "grad_norm": 0.9076124598026192, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 5818 + }, + { + "epoch": 0.05819, + "grad_norm": 0.8614969748668542, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 5819 + }, + { + "epoch": 0.0582, + "grad_norm": 0.7596720316772002, + "learning_rate": 0.003, + "loss": 4.123, + "step": 5820 + }, + { + "epoch": 0.05821, + "grad_norm": 0.7261229816186945, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5821 + }, + { + "epoch": 0.05822, + "grad_norm": 0.6389335810166767, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 5822 + }, + { + "epoch": 0.05823, + "grad_norm": 0.650431814244154, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 5823 + }, + { + "epoch": 0.05824, + "grad_norm": 0.7021908556345984, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5824 + }, + { + "epoch": 0.05825, + "grad_norm": 0.7075707130594027, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 5825 + }, + { + "epoch": 0.05826, + "grad_norm": 0.6632684340681474, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5826 + }, + { + "epoch": 0.05827, + "grad_norm": 0.6153406505756235, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 5827 + }, + { + "epoch": 0.05828, + "grad_norm": 0.5151089751197206, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5828 + }, + { + "epoch": 0.05829, + "grad_norm": 0.5657176931414479, + "learning_rate": 0.003, + "loss": 4.118, + "step": 5829 + }, + { + "epoch": 0.0583, + "grad_norm": 0.7071018057037146, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5830 + }, + { + "epoch": 0.05831, + "grad_norm": 1.0078552329007178, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5831 + }, + { + "epoch": 0.05832, + "grad_norm": 1.461038643232985, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 5832 + }, + { + "epoch": 0.05833, + "grad_norm": 0.515512286205366, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 5833 + }, + { + "epoch": 0.05834, + "grad_norm": 0.8693462452576697, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 5834 + }, + { + "epoch": 0.05835, + "grad_norm": 0.8652446702771586, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 5835 + }, + { + "epoch": 0.05836, + "grad_norm": 0.7247395539295343, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 5836 + }, + { + "epoch": 0.05837, + "grad_norm": 0.6702651347004529, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 5837 + }, + { + "epoch": 0.05838, + "grad_norm": 0.7045650962837332, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5838 + }, + { + "epoch": 0.05839, + "grad_norm": 0.785589688267228, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 5839 + }, + { + "epoch": 0.0584, + "grad_norm": 0.8133814023137367, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 5840 + }, + { + "epoch": 0.05841, + "grad_norm": 0.832422818436443, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 5841 + }, + { + "epoch": 0.05842, + "grad_norm": 0.8619710688010511, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 5842 + }, + { + "epoch": 0.05843, + "grad_norm": 0.9243101513524902, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 5843 + }, + { + "epoch": 0.05844, + "grad_norm": 0.8736875279868362, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 5844 + }, + { + "epoch": 0.05845, + "grad_norm": 0.8068837365535566, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5845 + }, + { + "epoch": 0.05846, + "grad_norm": 0.711931455880162, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 5846 + }, + { + "epoch": 0.05847, + "grad_norm": 0.7993811775050275, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 5847 + }, + { + "epoch": 0.05848, + "grad_norm": 0.754358214475092, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5848 + }, + { + "epoch": 0.05849, + "grad_norm": 0.8702234783293419, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 5849 + }, + { + "epoch": 0.0585, + "grad_norm": 0.9876577919554057, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 5850 + }, + { + "epoch": 0.05851, + "grad_norm": 1.0940891749043231, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 5851 + }, + { + "epoch": 0.05852, + "grad_norm": 0.9378030392921249, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5852 + }, + { + "epoch": 0.05853, + "grad_norm": 0.7566439324877372, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 5853 + }, + { + "epoch": 0.05854, + "grad_norm": 0.673296794630987, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 5854 + }, + { + "epoch": 0.05855, + "grad_norm": 0.578895079904098, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 5855 + }, + { + "epoch": 0.05856, + "grad_norm": 0.624693508028921, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 5856 + }, + { + "epoch": 0.05857, + "grad_norm": 0.7368589206607341, + "learning_rate": 0.003, + "loss": 4.128, + "step": 5857 + }, + { + "epoch": 0.05858, + "grad_norm": 0.8538648315734454, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5858 + }, + { + "epoch": 0.05859, + "grad_norm": 0.8556665179737083, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 5859 + }, + { + "epoch": 0.0586, + "grad_norm": 0.7826151609045512, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5860 + }, + { + "epoch": 0.05861, + "grad_norm": 0.7886112191620424, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 5861 + }, + { + "epoch": 0.05862, + "grad_norm": 0.9292901295633138, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 5862 + }, + { + "epoch": 0.05863, + "grad_norm": 0.8121851405516025, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 5863 + }, + { + "epoch": 0.05864, + "grad_norm": 0.6827602835508371, + "learning_rate": 0.003, + "loss": 4.107, + "step": 5864 + }, + { + "epoch": 0.05865, + "grad_norm": 0.7544438864360793, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5865 + }, + { + "epoch": 0.05866, + "grad_norm": 0.8838045357890466, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 5866 + }, + { + "epoch": 0.05867, + "grad_norm": 0.8168835101024318, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 5867 + }, + { + "epoch": 0.05868, + "grad_norm": 0.6536324933405178, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5868 + }, + { + "epoch": 0.05869, + "grad_norm": 0.683464201492265, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 5869 + }, + { + "epoch": 0.0587, + "grad_norm": 0.7963857853576817, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 5870 + }, + { + "epoch": 0.05871, + "grad_norm": 0.9792801347682168, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5871 + }, + { + "epoch": 0.05872, + "grad_norm": 1.4172796436924118, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 5872 + }, + { + "epoch": 0.05873, + "grad_norm": 0.7027081325886863, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 5873 + }, + { + "epoch": 0.05874, + "grad_norm": 0.7473953186030906, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 5874 + }, + { + "epoch": 0.05875, + "grad_norm": 0.809528681305329, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 5875 + }, + { + "epoch": 0.05876, + "grad_norm": 0.772298806074322, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5876 + }, + { + "epoch": 0.05877, + "grad_norm": 0.8257282875546335, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 5877 + }, + { + "epoch": 0.05878, + "grad_norm": 0.9455870472025797, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 5878 + }, + { + "epoch": 0.05879, + "grad_norm": 1.1569037258481754, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 5879 + }, + { + "epoch": 0.0588, + "grad_norm": 0.8577210204570376, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 5880 + }, + { + "epoch": 0.05881, + "grad_norm": 0.8032693830476965, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 5881 + }, + { + "epoch": 0.05882, + "grad_norm": 0.8937763672419452, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 5882 + }, + { + "epoch": 0.05883, + "grad_norm": 0.9593530484800846, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 5883 + }, + { + "epoch": 0.05884, + "grad_norm": 0.8900684740385257, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 5884 + }, + { + "epoch": 0.05885, + "grad_norm": 0.8643772869409747, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 5885 + }, + { + "epoch": 0.05886, + "grad_norm": 0.9219910015321089, + "learning_rate": 0.003, + "loss": 4.138, + "step": 5886 + }, + { + "epoch": 0.05887, + "grad_norm": 0.9867677093915318, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 5887 + }, + { + "epoch": 0.05888, + "grad_norm": 0.9465735583992112, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5888 + }, + { + "epoch": 0.05889, + "grad_norm": 0.8466592224711629, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 5889 + }, + { + "epoch": 0.0589, + "grad_norm": 0.886972464545185, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 5890 + }, + { + "epoch": 0.05891, + "grad_norm": 0.8652857070984449, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 5891 + }, + { + "epoch": 0.05892, + "grad_norm": 0.8664086289943967, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 5892 + }, + { + "epoch": 0.05893, + "grad_norm": 0.8708892579951463, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5893 + }, + { + "epoch": 0.05894, + "grad_norm": 0.8078932753206562, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 5894 + }, + { + "epoch": 0.05895, + "grad_norm": 0.7575994215774198, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 5895 + }, + { + "epoch": 0.05896, + "grad_norm": 0.8549826754505118, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 5896 + }, + { + "epoch": 0.05897, + "grad_norm": 1.0144405805900867, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 5897 + }, + { + "epoch": 0.05898, + "grad_norm": 0.9395405641055523, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 5898 + }, + { + "epoch": 0.05899, + "grad_norm": 0.81787258885681, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 5899 + }, + { + "epoch": 0.059, + "grad_norm": 0.7866920164370645, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 5900 + }, + { + "epoch": 0.05901, + "grad_norm": 0.910804015247061, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 5901 + }, + { + "epoch": 0.05902, + "grad_norm": 0.8747094738532013, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 5902 + }, + { + "epoch": 0.05903, + "grad_norm": 0.8330955629581883, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 5903 + }, + { + "epoch": 0.05904, + "grad_norm": 0.8149396050827394, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5904 + }, + { + "epoch": 0.05905, + "grad_norm": 0.8504004488105502, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 5905 + }, + { + "epoch": 0.05906, + "grad_norm": 0.8430930663249769, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 5906 + }, + { + "epoch": 0.05907, + "grad_norm": 0.7769082221377612, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 5907 + }, + { + "epoch": 0.05908, + "grad_norm": 0.7772454693694203, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 5908 + }, + { + "epoch": 0.05909, + "grad_norm": 0.7555830902663292, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 5909 + }, + { + "epoch": 0.0591, + "grad_norm": 0.852614526791537, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5910 + }, + { + "epoch": 0.05911, + "grad_norm": 0.9020994358765787, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 5911 + }, + { + "epoch": 0.05912, + "grad_norm": 0.8745419850417824, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 5912 + }, + { + "epoch": 0.05913, + "grad_norm": 0.8119839119136394, + "learning_rate": 0.003, + "loss": 4.1646, + "step": 5913 + }, + { + "epoch": 0.05914, + "grad_norm": 0.8566768807668204, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 5914 + }, + { + "epoch": 0.05915, + "grad_norm": 0.8415977723508733, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 5915 + }, + { + "epoch": 0.05916, + "grad_norm": 0.8568674252517323, + "learning_rate": 0.003, + "loss": 4.123, + "step": 5916 + }, + { + "epoch": 0.05917, + "grad_norm": 0.9558112684544686, + "learning_rate": 0.003, + "loss": 4.115, + "step": 5917 + }, + { + "epoch": 0.05918, + "grad_norm": 0.841796281005891, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5918 + }, + { + "epoch": 0.05919, + "grad_norm": 0.7555449033676591, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 5919 + }, + { + "epoch": 0.0592, + "grad_norm": 0.6802701845197375, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5920 + }, + { + "epoch": 0.05921, + "grad_norm": 0.7044168918621094, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 5921 + }, + { + "epoch": 0.05922, + "grad_norm": 0.6836628304390027, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 5922 + }, + { + "epoch": 0.05923, + "grad_norm": 0.6799364137142957, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 5923 + }, + { + "epoch": 0.05924, + "grad_norm": 0.6507013910949614, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 5924 + }, + { + "epoch": 0.05925, + "grad_norm": 0.7486761778135521, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 5925 + }, + { + "epoch": 0.05926, + "grad_norm": 1.047983816202794, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 5926 + }, + { + "epoch": 0.05927, + "grad_norm": 1.1278171371548802, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 5927 + }, + { + "epoch": 0.05928, + "grad_norm": 0.6906724039529817, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 5928 + }, + { + "epoch": 0.05929, + "grad_norm": 0.6279774605894174, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 5929 + }, + { + "epoch": 0.0593, + "grad_norm": 0.7848995148521671, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 5930 + }, + { + "epoch": 0.05931, + "grad_norm": 0.7630809841555396, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 5931 + }, + { + "epoch": 0.05932, + "grad_norm": 0.787461470705689, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 5932 + }, + { + "epoch": 0.05933, + "grad_norm": 0.9168988991566422, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5933 + }, + { + "epoch": 0.05934, + "grad_norm": 0.9816745251843256, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 5934 + }, + { + "epoch": 0.05935, + "grad_norm": 0.9381079865102622, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 5935 + }, + { + "epoch": 0.05936, + "grad_norm": 0.7956503064418853, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 5936 + }, + { + "epoch": 0.05937, + "grad_norm": 0.7624269514648494, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 5937 + }, + { + "epoch": 0.05938, + "grad_norm": 0.7591454779489837, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5938 + }, + { + "epoch": 0.05939, + "grad_norm": 0.8950444564122585, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 5939 + }, + { + "epoch": 0.0594, + "grad_norm": 1.0147229565164109, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 5940 + }, + { + "epoch": 0.05941, + "grad_norm": 0.8688261982567362, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 5941 + }, + { + "epoch": 0.05942, + "grad_norm": 0.8203220118126857, + "learning_rate": 0.003, + "loss": 4.131, + "step": 5942 + }, + { + "epoch": 0.05943, + "grad_norm": 0.6982096168862657, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 5943 + }, + { + "epoch": 0.05944, + "grad_norm": 0.7108234697493028, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5944 + }, + { + "epoch": 0.05945, + "grad_norm": 0.7441575726527577, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 5945 + }, + { + "epoch": 0.05946, + "grad_norm": 0.693966267768056, + "learning_rate": 0.003, + "loss": 4.107, + "step": 5946 + }, + { + "epoch": 0.05947, + "grad_norm": 0.7492920047746423, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 5947 + }, + { + "epoch": 0.05948, + "grad_norm": 0.9531204208790365, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 5948 + }, + { + "epoch": 0.05949, + "grad_norm": 1.0780044113765543, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 5949 + }, + { + "epoch": 0.0595, + "grad_norm": 1.1155592756719261, + "learning_rate": 0.003, + "loss": 4.138, + "step": 5950 + }, + { + "epoch": 0.05951, + "grad_norm": 0.9010475516605746, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 5951 + }, + { + "epoch": 0.05952, + "grad_norm": 0.8136117524297802, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 5952 + }, + { + "epoch": 0.05953, + "grad_norm": 0.9497081402575908, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 5953 + }, + { + "epoch": 0.05954, + "grad_norm": 1.2015146969192423, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 5954 + }, + { + "epoch": 0.05955, + "grad_norm": 1.1019093784349765, + "learning_rate": 0.003, + "loss": 4.13, + "step": 5955 + }, + { + "epoch": 0.05956, + "grad_norm": 0.9208936016687609, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5956 + }, + { + "epoch": 0.05957, + "grad_norm": 0.9862765579790758, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 5957 + }, + { + "epoch": 0.05958, + "grad_norm": 0.9376388993356572, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 5958 + }, + { + "epoch": 0.05959, + "grad_norm": 0.8845077986280037, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 5959 + }, + { + "epoch": 0.0596, + "grad_norm": 0.7693050418942645, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 5960 + }, + { + "epoch": 0.05961, + "grad_norm": 0.7082701488122094, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 5961 + }, + { + "epoch": 0.05962, + "grad_norm": 0.735178214010553, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 5962 + }, + { + "epoch": 0.05963, + "grad_norm": 0.7296949063274745, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 5963 + }, + { + "epoch": 0.05964, + "grad_norm": 0.6884040487243323, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 5964 + }, + { + "epoch": 0.05965, + "grad_norm": 0.6613609811586886, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 5965 + }, + { + "epoch": 0.05966, + "grad_norm": 0.6081326960838325, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 5966 + }, + { + "epoch": 0.05967, + "grad_norm": 0.7025319127350109, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 5967 + }, + { + "epoch": 0.05968, + "grad_norm": 0.8710659914493641, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 5968 + }, + { + "epoch": 0.05969, + "grad_norm": 1.1435903384618258, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 5969 + }, + { + "epoch": 0.0597, + "grad_norm": 0.9448264927162972, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 5970 + }, + { + "epoch": 0.05971, + "grad_norm": 0.7976448783438047, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 5971 + }, + { + "epoch": 0.05972, + "grad_norm": 0.6796283942705301, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 5972 + }, + { + "epoch": 0.05973, + "grad_norm": 0.7020776410943722, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 5973 + }, + { + "epoch": 0.05974, + "grad_norm": 0.6191112756816605, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 5974 + }, + { + "epoch": 0.05975, + "grad_norm": 0.5989871789758782, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 5975 + }, + { + "epoch": 0.05976, + "grad_norm": 0.6739238404846538, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 5976 + }, + { + "epoch": 0.05977, + "grad_norm": 0.7039494119632328, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 5977 + }, + { + "epoch": 0.05978, + "grad_norm": 0.6642661875712308, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 5978 + }, + { + "epoch": 0.05979, + "grad_norm": 0.6685425716431447, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 5979 + }, + { + "epoch": 0.0598, + "grad_norm": 0.6261981100897624, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 5980 + }, + { + "epoch": 0.05981, + "grad_norm": 0.5985779168949105, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5981 + }, + { + "epoch": 0.05982, + "grad_norm": 0.6136781919006569, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 5982 + }, + { + "epoch": 0.05983, + "grad_norm": 0.6537718426165939, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 5983 + }, + { + "epoch": 0.05984, + "grad_norm": 0.714227533885321, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5984 + }, + { + "epoch": 0.05985, + "grad_norm": 0.9506528476418039, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 5985 + }, + { + "epoch": 0.05986, + "grad_norm": 1.2746078074067864, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 5986 + }, + { + "epoch": 0.05987, + "grad_norm": 0.6791043539335656, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 5987 + }, + { + "epoch": 0.05988, + "grad_norm": 0.6346472513213476, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 5988 + }, + { + "epoch": 0.05989, + "grad_norm": 0.7500925942098075, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 5989 + }, + { + "epoch": 0.0599, + "grad_norm": 0.7561227959508446, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5990 + }, + { + "epoch": 0.05991, + "grad_norm": 0.8673565510034621, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 5991 + }, + { + "epoch": 0.05992, + "grad_norm": 0.8299139122744383, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 5992 + }, + { + "epoch": 0.05993, + "grad_norm": 0.8298749382371114, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 5993 + }, + { + "epoch": 0.05994, + "grad_norm": 0.9205333072442913, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 5994 + }, + { + "epoch": 0.05995, + "grad_norm": 1.028139546571208, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 5995 + }, + { + "epoch": 0.05996, + "grad_norm": 1.0308864569555287, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 5996 + }, + { + "epoch": 0.05997, + "grad_norm": 0.9909951484527934, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 5997 + }, + { + "epoch": 0.05998, + "grad_norm": 0.9523678206761688, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 5998 + }, + { + "epoch": 0.05999, + "grad_norm": 0.8423597146329871, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 5999 + }, + { + "epoch": 0.06, + "grad_norm": 0.7574405785566366, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 6000 + }, + { + "epoch": 0.06001, + "grad_norm": 0.7753384956998035, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 6001 + }, + { + "epoch": 0.06002, + "grad_norm": 0.6748903991582674, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6002 + }, + { + "epoch": 0.06003, + "grad_norm": 0.6986365521153964, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6003 + }, + { + "epoch": 0.06004, + "grad_norm": 0.7256641113413403, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 6004 + }, + { + "epoch": 0.06005, + "grad_norm": 0.9044735035650205, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6005 + }, + { + "epoch": 0.06006, + "grad_norm": 0.9755346452792312, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6006 + }, + { + "epoch": 0.06007, + "grad_norm": 1.0182340535787422, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 6007 + }, + { + "epoch": 0.06008, + "grad_norm": 0.9652283579037206, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6008 + }, + { + "epoch": 0.06009, + "grad_norm": 0.9368301156725635, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 6009 + }, + { + "epoch": 0.0601, + "grad_norm": 0.8091764219084102, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 6010 + }, + { + "epoch": 0.06011, + "grad_norm": 0.8468775113048598, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 6011 + }, + { + "epoch": 0.06012, + "grad_norm": 0.9029973593955182, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 6012 + }, + { + "epoch": 0.06013, + "grad_norm": 0.967030390027878, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6013 + }, + { + "epoch": 0.06014, + "grad_norm": 0.8441165732003634, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6014 + }, + { + "epoch": 0.06015, + "grad_norm": 0.8268570933571939, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 6015 + }, + { + "epoch": 0.06016, + "grad_norm": 0.9865898228103146, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 6016 + }, + { + "epoch": 0.06017, + "grad_norm": 1.1052131183203031, + "learning_rate": 0.003, + "loss": 4.1563, + "step": 6017 + }, + { + "epoch": 0.06018, + "grad_norm": 0.8257250500559014, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 6018 + }, + { + "epoch": 0.06019, + "grad_norm": 0.9972106824107583, + "learning_rate": 0.003, + "loss": 4.1339, + "step": 6019 + }, + { + "epoch": 0.0602, + "grad_norm": 0.9191337383387227, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 6020 + }, + { + "epoch": 0.06021, + "grad_norm": 0.8616408991185872, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 6021 + }, + { + "epoch": 0.06022, + "grad_norm": 0.846750898429168, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 6022 + }, + { + "epoch": 0.06023, + "grad_norm": 0.7999915956085327, + "learning_rate": 0.003, + "loss": 4.1365, + "step": 6023 + }, + { + "epoch": 0.06024, + "grad_norm": 0.7601013634660952, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 6024 + }, + { + "epoch": 0.06025, + "grad_norm": 0.7551843609513086, + "learning_rate": 0.003, + "loss": 4.164, + "step": 6025 + }, + { + "epoch": 0.06026, + "grad_norm": 0.7858596116292706, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 6026 + }, + { + "epoch": 0.06027, + "grad_norm": 0.9347452644357562, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 6027 + }, + { + "epoch": 0.06028, + "grad_norm": 1.235123054449198, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 6028 + }, + { + "epoch": 0.06029, + "grad_norm": 0.6475597045882941, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 6029 + }, + { + "epoch": 0.0603, + "grad_norm": 0.7127227106122453, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6030 + }, + { + "epoch": 0.06031, + "grad_norm": 0.6537901329104299, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 6031 + }, + { + "epoch": 0.06032, + "grad_norm": 0.6475267434508163, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6032 + }, + { + "epoch": 0.06033, + "grad_norm": 0.612865958283774, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 6033 + }, + { + "epoch": 0.06034, + "grad_norm": 0.6061910811704923, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 6034 + }, + { + "epoch": 0.06035, + "grad_norm": 0.6873968709597013, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 6035 + }, + { + "epoch": 0.06036, + "grad_norm": 0.739626150320404, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 6036 + }, + { + "epoch": 0.06037, + "grad_norm": 0.8044747493310122, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 6037 + }, + { + "epoch": 0.06038, + "grad_norm": 0.7763911045180569, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6038 + }, + { + "epoch": 0.06039, + "grad_norm": 0.6590222868385746, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6039 + }, + { + "epoch": 0.0604, + "grad_norm": 0.6350119436647222, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 6040 + }, + { + "epoch": 0.06041, + "grad_norm": 0.6853702741345424, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 6041 + }, + { + "epoch": 0.06042, + "grad_norm": 0.812479618716859, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6042 + }, + { + "epoch": 0.06043, + "grad_norm": 1.0164590695928413, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6043 + }, + { + "epoch": 0.06044, + "grad_norm": 1.0986220369036817, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 6044 + }, + { + "epoch": 0.06045, + "grad_norm": 0.7560995542851634, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 6045 + }, + { + "epoch": 0.06046, + "grad_norm": 0.7425869786438686, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 6046 + }, + { + "epoch": 0.06047, + "grad_norm": 0.7895703623032488, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 6047 + }, + { + "epoch": 0.06048, + "grad_norm": 0.7640848771237746, + "learning_rate": 0.003, + "loss": 4.097, + "step": 6048 + }, + { + "epoch": 0.06049, + "grad_norm": 0.8191791885730467, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6049 + }, + { + "epoch": 0.0605, + "grad_norm": 0.9304880173731273, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6050 + }, + { + "epoch": 0.06051, + "grad_norm": 1.0615295174903012, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 6051 + }, + { + "epoch": 0.06052, + "grad_norm": 0.9073447027396274, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6052 + }, + { + "epoch": 0.06053, + "grad_norm": 0.8100014586142358, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 6053 + }, + { + "epoch": 0.06054, + "grad_norm": 0.751311503119369, + "learning_rate": 0.003, + "loss": 4.129, + "step": 6054 + }, + { + "epoch": 0.06055, + "grad_norm": 0.7759515169731848, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 6055 + }, + { + "epoch": 0.06056, + "grad_norm": 0.8914450153166513, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6056 + }, + { + "epoch": 0.06057, + "grad_norm": 1.024211037077155, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 6057 + }, + { + "epoch": 0.06058, + "grad_norm": 0.8879874628037983, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6058 + }, + { + "epoch": 0.06059, + "grad_norm": 0.6906829909411741, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6059 + }, + { + "epoch": 0.0606, + "grad_norm": 0.8104217253555417, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6060 + }, + { + "epoch": 0.06061, + "grad_norm": 0.9436958041364509, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 6061 + }, + { + "epoch": 0.06062, + "grad_norm": 0.9932174935557901, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 6062 + }, + { + "epoch": 0.06063, + "grad_norm": 0.9401163733568618, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 6063 + }, + { + "epoch": 0.06064, + "grad_norm": 0.8183782327422785, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 6064 + }, + { + "epoch": 0.06065, + "grad_norm": 0.7768072754191233, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 6065 + }, + { + "epoch": 0.06066, + "grad_norm": 0.7909830641550517, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 6066 + }, + { + "epoch": 0.06067, + "grad_norm": 0.8761823413532669, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 6067 + }, + { + "epoch": 0.06068, + "grad_norm": 0.8406889367451033, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6068 + }, + { + "epoch": 0.06069, + "grad_norm": 0.8176623171241324, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 6069 + }, + { + "epoch": 0.0607, + "grad_norm": 0.8480636571382947, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 6070 + }, + { + "epoch": 0.06071, + "grad_norm": 0.9315249540383046, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 6071 + }, + { + "epoch": 0.06072, + "grad_norm": 0.9083643027747857, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 6072 + }, + { + "epoch": 0.06073, + "grad_norm": 0.921573456841131, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 6073 + }, + { + "epoch": 0.06074, + "grad_norm": 0.9225097184432787, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6074 + }, + { + "epoch": 0.06075, + "grad_norm": 0.7677224007706127, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 6075 + }, + { + "epoch": 0.06076, + "grad_norm": 0.6736519570783548, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6076 + }, + { + "epoch": 0.06077, + "grad_norm": 0.6250416081043346, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6077 + }, + { + "epoch": 0.06078, + "grad_norm": 0.6163479466470738, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6078 + }, + { + "epoch": 0.06079, + "grad_norm": 0.5999049345196155, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 6079 + }, + { + "epoch": 0.0608, + "grad_norm": 0.601162886689798, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6080 + }, + { + "epoch": 0.06081, + "grad_norm": 0.6240043403002632, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 6081 + }, + { + "epoch": 0.06082, + "grad_norm": 0.614892243322943, + "learning_rate": 0.003, + "loss": 4.1368, + "step": 6082 + }, + { + "epoch": 0.06083, + "grad_norm": 0.6688855801850269, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 6083 + }, + { + "epoch": 0.06084, + "grad_norm": 0.7690977511826896, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 6084 + }, + { + "epoch": 0.06085, + "grad_norm": 0.8541611833462315, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6085 + }, + { + "epoch": 0.06086, + "grad_norm": 0.7865155489463524, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 6086 + }, + { + "epoch": 0.06087, + "grad_norm": 0.7600736797530553, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 6087 + }, + { + "epoch": 0.06088, + "grad_norm": 0.6616973340921765, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 6088 + }, + { + "epoch": 0.06089, + "grad_norm": 0.6692763449618849, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 6089 + }, + { + "epoch": 0.0609, + "grad_norm": 0.8696023229449554, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 6090 + }, + { + "epoch": 0.06091, + "grad_norm": 0.9918899593451429, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6091 + }, + { + "epoch": 0.06092, + "grad_norm": 1.0384619852543377, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 6092 + }, + { + "epoch": 0.06093, + "grad_norm": 1.2074123219045159, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 6093 + }, + { + "epoch": 0.06094, + "grad_norm": 0.9389995680148735, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 6094 + }, + { + "epoch": 0.06095, + "grad_norm": 1.0313863077072658, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 6095 + }, + { + "epoch": 0.06096, + "grad_norm": 0.9094326514641806, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6096 + }, + { + "epoch": 0.06097, + "grad_norm": 0.8515923116062766, + "learning_rate": 0.003, + "loss": 4.136, + "step": 6097 + }, + { + "epoch": 0.06098, + "grad_norm": 0.8412113738456533, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 6098 + }, + { + "epoch": 0.06099, + "grad_norm": 0.7500388004828196, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 6099 + }, + { + "epoch": 0.061, + "grad_norm": 0.8105097769600541, + "learning_rate": 0.003, + "loss": 4.117, + "step": 6100 + }, + { + "epoch": 0.06101, + "grad_norm": 0.9271622318300181, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 6101 + }, + { + "epoch": 0.06102, + "grad_norm": 1.3781789498192056, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 6102 + }, + { + "epoch": 0.06103, + "grad_norm": 0.7983625286469737, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 6103 + }, + { + "epoch": 0.06104, + "grad_norm": 0.7600704963358303, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 6104 + }, + { + "epoch": 0.06105, + "grad_norm": 0.7601314430280259, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 6105 + }, + { + "epoch": 0.06106, + "grad_norm": 0.6953532621943936, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 6106 + }, + { + "epoch": 0.06107, + "grad_norm": 0.5743377587831264, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 6107 + }, + { + "epoch": 0.06108, + "grad_norm": 0.5287888866406477, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6108 + }, + { + "epoch": 0.06109, + "grad_norm": 0.4973630164083057, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 6109 + }, + { + "epoch": 0.0611, + "grad_norm": 0.5158580251680166, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 6110 + }, + { + "epoch": 0.06111, + "grad_norm": 0.5630354022802069, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6111 + }, + { + "epoch": 0.06112, + "grad_norm": 0.5923550014514884, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6112 + }, + { + "epoch": 0.06113, + "grad_norm": 0.6699929210319788, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 6113 + }, + { + "epoch": 0.06114, + "grad_norm": 0.7386163212187459, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 6114 + }, + { + "epoch": 0.06115, + "grad_norm": 0.8404980906457252, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 6115 + }, + { + "epoch": 0.06116, + "grad_norm": 0.9766819519937263, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 6116 + }, + { + "epoch": 0.06117, + "grad_norm": 1.1615586503825515, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6117 + }, + { + "epoch": 0.06118, + "grad_norm": 0.8758457447215398, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6118 + }, + { + "epoch": 0.06119, + "grad_norm": 0.8049538639676183, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 6119 + }, + { + "epoch": 0.0612, + "grad_norm": 0.9210216013334553, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 6120 + }, + { + "epoch": 0.06121, + "grad_norm": 0.9600155278920769, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6121 + }, + { + "epoch": 0.06122, + "grad_norm": 0.9978512170322461, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 6122 + }, + { + "epoch": 0.06123, + "grad_norm": 0.9967240552614798, + "learning_rate": 0.003, + "loss": 4.133, + "step": 6123 + }, + { + "epoch": 0.06124, + "grad_norm": 0.974088850622343, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 6124 + }, + { + "epoch": 0.06125, + "grad_norm": 0.9600836830620103, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 6125 + }, + { + "epoch": 0.06126, + "grad_norm": 0.8700792342844844, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 6126 + }, + { + "epoch": 0.06127, + "grad_norm": 0.9271241446223905, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6127 + }, + { + "epoch": 0.06128, + "grad_norm": 0.9422494813351958, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6128 + }, + { + "epoch": 0.06129, + "grad_norm": 1.059583879974644, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 6129 + }, + { + "epoch": 0.0613, + "grad_norm": 1.0947813947090914, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 6130 + }, + { + "epoch": 0.06131, + "grad_norm": 1.040398206093915, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 6131 + }, + { + "epoch": 0.06132, + "grad_norm": 1.2039883962519413, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 6132 + }, + { + "epoch": 0.06133, + "grad_norm": 0.8566788727645377, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 6133 + }, + { + "epoch": 0.06134, + "grad_norm": 0.855848356512777, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 6134 + }, + { + "epoch": 0.06135, + "grad_norm": 0.7583857270651828, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 6135 + }, + { + "epoch": 0.06136, + "grad_norm": 0.7061925305043327, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 6136 + }, + { + "epoch": 0.06137, + "grad_norm": 0.7565804099972333, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6137 + }, + { + "epoch": 0.06138, + "grad_norm": 0.7437903267496562, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 6138 + }, + { + "epoch": 0.06139, + "grad_norm": 0.7409304538323409, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 6139 + }, + { + "epoch": 0.0614, + "grad_norm": 0.7536092482113058, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 6140 + }, + { + "epoch": 0.06141, + "grad_norm": 0.7483522220373314, + "learning_rate": 0.003, + "loss": 4.091, + "step": 6141 + }, + { + "epoch": 0.06142, + "grad_norm": 0.7859436301216659, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6142 + }, + { + "epoch": 0.06143, + "grad_norm": 0.7728075385999797, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6143 + }, + { + "epoch": 0.06144, + "grad_norm": 0.7053828491804723, + "learning_rate": 0.003, + "loss": 4.102, + "step": 6144 + }, + { + "epoch": 0.06145, + "grad_norm": 0.7192002729715705, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6145 + }, + { + "epoch": 0.06146, + "grad_norm": 0.8251909641474612, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6146 + }, + { + "epoch": 0.06147, + "grad_norm": 0.9886778253887212, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 6147 + }, + { + "epoch": 0.06148, + "grad_norm": 1.155593610839946, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 6148 + }, + { + "epoch": 0.06149, + "grad_norm": 0.8070815232350429, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 6149 + }, + { + "epoch": 0.0615, + "grad_norm": 0.5204054968874773, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 6150 + }, + { + "epoch": 0.06151, + "grad_norm": 0.6505364378426297, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6151 + }, + { + "epoch": 0.06152, + "grad_norm": 0.7431735128096151, + "learning_rate": 0.003, + "loss": 4.105, + "step": 6152 + }, + { + "epoch": 0.06153, + "grad_norm": 0.7930578549673836, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6153 + }, + { + "epoch": 0.06154, + "grad_norm": 0.6504483402063065, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6154 + }, + { + "epoch": 0.06155, + "grad_norm": 0.6346938940236132, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 6155 + }, + { + "epoch": 0.06156, + "grad_norm": 0.6953414800898144, + "learning_rate": 0.003, + "loss": 4.103, + "step": 6156 + }, + { + "epoch": 0.06157, + "grad_norm": 0.6331665507038173, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 6157 + }, + { + "epoch": 0.06158, + "grad_norm": 0.5534718559856988, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 6158 + }, + { + "epoch": 0.06159, + "grad_norm": 0.6648240189265773, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6159 + }, + { + "epoch": 0.0616, + "grad_norm": 0.8461177280021167, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6160 + }, + { + "epoch": 0.06161, + "grad_norm": 1.0823146365874887, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 6161 + }, + { + "epoch": 0.06162, + "grad_norm": 1.0938191916380575, + "learning_rate": 0.003, + "loss": 4.112, + "step": 6162 + }, + { + "epoch": 0.06163, + "grad_norm": 0.9951722082510337, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 6163 + }, + { + "epoch": 0.06164, + "grad_norm": 1.0715869196930037, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 6164 + }, + { + "epoch": 0.06165, + "grad_norm": 0.9385585217959526, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 6165 + }, + { + "epoch": 0.06166, + "grad_norm": 1.1322976552908572, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 6166 + }, + { + "epoch": 0.06167, + "grad_norm": 0.782731511695347, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6167 + }, + { + "epoch": 0.06168, + "grad_norm": 0.7067978886921109, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 6168 + }, + { + "epoch": 0.06169, + "grad_norm": 0.8159491893240778, + "learning_rate": 0.003, + "loss": 4.118, + "step": 6169 + }, + { + "epoch": 0.0617, + "grad_norm": 0.8210084830581018, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 6170 + }, + { + "epoch": 0.06171, + "grad_norm": 0.8561350609180014, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 6171 + }, + { + "epoch": 0.06172, + "grad_norm": 0.7783998258196728, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6172 + }, + { + "epoch": 0.06173, + "grad_norm": 0.7897242515178252, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 6173 + }, + { + "epoch": 0.06174, + "grad_norm": 0.8018311525812623, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 6174 + }, + { + "epoch": 0.06175, + "grad_norm": 0.8029275912295462, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6175 + }, + { + "epoch": 0.06176, + "grad_norm": 0.8180919970020917, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6176 + }, + { + "epoch": 0.06177, + "grad_norm": 0.8315943443066572, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6177 + }, + { + "epoch": 0.06178, + "grad_norm": 0.83549072750134, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 6178 + }, + { + "epoch": 0.06179, + "grad_norm": 0.7234740322597832, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6179 + }, + { + "epoch": 0.0618, + "grad_norm": 0.6996794046268644, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6180 + }, + { + "epoch": 0.06181, + "grad_norm": 0.8092448833420763, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6181 + }, + { + "epoch": 0.06182, + "grad_norm": 0.967878116319822, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 6182 + }, + { + "epoch": 0.06183, + "grad_norm": 1.2411533171982605, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6183 + }, + { + "epoch": 0.06184, + "grad_norm": 0.9159714235499898, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6184 + }, + { + "epoch": 0.06185, + "grad_norm": 0.8292810109595354, + "learning_rate": 0.003, + "loss": 4.104, + "step": 6185 + }, + { + "epoch": 0.06186, + "grad_norm": 0.730826666849849, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 6186 + }, + { + "epoch": 0.06187, + "grad_norm": 0.7722325434278592, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6187 + }, + { + "epoch": 0.06188, + "grad_norm": 0.7338999457057217, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6188 + }, + { + "epoch": 0.06189, + "grad_norm": 0.6370529217825343, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 6189 + }, + { + "epoch": 0.0619, + "grad_norm": 0.5776382912555511, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 6190 + }, + { + "epoch": 0.06191, + "grad_norm": 0.5174409949409816, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 6191 + }, + { + "epoch": 0.06192, + "grad_norm": 0.5449741549919427, + "learning_rate": 0.003, + "loss": 4.09, + "step": 6192 + }, + { + "epoch": 0.06193, + "grad_norm": 0.5306253078122207, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 6193 + }, + { + "epoch": 0.06194, + "grad_norm": 0.6667082111771107, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 6194 + }, + { + "epoch": 0.06195, + "grad_norm": 1.0558812938178088, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6195 + }, + { + "epoch": 0.06196, + "grad_norm": 1.2872949574727126, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 6196 + }, + { + "epoch": 0.06197, + "grad_norm": 0.65715222645378, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 6197 + }, + { + "epoch": 0.06198, + "grad_norm": 0.8745169993094318, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 6198 + }, + { + "epoch": 0.06199, + "grad_norm": 0.9987733008927346, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 6199 + }, + { + "epoch": 0.062, + "grad_norm": 0.9846095446431814, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 6200 + }, + { + "epoch": 0.06201, + "grad_norm": 0.9485154036502214, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 6201 + }, + { + "epoch": 0.06202, + "grad_norm": 0.8778602800304401, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 6202 + }, + { + "epoch": 0.06203, + "grad_norm": 0.8448721905084925, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 6203 + }, + { + "epoch": 0.06204, + "grad_norm": 0.8474005822451897, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 6204 + }, + { + "epoch": 0.06205, + "grad_norm": 0.8173367189925055, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 6205 + }, + { + "epoch": 0.06206, + "grad_norm": 0.7111672778214739, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 6206 + }, + { + "epoch": 0.06207, + "grad_norm": 0.7478977056662977, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 6207 + }, + { + "epoch": 0.06208, + "grad_norm": 0.7799493342189914, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 6208 + }, + { + "epoch": 0.06209, + "grad_norm": 0.8467347312009388, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 6209 + }, + { + "epoch": 0.0621, + "grad_norm": 0.9239306201168563, + "learning_rate": 0.003, + "loss": 4.1621, + "step": 6210 + }, + { + "epoch": 0.06211, + "grad_norm": 0.9596942352000194, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 6211 + }, + { + "epoch": 0.06212, + "grad_norm": 0.8948519882056875, + "learning_rate": 0.003, + "loss": 4.134, + "step": 6212 + }, + { + "epoch": 0.06213, + "grad_norm": 0.7933444432212635, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 6213 + }, + { + "epoch": 0.06214, + "grad_norm": 0.809317950973503, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 6214 + }, + { + "epoch": 0.06215, + "grad_norm": 0.876601204166201, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 6215 + }, + { + "epoch": 0.06216, + "grad_norm": 0.9644768045541879, + "learning_rate": 0.003, + "loss": 4.162, + "step": 6216 + }, + { + "epoch": 0.06217, + "grad_norm": 1.043411427164528, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 6217 + }, + { + "epoch": 0.06218, + "grad_norm": 1.072871876469517, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 6218 + }, + { + "epoch": 0.06219, + "grad_norm": 0.8571177965245572, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 6219 + }, + { + "epoch": 0.0622, + "grad_norm": 0.7413732732655652, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 6220 + }, + { + "epoch": 0.06221, + "grad_norm": 0.68817509861139, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 6221 + }, + { + "epoch": 0.06222, + "grad_norm": 0.6601184764512894, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 6222 + }, + { + "epoch": 0.06223, + "grad_norm": 0.6189770004058355, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6223 + }, + { + "epoch": 0.06224, + "grad_norm": 0.5126985323827083, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 6224 + }, + { + "epoch": 0.06225, + "grad_norm": 0.549305548358553, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6225 + }, + { + "epoch": 0.06226, + "grad_norm": 0.6038854471945458, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 6226 + }, + { + "epoch": 0.06227, + "grad_norm": 0.7264264034452508, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6227 + }, + { + "epoch": 0.06228, + "grad_norm": 0.8139873862590992, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6228 + }, + { + "epoch": 0.06229, + "grad_norm": 0.9132950691176969, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 6229 + }, + { + "epoch": 0.0623, + "grad_norm": 0.8608762645557783, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6230 + }, + { + "epoch": 0.06231, + "grad_norm": 0.7323701212842743, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 6231 + }, + { + "epoch": 0.06232, + "grad_norm": 0.7533201360665337, + "learning_rate": 0.003, + "loss": 4.097, + "step": 6232 + }, + { + "epoch": 0.06233, + "grad_norm": 0.7962432286199019, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 6233 + }, + { + "epoch": 0.06234, + "grad_norm": 0.8001002251441555, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6234 + }, + { + "epoch": 0.06235, + "grad_norm": 0.8282187152840139, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 6235 + }, + { + "epoch": 0.06236, + "grad_norm": 0.8529888372634628, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 6236 + }, + { + "epoch": 0.06237, + "grad_norm": 0.8924649306049437, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6237 + }, + { + "epoch": 0.06238, + "grad_norm": 0.9517024598895053, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 6238 + }, + { + "epoch": 0.06239, + "grad_norm": 1.090896941369981, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 6239 + }, + { + "epoch": 0.0624, + "grad_norm": 0.8988025177459298, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 6240 + }, + { + "epoch": 0.06241, + "grad_norm": 0.8863798221813798, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6241 + }, + { + "epoch": 0.06242, + "grad_norm": 1.0387484289967834, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 6242 + }, + { + "epoch": 0.06243, + "grad_norm": 0.8979931905156554, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6243 + }, + { + "epoch": 0.06244, + "grad_norm": 0.8356118354466084, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 6244 + }, + { + "epoch": 0.06245, + "grad_norm": 0.8502805150632947, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 6245 + }, + { + "epoch": 0.06246, + "grad_norm": 0.9834636405295031, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 6246 + }, + { + "epoch": 0.06247, + "grad_norm": 1.1088932528341893, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 6247 + }, + { + "epoch": 0.06248, + "grad_norm": 1.018053258070267, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 6248 + }, + { + "epoch": 0.06249, + "grad_norm": 0.9988289602301562, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 6249 + }, + { + "epoch": 0.0625, + "grad_norm": 0.9321492794691144, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 6250 + }, + { + "epoch": 0.06251, + "grad_norm": 0.9254351424798546, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 6251 + }, + { + "epoch": 0.06252, + "grad_norm": 0.9487439513448277, + "learning_rate": 0.003, + "loss": 4.1425, + "step": 6252 + }, + { + "epoch": 0.06253, + "grad_norm": 0.8766138636695966, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 6253 + }, + { + "epoch": 0.06254, + "grad_norm": 0.8485254787917414, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 6254 + }, + { + "epoch": 0.06255, + "grad_norm": 0.7291384807358328, + "learning_rate": 0.003, + "loss": 4.131, + "step": 6255 + }, + { + "epoch": 0.06256, + "grad_norm": 0.7872134593442462, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 6256 + }, + { + "epoch": 0.06257, + "grad_norm": 0.9363917740480254, + "learning_rate": 0.003, + "loss": 4.086, + "step": 6257 + }, + { + "epoch": 0.06258, + "grad_norm": 1.0738791582349567, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 6258 + }, + { + "epoch": 0.06259, + "grad_norm": 1.213752532113539, + "learning_rate": 0.003, + "loss": 4.17, + "step": 6259 + }, + { + "epoch": 0.0626, + "grad_norm": 0.8439975974645768, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 6260 + }, + { + "epoch": 0.06261, + "grad_norm": 0.7967008971115906, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6261 + }, + { + "epoch": 0.06262, + "grad_norm": 0.8203007205386644, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6262 + }, + { + "epoch": 0.06263, + "grad_norm": 0.7563771758779613, + "learning_rate": 0.003, + "loss": 4.1435, + "step": 6263 + }, + { + "epoch": 0.06264, + "grad_norm": 0.868641577107568, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6264 + }, + { + "epoch": 0.06265, + "grad_norm": 0.8732294846971428, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 6265 + }, + { + "epoch": 0.06266, + "grad_norm": 0.7353975928746133, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6266 + }, + { + "epoch": 0.06267, + "grad_norm": 0.7283040841735945, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6267 + }, + { + "epoch": 0.06268, + "grad_norm": 0.6976022361269695, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 6268 + }, + { + "epoch": 0.06269, + "grad_norm": 0.5944661023232062, + "learning_rate": 0.003, + "loss": 4.094, + "step": 6269 + }, + { + "epoch": 0.0627, + "grad_norm": 0.6219309338060965, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 6270 + }, + { + "epoch": 0.06271, + "grad_norm": 0.6423109090656182, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 6271 + }, + { + "epoch": 0.06272, + "grad_norm": 0.6497521209683546, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 6272 + }, + { + "epoch": 0.06273, + "grad_norm": 0.6408710867659153, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 6273 + }, + { + "epoch": 0.06274, + "grad_norm": 0.6192887789752485, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6274 + }, + { + "epoch": 0.06275, + "grad_norm": 0.6261230406806092, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 6275 + }, + { + "epoch": 0.06276, + "grad_norm": 0.5976816107591381, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 6276 + }, + { + "epoch": 0.06277, + "grad_norm": 0.566538285689989, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 6277 + }, + { + "epoch": 0.06278, + "grad_norm": 0.6008204917237797, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6278 + }, + { + "epoch": 0.06279, + "grad_norm": 0.7578953746913875, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 6279 + }, + { + "epoch": 0.0628, + "grad_norm": 0.9512660338098725, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 6280 + }, + { + "epoch": 0.06281, + "grad_norm": 1.441612819184275, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6281 + }, + { + "epoch": 0.06282, + "grad_norm": 0.9355215042556002, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 6282 + }, + { + "epoch": 0.06283, + "grad_norm": 0.9775330592179509, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6283 + }, + { + "epoch": 0.06284, + "grad_norm": 1.152004214068161, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 6284 + }, + { + "epoch": 0.06285, + "grad_norm": 0.8632280844541238, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 6285 + }, + { + "epoch": 0.06286, + "grad_norm": 0.8688132213283226, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 6286 + }, + { + "epoch": 0.06287, + "grad_norm": 0.7521943262262261, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 6287 + }, + { + "epoch": 0.06288, + "grad_norm": 0.7590966433142204, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 6288 + }, + { + "epoch": 0.06289, + "grad_norm": 0.8547449463965621, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 6289 + }, + { + "epoch": 0.0629, + "grad_norm": 0.995426203878218, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6290 + }, + { + "epoch": 0.06291, + "grad_norm": 1.0501998061136046, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6291 + }, + { + "epoch": 0.06292, + "grad_norm": 0.8258937115933653, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6292 + }, + { + "epoch": 0.06293, + "grad_norm": 0.744681874493206, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 6293 + }, + { + "epoch": 0.06294, + "grad_norm": 0.6844598828137828, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 6294 + }, + { + "epoch": 0.06295, + "grad_norm": 0.709427198581793, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6295 + }, + { + "epoch": 0.06296, + "grad_norm": 0.7742480202450152, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 6296 + }, + { + "epoch": 0.06297, + "grad_norm": 0.9065810778029665, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 6297 + }, + { + "epoch": 0.06298, + "grad_norm": 1.0205749038805025, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6298 + }, + { + "epoch": 0.06299, + "grad_norm": 1.0014978552686065, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6299 + }, + { + "epoch": 0.063, + "grad_norm": 0.9906041215988247, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 6300 + }, + { + "epoch": 0.06301, + "grad_norm": 1.0237044308753132, + "learning_rate": 0.003, + "loss": 4.151, + "step": 6301 + }, + { + "epoch": 0.06302, + "grad_norm": 0.8716511028680038, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6302 + }, + { + "epoch": 0.06303, + "grad_norm": 0.9180601021684964, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 6303 + }, + { + "epoch": 0.06304, + "grad_norm": 0.8595930720169722, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 6304 + }, + { + "epoch": 0.06305, + "grad_norm": 0.8891037067579485, + "learning_rate": 0.003, + "loss": 4.1561, + "step": 6305 + }, + { + "epoch": 0.06306, + "grad_norm": 0.835929826094222, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 6306 + }, + { + "epoch": 0.06307, + "grad_norm": 0.8222800944063304, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 6307 + }, + { + "epoch": 0.06308, + "grad_norm": 0.7978612799259077, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 6308 + }, + { + "epoch": 0.06309, + "grad_norm": 0.8169417224723879, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 6309 + }, + { + "epoch": 0.0631, + "grad_norm": 0.7885494048599513, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 6310 + }, + { + "epoch": 0.06311, + "grad_norm": 0.7141125014011859, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 6311 + }, + { + "epoch": 0.06312, + "grad_norm": 0.8834006613536179, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6312 + }, + { + "epoch": 0.06313, + "grad_norm": 1.0782426771166267, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6313 + }, + { + "epoch": 0.06314, + "grad_norm": 0.9270596290945304, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 6314 + }, + { + "epoch": 0.06315, + "grad_norm": 0.866614344975497, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 6315 + }, + { + "epoch": 0.06316, + "grad_norm": 0.8025239448074284, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 6316 + }, + { + "epoch": 0.06317, + "grad_norm": 0.7483988975313902, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6317 + }, + { + "epoch": 0.06318, + "grad_norm": 0.7026343077048057, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 6318 + }, + { + "epoch": 0.06319, + "grad_norm": 0.6258340242384319, + "learning_rate": 0.003, + "loss": 4.1488, + "step": 6319 + }, + { + "epoch": 0.0632, + "grad_norm": 0.6022905615388534, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 6320 + }, + { + "epoch": 0.06321, + "grad_norm": 0.6537192595187358, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 6321 + }, + { + "epoch": 0.06322, + "grad_norm": 0.7679534667703762, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 6322 + }, + { + "epoch": 0.06323, + "grad_norm": 0.9777865112549871, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6323 + }, + { + "epoch": 0.06324, + "grad_norm": 1.1128170132105777, + "learning_rate": 0.003, + "loss": 4.133, + "step": 6324 + }, + { + "epoch": 0.06325, + "grad_norm": 0.9056105963919202, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 6325 + }, + { + "epoch": 0.06326, + "grad_norm": 0.9396881242509516, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 6326 + }, + { + "epoch": 0.06327, + "grad_norm": 0.9848953169243714, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6327 + }, + { + "epoch": 0.06328, + "grad_norm": 1.1839671933476679, + "learning_rate": 0.003, + "loss": 4.1609, + "step": 6328 + }, + { + "epoch": 0.06329, + "grad_norm": 0.7896842525783743, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6329 + }, + { + "epoch": 0.0633, + "grad_norm": 0.7869900869544297, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 6330 + }, + { + "epoch": 0.06331, + "grad_norm": 0.7444958476349693, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6331 + }, + { + "epoch": 0.06332, + "grad_norm": 0.7071721517704226, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 6332 + }, + { + "epoch": 0.06333, + "grad_norm": 0.6531451777187667, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 6333 + }, + { + "epoch": 0.06334, + "grad_norm": 0.5871978280186811, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6334 + }, + { + "epoch": 0.06335, + "grad_norm": 0.5799498876324073, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6335 + }, + { + "epoch": 0.06336, + "grad_norm": 0.5706118531443688, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 6336 + }, + { + "epoch": 0.06337, + "grad_norm": 0.7146855029967338, + "learning_rate": 0.003, + "loss": 4.132, + "step": 6337 + }, + { + "epoch": 0.06338, + "grad_norm": 0.9100626823469105, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 6338 + }, + { + "epoch": 0.06339, + "grad_norm": 1.2461283232196485, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 6339 + }, + { + "epoch": 0.0634, + "grad_norm": 0.8399519595087078, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 6340 + }, + { + "epoch": 0.06341, + "grad_norm": 0.6934313188966001, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 6341 + }, + { + "epoch": 0.06342, + "grad_norm": 0.7330598162651885, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6342 + }, + { + "epoch": 0.06343, + "grad_norm": 0.8671585484671817, + "learning_rate": 0.003, + "loss": 4.129, + "step": 6343 + }, + { + "epoch": 0.06344, + "grad_norm": 0.8753238271910776, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6344 + }, + { + "epoch": 0.06345, + "grad_norm": 0.880493691893159, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6345 + }, + { + "epoch": 0.06346, + "grad_norm": 0.8605398856134917, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6346 + }, + { + "epoch": 0.06347, + "grad_norm": 0.7991693733841341, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 6347 + }, + { + "epoch": 0.06348, + "grad_norm": 0.8445072089673568, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6348 + }, + { + "epoch": 0.06349, + "grad_norm": 1.013610824242872, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6349 + }, + { + "epoch": 0.0635, + "grad_norm": 1.1525443266019426, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 6350 + }, + { + "epoch": 0.06351, + "grad_norm": 0.9068123910947427, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 6351 + }, + { + "epoch": 0.06352, + "grad_norm": 0.9267793019839832, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 6352 + }, + { + "epoch": 0.06353, + "grad_norm": 0.9507357268589157, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6353 + }, + { + "epoch": 0.06354, + "grad_norm": 0.8933158322283256, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6354 + }, + { + "epoch": 0.06355, + "grad_norm": 0.8410685705766945, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 6355 + }, + { + "epoch": 0.06356, + "grad_norm": 0.8005298650181658, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6356 + }, + { + "epoch": 0.06357, + "grad_norm": 0.9221272355695364, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 6357 + }, + { + "epoch": 0.06358, + "grad_norm": 0.9028874971232729, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 6358 + }, + { + "epoch": 0.06359, + "grad_norm": 0.8042560641443137, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6359 + }, + { + "epoch": 0.0636, + "grad_norm": 0.9188716749774963, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 6360 + }, + { + "epoch": 0.06361, + "grad_norm": 0.95062392680164, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 6361 + }, + { + "epoch": 0.06362, + "grad_norm": 1.0546076113357394, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 6362 + }, + { + "epoch": 0.06363, + "grad_norm": 0.9203663169616263, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6363 + }, + { + "epoch": 0.06364, + "grad_norm": 0.7499435174266673, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 6364 + }, + { + "epoch": 0.06365, + "grad_norm": 0.6331693994024129, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 6365 + }, + { + "epoch": 0.06366, + "grad_norm": 0.5836138837711193, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 6366 + }, + { + "epoch": 0.06367, + "grad_norm": 0.5905679268129659, + "learning_rate": 0.003, + "loss": 4.102, + "step": 6367 + }, + { + "epoch": 0.06368, + "grad_norm": 0.5546690802747907, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 6368 + }, + { + "epoch": 0.06369, + "grad_norm": 0.5409525915376348, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 6369 + }, + { + "epoch": 0.0637, + "grad_norm": 0.6041426965862612, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 6370 + }, + { + "epoch": 0.06371, + "grad_norm": 0.6134926672713396, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 6371 + }, + { + "epoch": 0.06372, + "grad_norm": 0.5585920732749725, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 6372 + }, + { + "epoch": 0.06373, + "grad_norm": 0.5617821036929576, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 6373 + }, + { + "epoch": 0.06374, + "grad_norm": 0.5884001712501294, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6374 + }, + { + "epoch": 0.06375, + "grad_norm": 0.5836596540025951, + "learning_rate": 0.003, + "loss": 4.082, + "step": 6375 + }, + { + "epoch": 0.06376, + "grad_norm": 0.6350421742417018, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 6376 + }, + { + "epoch": 0.06377, + "grad_norm": 0.7167562203574344, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 6377 + }, + { + "epoch": 0.06378, + "grad_norm": 0.8951058433252564, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 6378 + }, + { + "epoch": 0.06379, + "grad_norm": 1.3635620798766852, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 6379 + }, + { + "epoch": 0.0638, + "grad_norm": 1.2248674028316016, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 6380 + }, + { + "epoch": 0.06381, + "grad_norm": 0.9100897202260229, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6381 + }, + { + "epoch": 0.06382, + "grad_norm": 0.8036704486012445, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 6382 + }, + { + "epoch": 0.06383, + "grad_norm": 0.743174482288171, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6383 + }, + { + "epoch": 0.06384, + "grad_norm": 0.8513937818435598, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6384 + }, + { + "epoch": 0.06385, + "grad_norm": 1.0786453242708027, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 6385 + }, + { + "epoch": 0.06386, + "grad_norm": 1.0238890384448516, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 6386 + }, + { + "epoch": 0.06387, + "grad_norm": 1.027593845316165, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6387 + }, + { + "epoch": 0.06388, + "grad_norm": 0.839413084100022, + "learning_rate": 0.003, + "loss": 4.104, + "step": 6388 + }, + { + "epoch": 0.06389, + "grad_norm": 0.7756185982524284, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 6389 + }, + { + "epoch": 0.0639, + "grad_norm": 0.8713850515068945, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 6390 + }, + { + "epoch": 0.06391, + "grad_norm": 1.0408467215495305, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 6391 + }, + { + "epoch": 0.06392, + "grad_norm": 1.0034083196993608, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 6392 + }, + { + "epoch": 0.06393, + "grad_norm": 0.8723434890415933, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 6393 + }, + { + "epoch": 0.06394, + "grad_norm": 0.8362827621512328, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 6394 + }, + { + "epoch": 0.06395, + "grad_norm": 1.0579394140223388, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 6395 + }, + { + "epoch": 0.06396, + "grad_norm": 1.0118284529885586, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 6396 + }, + { + "epoch": 0.06397, + "grad_norm": 1.0625315745598696, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 6397 + }, + { + "epoch": 0.06398, + "grad_norm": 0.9743429452862477, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 6398 + }, + { + "epoch": 0.06399, + "grad_norm": 0.9436595857616809, + "learning_rate": 0.003, + "loss": 4.1394, + "step": 6399 + }, + { + "epoch": 0.064, + "grad_norm": 1.018780717912782, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 6400 + }, + { + "epoch": 0.06401, + "grad_norm": 1.0102187640595444, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 6401 + }, + { + "epoch": 0.06402, + "grad_norm": 0.9216577375444452, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 6402 + }, + { + "epoch": 0.06403, + "grad_norm": 0.7975560893939748, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 6403 + }, + { + "epoch": 0.06404, + "grad_norm": 0.7840951281577243, + "learning_rate": 0.003, + "loss": 4.167, + "step": 6404 + }, + { + "epoch": 0.06405, + "grad_norm": 0.8211264085131749, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6405 + }, + { + "epoch": 0.06406, + "grad_norm": 0.8035929209152411, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 6406 + }, + { + "epoch": 0.06407, + "grad_norm": 0.8175467906216495, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 6407 + }, + { + "epoch": 0.06408, + "grad_norm": 1.001836035975777, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6408 + }, + { + "epoch": 0.06409, + "grad_norm": 1.2009899389705134, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 6409 + }, + { + "epoch": 0.0641, + "grad_norm": 0.8972232960554846, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 6410 + }, + { + "epoch": 0.06411, + "grad_norm": 0.8367366713616154, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 6411 + }, + { + "epoch": 0.06412, + "grad_norm": 0.8972350190353556, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 6412 + }, + { + "epoch": 0.06413, + "grad_norm": 1.0400706852657329, + "learning_rate": 0.003, + "loss": 4.16, + "step": 6413 + }, + { + "epoch": 0.06414, + "grad_norm": 1.020312511829939, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6414 + }, + { + "epoch": 0.06415, + "grad_norm": 1.028531990172739, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6415 + }, + { + "epoch": 0.06416, + "grad_norm": 0.8253303199298359, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 6416 + }, + { + "epoch": 0.06417, + "grad_norm": 0.6680387166207186, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 6417 + }, + { + "epoch": 0.06418, + "grad_norm": 0.6168723571778798, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6418 + }, + { + "epoch": 0.06419, + "grad_norm": 0.5368546076983477, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6419 + }, + { + "epoch": 0.0642, + "grad_norm": 0.6584077181684485, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 6420 + }, + { + "epoch": 0.06421, + "grad_norm": 0.8212576516224894, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 6421 + }, + { + "epoch": 0.06422, + "grad_norm": 1.1150116071999399, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6422 + }, + { + "epoch": 0.06423, + "grad_norm": 0.9785393710656575, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 6423 + }, + { + "epoch": 0.06424, + "grad_norm": 0.84840591881521, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6424 + }, + { + "epoch": 0.06425, + "grad_norm": 0.7746501989148333, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6425 + }, + { + "epoch": 0.06426, + "grad_norm": 0.7476002784072737, + "learning_rate": 0.003, + "loss": 4.08, + "step": 6426 + }, + { + "epoch": 0.06427, + "grad_norm": 0.7423456692827792, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 6427 + }, + { + "epoch": 0.06428, + "grad_norm": 0.7587323475557202, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6428 + }, + { + "epoch": 0.06429, + "grad_norm": 0.746195830101206, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6429 + }, + { + "epoch": 0.0643, + "grad_norm": 0.8255820271958185, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 6430 + }, + { + "epoch": 0.06431, + "grad_norm": 0.819510641482716, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 6431 + }, + { + "epoch": 0.06432, + "grad_norm": 0.6815867999333811, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 6432 + }, + { + "epoch": 0.06433, + "grad_norm": 0.6178051256251746, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 6433 + }, + { + "epoch": 0.06434, + "grad_norm": 0.6523329798252133, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6434 + }, + { + "epoch": 0.06435, + "grad_norm": 0.6691394409492107, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 6435 + }, + { + "epoch": 0.06436, + "grad_norm": 0.6367136162669638, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 6436 + }, + { + "epoch": 0.06437, + "grad_norm": 0.7151244539764642, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 6437 + }, + { + "epoch": 0.06438, + "grad_norm": 0.9303185507422409, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 6438 + }, + { + "epoch": 0.06439, + "grad_norm": 1.155547956758966, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 6439 + }, + { + "epoch": 0.0644, + "grad_norm": 0.8106227904496509, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 6440 + }, + { + "epoch": 0.06441, + "grad_norm": 0.6604648817786103, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 6441 + }, + { + "epoch": 0.06442, + "grad_norm": 0.7373200154591052, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 6442 + }, + { + "epoch": 0.06443, + "grad_norm": 0.8450728654117171, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 6443 + }, + { + "epoch": 0.06444, + "grad_norm": 1.0151152080080585, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 6444 + }, + { + "epoch": 0.06445, + "grad_norm": 1.0956617337515981, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6445 + }, + { + "epoch": 0.06446, + "grad_norm": 0.8878041786727323, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 6446 + }, + { + "epoch": 0.06447, + "grad_norm": 0.9130724650918877, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 6447 + }, + { + "epoch": 0.06448, + "grad_norm": 0.9311141225976478, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 6448 + }, + { + "epoch": 0.06449, + "grad_norm": 0.9266467844539056, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 6449 + }, + { + "epoch": 0.0645, + "grad_norm": 0.8842097985531295, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 6450 + }, + { + "epoch": 0.06451, + "grad_norm": 0.9231514896987845, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 6451 + }, + { + "epoch": 0.06452, + "grad_norm": 1.0541860942458612, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 6452 + }, + { + "epoch": 0.06453, + "grad_norm": 0.9145490815901766, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 6453 + }, + { + "epoch": 0.06454, + "grad_norm": 0.826571008490979, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6454 + }, + { + "epoch": 0.06455, + "grad_norm": 0.7020677977773648, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 6455 + }, + { + "epoch": 0.06456, + "grad_norm": 0.6693276104354243, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 6456 + }, + { + "epoch": 0.06457, + "grad_norm": 0.7996629696131489, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6457 + }, + { + "epoch": 0.06458, + "grad_norm": 0.9270705106073829, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6458 + }, + { + "epoch": 0.06459, + "grad_norm": 1.074497828894076, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 6459 + }, + { + "epoch": 0.0646, + "grad_norm": 0.9551554373739378, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6460 + }, + { + "epoch": 0.06461, + "grad_norm": 0.8033640397123838, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 6461 + }, + { + "epoch": 0.06462, + "grad_norm": 0.8571886698508542, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 6462 + }, + { + "epoch": 0.06463, + "grad_norm": 0.8476034867616584, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 6463 + }, + { + "epoch": 0.06464, + "grad_norm": 0.9308796780668124, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 6464 + }, + { + "epoch": 0.06465, + "grad_norm": 0.8577677132113817, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 6465 + }, + { + "epoch": 0.06466, + "grad_norm": 0.9685221965234145, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6466 + }, + { + "epoch": 0.06467, + "grad_norm": 0.947967210756858, + "learning_rate": 0.003, + "loss": 4.104, + "step": 6467 + }, + { + "epoch": 0.06468, + "grad_norm": 1.2313807236375918, + "learning_rate": 0.003, + "loss": 4.138, + "step": 6468 + }, + { + "epoch": 0.06469, + "grad_norm": 1.013089232311945, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 6469 + }, + { + "epoch": 0.0647, + "grad_norm": 0.7873532128436814, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6470 + }, + { + "epoch": 0.06471, + "grad_norm": 0.7728490694203982, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 6471 + }, + { + "epoch": 0.06472, + "grad_norm": 0.8569138079643775, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 6472 + }, + { + "epoch": 0.06473, + "grad_norm": 0.8903789705045198, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 6473 + }, + { + "epoch": 0.06474, + "grad_norm": 0.8414339883577647, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 6474 + }, + { + "epoch": 0.06475, + "grad_norm": 0.8185119706104915, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 6475 + }, + { + "epoch": 0.06476, + "grad_norm": 0.7735489074806816, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 6476 + }, + { + "epoch": 0.06477, + "grad_norm": 0.692541672437415, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 6477 + }, + { + "epoch": 0.06478, + "grad_norm": 0.6687914180889757, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 6478 + }, + { + "epoch": 0.06479, + "grad_norm": 0.6777500505762841, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6479 + }, + { + "epoch": 0.0648, + "grad_norm": 0.6486993051596719, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 6480 + }, + { + "epoch": 0.06481, + "grad_norm": 0.6227452111637245, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 6481 + }, + { + "epoch": 0.06482, + "grad_norm": 0.6060284390551969, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6482 + }, + { + "epoch": 0.06483, + "grad_norm": 0.7307292793303055, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 6483 + }, + { + "epoch": 0.06484, + "grad_norm": 0.8150760871337055, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 6484 + }, + { + "epoch": 0.06485, + "grad_norm": 0.9208909630123016, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 6485 + }, + { + "epoch": 0.06486, + "grad_norm": 1.1799194145477874, + "learning_rate": 0.003, + "loss": 4.14, + "step": 6486 + }, + { + "epoch": 0.06487, + "grad_norm": 0.8822420101848714, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6487 + }, + { + "epoch": 0.06488, + "grad_norm": 0.8457525546814305, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 6488 + }, + { + "epoch": 0.06489, + "grad_norm": 0.8600818951575393, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 6489 + }, + { + "epoch": 0.0649, + "grad_norm": 0.8039264767899832, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6490 + }, + { + "epoch": 0.06491, + "grad_norm": 0.8760540483969935, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 6491 + }, + { + "epoch": 0.06492, + "grad_norm": 0.82749202495857, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 6492 + }, + { + "epoch": 0.06493, + "grad_norm": 0.8943255943541918, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 6493 + }, + { + "epoch": 0.06494, + "grad_norm": 0.9825172024300408, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 6494 + }, + { + "epoch": 0.06495, + "grad_norm": 0.9527896010651128, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6495 + }, + { + "epoch": 0.06496, + "grad_norm": 0.8119745595177673, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 6496 + }, + { + "epoch": 0.06497, + "grad_norm": 0.6621216446190086, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 6497 + }, + { + "epoch": 0.06498, + "grad_norm": 0.6064887728246845, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 6498 + }, + { + "epoch": 0.06499, + "grad_norm": 0.5707453577641534, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6499 + }, + { + "epoch": 0.065, + "grad_norm": 0.5073270473104914, + "learning_rate": 0.003, + "loss": 4.095, + "step": 6500 + }, + { + "epoch": 0.06501, + "grad_norm": 0.5208394304660492, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 6501 + }, + { + "epoch": 0.06502, + "grad_norm": 0.5859156335669509, + "learning_rate": 0.003, + "loss": 4.098, + "step": 6502 + }, + { + "epoch": 0.06503, + "grad_norm": 0.6319278647879405, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6503 + }, + { + "epoch": 0.06504, + "grad_norm": 0.8345350267691076, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 6504 + }, + { + "epoch": 0.06505, + "grad_norm": 1.022594666565787, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 6505 + }, + { + "epoch": 0.06506, + "grad_norm": 1.1082118647709895, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 6506 + }, + { + "epoch": 0.06507, + "grad_norm": 0.9631150466533694, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 6507 + }, + { + "epoch": 0.06508, + "grad_norm": 0.9358226724585814, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 6508 + }, + { + "epoch": 0.06509, + "grad_norm": 0.903320053092754, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 6509 + }, + { + "epoch": 0.0651, + "grad_norm": 0.7988134798037461, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 6510 + }, + { + "epoch": 0.06511, + "grad_norm": 1.0115086044546666, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 6511 + }, + { + "epoch": 0.06512, + "grad_norm": 1.1339344083529093, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6512 + }, + { + "epoch": 0.06513, + "grad_norm": 0.9322613192351319, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6513 + }, + { + "epoch": 0.06514, + "grad_norm": 1.0489869510777938, + "learning_rate": 0.003, + "loss": 4.116, + "step": 6514 + }, + { + "epoch": 0.06515, + "grad_norm": 0.8551726865667978, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6515 + }, + { + "epoch": 0.06516, + "grad_norm": 0.8303598702103568, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6516 + }, + { + "epoch": 0.06517, + "grad_norm": 0.9044563142027361, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 6517 + }, + { + "epoch": 0.06518, + "grad_norm": 0.8703001988372849, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 6518 + }, + { + "epoch": 0.06519, + "grad_norm": 0.7847889819210776, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 6519 + }, + { + "epoch": 0.0652, + "grad_norm": 0.7604897553062402, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6520 + }, + { + "epoch": 0.06521, + "grad_norm": 0.7045491663860287, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 6521 + }, + { + "epoch": 0.06522, + "grad_norm": 0.5967359913316697, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6522 + }, + { + "epoch": 0.06523, + "grad_norm": 0.5994954551174616, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 6523 + }, + { + "epoch": 0.06524, + "grad_norm": 0.5929713355650309, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 6524 + }, + { + "epoch": 0.06525, + "grad_norm": 0.5725345126769195, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 6525 + }, + { + "epoch": 0.06526, + "grad_norm": 0.6769384736460269, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 6526 + }, + { + "epoch": 0.06527, + "grad_norm": 0.9123198261221063, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 6527 + }, + { + "epoch": 0.06528, + "grad_norm": 1.297782323536123, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 6528 + }, + { + "epoch": 0.06529, + "grad_norm": 0.9911311317864455, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 6529 + }, + { + "epoch": 0.0653, + "grad_norm": 1.0305368070734895, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 6530 + }, + { + "epoch": 0.06531, + "grad_norm": 0.9401086376942769, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 6531 + }, + { + "epoch": 0.06532, + "grad_norm": 0.9614097857861146, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 6532 + }, + { + "epoch": 0.06533, + "grad_norm": 1.0868494247089455, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 6533 + }, + { + "epoch": 0.06534, + "grad_norm": 1.0009233535813384, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 6534 + }, + { + "epoch": 0.06535, + "grad_norm": 1.0081228563907538, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 6535 + }, + { + "epoch": 0.06536, + "grad_norm": 0.9645958863688807, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6536 + }, + { + "epoch": 0.06537, + "grad_norm": 0.8769309680301732, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 6537 + }, + { + "epoch": 0.06538, + "grad_norm": 0.8137137300201961, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 6538 + }, + { + "epoch": 0.06539, + "grad_norm": 0.727085328822687, + "learning_rate": 0.003, + "loss": 4.096, + "step": 6539 + }, + { + "epoch": 0.0654, + "grad_norm": 0.7005089827719883, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 6540 + }, + { + "epoch": 0.06541, + "grad_norm": 0.6228763105598434, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 6541 + }, + { + "epoch": 0.06542, + "grad_norm": 0.7516112721925631, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 6542 + }, + { + "epoch": 0.06543, + "grad_norm": 1.0509494597212574, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 6543 + }, + { + "epoch": 0.06544, + "grad_norm": 1.1940069365357546, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 6544 + }, + { + "epoch": 0.06545, + "grad_norm": 0.8150174480762931, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 6545 + }, + { + "epoch": 0.06546, + "grad_norm": 0.9408126901018795, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6546 + }, + { + "epoch": 0.06547, + "grad_norm": 0.985809825256504, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6547 + }, + { + "epoch": 0.06548, + "grad_norm": 1.016850972033843, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 6548 + }, + { + "epoch": 0.06549, + "grad_norm": 0.9915312968382956, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 6549 + }, + { + "epoch": 0.0655, + "grad_norm": 0.7584853863997532, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 6550 + }, + { + "epoch": 0.06551, + "grad_norm": 0.7532414275895373, + "learning_rate": 0.003, + "loss": 4.094, + "step": 6551 + }, + { + "epoch": 0.06552, + "grad_norm": 0.7132878191635362, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 6552 + }, + { + "epoch": 0.06553, + "grad_norm": 0.7251259028750098, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 6553 + }, + { + "epoch": 0.06554, + "grad_norm": 0.6869756855220046, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6554 + }, + { + "epoch": 0.06555, + "grad_norm": 0.6789180373444202, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 6555 + }, + { + "epoch": 0.06556, + "grad_norm": 0.7688068758022717, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 6556 + }, + { + "epoch": 0.06557, + "grad_norm": 0.820506804409578, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 6557 + }, + { + "epoch": 0.06558, + "grad_norm": 1.0451835481907836, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 6558 + }, + { + "epoch": 0.06559, + "grad_norm": 1.0506338412291107, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 6559 + }, + { + "epoch": 0.0656, + "grad_norm": 0.8227023399945246, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6560 + }, + { + "epoch": 0.06561, + "grad_norm": 0.8263825289183352, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 6561 + }, + { + "epoch": 0.06562, + "grad_norm": 0.9118099179127191, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 6562 + }, + { + "epoch": 0.06563, + "grad_norm": 1.008868981921078, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 6563 + }, + { + "epoch": 0.06564, + "grad_norm": 0.8794384654321471, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6564 + }, + { + "epoch": 0.06565, + "grad_norm": 0.7799483866337282, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 6565 + }, + { + "epoch": 0.06566, + "grad_norm": 0.7247305841084201, + "learning_rate": 0.003, + "loss": 4.115, + "step": 6566 + }, + { + "epoch": 0.06567, + "grad_norm": 0.7781944614519534, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 6567 + }, + { + "epoch": 0.06568, + "grad_norm": 0.8470435504646778, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 6568 + }, + { + "epoch": 0.06569, + "grad_norm": 0.8975263588278858, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6569 + }, + { + "epoch": 0.0657, + "grad_norm": 0.9290752582286405, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 6570 + }, + { + "epoch": 0.06571, + "grad_norm": 0.9170887562858809, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6571 + }, + { + "epoch": 0.06572, + "grad_norm": 0.9006929422375534, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 6572 + }, + { + "epoch": 0.06573, + "grad_norm": 0.9322962718617643, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6573 + }, + { + "epoch": 0.06574, + "grad_norm": 0.7833680952020777, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 6574 + }, + { + "epoch": 0.06575, + "grad_norm": 0.7445025969000303, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 6575 + }, + { + "epoch": 0.06576, + "grad_norm": 0.7569699043405889, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 6576 + }, + { + "epoch": 0.06577, + "grad_norm": 0.9003358954292493, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 6577 + }, + { + "epoch": 0.06578, + "grad_norm": 1.0621423432401027, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 6578 + }, + { + "epoch": 0.06579, + "grad_norm": 1.1726402516475503, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 6579 + }, + { + "epoch": 0.0658, + "grad_norm": 0.8552429990963719, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 6580 + }, + { + "epoch": 0.06581, + "grad_norm": 0.8514271483398563, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 6581 + }, + { + "epoch": 0.06582, + "grad_norm": 0.9602464463060446, + "learning_rate": 0.003, + "loss": 4.13, + "step": 6582 + }, + { + "epoch": 0.06583, + "grad_norm": 0.9829530100461856, + "learning_rate": 0.003, + "loss": 4.101, + "step": 6583 + }, + { + "epoch": 0.06584, + "grad_norm": 0.9171258754858127, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 6584 + }, + { + "epoch": 0.06585, + "grad_norm": 0.84935508857, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6585 + }, + { + "epoch": 0.06586, + "grad_norm": 0.8705786102009245, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 6586 + }, + { + "epoch": 0.06587, + "grad_norm": 0.9714069205255046, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 6587 + }, + { + "epoch": 0.06588, + "grad_norm": 0.9127013611939936, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 6588 + }, + { + "epoch": 0.06589, + "grad_norm": 0.864633992723417, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 6589 + }, + { + "epoch": 0.0659, + "grad_norm": 0.7734108219438173, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 6590 + }, + { + "epoch": 0.06591, + "grad_norm": 0.7864629956935438, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 6591 + }, + { + "epoch": 0.06592, + "grad_norm": 0.7865367848858412, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 6592 + }, + { + "epoch": 0.06593, + "grad_norm": 0.795726928953009, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6593 + }, + { + "epoch": 0.06594, + "grad_norm": 0.9228436883463198, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 6594 + }, + { + "epoch": 0.06595, + "grad_norm": 0.999263149903344, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 6595 + }, + { + "epoch": 0.06596, + "grad_norm": 0.8576980929926036, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6596 + }, + { + "epoch": 0.06597, + "grad_norm": 0.7302418341288966, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 6597 + }, + { + "epoch": 0.06598, + "grad_norm": 0.7765021212690375, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 6598 + }, + { + "epoch": 0.06599, + "grad_norm": 0.8935858983881297, + "learning_rate": 0.003, + "loss": 4.124, + "step": 6599 + }, + { + "epoch": 0.066, + "grad_norm": 1.0268301145305159, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 6600 + }, + { + "epoch": 0.06601, + "grad_norm": 0.9441826567712125, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 6601 + }, + { + "epoch": 0.06602, + "grad_norm": 0.8826431920881573, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 6602 + }, + { + "epoch": 0.06603, + "grad_norm": 0.6782729476695759, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 6603 + }, + { + "epoch": 0.06604, + "grad_norm": 0.721923977940998, + "learning_rate": 0.003, + "loss": 4.1, + "step": 6604 + }, + { + "epoch": 0.06605, + "grad_norm": 0.8126707635116521, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 6605 + }, + { + "epoch": 0.06606, + "grad_norm": 0.9139955529917244, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 6606 + }, + { + "epoch": 0.06607, + "grad_norm": 0.9787805099146487, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6607 + }, + { + "epoch": 0.06608, + "grad_norm": 1.0120287022530385, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 6608 + }, + { + "epoch": 0.06609, + "grad_norm": 0.9553503023983144, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 6609 + }, + { + "epoch": 0.0661, + "grad_norm": 0.773415990938354, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 6610 + }, + { + "epoch": 0.06611, + "grad_norm": 0.7654329601140009, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 6611 + }, + { + "epoch": 0.06612, + "grad_norm": 0.7452194084183761, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 6612 + }, + { + "epoch": 0.06613, + "grad_norm": 0.8299039489543768, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 6613 + }, + { + "epoch": 0.06614, + "grad_norm": 0.8093687816033562, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 6614 + }, + { + "epoch": 0.06615, + "grad_norm": 0.8207953880857137, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 6615 + }, + { + "epoch": 0.06616, + "grad_norm": 0.8552296704671903, + "learning_rate": 0.003, + "loss": 4.151, + "step": 6616 + }, + { + "epoch": 0.06617, + "grad_norm": 0.9244751774090344, + "learning_rate": 0.003, + "loss": 4.131, + "step": 6617 + }, + { + "epoch": 0.06618, + "grad_norm": 0.971219029341592, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 6618 + }, + { + "epoch": 0.06619, + "grad_norm": 0.8506209302501583, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 6619 + }, + { + "epoch": 0.0662, + "grad_norm": 0.8713201673630117, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6620 + }, + { + "epoch": 0.06621, + "grad_norm": 0.8914133581100084, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 6621 + }, + { + "epoch": 0.06622, + "grad_norm": 0.9005033881236821, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 6622 + }, + { + "epoch": 0.06623, + "grad_norm": 1.0929062045922742, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 6623 + }, + { + "epoch": 0.06624, + "grad_norm": 1.0309406977127722, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 6624 + }, + { + "epoch": 0.06625, + "grad_norm": 0.9349272080200425, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6625 + }, + { + "epoch": 0.06626, + "grad_norm": 0.9406574766318107, + "learning_rate": 0.003, + "loss": 4.145, + "step": 6626 + }, + { + "epoch": 0.06627, + "grad_norm": 0.8215318070242014, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 6627 + }, + { + "epoch": 0.06628, + "grad_norm": 0.8411166912060155, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 6628 + }, + { + "epoch": 0.06629, + "grad_norm": 0.778414099268401, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 6629 + }, + { + "epoch": 0.0663, + "grad_norm": 0.7994566410875287, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 6630 + }, + { + "epoch": 0.06631, + "grad_norm": 0.9203945265715889, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6631 + }, + { + "epoch": 0.06632, + "grad_norm": 0.9593833055548883, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6632 + }, + { + "epoch": 0.06633, + "grad_norm": 0.8574352076483307, + "learning_rate": 0.003, + "loss": 4.139, + "step": 6633 + }, + { + "epoch": 0.06634, + "grad_norm": 0.7496261857566855, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 6634 + }, + { + "epoch": 0.06635, + "grad_norm": 0.6760158781401536, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 6635 + }, + { + "epoch": 0.06636, + "grad_norm": 0.6916307102085072, + "learning_rate": 0.003, + "loss": 4.093, + "step": 6636 + }, + { + "epoch": 0.06637, + "grad_norm": 0.6966845658280828, + "learning_rate": 0.003, + "loss": 4.117, + "step": 6637 + }, + { + "epoch": 0.06638, + "grad_norm": 0.6883147167526921, + "learning_rate": 0.003, + "loss": 4.083, + "step": 6638 + }, + { + "epoch": 0.06639, + "grad_norm": 0.6473901715950106, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 6639 + }, + { + "epoch": 0.0664, + "grad_norm": 0.7765981197562362, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6640 + }, + { + "epoch": 0.06641, + "grad_norm": 1.115678151117138, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6641 + }, + { + "epoch": 0.06642, + "grad_norm": 0.9790163721838566, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 6642 + }, + { + "epoch": 0.06643, + "grad_norm": 0.8361408902689975, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6643 + }, + { + "epoch": 0.06644, + "grad_norm": 0.7584196343566021, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6644 + }, + { + "epoch": 0.06645, + "grad_norm": 0.816065178623302, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 6645 + }, + { + "epoch": 0.06646, + "grad_norm": 0.8060094631825453, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 6646 + }, + { + "epoch": 0.06647, + "grad_norm": 0.757636506404951, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 6647 + }, + { + "epoch": 0.06648, + "grad_norm": 0.8404369742133107, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 6648 + }, + { + "epoch": 0.06649, + "grad_norm": 0.8868264232337089, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 6649 + }, + { + "epoch": 0.0665, + "grad_norm": 0.8472513295914429, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 6650 + }, + { + "epoch": 0.06651, + "grad_norm": 0.7624647071594521, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6651 + }, + { + "epoch": 0.06652, + "grad_norm": 0.70113702299994, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 6652 + }, + { + "epoch": 0.06653, + "grad_norm": 0.7170179450934846, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 6653 + }, + { + "epoch": 0.06654, + "grad_norm": 0.6408955553595038, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 6654 + }, + { + "epoch": 0.06655, + "grad_norm": 0.7303405365737384, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6655 + }, + { + "epoch": 0.06656, + "grad_norm": 0.8170077189908054, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 6656 + }, + { + "epoch": 0.06657, + "grad_norm": 1.0373997656423108, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 6657 + }, + { + "epoch": 0.06658, + "grad_norm": 1.2877634414846078, + "learning_rate": 0.003, + "loss": 4.117, + "step": 6658 + }, + { + "epoch": 0.06659, + "grad_norm": 0.7003902163594509, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 6659 + }, + { + "epoch": 0.0666, + "grad_norm": 0.7777938489458038, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 6660 + }, + { + "epoch": 0.06661, + "grad_norm": 0.759383450915605, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 6661 + }, + { + "epoch": 0.06662, + "grad_norm": 0.6357206487995134, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 6662 + }, + { + "epoch": 0.06663, + "grad_norm": 0.5997016664182901, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 6663 + }, + { + "epoch": 0.06664, + "grad_norm": 0.6928418683619769, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 6664 + }, + { + "epoch": 0.06665, + "grad_norm": 0.6725431198503523, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 6665 + }, + { + "epoch": 0.06666, + "grad_norm": 0.7185786516891489, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 6666 + }, + { + "epoch": 0.06667, + "grad_norm": 0.7583846091674911, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 6667 + }, + { + "epoch": 0.06668, + "grad_norm": 0.7727383750975167, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 6668 + }, + { + "epoch": 0.06669, + "grad_norm": 0.7990602062120472, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 6669 + }, + { + "epoch": 0.0667, + "grad_norm": 0.8932825213449198, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 6670 + }, + { + "epoch": 0.06671, + "grad_norm": 0.9161270016145233, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 6671 + }, + { + "epoch": 0.06672, + "grad_norm": 0.8948099688910036, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6672 + }, + { + "epoch": 0.06673, + "grad_norm": 0.8131754380634141, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 6673 + }, + { + "epoch": 0.06674, + "grad_norm": 0.9822219202437968, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 6674 + }, + { + "epoch": 0.06675, + "grad_norm": 1.4384917377302466, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 6675 + }, + { + "epoch": 0.06676, + "grad_norm": 1.0715294490450158, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 6676 + }, + { + "epoch": 0.06677, + "grad_norm": 0.9943884055205793, + "learning_rate": 0.003, + "loss": 4.143, + "step": 6677 + }, + { + "epoch": 0.06678, + "grad_norm": 1.1472283624566484, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 6678 + }, + { + "epoch": 0.06679, + "grad_norm": 0.837081045625505, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6679 + }, + { + "epoch": 0.0668, + "grad_norm": 0.8632696497966131, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 6680 + }, + { + "epoch": 0.06681, + "grad_norm": 0.881128794076206, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 6681 + }, + { + "epoch": 0.06682, + "grad_norm": 0.8577139639731307, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 6682 + }, + { + "epoch": 0.06683, + "grad_norm": 0.9681904531791742, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 6683 + }, + { + "epoch": 0.06684, + "grad_norm": 1.045816856105673, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6684 + }, + { + "epoch": 0.06685, + "grad_norm": 1.1727987324167213, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 6685 + }, + { + "epoch": 0.06686, + "grad_norm": 1.027054896660675, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 6686 + }, + { + "epoch": 0.06687, + "grad_norm": 0.9524644542845865, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 6687 + }, + { + "epoch": 0.06688, + "grad_norm": 0.8156341843328404, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 6688 + }, + { + "epoch": 0.06689, + "grad_norm": 0.7545435705045717, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 6689 + }, + { + "epoch": 0.0669, + "grad_norm": 0.7352426544757619, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6690 + }, + { + "epoch": 0.06691, + "grad_norm": 0.6665883072114859, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 6691 + }, + { + "epoch": 0.06692, + "grad_norm": 0.6725148756303375, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 6692 + }, + { + "epoch": 0.06693, + "grad_norm": 0.7130842452638179, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6693 + }, + { + "epoch": 0.06694, + "grad_norm": 0.8096745511420714, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 6694 + }, + { + "epoch": 0.06695, + "grad_norm": 0.904033151495042, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 6695 + }, + { + "epoch": 0.06696, + "grad_norm": 0.9485066116960444, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6696 + }, + { + "epoch": 0.06697, + "grad_norm": 0.9268917013823743, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 6697 + }, + { + "epoch": 0.06698, + "grad_norm": 0.9285500128581322, + "learning_rate": 0.003, + "loss": 4.108, + "step": 6698 + }, + { + "epoch": 0.06699, + "grad_norm": 1.0584540045258635, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6699 + }, + { + "epoch": 0.067, + "grad_norm": 1.0373189607989615, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 6700 + }, + { + "epoch": 0.06701, + "grad_norm": 0.9623285914421408, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 6701 + }, + { + "epoch": 0.06702, + "grad_norm": 0.9701353847551325, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 6702 + }, + { + "epoch": 0.06703, + "grad_norm": 0.8858538800012404, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 6703 + }, + { + "epoch": 0.06704, + "grad_norm": 0.9016995986468413, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 6704 + }, + { + "epoch": 0.06705, + "grad_norm": 0.925225581835428, + "learning_rate": 0.003, + "loss": 4.11, + "step": 6705 + }, + { + "epoch": 0.06706, + "grad_norm": 0.896855586772984, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 6706 + }, + { + "epoch": 0.06707, + "grad_norm": 0.7977187311047266, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 6707 + }, + { + "epoch": 0.06708, + "grad_norm": 0.8299509664392212, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 6708 + }, + { + "epoch": 0.06709, + "grad_norm": 0.9460884720074625, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 6709 + }, + { + "epoch": 0.0671, + "grad_norm": 0.8571195032444293, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 6710 + }, + { + "epoch": 0.06711, + "grad_norm": 0.9023183298505612, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 6711 + }, + { + "epoch": 0.06712, + "grad_norm": 1.1030390737820472, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6712 + }, + { + "epoch": 0.06713, + "grad_norm": 0.9582329185489545, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 6713 + }, + { + "epoch": 0.06714, + "grad_norm": 0.8269350900476661, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 6714 + }, + { + "epoch": 0.06715, + "grad_norm": 0.7110607112083103, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 6715 + }, + { + "epoch": 0.06716, + "grad_norm": 0.7100774219805601, + "learning_rate": 0.003, + "loss": 4.114, + "step": 6716 + }, + { + "epoch": 0.06717, + "grad_norm": 0.855459300866275, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 6717 + }, + { + "epoch": 0.06718, + "grad_norm": 0.9053889950699634, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 6718 + }, + { + "epoch": 0.06719, + "grad_norm": 0.8719291055963075, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 6719 + }, + { + "epoch": 0.0672, + "grad_norm": 0.8359415545682723, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 6720 + }, + { + "epoch": 0.06721, + "grad_norm": 0.8906370466300249, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 6721 + }, + { + "epoch": 0.06722, + "grad_norm": 1.0176449340197307, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 6722 + }, + { + "epoch": 0.06723, + "grad_norm": 0.9805817985293129, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 6723 + }, + { + "epoch": 0.06724, + "grad_norm": 0.9703246693491563, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6724 + }, + { + "epoch": 0.06725, + "grad_norm": 0.9242585229701251, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 6725 + }, + { + "epoch": 0.06726, + "grad_norm": 0.8600589534698727, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 6726 + }, + { + "epoch": 0.06727, + "grad_norm": 0.950790968925123, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 6727 + }, + { + "epoch": 0.06728, + "grad_norm": 1.1631164741777629, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 6728 + }, + { + "epoch": 0.06729, + "grad_norm": 0.7915866355485529, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 6729 + }, + { + "epoch": 0.0673, + "grad_norm": 0.8274594227494144, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 6730 + }, + { + "epoch": 0.06731, + "grad_norm": 0.6766878958185836, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 6731 + }, + { + "epoch": 0.06732, + "grad_norm": 0.6976909564398467, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 6732 + }, + { + "epoch": 0.06733, + "grad_norm": 0.7614581431023822, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 6733 + }, + { + "epoch": 0.06734, + "grad_norm": 1.057194128915908, + "learning_rate": 0.003, + "loss": 4.124, + "step": 6734 + }, + { + "epoch": 0.06735, + "grad_norm": 1.213622998381686, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 6735 + }, + { + "epoch": 0.06736, + "grad_norm": 0.734105473964094, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 6736 + }, + { + "epoch": 0.06737, + "grad_norm": 0.6158143832281706, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 6737 + }, + { + "epoch": 0.06738, + "grad_norm": 0.5821932035460526, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 6738 + }, + { + "epoch": 0.06739, + "grad_norm": 0.6445881524656328, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 6739 + }, + { + "epoch": 0.0674, + "grad_norm": 0.7083800475248121, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 6740 + }, + { + "epoch": 0.06741, + "grad_norm": 0.8121921490974171, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 6741 + }, + { + "epoch": 0.06742, + "grad_norm": 0.9424065146650864, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 6742 + }, + { + "epoch": 0.06743, + "grad_norm": 1.0699042142898014, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 6743 + }, + { + "epoch": 0.06744, + "grad_norm": 1.047616453763929, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 6744 + }, + { + "epoch": 0.06745, + "grad_norm": 0.897040372045614, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 6745 + }, + { + "epoch": 0.06746, + "grad_norm": 0.8107749141615237, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6746 + }, + { + "epoch": 0.06747, + "grad_norm": 0.8070857428863515, + "learning_rate": 0.003, + "loss": 4.106, + "step": 6747 + }, + { + "epoch": 0.06748, + "grad_norm": 0.8359697636031163, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 6748 + }, + { + "epoch": 0.06749, + "grad_norm": 0.827772533552076, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 6749 + }, + { + "epoch": 0.0675, + "grad_norm": 0.859730671541346, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 6750 + }, + { + "epoch": 0.06751, + "grad_norm": 0.7946849688098641, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6751 + }, + { + "epoch": 0.06752, + "grad_norm": 0.7787584913556823, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6752 + }, + { + "epoch": 0.06753, + "grad_norm": 0.7233106355478739, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 6753 + }, + { + "epoch": 0.06754, + "grad_norm": 0.5923969384902699, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 6754 + }, + { + "epoch": 0.06755, + "grad_norm": 0.620061843467779, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 6755 + }, + { + "epoch": 0.06756, + "grad_norm": 0.6491047345358799, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 6756 + }, + { + "epoch": 0.06757, + "grad_norm": 0.647987484033838, + "learning_rate": 0.003, + "loss": 4.109, + "step": 6757 + }, + { + "epoch": 0.06758, + "grad_norm": 0.6452806428391407, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 6758 + }, + { + "epoch": 0.06759, + "grad_norm": 0.7117568970383578, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 6759 + }, + { + "epoch": 0.0676, + "grad_norm": 0.9554260901760981, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 6760 + }, + { + "epoch": 0.06761, + "grad_norm": 1.4624912759082085, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 6761 + }, + { + "epoch": 0.06762, + "grad_norm": 0.5931283354167579, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 6762 + }, + { + "epoch": 0.06763, + "grad_norm": 0.8396616773751427, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 6763 + }, + { + "epoch": 0.06764, + "grad_norm": 0.9612365334414132, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6764 + }, + { + "epoch": 0.06765, + "grad_norm": 0.895825369939231, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 6765 + }, + { + "epoch": 0.06766, + "grad_norm": 0.9141273493748527, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 6766 + }, + { + "epoch": 0.06767, + "grad_norm": 1.0179298922982885, + "learning_rate": 0.003, + "loss": 4.127, + "step": 6767 + }, + { + "epoch": 0.06768, + "grad_norm": 1.0094086260801791, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 6768 + }, + { + "epoch": 0.06769, + "grad_norm": 1.0112198539456203, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 6769 + }, + { + "epoch": 0.0677, + "grad_norm": 0.9926517409036734, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 6770 + }, + { + "epoch": 0.06771, + "grad_norm": 0.8812766249718238, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 6771 + }, + { + "epoch": 0.06772, + "grad_norm": 0.8389743876880504, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 6772 + }, + { + "epoch": 0.06773, + "grad_norm": 0.7951452996932417, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6773 + }, + { + "epoch": 0.06774, + "grad_norm": 0.8837761133765221, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6774 + }, + { + "epoch": 0.06775, + "grad_norm": 0.957136507400834, + "learning_rate": 0.003, + "loss": 4.12, + "step": 6775 + }, + { + "epoch": 0.06776, + "grad_norm": 1.0481666958281863, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 6776 + }, + { + "epoch": 0.06777, + "grad_norm": 0.8631871107785212, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 6777 + }, + { + "epoch": 0.06778, + "grad_norm": 0.7576677776928321, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6778 + }, + { + "epoch": 0.06779, + "grad_norm": 0.7621818727362122, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 6779 + }, + { + "epoch": 0.0678, + "grad_norm": 0.7354090321987549, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 6780 + }, + { + "epoch": 0.06781, + "grad_norm": 0.796508496988399, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 6781 + }, + { + "epoch": 0.06782, + "grad_norm": 0.8416596876225658, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 6782 + }, + { + "epoch": 0.06783, + "grad_norm": 0.9169424759930719, + "learning_rate": 0.003, + "loss": 4.134, + "step": 6783 + }, + { + "epoch": 0.06784, + "grad_norm": 0.8178011962748643, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 6784 + }, + { + "epoch": 0.06785, + "grad_norm": 0.7835817405720389, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 6785 + }, + { + "epoch": 0.06786, + "grad_norm": 0.8242013907791396, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 6786 + }, + { + "epoch": 0.06787, + "grad_norm": 0.7961787365055251, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6787 + }, + { + "epoch": 0.06788, + "grad_norm": 0.8721165254426617, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 6788 + }, + { + "epoch": 0.06789, + "grad_norm": 1.1396955949704233, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 6789 + }, + { + "epoch": 0.0679, + "grad_norm": 1.0568810715247237, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6790 + }, + { + "epoch": 0.06791, + "grad_norm": 1.0143944196991193, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 6791 + }, + { + "epoch": 0.06792, + "grad_norm": 0.884818963074791, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 6792 + }, + { + "epoch": 0.06793, + "grad_norm": 0.8478088580149906, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 6793 + }, + { + "epoch": 0.06794, + "grad_norm": 0.8056286827298742, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 6794 + }, + { + "epoch": 0.06795, + "grad_norm": 0.7411740341729665, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 6795 + }, + { + "epoch": 0.06796, + "grad_norm": 0.7889975862725216, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 6796 + }, + { + "epoch": 0.06797, + "grad_norm": 0.9643036551361494, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 6797 + }, + { + "epoch": 0.06798, + "grad_norm": 0.9344988567161295, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6798 + }, + { + "epoch": 0.06799, + "grad_norm": 0.9452355458669397, + "learning_rate": 0.003, + "loss": 4.147, + "step": 6799 + }, + { + "epoch": 0.068, + "grad_norm": 0.8857614730909316, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 6800 + }, + { + "epoch": 0.06801, + "grad_norm": 0.809752329455322, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6801 + }, + { + "epoch": 0.06802, + "grad_norm": 0.8227719693453557, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 6802 + }, + { + "epoch": 0.06803, + "grad_norm": 0.7412508487470881, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6803 + }, + { + "epoch": 0.06804, + "grad_norm": 0.7054837765584971, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 6804 + }, + { + "epoch": 0.06805, + "grad_norm": 0.7568713263104572, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 6805 + }, + { + "epoch": 0.06806, + "grad_norm": 0.9817473684527529, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 6806 + }, + { + "epoch": 0.06807, + "grad_norm": 1.1600981786410642, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 6807 + }, + { + "epoch": 0.06808, + "grad_norm": 1.159887129703928, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 6808 + }, + { + "epoch": 0.06809, + "grad_norm": 0.8456896116982721, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 6809 + }, + { + "epoch": 0.0681, + "grad_norm": 0.858282322637232, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 6810 + }, + { + "epoch": 0.06811, + "grad_norm": 0.9611788401234072, + "learning_rate": 0.003, + "loss": 4.1494, + "step": 6811 + }, + { + "epoch": 0.06812, + "grad_norm": 0.8997598486326439, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 6812 + }, + { + "epoch": 0.06813, + "grad_norm": 0.7739640351300833, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 6813 + }, + { + "epoch": 0.06814, + "grad_norm": 0.811869754027927, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6814 + }, + { + "epoch": 0.06815, + "grad_norm": 0.7998801552358381, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 6815 + }, + { + "epoch": 0.06816, + "grad_norm": 0.7960758830426774, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 6816 + }, + { + "epoch": 0.06817, + "grad_norm": 0.863740767658441, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6817 + }, + { + "epoch": 0.06818, + "grad_norm": 0.8459457179335972, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 6818 + }, + { + "epoch": 0.06819, + "grad_norm": 0.8884032934270286, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6819 + }, + { + "epoch": 0.0682, + "grad_norm": 0.9225686576604342, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 6820 + }, + { + "epoch": 0.06821, + "grad_norm": 0.844036272472802, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 6821 + }, + { + "epoch": 0.06822, + "grad_norm": 0.8174293547589027, + "learning_rate": 0.003, + "loss": 4.093, + "step": 6822 + }, + { + "epoch": 0.06823, + "grad_norm": 0.836645021687918, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 6823 + }, + { + "epoch": 0.06824, + "grad_norm": 0.9752883654153214, + "learning_rate": 0.003, + "loss": 4.09, + "step": 6824 + }, + { + "epoch": 0.06825, + "grad_norm": 1.079600387760812, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 6825 + }, + { + "epoch": 0.06826, + "grad_norm": 0.8197898525120547, + "learning_rate": 0.003, + "loss": 4.099, + "step": 6826 + }, + { + "epoch": 0.06827, + "grad_norm": 0.9956479850125667, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 6827 + }, + { + "epoch": 0.06828, + "grad_norm": 1.152216850473556, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 6828 + }, + { + "epoch": 0.06829, + "grad_norm": 0.8571944350147577, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 6829 + }, + { + "epoch": 0.0683, + "grad_norm": 0.757993990915538, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 6830 + }, + { + "epoch": 0.06831, + "grad_norm": 0.864695692673034, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 6831 + }, + { + "epoch": 0.06832, + "grad_norm": 0.9705812470204103, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 6832 + }, + { + "epoch": 0.06833, + "grad_norm": 0.9213907790898336, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 6833 + }, + { + "epoch": 0.06834, + "grad_norm": 0.792235057764182, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 6834 + }, + { + "epoch": 0.06835, + "grad_norm": 0.7066677455983338, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 6835 + }, + { + "epoch": 0.06836, + "grad_norm": 0.6670788363565453, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 6836 + }, + { + "epoch": 0.06837, + "grad_norm": 0.6464856752678825, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 6837 + }, + { + "epoch": 0.06838, + "grad_norm": 0.6108028457392457, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 6838 + }, + { + "epoch": 0.06839, + "grad_norm": 0.6813565301110693, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 6839 + }, + { + "epoch": 0.0684, + "grad_norm": 0.794881760436682, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 6840 + }, + { + "epoch": 0.06841, + "grad_norm": 0.961174574044574, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 6841 + }, + { + "epoch": 0.06842, + "grad_norm": 1.0476574337633355, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 6842 + }, + { + "epoch": 0.06843, + "grad_norm": 0.8907102890021011, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 6843 + }, + { + "epoch": 0.06844, + "grad_norm": 0.8811872643928172, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 6844 + }, + { + "epoch": 0.06845, + "grad_norm": 0.9952993463626575, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 6845 + }, + { + "epoch": 0.06846, + "grad_norm": 1.2628089603731774, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 6846 + }, + { + "epoch": 0.06847, + "grad_norm": 0.901670711894547, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 6847 + }, + { + "epoch": 0.06848, + "grad_norm": 0.956123404918695, + "learning_rate": 0.003, + "loss": 4.119, + "step": 6848 + }, + { + "epoch": 0.06849, + "grad_norm": 1.0968210993157013, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 6849 + }, + { + "epoch": 0.0685, + "grad_norm": 0.936347183022355, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 6850 + }, + { + "epoch": 0.06851, + "grad_norm": 0.8968792133031543, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 6851 + }, + { + "epoch": 0.06852, + "grad_norm": 0.8507324682860469, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 6852 + }, + { + "epoch": 0.06853, + "grad_norm": 0.7108631167713592, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 6853 + }, + { + "epoch": 0.06854, + "grad_norm": 0.5899558073031355, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 6854 + }, + { + "epoch": 0.06855, + "grad_norm": 0.6714377951019126, + "learning_rate": 0.003, + "loss": 4.1342, + "step": 6855 + }, + { + "epoch": 0.06856, + "grad_norm": 0.7264783579127705, + "learning_rate": 0.003, + "loss": 4.133, + "step": 6856 + }, + { + "epoch": 0.06857, + "grad_norm": 0.8461334990738569, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 6857 + }, + { + "epoch": 0.06858, + "grad_norm": 0.9692288911903976, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 6858 + }, + { + "epoch": 0.06859, + "grad_norm": 0.9198005654114306, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 6859 + }, + { + "epoch": 0.0686, + "grad_norm": 0.6854066787324575, + "learning_rate": 0.003, + "loss": 4.096, + "step": 6860 + }, + { + "epoch": 0.06861, + "grad_norm": 0.6574071277975999, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 6861 + }, + { + "epoch": 0.06862, + "grad_norm": 0.930478597926848, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 6862 + }, + { + "epoch": 0.06863, + "grad_norm": 1.0592073313127546, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 6863 + }, + { + "epoch": 0.06864, + "grad_norm": 0.9388318477503523, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 6864 + }, + { + "epoch": 0.06865, + "grad_norm": 0.746039107432962, + "learning_rate": 0.003, + "loss": 4.097, + "step": 6865 + }, + { + "epoch": 0.06866, + "grad_norm": 0.7626047298181141, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 6866 + }, + { + "epoch": 0.06867, + "grad_norm": 0.8192485910979884, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 6867 + }, + { + "epoch": 0.06868, + "grad_norm": 0.8943577260151027, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 6868 + }, + { + "epoch": 0.06869, + "grad_norm": 1.0994368832659123, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 6869 + }, + { + "epoch": 0.0687, + "grad_norm": 0.8861962197230898, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6870 + }, + { + "epoch": 0.06871, + "grad_norm": 0.8815941784538627, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 6871 + }, + { + "epoch": 0.06872, + "grad_norm": 0.9924460757229603, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 6872 + }, + { + "epoch": 0.06873, + "grad_norm": 1.0509114751207171, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 6873 + }, + { + "epoch": 0.06874, + "grad_norm": 0.9532056718212061, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 6874 + }, + { + "epoch": 0.06875, + "grad_norm": 1.023680307403794, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 6875 + }, + { + "epoch": 0.06876, + "grad_norm": 0.9902008565775648, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6876 + }, + { + "epoch": 0.06877, + "grad_norm": 0.9531958381228303, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 6877 + }, + { + "epoch": 0.06878, + "grad_norm": 0.9710478342391432, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 6878 + }, + { + "epoch": 0.06879, + "grad_norm": 0.909639330602345, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 6879 + }, + { + "epoch": 0.0688, + "grad_norm": 0.8095671998298988, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 6880 + }, + { + "epoch": 0.06881, + "grad_norm": 0.9113186941401037, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 6881 + }, + { + "epoch": 0.06882, + "grad_norm": 0.9258946157268206, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 6882 + }, + { + "epoch": 0.06883, + "grad_norm": 1.0573211690290847, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 6883 + }, + { + "epoch": 0.06884, + "grad_norm": 1.0560959741409428, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 6884 + }, + { + "epoch": 0.06885, + "grad_norm": 0.9393005675185666, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 6885 + }, + { + "epoch": 0.06886, + "grad_norm": 0.8873501258981987, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 6886 + }, + { + "epoch": 0.06887, + "grad_norm": 0.9371568421249765, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 6887 + }, + { + "epoch": 0.06888, + "grad_norm": 0.960033318581805, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 6888 + }, + { + "epoch": 0.06889, + "grad_norm": 1.0030573810363002, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 6889 + }, + { + "epoch": 0.0689, + "grad_norm": 1.0144314201908178, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 6890 + }, + { + "epoch": 0.06891, + "grad_norm": 1.1068906368874076, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 6891 + }, + { + "epoch": 0.06892, + "grad_norm": 0.9054892903304268, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 6892 + }, + { + "epoch": 0.06893, + "grad_norm": 0.7540684414365276, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6893 + }, + { + "epoch": 0.06894, + "grad_norm": 0.6663108621237684, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 6894 + }, + { + "epoch": 0.06895, + "grad_norm": 0.5979163007296868, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 6895 + }, + { + "epoch": 0.06896, + "grad_norm": 0.5872771645065226, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 6896 + }, + { + "epoch": 0.06897, + "grad_norm": 0.6173404544816022, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 6897 + }, + { + "epoch": 0.06898, + "grad_norm": 0.762140883361191, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6898 + }, + { + "epoch": 0.06899, + "grad_norm": 1.0242084853160343, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 6899 + }, + { + "epoch": 0.069, + "grad_norm": 1.2323653252300188, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 6900 + }, + { + "epoch": 0.06901, + "grad_norm": 0.5820472612244612, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 6901 + }, + { + "epoch": 0.06902, + "grad_norm": 0.7237387908075824, + "learning_rate": 0.003, + "loss": 4.088, + "step": 6902 + }, + { + "epoch": 0.06903, + "grad_norm": 0.8664255236861678, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 6903 + }, + { + "epoch": 0.06904, + "grad_norm": 0.8559541713855746, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 6904 + }, + { + "epoch": 0.06905, + "grad_norm": 0.8471025018164605, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 6905 + }, + { + "epoch": 0.06906, + "grad_norm": 0.8690540035946629, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6906 + }, + { + "epoch": 0.06907, + "grad_norm": 0.8178132983249496, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 6907 + }, + { + "epoch": 0.06908, + "grad_norm": 0.7564578666671172, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6908 + }, + { + "epoch": 0.06909, + "grad_norm": 0.7790308250713, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 6909 + }, + { + "epoch": 0.0691, + "grad_norm": 0.7351729319289587, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 6910 + }, + { + "epoch": 0.06911, + "grad_norm": 0.7267083412234101, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 6911 + }, + { + "epoch": 0.06912, + "grad_norm": 0.6661025595948273, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 6912 + }, + { + "epoch": 0.06913, + "grad_norm": 0.674553980244978, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 6913 + }, + { + "epoch": 0.06914, + "grad_norm": 0.7348105212280468, + "learning_rate": 0.003, + "loss": 4.122, + "step": 6914 + }, + { + "epoch": 0.06915, + "grad_norm": 0.7493265450674638, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 6915 + }, + { + "epoch": 0.06916, + "grad_norm": 0.808956213639763, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 6916 + }, + { + "epoch": 0.06917, + "grad_norm": 0.8113376674561755, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 6917 + }, + { + "epoch": 0.06918, + "grad_norm": 0.8283810471683984, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 6918 + }, + { + "epoch": 0.06919, + "grad_norm": 0.8686223282408305, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 6919 + }, + { + "epoch": 0.0692, + "grad_norm": 0.7951171895581218, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 6920 + }, + { + "epoch": 0.06921, + "grad_norm": 0.9445545003116608, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 6921 + }, + { + "epoch": 0.06922, + "grad_norm": 0.9810026720945391, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 6922 + }, + { + "epoch": 0.06923, + "grad_norm": 0.9535589717974995, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 6923 + }, + { + "epoch": 0.06924, + "grad_norm": 0.9826106732792211, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 6924 + }, + { + "epoch": 0.06925, + "grad_norm": 0.958775832947445, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 6925 + }, + { + "epoch": 0.06926, + "grad_norm": 0.9026835208656796, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6926 + }, + { + "epoch": 0.06927, + "grad_norm": 0.8065987081198023, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 6927 + }, + { + "epoch": 0.06928, + "grad_norm": 0.7325070632512863, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 6928 + }, + { + "epoch": 0.06929, + "grad_norm": 0.7844696991539358, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 6929 + }, + { + "epoch": 0.0693, + "grad_norm": 0.7321103434207817, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 6930 + }, + { + "epoch": 0.06931, + "grad_norm": 0.6348052074469944, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 6931 + }, + { + "epoch": 0.06932, + "grad_norm": 0.6156078545430647, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 6932 + }, + { + "epoch": 0.06933, + "grad_norm": 0.7281337302342974, + "learning_rate": 0.003, + "loss": 4.121, + "step": 6933 + }, + { + "epoch": 0.06934, + "grad_norm": 0.9326406238175676, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 6934 + }, + { + "epoch": 0.06935, + "grad_norm": 1.348898577949277, + "learning_rate": 0.003, + "loss": 4.1261, + "step": 6935 + }, + { + "epoch": 0.06936, + "grad_norm": 0.8658725709964428, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 6936 + }, + { + "epoch": 0.06937, + "grad_norm": 0.8735566248577522, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 6937 + }, + { + "epoch": 0.06938, + "grad_norm": 0.9320220136220663, + "learning_rate": 0.003, + "loss": 4.079, + "step": 6938 + }, + { + "epoch": 0.06939, + "grad_norm": 0.9133167311611287, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 6939 + }, + { + "epoch": 0.0694, + "grad_norm": 0.9890705299908502, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 6940 + }, + { + "epoch": 0.06941, + "grad_norm": 1.3060885860531548, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6941 + }, + { + "epoch": 0.06942, + "grad_norm": 0.7699335742222013, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 6942 + }, + { + "epoch": 0.06943, + "grad_norm": 0.7643000254330296, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 6943 + }, + { + "epoch": 0.06944, + "grad_norm": 0.7529705113455923, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 6944 + }, + { + "epoch": 0.06945, + "grad_norm": 0.882823727029274, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 6945 + }, + { + "epoch": 0.06946, + "grad_norm": 1.0250594327340659, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 6946 + }, + { + "epoch": 0.06947, + "grad_norm": 1.1502658138369475, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 6947 + }, + { + "epoch": 0.06948, + "grad_norm": 0.8444220868745179, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6948 + }, + { + "epoch": 0.06949, + "grad_norm": 0.7711423294691977, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 6949 + }, + { + "epoch": 0.0695, + "grad_norm": 0.8610365574220729, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6950 + }, + { + "epoch": 0.06951, + "grad_norm": 0.9731685606617599, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 6951 + }, + { + "epoch": 0.06952, + "grad_norm": 1.1267620956159707, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 6952 + }, + { + "epoch": 0.06953, + "grad_norm": 0.9862000527924848, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 6953 + }, + { + "epoch": 0.06954, + "grad_norm": 0.9666325788788848, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6954 + }, + { + "epoch": 0.06955, + "grad_norm": 0.9599709329069231, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 6955 + }, + { + "epoch": 0.06956, + "grad_norm": 0.9866026018546916, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6956 + }, + { + "epoch": 0.06957, + "grad_norm": 0.9845733907627405, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 6957 + }, + { + "epoch": 0.06958, + "grad_norm": 0.9943021661775551, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 6958 + }, + { + "epoch": 0.06959, + "grad_norm": 1.0279554280201155, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 6959 + }, + { + "epoch": 0.0696, + "grad_norm": 1.1047587660616394, + "learning_rate": 0.003, + "loss": 4.143, + "step": 6960 + }, + { + "epoch": 0.06961, + "grad_norm": 0.9950639197668989, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 6961 + }, + { + "epoch": 0.06962, + "grad_norm": 1.027748422776989, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 6962 + }, + { + "epoch": 0.06963, + "grad_norm": 0.9781377480333107, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 6963 + }, + { + "epoch": 0.06964, + "grad_norm": 0.9482629019532105, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 6964 + }, + { + "epoch": 0.06965, + "grad_norm": 0.9356422484428207, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 6965 + }, + { + "epoch": 0.06966, + "grad_norm": 0.9056612242786197, + "learning_rate": 0.003, + "loss": 4.132, + "step": 6966 + }, + { + "epoch": 0.06967, + "grad_norm": 0.7203106230136754, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 6967 + }, + { + "epoch": 0.06968, + "grad_norm": 0.6599503868887251, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 6968 + }, + { + "epoch": 0.06969, + "grad_norm": 0.6987003782313111, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 6969 + }, + { + "epoch": 0.0697, + "grad_norm": 0.8342591105193845, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 6970 + }, + { + "epoch": 0.06971, + "grad_norm": 1.018817534245797, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 6971 + }, + { + "epoch": 0.06972, + "grad_norm": 1.2395063765633032, + "learning_rate": 0.003, + "loss": 4.082, + "step": 6972 + }, + { + "epoch": 0.06973, + "grad_norm": 0.8545094616038666, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 6973 + }, + { + "epoch": 0.06974, + "grad_norm": 0.8121109776419817, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 6974 + }, + { + "epoch": 0.06975, + "grad_norm": 0.7584233055394562, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 6975 + }, + { + "epoch": 0.06976, + "grad_norm": 0.715581943047214, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6976 + }, + { + "epoch": 0.06977, + "grad_norm": 0.8238558659092502, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 6977 + }, + { + "epoch": 0.06978, + "grad_norm": 0.9171894622076492, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 6978 + }, + { + "epoch": 0.06979, + "grad_norm": 0.8537393028092058, + "learning_rate": 0.003, + "loss": 4.13, + "step": 6979 + }, + { + "epoch": 0.0698, + "grad_norm": 0.6494522431850996, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 6980 + }, + { + "epoch": 0.06981, + "grad_norm": 0.6394410355891653, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 6981 + }, + { + "epoch": 0.06982, + "grad_norm": 0.655752618433206, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 6982 + }, + { + "epoch": 0.06983, + "grad_norm": 0.6522649714130653, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6983 + }, + { + "epoch": 0.06984, + "grad_norm": 0.6745056056178718, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6984 + }, + { + "epoch": 0.06985, + "grad_norm": 0.7167454235222351, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 6985 + }, + { + "epoch": 0.06986, + "grad_norm": 0.7745691656939917, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 6986 + }, + { + "epoch": 0.06987, + "grad_norm": 0.7824556430219963, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 6987 + }, + { + "epoch": 0.06988, + "grad_norm": 0.713073188685409, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 6988 + }, + { + "epoch": 0.06989, + "grad_norm": 0.6918866773648449, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 6989 + }, + { + "epoch": 0.0699, + "grad_norm": 0.6887212376040721, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 6990 + }, + { + "epoch": 0.06991, + "grad_norm": 0.728960146591993, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 6991 + }, + { + "epoch": 0.06992, + "grad_norm": 0.7773350928913978, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6992 + }, + { + "epoch": 0.06993, + "grad_norm": 1.0282095814384735, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 6993 + }, + { + "epoch": 0.06994, + "grad_norm": 1.2480624334513428, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 6994 + }, + { + "epoch": 0.06995, + "grad_norm": 0.8763541675945252, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 6995 + }, + { + "epoch": 0.06996, + "grad_norm": 0.825844759458018, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 6996 + }, + { + "epoch": 0.06997, + "grad_norm": 0.9173767828845192, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 6997 + }, + { + "epoch": 0.06998, + "grad_norm": 0.9510745593290524, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 6998 + }, + { + "epoch": 0.06999, + "grad_norm": 0.9779021385996265, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 6999 + }, + { + "epoch": 0.07, + "grad_norm": 1.1827894871717832, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7000 + }, + { + "epoch": 0.07001, + "grad_norm": 0.7823428891121994, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 7001 + }, + { + "epoch": 0.07002, + "grad_norm": 0.7502844132003458, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 7002 + }, + { + "epoch": 0.07003, + "grad_norm": 0.6431900619854782, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7003 + }, + { + "epoch": 0.07004, + "grad_norm": 0.6545830386408341, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7004 + }, + { + "epoch": 0.07005, + "grad_norm": 0.683875895653165, + "learning_rate": 0.003, + "loss": 4.096, + "step": 7005 + }, + { + "epoch": 0.07006, + "grad_norm": 0.7166463834547604, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 7006 + }, + { + "epoch": 0.07007, + "grad_norm": 0.7142802784000205, + "learning_rate": 0.003, + "loss": 4.097, + "step": 7007 + }, + { + "epoch": 0.07008, + "grad_norm": 0.8137440162068434, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7008 + }, + { + "epoch": 0.07009, + "grad_norm": 0.9813188383797056, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 7009 + }, + { + "epoch": 0.0701, + "grad_norm": 1.1360396784480946, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7010 + }, + { + "epoch": 0.07011, + "grad_norm": 1.122021954079669, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 7011 + }, + { + "epoch": 0.07012, + "grad_norm": 0.9025536928180181, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7012 + }, + { + "epoch": 0.07013, + "grad_norm": 0.8888035296110233, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7013 + }, + { + "epoch": 0.07014, + "grad_norm": 0.9010860607271816, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 7014 + }, + { + "epoch": 0.07015, + "grad_norm": 0.9906713805217798, + "learning_rate": 0.003, + "loss": 4.1, + "step": 7015 + }, + { + "epoch": 0.07016, + "grad_norm": 1.1465687996483969, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 7016 + }, + { + "epoch": 0.07017, + "grad_norm": 1.0247558863315256, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 7017 + }, + { + "epoch": 0.07018, + "grad_norm": 1.0613439192576388, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 7018 + }, + { + "epoch": 0.07019, + "grad_norm": 0.9990910934627305, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7019 + }, + { + "epoch": 0.0702, + "grad_norm": 0.9540598745983695, + "learning_rate": 0.003, + "loss": 4.103, + "step": 7020 + }, + { + "epoch": 0.07021, + "grad_norm": 0.9831178649652754, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 7021 + }, + { + "epoch": 0.07022, + "grad_norm": 0.9737401392446607, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 7022 + }, + { + "epoch": 0.07023, + "grad_norm": 1.1732936942454621, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 7023 + }, + { + "epoch": 0.07024, + "grad_norm": 0.8945111721502813, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7024 + }, + { + "epoch": 0.07025, + "grad_norm": 0.867087064716649, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 7025 + }, + { + "epoch": 0.07026, + "grad_norm": 0.8901204165200368, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7026 + }, + { + "epoch": 0.07027, + "grad_norm": 0.7820722242424083, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7027 + }, + { + "epoch": 0.07028, + "grad_norm": 0.8644909122379497, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7028 + }, + { + "epoch": 0.07029, + "grad_norm": 0.8953906928512193, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 7029 + }, + { + "epoch": 0.0703, + "grad_norm": 0.8991569997104605, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 7030 + }, + { + "epoch": 0.07031, + "grad_norm": 1.1088920936393387, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 7031 + }, + { + "epoch": 0.07032, + "grad_norm": 1.0567421197796187, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 7032 + }, + { + "epoch": 0.07033, + "grad_norm": 1.1300939504610534, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 7033 + }, + { + "epoch": 0.07034, + "grad_norm": 0.8874141777977914, + "learning_rate": 0.003, + "loss": 4.113, + "step": 7034 + }, + { + "epoch": 0.07035, + "grad_norm": 0.752043314504715, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 7035 + }, + { + "epoch": 0.07036, + "grad_norm": 0.7572808680785389, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7036 + }, + { + "epoch": 0.07037, + "grad_norm": 0.7616378265590469, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 7037 + }, + { + "epoch": 0.07038, + "grad_norm": 0.7319105489860215, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 7038 + }, + { + "epoch": 0.07039, + "grad_norm": 0.6986810220580801, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 7039 + }, + { + "epoch": 0.0704, + "grad_norm": 0.6919082493656014, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 7040 + }, + { + "epoch": 0.07041, + "grad_norm": 0.6752147196309027, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7041 + }, + { + "epoch": 0.07042, + "grad_norm": 0.7597403656610957, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 7042 + }, + { + "epoch": 0.07043, + "grad_norm": 0.8702208068936226, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 7043 + }, + { + "epoch": 0.07044, + "grad_norm": 0.997721683934861, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 7044 + }, + { + "epoch": 0.07045, + "grad_norm": 1.0590434408909246, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 7045 + }, + { + "epoch": 0.07046, + "grad_norm": 0.7262778214035177, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7046 + }, + { + "epoch": 0.07047, + "grad_norm": 0.6105175041151503, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 7047 + }, + { + "epoch": 0.07048, + "grad_norm": 0.6919653137461556, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 7048 + }, + { + "epoch": 0.07049, + "grad_norm": 0.7158702927734397, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 7049 + }, + { + "epoch": 0.0705, + "grad_norm": 0.7091325935961303, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 7050 + }, + { + "epoch": 0.07051, + "grad_norm": 0.8551664598496885, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7051 + }, + { + "epoch": 0.07052, + "grad_norm": 0.9183031935565956, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 7052 + }, + { + "epoch": 0.07053, + "grad_norm": 0.9122596997693425, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7053 + }, + { + "epoch": 0.07054, + "grad_norm": 1.0481518040269315, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7054 + }, + { + "epoch": 0.07055, + "grad_norm": 0.8836917687242147, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7055 + }, + { + "epoch": 0.07056, + "grad_norm": 0.8241724236620276, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 7056 + }, + { + "epoch": 0.07057, + "grad_norm": 0.8198719505381878, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 7057 + }, + { + "epoch": 0.07058, + "grad_norm": 1.0373640533731205, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7058 + }, + { + "epoch": 0.07059, + "grad_norm": 1.1298985200213496, + "learning_rate": 0.003, + "loss": 4.115, + "step": 7059 + }, + { + "epoch": 0.0706, + "grad_norm": 0.917680527288264, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7060 + }, + { + "epoch": 0.07061, + "grad_norm": 1.1487741663485906, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 7061 + }, + { + "epoch": 0.07062, + "grad_norm": 0.9765073918800828, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 7062 + }, + { + "epoch": 0.07063, + "grad_norm": 1.0320830994762968, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 7063 + }, + { + "epoch": 0.07064, + "grad_norm": 0.8433226609779286, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 7064 + }, + { + "epoch": 0.07065, + "grad_norm": 0.7826897468256804, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 7065 + }, + { + "epoch": 0.07066, + "grad_norm": 0.6869110843573996, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7066 + }, + { + "epoch": 0.07067, + "grad_norm": 0.6492482664561744, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 7067 + }, + { + "epoch": 0.07068, + "grad_norm": 0.6736220554775696, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 7068 + }, + { + "epoch": 0.07069, + "grad_norm": 0.803916767312363, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 7069 + }, + { + "epoch": 0.0707, + "grad_norm": 1.1648389249713418, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 7070 + }, + { + "epoch": 0.07071, + "grad_norm": 1.05402571939476, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7071 + }, + { + "epoch": 0.07072, + "grad_norm": 0.8354323461428795, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 7072 + }, + { + "epoch": 0.07073, + "grad_norm": 0.799265495194416, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 7073 + }, + { + "epoch": 0.07074, + "grad_norm": 0.8870813126594803, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 7074 + }, + { + "epoch": 0.07075, + "grad_norm": 0.9983580451793913, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 7075 + }, + { + "epoch": 0.07076, + "grad_norm": 0.9665581584289197, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7076 + }, + { + "epoch": 0.07077, + "grad_norm": 0.8830123185073006, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 7077 + }, + { + "epoch": 0.07078, + "grad_norm": 0.8590966319766036, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 7078 + }, + { + "epoch": 0.07079, + "grad_norm": 0.8754624480453634, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 7079 + }, + { + "epoch": 0.0708, + "grad_norm": 0.8357253588784314, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 7080 + }, + { + "epoch": 0.07081, + "grad_norm": 0.6853886151525752, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7081 + }, + { + "epoch": 0.07082, + "grad_norm": 0.746488253274503, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 7082 + }, + { + "epoch": 0.07083, + "grad_norm": 0.7589362058517924, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 7083 + }, + { + "epoch": 0.07084, + "grad_norm": 0.9173953249459876, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 7084 + }, + { + "epoch": 0.07085, + "grad_norm": 1.1273331172700023, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 7085 + }, + { + "epoch": 0.07086, + "grad_norm": 0.9129945566559451, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7086 + }, + { + "epoch": 0.07087, + "grad_norm": 0.782084497555971, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7087 + }, + { + "epoch": 0.07088, + "grad_norm": 0.7191976397005901, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7088 + }, + { + "epoch": 0.07089, + "grad_norm": 0.7254306742811141, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 7089 + }, + { + "epoch": 0.0709, + "grad_norm": 0.936953014758606, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 7090 + }, + { + "epoch": 0.07091, + "grad_norm": 1.1774180392198823, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 7091 + }, + { + "epoch": 0.07092, + "grad_norm": 0.8878181062918875, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 7092 + }, + { + "epoch": 0.07093, + "grad_norm": 0.9002651333364067, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 7093 + }, + { + "epoch": 0.07094, + "grad_norm": 0.9329648399723189, + "learning_rate": 0.003, + "loss": 4.1436, + "step": 7094 + }, + { + "epoch": 0.07095, + "grad_norm": 0.9632962134461859, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 7095 + }, + { + "epoch": 0.07096, + "grad_norm": 0.9420914940041324, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 7096 + }, + { + "epoch": 0.07097, + "grad_norm": 1.141780989450593, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 7097 + }, + { + "epoch": 0.07098, + "grad_norm": 1.200591969334309, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 7098 + }, + { + "epoch": 0.07099, + "grad_norm": 0.8009834169786036, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 7099 + }, + { + "epoch": 0.071, + "grad_norm": 0.7941069430643708, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 7100 + }, + { + "epoch": 0.07101, + "grad_norm": 0.8289765206226791, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 7101 + }, + { + "epoch": 0.07102, + "grad_norm": 0.7502407355058645, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7102 + }, + { + "epoch": 0.07103, + "grad_norm": 0.7118775207721677, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7103 + }, + { + "epoch": 0.07104, + "grad_norm": 0.7238556977567164, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7104 + }, + { + "epoch": 0.07105, + "grad_norm": 0.6978584610652161, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 7105 + }, + { + "epoch": 0.07106, + "grad_norm": 0.7371502133162132, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 7106 + }, + { + "epoch": 0.07107, + "grad_norm": 0.8360787228838016, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 7107 + }, + { + "epoch": 0.07108, + "grad_norm": 0.9175088055451589, + "learning_rate": 0.003, + "loss": 4.12, + "step": 7108 + }, + { + "epoch": 0.07109, + "grad_norm": 1.162644092744131, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7109 + }, + { + "epoch": 0.0711, + "grad_norm": 0.9424749205513693, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7110 + }, + { + "epoch": 0.07111, + "grad_norm": 1.017081150522671, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7111 + }, + { + "epoch": 0.07112, + "grad_norm": 1.0662436592570368, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7112 + }, + { + "epoch": 0.07113, + "grad_norm": 1.054393130448686, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 7113 + }, + { + "epoch": 0.07114, + "grad_norm": 0.9257779285013658, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 7114 + }, + { + "epoch": 0.07115, + "grad_norm": 0.9034958469747977, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7115 + }, + { + "epoch": 0.07116, + "grad_norm": 0.8243521449434865, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7116 + }, + { + "epoch": 0.07117, + "grad_norm": 0.7541154662606151, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 7117 + }, + { + "epoch": 0.07118, + "grad_norm": 0.8201537536206699, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7118 + }, + { + "epoch": 0.07119, + "grad_norm": 0.8555300267392018, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 7119 + }, + { + "epoch": 0.0712, + "grad_norm": 0.9725671974482196, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7120 + }, + { + "epoch": 0.07121, + "grad_norm": 0.9287551042992966, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 7121 + }, + { + "epoch": 0.07122, + "grad_norm": 0.8730257071456973, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 7122 + }, + { + "epoch": 0.07123, + "grad_norm": 0.9240063776082804, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7123 + }, + { + "epoch": 0.07124, + "grad_norm": 0.8897100879971401, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7124 + }, + { + "epoch": 0.07125, + "grad_norm": 0.7255092544625131, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 7125 + }, + { + "epoch": 0.07126, + "grad_norm": 0.64410022138994, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 7126 + }, + { + "epoch": 0.07127, + "grad_norm": 0.6757358271574939, + "learning_rate": 0.003, + "loss": 4.094, + "step": 7127 + }, + { + "epoch": 0.07128, + "grad_norm": 0.8691171001871797, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 7128 + }, + { + "epoch": 0.07129, + "grad_norm": 1.1128442451767677, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 7129 + }, + { + "epoch": 0.0713, + "grad_norm": 1.0139741345763373, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 7130 + }, + { + "epoch": 0.07131, + "grad_norm": 0.9159528876799511, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 7131 + }, + { + "epoch": 0.07132, + "grad_norm": 0.7898416755233202, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 7132 + }, + { + "epoch": 0.07133, + "grad_norm": 0.7848369474841521, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 7133 + }, + { + "epoch": 0.07134, + "grad_norm": 0.7751588573196545, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 7134 + }, + { + "epoch": 0.07135, + "grad_norm": 0.8777005315917537, + "learning_rate": 0.003, + "loss": 4.104, + "step": 7135 + }, + { + "epoch": 0.07136, + "grad_norm": 1.039823448097075, + "learning_rate": 0.003, + "loss": 4.15, + "step": 7136 + }, + { + "epoch": 0.07137, + "grad_norm": 0.9365493853127754, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 7137 + }, + { + "epoch": 0.07138, + "grad_norm": 1.0284030256383379, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7138 + }, + { + "epoch": 0.07139, + "grad_norm": 1.1288999030524856, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7139 + }, + { + "epoch": 0.0714, + "grad_norm": 1.057387996892944, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 7140 + }, + { + "epoch": 0.07141, + "grad_norm": 1.068538739459718, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7141 + }, + { + "epoch": 0.07142, + "grad_norm": 0.9626388123201881, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 7142 + }, + { + "epoch": 0.07143, + "grad_norm": 0.9093794766008101, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 7143 + }, + { + "epoch": 0.07144, + "grad_norm": 0.7746289446671863, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7144 + }, + { + "epoch": 0.07145, + "grad_norm": 0.7339182870303272, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7145 + }, + { + "epoch": 0.07146, + "grad_norm": 0.733959174720137, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7146 + }, + { + "epoch": 0.07147, + "grad_norm": 0.6952850136145936, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 7147 + }, + { + "epoch": 0.07148, + "grad_norm": 0.7615226403621356, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 7148 + }, + { + "epoch": 0.07149, + "grad_norm": 0.8936711792505057, + "learning_rate": 0.003, + "loss": 4.12, + "step": 7149 + }, + { + "epoch": 0.0715, + "grad_norm": 1.273915364248639, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 7150 + }, + { + "epoch": 0.07151, + "grad_norm": 0.7956720748067053, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7151 + }, + { + "epoch": 0.07152, + "grad_norm": 0.6838464347817911, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 7152 + }, + { + "epoch": 0.07153, + "grad_norm": 0.704544600749531, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7153 + }, + { + "epoch": 0.07154, + "grad_norm": 0.8027032921273827, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 7154 + }, + { + "epoch": 0.07155, + "grad_norm": 0.8232429663470329, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 7155 + }, + { + "epoch": 0.07156, + "grad_norm": 0.7266009407623805, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 7156 + }, + { + "epoch": 0.07157, + "grad_norm": 0.74357634876916, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 7157 + }, + { + "epoch": 0.07158, + "grad_norm": 0.6961339201769408, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 7158 + }, + { + "epoch": 0.07159, + "grad_norm": 0.7163813835599939, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 7159 + }, + { + "epoch": 0.0716, + "grad_norm": 0.6021787870861519, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 7160 + }, + { + "epoch": 0.07161, + "grad_norm": 0.5996751260865746, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7161 + }, + { + "epoch": 0.07162, + "grad_norm": 0.7383165393295259, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7162 + }, + { + "epoch": 0.07163, + "grad_norm": 0.8624369790330983, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 7163 + }, + { + "epoch": 0.07164, + "grad_norm": 0.9983212047149563, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7164 + }, + { + "epoch": 0.07165, + "grad_norm": 1.1984220097038951, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7165 + }, + { + "epoch": 0.07166, + "grad_norm": 1.0158195249204152, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 7166 + }, + { + "epoch": 0.07167, + "grad_norm": 1.2477237244225603, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 7167 + }, + { + "epoch": 0.07168, + "grad_norm": 0.8759979215569342, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 7168 + }, + { + "epoch": 0.07169, + "grad_norm": 0.9293872636545353, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 7169 + }, + { + "epoch": 0.0717, + "grad_norm": 0.9602891801389307, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 7170 + }, + { + "epoch": 0.07171, + "grad_norm": 1.1372428448834564, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 7171 + }, + { + "epoch": 0.07172, + "grad_norm": 1.1238803169197797, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 7172 + }, + { + "epoch": 0.07173, + "grad_norm": 0.9931562441621579, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 7173 + }, + { + "epoch": 0.07174, + "grad_norm": 0.9403706846070115, + "learning_rate": 0.003, + "loss": 4.098, + "step": 7174 + }, + { + "epoch": 0.07175, + "grad_norm": 0.9295449986782115, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 7175 + }, + { + "epoch": 0.07176, + "grad_norm": 0.911114790596918, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 7176 + }, + { + "epoch": 0.07177, + "grad_norm": 0.7839495659287221, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7177 + }, + { + "epoch": 0.07178, + "grad_norm": 0.7947159198866947, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 7178 + }, + { + "epoch": 0.07179, + "grad_norm": 0.8603443469449197, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 7179 + }, + { + "epoch": 0.0718, + "grad_norm": 0.9414623094280078, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 7180 + }, + { + "epoch": 0.07181, + "grad_norm": 0.9660522402559355, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7181 + }, + { + "epoch": 0.07182, + "grad_norm": 0.9278959841877216, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 7182 + }, + { + "epoch": 0.07183, + "grad_norm": 0.9339118033397995, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 7183 + }, + { + "epoch": 0.07184, + "grad_norm": 1.1778253262437426, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 7184 + }, + { + "epoch": 0.07185, + "grad_norm": 0.894326222770007, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 7185 + }, + { + "epoch": 0.07186, + "grad_norm": 1.0074070633488665, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 7186 + }, + { + "epoch": 0.07187, + "grad_norm": 0.990447843885946, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7187 + }, + { + "epoch": 0.07188, + "grad_norm": 0.964041753045374, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 7188 + }, + { + "epoch": 0.07189, + "grad_norm": 0.8917382273902423, + "learning_rate": 0.003, + "loss": 4.124, + "step": 7189 + }, + { + "epoch": 0.0719, + "grad_norm": 0.8732047816874139, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 7190 + }, + { + "epoch": 0.07191, + "grad_norm": 0.972536173342455, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 7191 + }, + { + "epoch": 0.07192, + "grad_norm": 1.0780834708500115, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7192 + }, + { + "epoch": 0.07193, + "grad_norm": 0.8395279271863723, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 7193 + }, + { + "epoch": 0.07194, + "grad_norm": 0.7648413650848259, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7194 + }, + { + "epoch": 0.07195, + "grad_norm": 0.8198768874993303, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 7195 + }, + { + "epoch": 0.07196, + "grad_norm": 0.9730451749490292, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7196 + }, + { + "epoch": 0.07197, + "grad_norm": 1.078359113639344, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 7197 + }, + { + "epoch": 0.07198, + "grad_norm": 0.872979660810934, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7198 + }, + { + "epoch": 0.07199, + "grad_norm": 0.7594430118398644, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 7199 + }, + { + "epoch": 0.072, + "grad_norm": 0.8418417578675667, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 7200 + }, + { + "epoch": 0.07201, + "grad_norm": 1.0033287381546427, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 7201 + }, + { + "epoch": 0.07202, + "grad_norm": 1.091821079155893, + "learning_rate": 0.003, + "loss": 4.137, + "step": 7202 + }, + { + "epoch": 0.07203, + "grad_norm": 0.8990512736019646, + "learning_rate": 0.003, + "loss": 4.097, + "step": 7203 + }, + { + "epoch": 0.07204, + "grad_norm": 0.8939835184490502, + "learning_rate": 0.003, + "loss": 4.134, + "step": 7204 + }, + { + "epoch": 0.07205, + "grad_norm": 1.0106951711454872, + "learning_rate": 0.003, + "loss": 4.136, + "step": 7205 + }, + { + "epoch": 0.07206, + "grad_norm": 1.0501076234127416, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 7206 + }, + { + "epoch": 0.07207, + "grad_norm": 0.8683748356773303, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7207 + }, + { + "epoch": 0.07208, + "grad_norm": 0.9807707638242467, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7208 + }, + { + "epoch": 0.07209, + "grad_norm": 1.313720322166701, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 7209 + }, + { + "epoch": 0.0721, + "grad_norm": 0.9036061216874292, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7210 + }, + { + "epoch": 0.07211, + "grad_norm": 0.8532801365525488, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7211 + }, + { + "epoch": 0.07212, + "grad_norm": 0.8177440544033299, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 7212 + }, + { + "epoch": 0.07213, + "grad_norm": 0.7877402885555771, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 7213 + }, + { + "epoch": 0.07214, + "grad_norm": 0.7962574349869463, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 7214 + }, + { + "epoch": 0.07215, + "grad_norm": 0.840821954584076, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7215 + }, + { + "epoch": 0.07216, + "grad_norm": 1.0202453719622064, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 7216 + }, + { + "epoch": 0.07217, + "grad_norm": 1.0210177210663058, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 7217 + }, + { + "epoch": 0.07218, + "grad_norm": 0.9291393542941405, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7218 + }, + { + "epoch": 0.07219, + "grad_norm": 0.9125365057037023, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7219 + }, + { + "epoch": 0.0722, + "grad_norm": 0.8684648065476432, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 7220 + }, + { + "epoch": 0.07221, + "grad_norm": 0.8203475478863702, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 7221 + }, + { + "epoch": 0.07222, + "grad_norm": 0.7337104645338505, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 7222 + }, + { + "epoch": 0.07223, + "grad_norm": 0.7540489703643396, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 7223 + }, + { + "epoch": 0.07224, + "grad_norm": 0.791141223365698, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 7224 + }, + { + "epoch": 0.07225, + "grad_norm": 0.8289724417410262, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7225 + }, + { + "epoch": 0.07226, + "grad_norm": 0.7971187373872961, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 7226 + }, + { + "epoch": 0.07227, + "grad_norm": 0.8010700923864582, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 7227 + }, + { + "epoch": 0.07228, + "grad_norm": 0.867491024834857, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 7228 + }, + { + "epoch": 0.07229, + "grad_norm": 0.9726581300415823, + "learning_rate": 0.003, + "loss": 4.105, + "step": 7229 + }, + { + "epoch": 0.0723, + "grad_norm": 1.0772680849810623, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 7230 + }, + { + "epoch": 0.07231, + "grad_norm": 1.0210940982142689, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 7231 + }, + { + "epoch": 0.07232, + "grad_norm": 0.8347979960369515, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 7232 + }, + { + "epoch": 0.07233, + "grad_norm": 0.7798104407644068, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 7233 + }, + { + "epoch": 0.07234, + "grad_norm": 0.7699473287189199, + "learning_rate": 0.003, + "loss": 4.081, + "step": 7234 + }, + { + "epoch": 0.07235, + "grad_norm": 0.8633138422643347, + "learning_rate": 0.003, + "loss": 4.133, + "step": 7235 + }, + { + "epoch": 0.07236, + "grad_norm": 1.0997904141948405, + "learning_rate": 0.003, + "loss": 4.138, + "step": 7236 + }, + { + "epoch": 0.07237, + "grad_norm": 0.9961327053280845, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 7237 + }, + { + "epoch": 0.07238, + "grad_norm": 0.8483548613066104, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 7238 + }, + { + "epoch": 0.07239, + "grad_norm": 0.7701019036531008, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 7239 + }, + { + "epoch": 0.0724, + "grad_norm": 0.7417934914192521, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 7240 + }, + { + "epoch": 0.07241, + "grad_norm": 0.7191524525608475, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 7241 + }, + { + "epoch": 0.07242, + "grad_norm": 0.8072871682611079, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 7242 + }, + { + "epoch": 0.07243, + "grad_norm": 0.8184708621816302, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 7243 + }, + { + "epoch": 0.07244, + "grad_norm": 0.7093054794981116, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 7244 + }, + { + "epoch": 0.07245, + "grad_norm": 0.675330894388966, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 7245 + }, + { + "epoch": 0.07246, + "grad_norm": 0.7323266139017566, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 7246 + }, + { + "epoch": 0.07247, + "grad_norm": 0.780943196371678, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 7247 + }, + { + "epoch": 0.07248, + "grad_norm": 0.899671652675055, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 7248 + }, + { + "epoch": 0.07249, + "grad_norm": 1.1022793278108771, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 7249 + }, + { + "epoch": 0.0725, + "grad_norm": 1.1608600206621666, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 7250 + }, + { + "epoch": 0.07251, + "grad_norm": 0.9173391017419072, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7251 + }, + { + "epoch": 0.07252, + "grad_norm": 0.8543519564138532, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7252 + }, + { + "epoch": 0.07253, + "grad_norm": 0.7929389671267115, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 7253 + }, + { + "epoch": 0.07254, + "grad_norm": 0.8566224768735904, + "learning_rate": 0.003, + "loss": 4.125, + "step": 7254 + }, + { + "epoch": 0.07255, + "grad_norm": 0.9509981257694082, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 7255 + }, + { + "epoch": 0.07256, + "grad_norm": 1.0323475737017942, + "learning_rate": 0.003, + "loss": 4.129, + "step": 7256 + }, + { + "epoch": 0.07257, + "grad_norm": 1.125862787361634, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 7257 + }, + { + "epoch": 0.07258, + "grad_norm": 0.9874067143143448, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7258 + }, + { + "epoch": 0.07259, + "grad_norm": 1.1005675012845613, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7259 + }, + { + "epoch": 0.0726, + "grad_norm": 0.8454037889791309, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 7260 + }, + { + "epoch": 0.07261, + "grad_norm": 0.856806639173543, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7261 + }, + { + "epoch": 0.07262, + "grad_norm": 1.1774126393139672, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 7262 + }, + { + "epoch": 0.07263, + "grad_norm": 0.98856001091982, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 7263 + }, + { + "epoch": 0.07264, + "grad_norm": 1.052918037838861, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 7264 + }, + { + "epoch": 0.07265, + "grad_norm": 0.9481848938716576, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 7265 + }, + { + "epoch": 0.07266, + "grad_norm": 0.8686413426498576, + "learning_rate": 0.003, + "loss": 4.13, + "step": 7266 + }, + { + "epoch": 0.07267, + "grad_norm": 0.8525273425159318, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 7267 + }, + { + "epoch": 0.07268, + "grad_norm": 0.9533691584951004, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 7268 + }, + { + "epoch": 0.07269, + "grad_norm": 0.9840691218693606, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 7269 + }, + { + "epoch": 0.0727, + "grad_norm": 1.022133557552684, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7270 + }, + { + "epoch": 0.07271, + "grad_norm": 0.9427872988130471, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 7271 + }, + { + "epoch": 0.07272, + "grad_norm": 0.8835705515323187, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 7272 + }, + { + "epoch": 0.07273, + "grad_norm": 0.7662262254321305, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 7273 + }, + { + "epoch": 0.07274, + "grad_norm": 0.6861881669104918, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 7274 + }, + { + "epoch": 0.07275, + "grad_norm": 0.6940748807512555, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 7275 + }, + { + "epoch": 0.07276, + "grad_norm": 0.7255703327321986, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7276 + }, + { + "epoch": 0.07277, + "grad_norm": 0.826898534170412, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 7277 + }, + { + "epoch": 0.07278, + "grad_norm": 0.7205286139763815, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 7278 + }, + { + "epoch": 0.07279, + "grad_norm": 0.7870100781810526, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 7279 + }, + { + "epoch": 0.0728, + "grad_norm": 1.1224906772039385, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 7280 + }, + { + "epoch": 0.07281, + "grad_norm": 1.064756942044224, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 7281 + }, + { + "epoch": 0.07282, + "grad_norm": 0.755328815035219, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 7282 + }, + { + "epoch": 0.07283, + "grad_norm": 0.7480564984223254, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 7283 + }, + { + "epoch": 0.07284, + "grad_norm": 0.7969334161992938, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 7284 + }, + { + "epoch": 0.07285, + "grad_norm": 0.8037999842824531, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 7285 + }, + { + "epoch": 0.07286, + "grad_norm": 0.8432356062305908, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 7286 + }, + { + "epoch": 0.07287, + "grad_norm": 0.7074733000617406, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7287 + }, + { + "epoch": 0.07288, + "grad_norm": 0.6862151638972047, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 7288 + }, + { + "epoch": 0.07289, + "grad_norm": 0.7256205161629711, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 7289 + }, + { + "epoch": 0.0729, + "grad_norm": 0.7506778983987582, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7290 + }, + { + "epoch": 0.07291, + "grad_norm": 0.9175171243997577, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 7291 + }, + { + "epoch": 0.07292, + "grad_norm": 1.1748119483133153, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 7292 + }, + { + "epoch": 0.07293, + "grad_norm": 0.946454415717222, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 7293 + }, + { + "epoch": 0.07294, + "grad_norm": 0.8588138723902505, + "learning_rate": 0.003, + "loss": 4.079, + "step": 7294 + }, + { + "epoch": 0.07295, + "grad_norm": 0.9200655141576249, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 7295 + }, + { + "epoch": 0.07296, + "grad_norm": 0.945584786641691, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 7296 + }, + { + "epoch": 0.07297, + "grad_norm": 0.9076831566613616, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7297 + }, + { + "epoch": 0.07298, + "grad_norm": 0.9132887839709575, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7298 + }, + { + "epoch": 0.07299, + "grad_norm": 1.0600320642602021, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 7299 + }, + { + "epoch": 0.073, + "grad_norm": 1.1555254508950479, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 7300 + }, + { + "epoch": 0.07301, + "grad_norm": 0.8091012187372432, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 7301 + }, + { + "epoch": 0.07302, + "grad_norm": 0.7969703558310967, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 7302 + }, + { + "epoch": 0.07303, + "grad_norm": 0.8742412667236584, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 7303 + }, + { + "epoch": 0.07304, + "grad_norm": 0.9295281273271904, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 7304 + }, + { + "epoch": 0.07305, + "grad_norm": 0.9710795112852725, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 7305 + }, + { + "epoch": 0.07306, + "grad_norm": 0.969596840488141, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 7306 + }, + { + "epoch": 0.07307, + "grad_norm": 1.1571657345709758, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 7307 + }, + { + "epoch": 0.07308, + "grad_norm": 0.8780928378824396, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 7308 + }, + { + "epoch": 0.07309, + "grad_norm": 0.7381472139260883, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7309 + }, + { + "epoch": 0.0731, + "grad_norm": 0.7936367882676433, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7310 + }, + { + "epoch": 0.07311, + "grad_norm": 0.8771254185491402, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 7311 + }, + { + "epoch": 0.07312, + "grad_norm": 0.9049549301553881, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 7312 + }, + { + "epoch": 0.07313, + "grad_norm": 0.8671243317374602, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 7313 + }, + { + "epoch": 0.07314, + "grad_norm": 0.8712423511074834, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 7314 + }, + { + "epoch": 0.07315, + "grad_norm": 0.9012186242781552, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7315 + }, + { + "epoch": 0.07316, + "grad_norm": 1.00783380326016, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 7316 + }, + { + "epoch": 0.07317, + "grad_norm": 0.9312278344910015, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 7317 + }, + { + "epoch": 0.07318, + "grad_norm": 0.9073283549997471, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 7318 + }, + { + "epoch": 0.07319, + "grad_norm": 0.9182845586703946, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 7319 + }, + { + "epoch": 0.0732, + "grad_norm": 1.0130838674437885, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7320 + }, + { + "epoch": 0.07321, + "grad_norm": 1.1175547994945165, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 7321 + }, + { + "epoch": 0.07322, + "grad_norm": 0.9871197560308737, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 7322 + }, + { + "epoch": 0.07323, + "grad_norm": 0.9778441606954984, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7323 + }, + { + "epoch": 0.07324, + "grad_norm": 0.8298462372409252, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 7324 + }, + { + "epoch": 0.07325, + "grad_norm": 0.7996270399874803, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 7325 + }, + { + "epoch": 0.07326, + "grad_norm": 0.8480564833291753, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 7326 + }, + { + "epoch": 0.07327, + "grad_norm": 0.9655678078823727, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 7327 + }, + { + "epoch": 0.07328, + "grad_norm": 0.9383954669497225, + "learning_rate": 0.003, + "loss": 4.081, + "step": 7328 + }, + { + "epoch": 0.07329, + "grad_norm": 0.9699028773325669, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7329 + }, + { + "epoch": 0.0733, + "grad_norm": 0.9768618937856877, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 7330 + }, + { + "epoch": 0.07331, + "grad_norm": 0.9467415617801669, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 7331 + }, + { + "epoch": 0.07332, + "grad_norm": 0.8567564686017997, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 7332 + }, + { + "epoch": 0.07333, + "grad_norm": 0.84392576775453, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 7333 + }, + { + "epoch": 0.07334, + "grad_norm": 0.8137901445810288, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7334 + }, + { + "epoch": 0.07335, + "grad_norm": 0.9003483859560396, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 7335 + }, + { + "epoch": 0.07336, + "grad_norm": 1.0011860597622482, + "learning_rate": 0.003, + "loss": 4.123, + "step": 7336 + }, + { + "epoch": 0.07337, + "grad_norm": 1.0810806000621536, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 7337 + }, + { + "epoch": 0.07338, + "grad_norm": 0.987420151810005, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 7338 + }, + { + "epoch": 0.07339, + "grad_norm": 0.8647937623676709, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 7339 + }, + { + "epoch": 0.0734, + "grad_norm": 0.823939663089722, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 7340 + }, + { + "epoch": 0.07341, + "grad_norm": 0.8020510247193195, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7341 + }, + { + "epoch": 0.07342, + "grad_norm": 0.8621826681855989, + "learning_rate": 0.003, + "loss": 4.126, + "step": 7342 + }, + { + "epoch": 0.07343, + "grad_norm": 0.96982132901119, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 7343 + }, + { + "epoch": 0.07344, + "grad_norm": 1.0089151394572116, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7344 + }, + { + "epoch": 0.07345, + "grad_norm": 1.0012381625820561, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 7345 + }, + { + "epoch": 0.07346, + "grad_norm": 1.0578613234075809, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 7346 + }, + { + "epoch": 0.07347, + "grad_norm": 0.9368355879723315, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 7347 + }, + { + "epoch": 0.07348, + "grad_norm": 0.98538363953966, + "learning_rate": 0.003, + "loss": 4.105, + "step": 7348 + }, + { + "epoch": 0.07349, + "grad_norm": 1.0421211993700539, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 7349 + }, + { + "epoch": 0.0735, + "grad_norm": 0.8354136210538131, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 7350 + }, + { + "epoch": 0.07351, + "grad_norm": 0.7274907728499738, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7351 + }, + { + "epoch": 0.07352, + "grad_norm": 0.7679883952167378, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 7352 + }, + { + "epoch": 0.07353, + "grad_norm": 0.8616796265202713, + "learning_rate": 0.003, + "loss": 4.126, + "step": 7353 + }, + { + "epoch": 0.07354, + "grad_norm": 0.9798851246937682, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 7354 + }, + { + "epoch": 0.07355, + "grad_norm": 1.0511225960032606, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 7355 + }, + { + "epoch": 0.07356, + "grad_norm": 0.9587450442062034, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 7356 + }, + { + "epoch": 0.07357, + "grad_norm": 0.8805872619152303, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 7357 + }, + { + "epoch": 0.07358, + "grad_norm": 0.849654200097335, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 7358 + }, + { + "epoch": 0.07359, + "grad_norm": 0.8872628692069536, + "learning_rate": 0.003, + "loss": 4.11, + "step": 7359 + }, + { + "epoch": 0.0736, + "grad_norm": 1.0989657984839474, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 7360 + }, + { + "epoch": 0.07361, + "grad_norm": 0.8695135396560348, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 7361 + }, + { + "epoch": 0.07362, + "grad_norm": 0.8528139392382852, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 7362 + }, + { + "epoch": 0.07363, + "grad_norm": 0.8772352719466499, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 7363 + }, + { + "epoch": 0.07364, + "grad_norm": 0.9572152378472161, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 7364 + }, + { + "epoch": 0.07365, + "grad_norm": 1.2418185340432195, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 7365 + }, + { + "epoch": 0.07366, + "grad_norm": 0.7569511177947986, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 7366 + }, + { + "epoch": 0.07367, + "grad_norm": 0.7408105950286488, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7367 + }, + { + "epoch": 0.07368, + "grad_norm": 0.9113307753760953, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 7368 + }, + { + "epoch": 0.07369, + "grad_norm": 1.1054144696822306, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 7369 + }, + { + "epoch": 0.0737, + "grad_norm": 0.9173693391865863, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 7370 + }, + { + "epoch": 0.07371, + "grad_norm": 0.7126602835933623, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 7371 + }, + { + "epoch": 0.07372, + "grad_norm": 0.7070143475341606, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7372 + }, + { + "epoch": 0.07373, + "grad_norm": 0.7912761016620928, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 7373 + }, + { + "epoch": 0.07374, + "grad_norm": 0.7382445142368581, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 7374 + }, + { + "epoch": 0.07375, + "grad_norm": 0.792351842888862, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 7375 + }, + { + "epoch": 0.07376, + "grad_norm": 0.8906583846931676, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 7376 + }, + { + "epoch": 0.07377, + "grad_norm": 0.9447731004050475, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 7377 + }, + { + "epoch": 0.07378, + "grad_norm": 0.8173782820222226, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 7378 + }, + { + "epoch": 0.07379, + "grad_norm": 0.8061166590769808, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7379 + }, + { + "epoch": 0.0738, + "grad_norm": 0.9117507479699275, + "learning_rate": 0.003, + "loss": 4.073, + "step": 7380 + }, + { + "epoch": 0.07381, + "grad_norm": 0.8833902157946827, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 7381 + }, + { + "epoch": 0.07382, + "grad_norm": 0.859726728050116, + "learning_rate": 0.003, + "loss": 4.134, + "step": 7382 + }, + { + "epoch": 0.07383, + "grad_norm": 0.8848213921508847, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7383 + }, + { + "epoch": 0.07384, + "grad_norm": 1.0298301316190803, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 7384 + }, + { + "epoch": 0.07385, + "grad_norm": 1.172557569337942, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7385 + }, + { + "epoch": 0.07386, + "grad_norm": 0.8669312655544931, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7386 + }, + { + "epoch": 0.07387, + "grad_norm": 0.8392913033776324, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 7387 + }, + { + "epoch": 0.07388, + "grad_norm": 0.7447523316906665, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 7388 + }, + { + "epoch": 0.07389, + "grad_norm": 0.6495275179258706, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 7389 + }, + { + "epoch": 0.0739, + "grad_norm": 0.6025258391763327, + "learning_rate": 0.003, + "loss": 4.08, + "step": 7390 + }, + { + "epoch": 0.07391, + "grad_norm": 0.6318049445146262, + "learning_rate": 0.003, + "loss": 4.105, + "step": 7391 + }, + { + "epoch": 0.07392, + "grad_norm": 0.7343060747804614, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7392 + }, + { + "epoch": 0.07393, + "grad_norm": 1.0000247571371808, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7393 + }, + { + "epoch": 0.07394, + "grad_norm": 1.5060068670788715, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 7394 + }, + { + "epoch": 0.07395, + "grad_norm": 0.6499300765174943, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7395 + }, + { + "epoch": 0.07396, + "grad_norm": 0.8568555410525963, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 7396 + }, + { + "epoch": 0.07397, + "grad_norm": 0.9063071745959117, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 7397 + }, + { + "epoch": 0.07398, + "grad_norm": 0.9019215085358515, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 7398 + }, + { + "epoch": 0.07399, + "grad_norm": 0.9546957281863762, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 7399 + }, + { + "epoch": 0.074, + "grad_norm": 0.8849692017723262, + "learning_rate": 0.003, + "loss": 4.118, + "step": 7400 + }, + { + "epoch": 0.07401, + "grad_norm": 0.8546205777376308, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 7401 + }, + { + "epoch": 0.07402, + "grad_norm": 0.7032841794462815, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7402 + }, + { + "epoch": 0.07403, + "grad_norm": 0.7639352112565744, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 7403 + }, + { + "epoch": 0.07404, + "grad_norm": 0.8684415164326015, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 7404 + }, + { + "epoch": 0.07405, + "grad_norm": 1.0166394680109212, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 7405 + }, + { + "epoch": 0.07406, + "grad_norm": 1.1062291356675835, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7406 + }, + { + "epoch": 0.07407, + "grad_norm": 0.9316736131089458, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 7407 + }, + { + "epoch": 0.07408, + "grad_norm": 0.8517821079613881, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7408 + }, + { + "epoch": 0.07409, + "grad_norm": 0.8096451130433067, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7409 + }, + { + "epoch": 0.0741, + "grad_norm": 0.7521940521106885, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 7410 + }, + { + "epoch": 0.07411, + "grad_norm": 0.7925702752865725, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 7411 + }, + { + "epoch": 0.07412, + "grad_norm": 0.8215884643463309, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7412 + }, + { + "epoch": 0.07413, + "grad_norm": 0.9250271828845864, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7413 + }, + { + "epoch": 0.07414, + "grad_norm": 1.1770944944492254, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 7414 + }, + { + "epoch": 0.07415, + "grad_norm": 1.0576816988803364, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 7415 + }, + { + "epoch": 0.07416, + "grad_norm": 0.9147150961080485, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 7416 + }, + { + "epoch": 0.07417, + "grad_norm": 0.8841312717180329, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 7417 + }, + { + "epoch": 0.07418, + "grad_norm": 1.0311827398926998, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 7418 + }, + { + "epoch": 0.07419, + "grad_norm": 1.0768166118437081, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 7419 + }, + { + "epoch": 0.0742, + "grad_norm": 0.988151970685207, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 7420 + }, + { + "epoch": 0.07421, + "grad_norm": 0.9184883764372472, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 7421 + }, + { + "epoch": 0.07422, + "grad_norm": 0.8685292845726943, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7422 + }, + { + "epoch": 0.07423, + "grad_norm": 0.8906473884300916, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 7423 + }, + { + "epoch": 0.07424, + "grad_norm": 1.0992726932466628, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 7424 + }, + { + "epoch": 0.07425, + "grad_norm": 1.169164042468968, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 7425 + }, + { + "epoch": 0.07426, + "grad_norm": 0.8731095700449631, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 7426 + }, + { + "epoch": 0.07427, + "grad_norm": 0.8974625019426375, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 7427 + }, + { + "epoch": 0.07428, + "grad_norm": 0.8619370872170999, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 7428 + }, + { + "epoch": 0.07429, + "grad_norm": 0.8606884636965143, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7429 + }, + { + "epoch": 0.0743, + "grad_norm": 0.9104655756270731, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 7430 + }, + { + "epoch": 0.07431, + "grad_norm": 1.0461049195727867, + "learning_rate": 0.003, + "loss": 4.096, + "step": 7431 + }, + { + "epoch": 0.07432, + "grad_norm": 1.030513769267123, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 7432 + }, + { + "epoch": 0.07433, + "grad_norm": 0.7826646595179413, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 7433 + }, + { + "epoch": 0.07434, + "grad_norm": 0.8449733596632347, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 7434 + }, + { + "epoch": 0.07435, + "grad_norm": 0.829034109699188, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7435 + }, + { + "epoch": 0.07436, + "grad_norm": 0.7276311889102272, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 7436 + }, + { + "epoch": 0.07437, + "grad_norm": 0.6962567087169638, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7437 + }, + { + "epoch": 0.07438, + "grad_norm": 0.754150337723482, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7438 + }, + { + "epoch": 0.07439, + "grad_norm": 0.8394913294621994, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7439 + }, + { + "epoch": 0.0744, + "grad_norm": 0.9210027114951105, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7440 + }, + { + "epoch": 0.07441, + "grad_norm": 0.9677662290482478, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 7441 + }, + { + "epoch": 0.07442, + "grad_norm": 0.9984463151295965, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 7442 + }, + { + "epoch": 0.07443, + "grad_norm": 0.8595131868015653, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 7443 + }, + { + "epoch": 0.07444, + "grad_norm": 0.7683813726273543, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 7444 + }, + { + "epoch": 0.07445, + "grad_norm": 0.753682686098993, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7445 + }, + { + "epoch": 0.07446, + "grad_norm": 0.8268762150034119, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 7446 + }, + { + "epoch": 0.07447, + "grad_norm": 1.094602286197476, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7447 + }, + { + "epoch": 0.07448, + "grad_norm": 0.9771739644265591, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 7448 + }, + { + "epoch": 0.07449, + "grad_norm": 0.8384542889699795, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 7449 + }, + { + "epoch": 0.0745, + "grad_norm": 0.7590190855947146, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7450 + }, + { + "epoch": 0.07451, + "grad_norm": 0.8937331488483308, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 7451 + }, + { + "epoch": 0.07452, + "grad_norm": 0.9693727157247541, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 7452 + }, + { + "epoch": 0.07453, + "grad_norm": 1.079219331582194, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 7453 + }, + { + "epoch": 0.07454, + "grad_norm": 0.9744818238494153, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 7454 + }, + { + "epoch": 0.07455, + "grad_norm": 0.9536322310393928, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 7455 + }, + { + "epoch": 0.07456, + "grad_norm": 1.0769846542491077, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7456 + }, + { + "epoch": 0.07457, + "grad_norm": 1.0409702107301997, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 7457 + }, + { + "epoch": 0.07458, + "grad_norm": 1.001295432389869, + "learning_rate": 0.003, + "loss": 4.12, + "step": 7458 + }, + { + "epoch": 0.07459, + "grad_norm": 0.8889128511582081, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 7459 + }, + { + "epoch": 0.0746, + "grad_norm": 0.9411931447305963, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 7460 + }, + { + "epoch": 0.07461, + "grad_norm": 0.8519572681750707, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 7461 + }, + { + "epoch": 0.07462, + "grad_norm": 0.7905764728430406, + "learning_rate": 0.003, + "loss": 4.076, + "step": 7462 + }, + { + "epoch": 0.07463, + "grad_norm": 0.8052221816614337, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 7463 + }, + { + "epoch": 0.07464, + "grad_norm": 0.9184878977339718, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 7464 + }, + { + "epoch": 0.07465, + "grad_norm": 0.9096124012536108, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 7465 + }, + { + "epoch": 0.07466, + "grad_norm": 0.8868750508524639, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 7466 + }, + { + "epoch": 0.07467, + "grad_norm": 1.1739238328463417, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7467 + }, + { + "epoch": 0.07468, + "grad_norm": 1.2731034343951069, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 7468 + }, + { + "epoch": 0.07469, + "grad_norm": 0.9599056408935575, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7469 + }, + { + "epoch": 0.0747, + "grad_norm": 0.8545661960822021, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7470 + }, + { + "epoch": 0.07471, + "grad_norm": 0.8017399886069515, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 7471 + }, + { + "epoch": 0.07472, + "grad_norm": 0.8220636823261487, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7472 + }, + { + "epoch": 0.07473, + "grad_norm": 0.8333913762759374, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 7473 + }, + { + "epoch": 0.07474, + "grad_norm": 0.9949440880989607, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 7474 + }, + { + "epoch": 0.07475, + "grad_norm": 1.0047709217104455, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 7475 + }, + { + "epoch": 0.07476, + "grad_norm": 0.9783980895093549, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 7476 + }, + { + "epoch": 0.07477, + "grad_norm": 1.089186008059605, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 7477 + }, + { + "epoch": 0.07478, + "grad_norm": 1.1072575591847578, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 7478 + }, + { + "epoch": 0.07479, + "grad_norm": 1.1298579787016074, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 7479 + }, + { + "epoch": 0.0748, + "grad_norm": 0.9998908890872064, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 7480 + }, + { + "epoch": 0.07481, + "grad_norm": 0.8169704618263066, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 7481 + }, + { + "epoch": 0.07482, + "grad_norm": 0.6487475847193562, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 7482 + }, + { + "epoch": 0.07483, + "grad_norm": 0.6863949014856395, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 7483 + }, + { + "epoch": 0.07484, + "grad_norm": 0.6587103027574833, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 7484 + }, + { + "epoch": 0.07485, + "grad_norm": 0.7160032679100043, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 7485 + }, + { + "epoch": 0.07486, + "grad_norm": 0.9672056856968704, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 7486 + }, + { + "epoch": 0.07487, + "grad_norm": 1.1696254951341385, + "learning_rate": 0.003, + "loss": 4.108, + "step": 7487 + }, + { + "epoch": 0.07488, + "grad_norm": 0.7863461489638554, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 7488 + }, + { + "epoch": 0.07489, + "grad_norm": 0.7384563910599533, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 7489 + }, + { + "epoch": 0.0749, + "grad_norm": 0.7572591385660193, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 7490 + }, + { + "epoch": 0.07491, + "grad_norm": 0.9074264182707067, + "learning_rate": 0.003, + "loss": 4.094, + "step": 7491 + }, + { + "epoch": 0.07492, + "grad_norm": 1.1199081661678163, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 7492 + }, + { + "epoch": 0.07493, + "grad_norm": 1.1575662283283366, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 7493 + }, + { + "epoch": 0.07494, + "grad_norm": 0.9434610130517279, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7494 + }, + { + "epoch": 0.07495, + "grad_norm": 0.8748804539205894, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 7495 + }, + { + "epoch": 0.07496, + "grad_norm": 0.8175617506999896, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7496 + }, + { + "epoch": 0.07497, + "grad_norm": 0.9144161517532557, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7497 + }, + { + "epoch": 0.07498, + "grad_norm": 0.9185423406505983, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 7498 + }, + { + "epoch": 0.07499, + "grad_norm": 0.8280717177948074, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 7499 + }, + { + "epoch": 0.075, + "grad_norm": 0.7960540221263401, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 7500 + }, + { + "epoch": 0.07501, + "grad_norm": 0.6399790221522081, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 7501 + }, + { + "epoch": 0.07502, + "grad_norm": 0.6669586492655686, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 7502 + }, + { + "epoch": 0.07503, + "grad_norm": 0.8085394369052066, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 7503 + }, + { + "epoch": 0.07504, + "grad_norm": 1.1164579548991722, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7504 + }, + { + "epoch": 0.07505, + "grad_norm": 1.1764299489396552, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7505 + }, + { + "epoch": 0.07506, + "grad_norm": 0.6778128482933068, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7506 + }, + { + "epoch": 0.07507, + "grad_norm": 0.6887355085828547, + "learning_rate": 0.003, + "loss": 4.075, + "step": 7507 + }, + { + "epoch": 0.07508, + "grad_norm": 0.8305506452463216, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 7508 + }, + { + "epoch": 0.07509, + "grad_norm": 0.9381005410193313, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7509 + }, + { + "epoch": 0.0751, + "grad_norm": 0.9286555796672374, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 7510 + }, + { + "epoch": 0.07511, + "grad_norm": 0.8493832587890462, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 7511 + }, + { + "epoch": 0.07512, + "grad_norm": 0.8203675397517558, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 7512 + }, + { + "epoch": 0.07513, + "grad_norm": 0.9987084064424658, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7513 + }, + { + "epoch": 0.07514, + "grad_norm": 1.0869538532848078, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7514 + }, + { + "epoch": 0.07515, + "grad_norm": 1.0403169728871233, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 7515 + }, + { + "epoch": 0.07516, + "grad_norm": 1.011790174275216, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 7516 + }, + { + "epoch": 0.07517, + "grad_norm": 0.8942557366391805, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7517 + }, + { + "epoch": 0.07518, + "grad_norm": 0.7974760716257956, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 7518 + }, + { + "epoch": 0.07519, + "grad_norm": 0.7849755190610048, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 7519 + }, + { + "epoch": 0.0752, + "grad_norm": 0.7855656056140153, + "learning_rate": 0.003, + "loss": 4.096, + "step": 7520 + }, + { + "epoch": 0.07521, + "grad_norm": 0.8813805509682716, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 7521 + }, + { + "epoch": 0.07522, + "grad_norm": 0.9965079513901988, + "learning_rate": 0.003, + "loss": 4.089, + "step": 7522 + }, + { + "epoch": 0.07523, + "grad_norm": 1.2142599219211911, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 7523 + }, + { + "epoch": 0.07524, + "grad_norm": 0.7326420889303906, + "learning_rate": 0.003, + "loss": 4.113, + "step": 7524 + }, + { + "epoch": 0.07525, + "grad_norm": 0.7133952113276287, + "learning_rate": 0.003, + "loss": 4.118, + "step": 7525 + }, + { + "epoch": 0.07526, + "grad_norm": 0.7827807395620044, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 7526 + }, + { + "epoch": 0.07527, + "grad_norm": 0.8767676778583422, + "learning_rate": 0.003, + "loss": 4.116, + "step": 7527 + }, + { + "epoch": 0.07528, + "grad_norm": 0.9065608532077762, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 7528 + }, + { + "epoch": 0.07529, + "grad_norm": 0.959630193106546, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 7529 + }, + { + "epoch": 0.0753, + "grad_norm": 0.9481228165926106, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7530 + }, + { + "epoch": 0.07531, + "grad_norm": 1.0127377126343031, + "learning_rate": 0.003, + "loss": 4.125, + "step": 7531 + }, + { + "epoch": 0.07532, + "grad_norm": 1.1381689608302172, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 7532 + }, + { + "epoch": 0.07533, + "grad_norm": 0.9088536862602051, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7533 + }, + { + "epoch": 0.07534, + "grad_norm": 0.9695169216091517, + "learning_rate": 0.003, + "loss": 4.103, + "step": 7534 + }, + { + "epoch": 0.07535, + "grad_norm": 0.9754230578467733, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 7535 + }, + { + "epoch": 0.07536, + "grad_norm": 0.9562163044893859, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 7536 + }, + { + "epoch": 0.07537, + "grad_norm": 1.1305394732850846, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 7537 + }, + { + "epoch": 0.07538, + "grad_norm": 0.9158848504564159, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 7538 + }, + { + "epoch": 0.07539, + "grad_norm": 0.9652651208133781, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 7539 + }, + { + "epoch": 0.0754, + "grad_norm": 0.9022392107906086, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 7540 + }, + { + "epoch": 0.07541, + "grad_norm": 0.8225042354592814, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 7541 + }, + { + "epoch": 0.07542, + "grad_norm": 1.0114606631383205, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 7542 + }, + { + "epoch": 0.07543, + "grad_norm": 1.2258446672383758, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 7543 + }, + { + "epoch": 0.07544, + "grad_norm": 0.9065708338931346, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 7544 + }, + { + "epoch": 0.07545, + "grad_norm": 0.8413566763887645, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 7545 + }, + { + "epoch": 0.07546, + "grad_norm": 0.931446753858521, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7546 + }, + { + "epoch": 0.07547, + "grad_norm": 0.8924179590547977, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 7547 + }, + { + "epoch": 0.07548, + "grad_norm": 0.7981884285863752, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 7548 + }, + { + "epoch": 0.07549, + "grad_norm": 0.847248132212302, + "learning_rate": 0.003, + "loss": 4.128, + "step": 7549 + }, + { + "epoch": 0.0755, + "grad_norm": 0.929686503004784, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 7550 + }, + { + "epoch": 0.07551, + "grad_norm": 1.0899432730160927, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 7551 + }, + { + "epoch": 0.07552, + "grad_norm": 0.9023134743903442, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 7552 + }, + { + "epoch": 0.07553, + "grad_norm": 0.9320609493389578, + "learning_rate": 0.003, + "loss": 4.124, + "step": 7553 + }, + { + "epoch": 0.07554, + "grad_norm": 0.9366611277818084, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 7554 + }, + { + "epoch": 0.07555, + "grad_norm": 1.1249548849939486, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 7555 + }, + { + "epoch": 0.07556, + "grad_norm": 0.8809337436260816, + "learning_rate": 0.003, + "loss": 4.124, + "step": 7556 + }, + { + "epoch": 0.07557, + "grad_norm": 0.8977538234717878, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 7557 + }, + { + "epoch": 0.07558, + "grad_norm": 0.9323094114843662, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7558 + }, + { + "epoch": 0.07559, + "grad_norm": 1.0365418541296352, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 7559 + }, + { + "epoch": 0.0756, + "grad_norm": 0.9700985533938646, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 7560 + }, + { + "epoch": 0.07561, + "grad_norm": 0.9945502187488869, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7561 + }, + { + "epoch": 0.07562, + "grad_norm": 1.0168375224426294, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 7562 + }, + { + "epoch": 0.07563, + "grad_norm": 0.9790876715890733, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7563 + }, + { + "epoch": 0.07564, + "grad_norm": 0.9910540055685099, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 7564 + }, + { + "epoch": 0.07565, + "grad_norm": 1.0341921379761243, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 7565 + }, + { + "epoch": 0.07566, + "grad_norm": 0.9443317462260504, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7566 + }, + { + "epoch": 0.07567, + "grad_norm": 1.0159322899398833, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 7567 + }, + { + "epoch": 0.07568, + "grad_norm": 1.1086490020888524, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 7568 + }, + { + "epoch": 0.07569, + "grad_norm": 0.861288591479629, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 7569 + }, + { + "epoch": 0.0757, + "grad_norm": 1.0504830493537256, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 7570 + }, + { + "epoch": 0.07571, + "grad_norm": 1.041967555005186, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 7571 + }, + { + "epoch": 0.07572, + "grad_norm": 1.0344004502845179, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 7572 + }, + { + "epoch": 0.07573, + "grad_norm": 0.763379777310832, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 7573 + }, + { + "epoch": 0.07574, + "grad_norm": 0.7213694523357767, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 7574 + }, + { + "epoch": 0.07575, + "grad_norm": 0.708796097196343, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 7575 + }, + { + "epoch": 0.07576, + "grad_norm": 0.7248673136274486, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 7576 + }, + { + "epoch": 0.07577, + "grad_norm": 0.8612577774285249, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 7577 + }, + { + "epoch": 0.07578, + "grad_norm": 1.005619129828009, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7578 + }, + { + "epoch": 0.07579, + "grad_norm": 1.0450194172356522, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7579 + }, + { + "epoch": 0.0758, + "grad_norm": 1.036801784903056, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7580 + }, + { + "epoch": 0.07581, + "grad_norm": 0.8846896117250778, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 7581 + }, + { + "epoch": 0.07582, + "grad_norm": 0.8553042305382026, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 7582 + }, + { + "epoch": 0.07583, + "grad_norm": 0.746157598396527, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 7583 + }, + { + "epoch": 0.07584, + "grad_norm": 0.703281013230497, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 7584 + }, + { + "epoch": 0.07585, + "grad_norm": 0.717776305908099, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 7585 + }, + { + "epoch": 0.07586, + "grad_norm": 0.7361220487355621, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 7586 + }, + { + "epoch": 0.07587, + "grad_norm": 0.7746201501489567, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 7587 + }, + { + "epoch": 0.07588, + "grad_norm": 0.883273441061299, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 7588 + }, + { + "epoch": 0.07589, + "grad_norm": 1.0296051718436425, + "learning_rate": 0.003, + "loss": 4.11, + "step": 7589 + }, + { + "epoch": 0.0759, + "grad_norm": 1.0366218527846427, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7590 + }, + { + "epoch": 0.07591, + "grad_norm": 0.9785797112728007, + "learning_rate": 0.003, + "loss": 4.113, + "step": 7591 + }, + { + "epoch": 0.07592, + "grad_norm": 1.0644976856872927, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 7592 + }, + { + "epoch": 0.07593, + "grad_norm": 0.8193582499715998, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7593 + }, + { + "epoch": 0.07594, + "grad_norm": 0.7786543529015693, + "learning_rate": 0.003, + "loss": 4.08, + "step": 7594 + }, + { + "epoch": 0.07595, + "grad_norm": 0.846428038730107, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 7595 + }, + { + "epoch": 0.07596, + "grad_norm": 0.8584163070813771, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 7596 + }, + { + "epoch": 0.07597, + "grad_norm": 0.9415235991639561, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 7597 + }, + { + "epoch": 0.07598, + "grad_norm": 1.0765153427852032, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 7598 + }, + { + "epoch": 0.07599, + "grad_norm": 0.804425870287582, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 7599 + }, + { + "epoch": 0.076, + "grad_norm": 0.8318708876658205, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 7600 + }, + { + "epoch": 0.07601, + "grad_norm": 1.1185425010615238, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 7601 + }, + { + "epoch": 0.07602, + "grad_norm": 1.047634075986158, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 7602 + }, + { + "epoch": 0.07603, + "grad_norm": 0.9009409871203874, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7603 + }, + { + "epoch": 0.07604, + "grad_norm": 0.8568906673551611, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 7604 + }, + { + "epoch": 0.07605, + "grad_norm": 0.7784225426074195, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 7605 + }, + { + "epoch": 0.07606, + "grad_norm": 0.839250801955735, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 7606 + }, + { + "epoch": 0.07607, + "grad_norm": 0.8782045362404554, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 7607 + }, + { + "epoch": 0.07608, + "grad_norm": 0.9200848229872217, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 7608 + }, + { + "epoch": 0.07609, + "grad_norm": 0.8082524390097369, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 7609 + }, + { + "epoch": 0.0761, + "grad_norm": 0.872081723964123, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7610 + }, + { + "epoch": 0.07611, + "grad_norm": 0.9341942398949311, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 7611 + }, + { + "epoch": 0.07612, + "grad_norm": 0.9227398398243862, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 7612 + }, + { + "epoch": 0.07613, + "grad_norm": 0.8683378115106041, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 7613 + }, + { + "epoch": 0.07614, + "grad_norm": 1.0063037001529425, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 7614 + }, + { + "epoch": 0.07615, + "grad_norm": 1.1435774924345972, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7615 + }, + { + "epoch": 0.07616, + "grad_norm": 0.8848928857504734, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7616 + }, + { + "epoch": 0.07617, + "grad_norm": 0.8941746199995625, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 7617 + }, + { + "epoch": 0.07618, + "grad_norm": 0.9902163474565946, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 7618 + }, + { + "epoch": 0.07619, + "grad_norm": 1.0531865923482646, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 7619 + }, + { + "epoch": 0.0762, + "grad_norm": 0.9255658320995923, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 7620 + }, + { + "epoch": 0.07621, + "grad_norm": 0.8649930523728451, + "learning_rate": 0.003, + "loss": 4.119, + "step": 7621 + }, + { + "epoch": 0.07622, + "grad_norm": 0.7989732615276255, + "learning_rate": 0.003, + "loss": 4.114, + "step": 7622 + }, + { + "epoch": 0.07623, + "grad_norm": 0.9110231085001621, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 7623 + }, + { + "epoch": 0.07624, + "grad_norm": 1.0834218353454426, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 7624 + }, + { + "epoch": 0.07625, + "grad_norm": 1.0325495122882569, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7625 + }, + { + "epoch": 0.07626, + "grad_norm": 0.8475068109381595, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7626 + }, + { + "epoch": 0.07627, + "grad_norm": 0.727600425780518, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7627 + }, + { + "epoch": 0.07628, + "grad_norm": 0.7632302827090037, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 7628 + }, + { + "epoch": 0.07629, + "grad_norm": 0.8788639933925416, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7629 + }, + { + "epoch": 0.0763, + "grad_norm": 1.0709031796647681, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 7630 + }, + { + "epoch": 0.07631, + "grad_norm": 1.074935960825108, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 7631 + }, + { + "epoch": 0.07632, + "grad_norm": 1.061277776789126, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7632 + }, + { + "epoch": 0.07633, + "grad_norm": 1.077834867567973, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 7633 + }, + { + "epoch": 0.07634, + "grad_norm": 0.8857955127757137, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 7634 + }, + { + "epoch": 0.07635, + "grad_norm": 0.7954993888470026, + "learning_rate": 0.003, + "loss": 4.106, + "step": 7635 + }, + { + "epoch": 0.07636, + "grad_norm": 0.7860087645669086, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 7636 + }, + { + "epoch": 0.07637, + "grad_norm": 0.8285890551300037, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 7637 + }, + { + "epoch": 0.07638, + "grad_norm": 0.8451951844334696, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 7638 + }, + { + "epoch": 0.07639, + "grad_norm": 0.8191635031988865, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7639 + }, + { + "epoch": 0.0764, + "grad_norm": 0.8579832231757687, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 7640 + }, + { + "epoch": 0.07641, + "grad_norm": 0.8068346229932589, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 7641 + }, + { + "epoch": 0.07642, + "grad_norm": 0.8465566767852095, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 7642 + }, + { + "epoch": 0.07643, + "grad_norm": 0.8119061434375141, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 7643 + }, + { + "epoch": 0.07644, + "grad_norm": 0.793320262503154, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 7644 + }, + { + "epoch": 0.07645, + "grad_norm": 0.9691323458548825, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7645 + }, + { + "epoch": 0.07646, + "grad_norm": 1.3768015813021042, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 7646 + }, + { + "epoch": 0.07647, + "grad_norm": 0.9773228446390375, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 7647 + }, + { + "epoch": 0.07648, + "grad_norm": 1.051227994942795, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 7648 + }, + { + "epoch": 0.07649, + "grad_norm": 0.8613190674836249, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 7649 + }, + { + "epoch": 0.0765, + "grad_norm": 0.8644471785420074, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7650 + }, + { + "epoch": 0.07651, + "grad_norm": 0.893229363014541, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 7651 + }, + { + "epoch": 0.07652, + "grad_norm": 0.8239815130553648, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7652 + }, + { + "epoch": 0.07653, + "grad_norm": 0.8454893514249822, + "learning_rate": 0.003, + "loss": 4.069, + "step": 7653 + }, + { + "epoch": 0.07654, + "grad_norm": 0.7622344925499536, + "learning_rate": 0.003, + "loss": 4.097, + "step": 7654 + }, + { + "epoch": 0.07655, + "grad_norm": 0.7645190916741668, + "learning_rate": 0.003, + "loss": 4.1, + "step": 7655 + }, + { + "epoch": 0.07656, + "grad_norm": 0.813005250974179, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 7656 + }, + { + "epoch": 0.07657, + "grad_norm": 0.8896111934604986, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 7657 + }, + { + "epoch": 0.07658, + "grad_norm": 0.8706430779245553, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 7658 + }, + { + "epoch": 0.07659, + "grad_norm": 0.8081972670264891, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 7659 + }, + { + "epoch": 0.0766, + "grad_norm": 0.7729443681983716, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 7660 + }, + { + "epoch": 0.07661, + "grad_norm": 0.862803077545191, + "learning_rate": 0.003, + "loss": 4.089, + "step": 7661 + }, + { + "epoch": 0.07662, + "grad_norm": 1.0006964442318478, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 7662 + }, + { + "epoch": 0.07663, + "grad_norm": 0.9936477236936112, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 7663 + }, + { + "epoch": 0.07664, + "grad_norm": 1.0569422427730744, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 7664 + }, + { + "epoch": 0.07665, + "grad_norm": 0.900621822760911, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7665 + }, + { + "epoch": 0.07666, + "grad_norm": 0.8581843257219371, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 7666 + }, + { + "epoch": 0.07667, + "grad_norm": 0.983726396544071, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 7667 + }, + { + "epoch": 0.07668, + "grad_norm": 1.1069406233194603, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 7668 + }, + { + "epoch": 0.07669, + "grad_norm": 1.0015927610296076, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7669 + }, + { + "epoch": 0.0767, + "grad_norm": 1.0573154199589114, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 7670 + }, + { + "epoch": 0.07671, + "grad_norm": 0.9825877248296772, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 7671 + }, + { + "epoch": 0.07672, + "grad_norm": 0.9631425499922999, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 7672 + }, + { + "epoch": 0.07673, + "grad_norm": 1.0326441105372977, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 7673 + }, + { + "epoch": 0.07674, + "grad_norm": 1.0631847500181333, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7674 + }, + { + "epoch": 0.07675, + "grad_norm": 1.101230295023731, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 7675 + }, + { + "epoch": 0.07676, + "grad_norm": 0.9273268109359967, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7676 + }, + { + "epoch": 0.07677, + "grad_norm": 0.8619308640763275, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 7677 + }, + { + "epoch": 0.07678, + "grad_norm": 0.9052969792740448, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 7678 + }, + { + "epoch": 0.07679, + "grad_norm": 0.9378634519819189, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 7679 + }, + { + "epoch": 0.0768, + "grad_norm": 0.8521164764973965, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 7680 + }, + { + "epoch": 0.07681, + "grad_norm": 0.8595477564055118, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7681 + }, + { + "epoch": 0.07682, + "grad_norm": 0.7754677865078394, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 7682 + }, + { + "epoch": 0.07683, + "grad_norm": 0.6840059827859887, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 7683 + }, + { + "epoch": 0.07684, + "grad_norm": 0.6575974583729506, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 7684 + }, + { + "epoch": 0.07685, + "grad_norm": 0.7235799579926786, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 7685 + }, + { + "epoch": 0.07686, + "grad_norm": 0.8377398222220133, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 7686 + }, + { + "epoch": 0.07687, + "grad_norm": 0.9082625532369022, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 7687 + }, + { + "epoch": 0.07688, + "grad_norm": 1.2331010320496771, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 7688 + }, + { + "epoch": 0.07689, + "grad_norm": 0.8212537354275941, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 7689 + }, + { + "epoch": 0.0769, + "grad_norm": 0.6665825438931752, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7690 + }, + { + "epoch": 0.07691, + "grad_norm": 0.7566566405069987, + "learning_rate": 0.003, + "loss": 4.103, + "step": 7691 + }, + { + "epoch": 0.07692, + "grad_norm": 0.8271311664477755, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 7692 + }, + { + "epoch": 0.07693, + "grad_norm": 0.8707736692236765, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 7693 + }, + { + "epoch": 0.07694, + "grad_norm": 0.8808612418292081, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 7694 + }, + { + "epoch": 0.07695, + "grad_norm": 1.1008954469557017, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7695 + }, + { + "epoch": 0.07696, + "grad_norm": 1.0606187645237288, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7696 + }, + { + "epoch": 0.07697, + "grad_norm": 1.0672701176658974, + "learning_rate": 0.003, + "loss": 4.138, + "step": 7697 + }, + { + "epoch": 0.07698, + "grad_norm": 0.881431735642276, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 7698 + }, + { + "epoch": 0.07699, + "grad_norm": 0.9075150605445227, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 7699 + }, + { + "epoch": 0.077, + "grad_norm": 0.9026931164648633, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 7700 + }, + { + "epoch": 0.07701, + "grad_norm": 1.0261275216945884, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 7701 + }, + { + "epoch": 0.07702, + "grad_norm": 1.0674364416061641, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 7702 + }, + { + "epoch": 0.07703, + "grad_norm": 0.8464641737302251, + "learning_rate": 0.003, + "loss": 4.109, + "step": 7703 + }, + { + "epoch": 0.07704, + "grad_norm": 0.7763702415342664, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 7704 + }, + { + "epoch": 0.07705, + "grad_norm": 0.7689660168469808, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 7705 + }, + { + "epoch": 0.07706, + "grad_norm": 0.8664018808766383, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7706 + }, + { + "epoch": 0.07707, + "grad_norm": 1.0667476859474658, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7707 + }, + { + "epoch": 0.07708, + "grad_norm": 1.2058062363442252, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 7708 + }, + { + "epoch": 0.07709, + "grad_norm": 1.0492356527458138, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 7709 + }, + { + "epoch": 0.0771, + "grad_norm": 0.9750451640390347, + "learning_rate": 0.003, + "loss": 4.086, + "step": 7710 + }, + { + "epoch": 0.07711, + "grad_norm": 0.8336813405535649, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 7711 + }, + { + "epoch": 0.07712, + "grad_norm": 0.8155486398893903, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 7712 + }, + { + "epoch": 0.07713, + "grad_norm": 0.8688818823181422, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 7713 + }, + { + "epoch": 0.07714, + "grad_norm": 0.9332384720820431, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 7714 + }, + { + "epoch": 0.07715, + "grad_norm": 1.1579449218264937, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7715 + }, + { + "epoch": 0.07716, + "grad_norm": 1.0775937449235249, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 7716 + }, + { + "epoch": 0.07717, + "grad_norm": 0.8872808956615507, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 7717 + }, + { + "epoch": 0.07718, + "grad_norm": 0.8502430408961995, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 7718 + }, + { + "epoch": 0.07719, + "grad_norm": 0.8840941630249397, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7719 + }, + { + "epoch": 0.0772, + "grad_norm": 0.9698792840195539, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 7720 + }, + { + "epoch": 0.07721, + "grad_norm": 0.9769844568177094, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7721 + }, + { + "epoch": 0.07722, + "grad_norm": 1.0502547313312145, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 7722 + }, + { + "epoch": 0.07723, + "grad_norm": 1.0515641978538086, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 7723 + }, + { + "epoch": 0.07724, + "grad_norm": 0.8196988574303467, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 7724 + }, + { + "epoch": 0.07725, + "grad_norm": 0.7914284658585109, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7725 + }, + { + "epoch": 0.07726, + "grad_norm": 0.8789170064349751, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 7726 + }, + { + "epoch": 0.07727, + "grad_norm": 1.010851883887785, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 7727 + }, + { + "epoch": 0.07728, + "grad_norm": 1.0707742849762627, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 7728 + }, + { + "epoch": 0.07729, + "grad_norm": 1.105072577657322, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 7729 + }, + { + "epoch": 0.0773, + "grad_norm": 0.8048148723787254, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 7730 + }, + { + "epoch": 0.07731, + "grad_norm": 0.804004108549886, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7731 + }, + { + "epoch": 0.07732, + "grad_norm": 0.9865333094479842, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 7732 + }, + { + "epoch": 0.07733, + "grad_norm": 1.187532994008932, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 7733 + }, + { + "epoch": 0.07734, + "grad_norm": 0.9116123033967836, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 7734 + }, + { + "epoch": 0.07735, + "grad_norm": 0.9647065687240539, + "learning_rate": 0.003, + "loss": 4.097, + "step": 7735 + }, + { + "epoch": 0.07736, + "grad_norm": 0.8163482733718864, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 7736 + }, + { + "epoch": 0.07737, + "grad_norm": 0.879564638557422, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 7737 + }, + { + "epoch": 0.07738, + "grad_norm": 0.922084879137366, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 7738 + }, + { + "epoch": 0.07739, + "grad_norm": 1.0445880792762618, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 7739 + }, + { + "epoch": 0.0774, + "grad_norm": 1.0375087691471465, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7740 + }, + { + "epoch": 0.07741, + "grad_norm": 0.9145323840861375, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 7741 + }, + { + "epoch": 0.07742, + "grad_norm": 0.8266492697370303, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 7742 + }, + { + "epoch": 0.07743, + "grad_norm": 0.9110104456277404, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 7743 + }, + { + "epoch": 0.07744, + "grad_norm": 1.0371705555718609, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7744 + }, + { + "epoch": 0.07745, + "grad_norm": 0.969698982306619, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 7745 + }, + { + "epoch": 0.07746, + "grad_norm": 1.0255113318693947, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 7746 + }, + { + "epoch": 0.07747, + "grad_norm": 0.8882628795779053, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 7747 + }, + { + "epoch": 0.07748, + "grad_norm": 0.7973386036195474, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 7748 + }, + { + "epoch": 0.07749, + "grad_norm": 0.8652226027466258, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 7749 + }, + { + "epoch": 0.0775, + "grad_norm": 0.870376274432496, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 7750 + }, + { + "epoch": 0.07751, + "grad_norm": 0.9893903819503679, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 7751 + }, + { + "epoch": 0.07752, + "grad_norm": 1.3482906255390648, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 7752 + }, + { + "epoch": 0.07753, + "grad_norm": 0.8068929176849469, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 7753 + }, + { + "epoch": 0.07754, + "grad_norm": 0.8014680097244021, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 7754 + }, + { + "epoch": 0.07755, + "grad_norm": 0.8754909218158006, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 7755 + }, + { + "epoch": 0.07756, + "grad_norm": 0.9676792258616127, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 7756 + }, + { + "epoch": 0.07757, + "grad_norm": 1.0045119040336394, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 7757 + }, + { + "epoch": 0.07758, + "grad_norm": 1.111589755893894, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 7758 + }, + { + "epoch": 0.07759, + "grad_norm": 0.8968827266122236, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7759 + }, + { + "epoch": 0.0776, + "grad_norm": 0.9286993706118074, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 7760 + }, + { + "epoch": 0.07761, + "grad_norm": 1.048945024253054, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 7761 + }, + { + "epoch": 0.07762, + "grad_norm": 0.9319693389175623, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 7762 + }, + { + "epoch": 0.07763, + "grad_norm": 0.813291989320193, + "learning_rate": 0.003, + "loss": 4.115, + "step": 7763 + }, + { + "epoch": 0.07764, + "grad_norm": 0.784853226242816, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 7764 + }, + { + "epoch": 0.07765, + "grad_norm": 0.7032656157830414, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 7765 + }, + { + "epoch": 0.07766, + "grad_norm": 0.6646150699887404, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 7766 + }, + { + "epoch": 0.07767, + "grad_norm": 0.7353992656309892, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 7767 + }, + { + "epoch": 0.07768, + "grad_norm": 0.880810970396917, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 7768 + }, + { + "epoch": 0.07769, + "grad_norm": 0.9578602477971352, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 7769 + }, + { + "epoch": 0.0777, + "grad_norm": 1.1156377007546823, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 7770 + }, + { + "epoch": 0.07771, + "grad_norm": 0.8301209539531904, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 7771 + }, + { + "epoch": 0.07772, + "grad_norm": 0.7560818213800261, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 7772 + }, + { + "epoch": 0.07773, + "grad_norm": 0.8404921927294836, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7773 + }, + { + "epoch": 0.07774, + "grad_norm": 0.8782760933360856, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 7774 + }, + { + "epoch": 0.07775, + "grad_norm": 0.7808631321910917, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7775 + }, + { + "epoch": 0.07776, + "grad_norm": 0.7541917305536269, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 7776 + }, + { + "epoch": 0.07777, + "grad_norm": 0.8093912075646889, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 7777 + }, + { + "epoch": 0.07778, + "grad_norm": 0.9732383115498489, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 7778 + }, + { + "epoch": 0.07779, + "grad_norm": 0.9921828686435009, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 7779 + }, + { + "epoch": 0.0778, + "grad_norm": 0.9497774890620421, + "learning_rate": 0.003, + "loss": 4.081, + "step": 7780 + }, + { + "epoch": 0.07781, + "grad_norm": 0.9033005759467783, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 7781 + }, + { + "epoch": 0.07782, + "grad_norm": 1.0370862427957597, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 7782 + }, + { + "epoch": 0.07783, + "grad_norm": 1.30997993322532, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 7783 + }, + { + "epoch": 0.07784, + "grad_norm": 0.9418736206229247, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 7784 + }, + { + "epoch": 0.07785, + "grad_norm": 1.0195051094835668, + "learning_rate": 0.003, + "loss": 4.133, + "step": 7785 + }, + { + "epoch": 0.07786, + "grad_norm": 1.0159016886924566, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 7786 + }, + { + "epoch": 0.07787, + "grad_norm": 1.0675515986885633, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 7787 + }, + { + "epoch": 0.07788, + "grad_norm": 1.2557775330748935, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 7788 + }, + { + "epoch": 0.07789, + "grad_norm": 1.081778503928991, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7789 + }, + { + "epoch": 0.0779, + "grad_norm": 0.8537282011122178, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 7790 + }, + { + "epoch": 0.07791, + "grad_norm": 0.7874541429569311, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 7791 + }, + { + "epoch": 0.07792, + "grad_norm": 0.7999390691767071, + "learning_rate": 0.003, + "loss": 4.077, + "step": 7792 + }, + { + "epoch": 0.07793, + "grad_norm": 0.7277192795700171, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 7793 + }, + { + "epoch": 0.07794, + "grad_norm": 0.9514831979826355, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 7794 + }, + { + "epoch": 0.07795, + "grad_norm": 1.012601597640212, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 7795 + }, + { + "epoch": 0.07796, + "grad_norm": 0.9201926222912106, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 7796 + }, + { + "epoch": 0.07797, + "grad_norm": 0.7756213744843218, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 7797 + }, + { + "epoch": 0.07798, + "grad_norm": 0.8500195316443565, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 7798 + }, + { + "epoch": 0.07799, + "grad_norm": 0.7663934706772912, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7799 + }, + { + "epoch": 0.078, + "grad_norm": 0.6950472452800454, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 7800 + }, + { + "epoch": 0.07801, + "grad_norm": 0.7781267723317771, + "learning_rate": 0.003, + "loss": 4.093, + "step": 7801 + }, + { + "epoch": 0.07802, + "grad_norm": 1.0163302306750792, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 7802 + }, + { + "epoch": 0.07803, + "grad_norm": 1.280595782619173, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 7803 + }, + { + "epoch": 0.07804, + "grad_norm": 0.8778051384904699, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 7804 + }, + { + "epoch": 0.07805, + "grad_norm": 0.899483151510845, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 7805 + }, + { + "epoch": 0.07806, + "grad_norm": 0.9396336994589124, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 7806 + }, + { + "epoch": 0.07807, + "grad_norm": 0.8665558808389224, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 7807 + }, + { + "epoch": 0.07808, + "grad_norm": 0.7981525957939095, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 7808 + }, + { + "epoch": 0.07809, + "grad_norm": 0.8751564912698292, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 7809 + }, + { + "epoch": 0.0781, + "grad_norm": 0.8166553655034294, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 7810 + }, + { + "epoch": 0.07811, + "grad_norm": 0.767979955899306, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 7811 + }, + { + "epoch": 0.07812, + "grad_norm": 0.7617431224556912, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 7812 + }, + { + "epoch": 0.07813, + "grad_norm": 0.7967942604337378, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 7813 + }, + { + "epoch": 0.07814, + "grad_norm": 0.9942681164347447, + "learning_rate": 0.003, + "loss": 4.141, + "step": 7814 + }, + { + "epoch": 0.07815, + "grad_norm": 1.1892095427285347, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 7815 + }, + { + "epoch": 0.07816, + "grad_norm": 0.876659865594819, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 7816 + }, + { + "epoch": 0.07817, + "grad_norm": 0.9631730515288163, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 7817 + }, + { + "epoch": 0.07818, + "grad_norm": 1.0715009376405462, + "learning_rate": 0.003, + "loss": 4.094, + "step": 7818 + }, + { + "epoch": 0.07819, + "grad_norm": 1.0520680251193615, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 7819 + }, + { + "epoch": 0.0782, + "grad_norm": 0.9439418798912824, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 7820 + }, + { + "epoch": 0.07821, + "grad_norm": 0.9069113626074072, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 7821 + }, + { + "epoch": 0.07822, + "grad_norm": 1.0012938180643256, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 7822 + }, + { + "epoch": 0.07823, + "grad_norm": 1.0048698890522239, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7823 + }, + { + "epoch": 0.07824, + "grad_norm": 0.9561751321839956, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 7824 + }, + { + "epoch": 0.07825, + "grad_norm": 0.9957350057887719, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 7825 + }, + { + "epoch": 0.07826, + "grad_norm": 1.079540426834375, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 7826 + }, + { + "epoch": 0.07827, + "grad_norm": 0.9385499777483688, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 7827 + }, + { + "epoch": 0.07828, + "grad_norm": 1.0935244049947113, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 7828 + }, + { + "epoch": 0.07829, + "grad_norm": 0.9621888212658791, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7829 + }, + { + "epoch": 0.0783, + "grad_norm": 1.0488809493534366, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 7830 + }, + { + "epoch": 0.07831, + "grad_norm": 1.1546610786818092, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 7831 + }, + { + "epoch": 0.07832, + "grad_norm": 0.9707818700073304, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7832 + }, + { + "epoch": 0.07833, + "grad_norm": 1.0395346936823906, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 7833 + }, + { + "epoch": 0.07834, + "grad_norm": 1.0441939813531944, + "learning_rate": 0.003, + "loss": 4.1592, + "step": 7834 + }, + { + "epoch": 0.07835, + "grad_norm": 0.9142633227917905, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 7835 + }, + { + "epoch": 0.07836, + "grad_norm": 0.7722831477807496, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 7836 + }, + { + "epoch": 0.07837, + "grad_norm": 0.7700267869758111, + "learning_rate": 0.003, + "loss": 4.102, + "step": 7837 + }, + { + "epoch": 0.07838, + "grad_norm": 0.8265142746844435, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7838 + }, + { + "epoch": 0.07839, + "grad_norm": 0.8648999060225745, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 7839 + }, + { + "epoch": 0.0784, + "grad_norm": 0.8130848977918211, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 7840 + }, + { + "epoch": 0.07841, + "grad_norm": 0.834862433110529, + "learning_rate": 0.003, + "loss": 4.086, + "step": 7841 + }, + { + "epoch": 0.07842, + "grad_norm": 0.721640102949708, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 7842 + }, + { + "epoch": 0.07843, + "grad_norm": 0.6116456054704642, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 7843 + }, + { + "epoch": 0.07844, + "grad_norm": 0.6234499687641849, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 7844 + }, + { + "epoch": 0.07845, + "grad_norm": 0.638928055891956, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 7845 + }, + { + "epoch": 0.07846, + "grad_norm": 0.7063602100511459, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 7846 + }, + { + "epoch": 0.07847, + "grad_norm": 0.7799338866735226, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7847 + }, + { + "epoch": 0.07848, + "grad_norm": 0.898204534805398, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 7848 + }, + { + "epoch": 0.07849, + "grad_norm": 1.012005052883131, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 7849 + }, + { + "epoch": 0.0785, + "grad_norm": 1.0529431619025262, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7850 + }, + { + "epoch": 0.07851, + "grad_norm": 0.9650822583370998, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 7851 + }, + { + "epoch": 0.07852, + "grad_norm": 1.0400516186857025, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 7852 + }, + { + "epoch": 0.07853, + "grad_norm": 0.9710595472244659, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 7853 + }, + { + "epoch": 0.07854, + "grad_norm": 1.0238516024921551, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 7854 + }, + { + "epoch": 0.07855, + "grad_norm": 1.09812820718059, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 7855 + }, + { + "epoch": 0.07856, + "grad_norm": 0.9164501284041578, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 7856 + }, + { + "epoch": 0.07857, + "grad_norm": 1.0913391162498283, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 7857 + }, + { + "epoch": 0.07858, + "grad_norm": 1.0601655274183024, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 7858 + }, + { + "epoch": 0.07859, + "grad_norm": 0.9500195926062109, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 7859 + }, + { + "epoch": 0.0786, + "grad_norm": 0.9248820458548533, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 7860 + }, + { + "epoch": 0.07861, + "grad_norm": 1.0564365177503385, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7861 + }, + { + "epoch": 0.07862, + "grad_norm": 1.0399061246279695, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 7862 + }, + { + "epoch": 0.07863, + "grad_norm": 0.9697750588673806, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 7863 + }, + { + "epoch": 0.07864, + "grad_norm": 1.0453089078095656, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 7864 + }, + { + "epoch": 0.07865, + "grad_norm": 1.009583329358567, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 7865 + }, + { + "epoch": 0.07866, + "grad_norm": 1.2318259015249517, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 7866 + }, + { + "epoch": 0.07867, + "grad_norm": 0.8590130093094929, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7867 + }, + { + "epoch": 0.07868, + "grad_norm": 0.9954328336568087, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 7868 + }, + { + "epoch": 0.07869, + "grad_norm": 1.1347719980582816, + "learning_rate": 0.003, + "loss": 4.111, + "step": 7869 + }, + { + "epoch": 0.0787, + "grad_norm": 0.9909944155141596, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 7870 + }, + { + "epoch": 0.07871, + "grad_norm": 1.027533472785972, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 7871 + }, + { + "epoch": 0.07872, + "grad_norm": 0.9509538791389344, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 7872 + }, + { + "epoch": 0.07873, + "grad_norm": 0.9743948500738193, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 7873 + }, + { + "epoch": 0.07874, + "grad_norm": 1.076018237392653, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 7874 + }, + { + "epoch": 0.07875, + "grad_norm": 0.991291863842112, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 7875 + }, + { + "epoch": 0.07876, + "grad_norm": 1.111246410563993, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 7876 + }, + { + "epoch": 0.07877, + "grad_norm": 0.9900378915913184, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 7877 + }, + { + "epoch": 0.07878, + "grad_norm": 0.8784205316420091, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 7878 + }, + { + "epoch": 0.07879, + "grad_norm": 0.9417046644892657, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 7879 + }, + { + "epoch": 0.0788, + "grad_norm": 1.037927767240949, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7880 + }, + { + "epoch": 0.07881, + "grad_norm": 1.2229688137388648, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7881 + }, + { + "epoch": 0.07882, + "grad_norm": 0.8555608429927865, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 7882 + }, + { + "epoch": 0.07883, + "grad_norm": 0.8611709341175292, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7883 + }, + { + "epoch": 0.07884, + "grad_norm": 0.77157174436383, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 7884 + }, + { + "epoch": 0.07885, + "grad_norm": 0.8455320691314467, + "learning_rate": 0.003, + "loss": 4.091, + "step": 7885 + }, + { + "epoch": 0.07886, + "grad_norm": 0.9014648447792224, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 7886 + }, + { + "epoch": 0.07887, + "grad_norm": 0.9366329641994356, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 7887 + }, + { + "epoch": 0.07888, + "grad_norm": 0.8053760771673537, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 7888 + }, + { + "epoch": 0.07889, + "grad_norm": 0.7828342950771663, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 7889 + }, + { + "epoch": 0.0789, + "grad_norm": 0.6756189697361212, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 7890 + }, + { + "epoch": 0.07891, + "grad_norm": 0.7330290536805933, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 7891 + }, + { + "epoch": 0.07892, + "grad_norm": 0.7188111647939824, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 7892 + }, + { + "epoch": 0.07893, + "grad_norm": 0.7586433320705798, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 7893 + }, + { + "epoch": 0.07894, + "grad_norm": 0.9966525848375563, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 7894 + }, + { + "epoch": 0.07895, + "grad_norm": 1.3442975582134642, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7895 + }, + { + "epoch": 0.07896, + "grad_norm": 0.729365353256303, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 7896 + }, + { + "epoch": 0.07897, + "grad_norm": 0.6629638171890595, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 7897 + }, + { + "epoch": 0.07898, + "grad_norm": 0.781346999366163, + "learning_rate": 0.003, + "loss": 4.094, + "step": 7898 + }, + { + "epoch": 0.07899, + "grad_norm": 0.793772205969321, + "learning_rate": 0.003, + "loss": 4.072, + "step": 7899 + }, + { + "epoch": 0.079, + "grad_norm": 0.9389063494976632, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7900 + }, + { + "epoch": 0.07901, + "grad_norm": 1.1343911954848018, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 7901 + }, + { + "epoch": 0.07902, + "grad_norm": 0.872291018671974, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 7902 + }, + { + "epoch": 0.07903, + "grad_norm": 0.8012280475594034, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 7903 + }, + { + "epoch": 0.07904, + "grad_norm": 0.8332934673059036, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 7904 + }, + { + "epoch": 0.07905, + "grad_norm": 0.9334487509044569, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7905 + }, + { + "epoch": 0.07906, + "grad_norm": 0.9873155951012756, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 7906 + }, + { + "epoch": 0.07907, + "grad_norm": 0.9584124647739157, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 7907 + }, + { + "epoch": 0.07908, + "grad_norm": 0.883721552574731, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 7908 + }, + { + "epoch": 0.07909, + "grad_norm": 0.870325820513853, + "learning_rate": 0.003, + "loss": 4.107, + "step": 7909 + }, + { + "epoch": 0.0791, + "grad_norm": 0.850686094030066, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 7910 + }, + { + "epoch": 0.07911, + "grad_norm": 0.9894998327802038, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 7911 + }, + { + "epoch": 0.07912, + "grad_norm": 1.0699204166192853, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 7912 + }, + { + "epoch": 0.07913, + "grad_norm": 1.1014338916357802, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 7913 + }, + { + "epoch": 0.07914, + "grad_norm": 0.8829338047891474, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 7914 + }, + { + "epoch": 0.07915, + "grad_norm": 0.8054943769351637, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 7915 + }, + { + "epoch": 0.07916, + "grad_norm": 0.8730309155139634, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 7916 + }, + { + "epoch": 0.07917, + "grad_norm": 0.8263944569948347, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7917 + }, + { + "epoch": 0.07918, + "grad_norm": 0.8908263616181593, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7918 + }, + { + "epoch": 0.07919, + "grad_norm": 1.199853472611651, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 7919 + }, + { + "epoch": 0.0792, + "grad_norm": 1.0475806506501837, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 7920 + }, + { + "epoch": 0.07921, + "grad_norm": 1.0110599460498464, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 7921 + }, + { + "epoch": 0.07922, + "grad_norm": 0.9433898207574585, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 7922 + }, + { + "epoch": 0.07923, + "grad_norm": 0.8348723789778444, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 7923 + }, + { + "epoch": 0.07924, + "grad_norm": 0.833140296619527, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 7924 + }, + { + "epoch": 0.07925, + "grad_norm": 0.9794878682061134, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 7925 + }, + { + "epoch": 0.07926, + "grad_norm": 1.2612535875768331, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 7926 + }, + { + "epoch": 0.07927, + "grad_norm": 0.7405583869267955, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7927 + }, + { + "epoch": 0.07928, + "grad_norm": 0.651164411585338, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 7928 + }, + { + "epoch": 0.07929, + "grad_norm": 0.7197387026951962, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 7929 + }, + { + "epoch": 0.0793, + "grad_norm": 0.8263085239318726, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7930 + }, + { + "epoch": 0.07931, + "grad_norm": 0.834126319233174, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 7931 + }, + { + "epoch": 0.07932, + "grad_norm": 0.7483212923735366, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 7932 + }, + { + "epoch": 0.07933, + "grad_norm": 0.8006850146197332, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 7933 + }, + { + "epoch": 0.07934, + "grad_norm": 0.8874417895778808, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 7934 + }, + { + "epoch": 0.07935, + "grad_norm": 0.9335481424769582, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 7935 + }, + { + "epoch": 0.07936, + "grad_norm": 0.9488902483540785, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7936 + }, + { + "epoch": 0.07937, + "grad_norm": 0.9210487807966519, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7937 + }, + { + "epoch": 0.07938, + "grad_norm": 0.8731093404086347, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 7938 + }, + { + "epoch": 0.07939, + "grad_norm": 0.8216594478774197, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 7939 + }, + { + "epoch": 0.0794, + "grad_norm": 0.7735027055172842, + "learning_rate": 0.003, + "loss": 4.08, + "step": 7940 + }, + { + "epoch": 0.07941, + "grad_norm": 0.9167989593985493, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7941 + }, + { + "epoch": 0.07942, + "grad_norm": 1.0793942883220813, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 7942 + }, + { + "epoch": 0.07943, + "grad_norm": 1.2258293772455378, + "learning_rate": 0.003, + "loss": 4.085, + "step": 7943 + }, + { + "epoch": 0.07944, + "grad_norm": 0.8299843295889483, + "learning_rate": 0.003, + "loss": 4.122, + "step": 7944 + }, + { + "epoch": 0.07945, + "grad_norm": 0.8715797526894794, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 7945 + }, + { + "epoch": 0.07946, + "grad_norm": 1.0070541220348046, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 7946 + }, + { + "epoch": 0.07947, + "grad_norm": 1.0559591654746912, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 7947 + }, + { + "epoch": 0.07948, + "grad_norm": 1.1246304232141864, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 7948 + }, + { + "epoch": 0.07949, + "grad_norm": 1.035646997049254, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 7949 + }, + { + "epoch": 0.0795, + "grad_norm": 1.1462152538006722, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 7950 + }, + { + "epoch": 0.07951, + "grad_norm": 1.045473251959421, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 7951 + }, + { + "epoch": 0.07952, + "grad_norm": 0.9603579292671465, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 7952 + }, + { + "epoch": 0.07953, + "grad_norm": 0.9846533672672768, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7953 + }, + { + "epoch": 0.07954, + "grad_norm": 1.0311061580889707, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 7954 + }, + { + "epoch": 0.07955, + "grad_norm": 1.1129291561662322, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 7955 + }, + { + "epoch": 0.07956, + "grad_norm": 1.3157210126198073, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 7956 + }, + { + "epoch": 0.07957, + "grad_norm": 0.9656061575557505, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 7957 + }, + { + "epoch": 0.07958, + "grad_norm": 1.0363457502099291, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 7958 + }, + { + "epoch": 0.07959, + "grad_norm": 0.9519287205545679, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 7959 + }, + { + "epoch": 0.0796, + "grad_norm": 0.8874892938612973, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 7960 + }, + { + "epoch": 0.07961, + "grad_norm": 0.8499204692701227, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 7961 + }, + { + "epoch": 0.07962, + "grad_norm": 0.7782187703657507, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 7962 + }, + { + "epoch": 0.07963, + "grad_norm": 0.8258839990795058, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 7963 + }, + { + "epoch": 0.07964, + "grad_norm": 0.8106755098862167, + "learning_rate": 0.003, + "loss": 4.092, + "step": 7964 + }, + { + "epoch": 0.07965, + "grad_norm": 0.8950386164564889, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7965 + }, + { + "epoch": 0.07966, + "grad_norm": 0.9664103829230193, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 7966 + }, + { + "epoch": 0.07967, + "grad_norm": 1.0360852176776754, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 7967 + }, + { + "epoch": 0.07968, + "grad_norm": 0.9475801661247285, + "learning_rate": 0.003, + "loss": 4.078, + "step": 7968 + }, + { + "epoch": 0.07969, + "grad_norm": 0.9150409614473582, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 7969 + }, + { + "epoch": 0.0797, + "grad_norm": 0.9289409320596667, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 7970 + }, + { + "epoch": 0.07971, + "grad_norm": 0.910519318449528, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7971 + }, + { + "epoch": 0.07972, + "grad_norm": 1.1071345490604145, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 7972 + }, + { + "epoch": 0.07973, + "grad_norm": 0.9397068669706325, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 7973 + }, + { + "epoch": 0.07974, + "grad_norm": 0.9921359001502011, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 7974 + }, + { + "epoch": 0.07975, + "grad_norm": 1.001495620220728, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 7975 + }, + { + "epoch": 0.07976, + "grad_norm": 0.803218925866123, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 7976 + }, + { + "epoch": 0.07977, + "grad_norm": 0.7357551979506947, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 7977 + }, + { + "epoch": 0.07978, + "grad_norm": 0.7441367320460851, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 7978 + }, + { + "epoch": 0.07979, + "grad_norm": 0.7100854694405272, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 7979 + }, + { + "epoch": 0.0798, + "grad_norm": 0.6890360258648266, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 7980 + }, + { + "epoch": 0.07981, + "grad_norm": 0.7537645173886749, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7981 + }, + { + "epoch": 0.07982, + "grad_norm": 0.8812941996862754, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 7982 + }, + { + "epoch": 0.07983, + "grad_norm": 0.973402696610273, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 7983 + }, + { + "epoch": 0.07984, + "grad_norm": 1.1594045747733872, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 7984 + }, + { + "epoch": 0.07985, + "grad_norm": 0.9192740280114771, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7985 + }, + { + "epoch": 0.07986, + "grad_norm": 0.9228179720848998, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 7986 + }, + { + "epoch": 0.07987, + "grad_norm": 1.003241676572412, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 7987 + }, + { + "epoch": 0.07988, + "grad_norm": 1.103347656691755, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 7988 + }, + { + "epoch": 0.07989, + "grad_norm": 0.9947380880010859, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 7989 + }, + { + "epoch": 0.0799, + "grad_norm": 1.0074317800416395, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 7990 + }, + { + "epoch": 0.07991, + "grad_norm": 1.0652610782678547, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 7991 + }, + { + "epoch": 0.07992, + "grad_norm": 0.9264232521401602, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 7992 + }, + { + "epoch": 0.07993, + "grad_norm": 0.9708215893637503, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 7993 + }, + { + "epoch": 0.07994, + "grad_norm": 0.9185573371093531, + "learning_rate": 0.003, + "loss": 4.108, + "step": 7994 + }, + { + "epoch": 0.07995, + "grad_norm": 1.058085876800838, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 7995 + }, + { + "epoch": 0.07996, + "grad_norm": 1.085515813637621, + "learning_rate": 0.003, + "loss": 4.104, + "step": 7996 + }, + { + "epoch": 0.07997, + "grad_norm": 1.0465886940521179, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 7997 + }, + { + "epoch": 0.07998, + "grad_norm": 0.8881454041255957, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 7998 + }, + { + "epoch": 0.07999, + "grad_norm": 0.8240517740099984, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 7999 + }, + { + "epoch": 0.08, + "grad_norm": 0.945864278006696, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 8000 + }, + { + "epoch": 0.08001, + "grad_norm": 1.0212774414123797, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8001 + }, + { + "epoch": 0.08002, + "grad_norm": 1.1204835979046452, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 8002 + }, + { + "epoch": 0.08003, + "grad_norm": 0.9520420391468847, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 8003 + }, + { + "epoch": 0.08004, + "grad_norm": 0.9448651646661593, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 8004 + }, + { + "epoch": 0.08005, + "grad_norm": 0.9170614863205321, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 8005 + }, + { + "epoch": 0.08006, + "grad_norm": 0.8779333843730515, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 8006 + }, + { + "epoch": 0.08007, + "grad_norm": 0.9266707111712874, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 8007 + }, + { + "epoch": 0.08008, + "grad_norm": 0.9901624722228317, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 8008 + }, + { + "epoch": 0.08009, + "grad_norm": 0.9651216411589141, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8009 + }, + { + "epoch": 0.0801, + "grad_norm": 1.0042760352203992, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8010 + }, + { + "epoch": 0.08011, + "grad_norm": 1.1768489281812973, + "learning_rate": 0.003, + "loss": 4.11, + "step": 8011 + }, + { + "epoch": 0.08012, + "grad_norm": 0.969355311788046, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 8012 + }, + { + "epoch": 0.08013, + "grad_norm": 1.1120606338609458, + "learning_rate": 0.003, + "loss": 4.109, + "step": 8013 + }, + { + "epoch": 0.08014, + "grad_norm": 0.9782866986068071, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8014 + }, + { + "epoch": 0.08015, + "grad_norm": 1.0758211992496314, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 8015 + }, + { + "epoch": 0.08016, + "grad_norm": 0.880271638602275, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 8016 + }, + { + "epoch": 0.08017, + "grad_norm": 0.7502263255876486, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 8017 + }, + { + "epoch": 0.08018, + "grad_norm": 0.7734348211878546, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8018 + }, + { + "epoch": 0.08019, + "grad_norm": 0.7684078035607653, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 8019 + }, + { + "epoch": 0.0802, + "grad_norm": 0.7708407552479869, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 8020 + }, + { + "epoch": 0.08021, + "grad_norm": 0.848673360124088, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8021 + }, + { + "epoch": 0.08022, + "grad_norm": 0.8845796435472464, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8022 + }, + { + "epoch": 0.08023, + "grad_norm": 1.016380065852941, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8023 + }, + { + "epoch": 0.08024, + "grad_norm": 1.0460153574644655, + "learning_rate": 0.003, + "loss": 4.107, + "step": 8024 + }, + { + "epoch": 0.08025, + "grad_norm": 1.0453001878006873, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 8025 + }, + { + "epoch": 0.08026, + "grad_norm": 0.9965656351708417, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8026 + }, + { + "epoch": 0.08027, + "grad_norm": 1.0457492474459331, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 8027 + }, + { + "epoch": 0.08028, + "grad_norm": 0.8875637028878758, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 8028 + }, + { + "epoch": 0.08029, + "grad_norm": 0.8862575547070733, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 8029 + }, + { + "epoch": 0.0803, + "grad_norm": 0.8727888553607356, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8030 + }, + { + "epoch": 0.08031, + "grad_norm": 0.91179863708708, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 8031 + }, + { + "epoch": 0.08032, + "grad_norm": 0.8917368648696833, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8032 + }, + { + "epoch": 0.08033, + "grad_norm": 0.9249513133662625, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 8033 + }, + { + "epoch": 0.08034, + "grad_norm": 1.056594929957501, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8034 + }, + { + "epoch": 0.08035, + "grad_norm": 1.2140600766666199, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 8035 + }, + { + "epoch": 0.08036, + "grad_norm": 0.9604814329204158, + "learning_rate": 0.003, + "loss": 4.14, + "step": 8036 + }, + { + "epoch": 0.08037, + "grad_norm": 0.9718783518855493, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8037 + }, + { + "epoch": 0.08038, + "grad_norm": 1.0788267825486346, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8038 + }, + { + "epoch": 0.08039, + "grad_norm": 1.0426960236669471, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8039 + }, + { + "epoch": 0.0804, + "grad_norm": 0.8990767990638859, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 8040 + }, + { + "epoch": 0.08041, + "grad_norm": 0.8206146604776955, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 8041 + }, + { + "epoch": 0.08042, + "grad_norm": 0.9315698270340985, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 8042 + }, + { + "epoch": 0.08043, + "grad_norm": 0.9892501547865945, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 8043 + }, + { + "epoch": 0.08044, + "grad_norm": 1.0447417437517303, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8044 + }, + { + "epoch": 0.08045, + "grad_norm": 1.0149509159686925, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8045 + }, + { + "epoch": 0.08046, + "grad_norm": 0.9358159055498079, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8046 + }, + { + "epoch": 0.08047, + "grad_norm": 0.8933533196863428, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8047 + }, + { + "epoch": 0.08048, + "grad_norm": 0.7664757213485384, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 8048 + }, + { + "epoch": 0.08049, + "grad_norm": 0.750556226478064, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8049 + }, + { + "epoch": 0.0805, + "grad_norm": 0.8246113468903432, + "learning_rate": 0.003, + "loss": 4.112, + "step": 8050 + }, + { + "epoch": 0.08051, + "grad_norm": 0.7466306423459229, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 8051 + }, + { + "epoch": 0.08052, + "grad_norm": 0.7461882411581569, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 8052 + }, + { + "epoch": 0.08053, + "grad_norm": 0.7594324683486439, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 8053 + }, + { + "epoch": 0.08054, + "grad_norm": 0.7783815553299155, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 8054 + }, + { + "epoch": 0.08055, + "grad_norm": 0.8662143832724243, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 8055 + }, + { + "epoch": 0.08056, + "grad_norm": 1.0448541115220396, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 8056 + }, + { + "epoch": 0.08057, + "grad_norm": 1.1089887487369232, + "learning_rate": 0.003, + "loss": 4.101, + "step": 8057 + }, + { + "epoch": 0.08058, + "grad_norm": 0.7539792902391563, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8058 + }, + { + "epoch": 0.08059, + "grad_norm": 0.7389100005311129, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8059 + }, + { + "epoch": 0.0806, + "grad_norm": 0.9262456642006766, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 8060 + }, + { + "epoch": 0.08061, + "grad_norm": 0.9795001713625814, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 8061 + }, + { + "epoch": 0.08062, + "grad_norm": 0.971892968973883, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 8062 + }, + { + "epoch": 0.08063, + "grad_norm": 0.9179217197155219, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 8063 + }, + { + "epoch": 0.08064, + "grad_norm": 0.8057836726183596, + "learning_rate": 0.003, + "loss": 4.083, + "step": 8064 + }, + { + "epoch": 0.08065, + "grad_norm": 0.7998960995576136, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 8065 + }, + { + "epoch": 0.08066, + "grad_norm": 0.8277143695899347, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 8066 + }, + { + "epoch": 0.08067, + "grad_norm": 0.8590956136100433, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8067 + }, + { + "epoch": 0.08068, + "grad_norm": 0.9394200313521635, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8068 + }, + { + "epoch": 0.08069, + "grad_norm": 1.090000328128135, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 8069 + }, + { + "epoch": 0.0807, + "grad_norm": 0.995341685424494, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8070 + }, + { + "epoch": 0.08071, + "grad_norm": 1.1683102498178706, + "learning_rate": 0.003, + "loss": 4.102, + "step": 8071 + }, + { + "epoch": 0.08072, + "grad_norm": 1.144707212294291, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8072 + }, + { + "epoch": 0.08073, + "grad_norm": 1.0813210978226437, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8073 + }, + { + "epoch": 0.08074, + "grad_norm": 0.9566230271673927, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 8074 + }, + { + "epoch": 0.08075, + "grad_norm": 0.95779239752232, + "learning_rate": 0.003, + "loss": 4.122, + "step": 8075 + }, + { + "epoch": 0.08076, + "grad_norm": 0.9663973763723217, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 8076 + }, + { + "epoch": 0.08077, + "grad_norm": 0.887449762689371, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 8077 + }, + { + "epoch": 0.08078, + "grad_norm": 0.8084711156241483, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8078 + }, + { + "epoch": 0.08079, + "grad_norm": 0.8723111575946362, + "learning_rate": 0.003, + "loss": 4.101, + "step": 8079 + }, + { + "epoch": 0.0808, + "grad_norm": 0.9387061640341692, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 8080 + }, + { + "epoch": 0.08081, + "grad_norm": 0.8355180753509858, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 8081 + }, + { + "epoch": 0.08082, + "grad_norm": 0.7557460372416088, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8082 + }, + { + "epoch": 0.08083, + "grad_norm": 0.7660773749158244, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 8083 + }, + { + "epoch": 0.08084, + "grad_norm": 0.8306004005127731, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 8084 + }, + { + "epoch": 0.08085, + "grad_norm": 0.8877741787140743, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 8085 + }, + { + "epoch": 0.08086, + "grad_norm": 0.8576233891044043, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 8086 + }, + { + "epoch": 0.08087, + "grad_norm": 0.9238061867576346, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8087 + }, + { + "epoch": 0.08088, + "grad_norm": 0.8883115357991481, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 8088 + }, + { + "epoch": 0.08089, + "grad_norm": 0.799907814286212, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 8089 + }, + { + "epoch": 0.0809, + "grad_norm": 0.8787224459097344, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 8090 + }, + { + "epoch": 0.08091, + "grad_norm": 1.1325378931178327, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 8091 + }, + { + "epoch": 0.08092, + "grad_norm": 1.1720491273192712, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8092 + }, + { + "epoch": 0.08093, + "grad_norm": 1.21086412881328, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 8093 + }, + { + "epoch": 0.08094, + "grad_norm": 1.2858697519995719, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 8094 + }, + { + "epoch": 0.08095, + "grad_norm": 0.984302984232329, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8095 + }, + { + "epoch": 0.08096, + "grad_norm": 0.996396135629695, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 8096 + }, + { + "epoch": 0.08097, + "grad_norm": 1.0159214458021701, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 8097 + }, + { + "epoch": 0.08098, + "grad_norm": 0.8969056697661771, + "learning_rate": 0.003, + "loss": 4.13, + "step": 8098 + }, + { + "epoch": 0.08099, + "grad_norm": 0.8810506298085342, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 8099 + }, + { + "epoch": 0.081, + "grad_norm": 0.9078212526479597, + "learning_rate": 0.003, + "loss": 4.106, + "step": 8100 + }, + { + "epoch": 0.08101, + "grad_norm": 0.949131998327021, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 8101 + }, + { + "epoch": 0.08102, + "grad_norm": 0.9893750375976563, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8102 + }, + { + "epoch": 0.08103, + "grad_norm": 0.9257798083015237, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8103 + }, + { + "epoch": 0.08104, + "grad_norm": 0.8487015013060492, + "learning_rate": 0.003, + "loss": 4.105, + "step": 8104 + }, + { + "epoch": 0.08105, + "grad_norm": 0.8526181950103013, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 8105 + }, + { + "epoch": 0.08106, + "grad_norm": 1.0786969740224188, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 8106 + }, + { + "epoch": 0.08107, + "grad_norm": 0.962314671540165, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 8107 + }, + { + "epoch": 0.08108, + "grad_norm": 0.9135863883972013, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8108 + }, + { + "epoch": 0.08109, + "grad_norm": 0.9401520709060528, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8109 + }, + { + "epoch": 0.0811, + "grad_norm": 0.9325748612282647, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 8110 + }, + { + "epoch": 0.08111, + "grad_norm": 0.743009273471666, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 8111 + }, + { + "epoch": 0.08112, + "grad_norm": 0.7374812555980911, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 8112 + }, + { + "epoch": 0.08113, + "grad_norm": 0.791116362217403, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 8113 + }, + { + "epoch": 0.08114, + "grad_norm": 0.8262660445589118, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8114 + }, + { + "epoch": 0.08115, + "grad_norm": 0.885927726638974, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 8115 + }, + { + "epoch": 0.08116, + "grad_norm": 1.0776387999266541, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 8116 + }, + { + "epoch": 0.08117, + "grad_norm": 1.0225223453354517, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 8117 + }, + { + "epoch": 0.08118, + "grad_norm": 0.9085632221138596, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8118 + }, + { + "epoch": 0.08119, + "grad_norm": 0.9197333700250085, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8119 + }, + { + "epoch": 0.0812, + "grad_norm": 1.0084531260316627, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8120 + }, + { + "epoch": 0.08121, + "grad_norm": 1.1783087596297812, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 8121 + }, + { + "epoch": 0.08122, + "grad_norm": 0.9639565880131747, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8122 + }, + { + "epoch": 0.08123, + "grad_norm": 0.9390472599240589, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 8123 + }, + { + "epoch": 0.08124, + "grad_norm": 0.8142870814265253, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 8124 + }, + { + "epoch": 0.08125, + "grad_norm": 0.788177789413209, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 8125 + }, + { + "epoch": 0.08126, + "grad_norm": 0.6912524998631848, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 8126 + }, + { + "epoch": 0.08127, + "grad_norm": 0.7226577365051999, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 8127 + }, + { + "epoch": 0.08128, + "grad_norm": 0.8021905080797842, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 8128 + }, + { + "epoch": 0.08129, + "grad_norm": 0.8332217552376096, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 8129 + }, + { + "epoch": 0.0813, + "grad_norm": 0.81623281600983, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8130 + }, + { + "epoch": 0.08131, + "grad_norm": 0.8995463311569188, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8131 + }, + { + "epoch": 0.08132, + "grad_norm": 1.2007744161280947, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8132 + }, + { + "epoch": 0.08133, + "grad_norm": 0.954576606375035, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8133 + }, + { + "epoch": 0.08134, + "grad_norm": 0.9799625763539351, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 8134 + }, + { + "epoch": 0.08135, + "grad_norm": 0.9363301593817257, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 8135 + }, + { + "epoch": 0.08136, + "grad_norm": 1.0237255724901253, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 8136 + }, + { + "epoch": 0.08137, + "grad_norm": 1.2813349124004845, + "learning_rate": 0.003, + "loss": 4.065, + "step": 8137 + }, + { + "epoch": 0.08138, + "grad_norm": 0.7579009503454592, + "learning_rate": 0.003, + "loss": 4.081, + "step": 8138 + }, + { + "epoch": 0.08139, + "grad_norm": 0.7301283473485153, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 8139 + }, + { + "epoch": 0.0814, + "grad_norm": 0.8224340230135018, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8140 + }, + { + "epoch": 0.08141, + "grad_norm": 1.0125830693831113, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 8141 + }, + { + "epoch": 0.08142, + "grad_norm": 1.0662048042861398, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 8142 + }, + { + "epoch": 0.08143, + "grad_norm": 1.0043431337989337, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8143 + }, + { + "epoch": 0.08144, + "grad_norm": 1.0395035240004515, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8144 + }, + { + "epoch": 0.08145, + "grad_norm": 0.9759886158509355, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 8145 + }, + { + "epoch": 0.08146, + "grad_norm": 0.9920073789649562, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 8146 + }, + { + "epoch": 0.08147, + "grad_norm": 1.180853967281421, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 8147 + }, + { + "epoch": 0.08148, + "grad_norm": 0.9563016175288455, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 8148 + }, + { + "epoch": 0.08149, + "grad_norm": 0.9375460459019592, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 8149 + }, + { + "epoch": 0.0815, + "grad_norm": 1.115020042058792, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 8150 + }, + { + "epoch": 0.08151, + "grad_norm": 0.9762709072033995, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 8151 + }, + { + "epoch": 0.08152, + "grad_norm": 0.8951996280257692, + "learning_rate": 0.003, + "loss": 4.109, + "step": 8152 + }, + { + "epoch": 0.08153, + "grad_norm": 0.8142401236996037, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8153 + }, + { + "epoch": 0.08154, + "grad_norm": 0.9155275103932833, + "learning_rate": 0.003, + "loss": 4.1, + "step": 8154 + }, + { + "epoch": 0.08155, + "grad_norm": 0.9719599421713861, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 8155 + }, + { + "epoch": 0.08156, + "grad_norm": 1.0665316378206227, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8156 + }, + { + "epoch": 0.08157, + "grad_norm": 1.135966914316917, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 8157 + }, + { + "epoch": 0.08158, + "grad_norm": 1.066902527807956, + "learning_rate": 0.003, + "loss": 4.134, + "step": 8158 + }, + { + "epoch": 0.08159, + "grad_norm": 0.9717089917395534, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 8159 + }, + { + "epoch": 0.0816, + "grad_norm": 0.869490845983841, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8160 + }, + { + "epoch": 0.08161, + "grad_norm": 0.6959074303014611, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8161 + }, + { + "epoch": 0.08162, + "grad_norm": 0.9986967763832338, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 8162 + }, + { + "epoch": 0.08163, + "grad_norm": 1.3187472317613815, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 8163 + }, + { + "epoch": 0.08164, + "grad_norm": 0.7687805545417059, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8164 + }, + { + "epoch": 0.08165, + "grad_norm": 0.7627252335014355, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8165 + }, + { + "epoch": 0.08166, + "grad_norm": 0.9603073324718235, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8166 + }, + { + "epoch": 0.08167, + "grad_norm": 1.1172406788641254, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 8167 + }, + { + "epoch": 0.08168, + "grad_norm": 0.8451325272847822, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 8168 + }, + { + "epoch": 0.08169, + "grad_norm": 0.9763570797625406, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 8169 + }, + { + "epoch": 0.0817, + "grad_norm": 1.0675923212188487, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 8170 + }, + { + "epoch": 0.08171, + "grad_norm": 1.0080405192206545, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8171 + }, + { + "epoch": 0.08172, + "grad_norm": 1.0205645004410517, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8172 + }, + { + "epoch": 0.08173, + "grad_norm": 1.2927512926089266, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 8173 + }, + { + "epoch": 0.08174, + "grad_norm": 0.8968040898130804, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 8174 + }, + { + "epoch": 0.08175, + "grad_norm": 0.8905007870405737, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8175 + }, + { + "epoch": 0.08176, + "grad_norm": 0.9541562811374488, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 8176 + }, + { + "epoch": 0.08177, + "grad_norm": 1.2033461912639642, + "learning_rate": 0.003, + "loss": 4.097, + "step": 8177 + }, + { + "epoch": 0.08178, + "grad_norm": 0.8207723835191075, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 8178 + }, + { + "epoch": 0.08179, + "grad_norm": 0.7950044143539097, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 8179 + }, + { + "epoch": 0.0818, + "grad_norm": 0.8278043595515765, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8180 + }, + { + "epoch": 0.08181, + "grad_norm": 0.7898952196220245, + "learning_rate": 0.003, + "loss": 4.081, + "step": 8181 + }, + { + "epoch": 0.08182, + "grad_norm": 0.8195849298752939, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 8182 + }, + { + "epoch": 0.08183, + "grad_norm": 0.9060159380953525, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8183 + }, + { + "epoch": 0.08184, + "grad_norm": 0.9695006051669894, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 8184 + }, + { + "epoch": 0.08185, + "grad_norm": 0.920077323996504, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 8185 + }, + { + "epoch": 0.08186, + "grad_norm": 0.8684848648180729, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8186 + }, + { + "epoch": 0.08187, + "grad_norm": 0.8950480855490158, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 8187 + }, + { + "epoch": 0.08188, + "grad_norm": 0.9082754445038497, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 8188 + }, + { + "epoch": 0.08189, + "grad_norm": 0.9345206345561519, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8189 + }, + { + "epoch": 0.0819, + "grad_norm": 0.9538918294206803, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 8190 + }, + { + "epoch": 0.08191, + "grad_norm": 1.1712240807589007, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 8191 + }, + { + "epoch": 0.08192, + "grad_norm": 0.8889111819447577, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 8192 + }, + { + "epoch": 0.08193, + "grad_norm": 0.9174624290372992, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8193 + }, + { + "epoch": 0.08194, + "grad_norm": 0.8802928153982901, + "learning_rate": 0.003, + "loss": 4.104, + "step": 8194 + }, + { + "epoch": 0.08195, + "grad_norm": 0.8173704619648245, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 8195 + }, + { + "epoch": 0.08196, + "grad_norm": 0.8797287670114331, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 8196 + }, + { + "epoch": 0.08197, + "grad_norm": 0.937319930269984, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8197 + }, + { + "epoch": 0.08198, + "grad_norm": 1.1242588728185605, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8198 + }, + { + "epoch": 0.08199, + "grad_norm": 0.8700428118547184, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8199 + }, + { + "epoch": 0.082, + "grad_norm": 0.8767979055978028, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 8200 + }, + { + "epoch": 0.08201, + "grad_norm": 0.8900605291980226, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 8201 + }, + { + "epoch": 0.08202, + "grad_norm": 0.9096496752774041, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 8202 + }, + { + "epoch": 0.08203, + "grad_norm": 0.9792634921450283, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 8203 + }, + { + "epoch": 0.08204, + "grad_norm": 1.0490129381297, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8204 + }, + { + "epoch": 0.08205, + "grad_norm": 1.0814614534865807, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 8205 + }, + { + "epoch": 0.08206, + "grad_norm": 0.9134513944524054, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 8206 + }, + { + "epoch": 0.08207, + "grad_norm": 1.00344956911268, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8207 + }, + { + "epoch": 0.08208, + "grad_norm": 0.9993535843590942, + "learning_rate": 0.003, + "loss": 4.1, + "step": 8208 + }, + { + "epoch": 0.08209, + "grad_norm": 1.0850834633621271, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 8209 + }, + { + "epoch": 0.0821, + "grad_norm": 1.113381435981985, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8210 + }, + { + "epoch": 0.08211, + "grad_norm": 1.1005650428474547, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8211 + }, + { + "epoch": 0.08212, + "grad_norm": 0.9040714007129831, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8212 + }, + { + "epoch": 0.08213, + "grad_norm": 0.9492708093617471, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 8213 + }, + { + "epoch": 0.08214, + "grad_norm": 1.117679266472159, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 8214 + }, + { + "epoch": 0.08215, + "grad_norm": 1.0190111526945924, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 8215 + }, + { + "epoch": 0.08216, + "grad_norm": 1.0878093466039667, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8216 + }, + { + "epoch": 0.08217, + "grad_norm": 0.9383248153244881, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 8217 + }, + { + "epoch": 0.08218, + "grad_norm": 0.8917500362995279, + "learning_rate": 0.003, + "loss": 4.092, + "step": 8218 + }, + { + "epoch": 0.08219, + "grad_norm": 0.9147124289798197, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 8219 + }, + { + "epoch": 0.0822, + "grad_norm": 1.0017630355328986, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 8220 + }, + { + "epoch": 0.08221, + "grad_norm": 1.1284879557676335, + "learning_rate": 0.003, + "loss": 4.122, + "step": 8221 + }, + { + "epoch": 0.08222, + "grad_norm": 1.003441827304382, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 8222 + }, + { + "epoch": 0.08223, + "grad_norm": 0.9619782927365945, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 8223 + }, + { + "epoch": 0.08224, + "grad_norm": 0.9756162129814597, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8224 + }, + { + "epoch": 0.08225, + "grad_norm": 0.9888357301690992, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 8225 + }, + { + "epoch": 0.08226, + "grad_norm": 0.9663381395639372, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 8226 + }, + { + "epoch": 0.08227, + "grad_norm": 1.0183404083190035, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 8227 + }, + { + "epoch": 0.08228, + "grad_norm": 1.140292961870533, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 8228 + }, + { + "epoch": 0.08229, + "grad_norm": 1.0203708219405985, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 8229 + }, + { + "epoch": 0.0823, + "grad_norm": 1.0163195425352944, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 8230 + }, + { + "epoch": 0.08231, + "grad_norm": 1.1712795476862792, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8231 + }, + { + "epoch": 0.08232, + "grad_norm": 1.0333865533848603, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8232 + }, + { + "epoch": 0.08233, + "grad_norm": 1.0259646369301396, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 8233 + }, + { + "epoch": 0.08234, + "grad_norm": 0.9259284266883602, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 8234 + }, + { + "epoch": 0.08235, + "grad_norm": 0.8058762523422451, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 8235 + }, + { + "epoch": 0.08236, + "grad_norm": 0.858811923745795, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 8236 + }, + { + "epoch": 0.08237, + "grad_norm": 0.9339055212007007, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 8237 + }, + { + "epoch": 0.08238, + "grad_norm": 0.9168589072673792, + "learning_rate": 0.003, + "loss": 4.104, + "step": 8238 + }, + { + "epoch": 0.08239, + "grad_norm": 0.9459787857366488, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 8239 + }, + { + "epoch": 0.0824, + "grad_norm": 1.0167156904516566, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 8240 + }, + { + "epoch": 0.08241, + "grad_norm": 1.074340977515434, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 8241 + }, + { + "epoch": 0.08242, + "grad_norm": 0.9810666497367793, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 8242 + }, + { + "epoch": 0.08243, + "grad_norm": 0.9995000094053831, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 8243 + }, + { + "epoch": 0.08244, + "grad_norm": 1.0082704251177956, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8244 + }, + { + "epoch": 0.08245, + "grad_norm": 1.245333937722013, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8245 + }, + { + "epoch": 0.08246, + "grad_norm": 0.9836272618729974, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8246 + }, + { + "epoch": 0.08247, + "grad_norm": 0.9425039973065011, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 8247 + }, + { + "epoch": 0.08248, + "grad_norm": 0.9823287792469457, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 8248 + }, + { + "epoch": 0.08249, + "grad_norm": 0.9588202707053501, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8249 + }, + { + "epoch": 0.0825, + "grad_norm": 0.9720158461900901, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8250 + }, + { + "epoch": 0.08251, + "grad_norm": 1.0653919600638315, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 8251 + }, + { + "epoch": 0.08252, + "grad_norm": 0.9189080871525903, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 8252 + }, + { + "epoch": 0.08253, + "grad_norm": 0.9618627197874235, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8253 + }, + { + "epoch": 0.08254, + "grad_norm": 0.9241157927248576, + "learning_rate": 0.003, + "loss": 4.085, + "step": 8254 + }, + { + "epoch": 0.08255, + "grad_norm": 1.05514630292859, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 8255 + }, + { + "epoch": 0.08256, + "grad_norm": 1.0556785701037765, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 8256 + }, + { + "epoch": 0.08257, + "grad_norm": 1.0625228873286099, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8257 + }, + { + "epoch": 0.08258, + "grad_norm": 0.8915720985402055, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 8258 + }, + { + "epoch": 0.08259, + "grad_norm": 0.7484394189351964, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 8259 + }, + { + "epoch": 0.0826, + "grad_norm": 0.7755690062742758, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 8260 + }, + { + "epoch": 0.08261, + "grad_norm": 0.7913615993039677, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 8261 + }, + { + "epoch": 0.08262, + "grad_norm": 0.7726032835071411, + "learning_rate": 0.003, + "loss": 4.123, + "step": 8262 + }, + { + "epoch": 0.08263, + "grad_norm": 0.8015224842829364, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8263 + }, + { + "epoch": 0.08264, + "grad_norm": 0.9833954108334119, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 8264 + }, + { + "epoch": 0.08265, + "grad_norm": 1.3245892086877982, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8265 + }, + { + "epoch": 0.08266, + "grad_norm": 0.6411106525224666, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 8266 + }, + { + "epoch": 0.08267, + "grad_norm": 0.6656579659129879, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 8267 + }, + { + "epoch": 0.08268, + "grad_norm": 0.7261153634127828, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8268 + }, + { + "epoch": 0.08269, + "grad_norm": 0.8544934625491182, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 8269 + }, + { + "epoch": 0.0827, + "grad_norm": 1.0458857129455876, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8270 + }, + { + "epoch": 0.08271, + "grad_norm": 0.9911064176984495, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 8271 + }, + { + "epoch": 0.08272, + "grad_norm": 0.9393573176422051, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 8272 + }, + { + "epoch": 0.08273, + "grad_norm": 0.8646719994209546, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 8273 + }, + { + "epoch": 0.08274, + "grad_norm": 0.833681481435964, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 8274 + }, + { + "epoch": 0.08275, + "grad_norm": 0.8499700263111327, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 8275 + }, + { + "epoch": 0.08276, + "grad_norm": 0.8761764628765608, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 8276 + }, + { + "epoch": 0.08277, + "grad_norm": 1.0009606164430283, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 8277 + }, + { + "epoch": 0.08278, + "grad_norm": 1.1219748051066134, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 8278 + }, + { + "epoch": 0.08279, + "grad_norm": 1.0353428874496953, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8279 + }, + { + "epoch": 0.0828, + "grad_norm": 1.032366749690802, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 8280 + }, + { + "epoch": 0.08281, + "grad_norm": 0.88361996817911, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 8281 + }, + { + "epoch": 0.08282, + "grad_norm": 0.7805065196095217, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 8282 + }, + { + "epoch": 0.08283, + "grad_norm": 0.7078532197092947, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 8283 + }, + { + "epoch": 0.08284, + "grad_norm": 0.8334930096391483, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 8284 + }, + { + "epoch": 0.08285, + "grad_norm": 0.9418443321092326, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8285 + }, + { + "epoch": 0.08286, + "grad_norm": 0.9196371932960127, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 8286 + }, + { + "epoch": 0.08287, + "grad_norm": 0.7266922490894868, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 8287 + }, + { + "epoch": 0.08288, + "grad_norm": 0.7577686283368227, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 8288 + }, + { + "epoch": 0.08289, + "grad_norm": 0.9147385160263024, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 8289 + }, + { + "epoch": 0.0829, + "grad_norm": 0.9970454387546689, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 8290 + }, + { + "epoch": 0.08291, + "grad_norm": 1.0525426776153293, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 8291 + }, + { + "epoch": 0.08292, + "grad_norm": 1.0216758548976381, + "learning_rate": 0.003, + "loss": 4.094, + "step": 8292 + }, + { + "epoch": 0.08293, + "grad_norm": 0.9754705735289286, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 8293 + }, + { + "epoch": 0.08294, + "grad_norm": 1.210783738859885, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 8294 + }, + { + "epoch": 0.08295, + "grad_norm": 1.0394514713837457, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8295 + }, + { + "epoch": 0.08296, + "grad_norm": 1.0556248494716691, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8296 + }, + { + "epoch": 0.08297, + "grad_norm": 1.0074376752131393, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8297 + }, + { + "epoch": 0.08298, + "grad_norm": 1.0587475783665163, + "learning_rate": 0.003, + "loss": 4.074, + "step": 8298 + }, + { + "epoch": 0.08299, + "grad_norm": 1.0609037924375446, + "learning_rate": 0.003, + "loss": 4.096, + "step": 8299 + }, + { + "epoch": 0.083, + "grad_norm": 0.9013525969374929, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 8300 + }, + { + "epoch": 0.08301, + "grad_norm": 0.8461074908827423, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8301 + }, + { + "epoch": 0.08302, + "grad_norm": 0.7808808902520736, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 8302 + }, + { + "epoch": 0.08303, + "grad_norm": 0.9240730304280933, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 8303 + }, + { + "epoch": 0.08304, + "grad_norm": 0.9410222419739239, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 8304 + }, + { + "epoch": 0.08305, + "grad_norm": 0.861091610388214, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8305 + }, + { + "epoch": 0.08306, + "grad_norm": 0.8150239293199747, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8306 + }, + { + "epoch": 0.08307, + "grad_norm": 0.9976234336488442, + "learning_rate": 0.003, + "loss": 4.089, + "step": 8307 + }, + { + "epoch": 0.08308, + "grad_norm": 1.1560035728708618, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8308 + }, + { + "epoch": 0.08309, + "grad_norm": 1.118929485449606, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 8309 + }, + { + "epoch": 0.0831, + "grad_norm": 0.8558868180407445, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8310 + }, + { + "epoch": 0.08311, + "grad_norm": 0.7250276766729903, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 8311 + }, + { + "epoch": 0.08312, + "grad_norm": 0.7096377864215281, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 8312 + }, + { + "epoch": 0.08313, + "grad_norm": 0.8317014820764579, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 8313 + }, + { + "epoch": 0.08314, + "grad_norm": 0.9061323428480831, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 8314 + }, + { + "epoch": 0.08315, + "grad_norm": 0.9783087646693294, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8315 + }, + { + "epoch": 0.08316, + "grad_norm": 1.1250180316240839, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8316 + }, + { + "epoch": 0.08317, + "grad_norm": 1.1731269535904751, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8317 + }, + { + "epoch": 0.08318, + "grad_norm": 1.0696733664284113, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 8318 + }, + { + "epoch": 0.08319, + "grad_norm": 0.9282818531386554, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 8319 + }, + { + "epoch": 0.0832, + "grad_norm": 1.108581382048844, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 8320 + }, + { + "epoch": 0.08321, + "grad_norm": 0.8810037509971127, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 8321 + }, + { + "epoch": 0.08322, + "grad_norm": 0.8426847706169508, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8322 + }, + { + "epoch": 0.08323, + "grad_norm": 0.8922307194607451, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 8323 + }, + { + "epoch": 0.08324, + "grad_norm": 1.0048199819448347, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 8324 + }, + { + "epoch": 0.08325, + "grad_norm": 1.1134236440962169, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 8325 + }, + { + "epoch": 0.08326, + "grad_norm": 0.9733022633639508, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 8326 + }, + { + "epoch": 0.08327, + "grad_norm": 1.0054631036308757, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 8327 + }, + { + "epoch": 0.08328, + "grad_norm": 1.038948492795079, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8328 + }, + { + "epoch": 0.08329, + "grad_norm": 1.0033030320157774, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 8329 + }, + { + "epoch": 0.0833, + "grad_norm": 0.9755743730426134, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8330 + }, + { + "epoch": 0.08331, + "grad_norm": 1.0383272686924012, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 8331 + }, + { + "epoch": 0.08332, + "grad_norm": 1.0593791914963704, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 8332 + }, + { + "epoch": 0.08333, + "grad_norm": 0.9791667631234542, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 8333 + }, + { + "epoch": 0.08334, + "grad_norm": 1.2529702981986344, + "learning_rate": 0.003, + "loss": 4.113, + "step": 8334 + }, + { + "epoch": 0.08335, + "grad_norm": 0.7364862616928067, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8335 + }, + { + "epoch": 0.08336, + "grad_norm": 0.7413619580049867, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 8336 + }, + { + "epoch": 0.08337, + "grad_norm": 0.7371250991151914, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 8337 + }, + { + "epoch": 0.08338, + "grad_norm": 0.7373545100133702, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8338 + }, + { + "epoch": 0.08339, + "grad_norm": 0.8437067988637226, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 8339 + }, + { + "epoch": 0.0834, + "grad_norm": 0.9548820887564872, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 8340 + }, + { + "epoch": 0.08341, + "grad_norm": 0.9986280957539666, + "learning_rate": 0.003, + "loss": 4.12, + "step": 8341 + }, + { + "epoch": 0.08342, + "grad_norm": 1.0692833169544227, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 8342 + }, + { + "epoch": 0.08343, + "grad_norm": 0.8537320866239007, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 8343 + }, + { + "epoch": 0.08344, + "grad_norm": 0.8521570469099049, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 8344 + }, + { + "epoch": 0.08345, + "grad_norm": 0.9526963753585949, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 8345 + }, + { + "epoch": 0.08346, + "grad_norm": 1.0064844318689625, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 8346 + }, + { + "epoch": 0.08347, + "grad_norm": 1.1864005892522749, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 8347 + }, + { + "epoch": 0.08348, + "grad_norm": 0.9763480219332881, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8348 + }, + { + "epoch": 0.08349, + "grad_norm": 1.1231533499463706, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8349 + }, + { + "epoch": 0.0835, + "grad_norm": 0.992323166083528, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 8350 + }, + { + "epoch": 0.08351, + "grad_norm": 1.0428277918008841, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 8351 + }, + { + "epoch": 0.08352, + "grad_norm": 1.020903771905812, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 8352 + }, + { + "epoch": 0.08353, + "grad_norm": 0.8986777589795104, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 8353 + }, + { + "epoch": 0.08354, + "grad_norm": 0.7969639569922113, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 8354 + }, + { + "epoch": 0.08355, + "grad_norm": 0.8429266001908198, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8355 + }, + { + "epoch": 0.08356, + "grad_norm": 0.8533752836936277, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 8356 + }, + { + "epoch": 0.08357, + "grad_norm": 0.8779051945830572, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 8357 + }, + { + "epoch": 0.08358, + "grad_norm": 1.1304931543239005, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 8358 + }, + { + "epoch": 0.08359, + "grad_norm": 1.1755980604584186, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8359 + }, + { + "epoch": 0.0836, + "grad_norm": 0.8820015012892566, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 8360 + }, + { + "epoch": 0.08361, + "grad_norm": 0.7968932521237145, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8361 + }, + { + "epoch": 0.08362, + "grad_norm": 0.7705615939438236, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 8362 + }, + { + "epoch": 0.08363, + "grad_norm": 0.8336591440869687, + "learning_rate": 0.003, + "loss": 4.117, + "step": 8363 + }, + { + "epoch": 0.08364, + "grad_norm": 0.8681252935444068, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 8364 + }, + { + "epoch": 0.08365, + "grad_norm": 0.7925044346326383, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8365 + }, + { + "epoch": 0.08366, + "grad_norm": 0.9118473686485629, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 8366 + }, + { + "epoch": 0.08367, + "grad_norm": 0.9863807106347243, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 8367 + }, + { + "epoch": 0.08368, + "grad_norm": 1.074863795592151, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 8368 + }, + { + "epoch": 0.08369, + "grad_norm": 1.0787611844668858, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 8369 + }, + { + "epoch": 0.0837, + "grad_norm": 1.085781743841614, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 8370 + }, + { + "epoch": 0.08371, + "grad_norm": 0.9121831664388997, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 8371 + }, + { + "epoch": 0.08372, + "grad_norm": 0.9425613813014645, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 8372 + }, + { + "epoch": 0.08373, + "grad_norm": 0.8475164269153014, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 8373 + }, + { + "epoch": 0.08374, + "grad_norm": 0.8365839586448178, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 8374 + }, + { + "epoch": 0.08375, + "grad_norm": 0.9788780251329858, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8375 + }, + { + "epoch": 0.08376, + "grad_norm": 1.1289608704104126, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8376 + }, + { + "epoch": 0.08377, + "grad_norm": 0.8654509843602431, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 8377 + }, + { + "epoch": 0.08378, + "grad_norm": 0.8972839474588501, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 8378 + }, + { + "epoch": 0.08379, + "grad_norm": 0.8716796230605194, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8379 + }, + { + "epoch": 0.0838, + "grad_norm": 0.8997359385164568, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8380 + }, + { + "epoch": 0.08381, + "grad_norm": 1.1955867056419545, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8381 + }, + { + "epoch": 0.08382, + "grad_norm": 1.100200909479212, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8382 + }, + { + "epoch": 0.08383, + "grad_norm": 1.0062419897932509, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 8383 + }, + { + "epoch": 0.08384, + "grad_norm": 1.04090096158985, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 8384 + }, + { + "epoch": 0.08385, + "grad_norm": 1.0428807613061897, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8385 + }, + { + "epoch": 0.08386, + "grad_norm": 1.092704813209603, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 8386 + }, + { + "epoch": 0.08387, + "grad_norm": 1.0018898539564087, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8387 + }, + { + "epoch": 0.08388, + "grad_norm": 1.0467770477939047, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 8388 + }, + { + "epoch": 0.08389, + "grad_norm": 0.9696307153718503, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8389 + }, + { + "epoch": 0.0839, + "grad_norm": 0.9834999621136172, + "learning_rate": 0.003, + "loss": 4.074, + "step": 8390 + }, + { + "epoch": 0.08391, + "grad_norm": 0.9555752100171525, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 8391 + }, + { + "epoch": 0.08392, + "grad_norm": 0.9505299882219872, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8392 + }, + { + "epoch": 0.08393, + "grad_norm": 1.0183911107738324, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 8393 + }, + { + "epoch": 0.08394, + "grad_norm": 0.9640779833177097, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 8394 + }, + { + "epoch": 0.08395, + "grad_norm": 0.8142550950148915, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 8395 + }, + { + "epoch": 0.08396, + "grad_norm": 0.8371291056267166, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8396 + }, + { + "epoch": 0.08397, + "grad_norm": 0.9596925071423696, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 8397 + }, + { + "epoch": 0.08398, + "grad_norm": 1.1271033499051497, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 8398 + }, + { + "epoch": 0.08399, + "grad_norm": 0.9775306799027432, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 8399 + }, + { + "epoch": 0.084, + "grad_norm": 0.9959146702560995, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 8400 + }, + { + "epoch": 0.08401, + "grad_norm": 1.0833441388968759, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8401 + }, + { + "epoch": 0.08402, + "grad_norm": 0.8392391540233762, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 8402 + }, + { + "epoch": 0.08403, + "grad_norm": 0.7315853340798093, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 8403 + }, + { + "epoch": 0.08404, + "grad_norm": 0.8000480577303863, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8404 + }, + { + "epoch": 0.08405, + "grad_norm": 0.8795325707481986, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 8405 + }, + { + "epoch": 0.08406, + "grad_norm": 0.9618122878448181, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 8406 + }, + { + "epoch": 0.08407, + "grad_norm": 1.046916378488563, + "learning_rate": 0.003, + "loss": 4.1411, + "step": 8407 + }, + { + "epoch": 0.08408, + "grad_norm": 1.1743216286266962, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 8408 + }, + { + "epoch": 0.08409, + "grad_norm": 0.8200986022886277, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 8409 + }, + { + "epoch": 0.0841, + "grad_norm": 0.7601736533748236, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 8410 + }, + { + "epoch": 0.08411, + "grad_norm": 0.7930919482404809, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 8411 + }, + { + "epoch": 0.08412, + "grad_norm": 0.8236918876291852, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 8412 + }, + { + "epoch": 0.08413, + "grad_norm": 0.7255830504638703, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8413 + }, + { + "epoch": 0.08414, + "grad_norm": 0.6364226525778399, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 8414 + }, + { + "epoch": 0.08415, + "grad_norm": 0.6283569979695853, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 8415 + }, + { + "epoch": 0.08416, + "grad_norm": 0.5584551617182978, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8416 + }, + { + "epoch": 0.08417, + "grad_norm": 0.6047478150553158, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8417 + }, + { + "epoch": 0.08418, + "grad_norm": 0.8178534299204772, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8418 + }, + { + "epoch": 0.08419, + "grad_norm": 1.2445876162198983, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 8419 + }, + { + "epoch": 0.0842, + "grad_norm": 1.2032672789944496, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 8420 + }, + { + "epoch": 0.08421, + "grad_norm": 0.9301781996834642, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 8421 + }, + { + "epoch": 0.08422, + "grad_norm": 1.043640415248368, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 8422 + }, + { + "epoch": 0.08423, + "grad_norm": 1.177651483822221, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 8423 + }, + { + "epoch": 0.08424, + "grad_norm": 0.8367571494130656, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 8424 + }, + { + "epoch": 0.08425, + "grad_norm": 0.7649028313402926, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 8425 + }, + { + "epoch": 0.08426, + "grad_norm": 0.7796505596514975, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 8426 + }, + { + "epoch": 0.08427, + "grad_norm": 0.8939607809315887, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 8427 + }, + { + "epoch": 0.08428, + "grad_norm": 1.1205858181962776, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8428 + }, + { + "epoch": 0.08429, + "grad_norm": 0.9351377901801635, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 8429 + }, + { + "epoch": 0.0843, + "grad_norm": 1.1917523797683225, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 8430 + }, + { + "epoch": 0.08431, + "grad_norm": 1.0769729281874565, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 8431 + }, + { + "epoch": 0.08432, + "grad_norm": 1.0675313880133899, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 8432 + }, + { + "epoch": 0.08433, + "grad_norm": 1.0814649771944809, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 8433 + }, + { + "epoch": 0.08434, + "grad_norm": 1.1374803833691784, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 8434 + }, + { + "epoch": 0.08435, + "grad_norm": 1.042128215570702, + "learning_rate": 0.003, + "loss": 4.103, + "step": 8435 + }, + { + "epoch": 0.08436, + "grad_norm": 1.121784999344552, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 8436 + }, + { + "epoch": 0.08437, + "grad_norm": 1.0213150253295122, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 8437 + }, + { + "epoch": 0.08438, + "grad_norm": 0.9921329826792672, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8438 + }, + { + "epoch": 0.08439, + "grad_norm": 0.9095075680838063, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 8439 + }, + { + "epoch": 0.0844, + "grad_norm": 0.9142515498264308, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 8440 + }, + { + "epoch": 0.08441, + "grad_norm": 0.9369650500084913, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 8441 + }, + { + "epoch": 0.08442, + "grad_norm": 1.0283524745080719, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 8442 + }, + { + "epoch": 0.08443, + "grad_norm": 1.3918574878910746, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 8443 + }, + { + "epoch": 0.08444, + "grad_norm": 1.0797096461735152, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 8444 + }, + { + "epoch": 0.08445, + "grad_norm": 1.0810997939018532, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 8445 + }, + { + "epoch": 0.08446, + "grad_norm": 1.0629108011545507, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 8446 + }, + { + "epoch": 0.08447, + "grad_norm": 1.1155516321536225, + "learning_rate": 0.003, + "loss": 4.118, + "step": 8447 + }, + { + "epoch": 0.08448, + "grad_norm": 0.8902814958665589, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 8448 + }, + { + "epoch": 0.08449, + "grad_norm": 0.8921467481709783, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 8449 + }, + { + "epoch": 0.0845, + "grad_norm": 0.9566675588218161, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8450 + }, + { + "epoch": 0.08451, + "grad_norm": 0.8487439459049356, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8451 + }, + { + "epoch": 0.08452, + "grad_norm": 0.8216051671856134, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 8452 + }, + { + "epoch": 0.08453, + "grad_norm": 0.71083164292099, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 8453 + }, + { + "epoch": 0.08454, + "grad_norm": 0.6932399778796303, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8454 + }, + { + "epoch": 0.08455, + "grad_norm": 0.6245955308624751, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 8455 + }, + { + "epoch": 0.08456, + "grad_norm": 0.710232969209134, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8456 + }, + { + "epoch": 0.08457, + "grad_norm": 0.8382923593356233, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 8457 + }, + { + "epoch": 0.08458, + "grad_norm": 1.1260446393029737, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 8458 + }, + { + "epoch": 0.08459, + "grad_norm": 1.1147365995229561, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 8459 + }, + { + "epoch": 0.0846, + "grad_norm": 0.952014546149753, + "learning_rate": 0.003, + "loss": 4.102, + "step": 8460 + }, + { + "epoch": 0.08461, + "grad_norm": 1.0093825873691928, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 8461 + }, + { + "epoch": 0.08462, + "grad_norm": 1.082164079498455, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 8462 + }, + { + "epoch": 0.08463, + "grad_norm": 0.9339020056620597, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 8463 + }, + { + "epoch": 0.08464, + "grad_norm": 1.1038150626697216, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 8464 + }, + { + "epoch": 0.08465, + "grad_norm": 0.8668252080144988, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 8465 + }, + { + "epoch": 0.08466, + "grad_norm": 1.0004324146635026, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 8466 + }, + { + "epoch": 0.08467, + "grad_norm": 1.1348520322515445, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 8467 + }, + { + "epoch": 0.08468, + "grad_norm": 0.839905944482792, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 8468 + }, + { + "epoch": 0.08469, + "grad_norm": 0.8611295736656586, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8469 + }, + { + "epoch": 0.0847, + "grad_norm": 1.0042989946600784, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8470 + }, + { + "epoch": 0.08471, + "grad_norm": 1.2058096444243045, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 8471 + }, + { + "epoch": 0.08472, + "grad_norm": 0.9217538372683004, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8472 + }, + { + "epoch": 0.08473, + "grad_norm": 0.9762750638792925, + "learning_rate": 0.003, + "loss": 4.078, + "step": 8473 + }, + { + "epoch": 0.08474, + "grad_norm": 0.9744705402329853, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8474 + }, + { + "epoch": 0.08475, + "grad_norm": 0.9522311246673508, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 8475 + }, + { + "epoch": 0.08476, + "grad_norm": 0.8380797124582146, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 8476 + }, + { + "epoch": 0.08477, + "grad_norm": 0.9117904343682559, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 8477 + }, + { + "epoch": 0.08478, + "grad_norm": 0.9942329046979015, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8478 + }, + { + "epoch": 0.08479, + "grad_norm": 1.1633063589143064, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 8479 + }, + { + "epoch": 0.0848, + "grad_norm": 0.8464171119669712, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 8480 + }, + { + "epoch": 0.08481, + "grad_norm": 0.7888781982408283, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 8481 + }, + { + "epoch": 0.08482, + "grad_norm": 0.6832850025599434, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8482 + }, + { + "epoch": 0.08483, + "grad_norm": 0.647607987664764, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 8483 + }, + { + "epoch": 0.08484, + "grad_norm": 0.8013913578651249, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8484 + }, + { + "epoch": 0.08485, + "grad_norm": 0.9673099824178485, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8485 + }, + { + "epoch": 0.08486, + "grad_norm": 1.3943328728885203, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 8486 + }, + { + "epoch": 0.08487, + "grad_norm": 0.9092491054040347, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 8487 + }, + { + "epoch": 0.08488, + "grad_norm": 0.9362284721268309, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 8488 + }, + { + "epoch": 0.08489, + "grad_norm": 1.0548388718733714, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8489 + }, + { + "epoch": 0.0849, + "grad_norm": 1.007922010373895, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 8490 + }, + { + "epoch": 0.08491, + "grad_norm": 0.950728262769965, + "learning_rate": 0.003, + "loss": 4.113, + "step": 8491 + }, + { + "epoch": 0.08492, + "grad_norm": 0.9646047582212579, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8492 + }, + { + "epoch": 0.08493, + "grad_norm": 1.1639240847412275, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 8493 + }, + { + "epoch": 0.08494, + "grad_norm": 0.8714062193959657, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 8494 + }, + { + "epoch": 0.08495, + "grad_norm": 0.8652047802203818, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 8495 + }, + { + "epoch": 0.08496, + "grad_norm": 0.76430691704875, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 8496 + }, + { + "epoch": 0.08497, + "grad_norm": 0.8622147074710873, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 8497 + }, + { + "epoch": 0.08498, + "grad_norm": 1.0688759380024537, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 8498 + }, + { + "epoch": 0.08499, + "grad_norm": 1.3345686099220913, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 8499 + }, + { + "epoch": 0.085, + "grad_norm": 1.0074196125905852, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 8500 + }, + { + "epoch": 0.08501, + "grad_norm": 1.0205831966113184, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8501 + }, + { + "epoch": 0.08502, + "grad_norm": 0.8448826348859688, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 8502 + }, + { + "epoch": 0.08503, + "grad_norm": 0.8429552443487274, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8503 + }, + { + "epoch": 0.08504, + "grad_norm": 0.8110637286450764, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 8504 + }, + { + "epoch": 0.08505, + "grad_norm": 0.7803465412418662, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 8505 + }, + { + "epoch": 0.08506, + "grad_norm": 0.7741778086269289, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 8506 + }, + { + "epoch": 0.08507, + "grad_norm": 0.9284973996173289, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 8507 + }, + { + "epoch": 0.08508, + "grad_norm": 1.219393337434796, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 8508 + }, + { + "epoch": 0.08509, + "grad_norm": 0.9541456980504919, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8509 + }, + { + "epoch": 0.0851, + "grad_norm": 1.0650148464854814, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 8510 + }, + { + "epoch": 0.08511, + "grad_norm": 1.0213709394078312, + "learning_rate": 0.003, + "loss": 4.106, + "step": 8511 + }, + { + "epoch": 0.08512, + "grad_norm": 1.02423392398546, + "learning_rate": 0.003, + "loss": 4.1333, + "step": 8512 + }, + { + "epoch": 0.08513, + "grad_norm": 0.9925200002172757, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8513 + }, + { + "epoch": 0.08514, + "grad_norm": 1.0517186955602054, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8514 + }, + { + "epoch": 0.08515, + "grad_norm": 1.051372765446143, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 8515 + }, + { + "epoch": 0.08516, + "grad_norm": 0.961139345826629, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 8516 + }, + { + "epoch": 0.08517, + "grad_norm": 1.1383847613698732, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8517 + }, + { + "epoch": 0.08518, + "grad_norm": 1.0614310515013552, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8518 + }, + { + "epoch": 0.08519, + "grad_norm": 0.9909535453090349, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8519 + }, + { + "epoch": 0.0852, + "grad_norm": 0.9606661139384731, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 8520 + }, + { + "epoch": 0.08521, + "grad_norm": 1.050436970571876, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 8521 + }, + { + "epoch": 0.08522, + "grad_norm": 0.9610076466521358, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 8522 + }, + { + "epoch": 0.08523, + "grad_norm": 0.8824080000960122, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8523 + }, + { + "epoch": 0.08524, + "grad_norm": 0.829155013045862, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8524 + }, + { + "epoch": 0.08525, + "grad_norm": 0.9207597550538581, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 8525 + }, + { + "epoch": 0.08526, + "grad_norm": 1.2774950411838009, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 8526 + }, + { + "epoch": 0.08527, + "grad_norm": 1.0681627254256, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 8527 + }, + { + "epoch": 0.08528, + "grad_norm": 1.1873985636491318, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8528 + }, + { + "epoch": 0.08529, + "grad_norm": 0.9005438289637654, + "learning_rate": 0.003, + "loss": 4.101, + "step": 8529 + }, + { + "epoch": 0.0853, + "grad_norm": 0.943871451895167, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 8530 + }, + { + "epoch": 0.08531, + "grad_norm": 0.9930340272788102, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 8531 + }, + { + "epoch": 0.08532, + "grad_norm": 1.0464636343499278, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 8532 + }, + { + "epoch": 0.08533, + "grad_norm": 1.1742467639306242, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 8533 + }, + { + "epoch": 0.08534, + "grad_norm": 0.9125996800869809, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 8534 + }, + { + "epoch": 0.08535, + "grad_norm": 0.9589681901028572, + "learning_rate": 0.003, + "loss": 4.089, + "step": 8535 + }, + { + "epoch": 0.08536, + "grad_norm": 1.003312873568851, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 8536 + }, + { + "epoch": 0.08537, + "grad_norm": 1.0579497731198382, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8537 + }, + { + "epoch": 0.08538, + "grad_norm": 1.029713587209702, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 8538 + }, + { + "epoch": 0.08539, + "grad_norm": 1.0616355170105047, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 8539 + }, + { + "epoch": 0.0854, + "grad_norm": 0.8632828924317381, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8540 + }, + { + "epoch": 0.08541, + "grad_norm": 0.8004540770861811, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 8541 + }, + { + "epoch": 0.08542, + "grad_norm": 0.8044230686905137, + "learning_rate": 0.003, + "loss": 4.111, + "step": 8542 + }, + { + "epoch": 0.08543, + "grad_norm": 0.892873539934706, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 8543 + }, + { + "epoch": 0.08544, + "grad_norm": 0.961046548554565, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 8544 + }, + { + "epoch": 0.08545, + "grad_norm": 1.0248813672026638, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8545 + }, + { + "epoch": 0.08546, + "grad_norm": 0.9312362207459531, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 8546 + }, + { + "epoch": 0.08547, + "grad_norm": 1.044567535355616, + "learning_rate": 0.003, + "loss": 4.113, + "step": 8547 + }, + { + "epoch": 0.08548, + "grad_norm": 1.3226993174635777, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 8548 + }, + { + "epoch": 0.08549, + "grad_norm": 1.1046724941315957, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 8549 + }, + { + "epoch": 0.0855, + "grad_norm": 1.0576792234018175, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 8550 + }, + { + "epoch": 0.08551, + "grad_norm": 0.9612912995427461, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 8551 + }, + { + "epoch": 0.08552, + "grad_norm": 1.054325780580547, + "learning_rate": 0.003, + "loss": 4.103, + "step": 8552 + }, + { + "epoch": 0.08553, + "grad_norm": 0.9892609567397835, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 8553 + }, + { + "epoch": 0.08554, + "grad_norm": 1.1262764733880528, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 8554 + }, + { + "epoch": 0.08555, + "grad_norm": 1.0051709334035899, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 8555 + }, + { + "epoch": 0.08556, + "grad_norm": 0.9899672021307735, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 8556 + }, + { + "epoch": 0.08557, + "grad_norm": 0.8190451863778596, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 8557 + }, + { + "epoch": 0.08558, + "grad_norm": 0.7138030450904894, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 8558 + }, + { + "epoch": 0.08559, + "grad_norm": 0.7010492899381877, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 8559 + }, + { + "epoch": 0.0856, + "grad_norm": 0.6839596608467561, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8560 + }, + { + "epoch": 0.08561, + "grad_norm": 0.8009807285545935, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 8561 + }, + { + "epoch": 0.08562, + "grad_norm": 0.9527168047081275, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 8562 + }, + { + "epoch": 0.08563, + "grad_norm": 1.203093505131277, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 8563 + }, + { + "epoch": 0.08564, + "grad_norm": 0.8506805712070058, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 8564 + }, + { + "epoch": 0.08565, + "grad_norm": 0.8867525740638995, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 8565 + }, + { + "epoch": 0.08566, + "grad_norm": 0.8312063974139594, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8566 + }, + { + "epoch": 0.08567, + "grad_norm": 0.9998824441733825, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 8567 + }, + { + "epoch": 0.08568, + "grad_norm": 0.932656507572589, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 8568 + }, + { + "epoch": 0.08569, + "grad_norm": 0.9868388758897252, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 8569 + }, + { + "epoch": 0.0857, + "grad_norm": 1.2304091841301328, + "learning_rate": 0.003, + "loss": 4.1263, + "step": 8570 + }, + { + "epoch": 0.08571, + "grad_norm": 1.0204084145493275, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 8571 + }, + { + "epoch": 0.08572, + "grad_norm": 1.034976370606038, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 8572 + }, + { + "epoch": 0.08573, + "grad_norm": 1.0524807086959482, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8573 + }, + { + "epoch": 0.08574, + "grad_norm": 0.877955924468013, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 8574 + }, + { + "epoch": 0.08575, + "grad_norm": 0.8300598690466507, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 8575 + }, + { + "epoch": 0.08576, + "grad_norm": 0.9308275145300718, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 8576 + }, + { + "epoch": 0.08577, + "grad_norm": 1.0891867015620795, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 8577 + }, + { + "epoch": 0.08578, + "grad_norm": 1.1639439297733545, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 8578 + }, + { + "epoch": 0.08579, + "grad_norm": 0.8114637979777802, + "learning_rate": 0.003, + "loss": 4.092, + "step": 8579 + }, + { + "epoch": 0.0858, + "grad_norm": 0.8219563908248404, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 8580 + }, + { + "epoch": 0.08581, + "grad_norm": 0.9514354441568278, + "learning_rate": 0.003, + "loss": 4.079, + "step": 8581 + }, + { + "epoch": 0.08582, + "grad_norm": 1.14862016991355, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 8582 + }, + { + "epoch": 0.08583, + "grad_norm": 0.9554904081552709, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 8583 + }, + { + "epoch": 0.08584, + "grad_norm": 1.1919961547318443, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 8584 + }, + { + "epoch": 0.08585, + "grad_norm": 0.9220063547173044, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 8585 + }, + { + "epoch": 0.08586, + "grad_norm": 0.7917210109825877, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 8586 + }, + { + "epoch": 0.08587, + "grad_norm": 0.6974880795367134, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 8587 + }, + { + "epoch": 0.08588, + "grad_norm": 0.7500850079717578, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 8588 + }, + { + "epoch": 0.08589, + "grad_norm": 0.9188442664538239, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 8589 + }, + { + "epoch": 0.0859, + "grad_norm": 0.9381614785643198, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8590 + }, + { + "epoch": 0.08591, + "grad_norm": 0.8381703502207993, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8591 + }, + { + "epoch": 0.08592, + "grad_norm": 0.9957238942437086, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 8592 + }, + { + "epoch": 0.08593, + "grad_norm": 1.092680779855448, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 8593 + }, + { + "epoch": 0.08594, + "grad_norm": 1.0701608564717937, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 8594 + }, + { + "epoch": 0.08595, + "grad_norm": 1.3280866738943609, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 8595 + }, + { + "epoch": 0.08596, + "grad_norm": 0.8332740836636172, + "learning_rate": 0.003, + "loss": 4.112, + "step": 8596 + }, + { + "epoch": 0.08597, + "grad_norm": 0.8133158050233575, + "learning_rate": 0.003, + "loss": 4.108, + "step": 8597 + }, + { + "epoch": 0.08598, + "grad_norm": 0.8677960828642273, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 8598 + }, + { + "epoch": 0.08599, + "grad_norm": 0.9453091797543772, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 8599 + }, + { + "epoch": 0.086, + "grad_norm": 1.101573205386506, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 8600 + }, + { + "epoch": 0.08601, + "grad_norm": 0.917760372367751, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8601 + }, + { + "epoch": 0.08602, + "grad_norm": 0.9154001124148273, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 8602 + }, + { + "epoch": 0.08603, + "grad_norm": 0.9398969707343389, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 8603 + }, + { + "epoch": 0.08604, + "grad_norm": 0.8825392522698661, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8604 + }, + { + "epoch": 0.08605, + "grad_norm": 0.8845588681011858, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 8605 + }, + { + "epoch": 0.08606, + "grad_norm": 0.8767686307766559, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 8606 + }, + { + "epoch": 0.08607, + "grad_norm": 0.9979908090017543, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 8607 + }, + { + "epoch": 0.08608, + "grad_norm": 1.167260745143609, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8608 + }, + { + "epoch": 0.08609, + "grad_norm": 1.0047803494363299, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 8609 + }, + { + "epoch": 0.0861, + "grad_norm": 1.1790950381285876, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 8610 + }, + { + "epoch": 0.08611, + "grad_norm": 0.9155454006163548, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 8611 + }, + { + "epoch": 0.08612, + "grad_norm": 1.1200626341917086, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 8612 + }, + { + "epoch": 0.08613, + "grad_norm": 1.0417155737120185, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 8613 + }, + { + "epoch": 0.08614, + "grad_norm": 0.99622616457739, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8614 + }, + { + "epoch": 0.08615, + "grad_norm": 1.1192206628076375, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 8615 + }, + { + "epoch": 0.08616, + "grad_norm": 1.2384683745781875, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 8616 + }, + { + "epoch": 0.08617, + "grad_norm": 1.0970291121489821, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8617 + }, + { + "epoch": 0.08618, + "grad_norm": 1.1560410206984835, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 8618 + }, + { + "epoch": 0.08619, + "grad_norm": 0.8941331840757659, + "learning_rate": 0.003, + "loss": 4.117, + "step": 8619 + }, + { + "epoch": 0.0862, + "grad_norm": 0.8951808980836901, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8620 + }, + { + "epoch": 0.08621, + "grad_norm": 0.9144905457902658, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 8621 + }, + { + "epoch": 0.08622, + "grad_norm": 0.846593765826877, + "learning_rate": 0.003, + "loss": 4.114, + "step": 8622 + }, + { + "epoch": 0.08623, + "grad_norm": 0.8296474727794095, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 8623 + }, + { + "epoch": 0.08624, + "grad_norm": 0.857115826300225, + "learning_rate": 0.003, + "loss": 4.086, + "step": 8624 + }, + { + "epoch": 0.08625, + "grad_norm": 0.9770816255055488, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 8625 + }, + { + "epoch": 0.08626, + "grad_norm": 1.054544039763614, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 8626 + }, + { + "epoch": 0.08627, + "grad_norm": 0.976150638507909, + "learning_rate": 0.003, + "loss": 4.097, + "step": 8627 + }, + { + "epoch": 0.08628, + "grad_norm": 0.9236639095745208, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 8628 + }, + { + "epoch": 0.08629, + "grad_norm": 0.8825959695592736, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 8629 + }, + { + "epoch": 0.0863, + "grad_norm": 0.8902473144031249, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 8630 + }, + { + "epoch": 0.08631, + "grad_norm": 0.9149383790153159, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 8631 + }, + { + "epoch": 0.08632, + "grad_norm": 0.9182837062894001, + "learning_rate": 0.003, + "loss": 4.084, + "step": 8632 + }, + { + "epoch": 0.08633, + "grad_norm": 1.0899428012063783, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 8633 + }, + { + "epoch": 0.08634, + "grad_norm": 1.0629044913328984, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 8634 + }, + { + "epoch": 0.08635, + "grad_norm": 0.9717726584129625, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8635 + }, + { + "epoch": 0.08636, + "grad_norm": 1.1164187292125398, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8636 + }, + { + "epoch": 0.08637, + "grad_norm": 1.3025486362018337, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 8637 + }, + { + "epoch": 0.08638, + "grad_norm": 0.8541936401852307, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8638 + }, + { + "epoch": 0.08639, + "grad_norm": 0.771724595769981, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8639 + }, + { + "epoch": 0.0864, + "grad_norm": 0.866492675972138, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8640 + }, + { + "epoch": 0.08641, + "grad_norm": 1.3350454450812863, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 8641 + }, + { + "epoch": 0.08642, + "grad_norm": 0.8473052034583577, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 8642 + }, + { + "epoch": 0.08643, + "grad_norm": 0.9423298490396024, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8643 + }, + { + "epoch": 0.08644, + "grad_norm": 0.9804863052777438, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8644 + }, + { + "epoch": 0.08645, + "grad_norm": 0.9176586395090423, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8645 + }, + { + "epoch": 0.08646, + "grad_norm": 1.051248679692915, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 8646 + }, + { + "epoch": 0.08647, + "grad_norm": 1.2067666871909615, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8647 + }, + { + "epoch": 0.08648, + "grad_norm": 0.8541324948901896, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 8648 + }, + { + "epoch": 0.08649, + "grad_norm": 0.952863199924152, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 8649 + }, + { + "epoch": 0.0865, + "grad_norm": 1.0020719863767489, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 8650 + }, + { + "epoch": 0.08651, + "grad_norm": 1.0541904737375236, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 8651 + }, + { + "epoch": 0.08652, + "grad_norm": 0.965588617649747, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8652 + }, + { + "epoch": 0.08653, + "grad_norm": 1.008009301735876, + "learning_rate": 0.003, + "loss": 4.083, + "step": 8653 + }, + { + "epoch": 0.08654, + "grad_norm": 1.0616740428111644, + "learning_rate": 0.003, + "loss": 4.099, + "step": 8654 + }, + { + "epoch": 0.08655, + "grad_norm": 0.9641551606343974, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 8655 + }, + { + "epoch": 0.08656, + "grad_norm": 0.92701904040581, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8656 + }, + { + "epoch": 0.08657, + "grad_norm": 0.9477211259009536, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 8657 + }, + { + "epoch": 0.08658, + "grad_norm": 0.8613906735799333, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8658 + }, + { + "epoch": 0.08659, + "grad_norm": 0.8582341328171613, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 8659 + }, + { + "epoch": 0.0866, + "grad_norm": 0.7554921765534043, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 8660 + }, + { + "epoch": 0.08661, + "grad_norm": 0.710138619445291, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 8661 + }, + { + "epoch": 0.08662, + "grad_norm": 0.6948787363147834, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 8662 + }, + { + "epoch": 0.08663, + "grad_norm": 0.8130674241859551, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 8663 + }, + { + "epoch": 0.08664, + "grad_norm": 0.8427386250145507, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8664 + }, + { + "epoch": 0.08665, + "grad_norm": 0.8151506821681499, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 8665 + }, + { + "epoch": 0.08666, + "grad_norm": 0.9564462744936523, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8666 + }, + { + "epoch": 0.08667, + "grad_norm": 1.1118303234126337, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 8667 + }, + { + "epoch": 0.08668, + "grad_norm": 0.9710629851216804, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 8668 + }, + { + "epoch": 0.08669, + "grad_norm": 1.5737744642483429, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8669 + }, + { + "epoch": 0.0867, + "grad_norm": 0.6740528009072599, + "learning_rate": 0.003, + "loss": 4.085, + "step": 8670 + }, + { + "epoch": 0.08671, + "grad_norm": 0.8368283275216355, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 8671 + }, + { + "epoch": 0.08672, + "grad_norm": 1.0436902189256585, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8672 + }, + { + "epoch": 0.08673, + "grad_norm": 1.0127357970798005, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8673 + }, + { + "epoch": 0.08674, + "grad_norm": 0.9615865521203145, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8674 + }, + { + "epoch": 0.08675, + "grad_norm": 1.0774654469343408, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 8675 + }, + { + "epoch": 0.08676, + "grad_norm": 1.0537181165753355, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 8676 + }, + { + "epoch": 0.08677, + "grad_norm": 1.0110231179636788, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 8677 + }, + { + "epoch": 0.08678, + "grad_norm": 0.9615912561390582, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 8678 + }, + { + "epoch": 0.08679, + "grad_norm": 0.9402198527708937, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8679 + }, + { + "epoch": 0.0868, + "grad_norm": 0.905661734228063, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 8680 + }, + { + "epoch": 0.08681, + "grad_norm": 1.1569378589826407, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 8681 + }, + { + "epoch": 0.08682, + "grad_norm": 0.9978572454734634, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 8682 + }, + { + "epoch": 0.08683, + "grad_norm": 0.9820424038569685, + "learning_rate": 0.003, + "loss": 4.116, + "step": 8683 + }, + { + "epoch": 0.08684, + "grad_norm": 0.9217316868751142, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 8684 + }, + { + "epoch": 0.08685, + "grad_norm": 1.13020305011687, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 8685 + }, + { + "epoch": 0.08686, + "grad_norm": 1.135596127784365, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 8686 + }, + { + "epoch": 0.08687, + "grad_norm": 0.8991895009637995, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 8687 + }, + { + "epoch": 0.08688, + "grad_norm": 0.7950591470094579, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8688 + }, + { + "epoch": 0.08689, + "grad_norm": 0.9175036427923986, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8689 + }, + { + "epoch": 0.0869, + "grad_norm": 1.1867185106176557, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 8690 + }, + { + "epoch": 0.08691, + "grad_norm": 1.150743996358858, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 8691 + }, + { + "epoch": 0.08692, + "grad_norm": 1.072456612250304, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8692 + }, + { + "epoch": 0.08693, + "grad_norm": 0.9820410634402862, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8693 + }, + { + "epoch": 0.08694, + "grad_norm": 0.996739398321439, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 8694 + }, + { + "epoch": 0.08695, + "grad_norm": 1.050712916361131, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8695 + }, + { + "epoch": 0.08696, + "grad_norm": 0.9400645560048955, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 8696 + }, + { + "epoch": 0.08697, + "grad_norm": 0.9974884973769946, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8697 + }, + { + "epoch": 0.08698, + "grad_norm": 1.0051997817192861, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8698 + }, + { + "epoch": 0.08699, + "grad_norm": 0.9163255483083583, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 8699 + }, + { + "epoch": 0.087, + "grad_norm": 1.006810901255363, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 8700 + }, + { + "epoch": 0.08701, + "grad_norm": 1.115045098588956, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 8701 + }, + { + "epoch": 0.08702, + "grad_norm": 0.9880723518291648, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 8702 + }, + { + "epoch": 0.08703, + "grad_norm": 0.8366949205399988, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 8703 + }, + { + "epoch": 0.08704, + "grad_norm": 0.7286032282861155, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 8704 + }, + { + "epoch": 0.08705, + "grad_norm": 0.818286090640339, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 8705 + }, + { + "epoch": 0.08706, + "grad_norm": 1.0020414264801158, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 8706 + }, + { + "epoch": 0.08707, + "grad_norm": 1.1058535805008598, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 8707 + }, + { + "epoch": 0.08708, + "grad_norm": 1.0637305918614082, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 8708 + }, + { + "epoch": 0.08709, + "grad_norm": 1.1622544435094628, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 8709 + }, + { + "epoch": 0.0871, + "grad_norm": 1.1107128106540358, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 8710 + }, + { + "epoch": 0.08711, + "grad_norm": 0.9646068373714444, + "learning_rate": 0.003, + "loss": 4.102, + "step": 8711 + }, + { + "epoch": 0.08712, + "grad_norm": 0.9931572732752244, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8712 + }, + { + "epoch": 0.08713, + "grad_norm": 0.9908778373110699, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 8713 + }, + { + "epoch": 0.08714, + "grad_norm": 1.0197596826627089, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 8714 + }, + { + "epoch": 0.08715, + "grad_norm": 1.014702738749466, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 8715 + }, + { + "epoch": 0.08716, + "grad_norm": 1.0045099245365388, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 8716 + }, + { + "epoch": 0.08717, + "grad_norm": 1.0710132715620833, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 8717 + }, + { + "epoch": 0.08718, + "grad_norm": 0.9699281274500614, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 8718 + }, + { + "epoch": 0.08719, + "grad_norm": 1.012536675784157, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 8719 + }, + { + "epoch": 0.0872, + "grad_norm": 0.9846464927862566, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 8720 + }, + { + "epoch": 0.08721, + "grad_norm": 1.051748323833279, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 8721 + }, + { + "epoch": 0.08722, + "grad_norm": 1.1390897577030044, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 8722 + }, + { + "epoch": 0.08723, + "grad_norm": 1.014378976687448, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 8723 + }, + { + "epoch": 0.08724, + "grad_norm": 1.0096210285347254, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 8724 + }, + { + "epoch": 0.08725, + "grad_norm": 0.9399244619167805, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 8725 + }, + { + "epoch": 0.08726, + "grad_norm": 0.9568485117772358, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 8726 + }, + { + "epoch": 0.08727, + "grad_norm": 0.9385590320694874, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 8727 + }, + { + "epoch": 0.08728, + "grad_norm": 0.9049360545768217, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 8728 + }, + { + "epoch": 0.08729, + "grad_norm": 0.9019422289302671, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8729 + }, + { + "epoch": 0.0873, + "grad_norm": 0.9330390334308929, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 8730 + }, + { + "epoch": 0.08731, + "grad_norm": 0.9809468283330093, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 8731 + }, + { + "epoch": 0.08732, + "grad_norm": 0.9010326157614988, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8732 + }, + { + "epoch": 0.08733, + "grad_norm": 0.7938742404264579, + "learning_rate": 0.003, + "loss": 4.081, + "step": 8733 + }, + { + "epoch": 0.08734, + "grad_norm": 0.8111233684875911, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8734 + }, + { + "epoch": 0.08735, + "grad_norm": 0.7272221907480916, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 8735 + }, + { + "epoch": 0.08736, + "grad_norm": 0.679447738423607, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 8736 + }, + { + "epoch": 0.08737, + "grad_norm": 0.8396545033444168, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 8737 + }, + { + "epoch": 0.08738, + "grad_norm": 1.1738244903309252, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 8738 + }, + { + "epoch": 0.08739, + "grad_norm": 1.0681589571108603, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 8739 + }, + { + "epoch": 0.0874, + "grad_norm": 1.05002028020519, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8740 + }, + { + "epoch": 0.08741, + "grad_norm": 0.9219236197609371, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 8741 + }, + { + "epoch": 0.08742, + "grad_norm": 1.000316386094372, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 8742 + }, + { + "epoch": 0.08743, + "grad_norm": 0.964186838515666, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 8743 + }, + { + "epoch": 0.08744, + "grad_norm": 0.8953102034049782, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 8744 + }, + { + "epoch": 0.08745, + "grad_norm": 1.0423780413564045, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8745 + }, + { + "epoch": 0.08746, + "grad_norm": 1.1156519078431824, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8746 + }, + { + "epoch": 0.08747, + "grad_norm": 0.9289740494359398, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8747 + }, + { + "epoch": 0.08748, + "grad_norm": 0.9221501347652119, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8748 + }, + { + "epoch": 0.08749, + "grad_norm": 1.137448507112466, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 8749 + }, + { + "epoch": 0.0875, + "grad_norm": 1.1606946555207784, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 8750 + }, + { + "epoch": 0.08751, + "grad_norm": 1.1418896933212783, + "learning_rate": 0.003, + "loss": 4.102, + "step": 8751 + }, + { + "epoch": 0.08752, + "grad_norm": 0.8892781422608295, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 8752 + }, + { + "epoch": 0.08753, + "grad_norm": 0.8235048326716268, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 8753 + }, + { + "epoch": 0.08754, + "grad_norm": 0.9178459509732138, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8754 + }, + { + "epoch": 0.08755, + "grad_norm": 1.043656892193055, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 8755 + }, + { + "epoch": 0.08756, + "grad_norm": 0.9634587521325285, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8756 + }, + { + "epoch": 0.08757, + "grad_norm": 1.0095694340504555, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 8757 + }, + { + "epoch": 0.08758, + "grad_norm": 1.0702638193359773, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 8758 + }, + { + "epoch": 0.08759, + "grad_norm": 1.1871597807200291, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8759 + }, + { + "epoch": 0.0876, + "grad_norm": 0.8734990186560198, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8760 + }, + { + "epoch": 0.08761, + "grad_norm": 0.9459790552922084, + "learning_rate": 0.003, + "loss": 4.125, + "step": 8761 + }, + { + "epoch": 0.08762, + "grad_norm": 1.170255371175725, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 8762 + }, + { + "epoch": 0.08763, + "grad_norm": 0.9807952883818362, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 8763 + }, + { + "epoch": 0.08764, + "grad_norm": 1.3716972394122473, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 8764 + }, + { + "epoch": 0.08765, + "grad_norm": 0.8981483803103426, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8765 + }, + { + "epoch": 0.08766, + "grad_norm": 0.9178267580818741, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 8766 + }, + { + "epoch": 0.08767, + "grad_norm": 0.8387751062720594, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 8767 + }, + { + "epoch": 0.08768, + "grad_norm": 0.765362163097001, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8768 + }, + { + "epoch": 0.08769, + "grad_norm": 0.8128170310066327, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 8769 + }, + { + "epoch": 0.0877, + "grad_norm": 0.9858951252865733, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 8770 + }, + { + "epoch": 0.08771, + "grad_norm": 1.1445554138270546, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8771 + }, + { + "epoch": 0.08772, + "grad_norm": 0.8621127850353041, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 8772 + }, + { + "epoch": 0.08773, + "grad_norm": 0.8525885089903674, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 8773 + }, + { + "epoch": 0.08774, + "grad_norm": 0.8871272119975309, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8774 + }, + { + "epoch": 0.08775, + "grad_norm": 0.991806274254639, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 8775 + }, + { + "epoch": 0.08776, + "grad_norm": 1.1533994303031723, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 8776 + }, + { + "epoch": 0.08777, + "grad_norm": 1.055322012206207, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 8777 + }, + { + "epoch": 0.08778, + "grad_norm": 1.071611255711352, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8778 + }, + { + "epoch": 0.08779, + "grad_norm": 0.9203284832312962, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 8779 + }, + { + "epoch": 0.0878, + "grad_norm": 0.8661451623284229, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 8780 + }, + { + "epoch": 0.08781, + "grad_norm": 0.9695475727064722, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 8781 + }, + { + "epoch": 0.08782, + "grad_norm": 0.9855488794840954, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 8782 + }, + { + "epoch": 0.08783, + "grad_norm": 1.0854676896438462, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 8783 + }, + { + "epoch": 0.08784, + "grad_norm": 0.92019280232576, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 8784 + }, + { + "epoch": 0.08785, + "grad_norm": 0.8916386749670293, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 8785 + }, + { + "epoch": 0.08786, + "grad_norm": 0.9384291792786592, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 8786 + }, + { + "epoch": 0.08787, + "grad_norm": 1.1242759626646945, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 8787 + }, + { + "epoch": 0.08788, + "grad_norm": 0.9566646203576128, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 8788 + }, + { + "epoch": 0.08789, + "grad_norm": 1.1892424954334855, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8789 + }, + { + "epoch": 0.0879, + "grad_norm": 0.9588075240707433, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 8790 + }, + { + "epoch": 0.08791, + "grad_norm": 1.06193895017149, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 8791 + }, + { + "epoch": 0.08792, + "grad_norm": 1.0293430845691305, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 8792 + }, + { + "epoch": 0.08793, + "grad_norm": 0.9914591638234613, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 8793 + }, + { + "epoch": 0.08794, + "grad_norm": 1.1989613224884543, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 8794 + }, + { + "epoch": 0.08795, + "grad_norm": 0.9861795106142962, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 8795 + }, + { + "epoch": 0.08796, + "grad_norm": 1.0041571125270279, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8796 + }, + { + "epoch": 0.08797, + "grad_norm": 1.0698702519124776, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 8797 + }, + { + "epoch": 0.08798, + "grad_norm": 1.1705625141993568, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 8798 + }, + { + "epoch": 0.08799, + "grad_norm": 0.7915192269652498, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8799 + }, + { + "epoch": 0.088, + "grad_norm": 0.6604270774270279, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 8800 + }, + { + "epoch": 0.08801, + "grad_norm": 0.7719368938463326, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 8801 + }, + { + "epoch": 0.08802, + "grad_norm": 0.8416958932367354, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 8802 + }, + { + "epoch": 0.08803, + "grad_norm": 1.0164740952605043, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 8803 + }, + { + "epoch": 0.08804, + "grad_norm": 1.1090228946685283, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 8804 + }, + { + "epoch": 0.08805, + "grad_norm": 0.9168056996303934, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 8805 + }, + { + "epoch": 0.08806, + "grad_norm": 0.9155273228148718, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 8806 + }, + { + "epoch": 0.08807, + "grad_norm": 0.8379245763047375, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 8807 + }, + { + "epoch": 0.08808, + "grad_norm": 0.8670250361401486, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 8808 + }, + { + "epoch": 0.08809, + "grad_norm": 1.0376022891758734, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 8809 + }, + { + "epoch": 0.0881, + "grad_norm": 1.065934829807962, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 8810 + }, + { + "epoch": 0.08811, + "grad_norm": 1.0298973690098348, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 8811 + }, + { + "epoch": 0.08812, + "grad_norm": 0.8686967291317386, + "learning_rate": 0.003, + "loss": 4.09, + "step": 8812 + }, + { + "epoch": 0.08813, + "grad_norm": 0.9680376583982786, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 8813 + }, + { + "epoch": 0.08814, + "grad_norm": 0.8851029994047319, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 8814 + }, + { + "epoch": 0.08815, + "grad_norm": 1.016504733957394, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 8815 + }, + { + "epoch": 0.08816, + "grad_norm": 1.039564849638083, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 8816 + }, + { + "epoch": 0.08817, + "grad_norm": 0.9769732491378709, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 8817 + }, + { + "epoch": 0.08818, + "grad_norm": 0.7419152032454192, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 8818 + }, + { + "epoch": 0.08819, + "grad_norm": 0.7449265064447682, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 8819 + }, + { + "epoch": 0.0882, + "grad_norm": 0.8134643932223908, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 8820 + }, + { + "epoch": 0.08821, + "grad_norm": 0.8600355770310976, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 8821 + }, + { + "epoch": 0.08822, + "grad_norm": 1.0780006611352508, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 8822 + }, + { + "epoch": 0.08823, + "grad_norm": 0.9907258089098394, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 8823 + }, + { + "epoch": 0.08824, + "grad_norm": 0.9262970676154467, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 8824 + }, + { + "epoch": 0.08825, + "grad_norm": 1.0361846461107758, + "learning_rate": 0.003, + "loss": 4.096, + "step": 8825 + }, + { + "epoch": 0.08826, + "grad_norm": 1.1352830686502071, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8826 + }, + { + "epoch": 0.08827, + "grad_norm": 0.9811751503381044, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 8827 + }, + { + "epoch": 0.08828, + "grad_norm": 1.1677797594356207, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 8828 + }, + { + "epoch": 0.08829, + "grad_norm": 0.8620817289548668, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 8829 + }, + { + "epoch": 0.0883, + "grad_norm": 0.9109610741986333, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8830 + }, + { + "epoch": 0.08831, + "grad_norm": 0.9792155464654494, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 8831 + }, + { + "epoch": 0.08832, + "grad_norm": 1.1813879043570434, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 8832 + }, + { + "epoch": 0.08833, + "grad_norm": 1.0032648409430165, + "learning_rate": 0.003, + "loss": 4.086, + "step": 8833 + }, + { + "epoch": 0.08834, + "grad_norm": 0.9769964414966447, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 8834 + }, + { + "epoch": 0.08835, + "grad_norm": 1.082737428043474, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 8835 + }, + { + "epoch": 0.08836, + "grad_norm": 0.984170203316877, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 8836 + }, + { + "epoch": 0.08837, + "grad_norm": 1.056134980376748, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8837 + }, + { + "epoch": 0.08838, + "grad_norm": 1.211754779715747, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 8838 + }, + { + "epoch": 0.08839, + "grad_norm": 0.9712974264137377, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 8839 + }, + { + "epoch": 0.0884, + "grad_norm": 1.1142216163765637, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 8840 + }, + { + "epoch": 0.08841, + "grad_norm": 1.0046432864292048, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 8841 + }, + { + "epoch": 0.08842, + "grad_norm": 1.056661438069542, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8842 + }, + { + "epoch": 0.08843, + "grad_norm": 1.2236595574756504, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 8843 + }, + { + "epoch": 0.08844, + "grad_norm": 1.03881071415002, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8844 + }, + { + "epoch": 0.08845, + "grad_norm": 1.1459947152275722, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 8845 + }, + { + "epoch": 0.08846, + "grad_norm": 1.0840904607136326, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 8846 + }, + { + "epoch": 0.08847, + "grad_norm": 0.9355054937743638, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 8847 + }, + { + "epoch": 0.08848, + "grad_norm": 0.9640890114169899, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 8848 + }, + { + "epoch": 0.08849, + "grad_norm": 1.0483305858052685, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 8849 + }, + { + "epoch": 0.0885, + "grad_norm": 1.0553477746801, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 8850 + }, + { + "epoch": 0.08851, + "grad_norm": 0.880408001509907, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 8851 + }, + { + "epoch": 0.08852, + "grad_norm": 0.8729293404205195, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8852 + }, + { + "epoch": 0.08853, + "grad_norm": 0.8295344564262668, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 8853 + }, + { + "epoch": 0.08854, + "grad_norm": 0.8793240956806683, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 8854 + }, + { + "epoch": 0.08855, + "grad_norm": 0.9216847733111198, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 8855 + }, + { + "epoch": 0.08856, + "grad_norm": 1.0905136756034943, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 8856 + }, + { + "epoch": 0.08857, + "grad_norm": 1.0871331725112339, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 8857 + }, + { + "epoch": 0.08858, + "grad_norm": 0.93110512066651, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 8858 + }, + { + "epoch": 0.08859, + "grad_norm": 0.9090747425713922, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 8859 + }, + { + "epoch": 0.0886, + "grad_norm": 0.9685128136141401, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 8860 + }, + { + "epoch": 0.08861, + "grad_norm": 0.9015137576718465, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 8861 + }, + { + "epoch": 0.08862, + "grad_norm": 1.0226210784829224, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 8862 + }, + { + "epoch": 0.08863, + "grad_norm": 1.2542964356410375, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 8863 + }, + { + "epoch": 0.08864, + "grad_norm": 0.763921705860707, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 8864 + }, + { + "epoch": 0.08865, + "grad_norm": 0.6665256002350811, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 8865 + }, + { + "epoch": 0.08866, + "grad_norm": 0.7496478357626905, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 8866 + }, + { + "epoch": 0.08867, + "grad_norm": 0.8460797471569417, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 8867 + }, + { + "epoch": 0.08868, + "grad_norm": 1.0223040210373233, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8868 + }, + { + "epoch": 0.08869, + "grad_norm": 1.1610482768655175, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 8869 + }, + { + "epoch": 0.0887, + "grad_norm": 0.9490101679373257, + "learning_rate": 0.003, + "loss": 4.074, + "step": 8870 + }, + { + "epoch": 0.08871, + "grad_norm": 1.0936328384500282, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 8871 + }, + { + "epoch": 0.08872, + "grad_norm": 1.0161161364379168, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 8872 + }, + { + "epoch": 0.08873, + "grad_norm": 1.092903916933006, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 8873 + }, + { + "epoch": 0.08874, + "grad_norm": 0.9262834148165788, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 8874 + }, + { + "epoch": 0.08875, + "grad_norm": 0.9453266208102923, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 8875 + }, + { + "epoch": 0.08876, + "grad_norm": 1.0239710759961396, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 8876 + }, + { + "epoch": 0.08877, + "grad_norm": 0.997162213443831, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 8877 + }, + { + "epoch": 0.08878, + "grad_norm": 1.078780262399322, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8878 + }, + { + "epoch": 0.08879, + "grad_norm": 1.1714017894712712, + "learning_rate": 0.003, + "loss": 4.088, + "step": 8879 + }, + { + "epoch": 0.0888, + "grad_norm": 1.0249330378361858, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 8880 + }, + { + "epoch": 0.08881, + "grad_norm": 1.0028152550391942, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 8881 + }, + { + "epoch": 0.08882, + "grad_norm": 1.1161280626520085, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 8882 + }, + { + "epoch": 0.08883, + "grad_norm": 0.9449109550267524, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 8883 + }, + { + "epoch": 0.08884, + "grad_norm": 0.9509758982206539, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 8884 + }, + { + "epoch": 0.08885, + "grad_norm": 1.0088259868099347, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 8885 + }, + { + "epoch": 0.08886, + "grad_norm": 1.0034750507067094, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 8886 + }, + { + "epoch": 0.08887, + "grad_norm": 1.1564868184220833, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8887 + }, + { + "epoch": 0.08888, + "grad_norm": 0.920047239707721, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8888 + }, + { + "epoch": 0.08889, + "grad_norm": 0.8744706857704945, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 8889 + }, + { + "epoch": 0.0889, + "grad_norm": 0.9285165665395756, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 8890 + }, + { + "epoch": 0.08891, + "grad_norm": 1.0099497011549061, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 8891 + }, + { + "epoch": 0.08892, + "grad_norm": 1.1488095492811832, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 8892 + }, + { + "epoch": 0.08893, + "grad_norm": 0.9091122428129211, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 8893 + }, + { + "epoch": 0.08894, + "grad_norm": 0.9429348565271667, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8894 + }, + { + "epoch": 0.08895, + "grad_norm": 1.008440730412949, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 8895 + }, + { + "epoch": 0.08896, + "grad_norm": 0.9039965851388357, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 8896 + }, + { + "epoch": 0.08897, + "grad_norm": 0.7170118026175338, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 8897 + }, + { + "epoch": 0.08898, + "grad_norm": 0.7903539062454961, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 8898 + }, + { + "epoch": 0.08899, + "grad_norm": 0.7245799759674336, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 8899 + }, + { + "epoch": 0.089, + "grad_norm": 0.6887895124684095, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 8900 + }, + { + "epoch": 0.08901, + "grad_norm": 0.6703863748825403, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 8901 + }, + { + "epoch": 0.08902, + "grad_norm": 0.6971104906184047, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 8902 + }, + { + "epoch": 0.08903, + "grad_norm": 0.8088806783931428, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 8903 + }, + { + "epoch": 0.08904, + "grad_norm": 1.0528694251570367, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 8904 + }, + { + "epoch": 0.08905, + "grad_norm": 1.333318141346917, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 8905 + }, + { + "epoch": 0.08906, + "grad_norm": 0.8796950703932015, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 8906 + }, + { + "epoch": 0.08907, + "grad_norm": 1.017711138999783, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 8907 + }, + { + "epoch": 0.08908, + "grad_norm": 1.2131433849659086, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 8908 + }, + { + "epoch": 0.08909, + "grad_norm": 0.910493543162808, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 8909 + }, + { + "epoch": 0.0891, + "grad_norm": 0.9813094815777755, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 8910 + }, + { + "epoch": 0.08911, + "grad_norm": 1.0796651179058687, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 8911 + }, + { + "epoch": 0.08912, + "grad_norm": 0.7841503223559056, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8912 + }, + { + "epoch": 0.08913, + "grad_norm": 0.8040837526844624, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 8913 + }, + { + "epoch": 0.08914, + "grad_norm": 0.9201111247555188, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 8914 + }, + { + "epoch": 0.08915, + "grad_norm": 1.2443487673269766, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 8915 + }, + { + "epoch": 0.08916, + "grad_norm": 0.9520842934826077, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 8916 + }, + { + "epoch": 0.08917, + "grad_norm": 0.8122356743140442, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 8917 + }, + { + "epoch": 0.08918, + "grad_norm": 0.849179194303752, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 8918 + }, + { + "epoch": 0.08919, + "grad_norm": 1.0835039742387556, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 8919 + }, + { + "epoch": 0.0892, + "grad_norm": 1.0060700478336813, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 8920 + }, + { + "epoch": 0.08921, + "grad_norm": 1.0073037824850515, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 8921 + }, + { + "epoch": 0.08922, + "grad_norm": 1.1074535105569765, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 8922 + }, + { + "epoch": 0.08923, + "grad_norm": 1.3376095505917398, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 8923 + }, + { + "epoch": 0.08924, + "grad_norm": 0.8791813131366455, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 8924 + }, + { + "epoch": 0.08925, + "grad_norm": 0.8009079909307454, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 8925 + }, + { + "epoch": 0.08926, + "grad_norm": 0.8350852915495754, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 8926 + }, + { + "epoch": 0.08927, + "grad_norm": 0.8752090800535209, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 8927 + }, + { + "epoch": 0.08928, + "grad_norm": 0.8879314625475655, + "learning_rate": 0.003, + "loss": 4.079, + "step": 8928 + }, + { + "epoch": 0.08929, + "grad_norm": 1.142327573748278, + "learning_rate": 0.003, + "loss": 4.091, + "step": 8929 + }, + { + "epoch": 0.0893, + "grad_norm": 1.090048369857334, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 8930 + }, + { + "epoch": 0.08931, + "grad_norm": 0.9435660692327841, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 8931 + }, + { + "epoch": 0.08932, + "grad_norm": 0.9782783037908521, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8932 + }, + { + "epoch": 0.08933, + "grad_norm": 1.0634692540653257, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 8933 + }, + { + "epoch": 0.08934, + "grad_norm": 1.0125558743673009, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 8934 + }, + { + "epoch": 0.08935, + "grad_norm": 0.9385240579578458, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 8935 + }, + { + "epoch": 0.08936, + "grad_norm": 0.9949684717325338, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 8936 + }, + { + "epoch": 0.08937, + "grad_norm": 1.0884204651036244, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8937 + }, + { + "epoch": 0.08938, + "grad_norm": 1.2595959013613358, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 8938 + }, + { + "epoch": 0.08939, + "grad_norm": 0.8718352236246899, + "learning_rate": 0.003, + "loss": 4.123, + "step": 8939 + }, + { + "epoch": 0.0894, + "grad_norm": 0.7377382096343973, + "learning_rate": 0.003, + "loss": 4.082, + "step": 8940 + }, + { + "epoch": 0.08941, + "grad_norm": 0.7815043419777372, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 8941 + }, + { + "epoch": 0.08942, + "grad_norm": 0.85637053050209, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8942 + }, + { + "epoch": 0.08943, + "grad_norm": 0.809694079364775, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 8943 + }, + { + "epoch": 0.08944, + "grad_norm": 0.8311218205671262, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 8944 + }, + { + "epoch": 0.08945, + "grad_norm": 0.9235692408794608, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8945 + }, + { + "epoch": 0.08946, + "grad_norm": 0.8988805187494495, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 8946 + }, + { + "epoch": 0.08947, + "grad_norm": 0.9750913166472722, + "learning_rate": 0.003, + "loss": 4.095, + "step": 8947 + }, + { + "epoch": 0.08948, + "grad_norm": 1.2358006562850736, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8948 + }, + { + "epoch": 0.08949, + "grad_norm": 1.0201308388850814, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 8949 + }, + { + "epoch": 0.0895, + "grad_norm": 1.3147224146265035, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 8950 + }, + { + "epoch": 0.08951, + "grad_norm": 0.850210561765072, + "learning_rate": 0.003, + "loss": 4.086, + "step": 8951 + }, + { + "epoch": 0.08952, + "grad_norm": 0.7409549525635194, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 8952 + }, + { + "epoch": 0.08953, + "grad_norm": 0.7712667656518064, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 8953 + }, + { + "epoch": 0.08954, + "grad_norm": 0.8389010542455372, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8954 + }, + { + "epoch": 0.08955, + "grad_norm": 1.0106297979397758, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 8955 + }, + { + "epoch": 0.08956, + "grad_norm": 1.1616116590632415, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 8956 + }, + { + "epoch": 0.08957, + "grad_norm": 0.9682455671909211, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 8957 + }, + { + "epoch": 0.08958, + "grad_norm": 0.9906512714609359, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 8958 + }, + { + "epoch": 0.08959, + "grad_norm": 1.0546522223319332, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 8959 + }, + { + "epoch": 0.0896, + "grad_norm": 1.1586648477198143, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 8960 + }, + { + "epoch": 0.08961, + "grad_norm": 1.0432105881764966, + "learning_rate": 0.003, + "loss": 4.08, + "step": 8961 + }, + { + "epoch": 0.08962, + "grad_norm": 1.0965840342756925, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 8962 + }, + { + "epoch": 0.08963, + "grad_norm": 1.0376163837487167, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 8963 + }, + { + "epoch": 0.08964, + "grad_norm": 1.0514679891633183, + "learning_rate": 0.003, + "loss": 4.098, + "step": 8964 + }, + { + "epoch": 0.08965, + "grad_norm": 1.0445039391438893, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 8965 + }, + { + "epoch": 0.08966, + "grad_norm": 1.0636480223637435, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 8966 + }, + { + "epoch": 0.08967, + "grad_norm": 0.8306886055128744, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 8967 + }, + { + "epoch": 0.08968, + "grad_norm": 1.0942130565636998, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 8968 + }, + { + "epoch": 0.08969, + "grad_norm": 1.0851056474023233, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 8969 + }, + { + "epoch": 0.0897, + "grad_norm": 1.0442994580535447, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 8970 + }, + { + "epoch": 0.08971, + "grad_norm": 1.0641712074518217, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 8971 + }, + { + "epoch": 0.08972, + "grad_norm": 0.8212488342541941, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 8972 + }, + { + "epoch": 0.08973, + "grad_norm": 0.9085039288779564, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 8973 + }, + { + "epoch": 0.08974, + "grad_norm": 1.05439509788842, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 8974 + }, + { + "epoch": 0.08975, + "grad_norm": 1.0378367702298532, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 8975 + }, + { + "epoch": 0.08976, + "grad_norm": 1.0587552419566737, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 8976 + }, + { + "epoch": 0.08977, + "grad_norm": 0.9882078968290161, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 8977 + }, + { + "epoch": 0.08978, + "grad_norm": 1.2634295327865903, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 8978 + }, + { + "epoch": 0.08979, + "grad_norm": 1.0868605949551717, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 8979 + }, + { + "epoch": 0.0898, + "grad_norm": 0.9024666538735595, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 8980 + }, + { + "epoch": 0.08981, + "grad_norm": 0.940643098285581, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 8981 + }, + { + "epoch": 0.08982, + "grad_norm": 1.0209299533352179, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 8982 + }, + { + "epoch": 0.08983, + "grad_norm": 1.1663864520745926, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 8983 + }, + { + "epoch": 0.08984, + "grad_norm": 1.0421843695021389, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 8984 + }, + { + "epoch": 0.08985, + "grad_norm": 0.9455839548731014, + "learning_rate": 0.003, + "loss": 4.1, + "step": 8985 + }, + { + "epoch": 0.08986, + "grad_norm": 1.0175793254211578, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 8986 + }, + { + "epoch": 0.08987, + "grad_norm": 1.0855539709461204, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 8987 + }, + { + "epoch": 0.08988, + "grad_norm": 1.2089023840888569, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 8988 + }, + { + "epoch": 0.08989, + "grad_norm": 0.9145464238654417, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 8989 + }, + { + "epoch": 0.0899, + "grad_norm": 0.9988559782011223, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 8990 + }, + { + "epoch": 0.08991, + "grad_norm": 0.9616640855485232, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 8991 + }, + { + "epoch": 0.08992, + "grad_norm": 1.0543039692078464, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 8992 + }, + { + "epoch": 0.08993, + "grad_norm": 0.952039320193524, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 8993 + }, + { + "epoch": 0.08994, + "grad_norm": 1.24776754345023, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 8994 + }, + { + "epoch": 0.08995, + "grad_norm": 0.8960768153711746, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 8995 + }, + { + "epoch": 0.08996, + "grad_norm": 0.8299433831762516, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 8996 + }, + { + "epoch": 0.08997, + "grad_norm": 0.8434475562979065, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 8997 + }, + { + "epoch": 0.08998, + "grad_norm": 0.9434883081551227, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 8998 + }, + { + "epoch": 0.08999, + "grad_norm": 1.028623116722872, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 8999 + }, + { + "epoch": 0.09, + "grad_norm": 1.1483053060085568, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9000 + }, + { + "epoch": 0.09001, + "grad_norm": 1.0462603971030422, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 9001 + }, + { + "epoch": 0.09002, + "grad_norm": 0.8866854703633072, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 9002 + }, + { + "epoch": 0.09003, + "grad_norm": 0.8897374118868974, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9003 + }, + { + "epoch": 0.09004, + "grad_norm": 1.0318758238899706, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9004 + }, + { + "epoch": 0.09005, + "grad_norm": 1.140885873667334, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 9005 + }, + { + "epoch": 0.09006, + "grad_norm": 0.891376466816432, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9006 + }, + { + "epoch": 0.09007, + "grad_norm": 1.1603687488320678, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 9007 + }, + { + "epoch": 0.09008, + "grad_norm": 0.8775142494741145, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9008 + }, + { + "epoch": 0.09009, + "grad_norm": 0.9654621607739925, + "learning_rate": 0.003, + "loss": 4.101, + "step": 9009 + }, + { + "epoch": 0.0901, + "grad_norm": 1.0294455899118256, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 9010 + }, + { + "epoch": 0.09011, + "grad_norm": 1.1050991894320772, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9011 + }, + { + "epoch": 0.09012, + "grad_norm": 1.1543143215597398, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 9012 + }, + { + "epoch": 0.09013, + "grad_norm": 0.8514403902861597, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 9013 + }, + { + "epoch": 0.09014, + "grad_norm": 0.869700266149385, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 9014 + }, + { + "epoch": 0.09015, + "grad_norm": 0.8302072302834347, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 9015 + }, + { + "epoch": 0.09016, + "grad_norm": 0.7449962289937087, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 9016 + }, + { + "epoch": 0.09017, + "grad_norm": 0.8441019310672158, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 9017 + }, + { + "epoch": 0.09018, + "grad_norm": 0.9224246439163316, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9018 + }, + { + "epoch": 0.09019, + "grad_norm": 1.2506310431634982, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 9019 + }, + { + "epoch": 0.0902, + "grad_norm": 1.0319725156855537, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 9020 + }, + { + "epoch": 0.09021, + "grad_norm": 1.0244238730812105, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 9021 + }, + { + "epoch": 0.09022, + "grad_norm": 1.0747193510062496, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9022 + }, + { + "epoch": 0.09023, + "grad_norm": 1.0829551491615659, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9023 + }, + { + "epoch": 0.09024, + "grad_norm": 0.8854265172004326, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 9024 + }, + { + "epoch": 0.09025, + "grad_norm": 0.9624141531169683, + "learning_rate": 0.003, + "loss": 4.095, + "step": 9025 + }, + { + "epoch": 0.09026, + "grad_norm": 0.9124575556477138, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 9026 + }, + { + "epoch": 0.09027, + "grad_norm": 0.7994369764537259, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9027 + }, + { + "epoch": 0.09028, + "grad_norm": 0.9507899093974508, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 9028 + }, + { + "epoch": 0.09029, + "grad_norm": 1.0373894988141181, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 9029 + }, + { + "epoch": 0.0903, + "grad_norm": 1.0795383394678428, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 9030 + }, + { + "epoch": 0.09031, + "grad_norm": 0.9371901971647943, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 9031 + }, + { + "epoch": 0.09032, + "grad_norm": 1.3118743983631005, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 9032 + }, + { + "epoch": 0.09033, + "grad_norm": 0.9602498029783673, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 9033 + }, + { + "epoch": 0.09034, + "grad_norm": 0.9950626958271721, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 9034 + }, + { + "epoch": 0.09035, + "grad_norm": 1.0561055356455764, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 9035 + }, + { + "epoch": 0.09036, + "grad_norm": 0.9351410832930079, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 9036 + }, + { + "epoch": 0.09037, + "grad_norm": 0.9375448148491637, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 9037 + }, + { + "epoch": 0.09038, + "grad_norm": 0.8772536980898831, + "learning_rate": 0.003, + "loss": 4.087, + "step": 9038 + }, + { + "epoch": 0.09039, + "grad_norm": 0.8994311716413951, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 9039 + }, + { + "epoch": 0.0904, + "grad_norm": 1.010507879277424, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9040 + }, + { + "epoch": 0.09041, + "grad_norm": 1.2199604338083574, + "learning_rate": 0.003, + "loss": 4.075, + "step": 9041 + }, + { + "epoch": 0.09042, + "grad_norm": 0.9306883379123116, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 9042 + }, + { + "epoch": 0.09043, + "grad_norm": 1.0192706712388095, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9043 + }, + { + "epoch": 0.09044, + "grad_norm": 1.0720290832366242, + "learning_rate": 0.003, + "loss": 4.082, + "step": 9044 + }, + { + "epoch": 0.09045, + "grad_norm": 0.9851527301184416, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 9045 + }, + { + "epoch": 0.09046, + "grad_norm": 1.0610574818669418, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9046 + }, + { + "epoch": 0.09047, + "grad_norm": 0.9500452554503603, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 9047 + }, + { + "epoch": 0.09048, + "grad_norm": 0.9313318824748757, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 9048 + }, + { + "epoch": 0.09049, + "grad_norm": 1.2054793227671492, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 9049 + }, + { + "epoch": 0.0905, + "grad_norm": 1.00585427743775, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9050 + }, + { + "epoch": 0.09051, + "grad_norm": 1.1034780299931397, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9051 + }, + { + "epoch": 0.09052, + "grad_norm": 1.0339569510475137, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 9052 + }, + { + "epoch": 0.09053, + "grad_norm": 0.9291730063765455, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 9053 + }, + { + "epoch": 0.09054, + "grad_norm": 0.7598831863125971, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9054 + }, + { + "epoch": 0.09055, + "grad_norm": 0.8785025929149661, + "learning_rate": 0.003, + "loss": 4.068, + "step": 9055 + }, + { + "epoch": 0.09056, + "grad_norm": 1.0578467685619801, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 9056 + }, + { + "epoch": 0.09057, + "grad_norm": 1.1700566880339585, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 9057 + }, + { + "epoch": 0.09058, + "grad_norm": 0.8486453469937698, + "learning_rate": 0.003, + "loss": 4.109, + "step": 9058 + }, + { + "epoch": 0.09059, + "grad_norm": 1.0260084269697587, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9059 + }, + { + "epoch": 0.0906, + "grad_norm": 1.0815529657850544, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 9060 + }, + { + "epoch": 0.09061, + "grad_norm": 0.9702489090387596, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 9061 + }, + { + "epoch": 0.09062, + "grad_norm": 1.0828394503861276, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 9062 + }, + { + "epoch": 0.09063, + "grad_norm": 0.9914403527753315, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 9063 + }, + { + "epoch": 0.09064, + "grad_norm": 1.0178247711714878, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 9064 + }, + { + "epoch": 0.09065, + "grad_norm": 1.0639916710421684, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 9065 + }, + { + "epoch": 0.09066, + "grad_norm": 0.9666784724080716, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 9066 + }, + { + "epoch": 0.09067, + "grad_norm": 1.1127616550466115, + "learning_rate": 0.003, + "loss": 4.089, + "step": 9067 + }, + { + "epoch": 0.09068, + "grad_norm": 0.9649238134507271, + "learning_rate": 0.003, + "loss": 4.098, + "step": 9068 + }, + { + "epoch": 0.09069, + "grad_norm": 0.9887864245443914, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 9069 + }, + { + "epoch": 0.0907, + "grad_norm": 1.138533659439355, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 9070 + }, + { + "epoch": 0.09071, + "grad_norm": 0.9538235049365951, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 9071 + }, + { + "epoch": 0.09072, + "grad_norm": 0.9399245548306119, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 9072 + }, + { + "epoch": 0.09073, + "grad_norm": 1.0410231512210018, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9073 + }, + { + "epoch": 0.09074, + "grad_norm": 0.9825288806775234, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9074 + }, + { + "epoch": 0.09075, + "grad_norm": 1.2536880541064686, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 9075 + }, + { + "epoch": 0.09076, + "grad_norm": 1.0802580187439756, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 9076 + }, + { + "epoch": 0.09077, + "grad_norm": 1.0000547727846416, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 9077 + }, + { + "epoch": 0.09078, + "grad_norm": 1.1810776381316752, + "learning_rate": 0.003, + "loss": 4.1, + "step": 9078 + }, + { + "epoch": 0.09079, + "grad_norm": 0.8675373011852131, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9079 + }, + { + "epoch": 0.0908, + "grad_norm": 0.8189772083275174, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9080 + }, + { + "epoch": 0.09081, + "grad_norm": 1.053936858891442, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 9081 + }, + { + "epoch": 0.09082, + "grad_norm": 1.1280890556667789, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9082 + }, + { + "epoch": 0.09083, + "grad_norm": 0.8002462051831681, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9083 + }, + { + "epoch": 0.09084, + "grad_norm": 0.7580618745930298, + "learning_rate": 0.003, + "loss": 4.068, + "step": 9084 + }, + { + "epoch": 0.09085, + "grad_norm": 0.8404228487181117, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 9085 + }, + { + "epoch": 0.09086, + "grad_norm": 0.9092656319720175, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 9086 + }, + { + "epoch": 0.09087, + "grad_norm": 1.1575206293331635, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 9087 + }, + { + "epoch": 0.09088, + "grad_norm": 1.1392497988218204, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 9088 + }, + { + "epoch": 0.09089, + "grad_norm": 1.0999951974096032, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9089 + }, + { + "epoch": 0.0909, + "grad_norm": 1.2058860154470423, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9090 + }, + { + "epoch": 0.09091, + "grad_norm": 0.8869769785320447, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 9091 + }, + { + "epoch": 0.09092, + "grad_norm": 0.968833257593624, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 9092 + }, + { + "epoch": 0.09093, + "grad_norm": 1.030403691084007, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9093 + }, + { + "epoch": 0.09094, + "grad_norm": 1.1575120263478036, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 9094 + }, + { + "epoch": 0.09095, + "grad_norm": 1.0658493375061597, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9095 + }, + { + "epoch": 0.09096, + "grad_norm": 1.039522944385023, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 9096 + }, + { + "epoch": 0.09097, + "grad_norm": 0.9449746348334745, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 9097 + }, + { + "epoch": 0.09098, + "grad_norm": 1.112543228949063, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9098 + }, + { + "epoch": 0.09099, + "grad_norm": 1.0356029258565156, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 9099 + }, + { + "epoch": 0.091, + "grad_norm": 0.9597566649353041, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 9100 + }, + { + "epoch": 0.09101, + "grad_norm": 0.909685008398296, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 9101 + }, + { + "epoch": 0.09102, + "grad_norm": 0.9939609422178486, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 9102 + }, + { + "epoch": 0.09103, + "grad_norm": 0.9981967075843049, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9103 + }, + { + "epoch": 0.09104, + "grad_norm": 0.901139384150308, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 9104 + }, + { + "epoch": 0.09105, + "grad_norm": 1.1303987033121574, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 9105 + }, + { + "epoch": 0.09106, + "grad_norm": 0.9605403802740986, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9106 + }, + { + "epoch": 0.09107, + "grad_norm": 1.0896925352972906, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9107 + }, + { + "epoch": 0.09108, + "grad_norm": 0.9390438282386121, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 9108 + }, + { + "epoch": 0.09109, + "grad_norm": 0.9831731988115366, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 9109 + }, + { + "epoch": 0.0911, + "grad_norm": 1.0442414761632293, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 9110 + }, + { + "epoch": 0.09111, + "grad_norm": 0.9165743816131849, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 9111 + }, + { + "epoch": 0.09112, + "grad_norm": 0.8395737037341123, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 9112 + }, + { + "epoch": 0.09113, + "grad_norm": 0.9188507446416299, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9113 + }, + { + "epoch": 0.09114, + "grad_norm": 1.192161128289549, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 9114 + }, + { + "epoch": 0.09115, + "grad_norm": 0.7878424986532916, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9115 + }, + { + "epoch": 0.09116, + "grad_norm": 0.756411216963468, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 9116 + }, + { + "epoch": 0.09117, + "grad_norm": 0.8675016418146388, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 9117 + }, + { + "epoch": 0.09118, + "grad_norm": 1.0711225834740048, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9118 + }, + { + "epoch": 0.09119, + "grad_norm": 1.3493838045641182, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9119 + }, + { + "epoch": 0.0912, + "grad_norm": 0.9734951792402872, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 9120 + }, + { + "epoch": 0.09121, + "grad_norm": 0.9595452425717873, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 9121 + }, + { + "epoch": 0.09122, + "grad_norm": 0.8608362956129894, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 9122 + }, + { + "epoch": 0.09123, + "grad_norm": 0.9267155554468813, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 9123 + }, + { + "epoch": 0.09124, + "grad_norm": 1.1438342201379448, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9124 + }, + { + "epoch": 0.09125, + "grad_norm": 0.912337828151199, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 9125 + }, + { + "epoch": 0.09126, + "grad_norm": 1.0294428453272435, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9126 + }, + { + "epoch": 0.09127, + "grad_norm": 0.9634503152628049, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9127 + }, + { + "epoch": 0.09128, + "grad_norm": 0.9285683958788655, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 9128 + }, + { + "epoch": 0.09129, + "grad_norm": 1.1437773291811142, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 9129 + }, + { + "epoch": 0.0913, + "grad_norm": 1.1492290125815292, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 9130 + }, + { + "epoch": 0.09131, + "grad_norm": 0.8288575691135607, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 9131 + }, + { + "epoch": 0.09132, + "grad_norm": 0.6963154193725737, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 9132 + }, + { + "epoch": 0.09133, + "grad_norm": 0.8615993381306877, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 9133 + }, + { + "epoch": 0.09134, + "grad_norm": 0.9872135875308844, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 9134 + }, + { + "epoch": 0.09135, + "grad_norm": 1.2143403531062356, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 9135 + }, + { + "epoch": 0.09136, + "grad_norm": 0.8677902987991879, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 9136 + }, + { + "epoch": 0.09137, + "grad_norm": 0.9650359637584537, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 9137 + }, + { + "epoch": 0.09138, + "grad_norm": 0.9888560898200366, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9138 + }, + { + "epoch": 0.09139, + "grad_norm": 0.9445593852088465, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9139 + }, + { + "epoch": 0.0914, + "grad_norm": 1.1631390018706205, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9140 + }, + { + "epoch": 0.09141, + "grad_norm": 0.9733863748246883, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 9141 + }, + { + "epoch": 0.09142, + "grad_norm": 1.144068894332549, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 9142 + }, + { + "epoch": 0.09143, + "grad_norm": 1.1162307413315125, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9143 + }, + { + "epoch": 0.09144, + "grad_norm": 0.9344229915073872, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 9144 + }, + { + "epoch": 0.09145, + "grad_norm": 0.934174583034849, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 9145 + }, + { + "epoch": 0.09146, + "grad_norm": 1.1330945945510411, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 9146 + }, + { + "epoch": 0.09147, + "grad_norm": 1.0873009118860704, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 9147 + }, + { + "epoch": 0.09148, + "grad_norm": 1.063907115512111, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 9148 + }, + { + "epoch": 0.09149, + "grad_norm": 0.9492654111198591, + "learning_rate": 0.003, + "loss": 4.053, + "step": 9149 + }, + { + "epoch": 0.0915, + "grad_norm": 0.8348158073927575, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9150 + }, + { + "epoch": 0.09151, + "grad_norm": 0.7907779988545944, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9151 + }, + { + "epoch": 0.09152, + "grad_norm": 0.820521579820349, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 9152 + }, + { + "epoch": 0.09153, + "grad_norm": 1.1762557932817632, + "learning_rate": 0.003, + "loss": 4.098, + "step": 9153 + }, + { + "epoch": 0.09154, + "grad_norm": 1.172758121597814, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9154 + }, + { + "epoch": 0.09155, + "grad_norm": 0.7806493977021605, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 9155 + }, + { + "epoch": 0.09156, + "grad_norm": 0.7334709809736035, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9156 + }, + { + "epoch": 0.09157, + "grad_norm": 0.7096998833904402, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 9157 + }, + { + "epoch": 0.09158, + "grad_norm": 0.6552858991619609, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 9158 + }, + { + "epoch": 0.09159, + "grad_norm": 0.6654141207947328, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 9159 + }, + { + "epoch": 0.0916, + "grad_norm": 0.7659679455492391, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 9160 + }, + { + "epoch": 0.09161, + "grad_norm": 0.8871298111496974, + "learning_rate": 0.003, + "loss": 4.069, + "step": 9161 + }, + { + "epoch": 0.09162, + "grad_norm": 1.0636906000745443, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 9162 + }, + { + "epoch": 0.09163, + "grad_norm": 0.9990259008300131, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 9163 + }, + { + "epoch": 0.09164, + "grad_norm": 0.9987224012321043, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 9164 + }, + { + "epoch": 0.09165, + "grad_norm": 1.0688733081179889, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 9165 + }, + { + "epoch": 0.09166, + "grad_norm": 0.9930601337362358, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9166 + }, + { + "epoch": 0.09167, + "grad_norm": 1.0601290119925093, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9167 + }, + { + "epoch": 0.09168, + "grad_norm": 1.1036579498016206, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 9168 + }, + { + "epoch": 0.09169, + "grad_norm": 1.101183482537871, + "learning_rate": 0.003, + "loss": 4.099, + "step": 9169 + }, + { + "epoch": 0.0917, + "grad_norm": 1.0752491199658576, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 9170 + }, + { + "epoch": 0.09171, + "grad_norm": 0.8939745974652651, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 9171 + }, + { + "epoch": 0.09172, + "grad_norm": 1.0352574625773354, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 9172 + }, + { + "epoch": 0.09173, + "grad_norm": 1.1944014003973107, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 9173 + }, + { + "epoch": 0.09174, + "grad_norm": 0.8767007669481375, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 9174 + }, + { + "epoch": 0.09175, + "grad_norm": 0.8817169634099795, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 9175 + }, + { + "epoch": 0.09176, + "grad_norm": 1.0813839509578795, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 9176 + }, + { + "epoch": 0.09177, + "grad_norm": 1.1959325088512973, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 9177 + }, + { + "epoch": 0.09178, + "grad_norm": 0.9670692426409625, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 9178 + }, + { + "epoch": 0.09179, + "grad_norm": 1.0878516037257497, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 9179 + }, + { + "epoch": 0.0918, + "grad_norm": 1.032818751415936, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 9180 + }, + { + "epoch": 0.09181, + "grad_norm": 1.1841768292302939, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 9181 + }, + { + "epoch": 0.09182, + "grad_norm": 0.8774737598295932, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 9182 + }, + { + "epoch": 0.09183, + "grad_norm": 0.9051881615495581, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9183 + }, + { + "epoch": 0.09184, + "grad_norm": 0.968743592106802, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 9184 + }, + { + "epoch": 0.09185, + "grad_norm": 1.0561355980332867, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 9185 + }, + { + "epoch": 0.09186, + "grad_norm": 1.007362715995797, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 9186 + }, + { + "epoch": 0.09187, + "grad_norm": 1.0062815389325381, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 9187 + }, + { + "epoch": 0.09188, + "grad_norm": 1.2814825677322759, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 9188 + }, + { + "epoch": 0.09189, + "grad_norm": 0.9675032030868871, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 9189 + }, + { + "epoch": 0.0919, + "grad_norm": 1.1111999953716745, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 9190 + }, + { + "epoch": 0.09191, + "grad_norm": 1.1548515391681171, + "learning_rate": 0.003, + "loss": 4.097, + "step": 9191 + }, + { + "epoch": 0.09192, + "grad_norm": 0.8490837287940421, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 9192 + }, + { + "epoch": 0.09193, + "grad_norm": 0.9222479120925116, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9193 + }, + { + "epoch": 0.09194, + "grad_norm": 0.9818940470607805, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 9194 + }, + { + "epoch": 0.09195, + "grad_norm": 1.0614643603489413, + "learning_rate": 0.003, + "loss": 4.091, + "step": 9195 + }, + { + "epoch": 0.09196, + "grad_norm": 1.1294918164841934, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 9196 + }, + { + "epoch": 0.09197, + "grad_norm": 1.0156458829127029, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9197 + }, + { + "epoch": 0.09198, + "grad_norm": 1.0305620102728879, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9198 + }, + { + "epoch": 0.09199, + "grad_norm": 0.9258303899781921, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9199 + }, + { + "epoch": 0.092, + "grad_norm": 1.052722229987556, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 9200 + }, + { + "epoch": 0.09201, + "grad_norm": 1.1423176862789393, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9201 + }, + { + "epoch": 0.09202, + "grad_norm": 0.90405167273345, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9202 + }, + { + "epoch": 0.09203, + "grad_norm": 0.8390172177360099, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 9203 + }, + { + "epoch": 0.09204, + "grad_norm": 1.0212772428161263, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 9204 + }, + { + "epoch": 0.09205, + "grad_norm": 1.3677660056338998, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 9205 + }, + { + "epoch": 0.09206, + "grad_norm": 0.8123508684473838, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 9206 + }, + { + "epoch": 0.09207, + "grad_norm": 0.7671163096769845, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9207 + }, + { + "epoch": 0.09208, + "grad_norm": 0.7996251518237344, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 9208 + }, + { + "epoch": 0.09209, + "grad_norm": 0.8386172650640057, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 9209 + }, + { + "epoch": 0.0921, + "grad_norm": 0.874525705380657, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9210 + }, + { + "epoch": 0.09211, + "grad_norm": 1.0250102800179441, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9211 + }, + { + "epoch": 0.09212, + "grad_norm": 1.252928643973771, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 9212 + }, + { + "epoch": 0.09213, + "grad_norm": 0.8470741103659993, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 9213 + }, + { + "epoch": 0.09214, + "grad_norm": 0.9021459877837885, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9214 + }, + { + "epoch": 0.09215, + "grad_norm": 1.089124576248861, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 9215 + }, + { + "epoch": 0.09216, + "grad_norm": 1.1489211291991535, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 9216 + }, + { + "epoch": 0.09217, + "grad_norm": 1.0292749044879108, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 9217 + }, + { + "epoch": 0.09218, + "grad_norm": 1.1023668228911894, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 9218 + }, + { + "epoch": 0.09219, + "grad_norm": 0.9991820142121032, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 9219 + }, + { + "epoch": 0.0922, + "grad_norm": 1.0771443271323655, + "learning_rate": 0.003, + "loss": 4.097, + "step": 9220 + }, + { + "epoch": 0.09221, + "grad_norm": 0.8927077451076355, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9221 + }, + { + "epoch": 0.09222, + "grad_norm": 0.921168419778457, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 9222 + }, + { + "epoch": 0.09223, + "grad_norm": 1.004730682614683, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9223 + }, + { + "epoch": 0.09224, + "grad_norm": 1.1490662225762478, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9224 + }, + { + "epoch": 0.09225, + "grad_norm": 0.8467914181281023, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9225 + }, + { + "epoch": 0.09226, + "grad_norm": 0.7997557193738468, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9226 + }, + { + "epoch": 0.09227, + "grad_norm": 0.8425166880932224, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9227 + }, + { + "epoch": 0.09228, + "grad_norm": 1.0077895082604096, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 9228 + }, + { + "epoch": 0.09229, + "grad_norm": 1.032289932362043, + "learning_rate": 0.003, + "loss": 4.094, + "step": 9229 + }, + { + "epoch": 0.0923, + "grad_norm": 1.2099726093934402, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 9230 + }, + { + "epoch": 0.09231, + "grad_norm": 0.9700410140732301, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 9231 + }, + { + "epoch": 0.09232, + "grad_norm": 1.145125305623255, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9232 + }, + { + "epoch": 0.09233, + "grad_norm": 0.9857080652159649, + "learning_rate": 0.003, + "loss": 4.112, + "step": 9233 + }, + { + "epoch": 0.09234, + "grad_norm": 0.853541587899831, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 9234 + }, + { + "epoch": 0.09235, + "grad_norm": 0.7853583653755962, + "learning_rate": 0.003, + "loss": 4.074, + "step": 9235 + }, + { + "epoch": 0.09236, + "grad_norm": 0.8600641615155024, + "learning_rate": 0.003, + "loss": 4.119, + "step": 9236 + }, + { + "epoch": 0.09237, + "grad_norm": 0.8802995694466931, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9237 + }, + { + "epoch": 0.09238, + "grad_norm": 0.889473257279724, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9238 + }, + { + "epoch": 0.09239, + "grad_norm": 0.965025798833757, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 9239 + }, + { + "epoch": 0.0924, + "grad_norm": 1.045798773841278, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9240 + }, + { + "epoch": 0.09241, + "grad_norm": 0.9585578674271291, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 9241 + }, + { + "epoch": 0.09242, + "grad_norm": 0.9326373066675198, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9242 + }, + { + "epoch": 0.09243, + "grad_norm": 1.1674477646861567, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 9243 + }, + { + "epoch": 0.09244, + "grad_norm": 1.111314854607193, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 9244 + }, + { + "epoch": 0.09245, + "grad_norm": 0.8498701271156277, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 9245 + }, + { + "epoch": 0.09246, + "grad_norm": 0.8355516969287872, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 9246 + }, + { + "epoch": 0.09247, + "grad_norm": 0.94621586277171, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 9247 + }, + { + "epoch": 0.09248, + "grad_norm": 1.2170065866247142, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9248 + }, + { + "epoch": 0.09249, + "grad_norm": 0.9279621491551032, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 9249 + }, + { + "epoch": 0.0925, + "grad_norm": 1.1809787158464518, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 9250 + }, + { + "epoch": 0.09251, + "grad_norm": 0.9888992006660906, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9251 + }, + { + "epoch": 0.09252, + "grad_norm": 0.9847921065738907, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9252 + }, + { + "epoch": 0.09253, + "grad_norm": 1.124411687709022, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9253 + }, + { + "epoch": 0.09254, + "grad_norm": 0.9386679683697067, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 9254 + }, + { + "epoch": 0.09255, + "grad_norm": 0.8865776906497654, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 9255 + }, + { + "epoch": 0.09256, + "grad_norm": 1.04809466760138, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9256 + }, + { + "epoch": 0.09257, + "grad_norm": 1.244274217288142, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 9257 + }, + { + "epoch": 0.09258, + "grad_norm": 0.8510022944180222, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 9258 + }, + { + "epoch": 0.09259, + "grad_norm": 0.9598837461861952, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 9259 + }, + { + "epoch": 0.0926, + "grad_norm": 0.9427395434917412, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 9260 + }, + { + "epoch": 0.09261, + "grad_norm": 0.8869685161322585, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 9261 + }, + { + "epoch": 0.09262, + "grad_norm": 0.8758797089936894, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9262 + }, + { + "epoch": 0.09263, + "grad_norm": 1.1037878707078868, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 9263 + }, + { + "epoch": 0.09264, + "grad_norm": 1.347143441300766, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 9264 + }, + { + "epoch": 0.09265, + "grad_norm": 0.8240409123116929, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 9265 + }, + { + "epoch": 0.09266, + "grad_norm": 0.9131281595783721, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 9266 + }, + { + "epoch": 0.09267, + "grad_norm": 1.0156568390080056, + "learning_rate": 0.003, + "loss": 4.075, + "step": 9267 + }, + { + "epoch": 0.09268, + "grad_norm": 1.2484963700197804, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 9268 + }, + { + "epoch": 0.09269, + "grad_norm": 1.0260778636305667, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 9269 + }, + { + "epoch": 0.0927, + "grad_norm": 1.125264451825767, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 9270 + }, + { + "epoch": 0.09271, + "grad_norm": 0.9337596088484723, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 9271 + }, + { + "epoch": 0.09272, + "grad_norm": 0.9755755501712275, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 9272 + }, + { + "epoch": 0.09273, + "grad_norm": 0.9365772539075061, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 9273 + }, + { + "epoch": 0.09274, + "grad_norm": 1.056284411019321, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 9274 + }, + { + "epoch": 0.09275, + "grad_norm": 1.0489238634825415, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9275 + }, + { + "epoch": 0.09276, + "grad_norm": 0.9904935266729148, + "learning_rate": 0.003, + "loss": 4.099, + "step": 9276 + }, + { + "epoch": 0.09277, + "grad_norm": 0.98765321752019, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 9277 + }, + { + "epoch": 0.09278, + "grad_norm": 1.0672587725016252, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 9278 + }, + { + "epoch": 0.09279, + "grad_norm": 1.0022153644343905, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 9279 + }, + { + "epoch": 0.0928, + "grad_norm": 1.114357821960331, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 9280 + }, + { + "epoch": 0.09281, + "grad_norm": 0.9933886219214135, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 9281 + }, + { + "epoch": 0.09282, + "grad_norm": 1.2306301044893613, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 9282 + }, + { + "epoch": 0.09283, + "grad_norm": 0.894007821463445, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 9283 + }, + { + "epoch": 0.09284, + "grad_norm": 0.995785402749101, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 9284 + }, + { + "epoch": 0.09285, + "grad_norm": 1.36921097660269, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9285 + }, + { + "epoch": 0.09286, + "grad_norm": 0.9689152429230696, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 9286 + }, + { + "epoch": 0.09287, + "grad_norm": 0.9373563728794481, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 9287 + }, + { + "epoch": 0.09288, + "grad_norm": 1.0159290553086127, + "learning_rate": 0.003, + "loss": 4.109, + "step": 9288 + }, + { + "epoch": 0.09289, + "grad_norm": 0.9130575266656131, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 9289 + }, + { + "epoch": 0.0929, + "grad_norm": 0.8501835536575353, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9290 + }, + { + "epoch": 0.09291, + "grad_norm": 0.8717656610305028, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9291 + }, + { + "epoch": 0.09292, + "grad_norm": 1.1198959017363963, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 9292 + }, + { + "epoch": 0.09293, + "grad_norm": 0.9126173242568293, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 9293 + }, + { + "epoch": 0.09294, + "grad_norm": 1.0289742761655396, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 9294 + }, + { + "epoch": 0.09295, + "grad_norm": 1.3069141395458677, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 9295 + }, + { + "epoch": 0.09296, + "grad_norm": 1.032110304900683, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 9296 + }, + { + "epoch": 0.09297, + "grad_norm": 1.0647117479317476, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 9297 + }, + { + "epoch": 0.09298, + "grad_norm": 1.1600003420191138, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 9298 + }, + { + "epoch": 0.09299, + "grad_norm": 0.8067138238009587, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9299 + }, + { + "epoch": 0.093, + "grad_norm": 0.8762900677643826, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9300 + }, + { + "epoch": 0.09301, + "grad_norm": 0.9503109211302515, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 9301 + }, + { + "epoch": 0.09302, + "grad_norm": 1.1335888930514397, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 9302 + }, + { + "epoch": 0.09303, + "grad_norm": 0.8457264426047458, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 9303 + }, + { + "epoch": 0.09304, + "grad_norm": 0.9912599414995296, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 9304 + }, + { + "epoch": 0.09305, + "grad_norm": 1.0180631509700129, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 9305 + }, + { + "epoch": 0.09306, + "grad_norm": 1.1194293041880785, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9306 + }, + { + "epoch": 0.09307, + "grad_norm": 1.1133493890272341, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 9307 + }, + { + "epoch": 0.09308, + "grad_norm": 0.9634161791067534, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 9308 + }, + { + "epoch": 0.09309, + "grad_norm": 0.9473486800894216, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 9309 + }, + { + "epoch": 0.0931, + "grad_norm": 1.0246728440886619, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 9310 + }, + { + "epoch": 0.09311, + "grad_norm": 0.9925465656864717, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9311 + }, + { + "epoch": 0.09312, + "grad_norm": 1.2423874194896622, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 9312 + }, + { + "epoch": 0.09313, + "grad_norm": 0.9586209397059013, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 9313 + }, + { + "epoch": 0.09314, + "grad_norm": 0.9862059787685488, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9314 + }, + { + "epoch": 0.09315, + "grad_norm": 0.9997247291164021, + "learning_rate": 0.003, + "loss": 4.092, + "step": 9315 + }, + { + "epoch": 0.09316, + "grad_norm": 0.9978795234093828, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 9316 + }, + { + "epoch": 0.09317, + "grad_norm": 1.0202358866446537, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 9317 + }, + { + "epoch": 0.09318, + "grad_norm": 1.0358808879366825, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 9318 + }, + { + "epoch": 0.09319, + "grad_norm": 0.9359347662036268, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 9319 + }, + { + "epoch": 0.0932, + "grad_norm": 1.0516397687574617, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 9320 + }, + { + "epoch": 0.09321, + "grad_norm": 1.027948417607539, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 9321 + }, + { + "epoch": 0.09322, + "grad_norm": 1.2238846396237006, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 9322 + }, + { + "epoch": 0.09323, + "grad_norm": 1.0710084791274084, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 9323 + }, + { + "epoch": 0.09324, + "grad_norm": 0.8664748503030784, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 9324 + }, + { + "epoch": 0.09325, + "grad_norm": 0.8244384523394639, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 9325 + }, + { + "epoch": 0.09326, + "grad_norm": 0.8366377301647347, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 9326 + }, + { + "epoch": 0.09327, + "grad_norm": 1.002471764215464, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 9327 + }, + { + "epoch": 0.09328, + "grad_norm": 1.2210088285550147, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9328 + }, + { + "epoch": 0.09329, + "grad_norm": 1.0013379117391867, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 9329 + }, + { + "epoch": 0.0933, + "grad_norm": 1.1944499976346676, + "learning_rate": 0.003, + "loss": 4.119, + "step": 9330 + }, + { + "epoch": 0.09331, + "grad_norm": 0.7632243261233325, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 9331 + }, + { + "epoch": 0.09332, + "grad_norm": 0.9200958475962796, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9332 + }, + { + "epoch": 0.09333, + "grad_norm": 1.403618569774701, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 9333 + }, + { + "epoch": 0.09334, + "grad_norm": 0.9530756028848038, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 9334 + }, + { + "epoch": 0.09335, + "grad_norm": 0.8373722239247557, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 9335 + }, + { + "epoch": 0.09336, + "grad_norm": 1.07671774620286, + "learning_rate": 0.003, + "loss": 4.082, + "step": 9336 + }, + { + "epoch": 0.09337, + "grad_norm": 1.1688556059409754, + "learning_rate": 0.003, + "loss": 4.102, + "step": 9337 + }, + { + "epoch": 0.09338, + "grad_norm": 0.9004553842817038, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 9338 + }, + { + "epoch": 0.09339, + "grad_norm": 0.8997829693537273, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 9339 + }, + { + "epoch": 0.0934, + "grad_norm": 1.0364306321074155, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 9340 + }, + { + "epoch": 0.09341, + "grad_norm": 1.1496436689471086, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 9341 + }, + { + "epoch": 0.09342, + "grad_norm": 0.8404709893004499, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 9342 + }, + { + "epoch": 0.09343, + "grad_norm": 0.8638397010633795, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 9343 + }, + { + "epoch": 0.09344, + "grad_norm": 1.009502025708084, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 9344 + }, + { + "epoch": 0.09345, + "grad_norm": 1.1049442915711438, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 9345 + }, + { + "epoch": 0.09346, + "grad_norm": 0.9264275595290813, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 9346 + }, + { + "epoch": 0.09347, + "grad_norm": 1.04657561770358, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 9347 + }, + { + "epoch": 0.09348, + "grad_norm": 1.1020959605289053, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 9348 + }, + { + "epoch": 0.09349, + "grad_norm": 0.8481624664881736, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 9349 + }, + { + "epoch": 0.0935, + "grad_norm": 0.8853005377836479, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 9350 + }, + { + "epoch": 0.09351, + "grad_norm": 0.8610406536309755, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9351 + }, + { + "epoch": 0.09352, + "grad_norm": 0.8557730123548091, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 9352 + }, + { + "epoch": 0.09353, + "grad_norm": 0.8862296744798557, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 9353 + }, + { + "epoch": 0.09354, + "grad_norm": 0.9143136262869864, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 9354 + }, + { + "epoch": 0.09355, + "grad_norm": 1.0652270291919952, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9355 + }, + { + "epoch": 0.09356, + "grad_norm": 1.2739415869779755, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9356 + }, + { + "epoch": 0.09357, + "grad_norm": 0.8507290447629781, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 9357 + }, + { + "epoch": 0.09358, + "grad_norm": 0.8382429448084058, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 9358 + }, + { + "epoch": 0.09359, + "grad_norm": 1.1026964823449827, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 9359 + }, + { + "epoch": 0.0936, + "grad_norm": 0.953788284156056, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9360 + }, + { + "epoch": 0.09361, + "grad_norm": 0.9725724420048601, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 9361 + }, + { + "epoch": 0.09362, + "grad_norm": 1.0587389958001008, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 9362 + }, + { + "epoch": 0.09363, + "grad_norm": 0.9453661891620305, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 9363 + }, + { + "epoch": 0.09364, + "grad_norm": 1.166103814168175, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9364 + }, + { + "epoch": 0.09365, + "grad_norm": 0.981304472296457, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 9365 + }, + { + "epoch": 0.09366, + "grad_norm": 1.3421162985123651, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 9366 + }, + { + "epoch": 0.09367, + "grad_norm": 1.112844598418381, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 9367 + }, + { + "epoch": 0.09368, + "grad_norm": 0.9409251616261554, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 9368 + }, + { + "epoch": 0.09369, + "grad_norm": 0.8769814653061887, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 9369 + }, + { + "epoch": 0.0937, + "grad_norm": 0.8344076839182143, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 9370 + }, + { + "epoch": 0.09371, + "grad_norm": 0.8198638715880406, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 9371 + }, + { + "epoch": 0.09372, + "grad_norm": 0.8780692813529192, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9372 + }, + { + "epoch": 0.09373, + "grad_norm": 0.919173903864044, + "learning_rate": 0.003, + "loss": 4.1259, + "step": 9373 + }, + { + "epoch": 0.09374, + "grad_norm": 1.1774099587785354, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 9374 + }, + { + "epoch": 0.09375, + "grad_norm": 0.9234795661462851, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9375 + }, + { + "epoch": 0.09376, + "grad_norm": 0.8716847861495836, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9376 + }, + { + "epoch": 0.09377, + "grad_norm": 0.8578647905954336, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9377 + }, + { + "epoch": 0.09378, + "grad_norm": 0.7674758289318644, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 9378 + }, + { + "epoch": 0.09379, + "grad_norm": 0.9036010042215633, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9379 + }, + { + "epoch": 0.0938, + "grad_norm": 1.2257343544089907, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9380 + }, + { + "epoch": 0.09381, + "grad_norm": 1.1510725609081562, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9381 + }, + { + "epoch": 0.09382, + "grad_norm": 1.0854982176578536, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 9382 + }, + { + "epoch": 0.09383, + "grad_norm": 1.0776699328606887, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9383 + }, + { + "epoch": 0.09384, + "grad_norm": 1.0496928305379307, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 9384 + }, + { + "epoch": 0.09385, + "grad_norm": 1.0158952106727617, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 9385 + }, + { + "epoch": 0.09386, + "grad_norm": 1.1009188691449934, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 9386 + }, + { + "epoch": 0.09387, + "grad_norm": 0.9331858264293696, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 9387 + }, + { + "epoch": 0.09388, + "grad_norm": 0.9977640976374138, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 9388 + }, + { + "epoch": 0.09389, + "grad_norm": 1.09021391133396, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9389 + }, + { + "epoch": 0.0939, + "grad_norm": 1.1012507495521715, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 9390 + }, + { + "epoch": 0.09391, + "grad_norm": 1.1340299983662043, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 9391 + }, + { + "epoch": 0.09392, + "grad_norm": 1.0755966661059326, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 9392 + }, + { + "epoch": 0.09393, + "grad_norm": 1.0564351543775305, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 9393 + }, + { + "epoch": 0.09394, + "grad_norm": 0.9221702675478415, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 9394 + }, + { + "epoch": 0.09395, + "grad_norm": 1.11867912752045, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9395 + }, + { + "epoch": 0.09396, + "grad_norm": 1.0650662730096374, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 9396 + }, + { + "epoch": 0.09397, + "grad_norm": 0.9989739888141916, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 9397 + }, + { + "epoch": 0.09398, + "grad_norm": 1.04745389684216, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 9398 + }, + { + "epoch": 0.09399, + "grad_norm": 0.9622692280566821, + "learning_rate": 0.003, + "loss": 4.096, + "step": 9399 + }, + { + "epoch": 0.094, + "grad_norm": 1.0129391627458666, + "learning_rate": 0.003, + "loss": 4.092, + "step": 9400 + }, + { + "epoch": 0.09401, + "grad_norm": 1.0179038955491484, + "learning_rate": 0.003, + "loss": 4.104, + "step": 9401 + }, + { + "epoch": 0.09402, + "grad_norm": 1.0037084366352333, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 9402 + }, + { + "epoch": 0.09403, + "grad_norm": 1.116193331964759, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 9403 + }, + { + "epoch": 0.09404, + "grad_norm": 1.114482641077499, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 9404 + }, + { + "epoch": 0.09405, + "grad_norm": 0.9656290178764857, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 9405 + }, + { + "epoch": 0.09406, + "grad_norm": 1.0418496678579732, + "learning_rate": 0.003, + "loss": 4.115, + "step": 9406 + }, + { + "epoch": 0.09407, + "grad_norm": 0.8701407322223268, + "learning_rate": 0.003, + "loss": 4.096, + "step": 9407 + }, + { + "epoch": 0.09408, + "grad_norm": 0.8498940051083433, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9408 + }, + { + "epoch": 0.09409, + "grad_norm": 0.9515337881634939, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 9409 + }, + { + "epoch": 0.0941, + "grad_norm": 0.8952625371005679, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 9410 + }, + { + "epoch": 0.09411, + "grad_norm": 0.9904717550526758, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 9411 + }, + { + "epoch": 0.09412, + "grad_norm": 1.0747473702120094, + "learning_rate": 0.003, + "loss": 4.116, + "step": 9412 + }, + { + "epoch": 0.09413, + "grad_norm": 1.0544939442524188, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9413 + }, + { + "epoch": 0.09414, + "grad_norm": 0.9446059373055282, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 9414 + }, + { + "epoch": 0.09415, + "grad_norm": 1.1814575295816008, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 9415 + }, + { + "epoch": 0.09416, + "grad_norm": 1.2784405713018843, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9416 + }, + { + "epoch": 0.09417, + "grad_norm": 1.0572081822689008, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 9417 + }, + { + "epoch": 0.09418, + "grad_norm": 1.2041574706110378, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9418 + }, + { + "epoch": 0.09419, + "grad_norm": 0.9168210345813573, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 9419 + }, + { + "epoch": 0.0942, + "grad_norm": 0.9165745133346884, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9420 + }, + { + "epoch": 0.09421, + "grad_norm": 0.8507495772840832, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 9421 + }, + { + "epoch": 0.09422, + "grad_norm": 0.9415383624526804, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9422 + }, + { + "epoch": 0.09423, + "grad_norm": 1.1234380596400146, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9423 + }, + { + "epoch": 0.09424, + "grad_norm": 0.9442798953705628, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9424 + }, + { + "epoch": 0.09425, + "grad_norm": 0.9135668408687693, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 9425 + }, + { + "epoch": 0.09426, + "grad_norm": 0.9577560361319429, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 9426 + }, + { + "epoch": 0.09427, + "grad_norm": 1.0865627308701054, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 9427 + }, + { + "epoch": 0.09428, + "grad_norm": 1.0884339478261547, + "learning_rate": 0.003, + "loss": 4.118, + "step": 9428 + }, + { + "epoch": 0.09429, + "grad_norm": 1.059814754063221, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 9429 + }, + { + "epoch": 0.0943, + "grad_norm": 1.0809288733936093, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 9430 + }, + { + "epoch": 0.09431, + "grad_norm": 0.9441751012722414, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 9431 + }, + { + "epoch": 0.09432, + "grad_norm": 1.0069652330680032, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9432 + }, + { + "epoch": 0.09433, + "grad_norm": 1.2770781548897892, + "learning_rate": 0.003, + "loss": 4.095, + "step": 9433 + }, + { + "epoch": 0.09434, + "grad_norm": 0.9107472747872962, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9434 + }, + { + "epoch": 0.09435, + "grad_norm": 0.9371330665814549, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9435 + }, + { + "epoch": 0.09436, + "grad_norm": 0.948797698466554, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 9436 + }, + { + "epoch": 0.09437, + "grad_norm": 1.0965285210238682, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 9437 + }, + { + "epoch": 0.09438, + "grad_norm": 0.9539276888088277, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 9438 + }, + { + "epoch": 0.09439, + "grad_norm": 0.9297934842516888, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 9439 + }, + { + "epoch": 0.0944, + "grad_norm": 0.8833885403510608, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 9440 + }, + { + "epoch": 0.09441, + "grad_norm": 0.7745310460762501, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 9441 + }, + { + "epoch": 0.09442, + "grad_norm": 0.7444800565907276, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 9442 + }, + { + "epoch": 0.09443, + "grad_norm": 0.8414080585325165, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 9443 + }, + { + "epoch": 0.09444, + "grad_norm": 1.0036008655746649, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 9444 + }, + { + "epoch": 0.09445, + "grad_norm": 1.2505778804758252, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 9445 + }, + { + "epoch": 0.09446, + "grad_norm": 1.0393979651836986, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9446 + }, + { + "epoch": 0.09447, + "grad_norm": 0.964114883135747, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 9447 + }, + { + "epoch": 0.09448, + "grad_norm": 1.16494849936199, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 9448 + }, + { + "epoch": 0.09449, + "grad_norm": 0.9548194627263663, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 9449 + }, + { + "epoch": 0.0945, + "grad_norm": 1.0233790768032425, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 9450 + }, + { + "epoch": 0.09451, + "grad_norm": 0.9840248139547673, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 9451 + }, + { + "epoch": 0.09452, + "grad_norm": 1.10386291686478, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 9452 + }, + { + "epoch": 0.09453, + "grad_norm": 1.0942515861458446, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9453 + }, + { + "epoch": 0.09454, + "grad_norm": 1.0520618341084693, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 9454 + }, + { + "epoch": 0.09455, + "grad_norm": 1.032921792376323, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9455 + }, + { + "epoch": 0.09456, + "grad_norm": 1.0398131988505905, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 9456 + }, + { + "epoch": 0.09457, + "grad_norm": 1.0491898439800345, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 9457 + }, + { + "epoch": 0.09458, + "grad_norm": 0.9354580793426419, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 9458 + }, + { + "epoch": 0.09459, + "grad_norm": 0.9716067137585749, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 9459 + }, + { + "epoch": 0.0946, + "grad_norm": 1.0690755013786215, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 9460 + }, + { + "epoch": 0.09461, + "grad_norm": 1.1380058613035746, + "learning_rate": 0.003, + "loss": 4.124, + "step": 9461 + }, + { + "epoch": 0.09462, + "grad_norm": 0.9795890325721478, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9462 + }, + { + "epoch": 0.09463, + "grad_norm": 1.0112480646782764, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 9463 + }, + { + "epoch": 0.09464, + "grad_norm": 1.4174770915576664, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9464 + }, + { + "epoch": 0.09465, + "grad_norm": 0.7118195304889618, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9465 + }, + { + "epoch": 0.09466, + "grad_norm": 0.7397324947507956, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 9466 + }, + { + "epoch": 0.09467, + "grad_norm": 0.9228651076473079, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 9467 + }, + { + "epoch": 0.09468, + "grad_norm": 1.053278439976445, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 9468 + }, + { + "epoch": 0.09469, + "grad_norm": 1.160147117243451, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9469 + }, + { + "epoch": 0.0947, + "grad_norm": 1.026133164845867, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9470 + }, + { + "epoch": 0.09471, + "grad_norm": 1.164591722645013, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 9471 + }, + { + "epoch": 0.09472, + "grad_norm": 0.7985197662831892, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9472 + }, + { + "epoch": 0.09473, + "grad_norm": 0.763553189156105, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 9473 + }, + { + "epoch": 0.09474, + "grad_norm": 0.8175418581057333, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9474 + }, + { + "epoch": 0.09475, + "grad_norm": 0.9100168795925652, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 9475 + }, + { + "epoch": 0.09476, + "grad_norm": 0.9026725830353436, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 9476 + }, + { + "epoch": 0.09477, + "grad_norm": 0.9624617938673868, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 9477 + }, + { + "epoch": 0.09478, + "grad_norm": 1.081533379210801, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 9478 + }, + { + "epoch": 0.09479, + "grad_norm": 1.0540513595530958, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 9479 + }, + { + "epoch": 0.0948, + "grad_norm": 1.0198438271547687, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 9480 + }, + { + "epoch": 0.09481, + "grad_norm": 1.2125887044399974, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 9481 + }, + { + "epoch": 0.09482, + "grad_norm": 1.1188988335754426, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 9482 + }, + { + "epoch": 0.09483, + "grad_norm": 0.9414423522388563, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9483 + }, + { + "epoch": 0.09484, + "grad_norm": 1.0501757386247388, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 9484 + }, + { + "epoch": 0.09485, + "grad_norm": 1.0797081595739608, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 9485 + }, + { + "epoch": 0.09486, + "grad_norm": 1.1593524413712497, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 9486 + }, + { + "epoch": 0.09487, + "grad_norm": 0.9008992419862395, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 9487 + }, + { + "epoch": 0.09488, + "grad_norm": 0.928425213219695, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 9488 + }, + { + "epoch": 0.09489, + "grad_norm": 1.1255438362788508, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 9489 + }, + { + "epoch": 0.0949, + "grad_norm": 0.9565865756324435, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 9490 + }, + { + "epoch": 0.09491, + "grad_norm": 1.091082485215994, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 9491 + }, + { + "epoch": 0.09492, + "grad_norm": 1.0432359393866104, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9492 + }, + { + "epoch": 0.09493, + "grad_norm": 0.9947831762502642, + "learning_rate": 0.003, + "loss": 4.068, + "step": 9493 + }, + { + "epoch": 0.09494, + "grad_norm": 1.170287104954015, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 9494 + }, + { + "epoch": 0.09495, + "grad_norm": 1.0499066150365186, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 9495 + }, + { + "epoch": 0.09496, + "grad_norm": 1.4045418939748382, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 9496 + }, + { + "epoch": 0.09497, + "grad_norm": 0.8509044383443939, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 9497 + }, + { + "epoch": 0.09498, + "grad_norm": 0.8693676163780726, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9498 + }, + { + "epoch": 0.09499, + "grad_norm": 0.8019980727919567, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 9499 + }, + { + "epoch": 0.095, + "grad_norm": 0.8765503155986619, + "learning_rate": 0.003, + "loss": 4.121, + "step": 9500 + }, + { + "epoch": 0.09501, + "grad_norm": 0.8949783455821654, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 9501 + }, + { + "epoch": 0.09502, + "grad_norm": 0.85473282604943, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 9502 + }, + { + "epoch": 0.09503, + "grad_norm": 0.8962207462857377, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9503 + }, + { + "epoch": 0.09504, + "grad_norm": 0.8523909551635448, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9504 + }, + { + "epoch": 0.09505, + "grad_norm": 0.9576305558046839, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 9505 + }, + { + "epoch": 0.09506, + "grad_norm": 1.2342117179715257, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 9506 + }, + { + "epoch": 0.09507, + "grad_norm": 1.2976078459163851, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 9507 + }, + { + "epoch": 0.09508, + "grad_norm": 1.185691690338481, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 9508 + }, + { + "epoch": 0.09509, + "grad_norm": 0.9674516966851515, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9509 + }, + { + "epoch": 0.0951, + "grad_norm": 1.014442271229994, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9510 + }, + { + "epoch": 0.09511, + "grad_norm": 1.0385934154900638, + "learning_rate": 0.003, + "loss": 4.083, + "step": 9511 + }, + { + "epoch": 0.09512, + "grad_norm": 0.9361766204688062, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9512 + }, + { + "epoch": 0.09513, + "grad_norm": 1.0033999785344312, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9513 + }, + { + "epoch": 0.09514, + "grad_norm": 1.05390650992189, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9514 + }, + { + "epoch": 0.09515, + "grad_norm": 1.116848612024721, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9515 + }, + { + "epoch": 0.09516, + "grad_norm": 0.9674724683645552, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 9516 + }, + { + "epoch": 0.09517, + "grad_norm": 1.0957320303666789, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 9517 + }, + { + "epoch": 0.09518, + "grad_norm": 1.0257060215691023, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 9518 + }, + { + "epoch": 0.09519, + "grad_norm": 1.0744943454127902, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 9519 + }, + { + "epoch": 0.0952, + "grad_norm": 1.2248690019464181, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 9520 + }, + { + "epoch": 0.09521, + "grad_norm": 0.9237408992684386, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 9521 + }, + { + "epoch": 0.09522, + "grad_norm": 0.8081645199251014, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9522 + }, + { + "epoch": 0.09523, + "grad_norm": 0.8001415036331109, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9523 + }, + { + "epoch": 0.09524, + "grad_norm": 0.8721020743119626, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 9524 + }, + { + "epoch": 0.09525, + "grad_norm": 1.0070519804673295, + "learning_rate": 0.003, + "loss": 4.092, + "step": 9525 + }, + { + "epoch": 0.09526, + "grad_norm": 1.0874508141227783, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 9526 + }, + { + "epoch": 0.09527, + "grad_norm": 0.9064561128680414, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 9527 + }, + { + "epoch": 0.09528, + "grad_norm": 0.8819852497888345, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 9528 + }, + { + "epoch": 0.09529, + "grad_norm": 1.0155582815835493, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 9529 + }, + { + "epoch": 0.0953, + "grad_norm": 1.2335728455983175, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9530 + }, + { + "epoch": 0.09531, + "grad_norm": 0.9890751026219354, + "learning_rate": 0.003, + "loss": 4.075, + "step": 9531 + }, + { + "epoch": 0.09532, + "grad_norm": 1.0139299938673432, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 9532 + }, + { + "epoch": 0.09533, + "grad_norm": 1.0406685870722243, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 9533 + }, + { + "epoch": 0.09534, + "grad_norm": 1.0415786211613325, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 9534 + }, + { + "epoch": 0.09535, + "grad_norm": 1.0266775971668027, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 9535 + }, + { + "epoch": 0.09536, + "grad_norm": 1.0680875423201488, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 9536 + }, + { + "epoch": 0.09537, + "grad_norm": 0.9532000734499076, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 9537 + }, + { + "epoch": 0.09538, + "grad_norm": 1.0875678049842619, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 9538 + }, + { + "epoch": 0.09539, + "grad_norm": 1.204316570645322, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 9539 + }, + { + "epoch": 0.0954, + "grad_norm": 1.030083701060855, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9540 + }, + { + "epoch": 0.09541, + "grad_norm": 1.1297064889435355, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 9541 + }, + { + "epoch": 0.09542, + "grad_norm": 0.9173234611129824, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 9542 + }, + { + "epoch": 0.09543, + "grad_norm": 0.9536043595896253, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 9543 + }, + { + "epoch": 0.09544, + "grad_norm": 1.111358012851114, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 9544 + }, + { + "epoch": 0.09545, + "grad_norm": 1.0573210212427635, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 9545 + }, + { + "epoch": 0.09546, + "grad_norm": 1.0671113439958861, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 9546 + }, + { + "epoch": 0.09547, + "grad_norm": 0.9551135025314722, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9547 + }, + { + "epoch": 0.09548, + "grad_norm": 1.089615403951393, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 9548 + }, + { + "epoch": 0.09549, + "grad_norm": 1.1596112981509654, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 9549 + }, + { + "epoch": 0.0955, + "grad_norm": 0.9088861727644186, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 9550 + }, + { + "epoch": 0.09551, + "grad_norm": 1.2391520142985264, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 9551 + }, + { + "epoch": 0.09552, + "grad_norm": 1.0362472449674538, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 9552 + }, + { + "epoch": 0.09553, + "grad_norm": 0.9059425211543064, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 9553 + }, + { + "epoch": 0.09554, + "grad_norm": 0.7685716396395997, + "learning_rate": 0.003, + "loss": 4.062, + "step": 9554 + }, + { + "epoch": 0.09555, + "grad_norm": 0.7760521570571651, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 9555 + }, + { + "epoch": 0.09556, + "grad_norm": 0.915451851496684, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 9556 + }, + { + "epoch": 0.09557, + "grad_norm": 1.0725359354533892, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 9557 + }, + { + "epoch": 0.09558, + "grad_norm": 1.1165179826378548, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 9558 + }, + { + "epoch": 0.09559, + "grad_norm": 1.0260042229050033, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 9559 + }, + { + "epoch": 0.0956, + "grad_norm": 1.1128206813614063, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 9560 + }, + { + "epoch": 0.09561, + "grad_norm": 0.9561200539995762, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 9561 + }, + { + "epoch": 0.09562, + "grad_norm": 0.7921512806544463, + "learning_rate": 0.003, + "loss": 4.052, + "step": 9562 + }, + { + "epoch": 0.09563, + "grad_norm": 0.9041768945308537, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 9563 + }, + { + "epoch": 0.09564, + "grad_norm": 1.1681721170614285, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 9564 + }, + { + "epoch": 0.09565, + "grad_norm": 1.229817627860069, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 9565 + }, + { + "epoch": 0.09566, + "grad_norm": 0.8408199819338241, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 9566 + }, + { + "epoch": 0.09567, + "grad_norm": 1.0091863337567137, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 9567 + }, + { + "epoch": 0.09568, + "grad_norm": 1.3094571250437879, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9568 + }, + { + "epoch": 0.09569, + "grad_norm": 0.8623005542878108, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9569 + }, + { + "epoch": 0.0957, + "grad_norm": 0.9413601755374859, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 9570 + }, + { + "epoch": 0.09571, + "grad_norm": 0.8957027980528774, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 9571 + }, + { + "epoch": 0.09572, + "grad_norm": 0.9187975967965829, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 9572 + }, + { + "epoch": 0.09573, + "grad_norm": 1.137687044149432, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 9573 + }, + { + "epoch": 0.09574, + "grad_norm": 1.1691163828761597, + "learning_rate": 0.003, + "loss": 4.095, + "step": 9574 + }, + { + "epoch": 0.09575, + "grad_norm": 0.9206397514109596, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 9575 + }, + { + "epoch": 0.09576, + "grad_norm": 0.9884030120613543, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9576 + }, + { + "epoch": 0.09577, + "grad_norm": 1.151909672565129, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 9577 + }, + { + "epoch": 0.09578, + "grad_norm": 0.943817382689356, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 9578 + }, + { + "epoch": 0.09579, + "grad_norm": 0.9266875861613927, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9579 + }, + { + "epoch": 0.0958, + "grad_norm": 1.1200656888418847, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9580 + }, + { + "epoch": 0.09581, + "grad_norm": 1.0924396133419563, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 9581 + }, + { + "epoch": 0.09582, + "grad_norm": 0.8011259972597407, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 9582 + }, + { + "epoch": 0.09583, + "grad_norm": 0.9706166643361759, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 9583 + }, + { + "epoch": 0.09584, + "grad_norm": 1.0375471914237828, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 9584 + }, + { + "epoch": 0.09585, + "grad_norm": 1.162671870104563, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9585 + }, + { + "epoch": 0.09586, + "grad_norm": 1.1269620104984417, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 9586 + }, + { + "epoch": 0.09587, + "grad_norm": 1.1157717297270022, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9587 + }, + { + "epoch": 0.09588, + "grad_norm": 1.1329330107551132, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 9588 + }, + { + "epoch": 0.09589, + "grad_norm": 1.1375709999149357, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 9589 + }, + { + "epoch": 0.0959, + "grad_norm": 1.0008878090061901, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 9590 + }, + { + "epoch": 0.09591, + "grad_norm": 1.0935601459276771, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 9591 + }, + { + "epoch": 0.09592, + "grad_norm": 0.9286570503636103, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9592 + }, + { + "epoch": 0.09593, + "grad_norm": 0.9832018964530126, + "learning_rate": 0.003, + "loss": 4.11, + "step": 9593 + }, + { + "epoch": 0.09594, + "grad_norm": 0.9244421706166079, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 9594 + }, + { + "epoch": 0.09595, + "grad_norm": 0.9104310402580331, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9595 + }, + { + "epoch": 0.09596, + "grad_norm": 0.9548164533755923, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9596 + }, + { + "epoch": 0.09597, + "grad_norm": 1.141997527843847, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 9597 + }, + { + "epoch": 0.09598, + "grad_norm": 1.245711271585723, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9598 + }, + { + "epoch": 0.09599, + "grad_norm": 0.8088955787411081, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 9599 + }, + { + "epoch": 0.096, + "grad_norm": 0.8952527018243382, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 9600 + }, + { + "epoch": 0.09601, + "grad_norm": 1.0349856846675602, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 9601 + }, + { + "epoch": 0.09602, + "grad_norm": 1.0594507709330971, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 9602 + }, + { + "epoch": 0.09603, + "grad_norm": 1.1028365715897552, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9603 + }, + { + "epoch": 0.09604, + "grad_norm": 0.8849548250049927, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 9604 + }, + { + "epoch": 0.09605, + "grad_norm": 0.922563197820632, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 9605 + }, + { + "epoch": 0.09606, + "grad_norm": 1.105340202233273, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9606 + }, + { + "epoch": 0.09607, + "grad_norm": 1.2220515830119882, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 9607 + }, + { + "epoch": 0.09608, + "grad_norm": 0.9494734780740829, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 9608 + }, + { + "epoch": 0.09609, + "grad_norm": 1.0816550584500007, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 9609 + }, + { + "epoch": 0.0961, + "grad_norm": 0.9492548605654175, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9610 + }, + { + "epoch": 0.09611, + "grad_norm": 0.9587253967158587, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 9611 + }, + { + "epoch": 0.09612, + "grad_norm": 1.1146970448641045, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 9612 + }, + { + "epoch": 0.09613, + "grad_norm": 1.0580108312474177, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 9613 + }, + { + "epoch": 0.09614, + "grad_norm": 1.0472858692305076, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 9614 + }, + { + "epoch": 0.09615, + "grad_norm": 1.2366350300452777, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 9615 + }, + { + "epoch": 0.09616, + "grad_norm": 0.8816270210973588, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9616 + }, + { + "epoch": 0.09617, + "grad_norm": 0.8722350756954568, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9617 + }, + { + "epoch": 0.09618, + "grad_norm": 0.9067366940665751, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 9618 + }, + { + "epoch": 0.09619, + "grad_norm": 1.0481783137642813, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 9619 + }, + { + "epoch": 0.0962, + "grad_norm": 1.015015664502483, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 9620 + }, + { + "epoch": 0.09621, + "grad_norm": 1.057917176675139, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 9621 + }, + { + "epoch": 0.09622, + "grad_norm": 1.062674626405321, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 9622 + }, + { + "epoch": 0.09623, + "grad_norm": 1.0840254275116858, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 9623 + }, + { + "epoch": 0.09624, + "grad_norm": 0.921050877590142, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9624 + }, + { + "epoch": 0.09625, + "grad_norm": 0.9745687452474494, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9625 + }, + { + "epoch": 0.09626, + "grad_norm": 1.0761942857238649, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9626 + }, + { + "epoch": 0.09627, + "grad_norm": 1.2836132658639128, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 9627 + }, + { + "epoch": 0.09628, + "grad_norm": 1.2249241628868797, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 9628 + }, + { + "epoch": 0.09629, + "grad_norm": 0.8425877511796831, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 9629 + }, + { + "epoch": 0.0963, + "grad_norm": 0.8336489764607532, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 9630 + }, + { + "epoch": 0.09631, + "grad_norm": 0.9683817459934598, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 9631 + }, + { + "epoch": 0.09632, + "grad_norm": 1.1120343603001173, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 9632 + }, + { + "epoch": 0.09633, + "grad_norm": 0.9976490542448117, + "learning_rate": 0.003, + "loss": 4.074, + "step": 9633 + }, + { + "epoch": 0.09634, + "grad_norm": 1.0472125777692542, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9634 + }, + { + "epoch": 0.09635, + "grad_norm": 1.046705158707178, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 9635 + }, + { + "epoch": 0.09636, + "grad_norm": 1.1240079021749696, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 9636 + }, + { + "epoch": 0.09637, + "grad_norm": 0.9390395253897545, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 9637 + }, + { + "epoch": 0.09638, + "grad_norm": 0.8986737735751393, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 9638 + }, + { + "epoch": 0.09639, + "grad_norm": 0.9270217680975166, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9639 + }, + { + "epoch": 0.0964, + "grad_norm": 0.9594265746905929, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 9640 + }, + { + "epoch": 0.09641, + "grad_norm": 1.0246638935142558, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9641 + }, + { + "epoch": 0.09642, + "grad_norm": 1.1350695971418943, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 9642 + }, + { + "epoch": 0.09643, + "grad_norm": 1.058296721974063, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 9643 + }, + { + "epoch": 0.09644, + "grad_norm": 1.1412333449738055, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 9644 + }, + { + "epoch": 0.09645, + "grad_norm": 1.0255174315928441, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 9645 + }, + { + "epoch": 0.09646, + "grad_norm": 1.1981325820421778, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 9646 + }, + { + "epoch": 0.09647, + "grad_norm": 0.96198785938754, + "learning_rate": 0.003, + "loss": 4.05, + "step": 9647 + }, + { + "epoch": 0.09648, + "grad_norm": 0.9103300780359126, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 9648 + }, + { + "epoch": 0.09649, + "grad_norm": 1.0146850485441834, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 9649 + }, + { + "epoch": 0.0965, + "grad_norm": 1.1014496175687365, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 9650 + }, + { + "epoch": 0.09651, + "grad_norm": 0.9799556586854834, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 9651 + }, + { + "epoch": 0.09652, + "grad_norm": 0.8859085759414587, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9652 + }, + { + "epoch": 0.09653, + "grad_norm": 0.9436219326106128, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 9653 + }, + { + "epoch": 0.09654, + "grad_norm": 1.049922096432769, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 9654 + }, + { + "epoch": 0.09655, + "grad_norm": 1.0512634494244577, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9655 + }, + { + "epoch": 0.09656, + "grad_norm": 1.2561052980696712, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 9656 + }, + { + "epoch": 0.09657, + "grad_norm": 0.9464113750920866, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9657 + }, + { + "epoch": 0.09658, + "grad_norm": 0.8936277941328696, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 9658 + }, + { + "epoch": 0.09659, + "grad_norm": 0.9074412530038821, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 9659 + }, + { + "epoch": 0.0966, + "grad_norm": 0.959500742463755, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 9660 + }, + { + "epoch": 0.09661, + "grad_norm": 1.025665764705048, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9661 + }, + { + "epoch": 0.09662, + "grad_norm": 0.9870720584477939, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 9662 + }, + { + "epoch": 0.09663, + "grad_norm": 1.1394465371192255, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 9663 + }, + { + "epoch": 0.09664, + "grad_norm": 1.0049834681547085, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 9664 + }, + { + "epoch": 0.09665, + "grad_norm": 1.1993551495956545, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9665 + }, + { + "epoch": 0.09666, + "grad_norm": 1.2099942084141782, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 9666 + }, + { + "epoch": 0.09667, + "grad_norm": 0.9917211003714878, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 9667 + }, + { + "epoch": 0.09668, + "grad_norm": 1.1434647616000775, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 9668 + }, + { + "epoch": 0.09669, + "grad_norm": 0.8296204462737752, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9669 + }, + { + "epoch": 0.0967, + "grad_norm": 0.8774506033092024, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9670 + }, + { + "epoch": 0.09671, + "grad_norm": 0.9598356684861605, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9671 + }, + { + "epoch": 0.09672, + "grad_norm": 1.1282693171893508, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 9672 + }, + { + "epoch": 0.09673, + "grad_norm": 0.9782521798986386, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 9673 + }, + { + "epoch": 0.09674, + "grad_norm": 1.0774337255597977, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 9674 + }, + { + "epoch": 0.09675, + "grad_norm": 1.1642571154544263, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 9675 + }, + { + "epoch": 0.09676, + "grad_norm": 1.163816620331313, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 9676 + }, + { + "epoch": 0.09677, + "grad_norm": 0.9786912961895912, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9677 + }, + { + "epoch": 0.09678, + "grad_norm": 1.073702735587886, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 9678 + }, + { + "epoch": 0.09679, + "grad_norm": 1.060204601252453, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 9679 + }, + { + "epoch": 0.0968, + "grad_norm": 1.0203738671671754, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 9680 + }, + { + "epoch": 0.09681, + "grad_norm": 1.047329246570657, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9681 + }, + { + "epoch": 0.09682, + "grad_norm": 1.0233270765342837, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 9682 + }, + { + "epoch": 0.09683, + "grad_norm": 1.1862154773782052, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 9683 + }, + { + "epoch": 0.09684, + "grad_norm": 0.9846506225870897, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 9684 + }, + { + "epoch": 0.09685, + "grad_norm": 0.9516168976016748, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 9685 + }, + { + "epoch": 0.09686, + "grad_norm": 1.003400150329955, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 9686 + }, + { + "epoch": 0.09687, + "grad_norm": 1.231150381428762, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 9687 + }, + { + "epoch": 0.09688, + "grad_norm": 0.8015979155657765, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 9688 + }, + { + "epoch": 0.09689, + "grad_norm": 0.7299334181220091, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9689 + }, + { + "epoch": 0.0969, + "grad_norm": 0.7282724727736201, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 9690 + }, + { + "epoch": 0.09691, + "grad_norm": 0.7638905566157688, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 9691 + }, + { + "epoch": 0.09692, + "grad_norm": 0.96832199777829, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 9692 + }, + { + "epoch": 0.09693, + "grad_norm": 1.1902987491844759, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 9693 + }, + { + "epoch": 0.09694, + "grad_norm": 0.8606385726141452, + "learning_rate": 0.003, + "loss": 4.079, + "step": 9694 + }, + { + "epoch": 0.09695, + "grad_norm": 0.8380888636245558, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9695 + }, + { + "epoch": 0.09696, + "grad_norm": 0.8309556717207915, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 9696 + }, + { + "epoch": 0.09697, + "grad_norm": 1.1345771693009117, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 9697 + }, + { + "epoch": 0.09698, + "grad_norm": 1.397021456563483, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 9698 + }, + { + "epoch": 0.09699, + "grad_norm": 0.8546336497231758, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 9699 + }, + { + "epoch": 0.097, + "grad_norm": 0.8327596035211329, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 9700 + }, + { + "epoch": 0.09701, + "grad_norm": 1.1146501724689344, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 9701 + }, + { + "epoch": 0.09702, + "grad_norm": 0.9998326345834255, + "learning_rate": 0.003, + "loss": 4.07, + "step": 9702 + }, + { + "epoch": 0.09703, + "grad_norm": 1.1440436859408962, + "learning_rate": 0.003, + "loss": 4.084, + "step": 9703 + }, + { + "epoch": 0.09704, + "grad_norm": 1.1173001326683785, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 9704 + }, + { + "epoch": 0.09705, + "grad_norm": 1.0523970352233503, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 9705 + }, + { + "epoch": 0.09706, + "grad_norm": 1.1713762482483892, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 9706 + }, + { + "epoch": 0.09707, + "grad_norm": 0.8804097816544441, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9707 + }, + { + "epoch": 0.09708, + "grad_norm": 1.0194966549341173, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9708 + }, + { + "epoch": 0.09709, + "grad_norm": 0.9694301047009535, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 9709 + }, + { + "epoch": 0.0971, + "grad_norm": 0.9481709323161568, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 9710 + }, + { + "epoch": 0.09711, + "grad_norm": 1.172731065283466, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 9711 + }, + { + "epoch": 0.09712, + "grad_norm": 0.8011684036455335, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 9712 + }, + { + "epoch": 0.09713, + "grad_norm": 0.9173643441590453, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9713 + }, + { + "epoch": 0.09714, + "grad_norm": 1.0566896957902046, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 9714 + }, + { + "epoch": 0.09715, + "grad_norm": 1.0004516075644887, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 9715 + }, + { + "epoch": 0.09716, + "grad_norm": 1.1696605958430277, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9716 + }, + { + "epoch": 0.09717, + "grad_norm": 1.1191433852342278, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 9717 + }, + { + "epoch": 0.09718, + "grad_norm": 1.1011799234185506, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 9718 + }, + { + "epoch": 0.09719, + "grad_norm": 1.1663962151665566, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 9719 + }, + { + "epoch": 0.0972, + "grad_norm": 1.1267499527720306, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 9720 + }, + { + "epoch": 0.09721, + "grad_norm": 0.9631878849174689, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 9721 + }, + { + "epoch": 0.09722, + "grad_norm": 1.1131135286770764, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 9722 + }, + { + "epoch": 0.09723, + "grad_norm": 1.1406866316885176, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 9723 + }, + { + "epoch": 0.09724, + "grad_norm": 0.9896761239836782, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 9724 + }, + { + "epoch": 0.09725, + "grad_norm": 1.0608334963018875, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 9725 + }, + { + "epoch": 0.09726, + "grad_norm": 1.0223160222808552, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 9726 + }, + { + "epoch": 0.09727, + "grad_norm": 1.1275241052880784, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9727 + }, + { + "epoch": 0.09728, + "grad_norm": 0.8895537961162207, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 9728 + }, + { + "epoch": 0.09729, + "grad_norm": 0.9705906071167186, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9729 + }, + { + "epoch": 0.0973, + "grad_norm": 0.9678791008979821, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 9730 + }, + { + "epoch": 0.09731, + "grad_norm": 0.9789358792375261, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9731 + }, + { + "epoch": 0.09732, + "grad_norm": 1.1262165489255724, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9732 + }, + { + "epoch": 0.09733, + "grad_norm": 0.9935366558924056, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 9733 + }, + { + "epoch": 0.09734, + "grad_norm": 0.9093887102489477, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 9734 + }, + { + "epoch": 0.09735, + "grad_norm": 0.8527078328559802, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 9735 + }, + { + "epoch": 0.09736, + "grad_norm": 0.9159904912201342, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 9736 + }, + { + "epoch": 0.09737, + "grad_norm": 1.0697597475828968, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9737 + }, + { + "epoch": 0.09738, + "grad_norm": 1.1285651318486103, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9738 + }, + { + "epoch": 0.09739, + "grad_norm": 0.9693482906116275, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 9739 + }, + { + "epoch": 0.0974, + "grad_norm": 0.8804291449230246, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 9740 + }, + { + "epoch": 0.09741, + "grad_norm": 0.8841645343019247, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 9741 + }, + { + "epoch": 0.09742, + "grad_norm": 0.9553532279141466, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9742 + }, + { + "epoch": 0.09743, + "grad_norm": 0.9782366425572494, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 9743 + }, + { + "epoch": 0.09744, + "grad_norm": 0.9869511307484383, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 9744 + }, + { + "epoch": 0.09745, + "grad_norm": 1.2099299560104868, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 9745 + }, + { + "epoch": 0.09746, + "grad_norm": 1.4209495817651363, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 9746 + }, + { + "epoch": 0.09747, + "grad_norm": 0.8357217470404129, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 9747 + }, + { + "epoch": 0.09748, + "grad_norm": 0.8752427718006248, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 9748 + }, + { + "epoch": 0.09749, + "grad_norm": 1.1935288313099783, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 9749 + }, + { + "epoch": 0.0975, + "grad_norm": 0.9694278343181028, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 9750 + }, + { + "epoch": 0.09751, + "grad_norm": 1.1606139746617636, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 9751 + }, + { + "epoch": 0.09752, + "grad_norm": 1.06007198810202, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9752 + }, + { + "epoch": 0.09753, + "grad_norm": 1.2442118361414638, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9753 + }, + { + "epoch": 0.09754, + "grad_norm": 0.8708103331423904, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 9754 + }, + { + "epoch": 0.09755, + "grad_norm": 0.8902103561683131, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 9755 + }, + { + "epoch": 0.09756, + "grad_norm": 0.8312788620338242, + "learning_rate": 0.003, + "loss": 4.073, + "step": 9756 + }, + { + "epoch": 0.09757, + "grad_norm": 0.9879969036527848, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9757 + }, + { + "epoch": 0.09758, + "grad_norm": 1.1228401769540215, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 9758 + }, + { + "epoch": 0.09759, + "grad_norm": 1.0242499113055676, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9759 + }, + { + "epoch": 0.0976, + "grad_norm": 1.2189767248747545, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9760 + }, + { + "epoch": 0.09761, + "grad_norm": 1.0854332954217905, + "learning_rate": 0.003, + "loss": 4.088, + "step": 9761 + }, + { + "epoch": 0.09762, + "grad_norm": 0.9142472209749145, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 9762 + }, + { + "epoch": 0.09763, + "grad_norm": 0.8999514242684113, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9763 + }, + { + "epoch": 0.09764, + "grad_norm": 0.9671425605430154, + "learning_rate": 0.003, + "loss": 4.057, + "step": 9764 + }, + { + "epoch": 0.09765, + "grad_norm": 1.0849879456729752, + "learning_rate": 0.003, + "loss": 4.117, + "step": 9765 + }, + { + "epoch": 0.09766, + "grad_norm": 1.0329063206869054, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 9766 + }, + { + "epoch": 0.09767, + "grad_norm": 1.0178074898813942, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 9767 + }, + { + "epoch": 0.09768, + "grad_norm": 1.2302824794252587, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 9768 + }, + { + "epoch": 0.09769, + "grad_norm": 0.8998354813027947, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9769 + }, + { + "epoch": 0.0977, + "grad_norm": 0.9191096965901105, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 9770 + }, + { + "epoch": 0.09771, + "grad_norm": 1.2023215373833045, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 9771 + }, + { + "epoch": 0.09772, + "grad_norm": 1.221541578031842, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 9772 + }, + { + "epoch": 0.09773, + "grad_norm": 1.0360835407587528, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 9773 + }, + { + "epoch": 0.09774, + "grad_norm": 0.9446876117619419, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 9774 + }, + { + "epoch": 0.09775, + "grad_norm": 1.0303074439810562, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 9775 + }, + { + "epoch": 0.09776, + "grad_norm": 1.1253618313034983, + "learning_rate": 0.003, + "loss": 4.078, + "step": 9776 + }, + { + "epoch": 0.09777, + "grad_norm": 0.9439094392869328, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9777 + }, + { + "epoch": 0.09778, + "grad_norm": 0.9618684632253788, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 9778 + }, + { + "epoch": 0.09779, + "grad_norm": 0.9811748296593858, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 9779 + }, + { + "epoch": 0.0978, + "grad_norm": 1.1858634020984171, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 9780 + }, + { + "epoch": 0.09781, + "grad_norm": 0.9338661709362392, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9781 + }, + { + "epoch": 0.09782, + "grad_norm": 0.9332022515801041, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 9782 + }, + { + "epoch": 0.09783, + "grad_norm": 1.1071223925652374, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 9783 + }, + { + "epoch": 0.09784, + "grad_norm": 1.2093941427447275, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 9784 + }, + { + "epoch": 0.09785, + "grad_norm": 1.102823929587096, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 9785 + }, + { + "epoch": 0.09786, + "grad_norm": 1.0035787179191338, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 9786 + }, + { + "epoch": 0.09787, + "grad_norm": 1.078102955585403, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 9787 + }, + { + "epoch": 0.09788, + "grad_norm": 1.0629142159355294, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9788 + }, + { + "epoch": 0.09789, + "grad_norm": 1.0804726075419981, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9789 + }, + { + "epoch": 0.0979, + "grad_norm": 1.1720178170225048, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 9790 + }, + { + "epoch": 0.09791, + "grad_norm": 1.1235469017472899, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 9791 + }, + { + "epoch": 0.09792, + "grad_norm": 1.2149830845028162, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 9792 + }, + { + "epoch": 0.09793, + "grad_norm": 0.9420274703508541, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 9793 + }, + { + "epoch": 0.09794, + "grad_norm": 1.033046463245827, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 9794 + }, + { + "epoch": 0.09795, + "grad_norm": 1.1348495739940019, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 9795 + }, + { + "epoch": 0.09796, + "grad_norm": 0.9757339572958833, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 9796 + }, + { + "epoch": 0.09797, + "grad_norm": 1.0351452973168425, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 9797 + }, + { + "epoch": 0.09798, + "grad_norm": 0.9782960688592283, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 9798 + }, + { + "epoch": 0.09799, + "grad_norm": 1.0164438910147562, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 9799 + }, + { + "epoch": 0.098, + "grad_norm": 1.1163964919065634, + "learning_rate": 0.003, + "loss": 4.089, + "step": 9800 + }, + { + "epoch": 0.09801, + "grad_norm": 1.0433207564204117, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 9801 + }, + { + "epoch": 0.09802, + "grad_norm": 1.006414968625939, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 9802 + }, + { + "epoch": 0.09803, + "grad_norm": 1.139600456125198, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 9803 + }, + { + "epoch": 0.09804, + "grad_norm": 1.0284067990368186, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 9804 + }, + { + "epoch": 0.09805, + "grad_norm": 1.1861908924400755, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 9805 + }, + { + "epoch": 0.09806, + "grad_norm": 0.8349801162214475, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 9806 + }, + { + "epoch": 0.09807, + "grad_norm": 0.8131825023363899, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 9807 + }, + { + "epoch": 0.09808, + "grad_norm": 0.8785674410167105, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 9808 + }, + { + "epoch": 0.09809, + "grad_norm": 1.134448089593032, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 9809 + }, + { + "epoch": 0.0981, + "grad_norm": 0.9440022631359163, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 9810 + }, + { + "epoch": 0.09811, + "grad_norm": 0.8747418766506534, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 9811 + }, + { + "epoch": 0.09812, + "grad_norm": 1.0271563855822783, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 9812 + }, + { + "epoch": 0.09813, + "grad_norm": 1.268250855234208, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 9813 + }, + { + "epoch": 0.09814, + "grad_norm": 0.9423284371731849, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9814 + }, + { + "epoch": 0.09815, + "grad_norm": 0.8587272806085712, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9815 + }, + { + "epoch": 0.09816, + "grad_norm": 0.8898000553630154, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 9816 + }, + { + "epoch": 0.09817, + "grad_norm": 0.8967370159090363, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 9817 + }, + { + "epoch": 0.09818, + "grad_norm": 0.9683907679710231, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 9818 + }, + { + "epoch": 0.09819, + "grad_norm": 1.1611928616933378, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 9819 + }, + { + "epoch": 0.0982, + "grad_norm": 0.9627567145629911, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 9820 + }, + { + "epoch": 0.09821, + "grad_norm": 1.0864907954219507, + "learning_rate": 0.003, + "loss": 4.076, + "step": 9821 + }, + { + "epoch": 0.09822, + "grad_norm": 1.3249140720038677, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 9822 + }, + { + "epoch": 0.09823, + "grad_norm": 0.8605074850339292, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 9823 + }, + { + "epoch": 0.09824, + "grad_norm": 0.9566425600412122, + "learning_rate": 0.003, + "loss": 4.097, + "step": 9824 + }, + { + "epoch": 0.09825, + "grad_norm": 1.0233355222981693, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9825 + }, + { + "epoch": 0.09826, + "grad_norm": 1.0786561064635678, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 9826 + }, + { + "epoch": 0.09827, + "grad_norm": 1.0514971996142433, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 9827 + }, + { + "epoch": 0.09828, + "grad_norm": 0.8623547190602233, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 9828 + }, + { + "epoch": 0.09829, + "grad_norm": 0.8781618673749649, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9829 + }, + { + "epoch": 0.0983, + "grad_norm": 0.8682088320906876, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9830 + }, + { + "epoch": 0.09831, + "grad_norm": 0.8891553003739594, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 9831 + }, + { + "epoch": 0.09832, + "grad_norm": 1.0895544072278256, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 9832 + }, + { + "epoch": 0.09833, + "grad_norm": 1.1615633672113128, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 9833 + }, + { + "epoch": 0.09834, + "grad_norm": 0.9838275297396515, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 9834 + }, + { + "epoch": 0.09835, + "grad_norm": 1.0531808921035346, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 9835 + }, + { + "epoch": 0.09836, + "grad_norm": 1.1225625590156698, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9836 + }, + { + "epoch": 0.09837, + "grad_norm": 1.047559273171705, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9837 + }, + { + "epoch": 0.09838, + "grad_norm": 1.0625120495027733, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 9838 + }, + { + "epoch": 0.09839, + "grad_norm": 1.1447218613408165, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 9839 + }, + { + "epoch": 0.0984, + "grad_norm": 0.9466230883186256, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 9840 + }, + { + "epoch": 0.09841, + "grad_norm": 0.9948433758958204, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 9841 + }, + { + "epoch": 0.09842, + "grad_norm": 1.0916342093786202, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9842 + }, + { + "epoch": 0.09843, + "grad_norm": 1.0810547158105994, + "learning_rate": 0.003, + "loss": 4.085, + "step": 9843 + }, + { + "epoch": 0.09844, + "grad_norm": 1.0666836813512826, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 9844 + }, + { + "epoch": 0.09845, + "grad_norm": 0.9364969036871016, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 9845 + }, + { + "epoch": 0.09846, + "grad_norm": 1.247977198047358, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 9846 + }, + { + "epoch": 0.09847, + "grad_norm": 1.0983305121758822, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 9847 + }, + { + "epoch": 0.09848, + "grad_norm": 0.9241457854629477, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 9848 + }, + { + "epoch": 0.09849, + "grad_norm": 0.8899279219685088, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 9849 + }, + { + "epoch": 0.0985, + "grad_norm": 0.9422816629698554, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 9850 + }, + { + "epoch": 0.09851, + "grad_norm": 0.9730853740755129, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9851 + }, + { + "epoch": 0.09852, + "grad_norm": 1.0334597856537537, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 9852 + }, + { + "epoch": 0.09853, + "grad_norm": 1.026542924409616, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 9853 + }, + { + "epoch": 0.09854, + "grad_norm": 1.191401021749489, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 9854 + }, + { + "epoch": 0.09855, + "grad_norm": 1.059077578402062, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9855 + }, + { + "epoch": 0.09856, + "grad_norm": 1.0498244668875298, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 9856 + }, + { + "epoch": 0.09857, + "grad_norm": 1.091092729660013, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 9857 + }, + { + "epoch": 0.09858, + "grad_norm": 0.9250542311162983, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 9858 + }, + { + "epoch": 0.09859, + "grad_norm": 1.0800039787638633, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 9859 + }, + { + "epoch": 0.0986, + "grad_norm": 1.041419214223572, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9860 + }, + { + "epoch": 0.09861, + "grad_norm": 0.9982888817464664, + "learning_rate": 0.003, + "loss": 4.088, + "step": 9861 + }, + { + "epoch": 0.09862, + "grad_norm": 1.1239145862062716, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 9862 + }, + { + "epoch": 0.09863, + "grad_norm": 0.9592019467214191, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9863 + }, + { + "epoch": 0.09864, + "grad_norm": 1.0352169689669506, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 9864 + }, + { + "epoch": 0.09865, + "grad_norm": 1.181841188508446, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 9865 + }, + { + "epoch": 0.09866, + "grad_norm": 0.9389016144596009, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 9866 + }, + { + "epoch": 0.09867, + "grad_norm": 0.8893717783577848, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 9867 + }, + { + "epoch": 0.09868, + "grad_norm": 0.9999511334783509, + "learning_rate": 0.003, + "loss": 4.071, + "step": 9868 + }, + { + "epoch": 0.09869, + "grad_norm": 1.3766579673159203, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 9869 + }, + { + "epoch": 0.0987, + "grad_norm": 0.8413732379628507, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 9870 + }, + { + "epoch": 0.09871, + "grad_norm": 0.7967590725793811, + "learning_rate": 0.003, + "loss": 4.1397, + "step": 9871 + }, + { + "epoch": 0.09872, + "grad_norm": 0.941707830097863, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 9872 + }, + { + "epoch": 0.09873, + "grad_norm": 0.9855273751841129, + "learning_rate": 0.003, + "loss": 4.086, + "step": 9873 + }, + { + "epoch": 0.09874, + "grad_norm": 0.8989862551825634, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9874 + }, + { + "epoch": 0.09875, + "grad_norm": 0.968402761982066, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 9875 + }, + { + "epoch": 0.09876, + "grad_norm": 1.0298264328815836, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 9876 + }, + { + "epoch": 0.09877, + "grad_norm": 1.0715370435837641, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 9877 + }, + { + "epoch": 0.09878, + "grad_norm": 0.9061930517874249, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9878 + }, + { + "epoch": 0.09879, + "grad_norm": 1.031786492000977, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 9879 + }, + { + "epoch": 0.0988, + "grad_norm": 1.3228565671620516, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 9880 + }, + { + "epoch": 0.09881, + "grad_norm": 1.035790465396129, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 9881 + }, + { + "epoch": 0.09882, + "grad_norm": 1.1699610602457922, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 9882 + }, + { + "epoch": 0.09883, + "grad_norm": 1.0555690693909043, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 9883 + }, + { + "epoch": 0.09884, + "grad_norm": 1.0713445946006626, + "learning_rate": 0.003, + "loss": 4.07, + "step": 9884 + }, + { + "epoch": 0.09885, + "grad_norm": 1.0685798383445284, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 9885 + }, + { + "epoch": 0.09886, + "grad_norm": 1.3401385584546417, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 9886 + }, + { + "epoch": 0.09887, + "grad_norm": 0.9107037359332328, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 9887 + }, + { + "epoch": 0.09888, + "grad_norm": 1.1599530666035878, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 9888 + }, + { + "epoch": 0.09889, + "grad_norm": 1.0285469001260248, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 9889 + }, + { + "epoch": 0.0989, + "grad_norm": 0.9967838845583898, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9890 + }, + { + "epoch": 0.09891, + "grad_norm": 1.2711079083653072, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 9891 + }, + { + "epoch": 0.09892, + "grad_norm": 0.9072541972447289, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 9892 + }, + { + "epoch": 0.09893, + "grad_norm": 0.9675167406986386, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 9893 + }, + { + "epoch": 0.09894, + "grad_norm": 0.9826844070764257, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 9894 + }, + { + "epoch": 0.09895, + "grad_norm": 1.1555014469207152, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 9895 + }, + { + "epoch": 0.09896, + "grad_norm": 0.8705076015969703, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 9896 + }, + { + "epoch": 0.09897, + "grad_norm": 1.0606028357144373, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 9897 + }, + { + "epoch": 0.09898, + "grad_norm": 1.0431107749585196, + "learning_rate": 0.003, + "loss": 4.079, + "step": 9898 + }, + { + "epoch": 0.09899, + "grad_norm": 1.1940626392877045, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9899 + }, + { + "epoch": 0.099, + "grad_norm": 1.0231246381632437, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 9900 + }, + { + "epoch": 0.09901, + "grad_norm": 1.1622210994726128, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9901 + }, + { + "epoch": 0.09902, + "grad_norm": 0.9443388666248479, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 9902 + }, + { + "epoch": 0.09903, + "grad_norm": 1.0384709982323672, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 9903 + }, + { + "epoch": 0.09904, + "grad_norm": 0.9723894387638847, + "learning_rate": 0.003, + "loss": 4.064, + "step": 9904 + }, + { + "epoch": 0.09905, + "grad_norm": 1.1523154218331038, + "learning_rate": 0.003, + "loss": 4.089, + "step": 9905 + }, + { + "epoch": 0.09906, + "grad_norm": 1.0547854344378345, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 9906 + }, + { + "epoch": 0.09907, + "grad_norm": 1.182257869541173, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9907 + }, + { + "epoch": 0.09908, + "grad_norm": 1.161665767642822, + "learning_rate": 0.003, + "loss": 4.102, + "step": 9908 + }, + { + "epoch": 0.09909, + "grad_norm": 0.9087642677557614, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 9909 + }, + { + "epoch": 0.0991, + "grad_norm": 1.0313878368391898, + "learning_rate": 0.003, + "loss": 4.093, + "step": 9910 + }, + { + "epoch": 0.09911, + "grad_norm": 1.2847136513646835, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 9911 + }, + { + "epoch": 0.09912, + "grad_norm": 0.8904708929002996, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 9912 + }, + { + "epoch": 0.09913, + "grad_norm": 1.0112171424842842, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9913 + }, + { + "epoch": 0.09914, + "grad_norm": 0.9794967717604602, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 9914 + }, + { + "epoch": 0.09915, + "grad_norm": 0.9423290437307489, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 9915 + }, + { + "epoch": 0.09916, + "grad_norm": 0.950907824636009, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 9916 + }, + { + "epoch": 0.09917, + "grad_norm": 1.0219884227601401, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 9917 + }, + { + "epoch": 0.09918, + "grad_norm": 1.032340531956216, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 9918 + }, + { + "epoch": 0.09919, + "grad_norm": 0.9004643268011698, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 9919 + }, + { + "epoch": 0.0992, + "grad_norm": 1.0606211698418486, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 9920 + }, + { + "epoch": 0.09921, + "grad_norm": 1.3653476752568316, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 9921 + }, + { + "epoch": 0.09922, + "grad_norm": 0.8159267949493182, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 9922 + }, + { + "epoch": 0.09923, + "grad_norm": 0.8093245515380362, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 9923 + }, + { + "epoch": 0.09924, + "grad_norm": 0.9136880506548397, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 9924 + }, + { + "epoch": 0.09925, + "grad_norm": 0.8854387914223482, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 9925 + }, + { + "epoch": 0.09926, + "grad_norm": 1.0055098961573377, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 9926 + }, + { + "epoch": 0.09927, + "grad_norm": 0.9799629554615119, + "learning_rate": 0.003, + "loss": 4.1182, + "step": 9927 + }, + { + "epoch": 0.09928, + "grad_norm": 1.097232317085835, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 9928 + }, + { + "epoch": 0.09929, + "grad_norm": 1.2181303331479505, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 9929 + }, + { + "epoch": 0.0993, + "grad_norm": 1.41744952211023, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 9930 + }, + { + "epoch": 0.09931, + "grad_norm": 0.8845168557408566, + "learning_rate": 0.003, + "loss": 4.081, + "step": 9931 + }, + { + "epoch": 0.09932, + "grad_norm": 0.9049290312336111, + "learning_rate": 0.003, + "loss": 4.088, + "step": 9932 + }, + { + "epoch": 0.09933, + "grad_norm": 1.1600752655306843, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 9933 + }, + { + "epoch": 0.09934, + "grad_norm": 1.0319033481213138, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 9934 + }, + { + "epoch": 0.09935, + "grad_norm": 1.2341422592136662, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 9935 + }, + { + "epoch": 0.09936, + "grad_norm": 1.0628728338770057, + "learning_rate": 0.003, + "loss": 4.088, + "step": 9936 + }, + { + "epoch": 0.09937, + "grad_norm": 0.8801535749313547, + "learning_rate": 0.003, + "loss": 4.052, + "step": 9937 + }, + { + "epoch": 0.09938, + "grad_norm": 0.8196753503087768, + "learning_rate": 0.003, + "loss": 4.09, + "step": 9938 + }, + { + "epoch": 0.09939, + "grad_norm": 0.9592522974642151, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 9939 + }, + { + "epoch": 0.0994, + "grad_norm": 1.0721297761126327, + "learning_rate": 0.003, + "loss": 4.115, + "step": 9940 + }, + { + "epoch": 0.09941, + "grad_norm": 1.0763492581657323, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 9941 + }, + { + "epoch": 0.09942, + "grad_norm": 1.019164330036407, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 9942 + }, + { + "epoch": 0.09943, + "grad_norm": 1.2248432437361976, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9943 + }, + { + "epoch": 0.09944, + "grad_norm": 1.0747625128630949, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 9944 + }, + { + "epoch": 0.09945, + "grad_norm": 0.9570097418411981, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 9945 + }, + { + "epoch": 0.09946, + "grad_norm": 0.9943121246735858, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 9946 + }, + { + "epoch": 0.09947, + "grad_norm": 1.2372355073906356, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 9947 + }, + { + "epoch": 0.09948, + "grad_norm": 0.8502611811001046, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 9948 + }, + { + "epoch": 0.09949, + "grad_norm": 0.9231335871633486, + "learning_rate": 0.003, + "loss": 4.091, + "step": 9949 + }, + { + "epoch": 0.0995, + "grad_norm": 1.0621020837735249, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 9950 + }, + { + "epoch": 0.09951, + "grad_norm": 1.1092075040539509, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 9951 + }, + { + "epoch": 0.09952, + "grad_norm": 1.1180875245705675, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9952 + }, + { + "epoch": 0.09953, + "grad_norm": 1.0583089223558193, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 9953 + }, + { + "epoch": 0.09954, + "grad_norm": 1.3656489755235695, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 9954 + }, + { + "epoch": 0.09955, + "grad_norm": 0.9216137223262066, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 9955 + }, + { + "epoch": 0.09956, + "grad_norm": 0.8331207661096087, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 9956 + }, + { + "epoch": 0.09957, + "grad_norm": 0.9411322885803222, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 9957 + }, + { + "epoch": 0.09958, + "grad_norm": 0.9765577007820736, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 9958 + }, + { + "epoch": 0.09959, + "grad_norm": 0.9833086407810087, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 9959 + }, + { + "epoch": 0.0996, + "grad_norm": 1.208131243797863, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 9960 + }, + { + "epoch": 0.09961, + "grad_norm": 0.9410870086120185, + "learning_rate": 0.003, + "loss": 4.094, + "step": 9961 + }, + { + "epoch": 0.09962, + "grad_norm": 0.9901662898897093, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 9962 + }, + { + "epoch": 0.09963, + "grad_norm": 1.0595283366933754, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 9963 + }, + { + "epoch": 0.09964, + "grad_norm": 1.1230131079391967, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 9964 + }, + { + "epoch": 0.09965, + "grad_norm": 1.408656633378828, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 9965 + }, + { + "epoch": 0.09966, + "grad_norm": 1.0494066026723656, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 9966 + }, + { + "epoch": 0.09967, + "grad_norm": 1.0510949632983806, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 9967 + }, + { + "epoch": 0.09968, + "grad_norm": 0.9184349300356729, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 9968 + }, + { + "epoch": 0.09969, + "grad_norm": 0.912619317530374, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 9969 + }, + { + "epoch": 0.0997, + "grad_norm": 0.9253745452194618, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 9970 + }, + { + "epoch": 0.09971, + "grad_norm": 0.9414528249595521, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 9971 + }, + { + "epoch": 0.09972, + "grad_norm": 1.212415753181934, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 9972 + }, + { + "epoch": 0.09973, + "grad_norm": 1.3032259078570259, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 9973 + }, + { + "epoch": 0.09974, + "grad_norm": 0.8834224407852096, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 9974 + }, + { + "epoch": 0.09975, + "grad_norm": 1.0257775503141435, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 9975 + }, + { + "epoch": 0.09976, + "grad_norm": 1.0027851181438205, + "learning_rate": 0.003, + "loss": 4.066, + "step": 9976 + }, + { + "epoch": 0.09977, + "grad_norm": 1.1054062839337309, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 9977 + }, + { + "epoch": 0.09978, + "grad_norm": 0.9983871843931804, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 9978 + }, + { + "epoch": 0.09979, + "grad_norm": 1.1469658774556046, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 9979 + }, + { + "epoch": 0.0998, + "grad_norm": 1.0417510639490304, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 9980 + }, + { + "epoch": 0.09981, + "grad_norm": 0.8997050282680549, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 9981 + }, + { + "epoch": 0.09982, + "grad_norm": 0.9795564166461124, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 9982 + }, + { + "epoch": 0.09983, + "grad_norm": 1.1265795885230583, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 9983 + }, + { + "epoch": 0.09984, + "grad_norm": 1.075755261601754, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 9984 + }, + { + "epoch": 0.09985, + "grad_norm": 1.1117966143075422, + "learning_rate": 0.003, + "loss": 4.105, + "step": 9985 + }, + { + "epoch": 0.09986, + "grad_norm": 0.8947693590206178, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 9986 + }, + { + "epoch": 0.09987, + "grad_norm": 0.964118687177187, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 9987 + }, + { + "epoch": 0.09988, + "grad_norm": 1.1466654575825213, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 9988 + }, + { + "epoch": 0.09989, + "grad_norm": 0.9371466526373202, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9989 + }, + { + "epoch": 0.0999, + "grad_norm": 1.1021242856154043, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 9990 + }, + { + "epoch": 0.09991, + "grad_norm": 1.4409488657540674, + "learning_rate": 0.003, + "loss": 4.121, + "step": 9991 + }, + { + "epoch": 0.09992, + "grad_norm": 0.9081140092271789, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 9992 + }, + { + "epoch": 0.09993, + "grad_norm": 1.0604218435107378, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 9993 + }, + { + "epoch": 0.09994, + "grad_norm": 1.2125286721128687, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 9994 + }, + { + "epoch": 0.09995, + "grad_norm": 1.0829501150938283, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 9995 + }, + { + "epoch": 0.09996, + "grad_norm": 1.0733185017791256, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 9996 + }, + { + "epoch": 0.09997, + "grad_norm": 1.021879556888008, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 9997 + }, + { + "epoch": 0.09998, + "grad_norm": 1.097873921598905, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 9998 + }, + { + "epoch": 0.09999, + "grad_norm": 1.0928565866659528, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 9999 + }, + { + "epoch": 0.1, + "grad_norm": 0.9146824902563694, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 10000 + }, + { + "epoch": 0.10001, + "grad_norm": 1.0268828767480918, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 10001 + }, + { + "epoch": 0.10002, + "grad_norm": 0.9456355767793398, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 10002 + }, + { + "epoch": 0.10003, + "grad_norm": 0.95383283914356, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 10003 + }, + { + "epoch": 0.10004, + "grad_norm": 1.045290937365292, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10004 + }, + { + "epoch": 0.10005, + "grad_norm": 0.9788174490820064, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10005 + }, + { + "epoch": 0.10006, + "grad_norm": 1.1152617835523273, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10006 + }, + { + "epoch": 0.10007, + "grad_norm": 1.0774775903166232, + "learning_rate": 0.003, + "loss": 4.068, + "step": 10007 + }, + { + "epoch": 0.10008, + "grad_norm": 1.2010923165312688, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10008 + }, + { + "epoch": 0.10009, + "grad_norm": 0.9615549697604012, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 10009 + }, + { + "epoch": 0.1001, + "grad_norm": 0.8606186327961719, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 10010 + }, + { + "epoch": 0.10011, + "grad_norm": 0.9641733347424303, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 10011 + }, + { + "epoch": 0.10012, + "grad_norm": 1.1026647643286422, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10012 + }, + { + "epoch": 0.10013, + "grad_norm": 1.0701035577711429, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 10013 + }, + { + "epoch": 0.10014, + "grad_norm": 1.0817395215435701, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 10014 + }, + { + "epoch": 0.10015, + "grad_norm": 1.4023794761788895, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 10015 + }, + { + "epoch": 0.10016, + "grad_norm": 0.8702730093122781, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 10016 + }, + { + "epoch": 0.10017, + "grad_norm": 0.7570220498972725, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10017 + }, + { + "epoch": 0.10018, + "grad_norm": 0.8707038500254759, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 10018 + }, + { + "epoch": 0.10019, + "grad_norm": 1.010728470425738, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 10019 + }, + { + "epoch": 0.1002, + "grad_norm": 0.9979899408508959, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 10020 + }, + { + "epoch": 0.10021, + "grad_norm": 1.0674101631325568, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 10021 + }, + { + "epoch": 0.10022, + "grad_norm": 1.0266913018745583, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10022 + }, + { + "epoch": 0.10023, + "grad_norm": 1.0638117707816785, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 10023 + }, + { + "epoch": 0.10024, + "grad_norm": 1.133369886133261, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 10024 + }, + { + "epoch": 0.10025, + "grad_norm": 1.102307846355811, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 10025 + }, + { + "epoch": 0.10026, + "grad_norm": 1.3400225966662347, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10026 + }, + { + "epoch": 0.10027, + "grad_norm": 1.0211547903184859, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 10027 + }, + { + "epoch": 0.10028, + "grad_norm": 1.1603850478475943, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 10028 + }, + { + "epoch": 0.10029, + "grad_norm": 1.148079705243166, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 10029 + }, + { + "epoch": 0.1003, + "grad_norm": 0.9117292209524029, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 10030 + }, + { + "epoch": 0.10031, + "grad_norm": 0.9445865002921869, + "learning_rate": 0.003, + "loss": 4.122, + "step": 10031 + }, + { + "epoch": 0.10032, + "grad_norm": 0.9161171568886268, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 10032 + }, + { + "epoch": 0.10033, + "grad_norm": 0.85950044194575, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 10033 + }, + { + "epoch": 0.10034, + "grad_norm": 0.9072001134697321, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 10034 + }, + { + "epoch": 0.10035, + "grad_norm": 1.1474126252804075, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10035 + }, + { + "epoch": 0.10036, + "grad_norm": 0.9761134534605453, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 10036 + }, + { + "epoch": 0.10037, + "grad_norm": 1.0237415513918258, + "learning_rate": 0.003, + "loss": 4.073, + "step": 10037 + }, + { + "epoch": 0.10038, + "grad_norm": 1.148438737648086, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 10038 + }, + { + "epoch": 0.10039, + "grad_norm": 1.0514373112344126, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10039 + }, + { + "epoch": 0.1004, + "grad_norm": 1.1847286970538577, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 10040 + }, + { + "epoch": 0.10041, + "grad_norm": 0.8922451988552241, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10041 + }, + { + "epoch": 0.10042, + "grad_norm": 0.8397773653104834, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10042 + }, + { + "epoch": 0.10043, + "grad_norm": 0.9089542086627763, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 10043 + }, + { + "epoch": 0.10044, + "grad_norm": 1.0073280068294317, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 10044 + }, + { + "epoch": 0.10045, + "grad_norm": 1.0440598241093555, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10045 + }, + { + "epoch": 0.10046, + "grad_norm": 1.0485056600329703, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 10046 + }, + { + "epoch": 0.10047, + "grad_norm": 1.2429836270576817, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 10047 + }, + { + "epoch": 0.10048, + "grad_norm": 1.2106150367464596, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 10048 + }, + { + "epoch": 0.10049, + "grad_norm": 1.2553698920040894, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 10049 + }, + { + "epoch": 0.1005, + "grad_norm": 1.0468634756369095, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 10050 + }, + { + "epoch": 0.10051, + "grad_norm": 1.118778052143475, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10051 + }, + { + "epoch": 0.10052, + "grad_norm": 1.179218879697393, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 10052 + }, + { + "epoch": 0.10053, + "grad_norm": 0.9200813106865441, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10053 + }, + { + "epoch": 0.10054, + "grad_norm": 0.8562476127183147, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 10054 + }, + { + "epoch": 0.10055, + "grad_norm": 0.9908634064538296, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10055 + }, + { + "epoch": 0.10056, + "grad_norm": 1.3614880394829219, + "learning_rate": 0.003, + "loss": 4.12, + "step": 10056 + }, + { + "epoch": 0.10057, + "grad_norm": 0.6795058628261522, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 10057 + }, + { + "epoch": 0.10058, + "grad_norm": 0.6725766270507677, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10058 + }, + { + "epoch": 0.10059, + "grad_norm": 0.7021007029637584, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10059 + }, + { + "epoch": 0.1006, + "grad_norm": 0.7029672751184473, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 10060 + }, + { + "epoch": 0.10061, + "grad_norm": 0.86274187466931, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 10061 + }, + { + "epoch": 0.10062, + "grad_norm": 1.1476605169788783, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 10062 + }, + { + "epoch": 0.10063, + "grad_norm": 1.1736377656589307, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 10063 + }, + { + "epoch": 0.10064, + "grad_norm": 1.0192130436130324, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 10064 + }, + { + "epoch": 0.10065, + "grad_norm": 1.2950673831269726, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 10065 + }, + { + "epoch": 0.10066, + "grad_norm": 0.8946009368675681, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10066 + }, + { + "epoch": 0.10067, + "grad_norm": 0.9034084243233744, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10067 + }, + { + "epoch": 0.10068, + "grad_norm": 0.9279795526086888, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10068 + }, + { + "epoch": 0.10069, + "grad_norm": 0.941069604433777, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 10069 + }, + { + "epoch": 0.1007, + "grad_norm": 0.9419306165269437, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10070 + }, + { + "epoch": 0.10071, + "grad_norm": 0.9249349363580369, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10071 + }, + { + "epoch": 0.10072, + "grad_norm": 1.0579196675587663, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10072 + }, + { + "epoch": 0.10073, + "grad_norm": 1.0641190190848893, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 10073 + }, + { + "epoch": 0.10074, + "grad_norm": 1.2438602971074477, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 10074 + }, + { + "epoch": 0.10075, + "grad_norm": 1.0564497520727791, + "learning_rate": 0.003, + "loss": 4.109, + "step": 10075 + }, + { + "epoch": 0.10076, + "grad_norm": 1.3090603530576794, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10076 + }, + { + "epoch": 0.10077, + "grad_norm": 0.7533461669269756, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 10077 + }, + { + "epoch": 0.10078, + "grad_norm": 0.7061355548387933, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 10078 + }, + { + "epoch": 0.10079, + "grad_norm": 0.8644525263853029, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10079 + }, + { + "epoch": 0.1008, + "grad_norm": 1.10452779458274, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 10080 + }, + { + "epoch": 0.10081, + "grad_norm": 1.3756511095540078, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 10081 + }, + { + "epoch": 0.10082, + "grad_norm": 1.038088737997971, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 10082 + }, + { + "epoch": 0.10083, + "grad_norm": 1.1207040034663152, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 10083 + }, + { + "epoch": 0.10084, + "grad_norm": 0.850821620028518, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 10084 + }, + { + "epoch": 0.10085, + "grad_norm": 1.0748025908828802, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10085 + }, + { + "epoch": 0.10086, + "grad_norm": 1.3549025239319474, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10086 + }, + { + "epoch": 0.10087, + "grad_norm": 1.0747448453376909, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 10087 + }, + { + "epoch": 0.10088, + "grad_norm": 1.2456306464923101, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 10088 + }, + { + "epoch": 0.10089, + "grad_norm": 0.9472976337664214, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 10089 + }, + { + "epoch": 0.1009, + "grad_norm": 0.9815403359612819, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 10090 + }, + { + "epoch": 0.10091, + "grad_norm": 0.9741684116233433, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10091 + }, + { + "epoch": 0.10092, + "grad_norm": 0.9476855315324259, + "learning_rate": 0.003, + "loss": 4.096, + "step": 10092 + }, + { + "epoch": 0.10093, + "grad_norm": 0.9811508699302622, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10093 + }, + { + "epoch": 0.10094, + "grad_norm": 1.1541788163650089, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 10094 + }, + { + "epoch": 0.10095, + "grad_norm": 1.1472639269086242, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 10095 + }, + { + "epoch": 0.10096, + "grad_norm": 1.0555714689128464, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10096 + }, + { + "epoch": 0.10097, + "grad_norm": 0.9083517008773805, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 10097 + }, + { + "epoch": 0.10098, + "grad_norm": 0.931319184757005, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10098 + }, + { + "epoch": 0.10099, + "grad_norm": 1.2061065602641496, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10099 + }, + { + "epoch": 0.101, + "grad_norm": 1.0790936601780008, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 10100 + }, + { + "epoch": 0.10101, + "grad_norm": 1.049700687920486, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10101 + }, + { + "epoch": 0.10102, + "grad_norm": 1.0588165543231847, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 10102 + }, + { + "epoch": 0.10103, + "grad_norm": 0.8785533801322097, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 10103 + }, + { + "epoch": 0.10104, + "grad_norm": 0.947148926018248, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 10104 + }, + { + "epoch": 0.10105, + "grad_norm": 1.0594563152689682, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10105 + }, + { + "epoch": 0.10106, + "grad_norm": 1.0225956027823544, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10106 + }, + { + "epoch": 0.10107, + "grad_norm": 1.0308601830350073, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 10107 + }, + { + "epoch": 0.10108, + "grad_norm": 1.0350078878450226, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 10108 + }, + { + "epoch": 0.10109, + "grad_norm": 1.1686695733740247, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 10109 + }, + { + "epoch": 0.1011, + "grad_norm": 1.2115978839317425, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10110 + }, + { + "epoch": 0.10111, + "grad_norm": 0.978402576765035, + "learning_rate": 0.003, + "loss": 4.103, + "step": 10111 + }, + { + "epoch": 0.10112, + "grad_norm": 1.0623646734983812, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 10112 + }, + { + "epoch": 0.10113, + "grad_norm": 0.9614458949786193, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10113 + }, + { + "epoch": 0.10114, + "grad_norm": 1.179118562016018, + "learning_rate": 0.003, + "loss": 4.063, + "step": 10114 + }, + { + "epoch": 0.10115, + "grad_norm": 1.0233642500878612, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 10115 + }, + { + "epoch": 0.10116, + "grad_norm": 1.00109985164738, + "learning_rate": 0.003, + "loss": 4.079, + "step": 10116 + }, + { + "epoch": 0.10117, + "grad_norm": 1.2153279913183614, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 10117 + }, + { + "epoch": 0.10118, + "grad_norm": 0.9253105426894055, + "learning_rate": 0.003, + "loss": 4.111, + "step": 10118 + }, + { + "epoch": 0.10119, + "grad_norm": 1.201988642511759, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10119 + }, + { + "epoch": 0.1012, + "grad_norm": 0.8945884294792387, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 10120 + }, + { + "epoch": 0.10121, + "grad_norm": 0.8959801151367283, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10121 + }, + { + "epoch": 0.10122, + "grad_norm": 1.0389605252243181, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 10122 + }, + { + "epoch": 0.10123, + "grad_norm": 1.1853235896574628, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10123 + }, + { + "epoch": 0.10124, + "grad_norm": 1.1175599651017967, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10124 + }, + { + "epoch": 0.10125, + "grad_norm": 1.1730052020802815, + "learning_rate": 0.003, + "loss": 4.101, + "step": 10125 + }, + { + "epoch": 0.10126, + "grad_norm": 1.0540782103922477, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10126 + }, + { + "epoch": 0.10127, + "grad_norm": 1.1471297550451278, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 10127 + }, + { + "epoch": 0.10128, + "grad_norm": 1.113435608277784, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 10128 + }, + { + "epoch": 0.10129, + "grad_norm": 1.1059156980127567, + "learning_rate": 0.003, + "loss": 4.105, + "step": 10129 + }, + { + "epoch": 0.1013, + "grad_norm": 0.8882002679244408, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10130 + }, + { + "epoch": 0.10131, + "grad_norm": 0.9154448816742513, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10131 + }, + { + "epoch": 0.10132, + "grad_norm": 0.9450551054713623, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 10132 + }, + { + "epoch": 0.10133, + "grad_norm": 0.9595142116108409, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10133 + }, + { + "epoch": 0.10134, + "grad_norm": 0.9736572446471938, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10134 + }, + { + "epoch": 0.10135, + "grad_norm": 1.0390522302798173, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10135 + }, + { + "epoch": 0.10136, + "grad_norm": 1.086694019268405, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10136 + }, + { + "epoch": 0.10137, + "grad_norm": 1.4802280932553884, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10137 + }, + { + "epoch": 0.10138, + "grad_norm": 0.8935572863107328, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10138 + }, + { + "epoch": 0.10139, + "grad_norm": 0.9129477796554756, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 10139 + }, + { + "epoch": 0.1014, + "grad_norm": 1.0712559348135158, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 10140 + }, + { + "epoch": 0.10141, + "grad_norm": 1.129298993945068, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 10141 + }, + { + "epoch": 0.10142, + "grad_norm": 1.0063431597085504, + "learning_rate": 0.003, + "loss": 4.069, + "step": 10142 + }, + { + "epoch": 0.10143, + "grad_norm": 1.0164759574916866, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 10143 + }, + { + "epoch": 0.10144, + "grad_norm": 1.102176942327319, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10144 + }, + { + "epoch": 0.10145, + "grad_norm": 0.8846991416216525, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 10145 + }, + { + "epoch": 0.10146, + "grad_norm": 0.7954991760065576, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 10146 + }, + { + "epoch": 0.10147, + "grad_norm": 0.9389225672902068, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 10147 + }, + { + "epoch": 0.10148, + "grad_norm": 0.9979062071789803, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 10148 + }, + { + "epoch": 0.10149, + "grad_norm": 0.9358394152637304, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10149 + }, + { + "epoch": 0.1015, + "grad_norm": 1.2030745401236895, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10150 + }, + { + "epoch": 0.10151, + "grad_norm": 1.19363323179694, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10151 + }, + { + "epoch": 0.10152, + "grad_norm": 1.051166326109995, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 10152 + }, + { + "epoch": 0.10153, + "grad_norm": 1.0612612415288574, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 10153 + }, + { + "epoch": 0.10154, + "grad_norm": 1.0366017630808055, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10154 + }, + { + "epoch": 0.10155, + "grad_norm": 1.0829838583007372, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 10155 + }, + { + "epoch": 0.10156, + "grad_norm": 1.2320653205953953, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 10156 + }, + { + "epoch": 0.10157, + "grad_norm": 1.1889966899388862, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 10157 + }, + { + "epoch": 0.10158, + "grad_norm": 1.0696254422253109, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10158 + }, + { + "epoch": 0.10159, + "grad_norm": 1.0566907846701998, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 10159 + }, + { + "epoch": 0.1016, + "grad_norm": 1.0525638580660244, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 10160 + }, + { + "epoch": 0.10161, + "grad_norm": 1.046979807820574, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10161 + }, + { + "epoch": 0.10162, + "grad_norm": 0.9848404406583341, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10162 + }, + { + "epoch": 0.10163, + "grad_norm": 1.0285044268005161, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10163 + }, + { + "epoch": 0.10164, + "grad_norm": 1.0192066947627048, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 10164 + }, + { + "epoch": 0.10165, + "grad_norm": 1.1070606086478363, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 10165 + }, + { + "epoch": 0.10166, + "grad_norm": 1.0624502625108392, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10166 + }, + { + "epoch": 0.10167, + "grad_norm": 1.1263826776460912, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10167 + }, + { + "epoch": 0.10168, + "grad_norm": 0.886296216319193, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10168 + }, + { + "epoch": 0.10169, + "grad_norm": 1.0451723259522945, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10169 + }, + { + "epoch": 0.1017, + "grad_norm": 1.2995896070350659, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10170 + }, + { + "epoch": 0.10171, + "grad_norm": 1.033886351063779, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10171 + }, + { + "epoch": 0.10172, + "grad_norm": 1.1153685517458167, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 10172 + }, + { + "epoch": 0.10173, + "grad_norm": 1.0772592426847833, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10173 + }, + { + "epoch": 0.10174, + "grad_norm": 1.13970670521581, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 10174 + }, + { + "epoch": 0.10175, + "grad_norm": 0.8756802427578665, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10175 + }, + { + "epoch": 0.10176, + "grad_norm": 0.8661317521414144, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 10176 + }, + { + "epoch": 0.10177, + "grad_norm": 0.9875363098650536, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 10177 + }, + { + "epoch": 0.10178, + "grad_norm": 1.1709545498623426, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10178 + }, + { + "epoch": 0.10179, + "grad_norm": 1.0144311108168578, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10179 + }, + { + "epoch": 0.1018, + "grad_norm": 1.2533092220049373, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10180 + }, + { + "epoch": 0.10181, + "grad_norm": 0.8942442154370184, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10181 + }, + { + "epoch": 0.10182, + "grad_norm": 1.0521845568594967, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 10182 + }, + { + "epoch": 0.10183, + "grad_norm": 1.1139008203198577, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10183 + }, + { + "epoch": 0.10184, + "grad_norm": 0.955052265077462, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 10184 + }, + { + "epoch": 0.10185, + "grad_norm": 1.2294865872901575, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 10185 + }, + { + "epoch": 0.10186, + "grad_norm": 1.0405948014225823, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10186 + }, + { + "epoch": 0.10187, + "grad_norm": 1.068606777390599, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 10187 + }, + { + "epoch": 0.10188, + "grad_norm": 1.0324240707550445, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10188 + }, + { + "epoch": 0.10189, + "grad_norm": 1.0255280020860909, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 10189 + }, + { + "epoch": 0.1019, + "grad_norm": 0.9186025646322838, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10190 + }, + { + "epoch": 0.10191, + "grad_norm": 0.88989047653086, + "learning_rate": 0.003, + "loss": 4.093, + "step": 10191 + }, + { + "epoch": 0.10192, + "grad_norm": 1.1633171294690876, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 10192 + }, + { + "epoch": 0.10193, + "grad_norm": 1.2427486358359723, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 10193 + }, + { + "epoch": 0.10194, + "grad_norm": 1.0586470094561826, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10194 + }, + { + "epoch": 0.10195, + "grad_norm": 1.0036822636108969, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 10195 + }, + { + "epoch": 0.10196, + "grad_norm": 0.9466215049233101, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 10196 + }, + { + "epoch": 0.10197, + "grad_norm": 0.8112351254656962, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10197 + }, + { + "epoch": 0.10198, + "grad_norm": 0.819974365220033, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 10198 + }, + { + "epoch": 0.10199, + "grad_norm": 0.9933425025429821, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10199 + }, + { + "epoch": 0.102, + "grad_norm": 1.0531125515747866, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10200 + }, + { + "epoch": 0.10201, + "grad_norm": 1.0550511348110323, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10201 + }, + { + "epoch": 0.10202, + "grad_norm": 1.0661497923251, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10202 + }, + { + "epoch": 0.10203, + "grad_norm": 1.220413862454087, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 10203 + }, + { + "epoch": 0.10204, + "grad_norm": 1.195159818899411, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 10204 + }, + { + "epoch": 0.10205, + "grad_norm": 1.0453481471485864, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 10205 + }, + { + "epoch": 0.10206, + "grad_norm": 1.2132010685629997, + "learning_rate": 0.003, + "loss": 4.102, + "step": 10206 + }, + { + "epoch": 0.10207, + "grad_norm": 0.8484088484915983, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 10207 + }, + { + "epoch": 0.10208, + "grad_norm": 0.7164134447580448, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 10208 + }, + { + "epoch": 0.10209, + "grad_norm": 0.8905702473288374, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 10209 + }, + { + "epoch": 0.1021, + "grad_norm": 1.0567582830849451, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10210 + }, + { + "epoch": 0.10211, + "grad_norm": 1.2831260130563622, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 10211 + }, + { + "epoch": 0.10212, + "grad_norm": 0.9130536063331267, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10212 + }, + { + "epoch": 0.10213, + "grad_norm": 0.9842667361773123, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 10213 + }, + { + "epoch": 0.10214, + "grad_norm": 1.2213307026626754, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10214 + }, + { + "epoch": 0.10215, + "grad_norm": 1.09106396917812, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 10215 + }, + { + "epoch": 0.10216, + "grad_norm": 0.8701974848383143, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 10216 + }, + { + "epoch": 0.10217, + "grad_norm": 0.9028785450684711, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 10217 + }, + { + "epoch": 0.10218, + "grad_norm": 0.9895049497286713, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 10218 + }, + { + "epoch": 0.10219, + "grad_norm": 0.9976561368167319, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 10219 + }, + { + "epoch": 0.1022, + "grad_norm": 1.1566379033564083, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 10220 + }, + { + "epoch": 0.10221, + "grad_norm": 0.9436125666687795, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 10221 + }, + { + "epoch": 0.10222, + "grad_norm": 1.1745810965985861, + "learning_rate": 0.003, + "loss": 4.096, + "step": 10222 + }, + { + "epoch": 0.10223, + "grad_norm": 1.1161877928907953, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 10223 + }, + { + "epoch": 0.10224, + "grad_norm": 1.1101784497581406, + "learning_rate": 0.003, + "loss": 4.098, + "step": 10224 + }, + { + "epoch": 0.10225, + "grad_norm": 1.166479137680796, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 10225 + }, + { + "epoch": 0.10226, + "grad_norm": 1.0999501771547044, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 10226 + }, + { + "epoch": 0.10227, + "grad_norm": 1.1405389566462554, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 10227 + }, + { + "epoch": 0.10228, + "grad_norm": 1.237424322483817, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 10228 + }, + { + "epoch": 0.10229, + "grad_norm": 0.9639553122447139, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 10229 + }, + { + "epoch": 0.1023, + "grad_norm": 1.0161290893618329, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10230 + }, + { + "epoch": 0.10231, + "grad_norm": 1.157806573153118, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 10231 + }, + { + "epoch": 0.10232, + "grad_norm": 1.1552991257125518, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 10232 + }, + { + "epoch": 0.10233, + "grad_norm": 1.0844941301760342, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 10233 + }, + { + "epoch": 0.10234, + "grad_norm": 0.9783976398030328, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 10234 + }, + { + "epoch": 0.10235, + "grad_norm": 1.0167830602919732, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10235 + }, + { + "epoch": 0.10236, + "grad_norm": 1.2164144746685364, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 10236 + }, + { + "epoch": 0.10237, + "grad_norm": 0.9709684203383907, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 10237 + }, + { + "epoch": 0.10238, + "grad_norm": 1.0206937521127843, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 10238 + }, + { + "epoch": 0.10239, + "grad_norm": 1.115053000183982, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 10239 + }, + { + "epoch": 0.1024, + "grad_norm": 0.9632605456430765, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10240 + }, + { + "epoch": 0.10241, + "grad_norm": 0.9223306760546657, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 10241 + }, + { + "epoch": 0.10242, + "grad_norm": 1.0350416435917151, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10242 + }, + { + "epoch": 0.10243, + "grad_norm": 0.9740341906220983, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 10243 + }, + { + "epoch": 0.10244, + "grad_norm": 1.0309607354313888, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10244 + }, + { + "epoch": 0.10245, + "grad_norm": 1.2098078378544046, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10245 + }, + { + "epoch": 0.10246, + "grad_norm": 1.181999436226006, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 10246 + }, + { + "epoch": 0.10247, + "grad_norm": 1.0574827260952755, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 10247 + }, + { + "epoch": 0.10248, + "grad_norm": 1.0257693998664565, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 10248 + }, + { + "epoch": 0.10249, + "grad_norm": 0.9554869482772075, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10249 + }, + { + "epoch": 0.1025, + "grad_norm": 1.0132475399789136, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10250 + }, + { + "epoch": 0.10251, + "grad_norm": 1.0895426483345603, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 10251 + }, + { + "epoch": 0.10252, + "grad_norm": 1.1015345937455574, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 10252 + }, + { + "epoch": 0.10253, + "grad_norm": 0.9037748854628284, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10253 + }, + { + "epoch": 0.10254, + "grad_norm": 1.0414946280019493, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10254 + }, + { + "epoch": 0.10255, + "grad_norm": 1.3181367319322463, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10255 + }, + { + "epoch": 0.10256, + "grad_norm": 0.8784215981738813, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 10256 + }, + { + "epoch": 0.10257, + "grad_norm": 0.794653621913254, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 10257 + }, + { + "epoch": 0.10258, + "grad_norm": 0.9681340816585842, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 10258 + }, + { + "epoch": 0.10259, + "grad_norm": 1.1841380702726292, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 10259 + }, + { + "epoch": 0.1026, + "grad_norm": 0.9894474812507661, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10260 + }, + { + "epoch": 0.10261, + "grad_norm": 0.8997159163161771, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 10261 + }, + { + "epoch": 0.10262, + "grad_norm": 1.088375858779392, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10262 + }, + { + "epoch": 0.10263, + "grad_norm": 1.0827529991090061, + "learning_rate": 0.003, + "loss": 4.072, + "step": 10263 + }, + { + "epoch": 0.10264, + "grad_norm": 1.099911495151273, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 10264 + }, + { + "epoch": 0.10265, + "grad_norm": 0.9649937187292063, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 10265 + }, + { + "epoch": 0.10266, + "grad_norm": 0.9430820433940578, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 10266 + }, + { + "epoch": 0.10267, + "grad_norm": 1.0166591619014769, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10267 + }, + { + "epoch": 0.10268, + "grad_norm": 1.0981876655466734, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10268 + }, + { + "epoch": 0.10269, + "grad_norm": 1.1636560229250192, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10269 + }, + { + "epoch": 0.1027, + "grad_norm": 1.147237455230559, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 10270 + }, + { + "epoch": 0.10271, + "grad_norm": 1.2545858285112752, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 10271 + }, + { + "epoch": 0.10272, + "grad_norm": 1.2305947119786966, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 10272 + }, + { + "epoch": 0.10273, + "grad_norm": 1.2638269237402653, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10273 + }, + { + "epoch": 0.10274, + "grad_norm": 0.786046900482246, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 10274 + }, + { + "epoch": 0.10275, + "grad_norm": 0.7962616812711191, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 10275 + }, + { + "epoch": 0.10276, + "grad_norm": 0.9054468958301576, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 10276 + }, + { + "epoch": 0.10277, + "grad_norm": 1.1773560503821552, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10277 + }, + { + "epoch": 0.10278, + "grad_norm": 0.8437933387987852, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 10278 + }, + { + "epoch": 0.10279, + "grad_norm": 1.0099305707822057, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10279 + }, + { + "epoch": 0.1028, + "grad_norm": 1.172542815070957, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10280 + }, + { + "epoch": 0.10281, + "grad_norm": 0.9226538744503163, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 10281 + }, + { + "epoch": 0.10282, + "grad_norm": 1.0087487540894944, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 10282 + }, + { + "epoch": 0.10283, + "grad_norm": 1.0526662717264217, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10283 + }, + { + "epoch": 0.10284, + "grad_norm": 1.2763147693807468, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 10284 + }, + { + "epoch": 0.10285, + "grad_norm": 1.1523737559077365, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 10285 + }, + { + "epoch": 0.10286, + "grad_norm": 0.9428247281357661, + "learning_rate": 0.003, + "loss": 4.085, + "step": 10286 + }, + { + "epoch": 0.10287, + "grad_norm": 1.0532765499789432, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 10287 + }, + { + "epoch": 0.10288, + "grad_norm": 1.0871678672283156, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 10288 + }, + { + "epoch": 0.10289, + "grad_norm": 1.1107409124743397, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 10289 + }, + { + "epoch": 0.1029, + "grad_norm": 0.9157068380119726, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 10290 + }, + { + "epoch": 0.10291, + "grad_norm": 0.9147142055941883, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 10291 + }, + { + "epoch": 0.10292, + "grad_norm": 1.015781317673214, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 10292 + }, + { + "epoch": 0.10293, + "grad_norm": 1.0357861344949726, + "learning_rate": 0.003, + "loss": 4.091, + "step": 10293 + }, + { + "epoch": 0.10294, + "grad_norm": 1.0614880964315667, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 10294 + }, + { + "epoch": 0.10295, + "grad_norm": 1.1224618732216065, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 10295 + }, + { + "epoch": 0.10296, + "grad_norm": 1.1959008142012484, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 10296 + }, + { + "epoch": 0.10297, + "grad_norm": 1.3089339465817695, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 10297 + }, + { + "epoch": 0.10298, + "grad_norm": 1.2566589944135738, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 10298 + }, + { + "epoch": 0.10299, + "grad_norm": 0.8853939220827068, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 10299 + }, + { + "epoch": 0.103, + "grad_norm": 0.9357792578389142, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10300 + }, + { + "epoch": 0.10301, + "grad_norm": 1.1368678165849317, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 10301 + }, + { + "epoch": 0.10302, + "grad_norm": 1.0614029064175152, + "learning_rate": 0.003, + "loss": 4.115, + "step": 10302 + }, + { + "epoch": 0.10303, + "grad_norm": 0.9536131022368639, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 10303 + }, + { + "epoch": 0.10304, + "grad_norm": 1.0489700654643546, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 10304 + }, + { + "epoch": 0.10305, + "grad_norm": 1.173340722230068, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 10305 + }, + { + "epoch": 0.10306, + "grad_norm": 0.9482417193908688, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 10306 + }, + { + "epoch": 0.10307, + "grad_norm": 0.8614319455407266, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 10307 + }, + { + "epoch": 0.10308, + "grad_norm": 0.8944477264218518, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10308 + }, + { + "epoch": 0.10309, + "grad_norm": 1.0220567330268846, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 10309 + }, + { + "epoch": 0.1031, + "grad_norm": 1.223436846411137, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10310 + }, + { + "epoch": 0.10311, + "grad_norm": 0.9505791636624836, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 10311 + }, + { + "epoch": 0.10312, + "grad_norm": 0.8911177782963302, + "learning_rate": 0.003, + "loss": 4.081, + "step": 10312 + }, + { + "epoch": 0.10313, + "grad_norm": 0.9244202750779421, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 10313 + }, + { + "epoch": 0.10314, + "grad_norm": 0.9999763464957447, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 10314 + }, + { + "epoch": 0.10315, + "grad_norm": 1.3083868898257345, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 10315 + }, + { + "epoch": 0.10316, + "grad_norm": 0.9300920090393802, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 10316 + }, + { + "epoch": 0.10317, + "grad_norm": 1.1215040272892525, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10317 + }, + { + "epoch": 0.10318, + "grad_norm": 1.1317715673516657, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 10318 + }, + { + "epoch": 0.10319, + "grad_norm": 1.0742094095501016, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 10319 + }, + { + "epoch": 0.1032, + "grad_norm": 1.1313280016207532, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 10320 + }, + { + "epoch": 0.10321, + "grad_norm": 1.0785567501408917, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 10321 + }, + { + "epoch": 0.10322, + "grad_norm": 1.116372226706656, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 10322 + }, + { + "epoch": 0.10323, + "grad_norm": 1.1381641966602283, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 10323 + }, + { + "epoch": 0.10324, + "grad_norm": 1.1469641020232408, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 10324 + }, + { + "epoch": 0.10325, + "grad_norm": 1.020356233262012, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 10325 + }, + { + "epoch": 0.10326, + "grad_norm": 1.09961944340241, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 10326 + }, + { + "epoch": 0.10327, + "grad_norm": 1.35925570455499, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10327 + }, + { + "epoch": 0.10328, + "grad_norm": 0.9169067271168425, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 10328 + }, + { + "epoch": 0.10329, + "grad_norm": 1.0577568254856755, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 10329 + }, + { + "epoch": 0.1033, + "grad_norm": 1.134734217462752, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 10330 + }, + { + "epoch": 0.10331, + "grad_norm": 1.0651170802221943, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10331 + }, + { + "epoch": 0.10332, + "grad_norm": 1.006008548176185, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 10332 + }, + { + "epoch": 0.10333, + "grad_norm": 1.1589315309415948, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10333 + }, + { + "epoch": 0.10334, + "grad_norm": 1.1452212070860508, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 10334 + }, + { + "epoch": 0.10335, + "grad_norm": 0.8896619267885125, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10335 + }, + { + "epoch": 0.10336, + "grad_norm": 0.985250176255216, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 10336 + }, + { + "epoch": 0.10337, + "grad_norm": 1.2128001129054007, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10337 + }, + { + "epoch": 0.10338, + "grad_norm": 0.9030879185357017, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 10338 + }, + { + "epoch": 0.10339, + "grad_norm": 1.0505055949856439, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10339 + }, + { + "epoch": 0.1034, + "grad_norm": 1.1890035409892301, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 10340 + }, + { + "epoch": 0.10341, + "grad_norm": 0.8832065619363271, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 10341 + }, + { + "epoch": 0.10342, + "grad_norm": 0.9181271214030564, + "learning_rate": 0.003, + "loss": 4.055, + "step": 10342 + }, + { + "epoch": 0.10343, + "grad_norm": 1.0026233570871133, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10343 + }, + { + "epoch": 0.10344, + "grad_norm": 1.195425512592561, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 10344 + }, + { + "epoch": 0.10345, + "grad_norm": 0.9124977352765757, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10345 + }, + { + "epoch": 0.10346, + "grad_norm": 1.0417341030487266, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 10346 + }, + { + "epoch": 0.10347, + "grad_norm": 1.125139684288503, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 10347 + }, + { + "epoch": 0.10348, + "grad_norm": 0.9637619839862117, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 10348 + }, + { + "epoch": 0.10349, + "grad_norm": 1.0000001930540041, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 10349 + }, + { + "epoch": 0.1035, + "grad_norm": 1.3264629530651777, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 10350 + }, + { + "epoch": 0.10351, + "grad_norm": 0.9410407767703273, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 10351 + }, + { + "epoch": 0.10352, + "grad_norm": 1.2064264493745134, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10352 + }, + { + "epoch": 0.10353, + "grad_norm": 0.9581655627934581, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 10353 + }, + { + "epoch": 0.10354, + "grad_norm": 1.0813766464475838, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 10354 + }, + { + "epoch": 0.10355, + "grad_norm": 1.0105125418694436, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 10355 + }, + { + "epoch": 0.10356, + "grad_norm": 1.190959028236916, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 10356 + }, + { + "epoch": 0.10357, + "grad_norm": 1.255268509152371, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10357 + }, + { + "epoch": 0.10358, + "grad_norm": 1.1990872530553214, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 10358 + }, + { + "epoch": 0.10359, + "grad_norm": 0.9754039308040845, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 10359 + }, + { + "epoch": 0.1036, + "grad_norm": 1.0169279962302413, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 10360 + }, + { + "epoch": 0.10361, + "grad_norm": 1.095290235333622, + "learning_rate": 0.003, + "loss": 4.052, + "step": 10361 + }, + { + "epoch": 0.10362, + "grad_norm": 1.0889728298226304, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 10362 + }, + { + "epoch": 0.10363, + "grad_norm": 0.9704203825231105, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10363 + }, + { + "epoch": 0.10364, + "grad_norm": 1.1099160526895995, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 10364 + }, + { + "epoch": 0.10365, + "grad_norm": 1.1233714244351318, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 10365 + }, + { + "epoch": 0.10366, + "grad_norm": 0.9225455273609658, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10366 + }, + { + "epoch": 0.10367, + "grad_norm": 1.0325933199333683, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 10367 + }, + { + "epoch": 0.10368, + "grad_norm": 1.1791979754809137, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 10368 + }, + { + "epoch": 0.10369, + "grad_norm": 0.8022118836361686, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10369 + }, + { + "epoch": 0.1037, + "grad_norm": 0.9774590280143128, + "learning_rate": 0.003, + "loss": 4.077, + "step": 10370 + }, + { + "epoch": 0.10371, + "grad_norm": 1.3320865254161802, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 10371 + }, + { + "epoch": 0.10372, + "grad_norm": 1.1731206938597318, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 10372 + }, + { + "epoch": 0.10373, + "grad_norm": 0.8445497416221235, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10373 + }, + { + "epoch": 0.10374, + "grad_norm": 0.8558340467037168, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10374 + }, + { + "epoch": 0.10375, + "grad_norm": 1.0298854077289796, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10375 + }, + { + "epoch": 0.10376, + "grad_norm": 1.3300837557126834, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10376 + }, + { + "epoch": 0.10377, + "grad_norm": 0.9249173973773479, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10377 + }, + { + "epoch": 0.10378, + "grad_norm": 1.0239269983800974, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10378 + }, + { + "epoch": 0.10379, + "grad_norm": 1.121667224761993, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 10379 + }, + { + "epoch": 0.1038, + "grad_norm": 1.0149496858844194, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 10380 + }, + { + "epoch": 0.10381, + "grad_norm": 1.0559874429370242, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10381 + }, + { + "epoch": 0.10382, + "grad_norm": 0.9631985714419412, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10382 + }, + { + "epoch": 0.10383, + "grad_norm": 0.9514945697162748, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 10383 + }, + { + "epoch": 0.10384, + "grad_norm": 1.1301104439298963, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 10384 + }, + { + "epoch": 0.10385, + "grad_norm": 1.0145183022082147, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10385 + }, + { + "epoch": 0.10386, + "grad_norm": 1.2581553816472262, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 10386 + }, + { + "epoch": 0.10387, + "grad_norm": 0.9484969624909415, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10387 + }, + { + "epoch": 0.10388, + "grad_norm": 0.9469902349525261, + "learning_rate": 0.003, + "loss": 4.081, + "step": 10388 + }, + { + "epoch": 0.10389, + "grad_norm": 0.9773642974595819, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 10389 + }, + { + "epoch": 0.1039, + "grad_norm": 1.0578749602427104, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 10390 + }, + { + "epoch": 0.10391, + "grad_norm": 1.0152407829650854, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10391 + }, + { + "epoch": 0.10392, + "grad_norm": 1.1687985660961098, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10392 + }, + { + "epoch": 0.10393, + "grad_norm": 1.274162871161583, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 10393 + }, + { + "epoch": 0.10394, + "grad_norm": 1.0376815910121797, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10394 + }, + { + "epoch": 0.10395, + "grad_norm": 1.1455649817000828, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 10395 + }, + { + "epoch": 0.10396, + "grad_norm": 1.048450933526513, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10396 + }, + { + "epoch": 0.10397, + "grad_norm": 1.0835103485182267, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 10397 + }, + { + "epoch": 0.10398, + "grad_norm": 1.1028674181870997, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10398 + }, + { + "epoch": 0.10399, + "grad_norm": 1.1462902932061454, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 10399 + }, + { + "epoch": 0.104, + "grad_norm": 1.070594058981979, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10400 + }, + { + "epoch": 0.10401, + "grad_norm": 1.041054683828324, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 10401 + }, + { + "epoch": 0.10402, + "grad_norm": 1.1525497854276674, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 10402 + }, + { + "epoch": 0.10403, + "grad_norm": 1.0190917702762687, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 10403 + }, + { + "epoch": 0.10404, + "grad_norm": 1.1285203858625503, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 10404 + }, + { + "epoch": 0.10405, + "grad_norm": 1.0516484986800476, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 10405 + }, + { + "epoch": 0.10406, + "grad_norm": 1.1198310400726428, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 10406 + }, + { + "epoch": 0.10407, + "grad_norm": 0.9507450061808831, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 10407 + }, + { + "epoch": 0.10408, + "grad_norm": 0.9195044187230283, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 10408 + }, + { + "epoch": 0.10409, + "grad_norm": 0.9570767287416899, + "learning_rate": 0.003, + "loss": 4.085, + "step": 10409 + }, + { + "epoch": 0.1041, + "grad_norm": 1.0343173813267, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 10410 + }, + { + "epoch": 0.10411, + "grad_norm": 1.1150088548083112, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10411 + }, + { + "epoch": 0.10412, + "grad_norm": 1.2434197115786008, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10412 + }, + { + "epoch": 0.10413, + "grad_norm": 0.9686766565831221, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 10413 + }, + { + "epoch": 0.10414, + "grad_norm": 1.0870609051427256, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 10414 + }, + { + "epoch": 0.10415, + "grad_norm": 1.1242148496718316, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10415 + }, + { + "epoch": 0.10416, + "grad_norm": 1.0292327009210644, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 10416 + }, + { + "epoch": 0.10417, + "grad_norm": 1.0822743668308503, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 10417 + }, + { + "epoch": 0.10418, + "grad_norm": 1.0278876159616779, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10418 + }, + { + "epoch": 0.10419, + "grad_norm": 1.0491246301200836, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 10419 + }, + { + "epoch": 0.1042, + "grad_norm": 0.9477998200019015, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 10420 + }, + { + "epoch": 0.10421, + "grad_norm": 1.0788587406474686, + "learning_rate": 0.003, + "loss": 4.08, + "step": 10421 + }, + { + "epoch": 0.10422, + "grad_norm": 0.9632100820284665, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10422 + }, + { + "epoch": 0.10423, + "grad_norm": 1.0019319499065928, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 10423 + }, + { + "epoch": 0.10424, + "grad_norm": 1.143735387277337, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 10424 + }, + { + "epoch": 0.10425, + "grad_norm": 0.7908128783667394, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 10425 + }, + { + "epoch": 0.10426, + "grad_norm": 0.9313781838068929, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 10426 + }, + { + "epoch": 0.10427, + "grad_norm": 1.1984211817477621, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10427 + }, + { + "epoch": 0.10428, + "grad_norm": 0.9431446163495729, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 10428 + }, + { + "epoch": 0.10429, + "grad_norm": 1.168965044587708, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 10429 + }, + { + "epoch": 0.1043, + "grad_norm": 1.3908380610858844, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10430 + }, + { + "epoch": 0.10431, + "grad_norm": 1.0676013928236892, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10431 + }, + { + "epoch": 0.10432, + "grad_norm": 1.066980716023993, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 10432 + }, + { + "epoch": 0.10433, + "grad_norm": 0.9948132055941239, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 10433 + }, + { + "epoch": 0.10434, + "grad_norm": 1.0508222592142207, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10434 + }, + { + "epoch": 0.10435, + "grad_norm": 1.0748934991259524, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 10435 + }, + { + "epoch": 0.10436, + "grad_norm": 0.9961213953784716, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 10436 + }, + { + "epoch": 0.10437, + "grad_norm": 1.2265320822152506, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 10437 + }, + { + "epoch": 0.10438, + "grad_norm": 0.9727388902004728, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 10438 + }, + { + "epoch": 0.10439, + "grad_norm": 0.9461742697962658, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 10439 + }, + { + "epoch": 0.1044, + "grad_norm": 0.9885970310180309, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 10440 + }, + { + "epoch": 0.10441, + "grad_norm": 1.0654469009945895, + "learning_rate": 0.003, + "loss": 4.075, + "step": 10441 + }, + { + "epoch": 0.10442, + "grad_norm": 0.9305298263072211, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 10442 + }, + { + "epoch": 0.10443, + "grad_norm": 1.0404155491200193, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 10443 + }, + { + "epoch": 0.10444, + "grad_norm": 1.1653476936243858, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10444 + }, + { + "epoch": 0.10445, + "grad_norm": 1.2134056802647015, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 10445 + }, + { + "epoch": 0.10446, + "grad_norm": 0.8997490311080218, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 10446 + }, + { + "epoch": 0.10447, + "grad_norm": 0.8352844337856086, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 10447 + }, + { + "epoch": 0.10448, + "grad_norm": 0.8792325987966317, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10448 + }, + { + "epoch": 0.10449, + "grad_norm": 1.1230065414769836, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10449 + }, + { + "epoch": 0.1045, + "grad_norm": 1.0569256969404204, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 10450 + }, + { + "epoch": 0.10451, + "grad_norm": 0.9741322946515599, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 10451 + }, + { + "epoch": 0.10452, + "grad_norm": 1.12094286850372, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 10452 + }, + { + "epoch": 0.10453, + "grad_norm": 1.0634000951508014, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 10453 + }, + { + "epoch": 0.10454, + "grad_norm": 1.0398692030033183, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 10454 + }, + { + "epoch": 0.10455, + "grad_norm": 1.0697825048342626, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 10455 + }, + { + "epoch": 0.10456, + "grad_norm": 0.9369648357510157, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10456 + }, + { + "epoch": 0.10457, + "grad_norm": 1.1077828883876948, + "learning_rate": 0.003, + "loss": 4.096, + "step": 10457 + }, + { + "epoch": 0.10458, + "grad_norm": 1.2486942415113471, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 10458 + }, + { + "epoch": 0.10459, + "grad_norm": 1.0800758371094688, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10459 + }, + { + "epoch": 0.1046, + "grad_norm": 1.1580347302708973, + "learning_rate": 0.003, + "loss": 4.105, + "step": 10460 + }, + { + "epoch": 0.10461, + "grad_norm": 1.1725452690555769, + "learning_rate": 0.003, + "loss": 4.068, + "step": 10461 + }, + { + "epoch": 0.10462, + "grad_norm": 1.0908592370020997, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10462 + }, + { + "epoch": 0.10463, + "grad_norm": 0.991908326018903, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 10463 + }, + { + "epoch": 0.10464, + "grad_norm": 1.2301306291338194, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 10464 + }, + { + "epoch": 0.10465, + "grad_norm": 0.9472521418396895, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10465 + }, + { + "epoch": 0.10466, + "grad_norm": 1.0771140483899684, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10466 + }, + { + "epoch": 0.10467, + "grad_norm": 1.0045428457733532, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 10467 + }, + { + "epoch": 0.10468, + "grad_norm": 1.063329814375369, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10468 + }, + { + "epoch": 0.10469, + "grad_norm": 0.8044646549832464, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10469 + }, + { + "epoch": 0.1047, + "grad_norm": 0.8705710123890662, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10470 + }, + { + "epoch": 0.10471, + "grad_norm": 0.9026927449559127, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 10471 + }, + { + "epoch": 0.10472, + "grad_norm": 0.9956602765823049, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10472 + }, + { + "epoch": 0.10473, + "grad_norm": 1.1589637831106856, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10473 + }, + { + "epoch": 0.10474, + "grad_norm": 1.184785189846938, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 10474 + }, + { + "epoch": 0.10475, + "grad_norm": 1.3550631194540537, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 10475 + }, + { + "epoch": 0.10476, + "grad_norm": 1.0320795978049657, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 10476 + }, + { + "epoch": 0.10477, + "grad_norm": 1.0604195369959852, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10477 + }, + { + "epoch": 0.10478, + "grad_norm": 1.0637193655219612, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10478 + }, + { + "epoch": 0.10479, + "grad_norm": 0.9630771797139417, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 10479 + }, + { + "epoch": 0.1048, + "grad_norm": 0.9484006284165586, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 10480 + }, + { + "epoch": 0.10481, + "grad_norm": 1.054567576660935, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10481 + }, + { + "epoch": 0.10482, + "grad_norm": 1.3360352292856175, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10482 + }, + { + "epoch": 0.10483, + "grad_norm": 0.9019567481761226, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 10483 + }, + { + "epoch": 0.10484, + "grad_norm": 0.991285568753618, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 10484 + }, + { + "epoch": 0.10485, + "grad_norm": 1.1370528625194423, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10485 + }, + { + "epoch": 0.10486, + "grad_norm": 0.9127707353956426, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 10486 + }, + { + "epoch": 0.10487, + "grad_norm": 0.9285607355356765, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10487 + }, + { + "epoch": 0.10488, + "grad_norm": 1.0231832309662972, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 10488 + }, + { + "epoch": 0.10489, + "grad_norm": 1.1181475303743535, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 10489 + }, + { + "epoch": 0.1049, + "grad_norm": 1.058957141573963, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 10490 + }, + { + "epoch": 0.10491, + "grad_norm": 1.4398699183245742, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 10491 + }, + { + "epoch": 0.10492, + "grad_norm": 0.7504173202433805, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 10492 + }, + { + "epoch": 0.10493, + "grad_norm": 0.8117939810615702, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 10493 + }, + { + "epoch": 0.10494, + "grad_norm": 1.0189444416069287, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 10494 + }, + { + "epoch": 0.10495, + "grad_norm": 1.0926650227724901, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 10495 + }, + { + "epoch": 0.10496, + "grad_norm": 0.9245404161543932, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 10496 + }, + { + "epoch": 0.10497, + "grad_norm": 0.8690029751405167, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 10497 + }, + { + "epoch": 0.10498, + "grad_norm": 0.9260719581793728, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10498 + }, + { + "epoch": 0.10499, + "grad_norm": 0.9504218723749869, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10499 + }, + { + "epoch": 0.105, + "grad_norm": 1.049731320459712, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 10500 + }, + { + "epoch": 0.10501, + "grad_norm": 1.2194994953775773, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 10501 + }, + { + "epoch": 0.10502, + "grad_norm": 1.3901211020384214, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 10502 + }, + { + "epoch": 0.10503, + "grad_norm": 0.8847049875069319, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 10503 + }, + { + "epoch": 0.10504, + "grad_norm": 0.9053464516632651, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 10504 + }, + { + "epoch": 0.10505, + "grad_norm": 1.0078443390204859, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 10505 + }, + { + "epoch": 0.10506, + "grad_norm": 1.0299360239934658, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 10506 + }, + { + "epoch": 0.10507, + "grad_norm": 1.0873339450933095, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10507 + }, + { + "epoch": 0.10508, + "grad_norm": 0.9753848523226265, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 10508 + }, + { + "epoch": 0.10509, + "grad_norm": 1.1237575917023788, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 10509 + }, + { + "epoch": 0.1051, + "grad_norm": 1.24958584132952, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 10510 + }, + { + "epoch": 0.10511, + "grad_norm": 1.2571755536635105, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10511 + }, + { + "epoch": 0.10512, + "grad_norm": 0.935825340827118, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 10512 + }, + { + "epoch": 0.10513, + "grad_norm": 0.954581530206918, + "learning_rate": 0.003, + "loss": 4.049, + "step": 10513 + }, + { + "epoch": 0.10514, + "grad_norm": 1.0318931238482127, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 10514 + }, + { + "epoch": 0.10515, + "grad_norm": 1.1905130618079427, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10515 + }, + { + "epoch": 0.10516, + "grad_norm": 0.9218196739218961, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 10516 + }, + { + "epoch": 0.10517, + "grad_norm": 1.0812257850070317, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 10517 + }, + { + "epoch": 0.10518, + "grad_norm": 1.0289312096687557, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 10518 + }, + { + "epoch": 0.10519, + "grad_norm": 1.0840110264499334, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 10519 + }, + { + "epoch": 0.1052, + "grad_norm": 1.0163700110837486, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 10520 + }, + { + "epoch": 0.10521, + "grad_norm": 1.168120560099558, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10521 + }, + { + "epoch": 0.10522, + "grad_norm": 0.941883900004278, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10522 + }, + { + "epoch": 0.10523, + "grad_norm": 0.9365433648184146, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 10523 + }, + { + "epoch": 0.10524, + "grad_norm": 1.0607376077744424, + "learning_rate": 0.003, + "loss": 4.079, + "step": 10524 + }, + { + "epoch": 0.10525, + "grad_norm": 1.1053780367918074, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 10525 + }, + { + "epoch": 0.10526, + "grad_norm": 1.1732250128516668, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 10526 + }, + { + "epoch": 0.10527, + "grad_norm": 1.0742015400018985, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 10527 + }, + { + "epoch": 0.10528, + "grad_norm": 1.366765506902064, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 10528 + }, + { + "epoch": 0.10529, + "grad_norm": 0.8202184376308155, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 10529 + }, + { + "epoch": 0.1053, + "grad_norm": 0.8208632856067515, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10530 + }, + { + "epoch": 0.10531, + "grad_norm": 0.8352417106297955, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 10531 + }, + { + "epoch": 0.10532, + "grad_norm": 0.9160049441948251, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 10532 + }, + { + "epoch": 0.10533, + "grad_norm": 1.0631425057807433, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 10533 + }, + { + "epoch": 0.10534, + "grad_norm": 1.1827574920147312, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 10534 + }, + { + "epoch": 0.10535, + "grad_norm": 0.8811514403995269, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 10535 + }, + { + "epoch": 0.10536, + "grad_norm": 1.3452482194853557, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 10536 + }, + { + "epoch": 0.10537, + "grad_norm": 1.2025873890846424, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10537 + }, + { + "epoch": 0.10538, + "grad_norm": 0.9467288343288578, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 10538 + }, + { + "epoch": 0.10539, + "grad_norm": 1.1008525955886186, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 10539 + }, + { + "epoch": 0.1054, + "grad_norm": 1.17784567298581, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 10540 + }, + { + "epoch": 0.10541, + "grad_norm": 0.9005444820875294, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10541 + }, + { + "epoch": 0.10542, + "grad_norm": 1.077049482321095, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 10542 + }, + { + "epoch": 0.10543, + "grad_norm": 1.0664201972433869, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 10543 + }, + { + "epoch": 0.10544, + "grad_norm": 1.0681106820442525, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 10544 + }, + { + "epoch": 0.10545, + "grad_norm": 1.2213402778855749, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 10545 + }, + { + "epoch": 0.10546, + "grad_norm": 0.9978122256357836, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 10546 + }, + { + "epoch": 0.10547, + "grad_norm": 1.2943816452235526, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 10547 + }, + { + "epoch": 0.10548, + "grad_norm": 0.8865941146378675, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 10548 + }, + { + "epoch": 0.10549, + "grad_norm": 1.0650208342194967, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10549 + }, + { + "epoch": 0.1055, + "grad_norm": 1.133121694024822, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10550 + }, + { + "epoch": 0.10551, + "grad_norm": 1.0356315787198624, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 10551 + }, + { + "epoch": 0.10552, + "grad_norm": 1.0470920531533148, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 10552 + }, + { + "epoch": 0.10553, + "grad_norm": 0.932425325526456, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 10553 + }, + { + "epoch": 0.10554, + "grad_norm": 1.0524772486523728, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10554 + }, + { + "epoch": 0.10555, + "grad_norm": 1.09832900977572, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 10555 + }, + { + "epoch": 0.10556, + "grad_norm": 0.9863036466375944, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10556 + }, + { + "epoch": 0.10557, + "grad_norm": 0.992725409801414, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 10557 + }, + { + "epoch": 0.10558, + "grad_norm": 1.0608804710461675, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 10558 + }, + { + "epoch": 0.10559, + "grad_norm": 1.0901193752068266, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 10559 + }, + { + "epoch": 0.1056, + "grad_norm": 1.2892535190614614, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 10560 + }, + { + "epoch": 0.10561, + "grad_norm": 0.9940038126547566, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 10561 + }, + { + "epoch": 0.10562, + "grad_norm": 0.9826732604285782, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 10562 + }, + { + "epoch": 0.10563, + "grad_norm": 1.1983939579421365, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 10563 + }, + { + "epoch": 0.10564, + "grad_norm": 1.0231507937173892, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 10564 + }, + { + "epoch": 0.10565, + "grad_norm": 1.125203308541098, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10565 + }, + { + "epoch": 0.10566, + "grad_norm": 0.8906519373180206, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 10566 + }, + { + "epoch": 0.10567, + "grad_norm": 0.962568584030796, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 10567 + }, + { + "epoch": 0.10568, + "grad_norm": 1.2402788749569134, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 10568 + }, + { + "epoch": 0.10569, + "grad_norm": 1.1134136146515001, + "learning_rate": 0.003, + "loss": 4.109, + "step": 10569 + }, + { + "epoch": 0.1057, + "grad_norm": 1.3133871676098183, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10570 + }, + { + "epoch": 0.10571, + "grad_norm": 0.951198930506272, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10571 + }, + { + "epoch": 0.10572, + "grad_norm": 0.9761781766255831, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10572 + }, + { + "epoch": 0.10573, + "grad_norm": 1.0133150920197171, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 10573 + }, + { + "epoch": 0.10574, + "grad_norm": 1.013912210371583, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10574 + }, + { + "epoch": 0.10575, + "grad_norm": 1.05313647300968, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 10575 + }, + { + "epoch": 0.10576, + "grad_norm": 0.9509674082920994, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 10576 + }, + { + "epoch": 0.10577, + "grad_norm": 1.0053376130317002, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 10577 + }, + { + "epoch": 0.10578, + "grad_norm": 1.0795367145271022, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 10578 + }, + { + "epoch": 0.10579, + "grad_norm": 1.1298331639731727, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10579 + }, + { + "epoch": 0.1058, + "grad_norm": 1.0884931502898405, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10580 + }, + { + "epoch": 0.10581, + "grad_norm": 1.0468541072064026, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10581 + }, + { + "epoch": 0.10582, + "grad_norm": 1.1974492928780258, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10582 + }, + { + "epoch": 0.10583, + "grad_norm": 0.882313206303918, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 10583 + }, + { + "epoch": 0.10584, + "grad_norm": 0.9033505969353639, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 10584 + }, + { + "epoch": 0.10585, + "grad_norm": 1.0056878191730187, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 10585 + }, + { + "epoch": 0.10586, + "grad_norm": 1.237777886872006, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 10586 + }, + { + "epoch": 0.10587, + "grad_norm": 0.9248181845601777, + "learning_rate": 0.003, + "loss": 4.107, + "step": 10587 + }, + { + "epoch": 0.10588, + "grad_norm": 1.0882857074285057, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 10588 + }, + { + "epoch": 0.10589, + "grad_norm": 1.1046332241765098, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10589 + }, + { + "epoch": 0.1059, + "grad_norm": 1.0054015979587632, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10590 + }, + { + "epoch": 0.10591, + "grad_norm": 1.1622872539084141, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 10591 + }, + { + "epoch": 0.10592, + "grad_norm": 0.8654066003416311, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 10592 + }, + { + "epoch": 0.10593, + "grad_norm": 0.9936129959819774, + "learning_rate": 0.003, + "loss": 4.058, + "step": 10593 + }, + { + "epoch": 0.10594, + "grad_norm": 1.1212371509536319, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10594 + }, + { + "epoch": 0.10595, + "grad_norm": 1.1599243180097774, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10595 + }, + { + "epoch": 0.10596, + "grad_norm": 1.0242071396552408, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 10596 + }, + { + "epoch": 0.10597, + "grad_norm": 0.9678502785231397, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10597 + }, + { + "epoch": 0.10598, + "grad_norm": 1.030853541727889, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10598 + }, + { + "epoch": 0.10599, + "grad_norm": 1.4289073689256773, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 10599 + }, + { + "epoch": 0.106, + "grad_norm": 1.0707379947995017, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10600 + }, + { + "epoch": 0.10601, + "grad_norm": 1.0119993011968624, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10601 + }, + { + "epoch": 0.10602, + "grad_norm": 1.1556786577647435, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 10602 + }, + { + "epoch": 0.10603, + "grad_norm": 0.9062233720174303, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 10603 + }, + { + "epoch": 0.10604, + "grad_norm": 0.7573877794731967, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 10604 + }, + { + "epoch": 0.10605, + "grad_norm": 0.8217624074209846, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10605 + }, + { + "epoch": 0.10606, + "grad_norm": 0.9714939673985139, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10606 + }, + { + "epoch": 0.10607, + "grad_norm": 1.0141974591485123, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 10607 + }, + { + "epoch": 0.10608, + "grad_norm": 1.0963103808271801, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10608 + }, + { + "epoch": 0.10609, + "grad_norm": 1.1431306058396495, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 10609 + }, + { + "epoch": 0.1061, + "grad_norm": 1.4008154829290844, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 10610 + }, + { + "epoch": 0.10611, + "grad_norm": 0.9470325380773612, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 10611 + }, + { + "epoch": 0.10612, + "grad_norm": 1.0311831527113837, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 10612 + }, + { + "epoch": 0.10613, + "grad_norm": 1.164743448303555, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 10613 + }, + { + "epoch": 0.10614, + "grad_norm": 0.9502120855552426, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 10614 + }, + { + "epoch": 0.10615, + "grad_norm": 1.1998095664589368, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 10615 + }, + { + "epoch": 0.10616, + "grad_norm": 1.0083901716861463, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 10616 + }, + { + "epoch": 0.10617, + "grad_norm": 1.0907852145771553, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 10617 + }, + { + "epoch": 0.10618, + "grad_norm": 1.0124541580039028, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 10618 + }, + { + "epoch": 0.10619, + "grad_norm": 1.113581821108827, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 10619 + }, + { + "epoch": 0.1062, + "grad_norm": 1.1026787669661373, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10620 + }, + { + "epoch": 0.10621, + "grad_norm": 0.9751223940659827, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 10621 + }, + { + "epoch": 0.10622, + "grad_norm": 1.0749021101818106, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 10622 + }, + { + "epoch": 0.10623, + "grad_norm": 0.945349520748991, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 10623 + }, + { + "epoch": 0.10624, + "grad_norm": 1.0556364929047508, + "learning_rate": 0.003, + "loss": 4.072, + "step": 10624 + }, + { + "epoch": 0.10625, + "grad_norm": 1.2238541346788792, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 10625 + }, + { + "epoch": 0.10626, + "grad_norm": 0.9013129245547551, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10626 + }, + { + "epoch": 0.10627, + "grad_norm": 0.864027919043482, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10627 + }, + { + "epoch": 0.10628, + "grad_norm": 0.9094857462874787, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 10628 + }, + { + "epoch": 0.10629, + "grad_norm": 1.0335557146674315, + "learning_rate": 0.003, + "loss": 4.078, + "step": 10629 + }, + { + "epoch": 0.1063, + "grad_norm": 1.292500999524564, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10630 + }, + { + "epoch": 0.10631, + "grad_norm": 0.9835133011239242, + "learning_rate": 0.003, + "loss": 4.101, + "step": 10631 + }, + { + "epoch": 0.10632, + "grad_norm": 1.0666622832328627, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 10632 + }, + { + "epoch": 0.10633, + "grad_norm": 1.0826186803880389, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 10633 + }, + { + "epoch": 0.10634, + "grad_norm": 1.0641990091605356, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10634 + }, + { + "epoch": 0.10635, + "grad_norm": 1.1612087316359914, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 10635 + }, + { + "epoch": 0.10636, + "grad_norm": 0.8748537891089967, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 10636 + }, + { + "epoch": 0.10637, + "grad_norm": 1.0481532412413115, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 10637 + }, + { + "epoch": 0.10638, + "grad_norm": 1.2606203743062692, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10638 + }, + { + "epoch": 0.10639, + "grad_norm": 1.1805082052116393, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 10639 + }, + { + "epoch": 0.1064, + "grad_norm": 1.2398870931992865, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 10640 + }, + { + "epoch": 0.10641, + "grad_norm": 0.9218771485309251, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 10641 + }, + { + "epoch": 0.10642, + "grad_norm": 1.1272822505994036, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10642 + }, + { + "epoch": 0.10643, + "grad_norm": 1.175175782769236, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 10643 + }, + { + "epoch": 0.10644, + "grad_norm": 0.8929122030535886, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 10644 + }, + { + "epoch": 0.10645, + "grad_norm": 1.0037622742587453, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 10645 + }, + { + "epoch": 0.10646, + "grad_norm": 1.1610195694424994, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 10646 + }, + { + "epoch": 0.10647, + "grad_norm": 0.852454983713457, + "learning_rate": 0.003, + "loss": 4.077, + "step": 10647 + }, + { + "epoch": 0.10648, + "grad_norm": 0.7401630263924073, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 10648 + }, + { + "epoch": 0.10649, + "grad_norm": 0.8311609147359353, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 10649 + }, + { + "epoch": 0.1065, + "grad_norm": 0.9504931584008245, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 10650 + }, + { + "epoch": 0.10651, + "grad_norm": 1.122066042533435, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 10651 + }, + { + "epoch": 0.10652, + "grad_norm": 1.2329797804688072, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 10652 + }, + { + "epoch": 0.10653, + "grad_norm": 0.8425687985433964, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 10653 + }, + { + "epoch": 0.10654, + "grad_norm": 0.9139911827111628, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 10654 + }, + { + "epoch": 0.10655, + "grad_norm": 1.011999739520214, + "learning_rate": 0.003, + "loss": 4.088, + "step": 10655 + }, + { + "epoch": 0.10656, + "grad_norm": 1.2012862999427212, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10656 + }, + { + "epoch": 0.10657, + "grad_norm": 1.0352708658419536, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 10657 + }, + { + "epoch": 0.10658, + "grad_norm": 1.2516087672765586, + "learning_rate": 0.003, + "loss": 4.07, + "step": 10658 + }, + { + "epoch": 0.10659, + "grad_norm": 0.8772559044208297, + "learning_rate": 0.003, + "loss": 4.086, + "step": 10659 + }, + { + "epoch": 0.1066, + "grad_norm": 0.9125719571822856, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 10660 + }, + { + "epoch": 0.10661, + "grad_norm": 0.8865786014446685, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 10661 + }, + { + "epoch": 0.10662, + "grad_norm": 1.0850987643365564, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10662 + }, + { + "epoch": 0.10663, + "grad_norm": 1.2748185200581434, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 10663 + }, + { + "epoch": 0.10664, + "grad_norm": 1.1162026613036358, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 10664 + }, + { + "epoch": 0.10665, + "grad_norm": 0.9494012979211818, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10665 + }, + { + "epoch": 0.10666, + "grad_norm": 1.02397050134285, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 10666 + }, + { + "epoch": 0.10667, + "grad_norm": 1.140121878572695, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10667 + }, + { + "epoch": 0.10668, + "grad_norm": 1.2059412023031264, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 10668 + }, + { + "epoch": 0.10669, + "grad_norm": 1.1076246091621595, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 10669 + }, + { + "epoch": 0.1067, + "grad_norm": 1.0177092098793927, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 10670 + }, + { + "epoch": 0.10671, + "grad_norm": 1.1723966711259484, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 10671 + }, + { + "epoch": 0.10672, + "grad_norm": 1.0135262749335747, + "learning_rate": 0.003, + "loss": 4.084, + "step": 10672 + }, + { + "epoch": 0.10673, + "grad_norm": 1.2974838543788962, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 10673 + }, + { + "epoch": 0.10674, + "grad_norm": 1.025359321526118, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 10674 + }, + { + "epoch": 0.10675, + "grad_norm": 1.1358099502014838, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 10675 + }, + { + "epoch": 0.10676, + "grad_norm": 0.9642777344120331, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 10676 + }, + { + "epoch": 0.10677, + "grad_norm": 1.1745840304279103, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10677 + }, + { + "epoch": 0.10678, + "grad_norm": 0.9325862438493242, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 10678 + }, + { + "epoch": 0.10679, + "grad_norm": 0.9072654723239565, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 10679 + }, + { + "epoch": 0.1068, + "grad_norm": 1.2030425278652832, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 10680 + }, + { + "epoch": 0.10681, + "grad_norm": 1.1857166606883949, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 10681 + }, + { + "epoch": 0.10682, + "grad_norm": 0.9635367808951002, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 10682 + }, + { + "epoch": 0.10683, + "grad_norm": 1.0311474575422612, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10683 + }, + { + "epoch": 0.10684, + "grad_norm": 1.0449055839802592, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 10684 + }, + { + "epoch": 0.10685, + "grad_norm": 0.9408793445391519, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 10685 + }, + { + "epoch": 0.10686, + "grad_norm": 1.0709142575491677, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10686 + }, + { + "epoch": 0.10687, + "grad_norm": 1.118246780411596, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 10687 + }, + { + "epoch": 0.10688, + "grad_norm": 1.0804038641938123, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 10688 + }, + { + "epoch": 0.10689, + "grad_norm": 1.0753929491965348, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 10689 + }, + { + "epoch": 0.1069, + "grad_norm": 1.0631896177874685, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10690 + }, + { + "epoch": 0.10691, + "grad_norm": 1.0556411062232636, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 10691 + }, + { + "epoch": 0.10692, + "grad_norm": 1.0547716927758737, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 10692 + }, + { + "epoch": 0.10693, + "grad_norm": 1.1858072902357788, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 10693 + }, + { + "epoch": 0.10694, + "grad_norm": 0.9053792623061231, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10694 + }, + { + "epoch": 0.10695, + "grad_norm": 1.1363465670513433, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 10695 + }, + { + "epoch": 0.10696, + "grad_norm": 1.1855669176217434, + "learning_rate": 0.003, + "loss": 4.094, + "step": 10696 + }, + { + "epoch": 0.10697, + "grad_norm": 1.1003628381077988, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10697 + }, + { + "epoch": 0.10698, + "grad_norm": 1.1351058188402894, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 10698 + }, + { + "epoch": 0.10699, + "grad_norm": 0.9768170554346906, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 10699 + }, + { + "epoch": 0.107, + "grad_norm": 1.0710799140402476, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 10700 + }, + { + "epoch": 0.10701, + "grad_norm": 1.0344209083339428, + "learning_rate": 0.003, + "loss": 4.072, + "step": 10701 + }, + { + "epoch": 0.10702, + "grad_norm": 1.061616138491857, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 10702 + }, + { + "epoch": 0.10703, + "grad_norm": 0.9847321842818494, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 10703 + }, + { + "epoch": 0.10704, + "grad_norm": 1.0378569487826044, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10704 + }, + { + "epoch": 0.10705, + "grad_norm": 1.227085863193695, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 10705 + }, + { + "epoch": 0.10706, + "grad_norm": 0.9995137410984207, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 10706 + }, + { + "epoch": 0.10707, + "grad_norm": 1.1986759073174438, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10707 + }, + { + "epoch": 0.10708, + "grad_norm": 1.1415366654865533, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 10708 + }, + { + "epoch": 0.10709, + "grad_norm": 1.5136241575025582, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 10709 + }, + { + "epoch": 0.1071, + "grad_norm": 0.8341804004870889, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10710 + }, + { + "epoch": 0.10711, + "grad_norm": 0.926630320291128, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 10711 + }, + { + "epoch": 0.10712, + "grad_norm": 0.9663161163591082, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 10712 + }, + { + "epoch": 0.10713, + "grad_norm": 1.2205745710480322, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 10713 + }, + { + "epoch": 0.10714, + "grad_norm": 1.0390508678036858, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 10714 + }, + { + "epoch": 0.10715, + "grad_norm": 1.0528320057494256, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 10715 + }, + { + "epoch": 0.10716, + "grad_norm": 1.2072562792099912, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 10716 + }, + { + "epoch": 0.10717, + "grad_norm": 0.8790537191222857, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10717 + }, + { + "epoch": 0.10718, + "grad_norm": 0.8808640271468053, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10718 + }, + { + "epoch": 0.10719, + "grad_norm": 1.0126623071945835, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 10719 + }, + { + "epoch": 0.1072, + "grad_norm": 1.319727283236711, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 10720 + }, + { + "epoch": 0.10721, + "grad_norm": 1.2466570549525533, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 10721 + }, + { + "epoch": 0.10722, + "grad_norm": 0.97840046732779, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 10722 + }, + { + "epoch": 0.10723, + "grad_norm": 1.0704044212362178, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10723 + }, + { + "epoch": 0.10724, + "grad_norm": 1.0468299829861158, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 10724 + }, + { + "epoch": 0.10725, + "grad_norm": 1.1732056432117681, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 10725 + }, + { + "epoch": 0.10726, + "grad_norm": 0.9618675203781902, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 10726 + }, + { + "epoch": 0.10727, + "grad_norm": 1.056036020432587, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10727 + }, + { + "epoch": 0.10728, + "grad_norm": 0.9599269707382877, + "learning_rate": 0.003, + "loss": 4.102, + "step": 10728 + }, + { + "epoch": 0.10729, + "grad_norm": 1.0564091003019227, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 10729 + }, + { + "epoch": 0.1073, + "grad_norm": 0.9973209788927452, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 10730 + }, + { + "epoch": 0.10731, + "grad_norm": 1.0212852406413646, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 10731 + }, + { + "epoch": 0.10732, + "grad_norm": 1.1209440659622514, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 10732 + }, + { + "epoch": 0.10733, + "grad_norm": 1.0618758913171396, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 10733 + }, + { + "epoch": 0.10734, + "grad_norm": 1.1488201145415744, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 10734 + }, + { + "epoch": 0.10735, + "grad_norm": 0.9043046551792133, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10735 + }, + { + "epoch": 0.10736, + "grad_norm": 0.892103611120565, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10736 + }, + { + "epoch": 0.10737, + "grad_norm": 0.970984267995846, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 10737 + }, + { + "epoch": 0.10738, + "grad_norm": 1.1262141754422643, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 10738 + }, + { + "epoch": 0.10739, + "grad_norm": 0.9769718344245536, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 10739 + }, + { + "epoch": 0.1074, + "grad_norm": 1.4962807603874, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 10740 + }, + { + "epoch": 0.10741, + "grad_norm": 1.155833820559741, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10741 + }, + { + "epoch": 0.10742, + "grad_norm": 1.174159756749479, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 10742 + }, + { + "epoch": 0.10743, + "grad_norm": 1.012862742000115, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10743 + }, + { + "epoch": 0.10744, + "grad_norm": 1.1370193003708244, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 10744 + }, + { + "epoch": 0.10745, + "grad_norm": 0.9464204645242252, + "learning_rate": 0.003, + "loss": 4.074, + "step": 10745 + }, + { + "epoch": 0.10746, + "grad_norm": 1.3033475431034287, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 10746 + }, + { + "epoch": 0.10747, + "grad_norm": 1.06969235369634, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10747 + }, + { + "epoch": 0.10748, + "grad_norm": 1.0142439992763128, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 10748 + }, + { + "epoch": 0.10749, + "grad_norm": 1.0685245746096876, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10749 + }, + { + "epoch": 0.1075, + "grad_norm": 0.8449126664648987, + "learning_rate": 0.003, + "loss": 4.09, + "step": 10750 + }, + { + "epoch": 0.10751, + "grad_norm": 0.9743842398056648, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 10751 + }, + { + "epoch": 0.10752, + "grad_norm": 1.1098706461499115, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10752 + }, + { + "epoch": 0.10753, + "grad_norm": 1.1485536672705534, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 10753 + }, + { + "epoch": 0.10754, + "grad_norm": 0.9857930113106791, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 10754 + }, + { + "epoch": 0.10755, + "grad_norm": 1.2508493100595617, + "learning_rate": 0.003, + "loss": 4.11, + "step": 10755 + }, + { + "epoch": 0.10756, + "grad_norm": 0.9426440254849656, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 10756 + }, + { + "epoch": 0.10757, + "grad_norm": 1.2917432581035484, + "learning_rate": 0.003, + "loss": 4.097, + "step": 10757 + }, + { + "epoch": 0.10758, + "grad_norm": 1.0086154909276714, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 10758 + }, + { + "epoch": 0.10759, + "grad_norm": 1.3374586895422267, + "learning_rate": 0.003, + "loss": 4.063, + "step": 10759 + }, + { + "epoch": 0.1076, + "grad_norm": 0.8816798826741424, + "learning_rate": 0.003, + "loss": 4.092, + "step": 10760 + }, + { + "epoch": 0.10761, + "grad_norm": 0.7655410742706316, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 10761 + }, + { + "epoch": 0.10762, + "grad_norm": 0.9097709470196474, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 10762 + }, + { + "epoch": 0.10763, + "grad_norm": 1.1233632654664232, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10763 + }, + { + "epoch": 0.10764, + "grad_norm": 0.8479098631341053, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10764 + }, + { + "epoch": 0.10765, + "grad_norm": 1.0699190076068852, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 10765 + }, + { + "epoch": 0.10766, + "grad_norm": 1.2865992907381223, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 10766 + }, + { + "epoch": 0.10767, + "grad_norm": 0.9373164003694404, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10767 + }, + { + "epoch": 0.10768, + "grad_norm": 1.0709980796464582, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 10768 + }, + { + "epoch": 0.10769, + "grad_norm": 1.3406893808408222, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10769 + }, + { + "epoch": 0.1077, + "grad_norm": 1.0224000701141334, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 10770 + }, + { + "epoch": 0.10771, + "grad_norm": 1.0526142707839026, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 10771 + }, + { + "epoch": 0.10772, + "grad_norm": 0.9443675613044936, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 10772 + }, + { + "epoch": 0.10773, + "grad_norm": 0.9631051318117351, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 10773 + }, + { + "epoch": 0.10774, + "grad_norm": 1.0549186900463854, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 10774 + }, + { + "epoch": 0.10775, + "grad_norm": 1.2375124249342755, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 10775 + }, + { + "epoch": 0.10776, + "grad_norm": 0.9660424900300653, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 10776 + }, + { + "epoch": 0.10777, + "grad_norm": 1.0703994783409576, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10777 + }, + { + "epoch": 0.10778, + "grad_norm": 1.0264385844846082, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 10778 + }, + { + "epoch": 0.10779, + "grad_norm": 1.3643862274983778, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 10779 + }, + { + "epoch": 0.1078, + "grad_norm": 0.8924594651101884, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 10780 + }, + { + "epoch": 0.10781, + "grad_norm": 0.9716988610336873, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10781 + }, + { + "epoch": 0.10782, + "grad_norm": 1.2408412829715514, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 10782 + }, + { + "epoch": 0.10783, + "grad_norm": 1.115281436405084, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 10783 + }, + { + "epoch": 0.10784, + "grad_norm": 0.9579470079302335, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 10784 + }, + { + "epoch": 0.10785, + "grad_norm": 0.9541684831885465, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10785 + }, + { + "epoch": 0.10786, + "grad_norm": 1.0880236074356275, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 10786 + }, + { + "epoch": 0.10787, + "grad_norm": 0.86760847992642, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 10787 + }, + { + "epoch": 0.10788, + "grad_norm": 0.9992650939886262, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 10788 + }, + { + "epoch": 0.10789, + "grad_norm": 1.2753021365318398, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 10789 + }, + { + "epoch": 0.1079, + "grad_norm": 1.0105579588137055, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 10790 + }, + { + "epoch": 0.10791, + "grad_norm": 1.243559766886193, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 10791 + }, + { + "epoch": 0.10792, + "grad_norm": 1.1291284966148707, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 10792 + }, + { + "epoch": 0.10793, + "grad_norm": 1.1994369192885233, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 10793 + }, + { + "epoch": 0.10794, + "grad_norm": 0.9428615077798324, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 10794 + }, + { + "epoch": 0.10795, + "grad_norm": 0.9872260672261326, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 10795 + }, + { + "epoch": 0.10796, + "grad_norm": 0.9935783558284699, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 10796 + }, + { + "epoch": 0.10797, + "grad_norm": 1.0810295695783516, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 10797 + }, + { + "epoch": 0.10798, + "grad_norm": 1.0770398752097612, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 10798 + }, + { + "epoch": 0.10799, + "grad_norm": 1.1063950750874518, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 10799 + }, + { + "epoch": 0.108, + "grad_norm": 1.0326094754145043, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 10800 + }, + { + "epoch": 0.10801, + "grad_norm": 1.0588200103733418, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 10801 + }, + { + "epoch": 0.10802, + "grad_norm": 1.1974743622830182, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 10802 + }, + { + "epoch": 0.10803, + "grad_norm": 1.0601372020160826, + "learning_rate": 0.003, + "loss": 4.093, + "step": 10803 + }, + { + "epoch": 0.10804, + "grad_norm": 1.1471872600753268, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 10804 + }, + { + "epoch": 0.10805, + "grad_norm": 1.2004049726774384, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10805 + }, + { + "epoch": 0.10806, + "grad_norm": 0.9872725686808971, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 10806 + }, + { + "epoch": 0.10807, + "grad_norm": 1.1116803409225564, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 10807 + }, + { + "epoch": 0.10808, + "grad_norm": 1.0580006606096264, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 10808 + }, + { + "epoch": 0.10809, + "grad_norm": 1.102724868900808, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10809 + }, + { + "epoch": 0.1081, + "grad_norm": 1.073448446686264, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 10810 + }, + { + "epoch": 0.10811, + "grad_norm": 0.9747513001123251, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 10811 + }, + { + "epoch": 0.10812, + "grad_norm": 0.967118322196395, + "learning_rate": 0.003, + "loss": 4.054, + "step": 10812 + }, + { + "epoch": 0.10813, + "grad_norm": 0.9459854684115462, + "learning_rate": 0.003, + "loss": 4.07, + "step": 10813 + }, + { + "epoch": 0.10814, + "grad_norm": 1.1274264516553678, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 10814 + }, + { + "epoch": 0.10815, + "grad_norm": 1.0099841714367115, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 10815 + }, + { + "epoch": 0.10816, + "grad_norm": 1.1528605178413807, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 10816 + }, + { + "epoch": 0.10817, + "grad_norm": 1.3025341293550339, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 10817 + }, + { + "epoch": 0.10818, + "grad_norm": 1.1569271056419455, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 10818 + }, + { + "epoch": 0.10819, + "grad_norm": 1.013481567794959, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 10819 + }, + { + "epoch": 0.1082, + "grad_norm": 1.007612982502211, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10820 + }, + { + "epoch": 0.10821, + "grad_norm": 0.9647199516942806, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10821 + }, + { + "epoch": 0.10822, + "grad_norm": 0.9587875659941222, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 10822 + }, + { + "epoch": 0.10823, + "grad_norm": 0.8512682493653407, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 10823 + }, + { + "epoch": 0.10824, + "grad_norm": 0.8911143440247272, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 10824 + }, + { + "epoch": 0.10825, + "grad_norm": 1.0269152554885081, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 10825 + }, + { + "epoch": 0.10826, + "grad_norm": 1.0423619545724405, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10826 + }, + { + "epoch": 0.10827, + "grad_norm": 1.1771294384202926, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 10827 + }, + { + "epoch": 0.10828, + "grad_norm": 1.2582773741455715, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 10828 + }, + { + "epoch": 0.10829, + "grad_norm": 1.0786632104433649, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 10829 + }, + { + "epoch": 0.1083, + "grad_norm": 1.1644041071796276, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 10830 + }, + { + "epoch": 0.10831, + "grad_norm": 1.023753602537439, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 10831 + }, + { + "epoch": 0.10832, + "grad_norm": 1.0261082103416639, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 10832 + }, + { + "epoch": 0.10833, + "grad_norm": 1.1215552384510263, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10833 + }, + { + "epoch": 0.10834, + "grad_norm": 1.065027071141165, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10834 + }, + { + "epoch": 0.10835, + "grad_norm": 1.0766952436089778, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10835 + }, + { + "epoch": 0.10836, + "grad_norm": 1.0098706745625174, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 10836 + }, + { + "epoch": 0.10837, + "grad_norm": 1.1013680159530501, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 10837 + }, + { + "epoch": 0.10838, + "grad_norm": 1.1065831185757904, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 10838 + }, + { + "epoch": 0.10839, + "grad_norm": 1.1637812774623542, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10839 + }, + { + "epoch": 0.1084, + "grad_norm": 1.148431636090298, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 10840 + }, + { + "epoch": 0.10841, + "grad_norm": 1.2366796360721557, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 10841 + }, + { + "epoch": 0.10842, + "grad_norm": 1.0577104738593404, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 10842 + }, + { + "epoch": 0.10843, + "grad_norm": 1.0484337010776401, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 10843 + }, + { + "epoch": 0.10844, + "grad_norm": 1.212192275320524, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 10844 + }, + { + "epoch": 0.10845, + "grad_norm": 0.9957500087946555, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 10845 + }, + { + "epoch": 0.10846, + "grad_norm": 1.0979094349107996, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 10846 + }, + { + "epoch": 0.10847, + "grad_norm": 0.9969902766004411, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 10847 + }, + { + "epoch": 0.10848, + "grad_norm": 1.252783740484669, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 10848 + }, + { + "epoch": 0.10849, + "grad_norm": 1.139595362233358, + "learning_rate": 0.003, + "loss": 4.065, + "step": 10849 + }, + { + "epoch": 0.1085, + "grad_norm": 1.1030600445586416, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10850 + }, + { + "epoch": 0.10851, + "grad_norm": 1.0384848869505103, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 10851 + }, + { + "epoch": 0.10852, + "grad_norm": 0.9121348149264054, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 10852 + }, + { + "epoch": 0.10853, + "grad_norm": 0.8855881354922143, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 10853 + }, + { + "epoch": 0.10854, + "grad_norm": 1.1175936911572262, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 10854 + }, + { + "epoch": 0.10855, + "grad_norm": 1.0245129092214602, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 10855 + }, + { + "epoch": 0.10856, + "grad_norm": 1.119836326344118, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 10856 + }, + { + "epoch": 0.10857, + "grad_norm": 0.9836679344334276, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10857 + }, + { + "epoch": 0.10858, + "grad_norm": 0.9511844758669555, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 10858 + }, + { + "epoch": 0.10859, + "grad_norm": 0.9264557593108973, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 10859 + }, + { + "epoch": 0.1086, + "grad_norm": 1.1703961872501316, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 10860 + }, + { + "epoch": 0.10861, + "grad_norm": 1.2581401538656547, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 10861 + }, + { + "epoch": 0.10862, + "grad_norm": 1.2444168962017241, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 10862 + }, + { + "epoch": 0.10863, + "grad_norm": 0.9504300595021578, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 10863 + }, + { + "epoch": 0.10864, + "grad_norm": 0.9554197526092899, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10864 + }, + { + "epoch": 0.10865, + "grad_norm": 1.2014661969325289, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 10865 + }, + { + "epoch": 0.10866, + "grad_norm": 0.952712085935606, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 10866 + }, + { + "epoch": 0.10867, + "grad_norm": 0.8805538915855081, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 10867 + }, + { + "epoch": 0.10868, + "grad_norm": 1.1935095308976902, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 10868 + }, + { + "epoch": 0.10869, + "grad_norm": 1.0856390824449236, + "learning_rate": 0.003, + "loss": 4.082, + "step": 10869 + }, + { + "epoch": 0.1087, + "grad_norm": 1.1514811577286823, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 10870 + }, + { + "epoch": 0.10871, + "grad_norm": 1.0182709548176088, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 10871 + }, + { + "epoch": 0.10872, + "grad_norm": 1.0620238002252582, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 10872 + }, + { + "epoch": 0.10873, + "grad_norm": 1.045749830039589, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 10873 + }, + { + "epoch": 0.10874, + "grad_norm": 1.2175684337392043, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10874 + }, + { + "epoch": 0.10875, + "grad_norm": 0.9961669916986353, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 10875 + }, + { + "epoch": 0.10876, + "grad_norm": 1.1138750412136187, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10876 + }, + { + "epoch": 0.10877, + "grad_norm": 0.9856995941163944, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10877 + }, + { + "epoch": 0.10878, + "grad_norm": 1.1563223710627515, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 10878 + }, + { + "epoch": 0.10879, + "grad_norm": 0.8985461595260961, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10879 + }, + { + "epoch": 0.1088, + "grad_norm": 0.9295949582223914, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 10880 + }, + { + "epoch": 0.10881, + "grad_norm": 1.0211027521470517, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 10881 + }, + { + "epoch": 0.10882, + "grad_norm": 1.0248546264914715, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 10882 + }, + { + "epoch": 0.10883, + "grad_norm": 1.0416183187131618, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 10883 + }, + { + "epoch": 0.10884, + "grad_norm": 1.164528405126475, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 10884 + }, + { + "epoch": 0.10885, + "grad_norm": 0.9203263062179105, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 10885 + }, + { + "epoch": 0.10886, + "grad_norm": 1.3593110655231138, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 10886 + }, + { + "epoch": 0.10887, + "grad_norm": 1.2827560760894003, + "learning_rate": 0.003, + "loss": 4.111, + "step": 10887 + }, + { + "epoch": 0.10888, + "grad_norm": 0.8780693627092314, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 10888 + }, + { + "epoch": 0.10889, + "grad_norm": 0.8007569440304311, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 10889 + }, + { + "epoch": 0.1089, + "grad_norm": 0.89074198244628, + "learning_rate": 0.003, + "loss": 4.058, + "step": 10890 + }, + { + "epoch": 0.10891, + "grad_norm": 0.9352229500355524, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 10891 + }, + { + "epoch": 0.10892, + "grad_norm": 1.0574403109866946, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 10892 + }, + { + "epoch": 0.10893, + "grad_norm": 1.213448491555586, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 10893 + }, + { + "epoch": 0.10894, + "grad_norm": 1.0633428552392232, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 10894 + }, + { + "epoch": 0.10895, + "grad_norm": 1.2172544183203573, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 10895 + }, + { + "epoch": 0.10896, + "grad_norm": 1.0344575054322562, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10896 + }, + { + "epoch": 0.10897, + "grad_norm": 1.2888555732863958, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 10897 + }, + { + "epoch": 0.10898, + "grad_norm": 0.9468165429324117, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 10898 + }, + { + "epoch": 0.10899, + "grad_norm": 0.9937944702570864, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 10899 + }, + { + "epoch": 0.109, + "grad_norm": 1.0784998893580096, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 10900 + }, + { + "epoch": 0.10901, + "grad_norm": 0.9586474239659144, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 10901 + }, + { + "epoch": 0.10902, + "grad_norm": 1.2612176071375183, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 10902 + }, + { + "epoch": 0.10903, + "grad_norm": 1.2552020722155313, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10903 + }, + { + "epoch": 0.10904, + "grad_norm": 1.1986973542170098, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 10904 + }, + { + "epoch": 0.10905, + "grad_norm": 1.1362325615742652, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10905 + }, + { + "epoch": 0.10906, + "grad_norm": 0.9680044227786018, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 10906 + }, + { + "epoch": 0.10907, + "grad_norm": 1.266038411305887, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 10907 + }, + { + "epoch": 0.10908, + "grad_norm": 0.8194233694583234, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 10908 + }, + { + "epoch": 0.10909, + "grad_norm": 0.9516173062517561, + "learning_rate": 0.003, + "loss": 4.065, + "step": 10909 + }, + { + "epoch": 0.1091, + "grad_norm": 1.190532497934995, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10910 + }, + { + "epoch": 0.10911, + "grad_norm": 1.062507805821197, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 10911 + }, + { + "epoch": 0.10912, + "grad_norm": 0.935468487016387, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10912 + }, + { + "epoch": 0.10913, + "grad_norm": 1.0332631560712264, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 10913 + }, + { + "epoch": 0.10914, + "grad_norm": 1.2241198062781187, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 10914 + }, + { + "epoch": 0.10915, + "grad_norm": 0.9200270302646114, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 10915 + }, + { + "epoch": 0.10916, + "grad_norm": 1.020811811570247, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 10916 + }, + { + "epoch": 0.10917, + "grad_norm": 1.1022325333127732, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 10917 + }, + { + "epoch": 0.10918, + "grad_norm": 1.1836549733574424, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 10918 + }, + { + "epoch": 0.10919, + "grad_norm": 1.170550896066117, + "learning_rate": 0.003, + "loss": 4.072, + "step": 10919 + }, + { + "epoch": 0.1092, + "grad_norm": 0.9028965097833227, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 10920 + }, + { + "epoch": 0.10921, + "grad_norm": 1.0067120812866543, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 10921 + }, + { + "epoch": 0.10922, + "grad_norm": 1.1613890341872082, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10922 + }, + { + "epoch": 0.10923, + "grad_norm": 1.0303548671060356, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10923 + }, + { + "epoch": 0.10924, + "grad_norm": 1.058935100561485, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 10924 + }, + { + "epoch": 0.10925, + "grad_norm": 0.982292907773461, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 10925 + }, + { + "epoch": 0.10926, + "grad_norm": 1.2578268516771849, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 10926 + }, + { + "epoch": 0.10927, + "grad_norm": 1.0731552123121946, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 10927 + }, + { + "epoch": 0.10928, + "grad_norm": 1.0945521448743394, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 10928 + }, + { + "epoch": 0.10929, + "grad_norm": 1.200417936854523, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 10929 + }, + { + "epoch": 0.1093, + "grad_norm": 0.958607621161436, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10930 + }, + { + "epoch": 0.10931, + "grad_norm": 1.066110489708051, + "learning_rate": 0.003, + "loss": 4.095, + "step": 10931 + }, + { + "epoch": 0.10932, + "grad_norm": 0.9762745373990046, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10932 + }, + { + "epoch": 0.10933, + "grad_norm": 1.2896825869107031, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 10933 + }, + { + "epoch": 0.10934, + "grad_norm": 0.9507052260715008, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 10934 + }, + { + "epoch": 0.10935, + "grad_norm": 1.3124453372335498, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 10935 + }, + { + "epoch": 0.10936, + "grad_norm": 0.852998866567921, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 10936 + }, + { + "epoch": 0.10937, + "grad_norm": 0.9231917675292967, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 10937 + }, + { + "epoch": 0.10938, + "grad_norm": 0.9979776375461109, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 10938 + }, + { + "epoch": 0.10939, + "grad_norm": 1.2108718834019867, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 10939 + }, + { + "epoch": 0.1094, + "grad_norm": 1.0380696008785413, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 10940 + }, + { + "epoch": 0.10941, + "grad_norm": 1.2623260185642127, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 10941 + }, + { + "epoch": 0.10942, + "grad_norm": 0.9668178565519607, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 10942 + }, + { + "epoch": 0.10943, + "grad_norm": 1.0862618944194453, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 10943 + }, + { + "epoch": 0.10944, + "grad_norm": 0.8509290193448615, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 10944 + }, + { + "epoch": 0.10945, + "grad_norm": 0.9300723483707174, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 10945 + }, + { + "epoch": 0.10946, + "grad_norm": 1.1516366707370755, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10946 + }, + { + "epoch": 0.10947, + "grad_norm": 1.0628248021466853, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 10947 + }, + { + "epoch": 0.10948, + "grad_norm": 1.1222961873727684, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 10948 + }, + { + "epoch": 0.10949, + "grad_norm": 1.1146700599169557, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 10949 + }, + { + "epoch": 0.1095, + "grad_norm": 0.9290653295123669, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 10950 + }, + { + "epoch": 0.10951, + "grad_norm": 1.0203598045295497, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 10951 + }, + { + "epoch": 0.10952, + "grad_norm": 1.3264874444048882, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10952 + }, + { + "epoch": 0.10953, + "grad_norm": 0.97658568631152, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 10953 + }, + { + "epoch": 0.10954, + "grad_norm": 1.3420224388646687, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 10954 + }, + { + "epoch": 0.10955, + "grad_norm": 1.0692922240105442, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 10955 + }, + { + "epoch": 0.10956, + "grad_norm": 1.1967597984232787, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 10956 + }, + { + "epoch": 0.10957, + "grad_norm": 1.0369332211726097, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10957 + }, + { + "epoch": 0.10958, + "grad_norm": 1.0667165567030539, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 10958 + }, + { + "epoch": 0.10959, + "grad_norm": 1.0467307328354403, + "learning_rate": 0.003, + "loss": 4.091, + "step": 10959 + }, + { + "epoch": 0.1096, + "grad_norm": 1.121746796410831, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 10960 + }, + { + "epoch": 0.10961, + "grad_norm": 1.1066837634594826, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 10961 + }, + { + "epoch": 0.10962, + "grad_norm": 1.0270607977048283, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 10962 + }, + { + "epoch": 0.10963, + "grad_norm": 0.9476690702771576, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 10963 + }, + { + "epoch": 0.10964, + "grad_norm": 1.0618940504526093, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 10964 + }, + { + "epoch": 0.10965, + "grad_norm": 1.2518627931195352, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10965 + }, + { + "epoch": 0.10966, + "grad_norm": 1.1576505895330684, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 10966 + }, + { + "epoch": 0.10967, + "grad_norm": 0.8989888307735112, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 10967 + }, + { + "epoch": 0.10968, + "grad_norm": 0.9840336646590621, + "learning_rate": 0.003, + "loss": 4.101, + "step": 10968 + }, + { + "epoch": 0.10969, + "grad_norm": 1.0738411321450865, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 10969 + }, + { + "epoch": 0.1097, + "grad_norm": 1.1647834960677774, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 10970 + }, + { + "epoch": 0.10971, + "grad_norm": 1.0416427485833692, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 10971 + }, + { + "epoch": 0.10972, + "grad_norm": 1.1140608228352962, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 10972 + }, + { + "epoch": 0.10973, + "grad_norm": 1.0655226246249958, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 10973 + }, + { + "epoch": 0.10974, + "grad_norm": 0.9914871581858634, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 10974 + }, + { + "epoch": 0.10975, + "grad_norm": 0.9698908639834418, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10975 + }, + { + "epoch": 0.10976, + "grad_norm": 1.241829245843147, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 10976 + }, + { + "epoch": 0.10977, + "grad_norm": 1.0944763139029807, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 10977 + }, + { + "epoch": 0.10978, + "grad_norm": 1.0969877929366048, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 10978 + }, + { + "epoch": 0.10979, + "grad_norm": 1.1257370588440034, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 10979 + }, + { + "epoch": 0.1098, + "grad_norm": 0.9569312789633604, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 10980 + }, + { + "epoch": 0.10981, + "grad_norm": 1.1566047577430134, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 10981 + }, + { + "epoch": 0.10982, + "grad_norm": 1.0417786216342104, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 10982 + }, + { + "epoch": 0.10983, + "grad_norm": 1.212286634724521, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 10983 + }, + { + "epoch": 0.10984, + "grad_norm": 1.0337908704581455, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 10984 + }, + { + "epoch": 0.10985, + "grad_norm": 1.1192563238020419, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 10985 + }, + { + "epoch": 0.10986, + "grad_norm": 0.965393768299253, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 10986 + }, + { + "epoch": 0.10987, + "grad_norm": 1.2578089988837124, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 10987 + }, + { + "epoch": 0.10988, + "grad_norm": 0.8914465824720181, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 10988 + }, + { + "epoch": 0.10989, + "grad_norm": 1.0285487793885861, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 10989 + }, + { + "epoch": 0.1099, + "grad_norm": 1.453015158533003, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 10990 + }, + { + "epoch": 0.10991, + "grad_norm": 1.0961809557509807, + "learning_rate": 0.003, + "loss": 4.094, + "step": 10991 + }, + { + "epoch": 0.10992, + "grad_norm": 1.2656872601856095, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 10992 + }, + { + "epoch": 0.10993, + "grad_norm": 1.0578502295128602, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 10993 + }, + { + "epoch": 0.10994, + "grad_norm": 0.9979520192054689, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 10994 + }, + { + "epoch": 0.10995, + "grad_norm": 1.1815914008752326, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 10995 + }, + { + "epoch": 0.10996, + "grad_norm": 1.0807357231606585, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 10996 + }, + { + "epoch": 0.10997, + "grad_norm": 1.012982864418392, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 10997 + }, + { + "epoch": 0.10998, + "grad_norm": 1.151115580492071, + "learning_rate": 0.003, + "loss": 4.1192, + "step": 10998 + }, + { + "epoch": 0.10999, + "grad_norm": 0.9517386329626287, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 10999 + }, + { + "epoch": 0.11, + "grad_norm": 0.9617818605873265, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 11000 + }, + { + "epoch": 0.11001, + "grad_norm": 1.0372420260839046, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 11001 + }, + { + "epoch": 0.11002, + "grad_norm": 1.130538854053489, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 11002 + }, + { + "epoch": 0.11003, + "grad_norm": 1.281765199831236, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 11003 + }, + { + "epoch": 0.11004, + "grad_norm": 1.0930629510140741, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 11004 + }, + { + "epoch": 0.11005, + "grad_norm": 1.0734309697476472, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11005 + }, + { + "epoch": 0.11006, + "grad_norm": 0.9376530387047713, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 11006 + }, + { + "epoch": 0.11007, + "grad_norm": 1.0045813713815963, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11007 + }, + { + "epoch": 0.11008, + "grad_norm": 1.1196806585546806, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 11008 + }, + { + "epoch": 0.11009, + "grad_norm": 1.0021801710470826, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 11009 + }, + { + "epoch": 0.1101, + "grad_norm": 1.1943219579368831, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 11010 + }, + { + "epoch": 0.11011, + "grad_norm": 0.9349293317995805, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 11011 + }, + { + "epoch": 0.11012, + "grad_norm": 1.0816389241323243, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11012 + }, + { + "epoch": 0.11013, + "grad_norm": 0.8440526839080905, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 11013 + }, + { + "epoch": 0.11014, + "grad_norm": 0.9622663832307063, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11014 + }, + { + "epoch": 0.11015, + "grad_norm": 1.0314368411196604, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 11015 + }, + { + "epoch": 0.11016, + "grad_norm": 1.0418344216916757, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11016 + }, + { + "epoch": 0.11017, + "grad_norm": 1.109456113280434, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 11017 + }, + { + "epoch": 0.11018, + "grad_norm": 1.1317173100404638, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11018 + }, + { + "epoch": 0.11019, + "grad_norm": 1.3365549493223279, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11019 + }, + { + "epoch": 0.1102, + "grad_norm": 1.140450402764878, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 11020 + }, + { + "epoch": 0.11021, + "grad_norm": 1.089641104327388, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 11021 + }, + { + "epoch": 0.11022, + "grad_norm": 1.0571029585661946, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 11022 + }, + { + "epoch": 0.11023, + "grad_norm": 0.9112995162650028, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 11023 + }, + { + "epoch": 0.11024, + "grad_norm": 1.0823248116168254, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 11024 + }, + { + "epoch": 0.11025, + "grad_norm": 1.269798404786744, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 11025 + }, + { + "epoch": 0.11026, + "grad_norm": 1.0179991697387427, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11026 + }, + { + "epoch": 0.11027, + "grad_norm": 1.283354079509163, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 11027 + }, + { + "epoch": 0.11028, + "grad_norm": 0.9881442149331964, + "learning_rate": 0.003, + "loss": 4.076, + "step": 11028 + }, + { + "epoch": 0.11029, + "grad_norm": 1.1506448604245498, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 11029 + }, + { + "epoch": 0.1103, + "grad_norm": 1.0701109805534972, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11030 + }, + { + "epoch": 0.11031, + "grad_norm": 1.045008921347305, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 11031 + }, + { + "epoch": 0.11032, + "grad_norm": 0.9940343193835519, + "learning_rate": 0.003, + "loss": 4.095, + "step": 11032 + }, + { + "epoch": 0.11033, + "grad_norm": 1.2452726277820456, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11033 + }, + { + "epoch": 0.11034, + "grad_norm": 0.8438970698928558, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11034 + }, + { + "epoch": 0.11035, + "grad_norm": 1.08785159976017, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11035 + }, + { + "epoch": 0.11036, + "grad_norm": 1.3447184537992016, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 11036 + }, + { + "epoch": 0.11037, + "grad_norm": 0.879399892817881, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 11037 + }, + { + "epoch": 0.11038, + "grad_norm": 0.9955476735433527, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11038 + }, + { + "epoch": 0.11039, + "grad_norm": 1.3410411619156724, + "learning_rate": 0.003, + "loss": 4.088, + "step": 11039 + }, + { + "epoch": 0.1104, + "grad_norm": 0.823843991929917, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 11040 + }, + { + "epoch": 0.11041, + "grad_norm": 0.957799568021374, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11041 + }, + { + "epoch": 0.11042, + "grad_norm": 1.2027831600136363, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11042 + }, + { + "epoch": 0.11043, + "grad_norm": 1.2711340978021695, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 11043 + }, + { + "epoch": 0.11044, + "grad_norm": 0.9926544740714466, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 11044 + }, + { + "epoch": 0.11045, + "grad_norm": 0.9737001715578901, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 11045 + }, + { + "epoch": 0.11046, + "grad_norm": 1.025534429126934, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 11046 + }, + { + "epoch": 0.11047, + "grad_norm": 1.0178943889938226, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 11047 + }, + { + "epoch": 0.11048, + "grad_norm": 1.1132590043193709, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11048 + }, + { + "epoch": 0.11049, + "grad_norm": 1.1688632704206605, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11049 + }, + { + "epoch": 0.1105, + "grad_norm": 0.9935437751692451, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11050 + }, + { + "epoch": 0.11051, + "grad_norm": 0.9625585983183765, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 11051 + }, + { + "epoch": 0.11052, + "grad_norm": 1.0492638687163323, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 11052 + }, + { + "epoch": 0.11053, + "grad_norm": 1.1494360439828093, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11053 + }, + { + "epoch": 0.11054, + "grad_norm": 1.1748136441795494, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 11054 + }, + { + "epoch": 0.11055, + "grad_norm": 1.1836800651405188, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11055 + }, + { + "epoch": 0.11056, + "grad_norm": 1.0175350898679119, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11056 + }, + { + "epoch": 0.11057, + "grad_norm": 1.1406708256575615, + "learning_rate": 0.003, + "loss": 4.068, + "step": 11057 + }, + { + "epoch": 0.11058, + "grad_norm": 0.9332127881921526, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 11058 + }, + { + "epoch": 0.11059, + "grad_norm": 0.8981176131082979, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 11059 + }, + { + "epoch": 0.1106, + "grad_norm": 1.067628186409739, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11060 + }, + { + "epoch": 0.11061, + "grad_norm": 1.2663751330065962, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 11061 + }, + { + "epoch": 0.11062, + "grad_norm": 0.9825686470798557, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11062 + }, + { + "epoch": 0.11063, + "grad_norm": 1.339545896310836, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11063 + }, + { + "epoch": 0.11064, + "grad_norm": 0.8470417821578471, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 11064 + }, + { + "epoch": 0.11065, + "grad_norm": 1.0311320495760354, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 11065 + }, + { + "epoch": 0.11066, + "grad_norm": 1.275046723295672, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 11066 + }, + { + "epoch": 0.11067, + "grad_norm": 0.9743425862414491, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 11067 + }, + { + "epoch": 0.11068, + "grad_norm": 1.1291392189378937, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 11068 + }, + { + "epoch": 0.11069, + "grad_norm": 1.0381149065784203, + "learning_rate": 0.003, + "loss": 4.059, + "step": 11069 + }, + { + "epoch": 0.1107, + "grad_norm": 1.0261726634476829, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 11070 + }, + { + "epoch": 0.11071, + "grad_norm": 1.2432042063184994, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11071 + }, + { + "epoch": 0.11072, + "grad_norm": 1.00271066051892, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 11072 + }, + { + "epoch": 0.11073, + "grad_norm": 1.2876725887592981, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 11073 + }, + { + "epoch": 0.11074, + "grad_norm": 1.0237671279976865, + "learning_rate": 0.003, + "loss": 4.075, + "step": 11074 + }, + { + "epoch": 0.11075, + "grad_norm": 1.1387702287135297, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 11075 + }, + { + "epoch": 0.11076, + "grad_norm": 1.082095577956758, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11076 + }, + { + "epoch": 0.11077, + "grad_norm": 1.075721944065049, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 11077 + }, + { + "epoch": 0.11078, + "grad_norm": 1.2229269824042177, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 11078 + }, + { + "epoch": 0.11079, + "grad_norm": 0.9430305345548962, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 11079 + }, + { + "epoch": 0.1108, + "grad_norm": 1.048335653430088, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11080 + }, + { + "epoch": 0.11081, + "grad_norm": 1.0976125542943163, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11081 + }, + { + "epoch": 0.11082, + "grad_norm": 1.1782408105639972, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11082 + }, + { + "epoch": 0.11083, + "grad_norm": 1.0710585939603285, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11083 + }, + { + "epoch": 0.11084, + "grad_norm": 1.294803945413708, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11084 + }, + { + "epoch": 0.11085, + "grad_norm": 1.078729240870453, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 11085 + }, + { + "epoch": 0.11086, + "grad_norm": 1.1632020820573332, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 11086 + }, + { + "epoch": 0.11087, + "grad_norm": 0.9669066808716378, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11087 + }, + { + "epoch": 0.11088, + "grad_norm": 1.0850182096865149, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 11088 + }, + { + "epoch": 0.11089, + "grad_norm": 1.2549161161581106, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 11089 + }, + { + "epoch": 0.1109, + "grad_norm": 1.0164081938030312, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 11090 + }, + { + "epoch": 0.11091, + "grad_norm": 1.1071292805706272, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 11091 + }, + { + "epoch": 0.11092, + "grad_norm": 1.1062151657057917, + "learning_rate": 0.003, + "loss": 4.074, + "step": 11092 + }, + { + "epoch": 0.11093, + "grad_norm": 1.1232047830134266, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 11093 + }, + { + "epoch": 0.11094, + "grad_norm": 0.8778638234580115, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 11094 + }, + { + "epoch": 0.11095, + "grad_norm": 0.930227028930981, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 11095 + }, + { + "epoch": 0.11096, + "grad_norm": 1.0920410338682343, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 11096 + }, + { + "epoch": 0.11097, + "grad_norm": 1.2024142449099442, + "learning_rate": 0.003, + "loss": 4.096, + "step": 11097 + }, + { + "epoch": 0.11098, + "grad_norm": 1.1127225562622876, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 11098 + }, + { + "epoch": 0.11099, + "grad_norm": 1.0323801131506445, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11099 + }, + { + "epoch": 0.111, + "grad_norm": 1.1710136677384504, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 11100 + }, + { + "epoch": 0.11101, + "grad_norm": 0.8840824981939511, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 11101 + }, + { + "epoch": 0.11102, + "grad_norm": 0.8862206579405567, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 11102 + }, + { + "epoch": 0.11103, + "grad_norm": 1.1293426398135202, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 11103 + }, + { + "epoch": 0.11104, + "grad_norm": 1.1343343190329958, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11104 + }, + { + "epoch": 0.11105, + "grad_norm": 1.073432893320314, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11105 + }, + { + "epoch": 0.11106, + "grad_norm": 1.0142594610929732, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 11106 + }, + { + "epoch": 0.11107, + "grad_norm": 1.154167579934993, + "learning_rate": 0.003, + "loss": 4.07, + "step": 11107 + }, + { + "epoch": 0.11108, + "grad_norm": 1.0955201894180502, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 11108 + }, + { + "epoch": 0.11109, + "grad_norm": 1.0748307909472112, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 11109 + }, + { + "epoch": 0.1111, + "grad_norm": 1.1984134922116756, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 11110 + }, + { + "epoch": 0.11111, + "grad_norm": 1.023062750460469, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 11111 + }, + { + "epoch": 0.11112, + "grad_norm": 1.148035200097653, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 11112 + }, + { + "epoch": 0.11113, + "grad_norm": 0.9643801573215892, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 11113 + }, + { + "epoch": 0.11114, + "grad_norm": 1.1451404200430764, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 11114 + }, + { + "epoch": 0.11115, + "grad_norm": 1.1395071268509465, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 11115 + }, + { + "epoch": 0.11116, + "grad_norm": 1.3169506565149234, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 11116 + }, + { + "epoch": 0.11117, + "grad_norm": 1.2272020840995974, + "learning_rate": 0.003, + "loss": 4.11, + "step": 11117 + }, + { + "epoch": 0.11118, + "grad_norm": 0.9150172409767166, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 11118 + }, + { + "epoch": 0.11119, + "grad_norm": 1.0060344474151803, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 11119 + }, + { + "epoch": 0.1112, + "grad_norm": 1.1316813330593096, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11120 + }, + { + "epoch": 0.11121, + "grad_norm": 1.1888005089421085, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 11121 + }, + { + "epoch": 0.11122, + "grad_norm": 1.1305955226981466, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 11122 + }, + { + "epoch": 0.11123, + "grad_norm": 1.0833462030744596, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 11123 + }, + { + "epoch": 0.11124, + "grad_norm": 1.0638925302665134, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 11124 + }, + { + "epoch": 0.11125, + "grad_norm": 1.0069967290408879, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 11125 + }, + { + "epoch": 0.11126, + "grad_norm": 1.2261766349487866, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11126 + }, + { + "epoch": 0.11127, + "grad_norm": 1.0028029989186422, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 11127 + }, + { + "epoch": 0.11128, + "grad_norm": 1.1780495201828052, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11128 + }, + { + "epoch": 0.11129, + "grad_norm": 0.8299165332619102, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11129 + }, + { + "epoch": 0.1113, + "grad_norm": 0.8202902758038285, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 11130 + }, + { + "epoch": 0.11131, + "grad_norm": 0.8321204036949327, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 11131 + }, + { + "epoch": 0.11132, + "grad_norm": 0.9729737967538941, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11132 + }, + { + "epoch": 0.11133, + "grad_norm": 1.3548737367613601, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 11133 + }, + { + "epoch": 0.11134, + "grad_norm": 1.1233322420616223, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 11134 + }, + { + "epoch": 0.11135, + "grad_norm": 1.2626200980787057, + "learning_rate": 0.003, + "loss": 4.087, + "step": 11135 + }, + { + "epoch": 0.11136, + "grad_norm": 0.9766533789117673, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 11136 + }, + { + "epoch": 0.11137, + "grad_norm": 1.021474893266102, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 11137 + }, + { + "epoch": 0.11138, + "grad_norm": 1.1274478767407865, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11138 + }, + { + "epoch": 0.11139, + "grad_norm": 1.1267855782303196, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 11139 + }, + { + "epoch": 0.1114, + "grad_norm": 1.0560581171106265, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 11140 + }, + { + "epoch": 0.11141, + "grad_norm": 1.1071428100121916, + "learning_rate": 0.003, + "loss": 4.089, + "step": 11141 + }, + { + "epoch": 0.11142, + "grad_norm": 1.049316649242421, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 11142 + }, + { + "epoch": 0.11143, + "grad_norm": 1.1684822588023638, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11143 + }, + { + "epoch": 0.11144, + "grad_norm": 1.0905646220362137, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 11144 + }, + { + "epoch": 0.11145, + "grad_norm": 1.1383713278971077, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 11145 + }, + { + "epoch": 0.11146, + "grad_norm": 1.0662303902298917, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 11146 + }, + { + "epoch": 0.11147, + "grad_norm": 1.136790129915702, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 11147 + }, + { + "epoch": 0.11148, + "grad_norm": 1.1811233013860687, + "learning_rate": 0.003, + "loss": 4.087, + "step": 11148 + }, + { + "epoch": 0.11149, + "grad_norm": 1.0329754282368866, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 11149 + }, + { + "epoch": 0.1115, + "grad_norm": 1.1162248343049477, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11150 + }, + { + "epoch": 0.11151, + "grad_norm": 1.0803414568954743, + "learning_rate": 0.003, + "loss": 4.102, + "step": 11151 + }, + { + "epoch": 0.11152, + "grad_norm": 0.9621437109509381, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 11152 + }, + { + "epoch": 0.11153, + "grad_norm": 1.0074645368691362, + "learning_rate": 0.003, + "loss": 4.094, + "step": 11153 + }, + { + "epoch": 0.11154, + "grad_norm": 1.3297829766491356, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 11154 + }, + { + "epoch": 0.11155, + "grad_norm": 0.917150674753661, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 11155 + }, + { + "epoch": 0.11156, + "grad_norm": 1.0716683342142723, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11156 + }, + { + "epoch": 0.11157, + "grad_norm": 0.9908296621704735, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 11157 + }, + { + "epoch": 0.11158, + "grad_norm": 1.195847729550385, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 11158 + }, + { + "epoch": 0.11159, + "grad_norm": 0.9140159093620859, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 11159 + }, + { + "epoch": 0.1116, + "grad_norm": 1.0693262870879578, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 11160 + }, + { + "epoch": 0.11161, + "grad_norm": 1.3174320581418724, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 11161 + }, + { + "epoch": 0.11162, + "grad_norm": 0.8934094400104304, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11162 + }, + { + "epoch": 0.11163, + "grad_norm": 0.9190953694678776, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 11163 + }, + { + "epoch": 0.11164, + "grad_norm": 0.9907049440370059, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 11164 + }, + { + "epoch": 0.11165, + "grad_norm": 1.1572973589976452, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11165 + }, + { + "epoch": 0.11166, + "grad_norm": 1.2329249741723933, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 11166 + }, + { + "epoch": 0.11167, + "grad_norm": 1.0903320835115502, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 11167 + }, + { + "epoch": 0.11168, + "grad_norm": 1.2283881015667557, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11168 + }, + { + "epoch": 0.11169, + "grad_norm": 1.052811405237688, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 11169 + }, + { + "epoch": 0.1117, + "grad_norm": 1.1879599635362192, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 11170 + }, + { + "epoch": 0.11171, + "grad_norm": 1.1562509517323158, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 11171 + }, + { + "epoch": 0.11172, + "grad_norm": 1.0746628555807403, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 11172 + }, + { + "epoch": 0.11173, + "grad_norm": 1.1465154807026283, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11173 + }, + { + "epoch": 0.11174, + "grad_norm": 1.1107132259817063, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11174 + }, + { + "epoch": 0.11175, + "grad_norm": 1.4460414667703407, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 11175 + }, + { + "epoch": 0.11176, + "grad_norm": 0.8687107072072395, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 11176 + }, + { + "epoch": 0.11177, + "grad_norm": 0.9451656942827614, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 11177 + }, + { + "epoch": 0.11178, + "grad_norm": 1.273691659752799, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 11178 + }, + { + "epoch": 0.11179, + "grad_norm": 1.1851241645193193, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11179 + }, + { + "epoch": 0.1118, + "grad_norm": 0.9995561306224513, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 11180 + }, + { + "epoch": 0.11181, + "grad_norm": 1.0561104495198952, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11181 + }, + { + "epoch": 0.11182, + "grad_norm": 1.2094738379122607, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 11182 + }, + { + "epoch": 0.11183, + "grad_norm": 0.9389470450058727, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11183 + }, + { + "epoch": 0.11184, + "grad_norm": 0.8960144214081149, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11184 + }, + { + "epoch": 0.11185, + "grad_norm": 0.9677063470994468, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 11185 + }, + { + "epoch": 0.11186, + "grad_norm": 1.0636641010940282, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11186 + }, + { + "epoch": 0.11187, + "grad_norm": 1.0042556361054829, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 11187 + }, + { + "epoch": 0.11188, + "grad_norm": 1.1804908575937623, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 11188 + }, + { + "epoch": 0.11189, + "grad_norm": 0.9760623848126584, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 11189 + }, + { + "epoch": 0.1119, + "grad_norm": 1.0655847280266737, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 11190 + }, + { + "epoch": 0.11191, + "grad_norm": 1.1804808531390547, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 11191 + }, + { + "epoch": 0.11192, + "grad_norm": 1.0417522219563624, + "learning_rate": 0.003, + "loss": 4.11, + "step": 11192 + }, + { + "epoch": 0.11193, + "grad_norm": 1.2328046581965018, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11193 + }, + { + "epoch": 0.11194, + "grad_norm": 0.9678481952515546, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 11194 + }, + { + "epoch": 0.11195, + "grad_norm": 1.0849728191438266, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11195 + }, + { + "epoch": 0.11196, + "grad_norm": 1.137568879220585, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11196 + }, + { + "epoch": 0.11197, + "grad_norm": 1.0788625907543243, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 11197 + }, + { + "epoch": 0.11198, + "grad_norm": 1.200279594942393, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11198 + }, + { + "epoch": 0.11199, + "grad_norm": 1.075283856632629, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11199 + }, + { + "epoch": 0.112, + "grad_norm": 1.3496552532591504, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11200 + }, + { + "epoch": 0.11201, + "grad_norm": 0.915277435554514, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11201 + }, + { + "epoch": 0.11202, + "grad_norm": 1.023416164318843, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 11202 + }, + { + "epoch": 0.11203, + "grad_norm": 1.0929941854292053, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 11203 + }, + { + "epoch": 0.11204, + "grad_norm": 0.9980228419712516, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 11204 + }, + { + "epoch": 0.11205, + "grad_norm": 1.1050330304505716, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 11205 + }, + { + "epoch": 0.11206, + "grad_norm": 1.031157908059373, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11206 + }, + { + "epoch": 0.11207, + "grad_norm": 0.9567238626654728, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11207 + }, + { + "epoch": 0.11208, + "grad_norm": 1.0268115166473828, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 11208 + }, + { + "epoch": 0.11209, + "grad_norm": 1.0830399806039088, + "learning_rate": 0.003, + "loss": 4.073, + "step": 11209 + }, + { + "epoch": 0.1121, + "grad_norm": 1.0100548761394696, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 11210 + }, + { + "epoch": 0.11211, + "grad_norm": 1.3731084883013778, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 11211 + }, + { + "epoch": 0.11212, + "grad_norm": 0.9029837035173026, + "learning_rate": 0.003, + "loss": 4.082, + "step": 11212 + }, + { + "epoch": 0.11213, + "grad_norm": 0.8742947566558211, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 11213 + }, + { + "epoch": 0.11214, + "grad_norm": 1.093073657307132, + "learning_rate": 0.003, + "loss": 4.112, + "step": 11214 + }, + { + "epoch": 0.11215, + "grad_norm": 1.150468961400977, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 11215 + }, + { + "epoch": 0.11216, + "grad_norm": 1.1623110083022907, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 11216 + }, + { + "epoch": 0.11217, + "grad_norm": 1.0136109306734828, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 11217 + }, + { + "epoch": 0.11218, + "grad_norm": 1.2225786094760291, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 11218 + }, + { + "epoch": 0.11219, + "grad_norm": 1.0818760771784373, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 11219 + }, + { + "epoch": 0.1122, + "grad_norm": 1.0582777227739197, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 11220 + }, + { + "epoch": 0.11221, + "grad_norm": 1.0362983069658394, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11221 + }, + { + "epoch": 0.11222, + "grad_norm": 1.2275266816383967, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 11222 + }, + { + "epoch": 0.11223, + "grad_norm": 1.0692292653315416, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 11223 + }, + { + "epoch": 0.11224, + "grad_norm": 1.1917186048611152, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 11224 + }, + { + "epoch": 0.11225, + "grad_norm": 1.0387982212954594, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 11225 + }, + { + "epoch": 0.11226, + "grad_norm": 1.3224755641339132, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 11226 + }, + { + "epoch": 0.11227, + "grad_norm": 1.119941539039678, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11227 + }, + { + "epoch": 0.11228, + "grad_norm": 1.2446054902205859, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 11228 + }, + { + "epoch": 0.11229, + "grad_norm": 1.159551158039648, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 11229 + }, + { + "epoch": 0.1123, + "grad_norm": 0.9846382272809041, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11230 + }, + { + "epoch": 0.11231, + "grad_norm": 1.0618869653269813, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 11231 + }, + { + "epoch": 0.11232, + "grad_norm": 1.0924187874072424, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11232 + }, + { + "epoch": 0.11233, + "grad_norm": 1.043653612630689, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 11233 + }, + { + "epoch": 0.11234, + "grad_norm": 1.2055624769452127, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 11234 + }, + { + "epoch": 0.11235, + "grad_norm": 1.0189350713213847, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 11235 + }, + { + "epoch": 0.11236, + "grad_norm": 1.1370600949836156, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 11236 + }, + { + "epoch": 0.11237, + "grad_norm": 1.0502229610345457, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 11237 + }, + { + "epoch": 0.11238, + "grad_norm": 1.3088946159549792, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11238 + }, + { + "epoch": 0.11239, + "grad_norm": 0.9674213958822012, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11239 + }, + { + "epoch": 0.1124, + "grad_norm": 1.1108252118839896, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 11240 + }, + { + "epoch": 0.11241, + "grad_norm": 1.0788904121254563, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 11241 + }, + { + "epoch": 0.11242, + "grad_norm": 0.8805638731782014, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 11242 + }, + { + "epoch": 0.11243, + "grad_norm": 0.953293260942598, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 11243 + }, + { + "epoch": 0.11244, + "grad_norm": 1.0944982541263293, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 11244 + }, + { + "epoch": 0.11245, + "grad_norm": 0.9878904899811639, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 11245 + }, + { + "epoch": 0.11246, + "grad_norm": 1.0169671093460908, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 11246 + }, + { + "epoch": 0.11247, + "grad_norm": 0.9576195622409958, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11247 + }, + { + "epoch": 0.11248, + "grad_norm": 1.1663956402495321, + "learning_rate": 0.003, + "loss": 4.109, + "step": 11248 + }, + { + "epoch": 0.11249, + "grad_norm": 1.0707941267852739, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 11249 + }, + { + "epoch": 0.1125, + "grad_norm": 1.459014431371671, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11250 + }, + { + "epoch": 0.11251, + "grad_norm": 1.1274606042467634, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11251 + }, + { + "epoch": 0.11252, + "grad_norm": 1.110095946102636, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 11252 + }, + { + "epoch": 0.11253, + "grad_norm": 1.1658531601254527, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11253 + }, + { + "epoch": 0.11254, + "grad_norm": 0.9050266401818737, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 11254 + }, + { + "epoch": 0.11255, + "grad_norm": 0.9758153208182531, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11255 + }, + { + "epoch": 0.11256, + "grad_norm": 1.0385188029198487, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11256 + }, + { + "epoch": 0.11257, + "grad_norm": 1.1873576303217848, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 11257 + }, + { + "epoch": 0.11258, + "grad_norm": 1.0026884103018077, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 11258 + }, + { + "epoch": 0.11259, + "grad_norm": 1.1656263767341002, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 11259 + }, + { + "epoch": 0.1126, + "grad_norm": 1.2347238493290456, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11260 + }, + { + "epoch": 0.11261, + "grad_norm": 1.1284817796438356, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11261 + }, + { + "epoch": 0.11262, + "grad_norm": 1.1365957031399199, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11262 + }, + { + "epoch": 0.11263, + "grad_norm": 0.9204181849384226, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 11263 + }, + { + "epoch": 0.11264, + "grad_norm": 0.8164884905461206, + "learning_rate": 0.003, + "loss": 4.072, + "step": 11264 + }, + { + "epoch": 0.11265, + "grad_norm": 0.8990747848873674, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 11265 + }, + { + "epoch": 0.11266, + "grad_norm": 1.0489106348607906, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 11266 + }, + { + "epoch": 0.11267, + "grad_norm": 1.1865102633489928, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 11267 + }, + { + "epoch": 0.11268, + "grad_norm": 1.1467484081323438, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11268 + }, + { + "epoch": 0.11269, + "grad_norm": 1.1552841477401594, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 11269 + }, + { + "epoch": 0.1127, + "grad_norm": 1.0772199439954333, + "learning_rate": 0.003, + "loss": 4.08, + "step": 11270 + }, + { + "epoch": 0.11271, + "grad_norm": 1.0118300121513137, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11271 + }, + { + "epoch": 0.11272, + "grad_norm": 1.1063044542278664, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 11272 + }, + { + "epoch": 0.11273, + "grad_norm": 0.9312828951421663, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 11273 + }, + { + "epoch": 0.11274, + "grad_norm": 0.9940260808226513, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11274 + }, + { + "epoch": 0.11275, + "grad_norm": 1.1446577070018433, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 11275 + }, + { + "epoch": 0.11276, + "grad_norm": 1.0018077554843705, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 11276 + }, + { + "epoch": 0.11277, + "grad_norm": 1.1677079749269628, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11277 + }, + { + "epoch": 0.11278, + "grad_norm": 1.1237391502408127, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 11278 + }, + { + "epoch": 0.11279, + "grad_norm": 1.2019139370235028, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 11279 + }, + { + "epoch": 0.1128, + "grad_norm": 1.2419978997736385, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11280 + }, + { + "epoch": 0.11281, + "grad_norm": 1.0872558401323524, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11281 + }, + { + "epoch": 0.11282, + "grad_norm": 1.209488425363245, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11282 + }, + { + "epoch": 0.11283, + "grad_norm": 0.9858160666512163, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 11283 + }, + { + "epoch": 0.11284, + "grad_norm": 1.00042028888944, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 11284 + }, + { + "epoch": 0.11285, + "grad_norm": 1.206729879952282, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 11285 + }, + { + "epoch": 0.11286, + "grad_norm": 1.058137393919995, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 11286 + }, + { + "epoch": 0.11287, + "grad_norm": 1.1303700238289203, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11287 + }, + { + "epoch": 0.11288, + "grad_norm": 1.0001149604553041, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 11288 + }, + { + "epoch": 0.11289, + "grad_norm": 0.8541549247427613, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 11289 + }, + { + "epoch": 0.1129, + "grad_norm": 1.0899293245321309, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 11290 + }, + { + "epoch": 0.11291, + "grad_norm": 1.2312123995841915, + "learning_rate": 0.003, + "loss": 4.101, + "step": 11291 + }, + { + "epoch": 0.11292, + "grad_norm": 1.0199608911110976, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 11292 + }, + { + "epoch": 0.11293, + "grad_norm": 1.1065648579263287, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 11293 + }, + { + "epoch": 0.11294, + "grad_norm": 0.9875106628926394, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11294 + }, + { + "epoch": 0.11295, + "grad_norm": 1.0024741305766218, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11295 + }, + { + "epoch": 0.11296, + "grad_norm": 1.0033426301135893, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 11296 + }, + { + "epoch": 0.11297, + "grad_norm": 1.1664712252997975, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 11297 + }, + { + "epoch": 0.11298, + "grad_norm": 1.034552101124066, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 11298 + }, + { + "epoch": 0.11299, + "grad_norm": 1.149005602372779, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 11299 + }, + { + "epoch": 0.113, + "grad_norm": 0.9280762195278215, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 11300 + }, + { + "epoch": 0.11301, + "grad_norm": 0.9174481792983767, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 11301 + }, + { + "epoch": 0.11302, + "grad_norm": 1.355403034506496, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 11302 + }, + { + "epoch": 0.11303, + "grad_norm": 1.2515777074789904, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 11303 + }, + { + "epoch": 0.11304, + "grad_norm": 1.026679034534568, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11304 + }, + { + "epoch": 0.11305, + "grad_norm": 1.3542331189203112, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11305 + }, + { + "epoch": 0.11306, + "grad_norm": 1.094094766775474, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 11306 + }, + { + "epoch": 0.11307, + "grad_norm": 1.149755997638481, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11307 + }, + { + "epoch": 0.11308, + "grad_norm": 0.8949282992535843, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 11308 + }, + { + "epoch": 0.11309, + "grad_norm": 0.8965306355096659, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 11309 + }, + { + "epoch": 0.1131, + "grad_norm": 0.8035470699791126, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 11310 + }, + { + "epoch": 0.11311, + "grad_norm": 0.901000387358812, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11311 + }, + { + "epoch": 0.11312, + "grad_norm": 1.1716063743041008, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 11312 + }, + { + "epoch": 0.11313, + "grad_norm": 1.2117160260271929, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 11313 + }, + { + "epoch": 0.11314, + "grad_norm": 1.1322856823309642, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 11314 + }, + { + "epoch": 0.11315, + "grad_norm": 1.1405877982475148, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11315 + }, + { + "epoch": 0.11316, + "grad_norm": 0.9413939056674395, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 11316 + }, + { + "epoch": 0.11317, + "grad_norm": 1.1013015545182125, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11317 + }, + { + "epoch": 0.11318, + "grad_norm": 1.0467991658776217, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 11318 + }, + { + "epoch": 0.11319, + "grad_norm": 1.1550280753143543, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 11319 + }, + { + "epoch": 0.1132, + "grad_norm": 1.1479696936429864, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 11320 + }, + { + "epoch": 0.11321, + "grad_norm": 1.0152156602974067, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 11321 + }, + { + "epoch": 0.11322, + "grad_norm": 1.075778583656459, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11322 + }, + { + "epoch": 0.11323, + "grad_norm": 1.0869881631034657, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 11323 + }, + { + "epoch": 0.11324, + "grad_norm": 1.1804229227385865, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 11324 + }, + { + "epoch": 0.11325, + "grad_norm": 0.8810791102719396, + "learning_rate": 0.003, + "loss": 4.091, + "step": 11325 + }, + { + "epoch": 0.11326, + "grad_norm": 1.072467901629413, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 11326 + }, + { + "epoch": 0.11327, + "grad_norm": 1.2032702548110166, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11327 + }, + { + "epoch": 0.11328, + "grad_norm": 1.2904192484723571, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 11328 + }, + { + "epoch": 0.11329, + "grad_norm": 1.0583280692957269, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 11329 + }, + { + "epoch": 0.1133, + "grad_norm": 0.9699677436263262, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11330 + }, + { + "epoch": 0.11331, + "grad_norm": 1.1451909643330345, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11331 + }, + { + "epoch": 0.11332, + "grad_norm": 1.046111033258354, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11332 + }, + { + "epoch": 0.11333, + "grad_norm": 0.9462258269015973, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11333 + }, + { + "epoch": 0.11334, + "grad_norm": 0.9957039438891643, + "learning_rate": 0.003, + "loss": 4.064, + "step": 11334 + }, + { + "epoch": 0.11335, + "grad_norm": 1.2195131489129374, + "learning_rate": 0.003, + "loss": 4.064, + "step": 11335 + }, + { + "epoch": 0.11336, + "grad_norm": 1.033509006350732, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 11336 + }, + { + "epoch": 0.11337, + "grad_norm": 1.2648851398653458, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 11337 + }, + { + "epoch": 0.11338, + "grad_norm": 1.177703282524506, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11338 + }, + { + "epoch": 0.11339, + "grad_norm": 1.0056354528106204, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 11339 + }, + { + "epoch": 0.1134, + "grad_norm": 1.3867214949166673, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 11340 + }, + { + "epoch": 0.11341, + "grad_norm": 0.9906858050160079, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 11341 + }, + { + "epoch": 0.11342, + "grad_norm": 1.0706164693016034, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11342 + }, + { + "epoch": 0.11343, + "grad_norm": 1.0292726784936213, + "learning_rate": 0.003, + "loss": 4.088, + "step": 11343 + }, + { + "epoch": 0.11344, + "grad_norm": 1.1232646849413133, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 11344 + }, + { + "epoch": 0.11345, + "grad_norm": 1.1124167099644198, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 11345 + }, + { + "epoch": 0.11346, + "grad_norm": 0.9798550547485815, + "learning_rate": 0.003, + "loss": 4.089, + "step": 11346 + }, + { + "epoch": 0.11347, + "grad_norm": 0.9799428380823298, + "learning_rate": 0.003, + "loss": 4.072, + "step": 11347 + }, + { + "epoch": 0.11348, + "grad_norm": 1.0124447590781391, + "learning_rate": 0.003, + "loss": 4.092, + "step": 11348 + }, + { + "epoch": 0.11349, + "grad_norm": 1.318597220274999, + "learning_rate": 0.003, + "loss": 4.098, + "step": 11349 + }, + { + "epoch": 0.1135, + "grad_norm": 1.0490907851050608, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 11350 + }, + { + "epoch": 0.11351, + "grad_norm": 1.1275921563276423, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 11351 + }, + { + "epoch": 0.11352, + "grad_norm": 1.2906503939683447, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 11352 + }, + { + "epoch": 0.11353, + "grad_norm": 1.1088910910099248, + "learning_rate": 0.003, + "loss": 4.068, + "step": 11353 + }, + { + "epoch": 0.11354, + "grad_norm": 1.1908101825995385, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 11354 + }, + { + "epoch": 0.11355, + "grad_norm": 0.9800870108001858, + "learning_rate": 0.003, + "loss": 4.114, + "step": 11355 + }, + { + "epoch": 0.11356, + "grad_norm": 1.2127513148741016, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 11356 + }, + { + "epoch": 0.11357, + "grad_norm": 1.0616934197142416, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 11357 + }, + { + "epoch": 0.11358, + "grad_norm": 1.1793154754632509, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 11358 + }, + { + "epoch": 0.11359, + "grad_norm": 1.0791143615314955, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 11359 + }, + { + "epoch": 0.1136, + "grad_norm": 1.0992938215616361, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 11360 + }, + { + "epoch": 0.11361, + "grad_norm": 0.9884345665722764, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 11361 + }, + { + "epoch": 0.11362, + "grad_norm": 1.1988548038821205, + "learning_rate": 0.003, + "loss": 4.085, + "step": 11362 + }, + { + "epoch": 0.11363, + "grad_norm": 0.9759992044198456, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 11363 + }, + { + "epoch": 0.11364, + "grad_norm": 1.2599692044663442, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 11364 + }, + { + "epoch": 0.11365, + "grad_norm": 0.8745803209168422, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 11365 + }, + { + "epoch": 0.11366, + "grad_norm": 0.8593192580434853, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 11366 + }, + { + "epoch": 0.11367, + "grad_norm": 1.030383110518904, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 11367 + }, + { + "epoch": 0.11368, + "grad_norm": 1.2702615917194533, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11368 + }, + { + "epoch": 0.11369, + "grad_norm": 1.3157323355023252, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 11369 + }, + { + "epoch": 0.1137, + "grad_norm": 0.9479350053364931, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 11370 + }, + { + "epoch": 0.11371, + "grad_norm": 1.0001937118939637, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11371 + }, + { + "epoch": 0.11372, + "grad_norm": 1.2520454896918638, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11372 + }, + { + "epoch": 0.11373, + "grad_norm": 1.0850205495773182, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 11373 + }, + { + "epoch": 0.11374, + "grad_norm": 1.1670106510434204, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 11374 + }, + { + "epoch": 0.11375, + "grad_norm": 0.8561088091995893, + "learning_rate": 0.003, + "loss": 4.088, + "step": 11375 + }, + { + "epoch": 0.11376, + "grad_norm": 0.7637915719569975, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 11376 + }, + { + "epoch": 0.11377, + "grad_norm": 0.9313796841061992, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11377 + }, + { + "epoch": 0.11378, + "grad_norm": 1.1937782116002877, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 11378 + }, + { + "epoch": 0.11379, + "grad_norm": 1.01019759802451, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 11379 + }, + { + "epoch": 0.1138, + "grad_norm": 1.2969923102509442, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 11380 + }, + { + "epoch": 0.11381, + "grad_norm": 1.1287939651559524, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 11381 + }, + { + "epoch": 0.11382, + "grad_norm": 1.2076230352917354, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 11382 + }, + { + "epoch": 0.11383, + "grad_norm": 1.0185750251810939, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 11383 + }, + { + "epoch": 0.11384, + "grad_norm": 1.0371687278136845, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 11384 + }, + { + "epoch": 0.11385, + "grad_norm": 1.2192193474269375, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11385 + }, + { + "epoch": 0.11386, + "grad_norm": 1.1899210254042143, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 11386 + }, + { + "epoch": 0.11387, + "grad_norm": 1.1919981601440022, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11387 + }, + { + "epoch": 0.11388, + "grad_norm": 1.0815046104847168, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 11388 + }, + { + "epoch": 0.11389, + "grad_norm": 1.1584442503144354, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 11389 + }, + { + "epoch": 0.1139, + "grad_norm": 1.1663439723281048, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11390 + }, + { + "epoch": 0.11391, + "grad_norm": 1.0722133583119948, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11391 + }, + { + "epoch": 0.11392, + "grad_norm": 1.2653674305987967, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11392 + }, + { + "epoch": 0.11393, + "grad_norm": 0.9632936012758943, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11393 + }, + { + "epoch": 0.11394, + "grad_norm": 1.2086534538177434, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 11394 + }, + { + "epoch": 0.11395, + "grad_norm": 1.2404795628469127, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 11395 + }, + { + "epoch": 0.11396, + "grad_norm": 1.1143481503393529, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 11396 + }, + { + "epoch": 0.11397, + "grad_norm": 1.2209668324766623, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11397 + }, + { + "epoch": 0.11398, + "grad_norm": 0.9180476890565811, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 11398 + }, + { + "epoch": 0.11399, + "grad_norm": 1.0341543566407236, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11399 + }, + { + "epoch": 0.114, + "grad_norm": 1.2550533601100675, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 11400 + }, + { + "epoch": 0.11401, + "grad_norm": 0.9408742974294715, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 11401 + }, + { + "epoch": 0.11402, + "grad_norm": 1.089989319062244, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 11402 + }, + { + "epoch": 0.11403, + "grad_norm": 0.9617519579250814, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 11403 + }, + { + "epoch": 0.11404, + "grad_norm": 1.094611527464326, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11404 + }, + { + "epoch": 0.11405, + "grad_norm": 1.2924390248074644, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11405 + }, + { + "epoch": 0.11406, + "grad_norm": 0.8613651877594999, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 11406 + }, + { + "epoch": 0.11407, + "grad_norm": 0.9020507368404247, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 11407 + }, + { + "epoch": 0.11408, + "grad_norm": 1.0194826586040544, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 11408 + }, + { + "epoch": 0.11409, + "grad_norm": 1.2170302680890877, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 11409 + }, + { + "epoch": 0.1141, + "grad_norm": 0.9748439760302462, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 11410 + }, + { + "epoch": 0.11411, + "grad_norm": 1.1438379739008626, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 11411 + }, + { + "epoch": 0.11412, + "grad_norm": 1.1914607143578906, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 11412 + }, + { + "epoch": 0.11413, + "grad_norm": 0.9667367633697801, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11413 + }, + { + "epoch": 0.11414, + "grad_norm": 1.0806930724974129, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 11414 + }, + { + "epoch": 0.11415, + "grad_norm": 1.0476544469337348, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11415 + }, + { + "epoch": 0.11416, + "grad_norm": 1.375644241403238, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11416 + }, + { + "epoch": 0.11417, + "grad_norm": 0.9556383642152451, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 11417 + }, + { + "epoch": 0.11418, + "grad_norm": 1.3551574310415802, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 11418 + }, + { + "epoch": 0.11419, + "grad_norm": 1.1236593786751847, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 11419 + }, + { + "epoch": 0.1142, + "grad_norm": 1.0487403125417163, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 11420 + }, + { + "epoch": 0.11421, + "grad_norm": 1.1149362265682679, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 11421 + }, + { + "epoch": 0.11422, + "grad_norm": 0.9316950413130666, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11422 + }, + { + "epoch": 0.11423, + "grad_norm": 1.112559363387894, + "learning_rate": 0.003, + "loss": 4.116, + "step": 11423 + }, + { + "epoch": 0.11424, + "grad_norm": 1.046947547930098, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 11424 + }, + { + "epoch": 0.11425, + "grad_norm": 1.025164897373827, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 11425 + }, + { + "epoch": 0.11426, + "grad_norm": 1.1000273644854461, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 11426 + }, + { + "epoch": 0.11427, + "grad_norm": 0.967175793628095, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11427 + }, + { + "epoch": 0.11428, + "grad_norm": 1.1723637978459762, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11428 + }, + { + "epoch": 0.11429, + "grad_norm": 1.0287812781648369, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 11429 + }, + { + "epoch": 0.1143, + "grad_norm": 1.3493177052407335, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 11430 + }, + { + "epoch": 0.11431, + "grad_norm": 1.0285149025991758, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 11431 + }, + { + "epoch": 0.11432, + "grad_norm": 1.3023078221513005, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 11432 + }, + { + "epoch": 0.11433, + "grad_norm": 1.2940193777377493, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 11433 + }, + { + "epoch": 0.11434, + "grad_norm": 1.1521329915424683, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 11434 + }, + { + "epoch": 0.11435, + "grad_norm": 1.09046750871652, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11435 + }, + { + "epoch": 0.11436, + "grad_norm": 1.239387425708267, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 11436 + }, + { + "epoch": 0.11437, + "grad_norm": 0.9839702313725753, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11437 + }, + { + "epoch": 0.11438, + "grad_norm": 1.0105437037979108, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11438 + }, + { + "epoch": 0.11439, + "grad_norm": 1.2805117104272348, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 11439 + }, + { + "epoch": 0.1144, + "grad_norm": 1.3017141223339386, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 11440 + }, + { + "epoch": 0.11441, + "grad_norm": 0.8997940091230753, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11441 + }, + { + "epoch": 0.11442, + "grad_norm": 0.9081136329240335, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11442 + }, + { + "epoch": 0.11443, + "grad_norm": 0.9592890689351932, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 11443 + }, + { + "epoch": 0.11444, + "grad_norm": 1.0751507227866715, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 11444 + }, + { + "epoch": 0.11445, + "grad_norm": 1.0817511739453243, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11445 + }, + { + "epoch": 0.11446, + "grad_norm": 0.985678283326643, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 11446 + }, + { + "epoch": 0.11447, + "grad_norm": 1.1095480987200779, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 11447 + }, + { + "epoch": 0.11448, + "grad_norm": 1.1999458551915834, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 11448 + }, + { + "epoch": 0.11449, + "grad_norm": 1.2755194245240447, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 11449 + }, + { + "epoch": 0.1145, + "grad_norm": 0.9554237876405631, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 11450 + }, + { + "epoch": 0.11451, + "grad_norm": 0.9921100619442162, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 11451 + }, + { + "epoch": 0.11452, + "grad_norm": 1.0983081877471728, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 11452 + }, + { + "epoch": 0.11453, + "grad_norm": 1.1593883506577896, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 11453 + }, + { + "epoch": 0.11454, + "grad_norm": 0.9269990739942203, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 11454 + }, + { + "epoch": 0.11455, + "grad_norm": 1.1980303015914466, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 11455 + }, + { + "epoch": 0.11456, + "grad_norm": 1.1480252892729894, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11456 + }, + { + "epoch": 0.11457, + "grad_norm": 1.1836346713452197, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11457 + }, + { + "epoch": 0.11458, + "grad_norm": 0.9351218802077039, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11458 + }, + { + "epoch": 0.11459, + "grad_norm": 1.1033123819477257, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11459 + }, + { + "epoch": 0.1146, + "grad_norm": 1.012445753388458, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 11460 + }, + { + "epoch": 0.11461, + "grad_norm": 1.3576147209988854, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11461 + }, + { + "epoch": 0.11462, + "grad_norm": 1.0883600335321508, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 11462 + }, + { + "epoch": 0.11463, + "grad_norm": 1.2657634520019636, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 11463 + }, + { + "epoch": 0.11464, + "grad_norm": 1.2311124665624869, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11464 + }, + { + "epoch": 0.11465, + "grad_norm": 1.1257407859233135, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 11465 + }, + { + "epoch": 0.11466, + "grad_norm": 0.944138764965183, + "learning_rate": 0.003, + "loss": 4.071, + "step": 11466 + }, + { + "epoch": 0.11467, + "grad_norm": 1.1813424893876345, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11467 + }, + { + "epoch": 0.11468, + "grad_norm": 1.0271819896606211, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 11468 + }, + { + "epoch": 0.11469, + "grad_norm": 1.0710135800874787, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11469 + }, + { + "epoch": 0.1147, + "grad_norm": 1.3291275440585504, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 11470 + }, + { + "epoch": 0.11471, + "grad_norm": 0.8813798372990844, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 11471 + }, + { + "epoch": 0.11472, + "grad_norm": 0.9608440586409432, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 11472 + }, + { + "epoch": 0.11473, + "grad_norm": 1.159326505594525, + "learning_rate": 0.003, + "loss": 4.079, + "step": 11473 + }, + { + "epoch": 0.11474, + "grad_norm": 0.982827792187837, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 11474 + }, + { + "epoch": 0.11475, + "grad_norm": 1.2625211478329064, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 11475 + }, + { + "epoch": 0.11476, + "grad_norm": 0.9599156828526338, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 11476 + }, + { + "epoch": 0.11477, + "grad_norm": 1.15290384723048, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 11477 + }, + { + "epoch": 0.11478, + "grad_norm": 1.1254972374254937, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 11478 + }, + { + "epoch": 0.11479, + "grad_norm": 1.039084825812637, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 11479 + }, + { + "epoch": 0.1148, + "grad_norm": 1.010880890786295, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 11480 + }, + { + "epoch": 0.11481, + "grad_norm": 1.125666168910612, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11481 + }, + { + "epoch": 0.11482, + "grad_norm": 1.069273471709598, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 11482 + }, + { + "epoch": 0.11483, + "grad_norm": 1.0996797520912263, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11483 + }, + { + "epoch": 0.11484, + "grad_norm": 1.053058869147153, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11484 + }, + { + "epoch": 0.11485, + "grad_norm": 0.95953616963765, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 11485 + }, + { + "epoch": 0.11486, + "grad_norm": 1.0673135991705158, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11486 + }, + { + "epoch": 0.11487, + "grad_norm": 1.2347727658118037, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 11487 + }, + { + "epoch": 0.11488, + "grad_norm": 1.1323197031207368, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 11488 + }, + { + "epoch": 0.11489, + "grad_norm": 1.233987085538438, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 11489 + }, + { + "epoch": 0.1149, + "grad_norm": 1.437730399768374, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 11490 + }, + { + "epoch": 0.11491, + "grad_norm": 1.0812904240647934, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 11491 + }, + { + "epoch": 0.11492, + "grad_norm": 1.2382203916879087, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 11492 + }, + { + "epoch": 0.11493, + "grad_norm": 1.1868789758176679, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 11493 + }, + { + "epoch": 0.11494, + "grad_norm": 1.0905374174299731, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 11494 + }, + { + "epoch": 0.11495, + "grad_norm": 1.0025408948887682, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 11495 + }, + { + "epoch": 0.11496, + "grad_norm": 1.059501013039697, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 11496 + }, + { + "epoch": 0.11497, + "grad_norm": 0.97928784748038, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 11497 + }, + { + "epoch": 0.11498, + "grad_norm": 0.9912519652552354, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 11498 + }, + { + "epoch": 0.11499, + "grad_norm": 1.2182826126289492, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11499 + }, + { + "epoch": 0.115, + "grad_norm": 1.0575900148374628, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 11500 + }, + { + "epoch": 0.11501, + "grad_norm": 1.042259946596932, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 11501 + }, + { + "epoch": 0.11502, + "grad_norm": 1.1491360742259606, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11502 + }, + { + "epoch": 0.11503, + "grad_norm": 1.1515986591053582, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11503 + }, + { + "epoch": 0.11504, + "grad_norm": 1.0082890004795189, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 11504 + }, + { + "epoch": 0.11505, + "grad_norm": 1.2398006653371065, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11505 + }, + { + "epoch": 0.11506, + "grad_norm": 0.9506533343856579, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 11506 + }, + { + "epoch": 0.11507, + "grad_norm": 1.2049014161424605, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 11507 + }, + { + "epoch": 0.11508, + "grad_norm": 1.016229736336884, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 11508 + }, + { + "epoch": 0.11509, + "grad_norm": 1.059800157746038, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 11509 + }, + { + "epoch": 0.1151, + "grad_norm": 1.2187976213839506, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11510 + }, + { + "epoch": 0.11511, + "grad_norm": 1.2526644591744729, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 11511 + }, + { + "epoch": 0.11512, + "grad_norm": 1.346175519230743, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 11512 + }, + { + "epoch": 0.11513, + "grad_norm": 1.0251209000081711, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 11513 + }, + { + "epoch": 0.11514, + "grad_norm": 1.1098866603814985, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 11514 + }, + { + "epoch": 0.11515, + "grad_norm": 1.1205216374867848, + "learning_rate": 0.003, + "loss": 4.118, + "step": 11515 + }, + { + "epoch": 0.11516, + "grad_norm": 1.133872234739474, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 11516 + }, + { + "epoch": 0.11517, + "grad_norm": 0.9546557729774939, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 11517 + }, + { + "epoch": 0.11518, + "grad_norm": 1.0000084540499257, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 11518 + }, + { + "epoch": 0.11519, + "grad_norm": 1.1378977666905323, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 11519 + }, + { + "epoch": 0.1152, + "grad_norm": 1.066034099269205, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11520 + }, + { + "epoch": 0.11521, + "grad_norm": 1.4441324390232206, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 11521 + }, + { + "epoch": 0.11522, + "grad_norm": 0.9289400412549517, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 11522 + }, + { + "epoch": 0.11523, + "grad_norm": 1.0067482412606141, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 11523 + }, + { + "epoch": 0.11524, + "grad_norm": 1.17887672911843, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11524 + }, + { + "epoch": 0.11525, + "grad_norm": 1.0209397605299606, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 11525 + }, + { + "epoch": 0.11526, + "grad_norm": 1.1822412291720787, + "learning_rate": 0.003, + "loss": 4.059, + "step": 11526 + }, + { + "epoch": 0.11527, + "grad_norm": 0.9907113033206698, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 11527 + }, + { + "epoch": 0.11528, + "grad_norm": 1.004037870295607, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 11528 + }, + { + "epoch": 0.11529, + "grad_norm": 1.27345356444765, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 11529 + }, + { + "epoch": 0.1153, + "grad_norm": 0.9091681488318872, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 11530 + }, + { + "epoch": 0.11531, + "grad_norm": 1.1485271799948789, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 11531 + }, + { + "epoch": 0.11532, + "grad_norm": 1.2286560991322542, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 11532 + }, + { + "epoch": 0.11533, + "grad_norm": 0.957779523859025, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 11533 + }, + { + "epoch": 0.11534, + "grad_norm": 1.0703209588477642, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 11534 + }, + { + "epoch": 0.11535, + "grad_norm": 1.4076865000833358, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 11535 + }, + { + "epoch": 0.11536, + "grad_norm": 0.8453178792943659, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11536 + }, + { + "epoch": 0.11537, + "grad_norm": 0.8850418916267287, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11537 + }, + { + "epoch": 0.11538, + "grad_norm": 1.0115039027362196, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 11538 + }, + { + "epoch": 0.11539, + "grad_norm": 1.296796036565217, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11539 + }, + { + "epoch": 0.1154, + "grad_norm": 0.9679605146867507, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 11540 + }, + { + "epoch": 0.11541, + "grad_norm": 1.1842908793146563, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11541 + }, + { + "epoch": 0.11542, + "grad_norm": 0.9950285761038729, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 11542 + }, + { + "epoch": 0.11543, + "grad_norm": 1.1300853554103478, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11543 + }, + { + "epoch": 0.11544, + "grad_norm": 1.1354787892059146, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 11544 + }, + { + "epoch": 0.11545, + "grad_norm": 1.0187525640548183, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 11545 + }, + { + "epoch": 0.11546, + "grad_norm": 1.3098113199196204, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 11546 + }, + { + "epoch": 0.11547, + "grad_norm": 1.0110391512040715, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11547 + }, + { + "epoch": 0.11548, + "grad_norm": 1.3003353713494008, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 11548 + }, + { + "epoch": 0.11549, + "grad_norm": 0.9518470943238035, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 11549 + }, + { + "epoch": 0.1155, + "grad_norm": 1.1331531114660882, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 11550 + }, + { + "epoch": 0.11551, + "grad_norm": 1.1341814792195717, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11551 + }, + { + "epoch": 0.11552, + "grad_norm": 1.1642435526875725, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 11552 + }, + { + "epoch": 0.11553, + "grad_norm": 1.1915747696382795, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 11553 + }, + { + "epoch": 0.11554, + "grad_norm": 0.9805179418386152, + "learning_rate": 0.003, + "loss": 4.069, + "step": 11554 + }, + { + "epoch": 0.11555, + "grad_norm": 1.1284658454331822, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11555 + }, + { + "epoch": 0.11556, + "grad_norm": 1.1631076906779456, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 11556 + }, + { + "epoch": 0.11557, + "grad_norm": 1.0466827929340412, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 11557 + }, + { + "epoch": 0.11558, + "grad_norm": 1.1745724748654662, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 11558 + }, + { + "epoch": 0.11559, + "grad_norm": 0.9459708236772346, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 11559 + }, + { + "epoch": 0.1156, + "grad_norm": 1.0021916398425177, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 11560 + }, + { + "epoch": 0.11561, + "grad_norm": 1.4207016748395007, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 11561 + }, + { + "epoch": 0.11562, + "grad_norm": 0.9699229174078287, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 11562 + }, + { + "epoch": 0.11563, + "grad_norm": 1.1311822024867861, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 11563 + }, + { + "epoch": 0.11564, + "grad_norm": 0.9679692818516854, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11564 + }, + { + "epoch": 0.11565, + "grad_norm": 1.1922138762146304, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 11565 + }, + { + "epoch": 0.11566, + "grad_norm": 1.237058073396377, + "learning_rate": 0.003, + "loss": 4.054, + "step": 11566 + }, + { + "epoch": 0.11567, + "grad_norm": 1.3887929548971565, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11567 + }, + { + "epoch": 0.11568, + "grad_norm": 0.958572461549701, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11568 + }, + { + "epoch": 0.11569, + "grad_norm": 1.0671355046838664, + "learning_rate": 0.003, + "loss": 4.082, + "step": 11569 + }, + { + "epoch": 0.1157, + "grad_norm": 1.2790464207002419, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 11570 + }, + { + "epoch": 0.11571, + "grad_norm": 0.9377850639923463, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 11571 + }, + { + "epoch": 0.11572, + "grad_norm": 0.8880366639899385, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11572 + }, + { + "epoch": 0.11573, + "grad_norm": 1.0787963181925324, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 11573 + }, + { + "epoch": 0.11574, + "grad_norm": 1.123504886496314, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11574 + }, + { + "epoch": 0.11575, + "grad_norm": 0.9183337157079674, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 11575 + }, + { + "epoch": 0.11576, + "grad_norm": 1.2491884156219748, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11576 + }, + { + "epoch": 0.11577, + "grad_norm": 1.1547263596011585, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 11577 + }, + { + "epoch": 0.11578, + "grad_norm": 1.1201197406092993, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 11578 + }, + { + "epoch": 0.11579, + "grad_norm": 1.2033338134892475, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 11579 + }, + { + "epoch": 0.1158, + "grad_norm": 1.0470671889660987, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 11580 + }, + { + "epoch": 0.11581, + "grad_norm": 1.2185434900378873, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 11581 + }, + { + "epoch": 0.11582, + "grad_norm": 1.041824094279636, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 11582 + }, + { + "epoch": 0.11583, + "grad_norm": 1.3025002104024346, + "learning_rate": 0.003, + "loss": 4.059, + "step": 11583 + }, + { + "epoch": 0.11584, + "grad_norm": 0.8944428556353944, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11584 + }, + { + "epoch": 0.11585, + "grad_norm": 0.9221289377596882, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 11585 + }, + { + "epoch": 0.11586, + "grad_norm": 1.1158151867038537, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 11586 + }, + { + "epoch": 0.11587, + "grad_norm": 1.0633801503841447, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11587 + }, + { + "epoch": 0.11588, + "grad_norm": 1.2175728246475466, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11588 + }, + { + "epoch": 0.11589, + "grad_norm": 0.9755229650012226, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 11589 + }, + { + "epoch": 0.1159, + "grad_norm": 1.2474696834282595, + "learning_rate": 0.003, + "loss": 4.064, + "step": 11590 + }, + { + "epoch": 0.11591, + "grad_norm": 0.9857271628762859, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 11591 + }, + { + "epoch": 0.11592, + "grad_norm": 1.1544073912447943, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 11592 + }, + { + "epoch": 0.11593, + "grad_norm": 1.1913649125562014, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 11593 + }, + { + "epoch": 0.11594, + "grad_norm": 1.2360990835362289, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 11594 + }, + { + "epoch": 0.11595, + "grad_norm": 1.1668662086330877, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 11595 + }, + { + "epoch": 0.11596, + "grad_norm": 1.2566056003469033, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 11596 + }, + { + "epoch": 0.11597, + "grad_norm": 1.0629373859203337, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 11597 + }, + { + "epoch": 0.11598, + "grad_norm": 1.1663361648055266, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 11598 + }, + { + "epoch": 0.11599, + "grad_norm": 1.2961480043008549, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 11599 + }, + { + "epoch": 0.116, + "grad_norm": 1.014250628737564, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11600 + }, + { + "epoch": 0.11601, + "grad_norm": 1.0304229179205031, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 11601 + }, + { + "epoch": 0.11602, + "grad_norm": 1.2069097235923103, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11602 + }, + { + "epoch": 0.11603, + "grad_norm": 1.0367709582145608, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 11603 + }, + { + "epoch": 0.11604, + "grad_norm": 1.1276340617378737, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 11604 + }, + { + "epoch": 0.11605, + "grad_norm": 0.9578098940648896, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11605 + }, + { + "epoch": 0.11606, + "grad_norm": 1.0784692357920973, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 11606 + }, + { + "epoch": 0.11607, + "grad_norm": 1.0911690675049772, + "learning_rate": 0.003, + "loss": 4.075, + "step": 11607 + }, + { + "epoch": 0.11608, + "grad_norm": 1.0784819343728043, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 11608 + }, + { + "epoch": 0.11609, + "grad_norm": 1.1460958124906055, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 11609 + }, + { + "epoch": 0.1161, + "grad_norm": 0.9943464723111869, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11610 + }, + { + "epoch": 0.11611, + "grad_norm": 1.098850299349954, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11611 + }, + { + "epoch": 0.11612, + "grad_norm": 1.1362782642396552, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 11612 + }, + { + "epoch": 0.11613, + "grad_norm": 1.1387512980756966, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 11613 + }, + { + "epoch": 0.11614, + "grad_norm": 1.345923900697499, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 11614 + }, + { + "epoch": 0.11615, + "grad_norm": 1.130939584436006, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 11615 + }, + { + "epoch": 0.11616, + "grad_norm": 0.886731255863067, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 11616 + }, + { + "epoch": 0.11617, + "grad_norm": 1.1544801702721095, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 11617 + }, + { + "epoch": 0.11618, + "grad_norm": 1.1811782827774127, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 11618 + }, + { + "epoch": 0.11619, + "grad_norm": 1.0060331338988424, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 11619 + }, + { + "epoch": 0.1162, + "grad_norm": 1.1521508346964646, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 11620 + }, + { + "epoch": 0.11621, + "grad_norm": 1.0257932213801997, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11621 + }, + { + "epoch": 0.11622, + "grad_norm": 1.0642025893798226, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 11622 + }, + { + "epoch": 0.11623, + "grad_norm": 1.0541204723690525, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 11623 + }, + { + "epoch": 0.11624, + "grad_norm": 0.963137136298594, + "learning_rate": 0.003, + "loss": 4.069, + "step": 11624 + }, + { + "epoch": 0.11625, + "grad_norm": 1.2253091846333735, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 11625 + }, + { + "epoch": 0.11626, + "grad_norm": 1.0179136549959853, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11626 + }, + { + "epoch": 0.11627, + "grad_norm": 1.284748849879381, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 11627 + }, + { + "epoch": 0.11628, + "grad_norm": 1.207376669095626, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 11628 + }, + { + "epoch": 0.11629, + "grad_norm": 1.201903728509045, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 11629 + }, + { + "epoch": 0.1163, + "grad_norm": 1.1275076903728707, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 11630 + }, + { + "epoch": 0.11631, + "grad_norm": 0.8270761377773259, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 11631 + }, + { + "epoch": 0.11632, + "grad_norm": 1.0228543931299996, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 11632 + }, + { + "epoch": 0.11633, + "grad_norm": 1.1302523592518674, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 11633 + }, + { + "epoch": 0.11634, + "grad_norm": 0.8638865702028005, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 11634 + }, + { + "epoch": 0.11635, + "grad_norm": 1.0027464596812203, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 11635 + }, + { + "epoch": 0.11636, + "grad_norm": 1.3485058979634321, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 11636 + }, + { + "epoch": 0.11637, + "grad_norm": 1.154943472996014, + "learning_rate": 0.003, + "loss": 4.08, + "step": 11637 + }, + { + "epoch": 0.11638, + "grad_norm": 1.0927699715554624, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 11638 + }, + { + "epoch": 0.11639, + "grad_norm": 0.9883426317196221, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 11639 + }, + { + "epoch": 0.1164, + "grad_norm": 1.1813503322922911, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 11640 + }, + { + "epoch": 0.11641, + "grad_norm": 0.9833000059546451, + "learning_rate": 0.003, + "loss": 4.083, + "step": 11641 + }, + { + "epoch": 0.11642, + "grad_norm": 1.017905346925556, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 11642 + }, + { + "epoch": 0.11643, + "grad_norm": 1.274217063915014, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 11643 + }, + { + "epoch": 0.11644, + "grad_norm": 1.0301408484730104, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11644 + }, + { + "epoch": 0.11645, + "grad_norm": 1.009275001586971, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 11645 + }, + { + "epoch": 0.11646, + "grad_norm": 1.2709057781981588, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 11646 + }, + { + "epoch": 0.11647, + "grad_norm": 1.1255501598073867, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 11647 + }, + { + "epoch": 0.11648, + "grad_norm": 0.984766142304668, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 11648 + }, + { + "epoch": 0.11649, + "grad_norm": 1.0563767234375925, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 11649 + }, + { + "epoch": 0.1165, + "grad_norm": 1.1606860639127619, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 11650 + }, + { + "epoch": 0.11651, + "grad_norm": 1.1015639702221316, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11651 + }, + { + "epoch": 0.11652, + "grad_norm": 1.143719872576751, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11652 + }, + { + "epoch": 0.11653, + "grad_norm": 1.2263249319837026, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 11653 + }, + { + "epoch": 0.11654, + "grad_norm": 0.8919799265058325, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 11654 + }, + { + "epoch": 0.11655, + "grad_norm": 1.1735764182230106, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 11655 + }, + { + "epoch": 0.11656, + "grad_norm": 1.033198482124397, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11656 + }, + { + "epoch": 0.11657, + "grad_norm": 1.3681255518075837, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 11657 + }, + { + "epoch": 0.11658, + "grad_norm": 1.2738790837565772, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11658 + }, + { + "epoch": 0.11659, + "grad_norm": 1.3607779431837075, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 11659 + }, + { + "epoch": 0.1166, + "grad_norm": 0.986669792330658, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 11660 + }, + { + "epoch": 0.11661, + "grad_norm": 1.152518378500052, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11661 + }, + { + "epoch": 0.11662, + "grad_norm": 1.1196506107063688, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 11662 + }, + { + "epoch": 0.11663, + "grad_norm": 0.995810438761169, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 11663 + }, + { + "epoch": 0.11664, + "grad_norm": 1.1216949623356607, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 11664 + }, + { + "epoch": 0.11665, + "grad_norm": 1.0930625371595246, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 11665 + }, + { + "epoch": 0.11666, + "grad_norm": 1.1430523888668103, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 11666 + }, + { + "epoch": 0.11667, + "grad_norm": 1.4096528741555352, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 11667 + }, + { + "epoch": 0.11668, + "grad_norm": 0.8572070174035606, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11668 + }, + { + "epoch": 0.11669, + "grad_norm": 0.9520455561841698, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 11669 + }, + { + "epoch": 0.1167, + "grad_norm": 1.1344283234440606, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11670 + }, + { + "epoch": 0.11671, + "grad_norm": 1.1367286001531283, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 11671 + }, + { + "epoch": 0.11672, + "grad_norm": 1.1019504330512884, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 11672 + }, + { + "epoch": 0.11673, + "grad_norm": 1.2257422864371508, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11673 + }, + { + "epoch": 0.11674, + "grad_norm": 1.0657527751866165, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 11674 + }, + { + "epoch": 0.11675, + "grad_norm": 1.1985652793128885, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 11675 + }, + { + "epoch": 0.11676, + "grad_norm": 0.9332198870740341, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 11676 + }, + { + "epoch": 0.11677, + "grad_norm": 1.0545279590864574, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 11677 + }, + { + "epoch": 0.11678, + "grad_norm": 1.3340048562889086, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 11678 + }, + { + "epoch": 0.11679, + "grad_norm": 0.9379675744284625, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 11679 + }, + { + "epoch": 0.1168, + "grad_norm": 1.1141130817934959, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 11680 + }, + { + "epoch": 0.11681, + "grad_norm": 1.177277877794882, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 11681 + }, + { + "epoch": 0.11682, + "grad_norm": 1.1277834165780665, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11682 + }, + { + "epoch": 0.11683, + "grad_norm": 1.281372903846131, + "learning_rate": 0.003, + "loss": 4.109, + "step": 11683 + }, + { + "epoch": 0.11684, + "grad_norm": 0.9181486510427459, + "learning_rate": 0.003, + "loss": 4.081, + "step": 11684 + }, + { + "epoch": 0.11685, + "grad_norm": 0.9626362886189451, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11685 + }, + { + "epoch": 0.11686, + "grad_norm": 1.0406179327979335, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11686 + }, + { + "epoch": 0.11687, + "grad_norm": 1.1605950057136238, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 11687 + }, + { + "epoch": 0.11688, + "grad_norm": 1.1183597832884749, + "learning_rate": 0.003, + "loss": 4.097, + "step": 11688 + }, + { + "epoch": 0.11689, + "grad_norm": 1.3177810263811713, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 11689 + }, + { + "epoch": 0.1169, + "grad_norm": 1.0339702155942279, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11690 + }, + { + "epoch": 0.11691, + "grad_norm": 1.2249081132818636, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 11691 + }, + { + "epoch": 0.11692, + "grad_norm": 1.266927910285917, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11692 + }, + { + "epoch": 0.11693, + "grad_norm": 1.0213845058392041, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 11693 + }, + { + "epoch": 0.11694, + "grad_norm": 1.0362441103808235, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11694 + }, + { + "epoch": 0.11695, + "grad_norm": 1.100736599721524, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 11695 + }, + { + "epoch": 0.11696, + "grad_norm": 1.0627450412294686, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 11696 + }, + { + "epoch": 0.11697, + "grad_norm": 1.2418770546939046, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11697 + }, + { + "epoch": 0.11698, + "grad_norm": 0.8793593011265435, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 11698 + }, + { + "epoch": 0.11699, + "grad_norm": 0.9356580319854434, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 11699 + }, + { + "epoch": 0.117, + "grad_norm": 1.0565849927073654, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 11700 + }, + { + "epoch": 0.11701, + "grad_norm": 1.0321570661745199, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11701 + }, + { + "epoch": 0.11702, + "grad_norm": 1.1302279305254166, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11702 + }, + { + "epoch": 0.11703, + "grad_norm": 1.1276832996289374, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 11703 + }, + { + "epoch": 0.11704, + "grad_norm": 1.643091586238816, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 11704 + }, + { + "epoch": 0.11705, + "grad_norm": 0.8483221837791483, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11705 + }, + { + "epoch": 0.11706, + "grad_norm": 1.118465992742277, + "learning_rate": 0.003, + "loss": 4.101, + "step": 11706 + }, + { + "epoch": 0.11707, + "grad_norm": 1.0063144729463553, + "learning_rate": 0.003, + "loss": 4.086, + "step": 11707 + }, + { + "epoch": 0.11708, + "grad_norm": 1.0977058727516777, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 11708 + }, + { + "epoch": 0.11709, + "grad_norm": 1.3282663240246637, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 11709 + }, + { + "epoch": 0.1171, + "grad_norm": 0.948967010561446, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 11710 + }, + { + "epoch": 0.11711, + "grad_norm": 1.1042229472219267, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 11711 + }, + { + "epoch": 0.11712, + "grad_norm": 1.1451972308255882, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 11712 + }, + { + "epoch": 0.11713, + "grad_norm": 1.1229877422007808, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 11713 + }, + { + "epoch": 0.11714, + "grad_norm": 1.3025407860311622, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 11714 + }, + { + "epoch": 0.11715, + "grad_norm": 0.9538833016975586, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 11715 + }, + { + "epoch": 0.11716, + "grad_norm": 1.1855847461520062, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 11716 + }, + { + "epoch": 0.11717, + "grad_norm": 1.1245699404440197, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11717 + }, + { + "epoch": 0.11718, + "grad_norm": 1.1777591935526606, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11718 + }, + { + "epoch": 0.11719, + "grad_norm": 1.1755063251134579, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 11719 + }, + { + "epoch": 0.1172, + "grad_norm": 1.1096620840332625, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 11720 + }, + { + "epoch": 0.11721, + "grad_norm": 1.0515005669697468, + "learning_rate": 0.003, + "loss": 4.073, + "step": 11721 + }, + { + "epoch": 0.11722, + "grad_norm": 1.1629975308450062, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 11722 + }, + { + "epoch": 0.11723, + "grad_norm": 0.975517625425596, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 11723 + }, + { + "epoch": 0.11724, + "grad_norm": 1.1223667286703718, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 11724 + }, + { + "epoch": 0.11725, + "grad_norm": 0.9892581009311286, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 11725 + }, + { + "epoch": 0.11726, + "grad_norm": 1.0867359120534805, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11726 + }, + { + "epoch": 0.11727, + "grad_norm": 1.0824253250195415, + "learning_rate": 0.003, + "loss": 4.101, + "step": 11727 + }, + { + "epoch": 0.11728, + "grad_norm": 1.1775735354805041, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 11728 + }, + { + "epoch": 0.11729, + "grad_norm": 1.218274123653387, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 11729 + }, + { + "epoch": 0.1173, + "grad_norm": 1.1566066956016399, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 11730 + }, + { + "epoch": 0.11731, + "grad_norm": 1.088035629745312, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 11731 + }, + { + "epoch": 0.11732, + "grad_norm": 1.165765137910104, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 11732 + }, + { + "epoch": 0.11733, + "grad_norm": 1.223175992941106, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11733 + }, + { + "epoch": 0.11734, + "grad_norm": 0.9196427956388414, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 11734 + }, + { + "epoch": 0.11735, + "grad_norm": 0.8899377969449798, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 11735 + }, + { + "epoch": 0.11736, + "grad_norm": 0.8856112456450769, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11736 + }, + { + "epoch": 0.11737, + "grad_norm": 0.8304790484767605, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 11737 + }, + { + "epoch": 0.11738, + "grad_norm": 0.8856599944783037, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11738 + }, + { + "epoch": 0.11739, + "grad_norm": 1.2201352791875548, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 11739 + }, + { + "epoch": 0.1174, + "grad_norm": 1.1462293051204187, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 11740 + }, + { + "epoch": 0.11741, + "grad_norm": 1.1459578959083037, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 11741 + }, + { + "epoch": 0.11742, + "grad_norm": 1.0533399880453962, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 11742 + }, + { + "epoch": 0.11743, + "grad_norm": 1.121969173337314, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11743 + }, + { + "epoch": 0.11744, + "grad_norm": 1.4689998599569376, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11744 + }, + { + "epoch": 0.11745, + "grad_norm": 1.3274205401294934, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 11745 + }, + { + "epoch": 0.11746, + "grad_norm": 0.8918388851683189, + "learning_rate": 0.003, + "loss": 4.049, + "step": 11746 + }, + { + "epoch": 0.11747, + "grad_norm": 1.0022823662035623, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 11747 + }, + { + "epoch": 0.11748, + "grad_norm": 1.2740630639286075, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 11748 + }, + { + "epoch": 0.11749, + "grad_norm": 1.1144566017947264, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 11749 + }, + { + "epoch": 0.1175, + "grad_norm": 1.2794518402465218, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 11750 + }, + { + "epoch": 0.11751, + "grad_norm": 0.9239685969957789, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 11751 + }, + { + "epoch": 0.11752, + "grad_norm": 0.9848298008138613, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 11752 + }, + { + "epoch": 0.11753, + "grad_norm": 1.2900820404962074, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 11753 + }, + { + "epoch": 0.11754, + "grad_norm": 1.06059796209915, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 11754 + }, + { + "epoch": 0.11755, + "grad_norm": 1.060399615268824, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 11755 + }, + { + "epoch": 0.11756, + "grad_norm": 1.2775892797667003, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 11756 + }, + { + "epoch": 0.11757, + "grad_norm": 0.972795804791523, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 11757 + }, + { + "epoch": 0.11758, + "grad_norm": 1.0490010562233407, + "learning_rate": 0.003, + "loss": 4.102, + "step": 11758 + }, + { + "epoch": 0.11759, + "grad_norm": 1.2564337969810733, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 11759 + }, + { + "epoch": 0.1176, + "grad_norm": 0.9491679826420742, + "learning_rate": 0.003, + "loss": 4.077, + "step": 11760 + }, + { + "epoch": 0.11761, + "grad_norm": 1.1248877574419467, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 11761 + }, + { + "epoch": 0.11762, + "grad_norm": 1.0079745325917062, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11762 + }, + { + "epoch": 0.11763, + "grad_norm": 1.3124075277396952, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 11763 + }, + { + "epoch": 0.11764, + "grad_norm": 0.9367086914805486, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 11764 + }, + { + "epoch": 0.11765, + "grad_norm": 1.1444616985411946, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 11765 + }, + { + "epoch": 0.11766, + "grad_norm": 1.3212645040441942, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11766 + }, + { + "epoch": 0.11767, + "grad_norm": 1.1244238476096746, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 11767 + }, + { + "epoch": 0.11768, + "grad_norm": 1.1773234175338592, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11768 + }, + { + "epoch": 0.11769, + "grad_norm": 0.9581425424892575, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 11769 + }, + { + "epoch": 0.1177, + "grad_norm": 1.0621317193895994, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 11770 + }, + { + "epoch": 0.11771, + "grad_norm": 1.0546416995775967, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 11771 + }, + { + "epoch": 0.11772, + "grad_norm": 1.4850667429157458, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 11772 + }, + { + "epoch": 0.11773, + "grad_norm": 1.1016435020141717, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 11773 + }, + { + "epoch": 0.11774, + "grad_norm": 1.356521945100985, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 11774 + }, + { + "epoch": 0.11775, + "grad_norm": 1.0785970233598485, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 11775 + }, + { + "epoch": 0.11776, + "grad_norm": 1.0520217338363091, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 11776 + }, + { + "epoch": 0.11777, + "grad_norm": 1.1807726280590305, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 11777 + }, + { + "epoch": 0.11778, + "grad_norm": 0.9514944018351177, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 11778 + }, + { + "epoch": 0.11779, + "grad_norm": 1.2589173671064429, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 11779 + }, + { + "epoch": 0.1178, + "grad_norm": 0.9943740001334537, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 11780 + }, + { + "epoch": 0.11781, + "grad_norm": 1.1756358853971083, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 11781 + }, + { + "epoch": 0.11782, + "grad_norm": 1.2088112431840263, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 11782 + }, + { + "epoch": 0.11783, + "grad_norm": 1.3874050801177034, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 11783 + }, + { + "epoch": 0.11784, + "grad_norm": 0.9053806520882297, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 11784 + }, + { + "epoch": 0.11785, + "grad_norm": 0.9013846202308751, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11785 + }, + { + "epoch": 0.11786, + "grad_norm": 0.9788339236866542, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11786 + }, + { + "epoch": 0.11787, + "grad_norm": 1.0873653791790059, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 11787 + }, + { + "epoch": 0.11788, + "grad_norm": 1.1421214463569709, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 11788 + }, + { + "epoch": 0.11789, + "grad_norm": 1.1779963108684894, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 11789 + }, + { + "epoch": 0.1179, + "grad_norm": 0.982025514320481, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 11790 + }, + { + "epoch": 0.11791, + "grad_norm": 1.0313215188232239, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 11791 + }, + { + "epoch": 0.11792, + "grad_norm": 1.0818733705067438, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 11792 + }, + { + "epoch": 0.11793, + "grad_norm": 1.0254561930890065, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 11793 + }, + { + "epoch": 0.11794, + "grad_norm": 1.0942218467450253, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 11794 + }, + { + "epoch": 0.11795, + "grad_norm": 1.3739924318419376, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 11795 + }, + { + "epoch": 0.11796, + "grad_norm": 1.0386170689594838, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 11796 + }, + { + "epoch": 0.11797, + "grad_norm": 1.468240521388495, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11797 + }, + { + "epoch": 0.11798, + "grad_norm": 0.8881530246130017, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 11798 + }, + { + "epoch": 0.11799, + "grad_norm": 0.8667721143037403, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 11799 + }, + { + "epoch": 0.118, + "grad_norm": 0.9828151445089266, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 11800 + }, + { + "epoch": 0.11801, + "grad_norm": 1.0962943506473461, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11801 + }, + { + "epoch": 0.11802, + "grad_norm": 1.032329474607015, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 11802 + }, + { + "epoch": 0.11803, + "grad_norm": 1.0895359905818427, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 11803 + }, + { + "epoch": 0.11804, + "grad_norm": 1.2567454842956758, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 11804 + }, + { + "epoch": 0.11805, + "grad_norm": 1.1423825854362604, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 11805 + }, + { + "epoch": 0.11806, + "grad_norm": 1.289721276916898, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 11806 + }, + { + "epoch": 0.11807, + "grad_norm": 1.0671857466036752, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 11807 + }, + { + "epoch": 0.11808, + "grad_norm": 1.343505970047031, + "learning_rate": 0.003, + "loss": 4.065, + "step": 11808 + }, + { + "epoch": 0.11809, + "grad_norm": 1.139936318061021, + "learning_rate": 0.003, + "loss": 4.093, + "step": 11809 + }, + { + "epoch": 0.1181, + "grad_norm": 1.244710701696588, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 11810 + }, + { + "epoch": 0.11811, + "grad_norm": 1.0503231155416128, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 11811 + }, + { + "epoch": 0.11812, + "grad_norm": 1.0926414145035097, + "learning_rate": 0.003, + "loss": 4.081, + "step": 11812 + }, + { + "epoch": 0.11813, + "grad_norm": 1.023841354122939, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 11813 + }, + { + "epoch": 0.11814, + "grad_norm": 1.1156410068020937, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 11814 + }, + { + "epoch": 0.11815, + "grad_norm": 1.1491708173222321, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 11815 + }, + { + "epoch": 0.11816, + "grad_norm": 1.284086474032784, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 11816 + }, + { + "epoch": 0.11817, + "grad_norm": 1.0748777372173741, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11817 + }, + { + "epoch": 0.11818, + "grad_norm": 1.029728756967552, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 11818 + }, + { + "epoch": 0.11819, + "grad_norm": 1.1223699498022326, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11819 + }, + { + "epoch": 0.1182, + "grad_norm": 1.2151216877560418, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 11820 + }, + { + "epoch": 0.11821, + "grad_norm": 0.9447500324212605, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11821 + }, + { + "epoch": 0.11822, + "grad_norm": 1.10373474981526, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 11822 + }, + { + "epoch": 0.11823, + "grad_norm": 1.1021794004120868, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 11823 + }, + { + "epoch": 0.11824, + "grad_norm": 1.1387450997635156, + "learning_rate": 0.003, + "loss": 4.068, + "step": 11824 + }, + { + "epoch": 0.11825, + "grad_norm": 1.2274566209269968, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 11825 + }, + { + "epoch": 0.11826, + "grad_norm": 0.9287450532497696, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 11826 + }, + { + "epoch": 0.11827, + "grad_norm": 1.1155557049195648, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 11827 + }, + { + "epoch": 0.11828, + "grad_norm": 1.1489952053902721, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 11828 + }, + { + "epoch": 0.11829, + "grad_norm": 1.0876359379671903, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 11829 + }, + { + "epoch": 0.1183, + "grad_norm": 1.4525289352556954, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11830 + }, + { + "epoch": 0.11831, + "grad_norm": 1.0619833667413887, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 11831 + }, + { + "epoch": 0.11832, + "grad_norm": 1.3638115504838992, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11832 + }, + { + "epoch": 0.11833, + "grad_norm": 1.1812529631150601, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 11833 + }, + { + "epoch": 0.11834, + "grad_norm": 1.375271052465075, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 11834 + }, + { + "epoch": 0.11835, + "grad_norm": 1.1164555597617594, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 11835 + }, + { + "epoch": 0.11836, + "grad_norm": 1.009193328635914, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11836 + }, + { + "epoch": 0.11837, + "grad_norm": 1.1768578245175483, + "learning_rate": 0.003, + "loss": 4.078, + "step": 11837 + }, + { + "epoch": 0.11838, + "grad_norm": 1.0258693957941911, + "learning_rate": 0.003, + "loss": 4.057, + "step": 11838 + }, + { + "epoch": 0.11839, + "grad_norm": 1.3065155218168176, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 11839 + }, + { + "epoch": 0.1184, + "grad_norm": 0.8799608760122808, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11840 + }, + { + "epoch": 0.11841, + "grad_norm": 0.98901493208036, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 11841 + }, + { + "epoch": 0.11842, + "grad_norm": 1.25068801265814, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11842 + }, + { + "epoch": 0.11843, + "grad_norm": 0.8406258499606613, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 11843 + }, + { + "epoch": 0.11844, + "grad_norm": 0.8570725315156429, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 11844 + }, + { + "epoch": 0.11845, + "grad_norm": 1.0311542246289087, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 11845 + }, + { + "epoch": 0.11846, + "grad_norm": 1.3311754584449043, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11846 + }, + { + "epoch": 0.11847, + "grad_norm": 1.0010647914601536, + "learning_rate": 0.003, + "loss": 4.082, + "step": 11847 + }, + { + "epoch": 0.11848, + "grad_norm": 1.485318441648701, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 11848 + }, + { + "epoch": 0.11849, + "grad_norm": 1.1214138373683373, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 11849 + }, + { + "epoch": 0.1185, + "grad_norm": 1.0196511782777165, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11850 + }, + { + "epoch": 0.11851, + "grad_norm": 1.1687837566468635, + "learning_rate": 0.003, + "loss": 4.069, + "step": 11851 + }, + { + "epoch": 0.11852, + "grad_norm": 1.0554697586410853, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 11852 + }, + { + "epoch": 0.11853, + "grad_norm": 1.0916125410803077, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 11853 + }, + { + "epoch": 0.11854, + "grad_norm": 1.0562018909120507, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 11854 + }, + { + "epoch": 0.11855, + "grad_norm": 1.150141586864, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 11855 + }, + { + "epoch": 0.11856, + "grad_norm": 1.133521226026218, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 11856 + }, + { + "epoch": 0.11857, + "grad_norm": 0.8634651948359986, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 11857 + }, + { + "epoch": 0.11858, + "grad_norm": 0.9642019874347069, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 11858 + }, + { + "epoch": 0.11859, + "grad_norm": 1.1319671295432074, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 11859 + }, + { + "epoch": 0.1186, + "grad_norm": 1.5953846007576176, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 11860 + }, + { + "epoch": 0.11861, + "grad_norm": 1.024932034441909, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11861 + }, + { + "epoch": 0.11862, + "grad_norm": 1.2986812785569133, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 11862 + }, + { + "epoch": 0.11863, + "grad_norm": 1.0985985673307168, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 11863 + }, + { + "epoch": 0.11864, + "grad_norm": 1.310174333179123, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 11864 + }, + { + "epoch": 0.11865, + "grad_norm": 0.899614133293594, + "learning_rate": 0.003, + "loss": 4.072, + "step": 11865 + }, + { + "epoch": 0.11866, + "grad_norm": 1.2292245748747577, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 11866 + }, + { + "epoch": 0.11867, + "grad_norm": 1.1814766245058643, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 11867 + }, + { + "epoch": 0.11868, + "grad_norm": 1.1547264563679283, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11868 + }, + { + "epoch": 0.11869, + "grad_norm": 1.135468755245357, + "learning_rate": 0.003, + "loss": 4.087, + "step": 11869 + }, + { + "epoch": 0.1187, + "grad_norm": 1.1283808115880198, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 11870 + }, + { + "epoch": 0.11871, + "grad_norm": 1.1651147439593341, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 11871 + }, + { + "epoch": 0.11872, + "grad_norm": 0.952732737427695, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 11872 + }, + { + "epoch": 0.11873, + "grad_norm": 1.0044383418061162, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11873 + }, + { + "epoch": 0.11874, + "grad_norm": 1.2403290296527454, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11874 + }, + { + "epoch": 0.11875, + "grad_norm": 1.051254612260916, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11875 + }, + { + "epoch": 0.11876, + "grad_norm": 1.0815830505822788, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11876 + }, + { + "epoch": 0.11877, + "grad_norm": 1.0620445404429997, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 11877 + }, + { + "epoch": 0.11878, + "grad_norm": 1.389493017030221, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 11878 + }, + { + "epoch": 0.11879, + "grad_norm": 1.085364632408075, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11879 + }, + { + "epoch": 0.1188, + "grad_norm": 1.0560306403664153, + "learning_rate": 0.003, + "loss": 4.076, + "step": 11880 + }, + { + "epoch": 0.11881, + "grad_norm": 1.2325914466530217, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11881 + }, + { + "epoch": 0.11882, + "grad_norm": 0.9430171665821426, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 11882 + }, + { + "epoch": 0.11883, + "grad_norm": 0.9922866907169047, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 11883 + }, + { + "epoch": 0.11884, + "grad_norm": 1.1876342527608525, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 11884 + }, + { + "epoch": 0.11885, + "grad_norm": 1.127731596475419, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11885 + }, + { + "epoch": 0.11886, + "grad_norm": 1.219941205295011, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 11886 + }, + { + "epoch": 0.11887, + "grad_norm": 1.1402853107785413, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 11887 + }, + { + "epoch": 0.11888, + "grad_norm": 1.2308931402462295, + "learning_rate": 0.003, + "loss": 4.109, + "step": 11888 + }, + { + "epoch": 0.11889, + "grad_norm": 1.0604024665201628, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 11889 + }, + { + "epoch": 0.1189, + "grad_norm": 1.221421625109362, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 11890 + }, + { + "epoch": 0.11891, + "grad_norm": 1.073227547779595, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 11891 + }, + { + "epoch": 0.11892, + "grad_norm": 1.091635809784193, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11892 + }, + { + "epoch": 0.11893, + "grad_norm": 1.0494296892537798, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11893 + }, + { + "epoch": 0.11894, + "grad_norm": 1.0319729573369925, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 11894 + }, + { + "epoch": 0.11895, + "grad_norm": 1.0088727056452342, + "learning_rate": 0.003, + "loss": 4.064, + "step": 11895 + }, + { + "epoch": 0.11896, + "grad_norm": 1.2211882582437528, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 11896 + }, + { + "epoch": 0.11897, + "grad_norm": 1.22459237408397, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 11897 + }, + { + "epoch": 0.11898, + "grad_norm": 1.2964883702512249, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 11898 + }, + { + "epoch": 0.11899, + "grad_norm": 1.1006710611829373, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 11899 + }, + { + "epoch": 0.119, + "grad_norm": 1.1283526124007004, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 11900 + }, + { + "epoch": 0.11901, + "grad_norm": 1.0477828176444812, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 11901 + }, + { + "epoch": 0.11902, + "grad_norm": 1.1105312518883845, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11902 + }, + { + "epoch": 0.11903, + "grad_norm": 0.884943929413305, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 11903 + }, + { + "epoch": 0.11904, + "grad_norm": 0.9284368125352005, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 11904 + }, + { + "epoch": 0.11905, + "grad_norm": 1.0589265482717447, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 11905 + }, + { + "epoch": 0.11906, + "grad_norm": 1.1092228565605042, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 11906 + }, + { + "epoch": 0.11907, + "grad_norm": 1.1941986399865445, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 11907 + }, + { + "epoch": 0.11908, + "grad_norm": 1.046439168125344, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 11908 + }, + { + "epoch": 0.11909, + "grad_norm": 1.3284280153633887, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 11909 + }, + { + "epoch": 0.1191, + "grad_norm": 1.1162343111004016, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 11910 + }, + { + "epoch": 0.11911, + "grad_norm": 1.2286716614076243, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11911 + }, + { + "epoch": 0.11912, + "grad_norm": 1.0845810445242683, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 11912 + }, + { + "epoch": 0.11913, + "grad_norm": 1.1757700274556278, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 11913 + }, + { + "epoch": 0.11914, + "grad_norm": 1.0720002142634966, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 11914 + }, + { + "epoch": 0.11915, + "grad_norm": 0.9585568712803956, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 11915 + }, + { + "epoch": 0.11916, + "grad_norm": 1.118410361514395, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 11916 + }, + { + "epoch": 0.11917, + "grad_norm": 1.1439516461494486, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 11917 + }, + { + "epoch": 0.11918, + "grad_norm": 0.8774537417656473, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11918 + }, + { + "epoch": 0.11919, + "grad_norm": 0.9912767987498607, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 11919 + }, + { + "epoch": 0.1192, + "grad_norm": 1.402160777194763, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 11920 + }, + { + "epoch": 0.11921, + "grad_norm": 1.068468107904343, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 11921 + }, + { + "epoch": 0.11922, + "grad_norm": 1.3659602551264145, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 11922 + }, + { + "epoch": 0.11923, + "grad_norm": 1.373546498881002, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11923 + }, + { + "epoch": 0.11924, + "grad_norm": 1.2060840228601442, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 11924 + }, + { + "epoch": 0.11925, + "grad_norm": 1.127806727390189, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 11925 + }, + { + "epoch": 0.11926, + "grad_norm": 1.0365064543575828, + "learning_rate": 0.003, + "loss": 4.08, + "step": 11926 + }, + { + "epoch": 0.11927, + "grad_norm": 1.1116861558348419, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 11927 + }, + { + "epoch": 0.11928, + "grad_norm": 0.8824401374304045, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 11928 + }, + { + "epoch": 0.11929, + "grad_norm": 0.91805897099131, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 11929 + }, + { + "epoch": 0.1193, + "grad_norm": 1.1680131795885302, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 11930 + }, + { + "epoch": 0.11931, + "grad_norm": 1.1886823191495501, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 11931 + }, + { + "epoch": 0.11932, + "grad_norm": 1.1887751402826594, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11932 + }, + { + "epoch": 0.11933, + "grad_norm": 1.1179307530288558, + "learning_rate": 0.003, + "loss": 4.058, + "step": 11933 + }, + { + "epoch": 0.11934, + "grad_norm": 1.0489444539389505, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 11934 + }, + { + "epoch": 0.11935, + "grad_norm": 1.237415215488098, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 11935 + }, + { + "epoch": 0.11936, + "grad_norm": 0.9957532742252633, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 11936 + }, + { + "epoch": 0.11937, + "grad_norm": 1.1891332584840746, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 11937 + }, + { + "epoch": 0.11938, + "grad_norm": 1.0463104442432303, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 11938 + }, + { + "epoch": 0.11939, + "grad_norm": 1.2373487906517007, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 11939 + }, + { + "epoch": 0.1194, + "grad_norm": 0.9982842826684045, + "learning_rate": 0.003, + "loss": 4.09, + "step": 11940 + }, + { + "epoch": 0.11941, + "grad_norm": 1.4640902331855827, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 11941 + }, + { + "epoch": 0.11942, + "grad_norm": 1.0480877070110093, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 11942 + }, + { + "epoch": 0.11943, + "grad_norm": 1.2282928382729854, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 11943 + }, + { + "epoch": 0.11944, + "grad_norm": 1.2343243580325551, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 11944 + }, + { + "epoch": 0.11945, + "grad_norm": 1.2027915615430216, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 11945 + }, + { + "epoch": 0.11946, + "grad_norm": 0.8917739824369894, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 11946 + }, + { + "epoch": 0.11947, + "grad_norm": 1.0447613041945965, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 11947 + }, + { + "epoch": 0.11948, + "grad_norm": 1.2178185052397315, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 11948 + }, + { + "epoch": 0.11949, + "grad_norm": 0.9378135352565747, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 11949 + }, + { + "epoch": 0.1195, + "grad_norm": 1.092216526784142, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 11950 + }, + { + "epoch": 0.11951, + "grad_norm": 1.100524960745319, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 11951 + }, + { + "epoch": 0.11952, + "grad_norm": 1.1444651267711, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 11952 + }, + { + "epoch": 0.11953, + "grad_norm": 1.203498741669757, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 11953 + }, + { + "epoch": 0.11954, + "grad_norm": 1.1427459985323243, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 11954 + }, + { + "epoch": 0.11955, + "grad_norm": 1.1361070526546504, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 11955 + }, + { + "epoch": 0.11956, + "grad_norm": 1.1205161898186382, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 11956 + }, + { + "epoch": 0.11957, + "grad_norm": 0.9815044846386909, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 11957 + }, + { + "epoch": 0.11958, + "grad_norm": 1.0013378538157867, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 11958 + }, + { + "epoch": 0.11959, + "grad_norm": 1.000998732925407, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 11959 + }, + { + "epoch": 0.1196, + "grad_norm": 1.0893670772900628, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 11960 + }, + { + "epoch": 0.11961, + "grad_norm": 1.0591646896060922, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 11961 + }, + { + "epoch": 0.11962, + "grad_norm": 1.093398185067271, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 11962 + }, + { + "epoch": 0.11963, + "grad_norm": 1.2085490823454053, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 11963 + }, + { + "epoch": 0.11964, + "grad_norm": 1.1371502930714275, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 11964 + }, + { + "epoch": 0.11965, + "grad_norm": 1.1255892314647256, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 11965 + }, + { + "epoch": 0.11966, + "grad_norm": 1.0853962886107014, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 11966 + }, + { + "epoch": 0.11967, + "grad_norm": 1.3747539499194141, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 11967 + }, + { + "epoch": 0.11968, + "grad_norm": 1.2647185472649858, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 11968 + }, + { + "epoch": 0.11969, + "grad_norm": 1.0741775586091273, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 11969 + }, + { + "epoch": 0.1197, + "grad_norm": 1.0350752138665191, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 11970 + }, + { + "epoch": 0.11971, + "grad_norm": 1.191119641426522, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 11971 + }, + { + "epoch": 0.11972, + "grad_norm": 0.9437137639131957, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 11972 + }, + { + "epoch": 0.11973, + "grad_norm": 1.283613714036086, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 11973 + }, + { + "epoch": 0.11974, + "grad_norm": 1.279899355911932, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 11974 + }, + { + "epoch": 0.11975, + "grad_norm": 1.0249399599556033, + "learning_rate": 0.003, + "loss": 4.063, + "step": 11975 + }, + { + "epoch": 0.11976, + "grad_norm": 1.321212623947207, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 11976 + }, + { + "epoch": 0.11977, + "grad_norm": 1.0165950102473211, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 11977 + }, + { + "epoch": 0.11978, + "grad_norm": 0.9277193711873348, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 11978 + }, + { + "epoch": 0.11979, + "grad_norm": 0.9280891520464905, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 11979 + }, + { + "epoch": 0.1198, + "grad_norm": 0.9840854182938272, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 11980 + }, + { + "epoch": 0.11981, + "grad_norm": 1.3684303366293669, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 11981 + }, + { + "epoch": 0.11982, + "grad_norm": 0.9945901016820448, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 11982 + }, + { + "epoch": 0.11983, + "grad_norm": 1.2752089237002326, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 11983 + }, + { + "epoch": 0.11984, + "grad_norm": 1.0771984928694656, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 11984 + }, + { + "epoch": 0.11985, + "grad_norm": 1.3299901938391445, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11985 + }, + { + "epoch": 0.11986, + "grad_norm": 0.9586548273984268, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 11986 + }, + { + "epoch": 0.11987, + "grad_norm": 0.9948402579369039, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 11987 + }, + { + "epoch": 0.11988, + "grad_norm": 1.3296600841391015, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11988 + }, + { + "epoch": 0.11989, + "grad_norm": 0.8762782098697669, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 11989 + }, + { + "epoch": 0.1199, + "grad_norm": 1.076562501910153, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 11990 + }, + { + "epoch": 0.11991, + "grad_norm": 1.4625601275009004, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 11991 + }, + { + "epoch": 0.11992, + "grad_norm": 1.00751868644314, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 11992 + }, + { + "epoch": 0.11993, + "grad_norm": 1.3989214406531236, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 11993 + }, + { + "epoch": 0.11994, + "grad_norm": 0.7532071646779008, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 11994 + }, + { + "epoch": 0.11995, + "grad_norm": 0.9108244533553659, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 11995 + }, + { + "epoch": 0.11996, + "grad_norm": 1.037693604414362, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 11996 + }, + { + "epoch": 0.11997, + "grad_norm": 1.510249381824229, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 11997 + }, + { + "epoch": 0.11998, + "grad_norm": 0.9025590524571965, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 11998 + }, + { + "epoch": 0.11999, + "grad_norm": 1.189616049192571, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 11999 + }, + { + "epoch": 0.12, + "grad_norm": 1.3350930490406336, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 12000 + }, + { + "epoch": 0.12001, + "grad_norm": 1.1180567164673338, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12001 + }, + { + "epoch": 0.12002, + "grad_norm": 1.1746116800621182, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12002 + }, + { + "epoch": 0.12003, + "grad_norm": 1.0368064271859383, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 12003 + }, + { + "epoch": 0.12004, + "grad_norm": 1.0681096635948948, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12004 + }, + { + "epoch": 0.12005, + "grad_norm": 1.1099414967667618, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 12005 + }, + { + "epoch": 0.12006, + "grad_norm": 0.9803751601666846, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 12006 + }, + { + "epoch": 0.12007, + "grad_norm": 1.2223238024022165, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12007 + }, + { + "epoch": 0.12008, + "grad_norm": 1.0685166133012711, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 12008 + }, + { + "epoch": 0.12009, + "grad_norm": 1.1599099213556685, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 12009 + }, + { + "epoch": 0.1201, + "grad_norm": 1.021768365708924, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 12010 + }, + { + "epoch": 0.12011, + "grad_norm": 1.1087618201802516, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 12011 + }, + { + "epoch": 0.12012, + "grad_norm": 1.3091376574754292, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12012 + }, + { + "epoch": 0.12013, + "grad_norm": 1.3389948200754302, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12013 + }, + { + "epoch": 0.12014, + "grad_norm": 1.0677259298014297, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 12014 + }, + { + "epoch": 0.12015, + "grad_norm": 1.3527692324615435, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12015 + }, + { + "epoch": 0.12016, + "grad_norm": 1.1878208635574032, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 12016 + }, + { + "epoch": 0.12017, + "grad_norm": 1.1556818900091732, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 12017 + }, + { + "epoch": 0.12018, + "grad_norm": 1.2874267600368892, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 12018 + }, + { + "epoch": 0.12019, + "grad_norm": 0.9336107902349933, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 12019 + }, + { + "epoch": 0.1202, + "grad_norm": 1.2932714597916453, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 12020 + }, + { + "epoch": 0.12021, + "grad_norm": 1.0130309654760399, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 12021 + }, + { + "epoch": 0.12022, + "grad_norm": 1.2879254166353613, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 12022 + }, + { + "epoch": 0.12023, + "grad_norm": 1.0818443860488969, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12023 + }, + { + "epoch": 0.12024, + "grad_norm": 1.0860566591981275, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12024 + }, + { + "epoch": 0.12025, + "grad_norm": 1.2135435164557746, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 12025 + }, + { + "epoch": 0.12026, + "grad_norm": 0.9841989092312695, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 12026 + }, + { + "epoch": 0.12027, + "grad_norm": 1.0266977635183048, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 12027 + }, + { + "epoch": 0.12028, + "grad_norm": 1.2117018689721994, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 12028 + }, + { + "epoch": 0.12029, + "grad_norm": 1.2276826573957662, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12029 + }, + { + "epoch": 0.1203, + "grad_norm": 1.1518195603304868, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12030 + }, + { + "epoch": 0.12031, + "grad_norm": 1.2467861618259808, + "learning_rate": 0.003, + "loss": 4.083, + "step": 12031 + }, + { + "epoch": 0.12032, + "grad_norm": 1.16415301239379, + "learning_rate": 0.003, + "loss": 4.071, + "step": 12032 + }, + { + "epoch": 0.12033, + "grad_norm": 1.0590101051657046, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 12033 + }, + { + "epoch": 0.12034, + "grad_norm": 1.2402649179536591, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 12034 + }, + { + "epoch": 0.12035, + "grad_norm": 1.0642608565899094, + "learning_rate": 0.003, + "loss": 4.046, + "step": 12035 + }, + { + "epoch": 0.12036, + "grad_norm": 1.1255827242322958, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 12036 + }, + { + "epoch": 0.12037, + "grad_norm": 1.1384628048440417, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 12037 + }, + { + "epoch": 0.12038, + "grad_norm": 1.1274178417869605, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12038 + }, + { + "epoch": 0.12039, + "grad_norm": 0.9995922903631983, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12039 + }, + { + "epoch": 0.1204, + "grad_norm": 1.026389899399592, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12040 + }, + { + "epoch": 0.12041, + "grad_norm": 1.1044160233714642, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 12041 + }, + { + "epoch": 0.12042, + "grad_norm": 1.080147358818234, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12042 + }, + { + "epoch": 0.12043, + "grad_norm": 1.121781114814016, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 12043 + }, + { + "epoch": 0.12044, + "grad_norm": 1.0909551039016203, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 12044 + }, + { + "epoch": 0.12045, + "grad_norm": 1.191158141010099, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12045 + }, + { + "epoch": 0.12046, + "grad_norm": 1.066915824442074, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12046 + }, + { + "epoch": 0.12047, + "grad_norm": 0.9596003185844033, + "learning_rate": 0.003, + "loss": 4.094, + "step": 12047 + }, + { + "epoch": 0.12048, + "grad_norm": 1.1701885942631547, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12048 + }, + { + "epoch": 0.12049, + "grad_norm": 1.0547752412751898, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 12049 + }, + { + "epoch": 0.1205, + "grad_norm": 1.410530280129304, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 12050 + }, + { + "epoch": 0.12051, + "grad_norm": 1.2104800826176374, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 12051 + }, + { + "epoch": 0.12052, + "grad_norm": 1.259195544404022, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 12052 + }, + { + "epoch": 0.12053, + "grad_norm": 1.0120480121923963, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12053 + }, + { + "epoch": 0.12054, + "grad_norm": 1.2133091190993877, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 12054 + }, + { + "epoch": 0.12055, + "grad_norm": 1.086573324709519, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 12055 + }, + { + "epoch": 0.12056, + "grad_norm": 0.9366357796969548, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12056 + }, + { + "epoch": 0.12057, + "grad_norm": 0.9413207797417531, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12057 + }, + { + "epoch": 0.12058, + "grad_norm": 1.054970577763672, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 12058 + }, + { + "epoch": 0.12059, + "grad_norm": 1.241755057586524, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 12059 + }, + { + "epoch": 0.1206, + "grad_norm": 1.178759157791092, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12060 + }, + { + "epoch": 0.12061, + "grad_norm": 1.072634607621007, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 12061 + }, + { + "epoch": 0.12062, + "grad_norm": 1.2335975090778577, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 12062 + }, + { + "epoch": 0.12063, + "grad_norm": 1.0043724859492582, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12063 + }, + { + "epoch": 0.12064, + "grad_norm": 1.1268181667654482, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12064 + }, + { + "epoch": 0.12065, + "grad_norm": 1.3467584212302015, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12065 + }, + { + "epoch": 0.12066, + "grad_norm": 1.1544538001555558, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 12066 + }, + { + "epoch": 0.12067, + "grad_norm": 1.255977573156133, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 12067 + }, + { + "epoch": 0.12068, + "grad_norm": 0.9350775949046741, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 12068 + }, + { + "epoch": 0.12069, + "grad_norm": 1.088070082817422, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12069 + }, + { + "epoch": 0.1207, + "grad_norm": 1.0486078245374844, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 12070 + }, + { + "epoch": 0.12071, + "grad_norm": 1.107745717985881, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12071 + }, + { + "epoch": 0.12072, + "grad_norm": 1.376209870161379, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 12072 + }, + { + "epoch": 0.12073, + "grad_norm": 1.0983080244094117, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 12073 + }, + { + "epoch": 0.12074, + "grad_norm": 1.2145115045698978, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 12074 + }, + { + "epoch": 0.12075, + "grad_norm": 1.0792027485780036, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 12075 + }, + { + "epoch": 0.12076, + "grad_norm": 0.9724230941494237, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 12076 + }, + { + "epoch": 0.12077, + "grad_norm": 1.0309662283880128, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12077 + }, + { + "epoch": 0.12078, + "grad_norm": 1.4142456890578097, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 12078 + }, + { + "epoch": 0.12079, + "grad_norm": 1.0604933955299787, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 12079 + }, + { + "epoch": 0.1208, + "grad_norm": 1.1335698224805126, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 12080 + }, + { + "epoch": 0.12081, + "grad_norm": 1.015134918616046, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12081 + }, + { + "epoch": 0.12082, + "grad_norm": 1.3097320253417457, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12082 + }, + { + "epoch": 0.12083, + "grad_norm": 1.0066362360319434, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 12083 + }, + { + "epoch": 0.12084, + "grad_norm": 1.0235728852892743, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 12084 + }, + { + "epoch": 0.12085, + "grad_norm": 1.0630954451681172, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 12085 + }, + { + "epoch": 0.12086, + "grad_norm": 1.218539724165097, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 12086 + }, + { + "epoch": 0.12087, + "grad_norm": 1.1620137919872433, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 12087 + }, + { + "epoch": 0.12088, + "grad_norm": 1.177282673119695, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12088 + }, + { + "epoch": 0.12089, + "grad_norm": 0.9972352639834798, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12089 + }, + { + "epoch": 0.1209, + "grad_norm": 1.0959043321646311, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12090 + }, + { + "epoch": 0.12091, + "grad_norm": 1.1874303616047068, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 12091 + }, + { + "epoch": 0.12092, + "grad_norm": 1.101021855644516, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12092 + }, + { + "epoch": 0.12093, + "grad_norm": 1.2633017678088525, + "learning_rate": 0.003, + "loss": 4.09, + "step": 12093 + }, + { + "epoch": 0.12094, + "grad_norm": 1.0519698321311366, + "learning_rate": 0.003, + "loss": 4.087, + "step": 12094 + }, + { + "epoch": 0.12095, + "grad_norm": 1.1665993943190327, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 12095 + }, + { + "epoch": 0.12096, + "grad_norm": 1.1096000107872868, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12096 + }, + { + "epoch": 0.12097, + "grad_norm": 1.0635730265142278, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 12097 + }, + { + "epoch": 0.12098, + "grad_norm": 1.2764783466361678, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 12098 + }, + { + "epoch": 0.12099, + "grad_norm": 0.9051264914681529, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 12099 + }, + { + "epoch": 0.121, + "grad_norm": 0.9249608490415164, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 12100 + }, + { + "epoch": 0.12101, + "grad_norm": 1.0835718975720605, + "learning_rate": 0.003, + "loss": 4.078, + "step": 12101 + }, + { + "epoch": 0.12102, + "grad_norm": 1.1642303094091533, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12102 + }, + { + "epoch": 0.12103, + "grad_norm": 1.1982430786318965, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12103 + }, + { + "epoch": 0.12104, + "grad_norm": 0.9021525338319302, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 12104 + }, + { + "epoch": 0.12105, + "grad_norm": 1.025465093748409, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 12105 + }, + { + "epoch": 0.12106, + "grad_norm": 1.0287080964025033, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 12106 + }, + { + "epoch": 0.12107, + "grad_norm": 1.1893433762623093, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 12107 + }, + { + "epoch": 0.12108, + "grad_norm": 1.1346629077604669, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 12108 + }, + { + "epoch": 0.12109, + "grad_norm": 1.384241655173453, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12109 + }, + { + "epoch": 0.1211, + "grad_norm": 1.1929830857172083, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12110 + }, + { + "epoch": 0.12111, + "grad_norm": 1.2157542584926002, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12111 + }, + { + "epoch": 0.12112, + "grad_norm": 1.1414637982634577, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 12112 + }, + { + "epoch": 0.12113, + "grad_norm": 1.0356391274044419, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 12113 + }, + { + "epoch": 0.12114, + "grad_norm": 1.182813601046115, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12114 + }, + { + "epoch": 0.12115, + "grad_norm": 0.9762033320326391, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12115 + }, + { + "epoch": 0.12116, + "grad_norm": 1.1921654583478314, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 12116 + }, + { + "epoch": 0.12117, + "grad_norm": 1.341606992951073, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 12117 + }, + { + "epoch": 0.12118, + "grad_norm": 0.8182518816856665, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12118 + }, + { + "epoch": 0.12119, + "grad_norm": 0.8704177926517407, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12119 + }, + { + "epoch": 0.1212, + "grad_norm": 1.0058197349809026, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 12120 + }, + { + "epoch": 0.12121, + "grad_norm": 1.069283234495121, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12121 + }, + { + "epoch": 0.12122, + "grad_norm": 0.9201542200711905, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 12122 + }, + { + "epoch": 0.12123, + "grad_norm": 1.110694878714642, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 12123 + }, + { + "epoch": 0.12124, + "grad_norm": 1.1291370286216942, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12124 + }, + { + "epoch": 0.12125, + "grad_norm": 1.161710409599256, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 12125 + }, + { + "epoch": 0.12126, + "grad_norm": 1.2729178933350158, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12126 + }, + { + "epoch": 0.12127, + "grad_norm": 1.1261437334380782, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 12127 + }, + { + "epoch": 0.12128, + "grad_norm": 1.093746833791882, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 12128 + }, + { + "epoch": 0.12129, + "grad_norm": 1.2284949310268252, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 12129 + }, + { + "epoch": 0.1213, + "grad_norm": 0.9984993555404615, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12130 + }, + { + "epoch": 0.12131, + "grad_norm": 1.0508566744500767, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 12131 + }, + { + "epoch": 0.12132, + "grad_norm": 1.140820512254794, + "learning_rate": 0.003, + "loss": 4.089, + "step": 12132 + }, + { + "epoch": 0.12133, + "grad_norm": 1.0677992380438892, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 12133 + }, + { + "epoch": 0.12134, + "grad_norm": 1.041338898545635, + "learning_rate": 0.003, + "loss": 4.092, + "step": 12134 + }, + { + "epoch": 0.12135, + "grad_norm": 1.3946407208815519, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12135 + }, + { + "epoch": 0.12136, + "grad_norm": 1.1006237334365854, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12136 + }, + { + "epoch": 0.12137, + "grad_norm": 1.458823136129887, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 12137 + }, + { + "epoch": 0.12138, + "grad_norm": 1.0559635113382873, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 12138 + }, + { + "epoch": 0.12139, + "grad_norm": 1.1285272904820276, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 12139 + }, + { + "epoch": 0.1214, + "grad_norm": 1.0897185720280367, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 12140 + }, + { + "epoch": 0.12141, + "grad_norm": 1.1955016181778377, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12141 + }, + { + "epoch": 0.12142, + "grad_norm": 1.325169066507099, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 12142 + }, + { + "epoch": 0.12143, + "grad_norm": 1.035485453079449, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 12143 + }, + { + "epoch": 0.12144, + "grad_norm": 1.226647122460927, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 12144 + }, + { + "epoch": 0.12145, + "grad_norm": 0.9832385513042932, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 12145 + }, + { + "epoch": 0.12146, + "grad_norm": 1.3042070469595108, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 12146 + }, + { + "epoch": 0.12147, + "grad_norm": 1.2735050426487993, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12147 + }, + { + "epoch": 0.12148, + "grad_norm": 1.011082663592818, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 12148 + }, + { + "epoch": 0.12149, + "grad_norm": 1.2474343009869946, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12149 + }, + { + "epoch": 0.1215, + "grad_norm": 1.079953110653683, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12150 + }, + { + "epoch": 0.12151, + "grad_norm": 1.1442982278151164, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 12151 + }, + { + "epoch": 0.12152, + "grad_norm": 0.9983388572149424, + "learning_rate": 0.003, + "loss": 4.055, + "step": 12152 + }, + { + "epoch": 0.12153, + "grad_norm": 1.2742688036210805, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 12153 + }, + { + "epoch": 0.12154, + "grad_norm": 1.1950400478983967, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12154 + }, + { + "epoch": 0.12155, + "grad_norm": 1.19847125467165, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 12155 + }, + { + "epoch": 0.12156, + "grad_norm": 1.0274718673323897, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 12156 + }, + { + "epoch": 0.12157, + "grad_norm": 1.100133073873854, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12157 + }, + { + "epoch": 0.12158, + "grad_norm": 1.1818660159283922, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 12158 + }, + { + "epoch": 0.12159, + "grad_norm": 1.13858706810257, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12159 + }, + { + "epoch": 0.1216, + "grad_norm": 0.9275078508631227, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12160 + }, + { + "epoch": 0.12161, + "grad_norm": 1.1328726130452051, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12161 + }, + { + "epoch": 0.12162, + "grad_norm": 1.2164427451782454, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12162 + }, + { + "epoch": 0.12163, + "grad_norm": 1.231293188312221, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 12163 + }, + { + "epoch": 0.12164, + "grad_norm": 1.000554865386488, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 12164 + }, + { + "epoch": 0.12165, + "grad_norm": 1.2202696872836503, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 12165 + }, + { + "epoch": 0.12166, + "grad_norm": 1.0390095032635105, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12166 + }, + { + "epoch": 0.12167, + "grad_norm": 1.0585462345095171, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 12167 + }, + { + "epoch": 0.12168, + "grad_norm": 1.2386723240033908, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12168 + }, + { + "epoch": 0.12169, + "grad_norm": 0.9660827919308314, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 12169 + }, + { + "epoch": 0.1217, + "grad_norm": 1.1345186857579614, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 12170 + }, + { + "epoch": 0.12171, + "grad_norm": 1.0463712239949596, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12171 + }, + { + "epoch": 0.12172, + "grad_norm": 1.1581387142451605, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 12172 + }, + { + "epoch": 0.12173, + "grad_norm": 1.0109749774734238, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12173 + }, + { + "epoch": 0.12174, + "grad_norm": 1.2056594441816655, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 12174 + }, + { + "epoch": 0.12175, + "grad_norm": 1.1351586399822327, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 12175 + }, + { + "epoch": 0.12176, + "grad_norm": 1.1631779505207185, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 12176 + }, + { + "epoch": 0.12177, + "grad_norm": 1.1536486927661425, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12177 + }, + { + "epoch": 0.12178, + "grad_norm": 1.0692593338617544, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12178 + }, + { + "epoch": 0.12179, + "grad_norm": 1.1624174038135675, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 12179 + }, + { + "epoch": 0.1218, + "grad_norm": 0.9717071055481967, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12180 + }, + { + "epoch": 0.12181, + "grad_norm": 1.098385052809246, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12181 + }, + { + "epoch": 0.12182, + "grad_norm": 1.1531436274372961, + "learning_rate": 0.003, + "loss": 4.081, + "step": 12182 + }, + { + "epoch": 0.12183, + "grad_norm": 0.9570030729100146, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 12183 + }, + { + "epoch": 0.12184, + "grad_norm": 1.0412301813920144, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 12184 + }, + { + "epoch": 0.12185, + "grad_norm": 1.3219954197115575, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12185 + }, + { + "epoch": 0.12186, + "grad_norm": 1.2067563045208698, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 12186 + }, + { + "epoch": 0.12187, + "grad_norm": 1.3637614562032707, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 12187 + }, + { + "epoch": 0.12188, + "grad_norm": 1.082708529547863, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 12188 + }, + { + "epoch": 0.12189, + "grad_norm": 1.2208834261957746, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12189 + }, + { + "epoch": 0.1219, + "grad_norm": 1.086438339445444, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12190 + }, + { + "epoch": 0.12191, + "grad_norm": 0.9994395476535417, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 12191 + }, + { + "epoch": 0.12192, + "grad_norm": 1.1637447492294328, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 12192 + }, + { + "epoch": 0.12193, + "grad_norm": 1.0482426916290108, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 12193 + }, + { + "epoch": 0.12194, + "grad_norm": 1.19499071880003, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 12194 + }, + { + "epoch": 0.12195, + "grad_norm": 1.0344994185916214, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12195 + }, + { + "epoch": 0.12196, + "grad_norm": 1.2946632998055925, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 12196 + }, + { + "epoch": 0.12197, + "grad_norm": 1.1635059304700333, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12197 + }, + { + "epoch": 0.12198, + "grad_norm": 0.927098380296469, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12198 + }, + { + "epoch": 0.12199, + "grad_norm": 1.057694142435552, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12199 + }, + { + "epoch": 0.122, + "grad_norm": 1.3472658537272546, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 12200 + }, + { + "epoch": 0.12201, + "grad_norm": 1.0889234944776072, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 12201 + }, + { + "epoch": 0.12202, + "grad_norm": 1.3188432472041998, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 12202 + }, + { + "epoch": 0.12203, + "grad_norm": 0.9595579361138602, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 12203 + }, + { + "epoch": 0.12204, + "grad_norm": 1.1922547447780105, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 12204 + }, + { + "epoch": 0.12205, + "grad_norm": 1.2992635103281398, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 12205 + }, + { + "epoch": 0.12206, + "grad_norm": 0.9557362537371793, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12206 + }, + { + "epoch": 0.12207, + "grad_norm": 0.9184107752252704, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 12207 + }, + { + "epoch": 0.12208, + "grad_norm": 1.178789407610477, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 12208 + }, + { + "epoch": 0.12209, + "grad_norm": 1.1231869598587525, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12209 + }, + { + "epoch": 0.1221, + "grad_norm": 1.2585732335258468, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12210 + }, + { + "epoch": 0.12211, + "grad_norm": 1.309785344594289, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 12211 + }, + { + "epoch": 0.12212, + "grad_norm": 1.1308828923601764, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12212 + }, + { + "epoch": 0.12213, + "grad_norm": 1.472297745706311, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12213 + }, + { + "epoch": 0.12214, + "grad_norm": 1.1166151733609975, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 12214 + }, + { + "epoch": 0.12215, + "grad_norm": 1.0804515266161436, + "learning_rate": 0.003, + "loss": 4.104, + "step": 12215 + }, + { + "epoch": 0.12216, + "grad_norm": 1.2400711237317346, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 12216 + }, + { + "epoch": 0.12217, + "grad_norm": 0.9418513988414021, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 12217 + }, + { + "epoch": 0.12218, + "grad_norm": 1.1394363437989878, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 12218 + }, + { + "epoch": 0.12219, + "grad_norm": 1.1372327507115731, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 12219 + }, + { + "epoch": 0.1222, + "grad_norm": 1.161677959167776, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12220 + }, + { + "epoch": 0.12221, + "grad_norm": 1.0789544842404515, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12221 + }, + { + "epoch": 0.12222, + "grad_norm": 1.0398693221757789, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12222 + }, + { + "epoch": 0.12223, + "grad_norm": 1.1267339048732326, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 12223 + }, + { + "epoch": 0.12224, + "grad_norm": 1.2760752189140523, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 12224 + }, + { + "epoch": 0.12225, + "grad_norm": 0.9262198806329611, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12225 + }, + { + "epoch": 0.12226, + "grad_norm": 1.043866646116881, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 12226 + }, + { + "epoch": 0.12227, + "grad_norm": 1.26294589943615, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 12227 + }, + { + "epoch": 0.12228, + "grad_norm": 0.7806838445270528, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12228 + }, + { + "epoch": 0.12229, + "grad_norm": 1.1547423675011483, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 12229 + }, + { + "epoch": 0.1223, + "grad_norm": 1.1320036145420767, + "learning_rate": 0.003, + "loss": 4.105, + "step": 12230 + }, + { + "epoch": 0.12231, + "grad_norm": 1.3764638273278833, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 12231 + }, + { + "epoch": 0.12232, + "grad_norm": 1.0128482401645624, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 12232 + }, + { + "epoch": 0.12233, + "grad_norm": 1.1726363376664437, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 12233 + }, + { + "epoch": 0.12234, + "grad_norm": 1.1040794996492702, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 12234 + }, + { + "epoch": 0.12235, + "grad_norm": 1.0358138471123548, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 12235 + }, + { + "epoch": 0.12236, + "grad_norm": 1.2145747382949974, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 12236 + }, + { + "epoch": 0.12237, + "grad_norm": 1.2700041322440572, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 12237 + }, + { + "epoch": 0.12238, + "grad_norm": 1.1154508261314582, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12238 + }, + { + "epoch": 0.12239, + "grad_norm": 1.0109089876091282, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 12239 + }, + { + "epoch": 0.1224, + "grad_norm": 1.185544974621423, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12240 + }, + { + "epoch": 0.12241, + "grad_norm": 1.1933483328999805, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12241 + }, + { + "epoch": 0.12242, + "grad_norm": 1.2439943885465512, + "learning_rate": 0.003, + "loss": 4.081, + "step": 12242 + }, + { + "epoch": 0.12243, + "grad_norm": 1.2486428182377738, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12243 + }, + { + "epoch": 0.12244, + "grad_norm": 1.2058088997505476, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 12244 + }, + { + "epoch": 0.12245, + "grad_norm": 1.0118112593781434, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 12245 + }, + { + "epoch": 0.12246, + "grad_norm": 1.2149480279494373, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12246 + }, + { + "epoch": 0.12247, + "grad_norm": 1.178407916487301, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 12247 + }, + { + "epoch": 0.12248, + "grad_norm": 1.0458635406153554, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 12248 + }, + { + "epoch": 0.12249, + "grad_norm": 1.1773227629424923, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 12249 + }, + { + "epoch": 0.1225, + "grad_norm": 1.1971688988027553, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12250 + }, + { + "epoch": 0.12251, + "grad_norm": 1.130549267275754, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 12251 + }, + { + "epoch": 0.12252, + "grad_norm": 0.9347182431246374, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12252 + }, + { + "epoch": 0.12253, + "grad_norm": 1.020618982790968, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 12253 + }, + { + "epoch": 0.12254, + "grad_norm": 1.2336898416551805, + "learning_rate": 0.003, + "loss": 4.08, + "step": 12254 + }, + { + "epoch": 0.12255, + "grad_norm": 0.9198700961085998, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 12255 + }, + { + "epoch": 0.12256, + "grad_norm": 1.2550574554890428, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 12256 + }, + { + "epoch": 0.12257, + "grad_norm": 1.1934871340723554, + "learning_rate": 0.003, + "loss": 4.1, + "step": 12257 + }, + { + "epoch": 0.12258, + "grad_norm": 1.3489157246382726, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 12258 + }, + { + "epoch": 0.12259, + "grad_norm": 1.0125660002793162, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12259 + }, + { + "epoch": 0.1226, + "grad_norm": 1.2472895843558622, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 12260 + }, + { + "epoch": 0.12261, + "grad_norm": 0.954366178907707, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 12261 + }, + { + "epoch": 0.12262, + "grad_norm": 1.1015996587580237, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 12262 + }, + { + "epoch": 0.12263, + "grad_norm": 1.1436633592224317, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 12263 + }, + { + "epoch": 0.12264, + "grad_norm": 1.1346288912154268, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12264 + }, + { + "epoch": 0.12265, + "grad_norm": 1.0618662450209828, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 12265 + }, + { + "epoch": 0.12266, + "grad_norm": 1.178050447666747, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 12266 + }, + { + "epoch": 0.12267, + "grad_norm": 1.0553507490994913, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 12267 + }, + { + "epoch": 0.12268, + "grad_norm": 1.2296144758772443, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 12268 + }, + { + "epoch": 0.12269, + "grad_norm": 1.0457920930485292, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12269 + }, + { + "epoch": 0.1227, + "grad_norm": 1.0840324335435874, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 12270 + }, + { + "epoch": 0.12271, + "grad_norm": 1.1445528256408435, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 12271 + }, + { + "epoch": 0.12272, + "grad_norm": 1.0789090092167062, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12272 + }, + { + "epoch": 0.12273, + "grad_norm": 1.2904987590364192, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 12273 + }, + { + "epoch": 0.12274, + "grad_norm": 1.2998917837299295, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12274 + }, + { + "epoch": 0.12275, + "grad_norm": 0.9954805985642087, + "learning_rate": 0.003, + "loss": 4.096, + "step": 12275 + }, + { + "epoch": 0.12276, + "grad_norm": 1.29013688740999, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 12276 + }, + { + "epoch": 0.12277, + "grad_norm": 0.8047671585111673, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12277 + }, + { + "epoch": 0.12278, + "grad_norm": 0.986527265960683, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 12278 + }, + { + "epoch": 0.12279, + "grad_norm": 1.2484795259575783, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12279 + }, + { + "epoch": 0.1228, + "grad_norm": 1.221383876281741, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12280 + }, + { + "epoch": 0.12281, + "grad_norm": 1.0856061468785485, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 12281 + }, + { + "epoch": 0.12282, + "grad_norm": 1.3831497350750577, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 12282 + }, + { + "epoch": 0.12283, + "grad_norm": 1.035678610826925, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12283 + }, + { + "epoch": 0.12284, + "grad_norm": 1.0961280068062669, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12284 + }, + { + "epoch": 0.12285, + "grad_norm": 1.2558527303666362, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12285 + }, + { + "epoch": 0.12286, + "grad_norm": 0.972650104375496, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 12286 + }, + { + "epoch": 0.12287, + "grad_norm": 1.0085264354475671, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12287 + }, + { + "epoch": 0.12288, + "grad_norm": 1.2253160289482012, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 12288 + }, + { + "epoch": 0.12289, + "grad_norm": 1.2852323689267404, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 12289 + }, + { + "epoch": 0.1229, + "grad_norm": 1.3072183218190938, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 12290 + }, + { + "epoch": 0.12291, + "grad_norm": 1.0026332220982133, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 12291 + }, + { + "epoch": 0.12292, + "grad_norm": 1.1161617368412051, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 12292 + }, + { + "epoch": 0.12293, + "grad_norm": 1.217967475014931, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 12293 + }, + { + "epoch": 0.12294, + "grad_norm": 1.2381766283955784, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 12294 + }, + { + "epoch": 0.12295, + "grad_norm": 0.9408012807630449, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12295 + }, + { + "epoch": 0.12296, + "grad_norm": 1.189999796632258, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12296 + }, + { + "epoch": 0.12297, + "grad_norm": 1.030538072411565, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 12297 + }, + { + "epoch": 0.12298, + "grad_norm": 1.253456405281475, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 12298 + }, + { + "epoch": 0.12299, + "grad_norm": 0.9763589760714394, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 12299 + }, + { + "epoch": 0.123, + "grad_norm": 1.1196444799820449, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 12300 + }, + { + "epoch": 0.12301, + "grad_norm": 1.3099114731176258, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 12301 + }, + { + "epoch": 0.12302, + "grad_norm": 1.1333606762203758, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 12302 + }, + { + "epoch": 0.12303, + "grad_norm": 1.1145907132725608, + "learning_rate": 0.003, + "loss": 4.1419, + "step": 12303 + }, + { + "epoch": 0.12304, + "grad_norm": 1.236368919516794, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 12304 + }, + { + "epoch": 0.12305, + "grad_norm": 1.1082195413575602, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 12305 + }, + { + "epoch": 0.12306, + "grad_norm": 1.0898656695938542, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12306 + }, + { + "epoch": 0.12307, + "grad_norm": 1.408135156441792, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12307 + }, + { + "epoch": 0.12308, + "grad_norm": 1.1627256450794383, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 12308 + }, + { + "epoch": 0.12309, + "grad_norm": 1.0927700416865638, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12309 + }, + { + "epoch": 0.1231, + "grad_norm": 1.2506427098071364, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12310 + }, + { + "epoch": 0.12311, + "grad_norm": 1.036971362561517, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 12311 + }, + { + "epoch": 0.12312, + "grad_norm": 1.151090599954385, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 12312 + }, + { + "epoch": 0.12313, + "grad_norm": 1.08749534761496, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 12313 + }, + { + "epoch": 0.12314, + "grad_norm": 1.0919178585163267, + "learning_rate": 0.003, + "loss": 4.067, + "step": 12314 + }, + { + "epoch": 0.12315, + "grad_norm": 1.0458368945872953, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 12315 + }, + { + "epoch": 0.12316, + "grad_norm": 1.1467188804604123, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 12316 + }, + { + "epoch": 0.12317, + "grad_norm": 1.1735741971873328, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12317 + }, + { + "epoch": 0.12318, + "grad_norm": 1.1233946426565826, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 12318 + }, + { + "epoch": 0.12319, + "grad_norm": 1.0783419878905267, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 12319 + }, + { + "epoch": 0.1232, + "grad_norm": 1.3161800684804994, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 12320 + }, + { + "epoch": 0.12321, + "grad_norm": 1.2003452285321743, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 12321 + }, + { + "epoch": 0.12322, + "grad_norm": 1.0232241990056046, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 12322 + }, + { + "epoch": 0.12323, + "grad_norm": 1.4150191393922857, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 12323 + }, + { + "epoch": 0.12324, + "grad_norm": 0.9549941813608207, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 12324 + }, + { + "epoch": 0.12325, + "grad_norm": 1.2920608946841499, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 12325 + }, + { + "epoch": 0.12326, + "grad_norm": 1.1018558427657896, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12326 + }, + { + "epoch": 0.12327, + "grad_norm": 1.052198948646688, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12327 + }, + { + "epoch": 0.12328, + "grad_norm": 1.340495755595897, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12328 + }, + { + "epoch": 0.12329, + "grad_norm": 1.0170314488453516, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 12329 + }, + { + "epoch": 0.1233, + "grad_norm": 1.1788094344761215, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 12330 + }, + { + "epoch": 0.12331, + "grad_norm": 1.197821830527221, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12331 + }, + { + "epoch": 0.12332, + "grad_norm": 1.0098326384816398, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12332 + }, + { + "epoch": 0.12333, + "grad_norm": 1.2040309792724453, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 12333 + }, + { + "epoch": 0.12334, + "grad_norm": 1.1531517961981639, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12334 + }, + { + "epoch": 0.12335, + "grad_norm": 1.3252120350812502, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 12335 + }, + { + "epoch": 0.12336, + "grad_norm": 0.9357893130632275, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12336 + }, + { + "epoch": 0.12337, + "grad_norm": 1.2881720330741127, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12337 + }, + { + "epoch": 0.12338, + "grad_norm": 1.327319438684096, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 12338 + }, + { + "epoch": 0.12339, + "grad_norm": 0.7620877981202213, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 12339 + }, + { + "epoch": 0.1234, + "grad_norm": 0.9047840231504427, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 12340 + }, + { + "epoch": 0.12341, + "grad_norm": 1.366151905719102, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 12341 + }, + { + "epoch": 0.12342, + "grad_norm": 1.0551986957803026, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 12342 + }, + { + "epoch": 0.12343, + "grad_norm": 1.2070422667122573, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 12343 + }, + { + "epoch": 0.12344, + "grad_norm": 1.0631378446641342, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 12344 + }, + { + "epoch": 0.12345, + "grad_norm": 1.158407287153162, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12345 + }, + { + "epoch": 0.12346, + "grad_norm": 1.012305861351394, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 12346 + }, + { + "epoch": 0.12347, + "grad_norm": 1.1892828019352273, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 12347 + }, + { + "epoch": 0.12348, + "grad_norm": 1.3626667884912291, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12348 + }, + { + "epoch": 0.12349, + "grad_norm": 1.1332303731996494, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 12349 + }, + { + "epoch": 0.1235, + "grad_norm": 1.0076252552415714, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 12350 + }, + { + "epoch": 0.12351, + "grad_norm": 1.3460749776655139, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 12351 + }, + { + "epoch": 0.12352, + "grad_norm": 0.8901594605028754, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 12352 + }, + { + "epoch": 0.12353, + "grad_norm": 1.143898844720561, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12353 + }, + { + "epoch": 0.12354, + "grad_norm": 1.3439182405422685, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12354 + }, + { + "epoch": 0.12355, + "grad_norm": 1.0678329059929563, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 12355 + }, + { + "epoch": 0.12356, + "grad_norm": 1.4136298856727907, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 12356 + }, + { + "epoch": 0.12357, + "grad_norm": 1.2423905187146829, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 12357 + }, + { + "epoch": 0.12358, + "grad_norm": 0.9767041739713436, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12358 + }, + { + "epoch": 0.12359, + "grad_norm": 1.1113949696387744, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12359 + }, + { + "epoch": 0.1236, + "grad_norm": 1.0883575699397783, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12360 + }, + { + "epoch": 0.12361, + "grad_norm": 1.0359979808099924, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12361 + }, + { + "epoch": 0.12362, + "grad_norm": 1.1557713223728756, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 12362 + }, + { + "epoch": 0.12363, + "grad_norm": 1.14932398730516, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 12363 + }, + { + "epoch": 0.12364, + "grad_norm": 1.1830974866306603, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12364 + }, + { + "epoch": 0.12365, + "grad_norm": 1.3246802002730393, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12365 + }, + { + "epoch": 0.12366, + "grad_norm": 1.0784211269847974, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 12366 + }, + { + "epoch": 0.12367, + "grad_norm": 1.0898334963731273, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 12367 + }, + { + "epoch": 0.12368, + "grad_norm": 1.1578889814224254, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12368 + }, + { + "epoch": 0.12369, + "grad_norm": 1.062330069966536, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 12369 + }, + { + "epoch": 0.1237, + "grad_norm": 1.2139385466118495, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 12370 + }, + { + "epoch": 0.12371, + "grad_norm": 1.0020124725683932, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 12371 + }, + { + "epoch": 0.12372, + "grad_norm": 1.1484049368261768, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 12372 + }, + { + "epoch": 0.12373, + "grad_norm": 1.1737217200295595, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 12373 + }, + { + "epoch": 0.12374, + "grad_norm": 0.9986106363869645, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 12374 + }, + { + "epoch": 0.12375, + "grad_norm": 1.3287108256165667, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12375 + }, + { + "epoch": 0.12376, + "grad_norm": 0.9346266004574802, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 12376 + }, + { + "epoch": 0.12377, + "grad_norm": 1.3672500381682775, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12377 + }, + { + "epoch": 0.12378, + "grad_norm": 1.1428084365292464, + "learning_rate": 0.003, + "loss": 4.11, + "step": 12378 + }, + { + "epoch": 0.12379, + "grad_norm": 1.488126705290297, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12379 + }, + { + "epoch": 0.1238, + "grad_norm": 0.9572206193227556, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12380 + }, + { + "epoch": 0.12381, + "grad_norm": 0.968112714151001, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 12381 + }, + { + "epoch": 0.12382, + "grad_norm": 1.0824430350452237, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 12382 + }, + { + "epoch": 0.12383, + "grad_norm": 1.1291656916632755, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 12383 + }, + { + "epoch": 0.12384, + "grad_norm": 1.004588532244519, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12384 + }, + { + "epoch": 0.12385, + "grad_norm": 1.1335076678685279, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 12385 + }, + { + "epoch": 0.12386, + "grad_norm": 1.016742229821469, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 12386 + }, + { + "epoch": 0.12387, + "grad_norm": 1.3524202383088189, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12387 + }, + { + "epoch": 0.12388, + "grad_norm": 1.0308871095461538, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12388 + }, + { + "epoch": 0.12389, + "grad_norm": 1.3224180622611557, + "learning_rate": 0.003, + "loss": 4.105, + "step": 12389 + }, + { + "epoch": 0.1239, + "grad_norm": 1.31342304381205, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 12390 + }, + { + "epoch": 0.12391, + "grad_norm": 1.1169817951291854, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12391 + }, + { + "epoch": 0.12392, + "grad_norm": 1.2741989835211562, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 12392 + }, + { + "epoch": 0.12393, + "grad_norm": 1.108532000671943, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 12393 + }, + { + "epoch": 0.12394, + "grad_norm": 1.066676166676712, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12394 + }, + { + "epoch": 0.12395, + "grad_norm": 1.1873521205310513, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 12395 + }, + { + "epoch": 0.12396, + "grad_norm": 1.1429627893687861, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 12396 + }, + { + "epoch": 0.12397, + "grad_norm": 1.2164184009580055, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 12397 + }, + { + "epoch": 0.12398, + "grad_norm": 1.0163836731887739, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 12398 + }, + { + "epoch": 0.12399, + "grad_norm": 1.0866617882267402, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 12399 + }, + { + "epoch": 0.124, + "grad_norm": 1.2176442736299438, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 12400 + }, + { + "epoch": 0.12401, + "grad_norm": 0.9756572130875887, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12401 + }, + { + "epoch": 0.12402, + "grad_norm": 1.0963849133526373, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 12402 + }, + { + "epoch": 0.12403, + "grad_norm": 1.007708889867285, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 12403 + }, + { + "epoch": 0.12404, + "grad_norm": 1.4254241170094033, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12404 + }, + { + "epoch": 0.12405, + "grad_norm": 1.1163735038983496, + "learning_rate": 0.003, + "loss": 4.086, + "step": 12405 + }, + { + "epoch": 0.12406, + "grad_norm": 1.208191698514626, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12406 + }, + { + "epoch": 0.12407, + "grad_norm": 1.3588304955679549, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 12407 + }, + { + "epoch": 0.12408, + "grad_norm": 1.050585779633214, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 12408 + }, + { + "epoch": 0.12409, + "grad_norm": 1.1268501042795123, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 12409 + }, + { + "epoch": 0.1241, + "grad_norm": 1.2536825066597725, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 12410 + }, + { + "epoch": 0.12411, + "grad_norm": 1.031264785916332, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 12411 + }, + { + "epoch": 0.12412, + "grad_norm": 1.184343588918058, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 12412 + }, + { + "epoch": 0.12413, + "grad_norm": 0.9025710769482064, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 12413 + }, + { + "epoch": 0.12414, + "grad_norm": 0.9955728246628766, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12414 + }, + { + "epoch": 0.12415, + "grad_norm": 1.2692295368173465, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 12415 + }, + { + "epoch": 0.12416, + "grad_norm": 0.9685638434293641, + "learning_rate": 0.003, + "loss": 4.095, + "step": 12416 + }, + { + "epoch": 0.12417, + "grad_norm": 1.1440919602921278, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 12417 + }, + { + "epoch": 0.12418, + "grad_norm": 1.1641712288650334, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 12418 + }, + { + "epoch": 0.12419, + "grad_norm": 1.0038079520291594, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12419 + }, + { + "epoch": 0.1242, + "grad_norm": 1.3473679768905762, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 12420 + }, + { + "epoch": 0.12421, + "grad_norm": 1.065665013632022, + "learning_rate": 0.003, + "loss": 4.101, + "step": 12421 + }, + { + "epoch": 0.12422, + "grad_norm": 1.3072996720657586, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 12422 + }, + { + "epoch": 0.12423, + "grad_norm": 1.1561999326808896, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 12423 + }, + { + "epoch": 0.12424, + "grad_norm": 1.1921792539983773, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 12424 + }, + { + "epoch": 0.12425, + "grad_norm": 1.1873182303310676, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 12425 + }, + { + "epoch": 0.12426, + "grad_norm": 1.0056230801509678, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 12426 + }, + { + "epoch": 0.12427, + "grad_norm": 1.1040079016766144, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 12427 + }, + { + "epoch": 0.12428, + "grad_norm": 1.2743386374313739, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 12428 + }, + { + "epoch": 0.12429, + "grad_norm": 1.0874576902520072, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 12429 + }, + { + "epoch": 0.1243, + "grad_norm": 1.123082312308546, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 12430 + }, + { + "epoch": 0.12431, + "grad_norm": 1.388091941410794, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 12431 + }, + { + "epoch": 0.12432, + "grad_norm": 1.0354864246784283, + "learning_rate": 0.003, + "loss": 4.089, + "step": 12432 + }, + { + "epoch": 0.12433, + "grad_norm": 1.3912030063770227, + "learning_rate": 0.003, + "loss": 4.096, + "step": 12433 + }, + { + "epoch": 0.12434, + "grad_norm": 0.9577813790528974, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 12434 + }, + { + "epoch": 0.12435, + "grad_norm": 1.0523873799438321, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 12435 + }, + { + "epoch": 0.12436, + "grad_norm": 1.211875416464265, + "learning_rate": 0.003, + "loss": 4.057, + "step": 12436 + }, + { + "epoch": 0.12437, + "grad_norm": 1.0160095230818849, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12437 + }, + { + "epoch": 0.12438, + "grad_norm": 1.1346874971281438, + "learning_rate": 0.003, + "loss": 4.059, + "step": 12438 + }, + { + "epoch": 0.12439, + "grad_norm": 1.062510295796496, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 12439 + }, + { + "epoch": 0.1244, + "grad_norm": 1.3567923004962648, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12440 + }, + { + "epoch": 0.12441, + "grad_norm": 1.2016760115144474, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 12441 + }, + { + "epoch": 0.12442, + "grad_norm": 0.9627995060103828, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 12442 + }, + { + "epoch": 0.12443, + "grad_norm": 1.2294875695669985, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12443 + }, + { + "epoch": 0.12444, + "grad_norm": 1.0422975531725298, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12444 + }, + { + "epoch": 0.12445, + "grad_norm": 1.058240202018281, + "learning_rate": 0.003, + "loss": 4.085, + "step": 12445 + }, + { + "epoch": 0.12446, + "grad_norm": 1.0860192893087985, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12446 + }, + { + "epoch": 0.12447, + "grad_norm": 1.0425529671685176, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 12447 + }, + { + "epoch": 0.12448, + "grad_norm": 1.153564456565648, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 12448 + }, + { + "epoch": 0.12449, + "grad_norm": 1.1238750683883074, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 12449 + }, + { + "epoch": 0.1245, + "grad_norm": 1.1043722679708285, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 12450 + }, + { + "epoch": 0.12451, + "grad_norm": 1.0674229419538308, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12451 + }, + { + "epoch": 0.12452, + "grad_norm": 1.1411490342650177, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 12452 + }, + { + "epoch": 0.12453, + "grad_norm": 1.0323760242004496, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 12453 + }, + { + "epoch": 0.12454, + "grad_norm": 1.3536660491480184, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 12454 + }, + { + "epoch": 0.12455, + "grad_norm": 1.1029696954685784, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12455 + }, + { + "epoch": 0.12456, + "grad_norm": 1.6733393686585805, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12456 + }, + { + "epoch": 0.12457, + "grad_norm": 1.1447453786927126, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 12457 + }, + { + "epoch": 0.12458, + "grad_norm": 0.9986680711717424, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 12458 + }, + { + "epoch": 0.12459, + "grad_norm": 1.1317782308250313, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 12459 + }, + { + "epoch": 0.1246, + "grad_norm": 1.361354226191513, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 12460 + }, + { + "epoch": 0.12461, + "grad_norm": 1.0677093299915452, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 12461 + }, + { + "epoch": 0.12462, + "grad_norm": 1.3928548303478745, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12462 + }, + { + "epoch": 0.12463, + "grad_norm": 0.9641189072985846, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 12463 + }, + { + "epoch": 0.12464, + "grad_norm": 1.0960604844788246, + "learning_rate": 0.003, + "loss": 4.09, + "step": 12464 + }, + { + "epoch": 0.12465, + "grad_norm": 1.1815436300582207, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 12465 + }, + { + "epoch": 0.12466, + "grad_norm": 1.2131619466594814, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 12466 + }, + { + "epoch": 0.12467, + "grad_norm": 1.0331125309199667, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 12467 + }, + { + "epoch": 0.12468, + "grad_norm": 1.163751673352337, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 12468 + }, + { + "epoch": 0.12469, + "grad_norm": 1.0031831435973153, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 12469 + }, + { + "epoch": 0.1247, + "grad_norm": 1.0910642628570784, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 12470 + }, + { + "epoch": 0.12471, + "grad_norm": 1.0918899476724453, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 12471 + }, + { + "epoch": 0.12472, + "grad_norm": 1.1964778879913656, + "learning_rate": 0.003, + "loss": 4.095, + "step": 12472 + }, + { + "epoch": 0.12473, + "grad_norm": 1.0079789042183729, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12473 + }, + { + "epoch": 0.12474, + "grad_norm": 1.386218901058728, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12474 + }, + { + "epoch": 0.12475, + "grad_norm": 0.9267867871365039, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 12475 + }, + { + "epoch": 0.12476, + "grad_norm": 1.2327000804525328, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 12476 + }, + { + "epoch": 0.12477, + "grad_norm": 1.2382234956781983, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12477 + }, + { + "epoch": 0.12478, + "grad_norm": 1.051207711453641, + "learning_rate": 0.003, + "loss": 4.051, + "step": 12478 + }, + { + "epoch": 0.12479, + "grad_norm": 1.4140646728284696, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12479 + }, + { + "epoch": 0.1248, + "grad_norm": 0.9227957116822071, + "learning_rate": 0.003, + "loss": 4.068, + "step": 12480 + }, + { + "epoch": 0.12481, + "grad_norm": 1.0383738395470508, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 12481 + }, + { + "epoch": 0.12482, + "grad_norm": 1.1713914515935469, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 12482 + }, + { + "epoch": 0.12483, + "grad_norm": 1.1476901482525905, + "learning_rate": 0.003, + "loss": 4.095, + "step": 12483 + }, + { + "epoch": 0.12484, + "grad_norm": 1.070386462226285, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 12484 + }, + { + "epoch": 0.12485, + "grad_norm": 1.3221204377348341, + "learning_rate": 0.003, + "loss": 4.102, + "step": 12485 + }, + { + "epoch": 0.12486, + "grad_norm": 0.9644922374383039, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12486 + }, + { + "epoch": 0.12487, + "grad_norm": 1.3121241168194553, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 12487 + }, + { + "epoch": 0.12488, + "grad_norm": 1.3450318680136746, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 12488 + }, + { + "epoch": 0.12489, + "grad_norm": 0.9534508800418893, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 12489 + }, + { + "epoch": 0.1249, + "grad_norm": 1.1599208964245398, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 12490 + }, + { + "epoch": 0.12491, + "grad_norm": 1.147692269571279, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 12491 + }, + { + "epoch": 0.12492, + "grad_norm": 1.1523079863182972, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 12492 + }, + { + "epoch": 0.12493, + "grad_norm": 1.1634536981359016, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 12493 + }, + { + "epoch": 0.12494, + "grad_norm": 0.9874636309717719, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 12494 + }, + { + "epoch": 0.12495, + "grad_norm": 1.1353746209396673, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12495 + }, + { + "epoch": 0.12496, + "grad_norm": 1.172186272262417, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 12496 + }, + { + "epoch": 0.12497, + "grad_norm": 1.1083322572035998, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12497 + }, + { + "epoch": 0.12498, + "grad_norm": 1.1604690432657063, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 12498 + }, + { + "epoch": 0.12499, + "grad_norm": 1.264096108808146, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12499 + }, + { + "epoch": 0.125, + "grad_norm": 1.075946107959381, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 12500 + }, + { + "epoch": 0.12501, + "grad_norm": 1.1792882356955536, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 12501 + }, + { + "epoch": 0.12502, + "grad_norm": 1.0507188272099386, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12502 + }, + { + "epoch": 0.12503, + "grad_norm": 1.19495514255372, + "learning_rate": 0.003, + "loss": 4.078, + "step": 12503 + }, + { + "epoch": 0.12504, + "grad_norm": 1.1288868298706587, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 12504 + }, + { + "epoch": 0.12505, + "grad_norm": 1.249344322822003, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 12505 + }, + { + "epoch": 0.12506, + "grad_norm": 1.08839290817881, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 12506 + }, + { + "epoch": 0.12507, + "grad_norm": 1.2123320378563451, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12507 + }, + { + "epoch": 0.12508, + "grad_norm": 1.0259058058605786, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 12508 + }, + { + "epoch": 0.12509, + "grad_norm": 1.0902172001824963, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 12509 + }, + { + "epoch": 0.1251, + "grad_norm": 1.3051601433884132, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 12510 + }, + { + "epoch": 0.12511, + "grad_norm": 1.0944379703777736, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12511 + }, + { + "epoch": 0.12512, + "grad_norm": 1.3104568659078275, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 12512 + }, + { + "epoch": 0.12513, + "grad_norm": 1.1745329664550175, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 12513 + }, + { + "epoch": 0.12514, + "grad_norm": 0.9551689514624544, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 12514 + }, + { + "epoch": 0.12515, + "grad_norm": 1.0262550517878668, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 12515 + }, + { + "epoch": 0.12516, + "grad_norm": 1.1407767421024977, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 12516 + }, + { + "epoch": 0.12517, + "grad_norm": 0.9962063019603865, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 12517 + }, + { + "epoch": 0.12518, + "grad_norm": 1.1970608765873851, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 12518 + }, + { + "epoch": 0.12519, + "grad_norm": 0.959190386912845, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 12519 + }, + { + "epoch": 0.1252, + "grad_norm": 1.1584921160079846, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 12520 + }, + { + "epoch": 0.12521, + "grad_norm": 1.4550002952801369, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 12521 + }, + { + "epoch": 0.12522, + "grad_norm": 1.1136148399771353, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 12522 + }, + { + "epoch": 0.12523, + "grad_norm": 1.0842256316001873, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 12523 + }, + { + "epoch": 0.12524, + "grad_norm": 1.134675623921096, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 12524 + }, + { + "epoch": 0.12525, + "grad_norm": 1.1593697013463147, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12525 + }, + { + "epoch": 0.12526, + "grad_norm": 1.0788774229439801, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 12526 + }, + { + "epoch": 0.12527, + "grad_norm": 1.2292604042409196, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 12527 + }, + { + "epoch": 0.12528, + "grad_norm": 1.1657288001713335, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 12528 + }, + { + "epoch": 0.12529, + "grad_norm": 1.0254204316259463, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 12529 + }, + { + "epoch": 0.1253, + "grad_norm": 1.1168542385335107, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 12530 + }, + { + "epoch": 0.12531, + "grad_norm": 1.3174987372246656, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 12531 + }, + { + "epoch": 0.12532, + "grad_norm": 1.072377108813044, + "learning_rate": 0.003, + "loss": 4.061, + "step": 12532 + }, + { + "epoch": 0.12533, + "grad_norm": 1.168048971647958, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12533 + }, + { + "epoch": 0.12534, + "grad_norm": 0.9203558918321404, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12534 + }, + { + "epoch": 0.12535, + "grad_norm": 1.0349617881965072, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12535 + }, + { + "epoch": 0.12536, + "grad_norm": 1.2139465679889232, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12536 + }, + { + "epoch": 0.12537, + "grad_norm": 1.0625145068279702, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12537 + }, + { + "epoch": 0.12538, + "grad_norm": 1.1817856187660896, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 12538 + }, + { + "epoch": 0.12539, + "grad_norm": 1.0838819731098341, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 12539 + }, + { + "epoch": 0.1254, + "grad_norm": 1.2713228931051814, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 12540 + }, + { + "epoch": 0.12541, + "grad_norm": 0.959739675663071, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 12541 + }, + { + "epoch": 0.12542, + "grad_norm": 1.0528085424308269, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 12542 + }, + { + "epoch": 0.12543, + "grad_norm": 1.1154825938444732, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12543 + }, + { + "epoch": 0.12544, + "grad_norm": 1.1691545603039764, + "learning_rate": 0.003, + "loss": 4.109, + "step": 12544 + }, + { + "epoch": 0.12545, + "grad_norm": 1.1275427125623834, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 12545 + }, + { + "epoch": 0.12546, + "grad_norm": 1.2910133967954116, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 12546 + }, + { + "epoch": 0.12547, + "grad_norm": 1.2408064762982007, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 12547 + }, + { + "epoch": 0.12548, + "grad_norm": 1.2395442220543138, + "learning_rate": 0.003, + "loss": 4.057, + "step": 12548 + }, + { + "epoch": 0.12549, + "grad_norm": 1.018837692744063, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12549 + }, + { + "epoch": 0.1255, + "grad_norm": 1.2228756800993543, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 12550 + }, + { + "epoch": 0.12551, + "grad_norm": 1.0041785080356689, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 12551 + }, + { + "epoch": 0.12552, + "grad_norm": 1.4947493162533745, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 12552 + }, + { + "epoch": 0.12553, + "grad_norm": 1.1094353181441732, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 12553 + }, + { + "epoch": 0.12554, + "grad_norm": 1.387389588948756, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12554 + }, + { + "epoch": 0.12555, + "grad_norm": 0.8422219554497506, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12555 + }, + { + "epoch": 0.12556, + "grad_norm": 1.0837224019829028, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 12556 + }, + { + "epoch": 0.12557, + "grad_norm": 1.4412117420958377, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 12557 + }, + { + "epoch": 0.12558, + "grad_norm": 0.8155523980416922, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12558 + }, + { + "epoch": 0.12559, + "grad_norm": 1.1309528187622053, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12559 + }, + { + "epoch": 0.1256, + "grad_norm": 1.3344558799240693, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12560 + }, + { + "epoch": 0.12561, + "grad_norm": 1.0022644446588989, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12561 + }, + { + "epoch": 0.12562, + "grad_norm": 1.1951526866976123, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 12562 + }, + { + "epoch": 0.12563, + "grad_norm": 1.184753483762079, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12563 + }, + { + "epoch": 0.12564, + "grad_norm": 1.0892418939395248, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12564 + }, + { + "epoch": 0.12565, + "grad_norm": 1.0306232708896363, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12565 + }, + { + "epoch": 0.12566, + "grad_norm": 1.059762840096475, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 12566 + }, + { + "epoch": 0.12567, + "grad_norm": 1.3303834179457787, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 12567 + }, + { + "epoch": 0.12568, + "grad_norm": 1.0285112629694402, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 12568 + }, + { + "epoch": 0.12569, + "grad_norm": 1.0888315853747654, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12569 + }, + { + "epoch": 0.1257, + "grad_norm": 1.123442724607301, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 12570 + }, + { + "epoch": 0.12571, + "grad_norm": 1.3191518473227668, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 12571 + }, + { + "epoch": 0.12572, + "grad_norm": 0.9436836707054405, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12572 + }, + { + "epoch": 0.12573, + "grad_norm": 1.3818106873341798, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 12573 + }, + { + "epoch": 0.12574, + "grad_norm": 1.335597219687906, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 12574 + }, + { + "epoch": 0.12575, + "grad_norm": 1.0665156220813965, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 12575 + }, + { + "epoch": 0.12576, + "grad_norm": 1.182121329164601, + "learning_rate": 0.003, + "loss": 4.081, + "step": 12576 + }, + { + "epoch": 0.12577, + "grad_norm": 1.1385534378456768, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 12577 + }, + { + "epoch": 0.12578, + "grad_norm": 1.0775480644504942, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12578 + }, + { + "epoch": 0.12579, + "grad_norm": 1.0227873834379646, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12579 + }, + { + "epoch": 0.1258, + "grad_norm": 1.381570570595422, + "learning_rate": 0.003, + "loss": 4.091, + "step": 12580 + }, + { + "epoch": 0.12581, + "grad_norm": 1.1654272026607977, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 12581 + }, + { + "epoch": 0.12582, + "grad_norm": 1.2843945997913482, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12582 + }, + { + "epoch": 0.12583, + "grad_norm": 1.1052825127887618, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 12583 + }, + { + "epoch": 0.12584, + "grad_norm": 1.2547493392225169, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12584 + }, + { + "epoch": 0.12585, + "grad_norm": 1.0500640052491619, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 12585 + }, + { + "epoch": 0.12586, + "grad_norm": 0.9939437347380568, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 12586 + }, + { + "epoch": 0.12587, + "grad_norm": 1.1801032533733091, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 12587 + }, + { + "epoch": 0.12588, + "grad_norm": 1.3091622637077749, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 12588 + }, + { + "epoch": 0.12589, + "grad_norm": 1.3128166278500608, + "learning_rate": 0.003, + "loss": 4.083, + "step": 12589 + }, + { + "epoch": 0.1259, + "grad_norm": 0.9637061554306544, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 12590 + }, + { + "epoch": 0.12591, + "grad_norm": 1.1711703489615166, + "learning_rate": 0.003, + "loss": 4.069, + "step": 12591 + }, + { + "epoch": 0.12592, + "grad_norm": 1.2492792198036187, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 12592 + }, + { + "epoch": 0.12593, + "grad_norm": 1.3356054194268845, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 12593 + }, + { + "epoch": 0.12594, + "grad_norm": 1.01457475403029, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 12594 + }, + { + "epoch": 0.12595, + "grad_norm": 1.2276265939904183, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 12595 + }, + { + "epoch": 0.12596, + "grad_norm": 0.9772130361552345, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 12596 + }, + { + "epoch": 0.12597, + "grad_norm": 1.224323778576316, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 12597 + }, + { + "epoch": 0.12598, + "grad_norm": 1.070538424084639, + "learning_rate": 0.003, + "loss": 4.098, + "step": 12598 + }, + { + "epoch": 0.12599, + "grad_norm": 1.1881950950666054, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 12599 + }, + { + "epoch": 0.126, + "grad_norm": 1.2033695005041851, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 12600 + }, + { + "epoch": 0.12601, + "grad_norm": 1.1206013859890769, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 12601 + }, + { + "epoch": 0.12602, + "grad_norm": 1.2669978901971495, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 12602 + }, + { + "epoch": 0.12603, + "grad_norm": 1.0868147281580245, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12603 + }, + { + "epoch": 0.12604, + "grad_norm": 1.189777558723123, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 12604 + }, + { + "epoch": 0.12605, + "grad_norm": 1.2022435550486126, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12605 + }, + { + "epoch": 0.12606, + "grad_norm": 1.1932237386510585, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 12606 + }, + { + "epoch": 0.12607, + "grad_norm": 1.1021135862889397, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 12607 + }, + { + "epoch": 0.12608, + "grad_norm": 1.1753432677944506, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 12608 + }, + { + "epoch": 0.12609, + "grad_norm": 1.1185593496277637, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 12609 + }, + { + "epoch": 0.1261, + "grad_norm": 1.0519538204701329, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 12610 + }, + { + "epoch": 0.12611, + "grad_norm": 1.035043160747932, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 12611 + }, + { + "epoch": 0.12612, + "grad_norm": 1.46921137359719, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 12612 + }, + { + "epoch": 0.12613, + "grad_norm": 1.0183327242129994, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12613 + }, + { + "epoch": 0.12614, + "grad_norm": 1.1321586476834622, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 12614 + }, + { + "epoch": 0.12615, + "grad_norm": 0.958765388845781, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 12615 + }, + { + "epoch": 0.12616, + "grad_norm": 1.1385791664107119, + "learning_rate": 0.003, + "loss": 4.052, + "step": 12616 + }, + { + "epoch": 0.12617, + "grad_norm": 1.0230861459562044, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 12617 + }, + { + "epoch": 0.12618, + "grad_norm": 1.0869030858149058, + "learning_rate": 0.003, + "loss": 4.059, + "step": 12618 + }, + { + "epoch": 0.12619, + "grad_norm": 1.4998950561962687, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 12619 + }, + { + "epoch": 0.1262, + "grad_norm": 1.0522470907633974, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 12620 + }, + { + "epoch": 0.12621, + "grad_norm": 1.2828423035677585, + "learning_rate": 0.003, + "loss": 4.08, + "step": 12621 + }, + { + "epoch": 0.12622, + "grad_norm": 1.0900994100221286, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12622 + }, + { + "epoch": 0.12623, + "grad_norm": 1.291938364401869, + "learning_rate": 0.003, + "loss": 4.087, + "step": 12623 + }, + { + "epoch": 0.12624, + "grad_norm": 0.9677638397011658, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12624 + }, + { + "epoch": 0.12625, + "grad_norm": 1.1489998438922944, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12625 + }, + { + "epoch": 0.12626, + "grad_norm": 1.2460095756081284, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 12626 + }, + { + "epoch": 0.12627, + "grad_norm": 0.9714846240349928, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12627 + }, + { + "epoch": 0.12628, + "grad_norm": 1.3984461874444771, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12628 + }, + { + "epoch": 0.12629, + "grad_norm": 1.1373515674926453, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 12629 + }, + { + "epoch": 0.1263, + "grad_norm": 1.2225090779305283, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 12630 + }, + { + "epoch": 0.12631, + "grad_norm": 1.1388744446609427, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 12631 + }, + { + "epoch": 0.12632, + "grad_norm": 1.2734851291663085, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 12632 + }, + { + "epoch": 0.12633, + "grad_norm": 0.9201830905127847, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 12633 + }, + { + "epoch": 0.12634, + "grad_norm": 1.0922115175002909, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12634 + }, + { + "epoch": 0.12635, + "grad_norm": 1.145360232044216, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 12635 + }, + { + "epoch": 0.12636, + "grad_norm": 1.1862988393375187, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 12636 + }, + { + "epoch": 0.12637, + "grad_norm": 1.0348188987624127, + "learning_rate": 0.003, + "loss": 4.08, + "step": 12637 + }, + { + "epoch": 0.12638, + "grad_norm": 1.1923097559505276, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12638 + }, + { + "epoch": 0.12639, + "grad_norm": 1.0703531305139602, + "learning_rate": 0.003, + "loss": 4.088, + "step": 12639 + }, + { + "epoch": 0.1264, + "grad_norm": 1.2905049113417022, + "learning_rate": 0.003, + "loss": 4.077, + "step": 12640 + }, + { + "epoch": 0.12641, + "grad_norm": 1.1902293742550116, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 12641 + }, + { + "epoch": 0.12642, + "grad_norm": 1.2410972077837368, + "learning_rate": 0.003, + "loss": 4.07, + "step": 12642 + }, + { + "epoch": 0.12643, + "grad_norm": 1.1406148977882309, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 12643 + }, + { + "epoch": 0.12644, + "grad_norm": 1.06509446797809, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 12644 + }, + { + "epoch": 0.12645, + "grad_norm": 1.0842259709739674, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 12645 + }, + { + "epoch": 0.12646, + "grad_norm": 1.2586305223074954, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 12646 + }, + { + "epoch": 0.12647, + "grad_norm": 1.275877459061793, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 12647 + }, + { + "epoch": 0.12648, + "grad_norm": 1.0925411969812069, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 12648 + }, + { + "epoch": 0.12649, + "grad_norm": 0.998476164112096, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 12649 + }, + { + "epoch": 0.1265, + "grad_norm": 1.0132017061136347, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 12650 + }, + { + "epoch": 0.12651, + "grad_norm": 1.1048870775158017, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12651 + }, + { + "epoch": 0.12652, + "grad_norm": 1.0423138819603341, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 12652 + }, + { + "epoch": 0.12653, + "grad_norm": 1.391018502997299, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 12653 + }, + { + "epoch": 0.12654, + "grad_norm": 1.140870115386, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 12654 + }, + { + "epoch": 0.12655, + "grad_norm": 1.426313352021284, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12655 + }, + { + "epoch": 0.12656, + "grad_norm": 1.0375477670940312, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 12656 + }, + { + "epoch": 0.12657, + "grad_norm": 1.145041205013688, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 12657 + }, + { + "epoch": 0.12658, + "grad_norm": 1.0887290250882993, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 12658 + }, + { + "epoch": 0.12659, + "grad_norm": 1.1273348516279116, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 12659 + }, + { + "epoch": 0.1266, + "grad_norm": 1.2027122471428053, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 12660 + }, + { + "epoch": 0.12661, + "grad_norm": 1.0416462226946577, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 12661 + }, + { + "epoch": 0.12662, + "grad_norm": 1.2468210679993077, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12662 + }, + { + "epoch": 0.12663, + "grad_norm": 1.0745200887126218, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 12663 + }, + { + "epoch": 0.12664, + "grad_norm": 1.1278433882285062, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12664 + }, + { + "epoch": 0.12665, + "grad_norm": 1.0909284791055842, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12665 + }, + { + "epoch": 0.12666, + "grad_norm": 1.133751811039444, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12666 + }, + { + "epoch": 0.12667, + "grad_norm": 1.1802982991607618, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 12667 + }, + { + "epoch": 0.12668, + "grad_norm": 1.1448431857062022, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 12668 + }, + { + "epoch": 0.12669, + "grad_norm": 1.0836278876953978, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 12669 + }, + { + "epoch": 0.1267, + "grad_norm": 1.3095005531198924, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 12670 + }, + { + "epoch": 0.12671, + "grad_norm": 1.1863330443749722, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 12671 + }, + { + "epoch": 0.12672, + "grad_norm": 1.3414469570535603, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 12672 + }, + { + "epoch": 0.12673, + "grad_norm": 0.9841922309060664, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 12673 + }, + { + "epoch": 0.12674, + "grad_norm": 1.1120261446531505, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12674 + }, + { + "epoch": 0.12675, + "grad_norm": 1.0960420521883425, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 12675 + }, + { + "epoch": 0.12676, + "grad_norm": 1.0953589001224968, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 12676 + }, + { + "epoch": 0.12677, + "grad_norm": 1.2852366440837981, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 12677 + }, + { + "epoch": 0.12678, + "grad_norm": 1.2366527749878387, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 12678 + }, + { + "epoch": 0.12679, + "grad_norm": 1.204052310910075, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12679 + }, + { + "epoch": 0.1268, + "grad_norm": 0.9691879137113483, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 12680 + }, + { + "epoch": 0.12681, + "grad_norm": 1.2505083596403377, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 12681 + }, + { + "epoch": 0.12682, + "grad_norm": 1.1328418494563997, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 12682 + }, + { + "epoch": 0.12683, + "grad_norm": 1.1689224771465725, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 12683 + }, + { + "epoch": 0.12684, + "grad_norm": 0.9215889407945064, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 12684 + }, + { + "epoch": 0.12685, + "grad_norm": 1.0174059550673156, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 12685 + }, + { + "epoch": 0.12686, + "grad_norm": 1.0789107426790325, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12686 + }, + { + "epoch": 0.12687, + "grad_norm": 1.138943276149499, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12687 + }, + { + "epoch": 0.12688, + "grad_norm": 1.3491267848129291, + "learning_rate": 0.003, + "loss": 4.073, + "step": 12688 + }, + { + "epoch": 0.12689, + "grad_norm": 1.3043096443469937, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 12689 + }, + { + "epoch": 0.1269, + "grad_norm": 1.041888803682189, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 12690 + }, + { + "epoch": 0.12691, + "grad_norm": 1.1928039104322192, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 12691 + }, + { + "epoch": 0.12692, + "grad_norm": 0.9677463033613477, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 12692 + }, + { + "epoch": 0.12693, + "grad_norm": 1.297628672717164, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 12693 + }, + { + "epoch": 0.12694, + "grad_norm": 1.1590402253727634, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 12694 + }, + { + "epoch": 0.12695, + "grad_norm": 1.0929268156695968, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12695 + }, + { + "epoch": 0.12696, + "grad_norm": 1.1738269618767545, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 12696 + }, + { + "epoch": 0.12697, + "grad_norm": 1.0869019305907497, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 12697 + }, + { + "epoch": 0.12698, + "grad_norm": 1.0608902265327735, + "learning_rate": 0.003, + "loss": 4.048, + "step": 12698 + }, + { + "epoch": 0.12699, + "grad_norm": 1.4341001860176978, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 12699 + }, + { + "epoch": 0.127, + "grad_norm": 0.9813981853354528, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12700 + }, + { + "epoch": 0.12701, + "grad_norm": 1.231682032088662, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 12701 + }, + { + "epoch": 0.12702, + "grad_norm": 1.0549722177713285, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 12702 + }, + { + "epoch": 0.12703, + "grad_norm": 1.158178022022144, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 12703 + }, + { + "epoch": 0.12704, + "grad_norm": 0.9903579204372447, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12704 + }, + { + "epoch": 0.12705, + "grad_norm": 1.2901104187242485, + "learning_rate": 0.003, + "loss": 4.083, + "step": 12705 + }, + { + "epoch": 0.12706, + "grad_norm": 1.2537091170376986, + "learning_rate": 0.003, + "loss": 4.066, + "step": 12706 + }, + { + "epoch": 0.12707, + "grad_norm": 1.234455267795835, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 12707 + }, + { + "epoch": 0.12708, + "grad_norm": 1.1892730196421701, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 12708 + }, + { + "epoch": 0.12709, + "grad_norm": 1.098330523777582, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 12709 + }, + { + "epoch": 0.1271, + "grad_norm": 1.3091469707490966, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12710 + }, + { + "epoch": 0.12711, + "grad_norm": 0.8126637106244213, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 12711 + }, + { + "epoch": 0.12712, + "grad_norm": 0.9937269329593112, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12712 + }, + { + "epoch": 0.12713, + "grad_norm": 1.4614438638500475, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 12713 + }, + { + "epoch": 0.12714, + "grad_norm": 1.0311941615428375, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 12714 + }, + { + "epoch": 0.12715, + "grad_norm": 1.368311455512496, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 12715 + }, + { + "epoch": 0.12716, + "grad_norm": 1.1194355453078446, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 12716 + }, + { + "epoch": 0.12717, + "grad_norm": 1.5343174739356966, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 12717 + }, + { + "epoch": 0.12718, + "grad_norm": 0.9189991945381788, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 12718 + }, + { + "epoch": 0.12719, + "grad_norm": 0.9429601349139406, + "learning_rate": 0.003, + "loss": 4.089, + "step": 12719 + }, + { + "epoch": 0.1272, + "grad_norm": 1.2671746093809337, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 12720 + }, + { + "epoch": 0.12721, + "grad_norm": 1.085669009766264, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12721 + }, + { + "epoch": 0.12722, + "grad_norm": 1.2856395614673346, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12722 + }, + { + "epoch": 0.12723, + "grad_norm": 1.1826019556959075, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12723 + }, + { + "epoch": 0.12724, + "grad_norm": 1.2137350740175774, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12724 + }, + { + "epoch": 0.12725, + "grad_norm": 0.9714405915883685, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12725 + }, + { + "epoch": 0.12726, + "grad_norm": 1.233171570676695, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12726 + }, + { + "epoch": 0.12727, + "grad_norm": 0.9854681450136724, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 12727 + }, + { + "epoch": 0.12728, + "grad_norm": 1.0247526093155872, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 12728 + }, + { + "epoch": 0.12729, + "grad_norm": 1.2216031150832127, + "learning_rate": 0.003, + "loss": 4.087, + "step": 12729 + }, + { + "epoch": 0.1273, + "grad_norm": 1.0900986705850872, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12730 + }, + { + "epoch": 0.12731, + "grad_norm": 1.3435657482823378, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 12731 + }, + { + "epoch": 0.12732, + "grad_norm": 1.0939415152116774, + "learning_rate": 0.003, + "loss": 4.053, + "step": 12732 + }, + { + "epoch": 0.12733, + "grad_norm": 1.1572450146336237, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 12733 + }, + { + "epoch": 0.12734, + "grad_norm": 1.2666318970790864, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 12734 + }, + { + "epoch": 0.12735, + "grad_norm": 1.0082757442143788, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12735 + }, + { + "epoch": 0.12736, + "grad_norm": 1.255147432277124, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 12736 + }, + { + "epoch": 0.12737, + "grad_norm": 1.1279108547109393, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 12737 + }, + { + "epoch": 0.12738, + "grad_norm": 1.3623171747528882, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 12738 + }, + { + "epoch": 0.12739, + "grad_norm": 1.1804668727447836, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 12739 + }, + { + "epoch": 0.1274, + "grad_norm": 1.2903582439030041, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 12740 + }, + { + "epoch": 0.12741, + "grad_norm": 1.1903708722838362, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 12741 + }, + { + "epoch": 0.12742, + "grad_norm": 0.9970718101508081, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 12742 + }, + { + "epoch": 0.12743, + "grad_norm": 1.228943373939233, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 12743 + }, + { + "epoch": 0.12744, + "grad_norm": 1.2865166226873788, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 12744 + }, + { + "epoch": 0.12745, + "grad_norm": 1.0020118590142613, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12745 + }, + { + "epoch": 0.12746, + "grad_norm": 1.3251412767916488, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 12746 + }, + { + "epoch": 0.12747, + "grad_norm": 1.092232438065853, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 12747 + }, + { + "epoch": 0.12748, + "grad_norm": 1.214629485605157, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 12748 + }, + { + "epoch": 0.12749, + "grad_norm": 1.1856415955145982, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12749 + }, + { + "epoch": 0.1275, + "grad_norm": 1.0913325877513864, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12750 + }, + { + "epoch": 0.12751, + "grad_norm": 1.4636934393582455, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 12751 + }, + { + "epoch": 0.12752, + "grad_norm": 1.0680594573992184, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 12752 + }, + { + "epoch": 0.12753, + "grad_norm": 1.1591841480104785, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 12753 + }, + { + "epoch": 0.12754, + "grad_norm": 1.3915949998580592, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 12754 + }, + { + "epoch": 0.12755, + "grad_norm": 1.100090177526697, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12755 + }, + { + "epoch": 0.12756, + "grad_norm": 1.326466604650826, + "learning_rate": 0.003, + "loss": 4.09, + "step": 12756 + }, + { + "epoch": 0.12757, + "grad_norm": 0.8779866180952189, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 12757 + }, + { + "epoch": 0.12758, + "grad_norm": 0.9123535258907629, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12758 + }, + { + "epoch": 0.12759, + "grad_norm": 1.196160302646205, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 12759 + }, + { + "epoch": 0.1276, + "grad_norm": 1.1557465866056045, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 12760 + }, + { + "epoch": 0.12761, + "grad_norm": 1.2354914565937516, + "learning_rate": 0.003, + "loss": 4.073, + "step": 12761 + }, + { + "epoch": 0.12762, + "grad_norm": 1.1764145362664384, + "learning_rate": 0.003, + "loss": 4.042, + "step": 12762 + }, + { + "epoch": 0.12763, + "grad_norm": 1.3445066594936244, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12763 + }, + { + "epoch": 0.12764, + "grad_norm": 0.9739010566952191, + "learning_rate": 0.003, + "loss": 4.093, + "step": 12764 + }, + { + "epoch": 0.12765, + "grad_norm": 1.0613937411936154, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12765 + }, + { + "epoch": 0.12766, + "grad_norm": 1.1363450780655793, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 12766 + }, + { + "epoch": 0.12767, + "grad_norm": 1.2576325751489537, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 12767 + }, + { + "epoch": 0.12768, + "grad_norm": 1.1040972214642275, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 12768 + }, + { + "epoch": 0.12769, + "grad_norm": 1.2165327989288754, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12769 + }, + { + "epoch": 0.1277, + "grad_norm": 0.9812695523482138, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12770 + }, + { + "epoch": 0.12771, + "grad_norm": 1.0937863210045036, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12771 + }, + { + "epoch": 0.12772, + "grad_norm": 1.1843615131462144, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12772 + }, + { + "epoch": 0.12773, + "grad_norm": 1.2602559036345033, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 12773 + }, + { + "epoch": 0.12774, + "grad_norm": 1.0963906426701393, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 12774 + }, + { + "epoch": 0.12775, + "grad_norm": 1.1176614549896196, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 12775 + }, + { + "epoch": 0.12776, + "grad_norm": 1.151217603095789, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 12776 + }, + { + "epoch": 0.12777, + "grad_norm": 1.238071600314445, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 12777 + }, + { + "epoch": 0.12778, + "grad_norm": 1.1705546677287786, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 12778 + }, + { + "epoch": 0.12779, + "grad_norm": 1.1871346099841646, + "learning_rate": 0.003, + "loss": 4.093, + "step": 12779 + }, + { + "epoch": 0.1278, + "grad_norm": 1.0061359267648406, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 12780 + }, + { + "epoch": 0.12781, + "grad_norm": 1.208483353168107, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12781 + }, + { + "epoch": 0.12782, + "grad_norm": 0.9963653813126722, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 12782 + }, + { + "epoch": 0.12783, + "grad_norm": 1.3875762642705003, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 12783 + }, + { + "epoch": 0.12784, + "grad_norm": 1.0252567643115786, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 12784 + }, + { + "epoch": 0.12785, + "grad_norm": 1.0876953323201521, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 12785 + }, + { + "epoch": 0.12786, + "grad_norm": 1.1023544017582871, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 12786 + }, + { + "epoch": 0.12787, + "grad_norm": 1.2189812506606332, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 12787 + }, + { + "epoch": 0.12788, + "grad_norm": 1.1337557681724755, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 12788 + }, + { + "epoch": 0.12789, + "grad_norm": 1.1118245184989517, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 12789 + }, + { + "epoch": 0.1279, + "grad_norm": 1.1495692894460081, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 12790 + }, + { + "epoch": 0.12791, + "grad_norm": 1.4006130182728584, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12791 + }, + { + "epoch": 0.12792, + "grad_norm": 1.149581584270721, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 12792 + }, + { + "epoch": 0.12793, + "grad_norm": 1.2295938062035259, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12793 + }, + { + "epoch": 0.12794, + "grad_norm": 0.9870563587036222, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 12794 + }, + { + "epoch": 0.12795, + "grad_norm": 1.3729576291106589, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12795 + }, + { + "epoch": 0.12796, + "grad_norm": 1.097502274613693, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 12796 + }, + { + "epoch": 0.12797, + "grad_norm": 1.2164641487977086, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 12797 + }, + { + "epoch": 0.12798, + "grad_norm": 0.9578732789955301, + "learning_rate": 0.003, + "loss": 4.048, + "step": 12798 + }, + { + "epoch": 0.12799, + "grad_norm": 1.0836746167399414, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 12799 + }, + { + "epoch": 0.128, + "grad_norm": 1.2216085200393374, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 12800 + }, + { + "epoch": 0.12801, + "grad_norm": 1.2345579740967534, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12801 + }, + { + "epoch": 0.12802, + "grad_norm": 1.2474953417169945, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 12802 + }, + { + "epoch": 0.12803, + "grad_norm": 1.0831676668779615, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12803 + }, + { + "epoch": 0.12804, + "grad_norm": 1.193916510472046, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12804 + }, + { + "epoch": 0.12805, + "grad_norm": 1.1035531417793012, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12805 + }, + { + "epoch": 0.12806, + "grad_norm": 1.2292925640284165, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 12806 + }, + { + "epoch": 0.12807, + "grad_norm": 0.9439588937890558, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 12807 + }, + { + "epoch": 0.12808, + "grad_norm": 1.335736102020312, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 12808 + }, + { + "epoch": 0.12809, + "grad_norm": 0.9787030435226739, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12809 + }, + { + "epoch": 0.1281, + "grad_norm": 1.3174519343980273, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 12810 + }, + { + "epoch": 0.12811, + "grad_norm": 1.2254434525804792, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 12811 + }, + { + "epoch": 0.12812, + "grad_norm": 1.210361715843662, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12812 + }, + { + "epoch": 0.12813, + "grad_norm": 1.1971184137498239, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 12813 + }, + { + "epoch": 0.12814, + "grad_norm": 1.082703240756199, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 12814 + }, + { + "epoch": 0.12815, + "grad_norm": 1.1930835496849195, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12815 + }, + { + "epoch": 0.12816, + "grad_norm": 1.0318640828326053, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 12816 + }, + { + "epoch": 0.12817, + "grad_norm": 1.3827741801067235, + "learning_rate": 0.003, + "loss": 4.079, + "step": 12817 + }, + { + "epoch": 0.12818, + "grad_norm": 0.8675507749901307, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 12818 + }, + { + "epoch": 0.12819, + "grad_norm": 1.122555691577188, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 12819 + }, + { + "epoch": 0.1282, + "grad_norm": 1.4578571399267979, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 12820 + }, + { + "epoch": 0.12821, + "grad_norm": 1.2230131071711259, + "learning_rate": 0.003, + "loss": 4.075, + "step": 12821 + }, + { + "epoch": 0.12822, + "grad_norm": 1.1470390517424534, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 12822 + }, + { + "epoch": 0.12823, + "grad_norm": 1.21824047768869, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 12823 + }, + { + "epoch": 0.12824, + "grad_norm": 1.1595940248211523, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12824 + }, + { + "epoch": 0.12825, + "grad_norm": 1.0071679971286354, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 12825 + }, + { + "epoch": 0.12826, + "grad_norm": 1.0886480802636633, + "learning_rate": 0.003, + "loss": 4.08, + "step": 12826 + }, + { + "epoch": 0.12827, + "grad_norm": 0.9879075812230554, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12827 + }, + { + "epoch": 0.12828, + "grad_norm": 1.25709114220997, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 12828 + }, + { + "epoch": 0.12829, + "grad_norm": 1.1897769499680302, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12829 + }, + { + "epoch": 0.1283, + "grad_norm": 1.2009363350007225, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 12830 + }, + { + "epoch": 0.12831, + "grad_norm": 1.0437214471375973, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 12831 + }, + { + "epoch": 0.12832, + "grad_norm": 1.3698860559323622, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12832 + }, + { + "epoch": 0.12833, + "grad_norm": 1.0702030003794927, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 12833 + }, + { + "epoch": 0.12834, + "grad_norm": 1.200090242630135, + "learning_rate": 0.003, + "loss": 4.067, + "step": 12834 + }, + { + "epoch": 0.12835, + "grad_norm": 1.2746786649376214, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 12835 + }, + { + "epoch": 0.12836, + "grad_norm": 0.9805497035425383, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 12836 + }, + { + "epoch": 0.12837, + "grad_norm": 1.1633036249671689, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 12837 + }, + { + "epoch": 0.12838, + "grad_norm": 1.329736020879512, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 12838 + }, + { + "epoch": 0.12839, + "grad_norm": 1.1938318223084425, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12839 + }, + { + "epoch": 0.1284, + "grad_norm": 1.120612214095928, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 12840 + }, + { + "epoch": 0.12841, + "grad_norm": 1.127307392558783, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 12841 + }, + { + "epoch": 0.12842, + "grad_norm": 1.3731566587701174, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 12842 + }, + { + "epoch": 0.12843, + "grad_norm": 1.0708552732989607, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 12843 + }, + { + "epoch": 0.12844, + "grad_norm": 1.0452607451544866, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 12844 + }, + { + "epoch": 0.12845, + "grad_norm": 1.299441269405997, + "learning_rate": 0.003, + "loss": 4.062, + "step": 12845 + }, + { + "epoch": 0.12846, + "grad_norm": 1.0305420303486308, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 12846 + }, + { + "epoch": 0.12847, + "grad_norm": 1.157120715208667, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12847 + }, + { + "epoch": 0.12848, + "grad_norm": 0.951748874423968, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 12848 + }, + { + "epoch": 0.12849, + "grad_norm": 1.2186662066141403, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12849 + }, + { + "epoch": 0.1285, + "grad_norm": 1.1616550894309363, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 12850 + }, + { + "epoch": 0.12851, + "grad_norm": 1.3611813068636822, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 12851 + }, + { + "epoch": 0.12852, + "grad_norm": 1.0525954515376, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 12852 + }, + { + "epoch": 0.12853, + "grad_norm": 1.3130415063478154, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12853 + }, + { + "epoch": 0.12854, + "grad_norm": 0.8970798116397892, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 12854 + }, + { + "epoch": 0.12855, + "grad_norm": 1.1852927993778317, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12855 + }, + { + "epoch": 0.12856, + "grad_norm": 1.1667596413715686, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12856 + }, + { + "epoch": 0.12857, + "grad_norm": 1.134386177464642, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 12857 + }, + { + "epoch": 0.12858, + "grad_norm": 1.3079045407666592, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 12858 + }, + { + "epoch": 0.12859, + "grad_norm": 0.8763922150653155, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 12859 + }, + { + "epoch": 0.1286, + "grad_norm": 1.011595626039733, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 12860 + }, + { + "epoch": 0.12861, + "grad_norm": 1.4613258416334778, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 12861 + }, + { + "epoch": 0.12862, + "grad_norm": 1.1180529142493827, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 12862 + }, + { + "epoch": 0.12863, + "grad_norm": 1.0748030892965381, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 12863 + }, + { + "epoch": 0.12864, + "grad_norm": 1.2663882454890067, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 12864 + }, + { + "epoch": 0.12865, + "grad_norm": 1.0771260362783301, + "learning_rate": 0.003, + "loss": 4.063, + "step": 12865 + }, + { + "epoch": 0.12866, + "grad_norm": 1.207007397316719, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12866 + }, + { + "epoch": 0.12867, + "grad_norm": 1.251880824085332, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 12867 + }, + { + "epoch": 0.12868, + "grad_norm": 1.0913493230948164, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12868 + }, + { + "epoch": 0.12869, + "grad_norm": 1.3775891247153929, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12869 + }, + { + "epoch": 0.1287, + "grad_norm": 1.2178575125106768, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 12870 + }, + { + "epoch": 0.12871, + "grad_norm": 1.0777722954407951, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 12871 + }, + { + "epoch": 0.12872, + "grad_norm": 1.2195294753087516, + "learning_rate": 0.003, + "loss": 4.068, + "step": 12872 + }, + { + "epoch": 0.12873, + "grad_norm": 1.0988131540261645, + "learning_rate": 0.003, + "loss": 4.072, + "step": 12873 + }, + { + "epoch": 0.12874, + "grad_norm": 1.1142189015320123, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 12874 + }, + { + "epoch": 0.12875, + "grad_norm": 1.3077766521456173, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 12875 + }, + { + "epoch": 0.12876, + "grad_norm": 1.1565760435368453, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 12876 + }, + { + "epoch": 0.12877, + "grad_norm": 1.6149809077401527, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 12877 + }, + { + "epoch": 0.12878, + "grad_norm": 0.8107627956085056, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 12878 + }, + { + "epoch": 0.12879, + "grad_norm": 0.855341741798638, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12879 + }, + { + "epoch": 0.1288, + "grad_norm": 1.0564515371668957, + "learning_rate": 0.003, + "loss": 4.057, + "step": 12880 + }, + { + "epoch": 0.12881, + "grad_norm": 1.17170624905697, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 12881 + }, + { + "epoch": 0.12882, + "grad_norm": 1.0267246945471074, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12882 + }, + { + "epoch": 0.12883, + "grad_norm": 1.3556623878864673, + "learning_rate": 0.003, + "loss": 4.088, + "step": 12883 + }, + { + "epoch": 0.12884, + "grad_norm": 1.0475880144786536, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 12884 + }, + { + "epoch": 0.12885, + "grad_norm": 1.5063259284449038, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12885 + }, + { + "epoch": 0.12886, + "grad_norm": 0.7899265655167541, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 12886 + }, + { + "epoch": 0.12887, + "grad_norm": 0.9058564323079618, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 12887 + }, + { + "epoch": 0.12888, + "grad_norm": 1.005475241367468, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12888 + }, + { + "epoch": 0.12889, + "grad_norm": 1.2714003535983294, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 12889 + }, + { + "epoch": 0.1289, + "grad_norm": 0.9377231923769617, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 12890 + }, + { + "epoch": 0.12891, + "grad_norm": 1.2929609564956943, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 12891 + }, + { + "epoch": 0.12892, + "grad_norm": 1.1241978401123458, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 12892 + }, + { + "epoch": 0.12893, + "grad_norm": 1.4405716398513726, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 12893 + }, + { + "epoch": 0.12894, + "grad_norm": 1.159914398922788, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 12894 + }, + { + "epoch": 0.12895, + "grad_norm": 1.1015643555910721, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 12895 + }, + { + "epoch": 0.12896, + "grad_norm": 1.2576780878173859, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12896 + }, + { + "epoch": 0.12897, + "grad_norm": 1.0783312786699863, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12897 + }, + { + "epoch": 0.12898, + "grad_norm": 1.076447530213746, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 12898 + }, + { + "epoch": 0.12899, + "grad_norm": 1.2113329644664113, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 12899 + }, + { + "epoch": 0.129, + "grad_norm": 1.4884997349144662, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 12900 + }, + { + "epoch": 0.12901, + "grad_norm": 1.055348011982021, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 12901 + }, + { + "epoch": 0.12902, + "grad_norm": 1.2716009511343596, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 12902 + }, + { + "epoch": 0.12903, + "grad_norm": 1.310704478906888, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 12903 + }, + { + "epoch": 0.12904, + "grad_norm": 1.0034390547908698, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 12904 + }, + { + "epoch": 0.12905, + "grad_norm": 1.2912107876545866, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 12905 + }, + { + "epoch": 0.12906, + "grad_norm": 0.9220033689146221, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 12906 + }, + { + "epoch": 0.12907, + "grad_norm": 1.2545148430744932, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 12907 + }, + { + "epoch": 0.12908, + "grad_norm": 1.061885091173034, + "learning_rate": 0.003, + "loss": 4.044, + "step": 12908 + }, + { + "epoch": 0.12909, + "grad_norm": 1.1981459208583543, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 12909 + }, + { + "epoch": 0.1291, + "grad_norm": 1.4887508899796054, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 12910 + }, + { + "epoch": 0.12911, + "grad_norm": 1.0484698691812218, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 12911 + }, + { + "epoch": 0.12912, + "grad_norm": 1.1974356627595546, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12912 + }, + { + "epoch": 0.12913, + "grad_norm": 1.1544594647944466, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 12913 + }, + { + "epoch": 0.12914, + "grad_norm": 1.06301186858534, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 12914 + }, + { + "epoch": 0.12915, + "grad_norm": 1.103201954615549, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 12915 + }, + { + "epoch": 0.12916, + "grad_norm": 1.1225680595769518, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 12916 + }, + { + "epoch": 0.12917, + "grad_norm": 1.216988199926607, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12917 + }, + { + "epoch": 0.12918, + "grad_norm": 1.0828569235398053, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 12918 + }, + { + "epoch": 0.12919, + "grad_norm": 1.078791826064448, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 12919 + }, + { + "epoch": 0.1292, + "grad_norm": 1.1184998086032465, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12920 + }, + { + "epoch": 0.12921, + "grad_norm": 1.2371745617415741, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 12921 + }, + { + "epoch": 0.12922, + "grad_norm": 1.231388001618202, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 12922 + }, + { + "epoch": 0.12923, + "grad_norm": 0.9319342231989519, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 12923 + }, + { + "epoch": 0.12924, + "grad_norm": 1.0626351285022981, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 12924 + }, + { + "epoch": 0.12925, + "grad_norm": 1.294675870585951, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 12925 + }, + { + "epoch": 0.12926, + "grad_norm": 1.0447166513830999, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 12926 + }, + { + "epoch": 0.12927, + "grad_norm": 1.3468163322260938, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12927 + }, + { + "epoch": 0.12928, + "grad_norm": 1.1187109002024622, + "learning_rate": 0.003, + "loss": 4.084, + "step": 12928 + }, + { + "epoch": 0.12929, + "grad_norm": 1.341429626291632, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 12929 + }, + { + "epoch": 0.1293, + "grad_norm": 1.1945670634171135, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 12930 + }, + { + "epoch": 0.12931, + "grad_norm": 1.2265556660339738, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 12931 + }, + { + "epoch": 0.12932, + "grad_norm": 1.0543481447222696, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 12932 + }, + { + "epoch": 0.12933, + "grad_norm": 1.2250952993934159, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 12933 + }, + { + "epoch": 0.12934, + "grad_norm": 1.0076811874522589, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 12934 + }, + { + "epoch": 0.12935, + "grad_norm": 1.212985332533582, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 12935 + }, + { + "epoch": 0.12936, + "grad_norm": 0.9350956920809937, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 12936 + }, + { + "epoch": 0.12937, + "grad_norm": 0.9610230910043541, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12937 + }, + { + "epoch": 0.12938, + "grad_norm": 1.233023364290806, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 12938 + }, + { + "epoch": 0.12939, + "grad_norm": 1.2941221452477212, + "learning_rate": 0.003, + "loss": 4.06, + "step": 12939 + }, + { + "epoch": 0.1294, + "grad_norm": 1.011201607814211, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 12940 + }, + { + "epoch": 0.12941, + "grad_norm": 1.151927786045976, + "learning_rate": 0.003, + "loss": 4.074, + "step": 12941 + }, + { + "epoch": 0.12942, + "grad_norm": 1.0484359269577173, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 12942 + }, + { + "epoch": 0.12943, + "grad_norm": 1.084312206261362, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 12943 + }, + { + "epoch": 0.12944, + "grad_norm": 1.2728509429992616, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 12944 + }, + { + "epoch": 0.12945, + "grad_norm": 1.0628719145322947, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12945 + }, + { + "epoch": 0.12946, + "grad_norm": 1.1191828860796316, + "learning_rate": 0.003, + "loss": 4.096, + "step": 12946 + }, + { + "epoch": 0.12947, + "grad_norm": 1.104488359673685, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 12947 + }, + { + "epoch": 0.12948, + "grad_norm": 1.054862388026599, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 12948 + }, + { + "epoch": 0.12949, + "grad_norm": 1.1190613173273303, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 12949 + }, + { + "epoch": 0.1295, + "grad_norm": 1.2062057801082409, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12950 + }, + { + "epoch": 0.12951, + "grad_norm": 1.2965595199159816, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 12951 + }, + { + "epoch": 0.12952, + "grad_norm": 1.3515064007048385, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 12952 + }, + { + "epoch": 0.12953, + "grad_norm": 1.1542883772874049, + "learning_rate": 0.003, + "loss": 4.088, + "step": 12953 + }, + { + "epoch": 0.12954, + "grad_norm": 1.1746570847636975, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 12954 + }, + { + "epoch": 0.12955, + "grad_norm": 1.2510953695065115, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 12955 + }, + { + "epoch": 0.12956, + "grad_norm": 1.06297432928455, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 12956 + }, + { + "epoch": 0.12957, + "grad_norm": 1.1473066198430277, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 12957 + }, + { + "epoch": 0.12958, + "grad_norm": 1.0201529204218371, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 12958 + }, + { + "epoch": 0.12959, + "grad_norm": 1.0898537437420681, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 12959 + }, + { + "epoch": 0.1296, + "grad_norm": 1.1055293846960481, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 12960 + }, + { + "epoch": 0.12961, + "grad_norm": 1.2351085707954883, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 12961 + }, + { + "epoch": 0.12962, + "grad_norm": 1.1854472228284336, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 12962 + }, + { + "epoch": 0.12963, + "grad_norm": 1.227653172728379, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 12963 + }, + { + "epoch": 0.12964, + "grad_norm": 1.2789582584621662, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 12964 + }, + { + "epoch": 0.12965, + "grad_norm": 1.2087528078557306, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12965 + }, + { + "epoch": 0.12966, + "grad_norm": 1.6112942427002772, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 12966 + }, + { + "epoch": 0.12967, + "grad_norm": 0.8976210470667042, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12967 + }, + { + "epoch": 0.12968, + "grad_norm": 0.8608094199061171, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12968 + }, + { + "epoch": 0.12969, + "grad_norm": 1.1370110839187368, + "learning_rate": 0.003, + "loss": 4.063, + "step": 12969 + }, + { + "epoch": 0.1297, + "grad_norm": 1.4481956638153832, + "learning_rate": 0.003, + "loss": 4.05, + "step": 12970 + }, + { + "epoch": 0.12971, + "grad_norm": 1.2543473501113689, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 12971 + }, + { + "epoch": 0.12972, + "grad_norm": 1.0145533995976639, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12972 + }, + { + "epoch": 0.12973, + "grad_norm": 1.1360402979394963, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 12973 + }, + { + "epoch": 0.12974, + "grad_norm": 1.1046758374769596, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 12974 + }, + { + "epoch": 0.12975, + "grad_norm": 1.2369109496401294, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 12975 + }, + { + "epoch": 0.12976, + "grad_norm": 0.9704420219201002, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 12976 + }, + { + "epoch": 0.12977, + "grad_norm": 1.1921286204207069, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 12977 + }, + { + "epoch": 0.12978, + "grad_norm": 0.8809197230731804, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 12978 + }, + { + "epoch": 0.12979, + "grad_norm": 1.1979650800023425, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 12979 + }, + { + "epoch": 0.1298, + "grad_norm": 1.1362702093641754, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 12980 + }, + { + "epoch": 0.12981, + "grad_norm": 1.2225938830516137, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 12981 + }, + { + "epoch": 0.12982, + "grad_norm": 1.3885255723407812, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 12982 + }, + { + "epoch": 0.12983, + "grad_norm": 1.172983719413079, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 12983 + }, + { + "epoch": 0.12984, + "grad_norm": 1.1540178894424364, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 12984 + }, + { + "epoch": 0.12985, + "grad_norm": 1.0860372470568143, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 12985 + }, + { + "epoch": 0.12986, + "grad_norm": 1.1949009167059308, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 12986 + }, + { + "epoch": 0.12987, + "grad_norm": 1.1886249054151257, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 12987 + }, + { + "epoch": 0.12988, + "grad_norm": 1.0545386379249677, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12988 + }, + { + "epoch": 0.12989, + "grad_norm": 1.4904358019801487, + "learning_rate": 0.003, + "loss": 4.076, + "step": 12989 + }, + { + "epoch": 0.1299, + "grad_norm": 0.9820795206052854, + "learning_rate": 0.003, + "loss": 4.082, + "step": 12990 + }, + { + "epoch": 0.12991, + "grad_norm": 1.0013014198246373, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 12991 + }, + { + "epoch": 0.12992, + "grad_norm": 1.0991845081239433, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 12992 + }, + { + "epoch": 0.12993, + "grad_norm": 1.0433208295006966, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 12993 + }, + { + "epoch": 0.12994, + "grad_norm": 1.3200638135840004, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 12994 + }, + { + "epoch": 0.12995, + "grad_norm": 0.9502599445554497, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 12995 + }, + { + "epoch": 0.12996, + "grad_norm": 1.2478926843509446, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 12996 + }, + { + "epoch": 0.12997, + "grad_norm": 1.1228555540957137, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 12997 + }, + { + "epoch": 0.12998, + "grad_norm": 1.2809069083667275, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 12998 + }, + { + "epoch": 0.12999, + "grad_norm": 1.0503689952353794, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 12999 + }, + { + "epoch": 0.13, + "grad_norm": 1.3115533907016008, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13000 + }, + { + "epoch": 0.13001, + "grad_norm": 1.5532057915451387, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13001 + }, + { + "epoch": 0.13002, + "grad_norm": 1.0069991586277387, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13002 + }, + { + "epoch": 0.13003, + "grad_norm": 1.113897741771145, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13003 + }, + { + "epoch": 0.13004, + "grad_norm": 1.3576791887916393, + "learning_rate": 0.003, + "loss": 4.069, + "step": 13004 + }, + { + "epoch": 0.13005, + "grad_norm": 1.1468295753829618, + "learning_rate": 0.003, + "loss": 4.065, + "step": 13005 + }, + { + "epoch": 0.13006, + "grad_norm": 1.1026555314244328, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13006 + }, + { + "epoch": 0.13007, + "grad_norm": 1.2521757772939597, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 13007 + }, + { + "epoch": 0.13008, + "grad_norm": 1.1901151612620091, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 13008 + }, + { + "epoch": 0.13009, + "grad_norm": 1.075477985606486, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 13009 + }, + { + "epoch": 0.1301, + "grad_norm": 1.2593027138009543, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13010 + }, + { + "epoch": 0.13011, + "grad_norm": 1.289946689484295, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13011 + }, + { + "epoch": 0.13012, + "grad_norm": 1.1376906437862906, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 13012 + }, + { + "epoch": 0.13013, + "grad_norm": 1.1651300549158004, + "learning_rate": 0.003, + "loss": 4.102, + "step": 13013 + }, + { + "epoch": 0.13014, + "grad_norm": 0.9765529309917151, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13014 + }, + { + "epoch": 0.13015, + "grad_norm": 1.1432089220373196, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13015 + }, + { + "epoch": 0.13016, + "grad_norm": 1.2872956102777362, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 13016 + }, + { + "epoch": 0.13017, + "grad_norm": 1.032176218481528, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 13017 + }, + { + "epoch": 0.13018, + "grad_norm": 1.3836158601570603, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 13018 + }, + { + "epoch": 0.13019, + "grad_norm": 1.2484311132058388, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 13019 + }, + { + "epoch": 0.1302, + "grad_norm": 1.133818858935883, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13020 + }, + { + "epoch": 0.13021, + "grad_norm": 1.176538778672168, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13021 + }, + { + "epoch": 0.13022, + "grad_norm": 1.13476736669025, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13022 + }, + { + "epoch": 0.13023, + "grad_norm": 1.2191650827911746, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 13023 + }, + { + "epoch": 0.13024, + "grad_norm": 0.9137587500950195, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13024 + }, + { + "epoch": 0.13025, + "grad_norm": 1.056799876749151, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13025 + }, + { + "epoch": 0.13026, + "grad_norm": 1.32997324639056, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 13026 + }, + { + "epoch": 0.13027, + "grad_norm": 1.0335813697420737, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 13027 + }, + { + "epoch": 0.13028, + "grad_norm": 1.2129057627304443, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 13028 + }, + { + "epoch": 0.13029, + "grad_norm": 1.0463769845927975, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13029 + }, + { + "epoch": 0.1303, + "grad_norm": 0.9371220497761807, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 13030 + }, + { + "epoch": 0.13031, + "grad_norm": 1.0876623507048622, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 13031 + }, + { + "epoch": 0.13032, + "grad_norm": 1.220571969379598, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13032 + }, + { + "epoch": 0.13033, + "grad_norm": 1.283113920569638, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13033 + }, + { + "epoch": 0.13034, + "grad_norm": 1.2716846620882998, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 13034 + }, + { + "epoch": 0.13035, + "grad_norm": 0.9581416755949399, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 13035 + }, + { + "epoch": 0.13036, + "grad_norm": 1.3266757217606555, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13036 + }, + { + "epoch": 0.13037, + "grad_norm": 1.0374416485559421, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13037 + }, + { + "epoch": 0.13038, + "grad_norm": 1.1415059688805005, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 13038 + }, + { + "epoch": 0.13039, + "grad_norm": 1.0862234750529718, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 13039 + }, + { + "epoch": 0.1304, + "grad_norm": 1.3078031355729784, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 13040 + }, + { + "epoch": 0.13041, + "grad_norm": 1.2283871938375608, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13041 + }, + { + "epoch": 0.13042, + "grad_norm": 1.1038751327724663, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13042 + }, + { + "epoch": 0.13043, + "grad_norm": 1.1386760904372004, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 13043 + }, + { + "epoch": 0.13044, + "grad_norm": 1.223682986720779, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 13044 + }, + { + "epoch": 0.13045, + "grad_norm": 1.1310262796396329, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 13045 + }, + { + "epoch": 0.13046, + "grad_norm": 1.1694310039446454, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 13046 + }, + { + "epoch": 0.13047, + "grad_norm": 1.2884607512361166, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13047 + }, + { + "epoch": 0.13048, + "grad_norm": 1.0162810931708417, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13048 + }, + { + "epoch": 0.13049, + "grad_norm": 1.1075577086696535, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 13049 + }, + { + "epoch": 0.1305, + "grad_norm": 0.9801540277850392, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 13050 + }, + { + "epoch": 0.13051, + "grad_norm": 1.0137626493739156, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 13051 + }, + { + "epoch": 0.13052, + "grad_norm": 1.347414761849935, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 13052 + }, + { + "epoch": 0.13053, + "grad_norm": 1.0757138709212104, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 13053 + }, + { + "epoch": 0.13054, + "grad_norm": 1.1405350325370918, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13054 + }, + { + "epoch": 0.13055, + "grad_norm": 1.0298726527013953, + "learning_rate": 0.003, + "loss": 4.054, + "step": 13055 + }, + { + "epoch": 0.13056, + "grad_norm": 1.4076665701517697, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13056 + }, + { + "epoch": 0.13057, + "grad_norm": 0.9379627386101226, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 13057 + }, + { + "epoch": 0.13058, + "grad_norm": 1.3352121990417045, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 13058 + }, + { + "epoch": 0.13059, + "grad_norm": 1.1498020719130333, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 13059 + }, + { + "epoch": 0.1306, + "grad_norm": 1.4026732415806318, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 13060 + }, + { + "epoch": 0.13061, + "grad_norm": 0.9904325540933382, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 13061 + }, + { + "epoch": 0.13062, + "grad_norm": 1.2017031300002665, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 13062 + }, + { + "epoch": 0.13063, + "grad_norm": 1.0898718183097498, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 13063 + }, + { + "epoch": 0.13064, + "grad_norm": 1.1900076841116998, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 13064 + }, + { + "epoch": 0.13065, + "grad_norm": 1.2327062041730883, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13065 + }, + { + "epoch": 0.13066, + "grad_norm": 1.1035502075789365, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 13066 + }, + { + "epoch": 0.13067, + "grad_norm": 0.9854179039669931, + "learning_rate": 0.003, + "loss": 4.119, + "step": 13067 + }, + { + "epoch": 0.13068, + "grad_norm": 1.1314639070759815, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13068 + }, + { + "epoch": 0.13069, + "grad_norm": 1.0365667525729436, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 13069 + }, + { + "epoch": 0.1307, + "grad_norm": 1.1475215298763473, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13070 + }, + { + "epoch": 0.13071, + "grad_norm": 1.0185681100892678, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13071 + }, + { + "epoch": 0.13072, + "grad_norm": 1.220724504459399, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13072 + }, + { + "epoch": 0.13073, + "grad_norm": 1.2797075993239362, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 13073 + }, + { + "epoch": 0.13074, + "grad_norm": 1.028420780164218, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13074 + }, + { + "epoch": 0.13075, + "grad_norm": 1.4298434252273828, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 13075 + }, + { + "epoch": 0.13076, + "grad_norm": 0.9944607543403813, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 13076 + }, + { + "epoch": 0.13077, + "grad_norm": 1.3787263022153806, + "learning_rate": 0.003, + "loss": 4.088, + "step": 13077 + }, + { + "epoch": 0.13078, + "grad_norm": 1.1237099249797804, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 13078 + }, + { + "epoch": 0.13079, + "grad_norm": 1.0764073357883925, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13079 + }, + { + "epoch": 0.1308, + "grad_norm": 1.4621915905997236, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 13080 + }, + { + "epoch": 0.13081, + "grad_norm": 1.2041644590549643, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 13081 + }, + { + "epoch": 0.13082, + "grad_norm": 1.2039357142163236, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 13082 + }, + { + "epoch": 0.13083, + "grad_norm": 1.2885711119752967, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13083 + }, + { + "epoch": 0.13084, + "grad_norm": 1.1357064756724613, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 13084 + }, + { + "epoch": 0.13085, + "grad_norm": 1.14261875842365, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 13085 + }, + { + "epoch": 0.13086, + "grad_norm": 1.1684335033953381, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 13086 + }, + { + "epoch": 0.13087, + "grad_norm": 1.1087092618521905, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 13087 + }, + { + "epoch": 0.13088, + "grad_norm": 1.2939717696394601, + "learning_rate": 0.003, + "loss": 4.09, + "step": 13088 + }, + { + "epoch": 0.13089, + "grad_norm": 1.0429463650533524, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13089 + }, + { + "epoch": 0.1309, + "grad_norm": 1.3643289914345351, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 13090 + }, + { + "epoch": 0.13091, + "grad_norm": 1.1016290315760682, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 13091 + }, + { + "epoch": 0.13092, + "grad_norm": 1.2184391063916458, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13092 + }, + { + "epoch": 0.13093, + "grad_norm": 1.1907130843556937, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13093 + }, + { + "epoch": 0.13094, + "grad_norm": 1.0836391576772713, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13094 + }, + { + "epoch": 0.13095, + "grad_norm": 1.3836963149703367, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13095 + }, + { + "epoch": 0.13096, + "grad_norm": 1.1130676307420784, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13096 + }, + { + "epoch": 0.13097, + "grad_norm": 1.1647929026892412, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 13097 + }, + { + "epoch": 0.13098, + "grad_norm": 1.2042645357212667, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 13098 + }, + { + "epoch": 0.13099, + "grad_norm": 1.2712696162753325, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 13099 + }, + { + "epoch": 0.131, + "grad_norm": 0.9122188658437458, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 13100 + }, + { + "epoch": 0.13101, + "grad_norm": 1.199374985405497, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 13101 + }, + { + "epoch": 0.13102, + "grad_norm": 0.9794571983598035, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 13102 + }, + { + "epoch": 0.13103, + "grad_norm": 1.3212059710839936, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 13103 + }, + { + "epoch": 0.13104, + "grad_norm": 0.969708064207809, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 13104 + }, + { + "epoch": 0.13105, + "grad_norm": 1.4038448146151083, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13105 + }, + { + "epoch": 0.13106, + "grad_norm": 1.3397166893368102, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13106 + }, + { + "epoch": 0.13107, + "grad_norm": 1.1263536237964167, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 13107 + }, + { + "epoch": 0.13108, + "grad_norm": 1.1421597084589206, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13108 + }, + { + "epoch": 0.13109, + "grad_norm": 1.0402132670011577, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 13109 + }, + { + "epoch": 0.1311, + "grad_norm": 1.103323397910269, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13110 + }, + { + "epoch": 0.13111, + "grad_norm": 1.1138973357958888, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 13111 + }, + { + "epoch": 0.13112, + "grad_norm": 1.2533795396144758, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 13112 + }, + { + "epoch": 0.13113, + "grad_norm": 1.2492591473176269, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 13113 + }, + { + "epoch": 0.13114, + "grad_norm": 1.2822561331818683, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13114 + }, + { + "epoch": 0.13115, + "grad_norm": 0.9692542170042184, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 13115 + }, + { + "epoch": 0.13116, + "grad_norm": 1.252201337846062, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 13116 + }, + { + "epoch": 0.13117, + "grad_norm": 1.124907025503089, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 13117 + }, + { + "epoch": 0.13118, + "grad_norm": 1.140182893728732, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 13118 + }, + { + "epoch": 0.13119, + "grad_norm": 1.1603913986093715, + "learning_rate": 0.003, + "loss": 4.063, + "step": 13119 + }, + { + "epoch": 0.1312, + "grad_norm": 1.3401520585930122, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 13120 + }, + { + "epoch": 0.13121, + "grad_norm": 1.045707895208328, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 13121 + }, + { + "epoch": 0.13122, + "grad_norm": 1.3487202343862288, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 13122 + }, + { + "epoch": 0.13123, + "grad_norm": 1.2569422638761363, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13123 + }, + { + "epoch": 0.13124, + "grad_norm": 1.1886828284004596, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13124 + }, + { + "epoch": 0.13125, + "grad_norm": 1.0547772494125238, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13125 + }, + { + "epoch": 0.13126, + "grad_norm": 1.0628137619252966, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 13126 + }, + { + "epoch": 0.13127, + "grad_norm": 1.3267722281858412, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13127 + }, + { + "epoch": 0.13128, + "grad_norm": 0.9974671959993655, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13128 + }, + { + "epoch": 0.13129, + "grad_norm": 1.2037968613620844, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 13129 + }, + { + "epoch": 0.1313, + "grad_norm": 1.0272222870037793, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 13130 + }, + { + "epoch": 0.13131, + "grad_norm": 1.4273829392694313, + "learning_rate": 0.003, + "loss": 4.087, + "step": 13131 + }, + { + "epoch": 0.13132, + "grad_norm": 0.9129965729121428, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 13132 + }, + { + "epoch": 0.13133, + "grad_norm": 1.4042378210051416, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 13133 + }, + { + "epoch": 0.13134, + "grad_norm": 1.2315959320398058, + "learning_rate": 0.003, + "loss": 4.065, + "step": 13134 + }, + { + "epoch": 0.13135, + "grad_norm": 1.145000028319332, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 13135 + }, + { + "epoch": 0.13136, + "grad_norm": 1.3626312509951313, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 13136 + }, + { + "epoch": 0.13137, + "grad_norm": 0.9208041507990571, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 13137 + }, + { + "epoch": 0.13138, + "grad_norm": 0.9949139138439406, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 13138 + }, + { + "epoch": 0.13139, + "grad_norm": 1.1330070542618937, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 13139 + }, + { + "epoch": 0.1314, + "grad_norm": 1.0277347495649414, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 13140 + }, + { + "epoch": 0.13141, + "grad_norm": 1.2157307491872351, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 13141 + }, + { + "epoch": 0.13142, + "grad_norm": 1.1540030738824, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 13142 + }, + { + "epoch": 0.13143, + "grad_norm": 1.083663671224984, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13143 + }, + { + "epoch": 0.13144, + "grad_norm": 1.1693101820813463, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 13144 + }, + { + "epoch": 0.13145, + "grad_norm": 1.0520615237330742, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 13145 + }, + { + "epoch": 0.13146, + "grad_norm": 1.3595372029185335, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13146 + }, + { + "epoch": 0.13147, + "grad_norm": 1.2011223687753416, + "learning_rate": 0.003, + "loss": 4.09, + "step": 13147 + }, + { + "epoch": 0.13148, + "grad_norm": 1.3437084532597734, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 13148 + }, + { + "epoch": 0.13149, + "grad_norm": 1.0555719973390898, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13149 + }, + { + "epoch": 0.1315, + "grad_norm": 1.10074775022198, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 13150 + }, + { + "epoch": 0.13151, + "grad_norm": 1.26607507230037, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 13151 + }, + { + "epoch": 0.13152, + "grad_norm": 1.107632811054954, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 13152 + }, + { + "epoch": 0.13153, + "grad_norm": 1.1120740461445913, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 13153 + }, + { + "epoch": 0.13154, + "grad_norm": 1.0197773346455092, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13154 + }, + { + "epoch": 0.13155, + "grad_norm": 1.3563669767266036, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 13155 + }, + { + "epoch": 0.13156, + "grad_norm": 1.2096978560595582, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 13156 + }, + { + "epoch": 0.13157, + "grad_norm": 1.201207543586879, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 13157 + }, + { + "epoch": 0.13158, + "grad_norm": 1.2986901429844693, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 13158 + }, + { + "epoch": 0.13159, + "grad_norm": 0.9337125919864209, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13159 + }, + { + "epoch": 0.1316, + "grad_norm": 1.093091216961512, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13160 + }, + { + "epoch": 0.13161, + "grad_norm": 1.2966463755097752, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 13161 + }, + { + "epoch": 0.13162, + "grad_norm": 1.0873521204134697, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 13162 + }, + { + "epoch": 0.13163, + "grad_norm": 1.1664387280121102, + "learning_rate": 0.003, + "loss": 4.117, + "step": 13163 + }, + { + "epoch": 0.13164, + "grad_norm": 1.0579034446025735, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13164 + }, + { + "epoch": 0.13165, + "grad_norm": 1.2343600101983327, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13165 + }, + { + "epoch": 0.13166, + "grad_norm": 1.30867519231371, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 13166 + }, + { + "epoch": 0.13167, + "grad_norm": 1.4435775289061636, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13167 + }, + { + "epoch": 0.13168, + "grad_norm": 1.2152487189165901, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 13168 + }, + { + "epoch": 0.13169, + "grad_norm": 0.9907585100666397, + "learning_rate": 0.003, + "loss": 4.044, + "step": 13169 + }, + { + "epoch": 0.1317, + "grad_norm": 1.166126698203818, + "learning_rate": 0.003, + "loss": 4.085, + "step": 13170 + }, + { + "epoch": 0.13171, + "grad_norm": 1.1213359245000933, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 13171 + }, + { + "epoch": 0.13172, + "grad_norm": 1.3021235886651639, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 13172 + }, + { + "epoch": 0.13173, + "grad_norm": 1.0134856626832358, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 13173 + }, + { + "epoch": 0.13174, + "grad_norm": 1.2731135211937619, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 13174 + }, + { + "epoch": 0.13175, + "grad_norm": 0.9865560884011636, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13175 + }, + { + "epoch": 0.13176, + "grad_norm": 1.4327052210028386, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13176 + }, + { + "epoch": 0.13177, + "grad_norm": 1.148008225907461, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 13177 + }, + { + "epoch": 0.13178, + "grad_norm": 1.218790609154238, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 13178 + }, + { + "epoch": 0.13179, + "grad_norm": 1.1822536749554937, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13179 + }, + { + "epoch": 0.1318, + "grad_norm": 0.9070379499444476, + "learning_rate": 0.003, + "loss": 4.055, + "step": 13180 + }, + { + "epoch": 0.13181, + "grad_norm": 0.9945075791236695, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 13181 + }, + { + "epoch": 0.13182, + "grad_norm": 1.240539209282678, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 13182 + }, + { + "epoch": 0.13183, + "grad_norm": 1.1850668198222947, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 13183 + }, + { + "epoch": 0.13184, + "grad_norm": 1.1527421237373985, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13184 + }, + { + "epoch": 0.13185, + "grad_norm": 1.0952135964750374, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 13185 + }, + { + "epoch": 0.13186, + "grad_norm": 1.2506033733384114, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 13186 + }, + { + "epoch": 0.13187, + "grad_norm": 0.8635762234024932, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13187 + }, + { + "epoch": 0.13188, + "grad_norm": 0.9892073305520283, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13188 + }, + { + "epoch": 0.13189, + "grad_norm": 1.2420881455372659, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13189 + }, + { + "epoch": 0.1319, + "grad_norm": 1.017536179497929, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 13190 + }, + { + "epoch": 0.13191, + "grad_norm": 1.1841956984720243, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 13191 + }, + { + "epoch": 0.13192, + "grad_norm": 1.2260648938645817, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 13192 + }, + { + "epoch": 0.13193, + "grad_norm": 1.1065175186224798, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13193 + }, + { + "epoch": 0.13194, + "grad_norm": 1.4416607096483425, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 13194 + }, + { + "epoch": 0.13195, + "grad_norm": 1.316389643719257, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13195 + }, + { + "epoch": 0.13196, + "grad_norm": 1.1352159192036904, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 13196 + }, + { + "epoch": 0.13197, + "grad_norm": 1.1568917969318886, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 13197 + }, + { + "epoch": 0.13198, + "grad_norm": 1.1256710643386678, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 13198 + }, + { + "epoch": 0.13199, + "grad_norm": 1.2156018543905163, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13199 + }, + { + "epoch": 0.132, + "grad_norm": 1.342164752783955, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 13200 + }, + { + "epoch": 0.13201, + "grad_norm": 1.0440248276876525, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 13201 + }, + { + "epoch": 0.13202, + "grad_norm": 1.3321336678515836, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 13202 + }, + { + "epoch": 0.13203, + "grad_norm": 1.1136943829841504, + "learning_rate": 0.003, + "loss": 4.084, + "step": 13203 + }, + { + "epoch": 0.13204, + "grad_norm": 1.0896162709661188, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 13204 + }, + { + "epoch": 0.13205, + "grad_norm": 1.1487113403901221, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 13205 + }, + { + "epoch": 0.13206, + "grad_norm": 1.139305369933213, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 13206 + }, + { + "epoch": 0.13207, + "grad_norm": 1.1835547312488282, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 13207 + }, + { + "epoch": 0.13208, + "grad_norm": 1.33247995754553, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 13208 + }, + { + "epoch": 0.13209, + "grad_norm": 1.3513182675506308, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13209 + }, + { + "epoch": 0.1321, + "grad_norm": 1.2551941924992103, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 13210 + }, + { + "epoch": 0.13211, + "grad_norm": 0.9604208742486993, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 13211 + }, + { + "epoch": 0.13212, + "grad_norm": 1.206886761240646, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13212 + }, + { + "epoch": 0.13213, + "grad_norm": 1.0990701998695058, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 13213 + }, + { + "epoch": 0.13214, + "grad_norm": 1.1918705286316655, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13214 + }, + { + "epoch": 0.13215, + "grad_norm": 1.162824123790289, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13215 + }, + { + "epoch": 0.13216, + "grad_norm": 1.2346845464822072, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 13216 + }, + { + "epoch": 0.13217, + "grad_norm": 1.1794704948545738, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 13217 + }, + { + "epoch": 0.13218, + "grad_norm": 1.1513387641304558, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 13218 + }, + { + "epoch": 0.13219, + "grad_norm": 1.2180329095543465, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 13219 + }, + { + "epoch": 0.1322, + "grad_norm": 1.484575446629248, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 13220 + }, + { + "epoch": 0.13221, + "grad_norm": 0.9424781728975216, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13221 + }, + { + "epoch": 0.13222, + "grad_norm": 1.0684150192432058, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13222 + }, + { + "epoch": 0.13223, + "grad_norm": 1.2191264328565914, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13223 + }, + { + "epoch": 0.13224, + "grad_norm": 0.9350032787837502, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 13224 + }, + { + "epoch": 0.13225, + "grad_norm": 1.2614731457692834, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13225 + }, + { + "epoch": 0.13226, + "grad_norm": 1.4290705818533183, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 13226 + }, + { + "epoch": 0.13227, + "grad_norm": 1.0077627912110383, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13227 + }, + { + "epoch": 0.13228, + "grad_norm": 1.2211893006066725, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 13228 + }, + { + "epoch": 0.13229, + "grad_norm": 1.009855936876147, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 13229 + }, + { + "epoch": 0.1323, + "grad_norm": 1.37922530451958, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 13230 + }, + { + "epoch": 0.13231, + "grad_norm": 1.0292004867682374, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 13231 + }, + { + "epoch": 0.13232, + "grad_norm": 1.1769272669945927, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 13232 + }, + { + "epoch": 0.13233, + "grad_norm": 1.1612880435231938, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 13233 + }, + { + "epoch": 0.13234, + "grad_norm": 1.2253876757763627, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13234 + }, + { + "epoch": 0.13235, + "grad_norm": 1.3619907271565719, + "learning_rate": 0.003, + "loss": 4.076, + "step": 13235 + }, + { + "epoch": 0.13236, + "grad_norm": 0.9839652554544269, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 13236 + }, + { + "epoch": 0.13237, + "grad_norm": 1.1875127600863051, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13237 + }, + { + "epoch": 0.13238, + "grad_norm": 1.1055518959048614, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 13238 + }, + { + "epoch": 0.13239, + "grad_norm": 1.1295348692789726, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13239 + }, + { + "epoch": 0.1324, + "grad_norm": 1.0981871470208329, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 13240 + }, + { + "epoch": 0.13241, + "grad_norm": 1.6840769284460566, + "learning_rate": 0.003, + "loss": 4.074, + "step": 13241 + }, + { + "epoch": 0.13242, + "grad_norm": 0.9366529133323972, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 13242 + }, + { + "epoch": 0.13243, + "grad_norm": 1.0424880225179558, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 13243 + }, + { + "epoch": 0.13244, + "grad_norm": 1.0848485150228266, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13244 + }, + { + "epoch": 0.13245, + "grad_norm": 1.1482278902934118, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13245 + }, + { + "epoch": 0.13246, + "grad_norm": 1.1663589694334175, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 13246 + }, + { + "epoch": 0.13247, + "grad_norm": 1.2100677855701005, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 13247 + }, + { + "epoch": 0.13248, + "grad_norm": 1.4384682116904701, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 13248 + }, + { + "epoch": 0.13249, + "grad_norm": 1.2706725172309934, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 13249 + }, + { + "epoch": 0.1325, + "grad_norm": 1.2777569039571888, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 13250 + }, + { + "epoch": 0.13251, + "grad_norm": 1.286598133095499, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13251 + }, + { + "epoch": 0.13252, + "grad_norm": 1.1308979039030347, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 13252 + }, + { + "epoch": 0.13253, + "grad_norm": 1.1684796593582514, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 13253 + }, + { + "epoch": 0.13254, + "grad_norm": 1.2589559908787755, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 13254 + }, + { + "epoch": 0.13255, + "grad_norm": 1.113630615783083, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 13255 + }, + { + "epoch": 0.13256, + "grad_norm": 1.2670567957920973, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 13256 + }, + { + "epoch": 0.13257, + "grad_norm": 1.0472226710354418, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 13257 + }, + { + "epoch": 0.13258, + "grad_norm": 1.2427763650400145, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13258 + }, + { + "epoch": 0.13259, + "grad_norm": 0.8243964260074956, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 13259 + }, + { + "epoch": 0.1326, + "grad_norm": 1.063344790053129, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13260 + }, + { + "epoch": 0.13261, + "grad_norm": 1.541367385463494, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 13261 + }, + { + "epoch": 0.13262, + "grad_norm": 0.9808939301570836, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 13262 + }, + { + "epoch": 0.13263, + "grad_norm": 1.2734804829811215, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13263 + }, + { + "epoch": 0.13264, + "grad_norm": 1.124647905227643, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13264 + }, + { + "epoch": 0.13265, + "grad_norm": 1.319801420352576, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13265 + }, + { + "epoch": 0.13266, + "grad_norm": 1.0439692891652967, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 13266 + }, + { + "epoch": 0.13267, + "grad_norm": 1.074415191311497, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13267 + }, + { + "epoch": 0.13268, + "grad_norm": 1.2097445537659586, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13268 + }, + { + "epoch": 0.13269, + "grad_norm": 1.2050238004257336, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 13269 + }, + { + "epoch": 0.1327, + "grad_norm": 1.4526043801125104, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 13270 + }, + { + "epoch": 0.13271, + "grad_norm": 1.0282734940110956, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 13271 + }, + { + "epoch": 0.13272, + "grad_norm": 1.2436480572127762, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 13272 + }, + { + "epoch": 0.13273, + "grad_norm": 1.23181245795007, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 13273 + }, + { + "epoch": 0.13274, + "grad_norm": 1.1028753704920953, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 13274 + }, + { + "epoch": 0.13275, + "grad_norm": 1.1569503153916652, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13275 + }, + { + "epoch": 0.13276, + "grad_norm": 1.066571379296384, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 13276 + }, + { + "epoch": 0.13277, + "grad_norm": 1.1392178172837175, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 13277 + }, + { + "epoch": 0.13278, + "grad_norm": 1.2438477369686805, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13278 + }, + { + "epoch": 0.13279, + "grad_norm": 1.2866787913315778, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 13279 + }, + { + "epoch": 0.1328, + "grad_norm": 1.1591452088705272, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 13280 + }, + { + "epoch": 0.13281, + "grad_norm": 1.1830361582764495, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 13281 + }, + { + "epoch": 0.13282, + "grad_norm": 1.2505311700849044, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 13282 + }, + { + "epoch": 0.13283, + "grad_norm": 1.2096547188191793, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13283 + }, + { + "epoch": 0.13284, + "grad_norm": 1.0036341444276413, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 13284 + }, + { + "epoch": 0.13285, + "grad_norm": 1.4806163538861903, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 13285 + }, + { + "epoch": 0.13286, + "grad_norm": 1.0813766591338672, + "learning_rate": 0.003, + "loss": 4.106, + "step": 13286 + }, + { + "epoch": 0.13287, + "grad_norm": 1.427730546993291, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 13287 + }, + { + "epoch": 0.13288, + "grad_norm": 1.0527188181403344, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 13288 + }, + { + "epoch": 0.13289, + "grad_norm": 1.2044844905582028, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 13289 + }, + { + "epoch": 0.1329, + "grad_norm": 1.2217949119878093, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 13290 + }, + { + "epoch": 0.13291, + "grad_norm": 1.0952651198626726, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 13291 + }, + { + "epoch": 0.13292, + "grad_norm": 1.093891287004927, + "learning_rate": 0.003, + "loss": 4.084, + "step": 13292 + }, + { + "epoch": 0.13293, + "grad_norm": 1.2853429222571366, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 13293 + }, + { + "epoch": 0.13294, + "grad_norm": 1.0565395357452285, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 13294 + }, + { + "epoch": 0.13295, + "grad_norm": 1.129468602840372, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 13295 + }, + { + "epoch": 0.13296, + "grad_norm": 1.285359229503194, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 13296 + }, + { + "epoch": 0.13297, + "grad_norm": 1.2142336058007384, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 13297 + }, + { + "epoch": 0.13298, + "grad_norm": 1.1731333515874773, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 13298 + }, + { + "epoch": 0.13299, + "grad_norm": 1.1489028482637664, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 13299 + }, + { + "epoch": 0.133, + "grad_norm": 1.445604589879857, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 13300 + }, + { + "epoch": 0.13301, + "grad_norm": 1.067028670997227, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 13301 + }, + { + "epoch": 0.13302, + "grad_norm": 1.1368262006373977, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 13302 + }, + { + "epoch": 0.13303, + "grad_norm": 1.2180401571199988, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13303 + }, + { + "epoch": 0.13304, + "grad_norm": 1.1993665889601892, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 13304 + }, + { + "epoch": 0.13305, + "grad_norm": 1.032978558822638, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 13305 + }, + { + "epoch": 0.13306, + "grad_norm": 1.2657793854465964, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13306 + }, + { + "epoch": 0.13307, + "grad_norm": 1.1347037754378475, + "learning_rate": 0.003, + "loss": 4.058, + "step": 13307 + }, + { + "epoch": 0.13308, + "grad_norm": 1.0149496856090754, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 13308 + }, + { + "epoch": 0.13309, + "grad_norm": 1.4440999617014638, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 13309 + }, + { + "epoch": 0.1331, + "grad_norm": 0.9308118300964536, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 13310 + }, + { + "epoch": 0.13311, + "grad_norm": 1.2301159217166553, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 13311 + }, + { + "epoch": 0.13312, + "grad_norm": 1.1980100279002193, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 13312 + }, + { + "epoch": 0.13313, + "grad_norm": 1.1301051915096612, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13313 + }, + { + "epoch": 0.13314, + "grad_norm": 1.1689863478212108, + "learning_rate": 0.003, + "loss": 4.071, + "step": 13314 + }, + { + "epoch": 0.13315, + "grad_norm": 1.062336883714272, + "learning_rate": 0.003, + "loss": 4.087, + "step": 13315 + }, + { + "epoch": 0.13316, + "grad_norm": 1.3501774067650583, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13316 + }, + { + "epoch": 0.13317, + "grad_norm": 1.1449323114599748, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 13317 + }, + { + "epoch": 0.13318, + "grad_norm": 1.1220465724242124, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13318 + }, + { + "epoch": 0.13319, + "grad_norm": 1.259102102799351, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 13319 + }, + { + "epoch": 0.1332, + "grad_norm": 1.12159952245913, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 13320 + }, + { + "epoch": 0.13321, + "grad_norm": 1.2950597137586786, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13321 + }, + { + "epoch": 0.13322, + "grad_norm": 1.0369680795149159, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 13322 + }, + { + "epoch": 0.13323, + "grad_norm": 1.2306452875554148, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 13323 + }, + { + "epoch": 0.13324, + "grad_norm": 1.2988148242167328, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 13324 + }, + { + "epoch": 0.13325, + "grad_norm": 1.0726268592798407, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13325 + }, + { + "epoch": 0.13326, + "grad_norm": 1.201112262648506, + "learning_rate": 0.003, + "loss": 4.102, + "step": 13326 + }, + { + "epoch": 0.13327, + "grad_norm": 1.1806783452463878, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 13327 + }, + { + "epoch": 0.13328, + "grad_norm": 1.3125564729665666, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13328 + }, + { + "epoch": 0.13329, + "grad_norm": 1.0701162709873437, + "learning_rate": 0.003, + "loss": 4.092, + "step": 13329 + }, + { + "epoch": 0.1333, + "grad_norm": 1.08267371694619, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 13330 + }, + { + "epoch": 0.13331, + "grad_norm": 1.1817818754448577, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 13331 + }, + { + "epoch": 0.13332, + "grad_norm": 1.1969977050107905, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 13332 + }, + { + "epoch": 0.13333, + "grad_norm": 1.3098841770978693, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 13333 + }, + { + "epoch": 0.13334, + "grad_norm": 0.9831900150336861, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 13334 + }, + { + "epoch": 0.13335, + "grad_norm": 1.0387924753083626, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 13335 + }, + { + "epoch": 0.13336, + "grad_norm": 1.3004967745171618, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 13336 + }, + { + "epoch": 0.13337, + "grad_norm": 1.1178283993965488, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13337 + }, + { + "epoch": 0.13338, + "grad_norm": 1.0357008083180437, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 13338 + }, + { + "epoch": 0.13339, + "grad_norm": 1.0711535445470308, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 13339 + }, + { + "epoch": 0.1334, + "grad_norm": 1.1066555227662631, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 13340 + }, + { + "epoch": 0.13341, + "grad_norm": 1.316474283196772, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 13341 + }, + { + "epoch": 0.13342, + "grad_norm": 1.245714138650924, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 13342 + }, + { + "epoch": 0.13343, + "grad_norm": 1.1062513626932633, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13343 + }, + { + "epoch": 0.13344, + "grad_norm": 1.088863163471069, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 13344 + }, + { + "epoch": 0.13345, + "grad_norm": 0.9775251027358349, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 13345 + }, + { + "epoch": 0.13346, + "grad_norm": 1.3447630935973398, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 13346 + }, + { + "epoch": 0.13347, + "grad_norm": 1.2367279065358332, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13347 + }, + { + "epoch": 0.13348, + "grad_norm": 1.0752274257332177, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13348 + }, + { + "epoch": 0.13349, + "grad_norm": 1.344950271232708, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 13349 + }, + { + "epoch": 0.1335, + "grad_norm": 1.2571772033099773, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 13350 + }, + { + "epoch": 0.13351, + "grad_norm": 1.235970763029316, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13351 + }, + { + "epoch": 0.13352, + "grad_norm": 1.295337004405207, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 13352 + }, + { + "epoch": 0.13353, + "grad_norm": 1.180645390644191, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 13353 + }, + { + "epoch": 0.13354, + "grad_norm": 1.1978662749356284, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 13354 + }, + { + "epoch": 0.13355, + "grad_norm": 1.1867509477213356, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 13355 + }, + { + "epoch": 0.13356, + "grad_norm": 1.245461161844399, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 13356 + }, + { + "epoch": 0.13357, + "grad_norm": 1.1018784359679767, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 13357 + }, + { + "epoch": 0.13358, + "grad_norm": 1.1035812117049424, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 13358 + }, + { + "epoch": 0.13359, + "grad_norm": 1.1254309321450855, + "learning_rate": 0.003, + "loss": 4.088, + "step": 13359 + }, + { + "epoch": 0.1336, + "grad_norm": 1.2553998155691877, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 13360 + }, + { + "epoch": 0.13361, + "grad_norm": 1.0222119041098203, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 13361 + }, + { + "epoch": 0.13362, + "grad_norm": 1.3022338392042456, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13362 + }, + { + "epoch": 0.13363, + "grad_norm": 0.9521752871091076, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13363 + }, + { + "epoch": 0.13364, + "grad_norm": 1.229941668303482, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 13364 + }, + { + "epoch": 0.13365, + "grad_norm": 1.2541762933410678, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13365 + }, + { + "epoch": 0.13366, + "grad_norm": 1.4218542821693612, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 13366 + }, + { + "epoch": 0.13367, + "grad_norm": 1.3063230638440648, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 13367 + }, + { + "epoch": 0.13368, + "grad_norm": 1.0202774961705887, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 13368 + }, + { + "epoch": 0.13369, + "grad_norm": 1.1992099294907612, + "learning_rate": 0.003, + "loss": 4.082, + "step": 13369 + }, + { + "epoch": 0.1337, + "grad_norm": 0.9642267936227421, + "learning_rate": 0.003, + "loss": 4.072, + "step": 13370 + }, + { + "epoch": 0.13371, + "grad_norm": 1.281749927644538, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 13371 + }, + { + "epoch": 0.13372, + "grad_norm": 1.1878458622250818, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 13372 + }, + { + "epoch": 0.13373, + "grad_norm": 1.5797165771743982, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 13373 + }, + { + "epoch": 0.13374, + "grad_norm": 1.0396410036095307, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 13374 + }, + { + "epoch": 0.13375, + "grad_norm": 1.2712754912215498, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 13375 + }, + { + "epoch": 0.13376, + "grad_norm": 1.2183517382921254, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13376 + }, + { + "epoch": 0.13377, + "grad_norm": 1.0613623993549737, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 13377 + }, + { + "epoch": 0.13378, + "grad_norm": 1.3355682372522968, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13378 + }, + { + "epoch": 0.13379, + "grad_norm": 1.0938413286089992, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 13379 + }, + { + "epoch": 0.1338, + "grad_norm": 1.1103890552746947, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 13380 + }, + { + "epoch": 0.13381, + "grad_norm": 1.2699701548903857, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 13381 + }, + { + "epoch": 0.13382, + "grad_norm": 1.1233842540217829, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 13382 + }, + { + "epoch": 0.13383, + "grad_norm": 1.2765612797715782, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 13383 + }, + { + "epoch": 0.13384, + "grad_norm": 1.2006039862386173, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13384 + }, + { + "epoch": 0.13385, + "grad_norm": 1.1642926025532412, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13385 + }, + { + "epoch": 0.13386, + "grad_norm": 1.2571178812083903, + "learning_rate": 0.003, + "loss": 4.081, + "step": 13386 + }, + { + "epoch": 0.13387, + "grad_norm": 0.9608708361775664, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13387 + }, + { + "epoch": 0.13388, + "grad_norm": 1.1366847605441799, + "learning_rate": 0.003, + "loss": 4.087, + "step": 13388 + }, + { + "epoch": 0.13389, + "grad_norm": 1.2306720467075594, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13389 + }, + { + "epoch": 0.1339, + "grad_norm": 1.1164334604497055, + "learning_rate": 0.003, + "loss": 4.07, + "step": 13390 + }, + { + "epoch": 0.13391, + "grad_norm": 1.0426308201644983, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 13391 + }, + { + "epoch": 0.13392, + "grad_norm": 1.210123565635377, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 13392 + }, + { + "epoch": 0.13393, + "grad_norm": 1.2414495817178337, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 13393 + }, + { + "epoch": 0.13394, + "grad_norm": 1.3451434252302206, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13394 + }, + { + "epoch": 0.13395, + "grad_norm": 1.1878714124216885, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13395 + }, + { + "epoch": 0.13396, + "grad_norm": 1.2257200245169122, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13396 + }, + { + "epoch": 0.13397, + "grad_norm": 1.0356620543924908, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 13397 + }, + { + "epoch": 0.13398, + "grad_norm": 1.282872635708081, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 13398 + }, + { + "epoch": 0.13399, + "grad_norm": 1.1479540377003152, + "learning_rate": 0.003, + "loss": 4.036, + "step": 13399 + }, + { + "epoch": 0.134, + "grad_norm": 1.1005041773174578, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 13400 + }, + { + "epoch": 0.13401, + "grad_norm": 1.268179261420047, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 13401 + }, + { + "epoch": 0.13402, + "grad_norm": 1.1738455552542615, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 13402 + }, + { + "epoch": 0.13403, + "grad_norm": 1.3879437579648175, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 13403 + }, + { + "epoch": 0.13404, + "grad_norm": 1.2541279008654285, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 13404 + }, + { + "epoch": 0.13405, + "grad_norm": 1.2605087317660528, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13405 + }, + { + "epoch": 0.13406, + "grad_norm": 1.2681078901785494, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13406 + }, + { + "epoch": 0.13407, + "grad_norm": 1.2605010221084334, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13407 + }, + { + "epoch": 0.13408, + "grad_norm": 0.8239793765808275, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 13408 + }, + { + "epoch": 0.13409, + "grad_norm": 0.9311582810932972, + "learning_rate": 0.003, + "loss": 4.066, + "step": 13409 + }, + { + "epoch": 0.1341, + "grad_norm": 1.1252731190236187, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 13410 + }, + { + "epoch": 0.13411, + "grad_norm": 1.1060331467637572, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13411 + }, + { + "epoch": 0.13412, + "grad_norm": 0.9573062747071894, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 13412 + }, + { + "epoch": 0.13413, + "grad_norm": 1.5146959163459248, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 13413 + }, + { + "epoch": 0.13414, + "grad_norm": 1.1917831643750696, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 13414 + }, + { + "epoch": 0.13415, + "grad_norm": 1.2904828536059358, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13415 + }, + { + "epoch": 0.13416, + "grad_norm": 0.9840594781612224, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13416 + }, + { + "epoch": 0.13417, + "grad_norm": 1.2528731806628532, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 13417 + }, + { + "epoch": 0.13418, + "grad_norm": 1.0942044526081434, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13418 + }, + { + "epoch": 0.13419, + "grad_norm": 1.3026400484830545, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13419 + }, + { + "epoch": 0.1342, + "grad_norm": 1.058350961997098, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 13420 + }, + { + "epoch": 0.13421, + "grad_norm": 1.0299704476949245, + "learning_rate": 0.003, + "loss": 4.086, + "step": 13421 + }, + { + "epoch": 0.13422, + "grad_norm": 1.1181419868091191, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 13422 + }, + { + "epoch": 0.13423, + "grad_norm": 1.2457760033592935, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13423 + }, + { + "epoch": 0.13424, + "grad_norm": 1.1283267972883926, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 13424 + }, + { + "epoch": 0.13425, + "grad_norm": 1.2522641300795905, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13425 + }, + { + "epoch": 0.13426, + "grad_norm": 0.9834937636747162, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 13426 + }, + { + "epoch": 0.13427, + "grad_norm": 1.2090305796613712, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 13427 + }, + { + "epoch": 0.13428, + "grad_norm": 1.194819039867254, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 13428 + }, + { + "epoch": 0.13429, + "grad_norm": 1.3388245141106132, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 13429 + }, + { + "epoch": 0.1343, + "grad_norm": 1.1759131578031872, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 13430 + }, + { + "epoch": 0.13431, + "grad_norm": 1.4354047016128604, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 13431 + }, + { + "epoch": 0.13432, + "grad_norm": 0.9462723573405492, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 13432 + }, + { + "epoch": 0.13433, + "grad_norm": 1.1738535078072172, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13433 + }, + { + "epoch": 0.13434, + "grad_norm": 1.1280589093406457, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 13434 + }, + { + "epoch": 0.13435, + "grad_norm": 1.1328603065697254, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 13435 + }, + { + "epoch": 0.13436, + "grad_norm": 1.3643358412921909, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 13436 + }, + { + "epoch": 0.13437, + "grad_norm": 1.1349177609772076, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13437 + }, + { + "epoch": 0.13438, + "grad_norm": 1.3786563909537977, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 13438 + }, + { + "epoch": 0.13439, + "grad_norm": 0.9610016782956333, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13439 + }, + { + "epoch": 0.1344, + "grad_norm": 1.2297517952028227, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13440 + }, + { + "epoch": 0.13441, + "grad_norm": 1.1601578060683668, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 13441 + }, + { + "epoch": 0.13442, + "grad_norm": 1.3373133683032783, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 13442 + }, + { + "epoch": 0.13443, + "grad_norm": 1.0474449161363737, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 13443 + }, + { + "epoch": 0.13444, + "grad_norm": 1.4151051682755142, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 13444 + }, + { + "epoch": 0.13445, + "grad_norm": 1.0537975131356527, + "learning_rate": 0.003, + "loss": 4.07, + "step": 13445 + }, + { + "epoch": 0.13446, + "grad_norm": 1.186175713874241, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13446 + }, + { + "epoch": 0.13447, + "grad_norm": 1.061759357207979, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 13447 + }, + { + "epoch": 0.13448, + "grad_norm": 1.415112960218441, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13448 + }, + { + "epoch": 0.13449, + "grad_norm": 1.1728851918493217, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 13449 + }, + { + "epoch": 0.1345, + "grad_norm": 1.4236709724040442, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13450 + }, + { + "epoch": 0.13451, + "grad_norm": 0.9223059572756359, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 13451 + }, + { + "epoch": 0.13452, + "grad_norm": 1.2323279228543833, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13452 + }, + { + "epoch": 0.13453, + "grad_norm": 1.291375695669297, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 13453 + }, + { + "epoch": 0.13454, + "grad_norm": 1.157160154775136, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13454 + }, + { + "epoch": 0.13455, + "grad_norm": 1.2368519621370178, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 13455 + }, + { + "epoch": 0.13456, + "grad_norm": 1.3142726252125636, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 13456 + }, + { + "epoch": 0.13457, + "grad_norm": 1.162434038095761, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 13457 + }, + { + "epoch": 0.13458, + "grad_norm": 1.1078576104450124, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 13458 + }, + { + "epoch": 0.13459, + "grad_norm": 1.2383130680770433, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 13459 + }, + { + "epoch": 0.1346, + "grad_norm": 1.1125401488562099, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13460 + }, + { + "epoch": 0.13461, + "grad_norm": 1.2140055487172725, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13461 + }, + { + "epoch": 0.13462, + "grad_norm": 1.1433151082950783, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 13462 + }, + { + "epoch": 0.13463, + "grad_norm": 1.1821004125728793, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13463 + }, + { + "epoch": 0.13464, + "grad_norm": 1.2662168014729447, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 13464 + }, + { + "epoch": 0.13465, + "grad_norm": 0.9888495946718485, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 13465 + }, + { + "epoch": 0.13466, + "grad_norm": 1.1220154671518745, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 13466 + }, + { + "epoch": 0.13467, + "grad_norm": 1.0919519821871781, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13467 + }, + { + "epoch": 0.13468, + "grad_norm": 1.340826365379991, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 13468 + }, + { + "epoch": 0.13469, + "grad_norm": 1.0121555399016573, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 13469 + }, + { + "epoch": 0.1347, + "grad_norm": 1.3067290538201393, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 13470 + }, + { + "epoch": 0.13471, + "grad_norm": 1.0880686216576068, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 13471 + }, + { + "epoch": 0.13472, + "grad_norm": 1.1831367938514399, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13472 + }, + { + "epoch": 0.13473, + "grad_norm": 1.3612868750769027, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 13473 + }, + { + "epoch": 0.13474, + "grad_norm": 0.9508255251277685, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 13474 + }, + { + "epoch": 0.13475, + "grad_norm": 1.2856856658658182, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13475 + }, + { + "epoch": 0.13476, + "grad_norm": 1.1606275811173252, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 13476 + }, + { + "epoch": 0.13477, + "grad_norm": 1.1524479186013914, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 13477 + }, + { + "epoch": 0.13478, + "grad_norm": 1.235184313994686, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 13478 + }, + { + "epoch": 0.13479, + "grad_norm": 1.1070378269991312, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13479 + }, + { + "epoch": 0.1348, + "grad_norm": 1.0632334092285003, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 13480 + }, + { + "epoch": 0.13481, + "grad_norm": 1.2234028229811578, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13481 + }, + { + "epoch": 0.13482, + "grad_norm": 1.146489255781058, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13482 + }, + { + "epoch": 0.13483, + "grad_norm": 1.1637185523813744, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13483 + }, + { + "epoch": 0.13484, + "grad_norm": 1.1185100360085258, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 13484 + }, + { + "epoch": 0.13485, + "grad_norm": 1.193469537143719, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 13485 + }, + { + "epoch": 0.13486, + "grad_norm": 1.135558165767479, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 13486 + }, + { + "epoch": 0.13487, + "grad_norm": 1.4057601165947837, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 13487 + }, + { + "epoch": 0.13488, + "grad_norm": 1.1656899608751365, + "learning_rate": 0.003, + "loss": 4.078, + "step": 13488 + }, + { + "epoch": 0.13489, + "grad_norm": 1.5043333136087835, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 13489 + }, + { + "epoch": 0.1349, + "grad_norm": 1.0163820255010707, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13490 + }, + { + "epoch": 0.13491, + "grad_norm": 1.3061494100348068, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 13491 + }, + { + "epoch": 0.13492, + "grad_norm": 1.1228916805737252, + "learning_rate": 0.003, + "loss": 4.084, + "step": 13492 + }, + { + "epoch": 0.13493, + "grad_norm": 1.2622587830241703, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13493 + }, + { + "epoch": 0.13494, + "grad_norm": 1.2240498827665462, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13494 + }, + { + "epoch": 0.13495, + "grad_norm": 1.1536114873612044, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 13495 + }, + { + "epoch": 0.13496, + "grad_norm": 1.2557495984348974, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 13496 + }, + { + "epoch": 0.13497, + "grad_norm": 1.2383652825597389, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13497 + }, + { + "epoch": 0.13498, + "grad_norm": 1.101293070079644, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13498 + }, + { + "epoch": 0.13499, + "grad_norm": 1.2268027447598873, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 13499 + }, + { + "epoch": 0.135, + "grad_norm": 1.4412497255523504, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13500 + }, + { + "epoch": 0.13501, + "grad_norm": 1.1916063975445812, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 13501 + }, + { + "epoch": 0.13502, + "grad_norm": 1.178800704541382, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13502 + }, + { + "epoch": 0.13503, + "grad_norm": 1.1767774893647651, + "learning_rate": 0.003, + "loss": 4.064, + "step": 13503 + }, + { + "epoch": 0.13504, + "grad_norm": 1.1528621548637201, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 13504 + }, + { + "epoch": 0.13505, + "grad_norm": 1.3891953153521712, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 13505 + }, + { + "epoch": 0.13506, + "grad_norm": 0.9887205718992813, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13506 + }, + { + "epoch": 0.13507, + "grad_norm": 1.139102638279225, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 13507 + }, + { + "epoch": 0.13508, + "grad_norm": 1.0348166624357527, + "learning_rate": 0.003, + "loss": 4.058, + "step": 13508 + }, + { + "epoch": 0.13509, + "grad_norm": 1.266192907540622, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13509 + }, + { + "epoch": 0.1351, + "grad_norm": 1.235255795201123, + "learning_rate": 0.003, + "loss": 4.078, + "step": 13510 + }, + { + "epoch": 0.13511, + "grad_norm": 1.2451094931773148, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13511 + }, + { + "epoch": 0.13512, + "grad_norm": 1.038734480871431, + "learning_rate": 0.003, + "loss": 4.066, + "step": 13512 + }, + { + "epoch": 0.13513, + "grad_norm": 1.3362978558878509, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13513 + }, + { + "epoch": 0.13514, + "grad_norm": 1.0116571880327812, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 13514 + }, + { + "epoch": 0.13515, + "grad_norm": 1.5055660158456143, + "learning_rate": 0.003, + "loss": 4.094, + "step": 13515 + }, + { + "epoch": 0.13516, + "grad_norm": 1.0753440149432287, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 13516 + }, + { + "epoch": 0.13517, + "grad_norm": 1.475048709045603, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 13517 + }, + { + "epoch": 0.13518, + "grad_norm": 1.057908709973592, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 13518 + }, + { + "epoch": 0.13519, + "grad_norm": 1.2103503090775813, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13519 + }, + { + "epoch": 0.1352, + "grad_norm": 1.2676298070510073, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 13520 + }, + { + "epoch": 0.13521, + "grad_norm": 1.1815116642938954, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 13521 + }, + { + "epoch": 0.13522, + "grad_norm": 1.085564867833204, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13522 + }, + { + "epoch": 0.13523, + "grad_norm": 1.0268095579877041, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13523 + }, + { + "epoch": 0.13524, + "grad_norm": 1.3869619517893401, + "learning_rate": 0.003, + "loss": 4.099, + "step": 13524 + }, + { + "epoch": 0.13525, + "grad_norm": 0.9848067653607371, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 13525 + }, + { + "epoch": 0.13526, + "grad_norm": 1.4095299409827127, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 13526 + }, + { + "epoch": 0.13527, + "grad_norm": 1.1064622843634515, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 13527 + }, + { + "epoch": 0.13528, + "grad_norm": 1.2619597308926152, + "learning_rate": 0.003, + "loss": 4.077, + "step": 13528 + }, + { + "epoch": 0.13529, + "grad_norm": 1.3763006823062767, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 13529 + }, + { + "epoch": 0.1353, + "grad_norm": 1.1529577761975283, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 13530 + }, + { + "epoch": 0.13531, + "grad_norm": 1.4151078239343364, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 13531 + }, + { + "epoch": 0.13532, + "grad_norm": 1.0389544412766625, + "learning_rate": 0.003, + "loss": 4.099, + "step": 13532 + }, + { + "epoch": 0.13533, + "grad_norm": 1.1946282506821346, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 13533 + }, + { + "epoch": 0.13534, + "grad_norm": 1.2747275025398543, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 13534 + }, + { + "epoch": 0.13535, + "grad_norm": 1.1473872652809254, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13535 + }, + { + "epoch": 0.13536, + "grad_norm": 1.1398330387721207, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13536 + }, + { + "epoch": 0.13537, + "grad_norm": 1.437869757657236, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13537 + }, + { + "epoch": 0.13538, + "grad_norm": 1.0497105637254676, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 13538 + }, + { + "epoch": 0.13539, + "grad_norm": 1.1803193267606635, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13539 + }, + { + "epoch": 0.1354, + "grad_norm": 1.0150269320384628, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 13540 + }, + { + "epoch": 0.13541, + "grad_norm": 1.2965251686076023, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 13541 + }, + { + "epoch": 0.13542, + "grad_norm": 1.0030909278074205, + "learning_rate": 0.003, + "loss": 4.092, + "step": 13542 + }, + { + "epoch": 0.13543, + "grad_norm": 1.2468250804520131, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13543 + }, + { + "epoch": 0.13544, + "grad_norm": 1.0310918920010577, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13544 + }, + { + "epoch": 0.13545, + "grad_norm": 1.225367536185542, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 13545 + }, + { + "epoch": 0.13546, + "grad_norm": 1.0087671343539084, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 13546 + }, + { + "epoch": 0.13547, + "grad_norm": 1.3541882850349083, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 13547 + }, + { + "epoch": 0.13548, + "grad_norm": 1.1022860089932935, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13548 + }, + { + "epoch": 0.13549, + "grad_norm": 1.3990431601206805, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 13549 + }, + { + "epoch": 0.1355, + "grad_norm": 1.209753798907345, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 13550 + }, + { + "epoch": 0.13551, + "grad_norm": 1.2227913643789063, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13551 + }, + { + "epoch": 0.13552, + "grad_norm": 1.1239293994750092, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 13552 + }, + { + "epoch": 0.13553, + "grad_norm": 1.1809798944838255, + "learning_rate": 0.003, + "loss": 4.11, + "step": 13553 + }, + { + "epoch": 0.13554, + "grad_norm": 1.1052256243036818, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 13554 + }, + { + "epoch": 0.13555, + "grad_norm": 1.2460361801722424, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 13555 + }, + { + "epoch": 0.13556, + "grad_norm": 1.0764279421582164, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 13556 + }, + { + "epoch": 0.13557, + "grad_norm": 1.1564737937797744, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 13557 + }, + { + "epoch": 0.13558, + "grad_norm": 1.1661195493428866, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 13558 + }, + { + "epoch": 0.13559, + "grad_norm": 1.3724361794108662, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 13559 + }, + { + "epoch": 0.1356, + "grad_norm": 1.137116772236948, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 13560 + }, + { + "epoch": 0.13561, + "grad_norm": 1.3021859483441538, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13561 + }, + { + "epoch": 0.13562, + "grad_norm": 0.9913114792500876, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 13562 + }, + { + "epoch": 0.13563, + "grad_norm": 1.3501726665731697, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13563 + }, + { + "epoch": 0.13564, + "grad_norm": 1.1109480504606148, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13564 + }, + { + "epoch": 0.13565, + "grad_norm": 1.1932453808361558, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13565 + }, + { + "epoch": 0.13566, + "grad_norm": 1.2968291961796083, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 13566 + }, + { + "epoch": 0.13567, + "grad_norm": 0.8928736539916826, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13567 + }, + { + "epoch": 0.13568, + "grad_norm": 0.9474463361605496, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 13568 + }, + { + "epoch": 0.13569, + "grad_norm": 1.3066183160874658, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 13569 + }, + { + "epoch": 0.1357, + "grad_norm": 1.0298676653080354, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 13570 + }, + { + "epoch": 0.13571, + "grad_norm": 1.2340368544245897, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 13571 + }, + { + "epoch": 0.13572, + "grad_norm": 1.4403037500764257, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 13572 + }, + { + "epoch": 0.13573, + "grad_norm": 1.0939889578066653, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13573 + }, + { + "epoch": 0.13574, + "grad_norm": 1.3088511673177976, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13574 + }, + { + "epoch": 0.13575, + "grad_norm": 1.2768363323140435, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 13575 + }, + { + "epoch": 0.13576, + "grad_norm": 1.256661574995121, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13576 + }, + { + "epoch": 0.13577, + "grad_norm": 1.0588483065511913, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 13577 + }, + { + "epoch": 0.13578, + "grad_norm": 1.2630132494145923, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 13578 + }, + { + "epoch": 0.13579, + "grad_norm": 1.1653312093577335, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13579 + }, + { + "epoch": 0.1358, + "grad_norm": 1.223415623412137, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13580 + }, + { + "epoch": 0.13581, + "grad_norm": 1.3238249676184883, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 13581 + }, + { + "epoch": 0.13582, + "grad_norm": 1.03416907167867, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 13582 + }, + { + "epoch": 0.13583, + "grad_norm": 1.3594672938144423, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 13583 + }, + { + "epoch": 0.13584, + "grad_norm": 1.077123244373863, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13584 + }, + { + "epoch": 0.13585, + "grad_norm": 1.1672176890992012, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13585 + }, + { + "epoch": 0.13586, + "grad_norm": 1.2049053246218717, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 13586 + }, + { + "epoch": 0.13587, + "grad_norm": 1.3865760587332796, + "learning_rate": 0.003, + "loss": 4.062, + "step": 13587 + }, + { + "epoch": 0.13588, + "grad_norm": 1.1607163231788171, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13588 + }, + { + "epoch": 0.13589, + "grad_norm": 1.174397464749926, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13589 + }, + { + "epoch": 0.1359, + "grad_norm": 1.2792718395574298, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 13590 + }, + { + "epoch": 0.13591, + "grad_norm": 0.9721022972830015, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13591 + }, + { + "epoch": 0.13592, + "grad_norm": 1.365827999405511, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 13592 + }, + { + "epoch": 0.13593, + "grad_norm": 1.2376559086323935, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 13593 + }, + { + "epoch": 0.13594, + "grad_norm": 1.1000453997314037, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 13594 + }, + { + "epoch": 0.13595, + "grad_norm": 1.0605799757249121, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13595 + }, + { + "epoch": 0.13596, + "grad_norm": 1.0047523449138953, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13596 + }, + { + "epoch": 0.13597, + "grad_norm": 1.1212769914855838, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 13597 + }, + { + "epoch": 0.13598, + "grad_norm": 1.0643512786066576, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 13598 + }, + { + "epoch": 0.13599, + "grad_norm": 1.1137772972856694, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13599 + }, + { + "epoch": 0.136, + "grad_norm": 1.238310229737092, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13600 + }, + { + "epoch": 0.13601, + "grad_norm": 1.2288065213057453, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 13601 + }, + { + "epoch": 0.13602, + "grad_norm": 1.2127205294525991, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 13602 + }, + { + "epoch": 0.13603, + "grad_norm": 1.0688204869182307, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 13603 + }, + { + "epoch": 0.13604, + "grad_norm": 1.3238519061580016, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 13604 + }, + { + "epoch": 0.13605, + "grad_norm": 1.065978057108014, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13605 + }, + { + "epoch": 0.13606, + "grad_norm": 1.5113214195065328, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 13606 + }, + { + "epoch": 0.13607, + "grad_norm": 1.1835044192072945, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 13607 + }, + { + "epoch": 0.13608, + "grad_norm": 1.3389232557551585, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 13608 + }, + { + "epoch": 0.13609, + "grad_norm": 1.08573694872505, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13609 + }, + { + "epoch": 0.1361, + "grad_norm": 1.3183382408270754, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13610 + }, + { + "epoch": 0.13611, + "grad_norm": 1.088614267022385, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13611 + }, + { + "epoch": 0.13612, + "grad_norm": 1.584591299836842, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 13612 + }, + { + "epoch": 0.13613, + "grad_norm": 1.0241167509813345, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 13613 + }, + { + "epoch": 0.13614, + "grad_norm": 1.1410279990397971, + "learning_rate": 0.003, + "loss": 4.072, + "step": 13614 + }, + { + "epoch": 0.13615, + "grad_norm": 1.140403522308166, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 13615 + }, + { + "epoch": 0.13616, + "grad_norm": 1.1882408037137873, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 13616 + }, + { + "epoch": 0.13617, + "grad_norm": 1.2733678401188344, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13617 + }, + { + "epoch": 0.13618, + "grad_norm": 1.182541789342089, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 13618 + }, + { + "epoch": 0.13619, + "grad_norm": 1.8158182544769947, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 13619 + }, + { + "epoch": 0.1362, + "grad_norm": 0.9262895261876131, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 13620 + }, + { + "epoch": 0.13621, + "grad_norm": 1.0048294050605548, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13621 + }, + { + "epoch": 0.13622, + "grad_norm": 1.2932780705564038, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13622 + }, + { + "epoch": 0.13623, + "grad_norm": 1.0931089753864223, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 13623 + }, + { + "epoch": 0.13624, + "grad_norm": 1.2219390368450898, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13624 + }, + { + "epoch": 0.13625, + "grad_norm": 1.1703642070006044, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 13625 + }, + { + "epoch": 0.13626, + "grad_norm": 1.3588552313638005, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 13626 + }, + { + "epoch": 0.13627, + "grad_norm": 0.991279636178976, + "learning_rate": 0.003, + "loss": 4.056, + "step": 13627 + }, + { + "epoch": 0.13628, + "grad_norm": 1.2418972652541356, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 13628 + }, + { + "epoch": 0.13629, + "grad_norm": 0.9762054181088988, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 13629 + }, + { + "epoch": 0.1363, + "grad_norm": 1.0594583167811744, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 13630 + }, + { + "epoch": 0.13631, + "grad_norm": 1.3725521661001745, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 13631 + }, + { + "epoch": 0.13632, + "grad_norm": 0.9807824018898389, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 13632 + }, + { + "epoch": 0.13633, + "grad_norm": 1.2084168231669876, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13633 + }, + { + "epoch": 0.13634, + "grad_norm": 1.1830272923144818, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 13634 + }, + { + "epoch": 0.13635, + "grad_norm": 1.3078598922535416, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 13635 + }, + { + "epoch": 0.13636, + "grad_norm": 1.1001066668138082, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 13636 + }, + { + "epoch": 0.13637, + "grad_norm": 1.1638935784962923, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 13637 + }, + { + "epoch": 0.13638, + "grad_norm": 1.0385485798210434, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13638 + }, + { + "epoch": 0.13639, + "grad_norm": 1.1898560868781058, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 13639 + }, + { + "epoch": 0.1364, + "grad_norm": 1.0299533186076069, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 13640 + }, + { + "epoch": 0.13641, + "grad_norm": 1.2105958219454982, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 13641 + }, + { + "epoch": 0.13642, + "grad_norm": 1.2786757826782007, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 13642 + }, + { + "epoch": 0.13643, + "grad_norm": 1.107637562382635, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 13643 + }, + { + "epoch": 0.13644, + "grad_norm": 1.4947670952049972, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 13644 + }, + { + "epoch": 0.13645, + "grad_norm": 1.043630530838436, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 13645 + }, + { + "epoch": 0.13646, + "grad_norm": 1.1074987074965126, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13646 + }, + { + "epoch": 0.13647, + "grad_norm": 1.3183004868184747, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 13647 + }, + { + "epoch": 0.13648, + "grad_norm": 1.3334267070048087, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 13648 + }, + { + "epoch": 0.13649, + "grad_norm": 0.9769178281211383, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 13649 + }, + { + "epoch": 0.1365, + "grad_norm": 0.8940528609794745, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 13650 + }, + { + "epoch": 0.13651, + "grad_norm": 1.0203903879437717, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 13651 + }, + { + "epoch": 0.13652, + "grad_norm": 1.47641862136386, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13652 + }, + { + "epoch": 0.13653, + "grad_norm": 1.023482671599721, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 13653 + }, + { + "epoch": 0.13654, + "grad_norm": 1.4599717145775253, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 13654 + }, + { + "epoch": 0.13655, + "grad_norm": 1.1430378029023696, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13655 + }, + { + "epoch": 0.13656, + "grad_norm": 1.2479062092081508, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13656 + }, + { + "epoch": 0.13657, + "grad_norm": 1.1586196485600007, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 13657 + }, + { + "epoch": 0.13658, + "grad_norm": 1.2402177149121234, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 13658 + }, + { + "epoch": 0.13659, + "grad_norm": 1.1070055244833996, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13659 + }, + { + "epoch": 0.1366, + "grad_norm": 1.3387286615081702, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 13660 + }, + { + "epoch": 0.13661, + "grad_norm": 0.8802671899533874, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 13661 + }, + { + "epoch": 0.13662, + "grad_norm": 1.0961436556444828, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 13662 + }, + { + "epoch": 0.13663, + "grad_norm": 1.1659633698089749, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 13663 + }, + { + "epoch": 0.13664, + "grad_norm": 1.1625088068087674, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 13664 + }, + { + "epoch": 0.13665, + "grad_norm": 1.1952845956658587, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 13665 + }, + { + "epoch": 0.13666, + "grad_norm": 1.3022934851628816, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13666 + }, + { + "epoch": 0.13667, + "grad_norm": 1.1024907017076262, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 13667 + }, + { + "epoch": 0.13668, + "grad_norm": 1.1053766957068363, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 13668 + }, + { + "epoch": 0.13669, + "grad_norm": 1.0384195487713894, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 13669 + }, + { + "epoch": 0.1367, + "grad_norm": 1.654512344419585, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13670 + }, + { + "epoch": 0.13671, + "grad_norm": 1.3646809877437882, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 13671 + }, + { + "epoch": 0.13672, + "grad_norm": 1.0469603430029577, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13672 + }, + { + "epoch": 0.13673, + "grad_norm": 1.3395441818598068, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 13673 + }, + { + "epoch": 0.13674, + "grad_norm": 1.1458122349135758, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 13674 + }, + { + "epoch": 0.13675, + "grad_norm": 1.1271907544979884, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 13675 + }, + { + "epoch": 0.13676, + "grad_norm": 1.114889662851139, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13676 + }, + { + "epoch": 0.13677, + "grad_norm": 1.182067015827954, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 13677 + }, + { + "epoch": 0.13678, + "grad_norm": 1.115747324972659, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 13678 + }, + { + "epoch": 0.13679, + "grad_norm": 1.1011917999804097, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 13679 + }, + { + "epoch": 0.1368, + "grad_norm": 1.1142952525222676, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 13680 + }, + { + "epoch": 0.13681, + "grad_norm": 1.0676043236090158, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 13681 + }, + { + "epoch": 0.13682, + "grad_norm": 1.156689715671784, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 13682 + }, + { + "epoch": 0.13683, + "grad_norm": 1.1911333513095246, + "learning_rate": 0.003, + "loss": 4.067, + "step": 13683 + }, + { + "epoch": 0.13684, + "grad_norm": 1.2668964686340978, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 13684 + }, + { + "epoch": 0.13685, + "grad_norm": 1.2466397953616286, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 13685 + }, + { + "epoch": 0.13686, + "grad_norm": 1.064899811062785, + "learning_rate": 0.003, + "loss": 4.072, + "step": 13686 + }, + { + "epoch": 0.13687, + "grad_norm": 1.298345882337796, + "learning_rate": 0.003, + "loss": 4.065, + "step": 13687 + }, + { + "epoch": 0.13688, + "grad_norm": 1.0430799466261411, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13688 + }, + { + "epoch": 0.13689, + "grad_norm": 1.3666586493580102, + "learning_rate": 0.003, + "loss": 4.095, + "step": 13689 + }, + { + "epoch": 0.1369, + "grad_norm": 1.0370636511576221, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 13690 + }, + { + "epoch": 0.13691, + "grad_norm": 1.1996813727422466, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 13691 + }, + { + "epoch": 0.13692, + "grad_norm": 1.2216185605014462, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13692 + }, + { + "epoch": 0.13693, + "grad_norm": 1.2998804361911467, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 13693 + }, + { + "epoch": 0.13694, + "grad_norm": 1.017742494137553, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 13694 + }, + { + "epoch": 0.13695, + "grad_norm": 1.3318604210617848, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 13695 + }, + { + "epoch": 0.13696, + "grad_norm": 1.2027103780835313, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13696 + }, + { + "epoch": 0.13697, + "grad_norm": 1.2468422630000149, + "learning_rate": 0.003, + "loss": 4.048, + "step": 13697 + }, + { + "epoch": 0.13698, + "grad_norm": 1.1135879945021516, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 13698 + }, + { + "epoch": 0.13699, + "grad_norm": 1.33820725858327, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 13699 + }, + { + "epoch": 0.137, + "grad_norm": 1.115047651613133, + "learning_rate": 0.003, + "loss": 4.057, + "step": 13700 + }, + { + "epoch": 0.13701, + "grad_norm": 1.0425628725738034, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 13701 + }, + { + "epoch": 0.13702, + "grad_norm": 1.3163601188035263, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13702 + }, + { + "epoch": 0.13703, + "grad_norm": 1.078429519381817, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 13703 + }, + { + "epoch": 0.13704, + "grad_norm": 1.1876842179490408, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 13704 + }, + { + "epoch": 0.13705, + "grad_norm": 1.0233911271089657, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 13705 + }, + { + "epoch": 0.13706, + "grad_norm": 1.3133228580224459, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13706 + }, + { + "epoch": 0.13707, + "grad_norm": 0.876190413993581, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 13707 + }, + { + "epoch": 0.13708, + "grad_norm": 1.1224763276316478, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 13708 + }, + { + "epoch": 0.13709, + "grad_norm": 1.3941391601536905, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13709 + }, + { + "epoch": 0.1371, + "grad_norm": 1.3076888362270127, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 13710 + }, + { + "epoch": 0.13711, + "grad_norm": 1.2454693390615892, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 13711 + }, + { + "epoch": 0.13712, + "grad_norm": 1.3841143185327094, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 13712 + }, + { + "epoch": 0.13713, + "grad_norm": 1.2374480524788867, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 13713 + }, + { + "epoch": 0.13714, + "grad_norm": 1.1199426571591335, + "learning_rate": 0.003, + "loss": 4.061, + "step": 13714 + }, + { + "epoch": 0.13715, + "grad_norm": 1.0598577289200253, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 13715 + }, + { + "epoch": 0.13716, + "grad_norm": 1.2854522109147988, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 13716 + }, + { + "epoch": 0.13717, + "grad_norm": 0.9743690945461065, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 13717 + }, + { + "epoch": 0.13718, + "grad_norm": 1.4056823819656654, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 13718 + }, + { + "epoch": 0.13719, + "grad_norm": 1.2471282630318425, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 13719 + }, + { + "epoch": 0.1372, + "grad_norm": 1.1003533404374917, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13720 + }, + { + "epoch": 0.13721, + "grad_norm": 1.2369300729489772, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 13721 + }, + { + "epoch": 0.13722, + "grad_norm": 1.0168342614465826, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13722 + }, + { + "epoch": 0.13723, + "grad_norm": 1.2102545293950289, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 13723 + }, + { + "epoch": 0.13724, + "grad_norm": 1.0154559928282219, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 13724 + }, + { + "epoch": 0.13725, + "grad_norm": 1.3594160651554588, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 13725 + }, + { + "epoch": 0.13726, + "grad_norm": 0.8551637541897258, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 13726 + }, + { + "epoch": 0.13727, + "grad_norm": 1.017273693667817, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 13727 + }, + { + "epoch": 0.13728, + "grad_norm": 1.4490405897764906, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 13728 + }, + { + "epoch": 0.13729, + "grad_norm": 1.2732754804061845, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 13729 + }, + { + "epoch": 0.1373, + "grad_norm": 1.3607592856126363, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 13730 + }, + { + "epoch": 0.13731, + "grad_norm": 1.4082351054771112, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13731 + }, + { + "epoch": 0.13732, + "grad_norm": 1.416419540367466, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13732 + }, + { + "epoch": 0.13733, + "grad_norm": 1.2685878698524238, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 13733 + }, + { + "epoch": 0.13734, + "grad_norm": 1.207884398659683, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13734 + }, + { + "epoch": 0.13735, + "grad_norm": 1.10048154946459, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 13735 + }, + { + "epoch": 0.13736, + "grad_norm": 1.1877902199973267, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 13736 + }, + { + "epoch": 0.13737, + "grad_norm": 0.9533065968877271, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13737 + }, + { + "epoch": 0.13738, + "grad_norm": 1.1984993751824762, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 13738 + }, + { + "epoch": 0.13739, + "grad_norm": 1.1171422689859538, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13739 + }, + { + "epoch": 0.1374, + "grad_norm": 1.1911646618176164, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 13740 + }, + { + "epoch": 0.13741, + "grad_norm": 1.6440883465028686, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 13741 + }, + { + "epoch": 0.13742, + "grad_norm": 0.8918587280721783, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 13742 + }, + { + "epoch": 0.13743, + "grad_norm": 1.2441672699053672, + "learning_rate": 0.003, + "loss": 4.047, + "step": 13743 + }, + { + "epoch": 0.13744, + "grad_norm": 1.2075110129430175, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 13744 + }, + { + "epoch": 0.13745, + "grad_norm": 1.3501456347761631, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 13745 + }, + { + "epoch": 0.13746, + "grad_norm": 1.0841040758697305, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 13746 + }, + { + "epoch": 0.13747, + "grad_norm": 0.9834614633510931, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13747 + }, + { + "epoch": 0.13748, + "grad_norm": 1.276521807710027, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 13748 + }, + { + "epoch": 0.13749, + "grad_norm": 1.1909654186897163, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13749 + }, + { + "epoch": 0.1375, + "grad_norm": 1.2622123414540236, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 13750 + }, + { + "epoch": 0.13751, + "grad_norm": 1.0297963826688237, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 13751 + }, + { + "epoch": 0.13752, + "grad_norm": 1.2260649577899476, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13752 + }, + { + "epoch": 0.13753, + "grad_norm": 1.2058677000367255, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 13753 + }, + { + "epoch": 0.13754, + "grad_norm": 1.1248417684321146, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13754 + }, + { + "epoch": 0.13755, + "grad_norm": 1.2510522500262307, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 13755 + }, + { + "epoch": 0.13756, + "grad_norm": 1.1205481519257474, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13756 + }, + { + "epoch": 0.13757, + "grad_norm": 1.239229758552996, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 13757 + }, + { + "epoch": 0.13758, + "grad_norm": 1.0140626079928357, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 13758 + }, + { + "epoch": 0.13759, + "grad_norm": 1.3690637911522001, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 13759 + }, + { + "epoch": 0.1376, + "grad_norm": 0.9507319118956362, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 13760 + }, + { + "epoch": 0.13761, + "grad_norm": 1.2483510717007327, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13761 + }, + { + "epoch": 0.13762, + "grad_norm": 1.4389340645739723, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 13762 + }, + { + "epoch": 0.13763, + "grad_norm": 1.286178570489444, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 13763 + }, + { + "epoch": 0.13764, + "grad_norm": 1.1319149093379803, + "learning_rate": 0.003, + "loss": 4.095, + "step": 13764 + }, + { + "epoch": 0.13765, + "grad_norm": 1.2618229318089147, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 13765 + }, + { + "epoch": 0.13766, + "grad_norm": 1.1658637765032893, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 13766 + }, + { + "epoch": 0.13767, + "grad_norm": 1.18681907560757, + "learning_rate": 0.003, + "loss": 4.052, + "step": 13767 + }, + { + "epoch": 0.13768, + "grad_norm": 1.5055157878611105, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 13768 + }, + { + "epoch": 0.13769, + "grad_norm": 1.0051737452007148, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 13769 + }, + { + "epoch": 0.1377, + "grad_norm": 1.229234248921126, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13770 + }, + { + "epoch": 0.13771, + "grad_norm": 1.0040775586458914, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 13771 + }, + { + "epoch": 0.13772, + "grad_norm": 1.4109407557293536, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 13772 + }, + { + "epoch": 0.13773, + "grad_norm": 1.3305022959252513, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 13773 + }, + { + "epoch": 0.13774, + "grad_norm": 1.2595580781737046, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 13774 + }, + { + "epoch": 0.13775, + "grad_norm": 1.0225571862727028, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 13775 + }, + { + "epoch": 0.13776, + "grad_norm": 1.0741113836973304, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13776 + }, + { + "epoch": 0.13777, + "grad_norm": 1.0102454669852285, + "learning_rate": 0.003, + "loss": 4.047, + "step": 13777 + }, + { + "epoch": 0.13778, + "grad_norm": 1.298853127007687, + "learning_rate": 0.003, + "loss": 4.049, + "step": 13778 + }, + { + "epoch": 0.13779, + "grad_norm": 1.193382865103929, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 13779 + }, + { + "epoch": 0.1378, + "grad_norm": 1.6791272922067941, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 13780 + }, + { + "epoch": 0.13781, + "grad_norm": 1.1043189830034508, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 13781 + }, + { + "epoch": 0.13782, + "grad_norm": 1.0874160542628672, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 13782 + }, + { + "epoch": 0.13783, + "grad_norm": 1.2642927731824327, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 13783 + }, + { + "epoch": 0.13784, + "grad_norm": 1.010104274695837, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 13784 + }, + { + "epoch": 0.13785, + "grad_norm": 1.1285935067299098, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 13785 + }, + { + "epoch": 0.13786, + "grad_norm": 1.2335300965208245, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 13786 + }, + { + "epoch": 0.13787, + "grad_norm": 1.2342508255116131, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 13787 + }, + { + "epoch": 0.13788, + "grad_norm": 1.0984236689135678, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 13788 + }, + { + "epoch": 0.13789, + "grad_norm": 1.1947137028784856, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 13789 + }, + { + "epoch": 0.1379, + "grad_norm": 1.2858819082052524, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 13790 + }, + { + "epoch": 0.13791, + "grad_norm": 1.0612760980741958, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 13791 + }, + { + "epoch": 0.13792, + "grad_norm": 1.315051291837711, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 13792 + }, + { + "epoch": 0.13793, + "grad_norm": 1.2945660930882192, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13793 + }, + { + "epoch": 0.13794, + "grad_norm": 1.1064503623568218, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 13794 + }, + { + "epoch": 0.13795, + "grad_norm": 1.1363722862351189, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 13795 + }, + { + "epoch": 0.13796, + "grad_norm": 1.2670966849911767, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 13796 + }, + { + "epoch": 0.13797, + "grad_norm": 1.2022932708678868, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 13797 + }, + { + "epoch": 0.13798, + "grad_norm": 1.3045342419357953, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 13798 + }, + { + "epoch": 0.13799, + "grad_norm": 1.1236873174766076, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 13799 + }, + { + "epoch": 0.138, + "grad_norm": 1.3005949834912098, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 13800 + }, + { + "epoch": 0.13801, + "grad_norm": 1.053257699141777, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 13801 + }, + { + "epoch": 0.13802, + "grad_norm": 1.2482038722674353, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 13802 + }, + { + "epoch": 0.13803, + "grad_norm": 1.122974392828506, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13803 + }, + { + "epoch": 0.13804, + "grad_norm": 1.4588114593504007, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 13804 + }, + { + "epoch": 0.13805, + "grad_norm": 1.2518257379913156, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 13805 + }, + { + "epoch": 0.13806, + "grad_norm": 1.0687988535135047, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 13806 + }, + { + "epoch": 0.13807, + "grad_norm": 1.3961170728719694, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 13807 + }, + { + "epoch": 0.13808, + "grad_norm": 1.2103978593187432, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13808 + }, + { + "epoch": 0.13809, + "grad_norm": 1.2205250904177936, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 13809 + }, + { + "epoch": 0.1381, + "grad_norm": 1.144928087725286, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 13810 + }, + { + "epoch": 0.13811, + "grad_norm": 1.243183918413564, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 13811 + }, + { + "epoch": 0.13812, + "grad_norm": 1.1858159425116392, + "learning_rate": 0.003, + "loss": 4.082, + "step": 13812 + }, + { + "epoch": 0.13813, + "grad_norm": 1.2392840620813086, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 13813 + }, + { + "epoch": 0.13814, + "grad_norm": 1.1961079829938566, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 13814 + }, + { + "epoch": 0.13815, + "grad_norm": 1.2286148445830436, + "learning_rate": 0.003, + "loss": 4.044, + "step": 13815 + }, + { + "epoch": 0.13816, + "grad_norm": 1.059480739896095, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 13816 + }, + { + "epoch": 0.13817, + "grad_norm": 1.45459892002146, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 13817 + }, + { + "epoch": 0.13818, + "grad_norm": 0.8566130369181839, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 13818 + }, + { + "epoch": 0.13819, + "grad_norm": 1.1883964343787299, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 13819 + }, + { + "epoch": 0.1382, + "grad_norm": 1.1835827104338559, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 13820 + }, + { + "epoch": 0.13821, + "grad_norm": 1.2119907320687489, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13821 + }, + { + "epoch": 0.13822, + "grad_norm": 1.0867692976875394, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13822 + }, + { + "epoch": 0.13823, + "grad_norm": 1.0277948363309009, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 13823 + }, + { + "epoch": 0.13824, + "grad_norm": 1.3087394817360376, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 13824 + }, + { + "epoch": 0.13825, + "grad_norm": 1.3231505662393632, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 13825 + }, + { + "epoch": 0.13826, + "grad_norm": 1.0199844943105496, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 13826 + }, + { + "epoch": 0.13827, + "grad_norm": 1.2483024765730963, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 13827 + }, + { + "epoch": 0.13828, + "grad_norm": 0.9830237717557309, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 13828 + }, + { + "epoch": 0.13829, + "grad_norm": 1.1549798333543255, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 13829 + }, + { + "epoch": 0.1383, + "grad_norm": 1.2256141253549033, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13830 + }, + { + "epoch": 0.13831, + "grad_norm": 1.3893703266288122, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13831 + }, + { + "epoch": 0.13832, + "grad_norm": 1.04461634398326, + "learning_rate": 0.003, + "loss": 4.055, + "step": 13832 + }, + { + "epoch": 0.13833, + "grad_norm": 1.2570903342063884, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 13833 + }, + { + "epoch": 0.13834, + "grad_norm": 1.2108687512011658, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 13834 + }, + { + "epoch": 0.13835, + "grad_norm": 1.1861073808987619, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13835 + }, + { + "epoch": 0.13836, + "grad_norm": 1.2103893164541015, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 13836 + }, + { + "epoch": 0.13837, + "grad_norm": 1.1816006360159812, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 13837 + }, + { + "epoch": 0.13838, + "grad_norm": 1.0234266452644853, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 13838 + }, + { + "epoch": 0.13839, + "grad_norm": 1.1374835719429075, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 13839 + }, + { + "epoch": 0.1384, + "grad_norm": 1.113220292818902, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13840 + }, + { + "epoch": 0.13841, + "grad_norm": 1.2108498192604105, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 13841 + }, + { + "epoch": 0.13842, + "grad_norm": 1.182047080437874, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 13842 + }, + { + "epoch": 0.13843, + "grad_norm": 1.0365640423934581, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 13843 + }, + { + "epoch": 0.13844, + "grad_norm": 1.2120314765030384, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 13844 + }, + { + "epoch": 0.13845, + "grad_norm": 1.4391785218167055, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 13845 + }, + { + "epoch": 0.13846, + "grad_norm": 1.398202203052698, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 13846 + }, + { + "epoch": 0.13847, + "grad_norm": 1.2374107107581187, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 13847 + }, + { + "epoch": 0.13848, + "grad_norm": 1.142601397450018, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13848 + }, + { + "epoch": 0.13849, + "grad_norm": 1.4394250915044238, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13849 + }, + { + "epoch": 0.1385, + "grad_norm": 0.9071756635676798, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13850 + }, + { + "epoch": 0.13851, + "grad_norm": 1.155898767954042, + "learning_rate": 0.003, + "loss": 4.05, + "step": 13851 + }, + { + "epoch": 0.13852, + "grad_norm": 1.3854446050083606, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 13852 + }, + { + "epoch": 0.13853, + "grad_norm": 1.3192226385987496, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 13853 + }, + { + "epoch": 0.13854, + "grad_norm": 1.2084015385701228, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 13854 + }, + { + "epoch": 0.13855, + "grad_norm": 1.25301652300438, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13855 + }, + { + "epoch": 0.13856, + "grad_norm": 1.0665196097239276, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 13856 + }, + { + "epoch": 0.13857, + "grad_norm": 1.1618993258053154, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 13857 + }, + { + "epoch": 0.13858, + "grad_norm": 1.3561969206027245, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 13858 + }, + { + "epoch": 0.13859, + "grad_norm": 1.091509768863945, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 13859 + }, + { + "epoch": 0.1386, + "grad_norm": 1.1411802259550208, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 13860 + }, + { + "epoch": 0.13861, + "grad_norm": 1.1427719793225433, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 13861 + }, + { + "epoch": 0.13862, + "grad_norm": 1.1738927424268664, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 13862 + }, + { + "epoch": 0.13863, + "grad_norm": 1.1997101675962227, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 13863 + }, + { + "epoch": 0.13864, + "grad_norm": 1.1630761787979964, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13864 + }, + { + "epoch": 0.13865, + "grad_norm": 1.139098753069408, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 13865 + }, + { + "epoch": 0.13866, + "grad_norm": 1.1557888439300514, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 13866 + }, + { + "epoch": 0.13867, + "grad_norm": 1.2623839567530815, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 13867 + }, + { + "epoch": 0.13868, + "grad_norm": 1.1837947341599429, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13868 + }, + { + "epoch": 0.13869, + "grad_norm": 1.4933916759250025, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 13869 + }, + { + "epoch": 0.1387, + "grad_norm": 1.104622387180865, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13870 + }, + { + "epoch": 0.13871, + "grad_norm": 1.2509532855231675, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13871 + }, + { + "epoch": 0.13872, + "grad_norm": 1.3209278407919876, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 13872 + }, + { + "epoch": 0.13873, + "grad_norm": 1.1340281379719, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 13873 + }, + { + "epoch": 0.13874, + "grad_norm": 1.2491776894445892, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 13874 + }, + { + "epoch": 0.13875, + "grad_norm": 0.9903960367573045, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 13875 + }, + { + "epoch": 0.13876, + "grad_norm": 1.352282192489039, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 13876 + }, + { + "epoch": 0.13877, + "grad_norm": 1.0005486288387502, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 13877 + }, + { + "epoch": 0.13878, + "grad_norm": 1.4365204816386525, + "learning_rate": 0.003, + "loss": 4.053, + "step": 13878 + }, + { + "epoch": 0.13879, + "grad_norm": 0.8805469147799052, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 13879 + }, + { + "epoch": 0.1388, + "grad_norm": 1.4887720150086994, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 13880 + }, + { + "epoch": 0.13881, + "grad_norm": 1.410762056398085, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 13881 + }, + { + "epoch": 0.13882, + "grad_norm": 1.3053880278160495, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 13882 + }, + { + "epoch": 0.13883, + "grad_norm": 1.3465284752908813, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 13883 + }, + { + "epoch": 0.13884, + "grad_norm": 1.2563042257227426, + "learning_rate": 0.003, + "loss": 4.096, + "step": 13884 + }, + { + "epoch": 0.13885, + "grad_norm": 1.2592782036434007, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 13885 + }, + { + "epoch": 0.13886, + "grad_norm": 1.0859532416025257, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13886 + }, + { + "epoch": 0.13887, + "grad_norm": 1.491418353096719, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 13887 + }, + { + "epoch": 0.13888, + "grad_norm": 0.8654041738780719, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 13888 + }, + { + "epoch": 0.13889, + "grad_norm": 0.9210784584786764, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 13889 + }, + { + "epoch": 0.1389, + "grad_norm": 1.099693943923993, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 13890 + }, + { + "epoch": 0.13891, + "grad_norm": 1.1958793076649281, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 13891 + }, + { + "epoch": 0.13892, + "grad_norm": 1.0885220014006605, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 13892 + }, + { + "epoch": 0.13893, + "grad_norm": 1.2725398146978182, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 13893 + }, + { + "epoch": 0.13894, + "grad_norm": 1.2371715001604757, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 13894 + }, + { + "epoch": 0.13895, + "grad_norm": 1.4099389084174885, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 13895 + }, + { + "epoch": 0.13896, + "grad_norm": 1.0048792330671146, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 13896 + }, + { + "epoch": 0.13897, + "grad_norm": 1.1219108421533248, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 13897 + }, + { + "epoch": 0.13898, + "grad_norm": 1.1050201127531942, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 13898 + }, + { + "epoch": 0.13899, + "grad_norm": 1.3441208780893406, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13899 + }, + { + "epoch": 0.139, + "grad_norm": 1.0851640605237316, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 13900 + }, + { + "epoch": 0.13901, + "grad_norm": 1.3455747589350209, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 13901 + }, + { + "epoch": 0.13902, + "grad_norm": 1.189938938341629, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 13902 + }, + { + "epoch": 0.13903, + "grad_norm": 1.12316810081831, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13903 + }, + { + "epoch": 0.13904, + "grad_norm": 1.229408118975543, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 13904 + }, + { + "epoch": 0.13905, + "grad_norm": 1.0338220238270672, + "learning_rate": 0.003, + "loss": 4.085, + "step": 13905 + }, + { + "epoch": 0.13906, + "grad_norm": 1.320533597961583, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13906 + }, + { + "epoch": 0.13907, + "grad_norm": 1.0975575164129683, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 13907 + }, + { + "epoch": 0.13908, + "grad_norm": 1.4485545309639958, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 13908 + }, + { + "epoch": 0.13909, + "grad_norm": 1.1849479646868737, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 13909 + }, + { + "epoch": 0.1391, + "grad_norm": 1.3591596916616118, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 13910 + }, + { + "epoch": 0.13911, + "grad_norm": 1.193285184647018, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 13911 + }, + { + "epoch": 0.13912, + "grad_norm": 1.2602725825475507, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 13912 + }, + { + "epoch": 0.13913, + "grad_norm": 1.1496508245124075, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 13913 + }, + { + "epoch": 0.13914, + "grad_norm": 1.1308946707113137, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 13914 + }, + { + "epoch": 0.13915, + "grad_norm": 1.0958359046447819, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 13915 + }, + { + "epoch": 0.13916, + "grad_norm": 1.260606596749469, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 13916 + }, + { + "epoch": 0.13917, + "grad_norm": 1.0727415215522746, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 13917 + }, + { + "epoch": 0.13918, + "grad_norm": 1.1109603984637593, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 13918 + }, + { + "epoch": 0.13919, + "grad_norm": 1.2333451431442168, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 13919 + }, + { + "epoch": 0.1392, + "grad_norm": 1.4377985912806581, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13920 + }, + { + "epoch": 0.13921, + "grad_norm": 1.0271248694608675, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 13921 + }, + { + "epoch": 0.13922, + "grad_norm": 1.4524105427120009, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 13922 + }, + { + "epoch": 0.13923, + "grad_norm": 1.0931444063543732, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 13923 + }, + { + "epoch": 0.13924, + "grad_norm": 1.377676623237047, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 13924 + }, + { + "epoch": 0.13925, + "grad_norm": 1.1584891813511757, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 13925 + }, + { + "epoch": 0.13926, + "grad_norm": 1.464207200492954, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 13926 + }, + { + "epoch": 0.13927, + "grad_norm": 1.0267295476732703, + "learning_rate": 0.003, + "loss": 4.065, + "step": 13927 + }, + { + "epoch": 0.13928, + "grad_norm": 1.3087279077958458, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 13928 + }, + { + "epoch": 0.13929, + "grad_norm": 1.1030413931333924, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 13929 + }, + { + "epoch": 0.1393, + "grad_norm": 1.1963510192727531, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13930 + }, + { + "epoch": 0.13931, + "grad_norm": 1.3552900115306397, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13931 + }, + { + "epoch": 0.13932, + "grad_norm": 1.117443482134989, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 13932 + }, + { + "epoch": 0.13933, + "grad_norm": 1.4176894352836409, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 13933 + }, + { + "epoch": 0.13934, + "grad_norm": 0.9146997303307944, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 13934 + }, + { + "epoch": 0.13935, + "grad_norm": 1.1075004455081943, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 13935 + }, + { + "epoch": 0.13936, + "grad_norm": 1.3590514377512752, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 13936 + }, + { + "epoch": 0.13937, + "grad_norm": 1.2083401040806896, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 13937 + }, + { + "epoch": 0.13938, + "grad_norm": 0.9787905320757424, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13938 + }, + { + "epoch": 0.13939, + "grad_norm": 1.251635670336556, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 13939 + }, + { + "epoch": 0.1394, + "grad_norm": 1.3134870525772442, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 13940 + }, + { + "epoch": 0.13941, + "grad_norm": 1.2293564795195282, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13941 + }, + { + "epoch": 0.13942, + "grad_norm": 1.060690054358625, + "learning_rate": 0.003, + "loss": 4.073, + "step": 13942 + }, + { + "epoch": 0.13943, + "grad_norm": 1.090686083165746, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 13943 + }, + { + "epoch": 0.13944, + "grad_norm": 1.301586014251149, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 13944 + }, + { + "epoch": 0.13945, + "grad_norm": 0.9965137088015933, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 13945 + }, + { + "epoch": 0.13946, + "grad_norm": 1.3698183694761605, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 13946 + }, + { + "epoch": 0.13947, + "grad_norm": 1.0772867287069972, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 13947 + }, + { + "epoch": 0.13948, + "grad_norm": 1.3466737330984049, + "learning_rate": 0.003, + "loss": 4.089, + "step": 13948 + }, + { + "epoch": 0.13949, + "grad_norm": 1.231606661963063, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 13949 + }, + { + "epoch": 0.1395, + "grad_norm": 1.1341519641675695, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 13950 + }, + { + "epoch": 0.13951, + "grad_norm": 1.1325294230135632, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 13951 + }, + { + "epoch": 0.13952, + "grad_norm": 0.995793562007214, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 13952 + }, + { + "epoch": 0.13953, + "grad_norm": 1.0873814056283444, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 13953 + }, + { + "epoch": 0.13954, + "grad_norm": 1.0919384599880155, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13954 + }, + { + "epoch": 0.13955, + "grad_norm": 1.3122930160032409, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 13955 + }, + { + "epoch": 0.13956, + "grad_norm": 1.2936895953678114, + "learning_rate": 0.003, + "loss": 4.083, + "step": 13956 + }, + { + "epoch": 0.13957, + "grad_norm": 1.1638919567860373, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 13957 + }, + { + "epoch": 0.13958, + "grad_norm": 1.3219373768753289, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 13958 + }, + { + "epoch": 0.13959, + "grad_norm": 1.3180085740090488, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 13959 + }, + { + "epoch": 0.1396, + "grad_norm": 1.1477439233655948, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 13960 + }, + { + "epoch": 0.13961, + "grad_norm": 1.1681959988248531, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 13961 + }, + { + "epoch": 0.13962, + "grad_norm": 1.0433457251588483, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 13962 + }, + { + "epoch": 0.13963, + "grad_norm": 1.391823485061002, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 13963 + }, + { + "epoch": 0.13964, + "grad_norm": 1.0210707733989797, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13964 + }, + { + "epoch": 0.13965, + "grad_norm": 1.4127300935023928, + "learning_rate": 0.003, + "loss": 4.046, + "step": 13965 + }, + { + "epoch": 0.13966, + "grad_norm": 1.1296663061712948, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 13966 + }, + { + "epoch": 0.13967, + "grad_norm": 1.2884345926712524, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 13967 + }, + { + "epoch": 0.13968, + "grad_norm": 1.2284051362680708, + "learning_rate": 0.003, + "loss": 4.038, + "step": 13968 + }, + { + "epoch": 0.13969, + "grad_norm": 1.168348777197498, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 13969 + }, + { + "epoch": 0.1397, + "grad_norm": 1.244855212775038, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 13970 + }, + { + "epoch": 0.13971, + "grad_norm": 1.2413849979162377, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 13971 + }, + { + "epoch": 0.13972, + "grad_norm": 1.254429545205275, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 13972 + }, + { + "epoch": 0.13973, + "grad_norm": 1.0510032319940372, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 13973 + }, + { + "epoch": 0.13974, + "grad_norm": 1.240022161138721, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 13974 + }, + { + "epoch": 0.13975, + "grad_norm": 1.1236021708137216, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 13975 + }, + { + "epoch": 0.13976, + "grad_norm": 1.282349149962916, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 13976 + }, + { + "epoch": 0.13977, + "grad_norm": 1.273656352075562, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 13977 + }, + { + "epoch": 0.13978, + "grad_norm": 1.731791547986483, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 13978 + }, + { + "epoch": 0.13979, + "grad_norm": 0.9412192924359677, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 13979 + }, + { + "epoch": 0.1398, + "grad_norm": 1.1107332178250728, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 13980 + }, + { + "epoch": 0.13981, + "grad_norm": 1.616234813264336, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 13981 + }, + { + "epoch": 0.13982, + "grad_norm": 0.9711871798503602, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 13982 + }, + { + "epoch": 0.13983, + "grad_norm": 1.1601552691916304, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 13983 + }, + { + "epoch": 0.13984, + "grad_norm": 1.2597883044670368, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 13984 + }, + { + "epoch": 0.13985, + "grad_norm": 1.2256076832195404, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 13985 + }, + { + "epoch": 0.13986, + "grad_norm": 1.1600366740074843, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 13986 + }, + { + "epoch": 0.13987, + "grad_norm": 1.3271071059940935, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 13987 + }, + { + "epoch": 0.13988, + "grad_norm": 1.0963944376064725, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 13988 + }, + { + "epoch": 0.13989, + "grad_norm": 1.1583167783910462, + "learning_rate": 0.003, + "loss": 4.066, + "step": 13989 + }, + { + "epoch": 0.1399, + "grad_norm": 1.2763012872823292, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13990 + }, + { + "epoch": 0.13991, + "grad_norm": 1.1761647857400959, + "learning_rate": 0.003, + "loss": 4.082, + "step": 13991 + }, + { + "epoch": 0.13992, + "grad_norm": 1.284745873124637, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 13992 + }, + { + "epoch": 0.13993, + "grad_norm": 1.1017391051460177, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13993 + }, + { + "epoch": 0.13994, + "grad_norm": 1.2918818495727538, + "learning_rate": 0.003, + "loss": 4.075, + "step": 13994 + }, + { + "epoch": 0.13995, + "grad_norm": 1.015154283188355, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 13995 + }, + { + "epoch": 0.13996, + "grad_norm": 1.1298280746391378, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 13996 + }, + { + "epoch": 0.13997, + "grad_norm": 1.0422809770483068, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 13997 + }, + { + "epoch": 0.13998, + "grad_norm": 1.3048376609139016, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 13998 + }, + { + "epoch": 0.13999, + "grad_norm": 1.158199861113384, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 13999 + }, + { + "epoch": 0.14, + "grad_norm": 1.4235527130575911, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 14000 + }, + { + "epoch": 0.14001, + "grad_norm": 1.1218686754587923, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14001 + }, + { + "epoch": 0.14002, + "grad_norm": 1.073910869987476, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14002 + }, + { + "epoch": 0.14003, + "grad_norm": 1.341400623868328, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 14003 + }, + { + "epoch": 0.14004, + "grad_norm": 1.0756948186119029, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14004 + }, + { + "epoch": 0.14005, + "grad_norm": 1.2865374378748764, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 14005 + }, + { + "epoch": 0.14006, + "grad_norm": 1.0091892120642572, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14006 + }, + { + "epoch": 0.14007, + "grad_norm": 1.199666970298566, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14007 + }, + { + "epoch": 0.14008, + "grad_norm": 1.1846263521850238, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 14008 + }, + { + "epoch": 0.14009, + "grad_norm": 1.1943126712677172, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 14009 + }, + { + "epoch": 0.1401, + "grad_norm": 1.1383321883966389, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14010 + }, + { + "epoch": 0.14011, + "grad_norm": 1.2282899338116846, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 14011 + }, + { + "epoch": 0.14012, + "grad_norm": 1.0289904331903552, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14012 + }, + { + "epoch": 0.14013, + "grad_norm": 1.238851255206333, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 14013 + }, + { + "epoch": 0.14014, + "grad_norm": 1.235452537507704, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14014 + }, + { + "epoch": 0.14015, + "grad_norm": 1.1040128763356043, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 14015 + }, + { + "epoch": 0.14016, + "grad_norm": 1.3077179023264822, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 14016 + }, + { + "epoch": 0.14017, + "grad_norm": 1.070994763136842, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14017 + }, + { + "epoch": 0.14018, + "grad_norm": 1.5795010445329243, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 14018 + }, + { + "epoch": 0.14019, + "grad_norm": 1.110038500243441, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14019 + }, + { + "epoch": 0.1402, + "grad_norm": 1.282640788277247, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14020 + }, + { + "epoch": 0.14021, + "grad_norm": 1.297107847097581, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 14021 + }, + { + "epoch": 0.14022, + "grad_norm": 1.1023765488365382, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 14022 + }, + { + "epoch": 0.14023, + "grad_norm": 1.460130837344323, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 14023 + }, + { + "epoch": 0.14024, + "grad_norm": 1.2242712859475282, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14024 + }, + { + "epoch": 0.14025, + "grad_norm": 1.1691008234239406, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 14025 + }, + { + "epoch": 0.14026, + "grad_norm": 1.2835431334905392, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 14026 + }, + { + "epoch": 0.14027, + "grad_norm": 1.182980232383101, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 14027 + }, + { + "epoch": 0.14028, + "grad_norm": 1.2690266061757915, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14028 + }, + { + "epoch": 0.14029, + "grad_norm": 1.0544734376417177, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 14029 + }, + { + "epoch": 0.1403, + "grad_norm": 1.3424873029789186, + "learning_rate": 0.003, + "loss": 4.069, + "step": 14030 + }, + { + "epoch": 0.14031, + "grad_norm": 1.0456050868891826, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14031 + }, + { + "epoch": 0.14032, + "grad_norm": 1.261805056951264, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 14032 + }, + { + "epoch": 0.14033, + "grad_norm": 0.980685828701439, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14033 + }, + { + "epoch": 0.14034, + "grad_norm": 1.4253951982871123, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14034 + }, + { + "epoch": 0.14035, + "grad_norm": 1.090918477219447, + "learning_rate": 0.003, + "loss": 4.099, + "step": 14035 + }, + { + "epoch": 0.14036, + "grad_norm": 1.2198947919818297, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14036 + }, + { + "epoch": 0.14037, + "grad_norm": 1.1548789215429767, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 14037 + }, + { + "epoch": 0.14038, + "grad_norm": 1.194905070744935, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 14038 + }, + { + "epoch": 0.14039, + "grad_norm": 1.335307512644602, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 14039 + }, + { + "epoch": 0.1404, + "grad_norm": 1.3541627094142514, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 14040 + }, + { + "epoch": 0.14041, + "grad_norm": 1.2668978861497087, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 14041 + }, + { + "epoch": 0.14042, + "grad_norm": 1.0961704138846822, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14042 + }, + { + "epoch": 0.14043, + "grad_norm": 1.310633574915397, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 14043 + }, + { + "epoch": 0.14044, + "grad_norm": 1.3539709644080737, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 14044 + }, + { + "epoch": 0.14045, + "grad_norm": 1.1545400183981318, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 14045 + }, + { + "epoch": 0.14046, + "grad_norm": 1.3388500028882995, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14046 + }, + { + "epoch": 0.14047, + "grad_norm": 0.9722385197522674, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 14047 + }, + { + "epoch": 0.14048, + "grad_norm": 1.3184871593401073, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 14048 + }, + { + "epoch": 0.14049, + "grad_norm": 1.0804800429589527, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14049 + }, + { + "epoch": 0.1405, + "grad_norm": 1.4357860866189762, + "learning_rate": 0.003, + "loss": 4.081, + "step": 14050 + }, + { + "epoch": 0.14051, + "grad_norm": 1.1952804183318617, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 14051 + }, + { + "epoch": 0.14052, + "grad_norm": 1.4024191113504245, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 14052 + }, + { + "epoch": 0.14053, + "grad_norm": 1.0332715115039801, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 14053 + }, + { + "epoch": 0.14054, + "grad_norm": 1.2680229785791972, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 14054 + }, + { + "epoch": 0.14055, + "grad_norm": 1.220671894936996, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 14055 + }, + { + "epoch": 0.14056, + "grad_norm": 1.2233171041806348, + "learning_rate": 0.003, + "loss": 4.084, + "step": 14056 + }, + { + "epoch": 0.14057, + "grad_norm": 1.3157773723684758, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 14057 + }, + { + "epoch": 0.14058, + "grad_norm": 1.124405796378944, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14058 + }, + { + "epoch": 0.14059, + "grad_norm": 1.0499379920133691, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14059 + }, + { + "epoch": 0.1406, + "grad_norm": 1.3067068364078773, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14060 + }, + { + "epoch": 0.14061, + "grad_norm": 1.2410227155176823, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14061 + }, + { + "epoch": 0.14062, + "grad_norm": 1.0952452557827719, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 14062 + }, + { + "epoch": 0.14063, + "grad_norm": 1.1135075479715215, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 14063 + }, + { + "epoch": 0.14064, + "grad_norm": 1.1132489274054158, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14064 + }, + { + "epoch": 0.14065, + "grad_norm": 1.2691091680210476, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14065 + }, + { + "epoch": 0.14066, + "grad_norm": 1.1945059279642607, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14066 + }, + { + "epoch": 0.14067, + "grad_norm": 1.0849499469417416, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 14067 + }, + { + "epoch": 0.14068, + "grad_norm": 1.4788611387245632, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 14068 + }, + { + "epoch": 0.14069, + "grad_norm": 1.1060156621939714, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14069 + }, + { + "epoch": 0.1407, + "grad_norm": 1.2579130611915266, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14070 + }, + { + "epoch": 0.14071, + "grad_norm": 1.0615390463572114, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 14071 + }, + { + "epoch": 0.14072, + "grad_norm": 1.290325586202983, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14072 + }, + { + "epoch": 0.14073, + "grad_norm": 0.9851816152194139, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 14073 + }, + { + "epoch": 0.14074, + "grad_norm": 1.3244838979814133, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14074 + }, + { + "epoch": 0.14075, + "grad_norm": 0.9445330443948831, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 14075 + }, + { + "epoch": 0.14076, + "grad_norm": 1.1164429458611043, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14076 + }, + { + "epoch": 0.14077, + "grad_norm": 1.267983868741483, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 14077 + }, + { + "epoch": 0.14078, + "grad_norm": 1.1928263559105565, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 14078 + }, + { + "epoch": 0.14079, + "grad_norm": 1.122771365310409, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 14079 + }, + { + "epoch": 0.1408, + "grad_norm": 1.017592936371798, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 14080 + }, + { + "epoch": 0.14081, + "grad_norm": 1.449617758954173, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14081 + }, + { + "epoch": 0.14082, + "grad_norm": 1.1594842458389811, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 14082 + }, + { + "epoch": 0.14083, + "grad_norm": 1.257819534262536, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14083 + }, + { + "epoch": 0.14084, + "grad_norm": 1.381279265951169, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14084 + }, + { + "epoch": 0.14085, + "grad_norm": 1.0803274942689212, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 14085 + }, + { + "epoch": 0.14086, + "grad_norm": 1.1315871871624874, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14086 + }, + { + "epoch": 0.14087, + "grad_norm": 1.1748763201060617, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14087 + }, + { + "epoch": 0.14088, + "grad_norm": 1.2618259030019516, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14088 + }, + { + "epoch": 0.14089, + "grad_norm": 1.2343867938713182, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14089 + }, + { + "epoch": 0.1409, + "grad_norm": 0.9682727798543243, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 14090 + }, + { + "epoch": 0.14091, + "grad_norm": 1.4059866113262258, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 14091 + }, + { + "epoch": 0.14092, + "grad_norm": 1.054421179290126, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 14092 + }, + { + "epoch": 0.14093, + "grad_norm": 1.351264856046475, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14093 + }, + { + "epoch": 0.14094, + "grad_norm": 1.1926974079925088, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 14094 + }, + { + "epoch": 0.14095, + "grad_norm": 1.199123877993281, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 14095 + }, + { + "epoch": 0.14096, + "grad_norm": 1.416646770179057, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 14096 + }, + { + "epoch": 0.14097, + "grad_norm": 1.0840995665200288, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 14097 + }, + { + "epoch": 0.14098, + "grad_norm": 1.4593143604911771, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 14098 + }, + { + "epoch": 0.14099, + "grad_norm": 1.372636577899513, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 14099 + }, + { + "epoch": 0.141, + "grad_norm": 1.204041439925, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 14100 + }, + { + "epoch": 0.14101, + "grad_norm": 1.2853867147082851, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 14101 + }, + { + "epoch": 0.14102, + "grad_norm": 1.0729750244090508, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14102 + }, + { + "epoch": 0.14103, + "grad_norm": 1.328516026937791, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 14103 + }, + { + "epoch": 0.14104, + "grad_norm": 1.4265493338851565, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 14104 + }, + { + "epoch": 0.14105, + "grad_norm": 1.2046846966356402, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 14105 + }, + { + "epoch": 0.14106, + "grad_norm": 1.0158974556644869, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 14106 + }, + { + "epoch": 0.14107, + "grad_norm": 1.4414729489269356, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 14107 + }, + { + "epoch": 0.14108, + "grad_norm": 0.9747188424913261, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 14108 + }, + { + "epoch": 0.14109, + "grad_norm": 1.117458651506205, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14109 + }, + { + "epoch": 0.1411, + "grad_norm": 1.1707868125332495, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 14110 + }, + { + "epoch": 0.14111, + "grad_norm": 1.1546525831558394, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 14111 + }, + { + "epoch": 0.14112, + "grad_norm": 1.1526070159060042, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 14112 + }, + { + "epoch": 0.14113, + "grad_norm": 1.1098644078796183, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 14113 + }, + { + "epoch": 0.14114, + "grad_norm": 1.0113903919831815, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14114 + }, + { + "epoch": 0.14115, + "grad_norm": 1.2719335196897958, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 14115 + }, + { + "epoch": 0.14116, + "grad_norm": 1.0879888158201663, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 14116 + }, + { + "epoch": 0.14117, + "grad_norm": 1.4061852816368725, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14117 + }, + { + "epoch": 0.14118, + "grad_norm": 1.155127036615318, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 14118 + }, + { + "epoch": 0.14119, + "grad_norm": 1.4390313206171075, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 14119 + }, + { + "epoch": 0.1412, + "grad_norm": 1.2243146873664432, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 14120 + }, + { + "epoch": 0.14121, + "grad_norm": 1.3477283252085275, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14121 + }, + { + "epoch": 0.14122, + "grad_norm": 1.244334698337371, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 14122 + }, + { + "epoch": 0.14123, + "grad_norm": 1.1362973216272718, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14123 + }, + { + "epoch": 0.14124, + "grad_norm": 1.1199342054580403, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 14124 + }, + { + "epoch": 0.14125, + "grad_norm": 1.1351963229862383, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 14125 + }, + { + "epoch": 0.14126, + "grad_norm": 1.2885471864534528, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14126 + }, + { + "epoch": 0.14127, + "grad_norm": 1.0621581921969891, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14127 + }, + { + "epoch": 0.14128, + "grad_norm": 1.5233796698785904, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14128 + }, + { + "epoch": 0.14129, + "grad_norm": 1.0809104836398291, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 14129 + }, + { + "epoch": 0.1413, + "grad_norm": 1.2136014062527531, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 14130 + }, + { + "epoch": 0.14131, + "grad_norm": 1.3219523759291874, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14131 + }, + { + "epoch": 0.14132, + "grad_norm": 1.0915606916320713, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 14132 + }, + { + "epoch": 0.14133, + "grad_norm": 1.1791363407813595, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 14133 + }, + { + "epoch": 0.14134, + "grad_norm": 1.1719934454427836, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 14134 + }, + { + "epoch": 0.14135, + "grad_norm": 1.2671170437913437, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 14135 + }, + { + "epoch": 0.14136, + "grad_norm": 1.2002924475133558, + "learning_rate": 0.003, + "loss": 4.057, + "step": 14136 + }, + { + "epoch": 0.14137, + "grad_norm": 1.2703077591986773, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14137 + }, + { + "epoch": 0.14138, + "grad_norm": 1.2298921842905388, + "learning_rate": 0.003, + "loss": 4.077, + "step": 14138 + }, + { + "epoch": 0.14139, + "grad_norm": 1.423085309232018, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14139 + }, + { + "epoch": 0.1414, + "grad_norm": 1.0237604249414567, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 14140 + }, + { + "epoch": 0.14141, + "grad_norm": 1.2418351630645756, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14141 + }, + { + "epoch": 0.14142, + "grad_norm": 1.3108708654007848, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14142 + }, + { + "epoch": 0.14143, + "grad_norm": 1.1746333607980173, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 14143 + }, + { + "epoch": 0.14144, + "grad_norm": 1.2919370602961495, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 14144 + }, + { + "epoch": 0.14145, + "grad_norm": 1.020497759641275, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14145 + }, + { + "epoch": 0.14146, + "grad_norm": 1.366765879723746, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 14146 + }, + { + "epoch": 0.14147, + "grad_norm": 1.068908212949406, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 14147 + }, + { + "epoch": 0.14148, + "grad_norm": 1.0305235636939665, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 14148 + }, + { + "epoch": 0.14149, + "grad_norm": 1.2279229983109714, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 14149 + }, + { + "epoch": 0.1415, + "grad_norm": 1.0316938377792986, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14150 + }, + { + "epoch": 0.14151, + "grad_norm": 1.377595721666909, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 14151 + }, + { + "epoch": 0.14152, + "grad_norm": 1.1522508689722177, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14152 + }, + { + "epoch": 0.14153, + "grad_norm": 1.2934979529418558, + "learning_rate": 0.003, + "loss": 4.046, + "step": 14153 + }, + { + "epoch": 0.14154, + "grad_norm": 1.037657039612129, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 14154 + }, + { + "epoch": 0.14155, + "grad_norm": 1.4118107667174071, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 14155 + }, + { + "epoch": 0.14156, + "grad_norm": 0.9756006396870087, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 14156 + }, + { + "epoch": 0.14157, + "grad_norm": 1.19230919409468, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 14157 + }, + { + "epoch": 0.14158, + "grad_norm": 1.2705488618993246, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14158 + }, + { + "epoch": 0.14159, + "grad_norm": 1.1206704851826939, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14159 + }, + { + "epoch": 0.1416, + "grad_norm": 1.3111408934130488, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 14160 + }, + { + "epoch": 0.14161, + "grad_norm": 0.9504522977031813, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 14161 + }, + { + "epoch": 0.14162, + "grad_norm": 1.188242443855758, + "learning_rate": 0.003, + "loss": 4.055, + "step": 14162 + }, + { + "epoch": 0.14163, + "grad_norm": 1.2109152723685594, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14163 + }, + { + "epoch": 0.14164, + "grad_norm": 1.1758764862974531, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14164 + }, + { + "epoch": 0.14165, + "grad_norm": 1.1963761766831424, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14165 + }, + { + "epoch": 0.14166, + "grad_norm": 1.2296011716516042, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14166 + }, + { + "epoch": 0.14167, + "grad_norm": 1.1056886732813094, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 14167 + }, + { + "epoch": 0.14168, + "grad_norm": 1.1019334916953336, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14168 + }, + { + "epoch": 0.14169, + "grad_norm": 1.1489295640476946, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 14169 + }, + { + "epoch": 0.1417, + "grad_norm": 1.034958735378265, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 14170 + }, + { + "epoch": 0.14171, + "grad_norm": 1.164901638667101, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 14171 + }, + { + "epoch": 0.14172, + "grad_norm": 1.3051868501269035, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 14172 + }, + { + "epoch": 0.14173, + "grad_norm": 1.6416928856078798, + "learning_rate": 0.003, + "loss": 4.092, + "step": 14173 + }, + { + "epoch": 0.14174, + "grad_norm": 0.992803965415215, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14174 + }, + { + "epoch": 0.14175, + "grad_norm": 1.2586048849110307, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 14175 + }, + { + "epoch": 0.14176, + "grad_norm": 1.3636801383756292, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 14176 + }, + { + "epoch": 0.14177, + "grad_norm": 1.3046002495581674, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 14177 + }, + { + "epoch": 0.14178, + "grad_norm": 1.0368564735366113, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 14178 + }, + { + "epoch": 0.14179, + "grad_norm": 1.2192823265187933, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 14179 + }, + { + "epoch": 0.1418, + "grad_norm": 1.2397683265384485, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 14180 + }, + { + "epoch": 0.14181, + "grad_norm": 1.2463906763765813, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14181 + }, + { + "epoch": 0.14182, + "grad_norm": 1.280643109256286, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14182 + }, + { + "epoch": 0.14183, + "grad_norm": 1.1470938139266507, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14183 + }, + { + "epoch": 0.14184, + "grad_norm": 1.2085749166227202, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 14184 + }, + { + "epoch": 0.14185, + "grad_norm": 1.2226127200246437, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 14185 + }, + { + "epoch": 0.14186, + "grad_norm": 1.1675834090908574, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 14186 + }, + { + "epoch": 0.14187, + "grad_norm": 1.273509064648255, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 14187 + }, + { + "epoch": 0.14188, + "grad_norm": 1.220538086442717, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 14188 + }, + { + "epoch": 0.14189, + "grad_norm": 1.27082425692112, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 14189 + }, + { + "epoch": 0.1419, + "grad_norm": 1.2715796466945537, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 14190 + }, + { + "epoch": 0.14191, + "grad_norm": 1.0415153909241641, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 14191 + }, + { + "epoch": 0.14192, + "grad_norm": 1.2253256958990855, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 14192 + }, + { + "epoch": 0.14193, + "grad_norm": 1.03874298384842, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 14193 + }, + { + "epoch": 0.14194, + "grad_norm": 1.1986931317570426, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 14194 + }, + { + "epoch": 0.14195, + "grad_norm": 1.1557863525220406, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 14195 + }, + { + "epoch": 0.14196, + "grad_norm": 1.0770610880955502, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 14196 + }, + { + "epoch": 0.14197, + "grad_norm": 1.132517261030939, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 14197 + }, + { + "epoch": 0.14198, + "grad_norm": 1.3532234854028709, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 14198 + }, + { + "epoch": 0.14199, + "grad_norm": 1.1347193448807606, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14199 + }, + { + "epoch": 0.142, + "grad_norm": 1.3727967137741413, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 14200 + }, + { + "epoch": 0.14201, + "grad_norm": 1.0442182792604975, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 14201 + }, + { + "epoch": 0.14202, + "grad_norm": 1.2585233889585916, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 14202 + }, + { + "epoch": 0.14203, + "grad_norm": 1.3628434262296514, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 14203 + }, + { + "epoch": 0.14204, + "grad_norm": 0.9471467248764754, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 14204 + }, + { + "epoch": 0.14205, + "grad_norm": 1.2167492529019486, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14205 + }, + { + "epoch": 0.14206, + "grad_norm": 1.0440167721462932, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 14206 + }, + { + "epoch": 0.14207, + "grad_norm": 1.360618441047889, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 14207 + }, + { + "epoch": 0.14208, + "grad_norm": 1.2537038117019181, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14208 + }, + { + "epoch": 0.14209, + "grad_norm": 1.081791751264513, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14209 + }, + { + "epoch": 0.1421, + "grad_norm": 1.254534712247267, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 14210 + }, + { + "epoch": 0.14211, + "grad_norm": 1.5122240518817431, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14211 + }, + { + "epoch": 0.14212, + "grad_norm": 1.0112483003151058, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 14212 + }, + { + "epoch": 0.14213, + "grad_norm": 1.5511275971569494, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 14213 + }, + { + "epoch": 0.14214, + "grad_norm": 1.1564205582922507, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14214 + }, + { + "epoch": 0.14215, + "grad_norm": 1.4372825515145655, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 14215 + }, + { + "epoch": 0.14216, + "grad_norm": 1.1860158195555228, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 14216 + }, + { + "epoch": 0.14217, + "grad_norm": 1.3192104639179023, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14217 + }, + { + "epoch": 0.14218, + "grad_norm": 0.9797169344125047, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14218 + }, + { + "epoch": 0.14219, + "grad_norm": 1.455155509542232, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14219 + }, + { + "epoch": 0.1422, + "grad_norm": 1.192258717354992, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 14220 + }, + { + "epoch": 0.14221, + "grad_norm": 1.4572850914795752, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 14221 + }, + { + "epoch": 0.14222, + "grad_norm": 1.0724039986564513, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 14222 + }, + { + "epoch": 0.14223, + "grad_norm": 1.2557558241219466, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14223 + }, + { + "epoch": 0.14224, + "grad_norm": 1.0698137237090464, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 14224 + }, + { + "epoch": 0.14225, + "grad_norm": 1.359659837901035, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 14225 + }, + { + "epoch": 0.14226, + "grad_norm": 1.121514307256611, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14226 + }, + { + "epoch": 0.14227, + "grad_norm": 1.2797986124941525, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14227 + }, + { + "epoch": 0.14228, + "grad_norm": 1.2148723431841384, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 14228 + }, + { + "epoch": 0.14229, + "grad_norm": 1.7105353367182838, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14229 + }, + { + "epoch": 0.1423, + "grad_norm": 0.8998733141850936, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14230 + }, + { + "epoch": 0.14231, + "grad_norm": 1.0518522620748678, + "learning_rate": 0.003, + "loss": 4.089, + "step": 14231 + }, + { + "epoch": 0.14232, + "grad_norm": 1.350838942083396, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 14232 + }, + { + "epoch": 0.14233, + "grad_norm": 1.0179680902312946, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 14233 + }, + { + "epoch": 0.14234, + "grad_norm": 1.2698091867895445, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14234 + }, + { + "epoch": 0.14235, + "grad_norm": 1.2965428655396696, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 14235 + }, + { + "epoch": 0.14236, + "grad_norm": 1.0899343077133432, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14236 + }, + { + "epoch": 0.14237, + "grad_norm": 1.3340083275697776, + "learning_rate": 0.003, + "loss": 4.066, + "step": 14237 + }, + { + "epoch": 0.14238, + "grad_norm": 1.246926879826877, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 14238 + }, + { + "epoch": 0.14239, + "grad_norm": 1.1718419258346218, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 14239 + }, + { + "epoch": 0.1424, + "grad_norm": 1.325326309558883, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 14240 + }, + { + "epoch": 0.14241, + "grad_norm": 0.9120505784288971, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 14241 + }, + { + "epoch": 0.14242, + "grad_norm": 1.0852945738282669, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 14242 + }, + { + "epoch": 0.14243, + "grad_norm": 1.2867726876794063, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 14243 + }, + { + "epoch": 0.14244, + "grad_norm": 1.0743183760770276, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 14244 + }, + { + "epoch": 0.14245, + "grad_norm": 1.3665750595546649, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14245 + }, + { + "epoch": 0.14246, + "grad_norm": 1.235528909124935, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 14246 + }, + { + "epoch": 0.14247, + "grad_norm": 1.1471484614944625, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14247 + }, + { + "epoch": 0.14248, + "grad_norm": 1.1158396969741906, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14248 + }, + { + "epoch": 0.14249, + "grad_norm": 1.2451790763660837, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14249 + }, + { + "epoch": 0.1425, + "grad_norm": 1.1660921278797662, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 14250 + }, + { + "epoch": 0.14251, + "grad_norm": 1.337552507610175, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14251 + }, + { + "epoch": 0.14252, + "grad_norm": 1.1237501224202473, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14252 + }, + { + "epoch": 0.14253, + "grad_norm": 1.2401862552002867, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 14253 + }, + { + "epoch": 0.14254, + "grad_norm": 1.2076917825407354, + "learning_rate": 0.003, + "loss": 4.062, + "step": 14254 + }, + { + "epoch": 0.14255, + "grad_norm": 1.1075968289184035, + "learning_rate": 0.003, + "loss": 4.057, + "step": 14255 + }, + { + "epoch": 0.14256, + "grad_norm": 1.2753987269843883, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14256 + }, + { + "epoch": 0.14257, + "grad_norm": 1.1585523826878703, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 14257 + }, + { + "epoch": 0.14258, + "grad_norm": 1.5092524682885513, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 14258 + }, + { + "epoch": 0.14259, + "grad_norm": 0.9875483613399724, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 14259 + }, + { + "epoch": 0.1426, + "grad_norm": 1.326626446023979, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14260 + }, + { + "epoch": 0.14261, + "grad_norm": 1.1717803350308014, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14261 + }, + { + "epoch": 0.14262, + "grad_norm": 1.1204453946088764, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 14262 + }, + { + "epoch": 0.14263, + "grad_norm": 1.1459022877545624, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14263 + }, + { + "epoch": 0.14264, + "grad_norm": 1.0570187331453484, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14264 + }, + { + "epoch": 0.14265, + "grad_norm": 1.2507717034006678, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14265 + }, + { + "epoch": 0.14266, + "grad_norm": 1.2440790086756464, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 14266 + }, + { + "epoch": 0.14267, + "grad_norm": 1.1507090337948278, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14267 + }, + { + "epoch": 0.14268, + "grad_norm": 0.933785452272276, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14268 + }, + { + "epoch": 0.14269, + "grad_norm": 1.33321097143748, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 14269 + }, + { + "epoch": 0.1427, + "grad_norm": 1.0385155213078152, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14270 + }, + { + "epoch": 0.14271, + "grad_norm": 1.279339007087537, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14271 + }, + { + "epoch": 0.14272, + "grad_norm": 1.3206856987994857, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 14272 + }, + { + "epoch": 0.14273, + "grad_norm": 1.3968104502893839, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14273 + }, + { + "epoch": 0.14274, + "grad_norm": 1.233775105693856, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 14274 + }, + { + "epoch": 0.14275, + "grad_norm": 1.0425653767912715, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 14275 + }, + { + "epoch": 0.14276, + "grad_norm": 1.4118044096015665, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 14276 + }, + { + "epoch": 0.14277, + "grad_norm": 1.471296450615494, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 14277 + }, + { + "epoch": 0.14278, + "grad_norm": 1.1820443927859594, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14278 + }, + { + "epoch": 0.14279, + "grad_norm": 1.1725352181243902, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 14279 + }, + { + "epoch": 0.1428, + "grad_norm": 1.3310836236444723, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 14280 + }, + { + "epoch": 0.14281, + "grad_norm": 1.2274869974280136, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14281 + }, + { + "epoch": 0.14282, + "grad_norm": 1.1488388991529206, + "learning_rate": 0.003, + "loss": 4.077, + "step": 14282 + }, + { + "epoch": 0.14283, + "grad_norm": 1.0685417106519242, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14283 + }, + { + "epoch": 0.14284, + "grad_norm": 1.1333158664486487, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 14284 + }, + { + "epoch": 0.14285, + "grad_norm": 1.1609384114256422, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 14285 + }, + { + "epoch": 0.14286, + "grad_norm": 1.273127080220649, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 14286 + }, + { + "epoch": 0.14287, + "grad_norm": 1.592837902156321, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 14287 + }, + { + "epoch": 0.14288, + "grad_norm": 1.1149129312014674, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 14288 + }, + { + "epoch": 0.14289, + "grad_norm": 1.3242356847900765, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 14289 + }, + { + "epoch": 0.1429, + "grad_norm": 1.3328202982701185, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 14290 + }, + { + "epoch": 0.14291, + "grad_norm": 0.990076164424594, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14291 + }, + { + "epoch": 0.14292, + "grad_norm": 1.292375579717724, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14292 + }, + { + "epoch": 0.14293, + "grad_norm": 0.9475587721700227, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 14293 + }, + { + "epoch": 0.14294, + "grad_norm": 1.3188992823297991, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 14294 + }, + { + "epoch": 0.14295, + "grad_norm": 1.1802765919358251, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14295 + }, + { + "epoch": 0.14296, + "grad_norm": 1.470604664552602, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14296 + }, + { + "epoch": 0.14297, + "grad_norm": 1.2692534422761599, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 14297 + }, + { + "epoch": 0.14298, + "grad_norm": 1.190711026045534, + "learning_rate": 0.003, + "loss": 4.054, + "step": 14298 + }, + { + "epoch": 0.14299, + "grad_norm": 1.1527200359274339, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 14299 + }, + { + "epoch": 0.143, + "grad_norm": 1.2253939632615483, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 14300 + }, + { + "epoch": 0.14301, + "grad_norm": 1.1452804353444426, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 14301 + }, + { + "epoch": 0.14302, + "grad_norm": 1.2653095717075904, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 14302 + }, + { + "epoch": 0.14303, + "grad_norm": 1.2283386711500326, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14303 + }, + { + "epoch": 0.14304, + "grad_norm": 1.2260239140907923, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 14304 + }, + { + "epoch": 0.14305, + "grad_norm": 1.241379595591714, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14305 + }, + { + "epoch": 0.14306, + "grad_norm": 1.2089446349096906, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14306 + }, + { + "epoch": 0.14307, + "grad_norm": 1.317381627080003, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 14307 + }, + { + "epoch": 0.14308, + "grad_norm": 1.2697092334474644, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 14308 + }, + { + "epoch": 0.14309, + "grad_norm": 1.178546586293183, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 14309 + }, + { + "epoch": 0.1431, + "grad_norm": 1.1018919994675551, + "learning_rate": 0.003, + "loss": 4.056, + "step": 14310 + }, + { + "epoch": 0.14311, + "grad_norm": 1.087360520477749, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 14311 + }, + { + "epoch": 0.14312, + "grad_norm": 1.1353127920896684, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 14312 + }, + { + "epoch": 0.14313, + "grad_norm": 1.3300518405090815, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 14313 + }, + { + "epoch": 0.14314, + "grad_norm": 1.1134049920041054, + "learning_rate": 0.003, + "loss": 4.053, + "step": 14314 + }, + { + "epoch": 0.14315, + "grad_norm": 1.4393999140109497, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 14315 + }, + { + "epoch": 0.14316, + "grad_norm": 0.9685541012410513, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 14316 + }, + { + "epoch": 0.14317, + "grad_norm": 1.2922155314825943, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14317 + }, + { + "epoch": 0.14318, + "grad_norm": 1.0866212897670042, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14318 + }, + { + "epoch": 0.14319, + "grad_norm": 1.1904422597157893, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 14319 + }, + { + "epoch": 0.1432, + "grad_norm": 1.0813133187971877, + "learning_rate": 0.003, + "loss": 4.078, + "step": 14320 + }, + { + "epoch": 0.14321, + "grad_norm": 1.2934273949969206, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 14321 + }, + { + "epoch": 0.14322, + "grad_norm": 1.3390521441092138, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 14322 + }, + { + "epoch": 0.14323, + "grad_norm": 1.052371538996439, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 14323 + }, + { + "epoch": 0.14324, + "grad_norm": 1.5072536866034199, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 14324 + }, + { + "epoch": 0.14325, + "grad_norm": 1.2621890886460423, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14325 + }, + { + "epoch": 0.14326, + "grad_norm": 1.2722920221123368, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 14326 + }, + { + "epoch": 0.14327, + "grad_norm": 1.3018127539015547, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 14327 + }, + { + "epoch": 0.14328, + "grad_norm": 1.0657898922750024, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 14328 + }, + { + "epoch": 0.14329, + "grad_norm": 1.2160029214303807, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 14329 + }, + { + "epoch": 0.1433, + "grad_norm": 1.2569127274770167, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 14330 + }, + { + "epoch": 0.14331, + "grad_norm": 1.1996116025902857, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 14331 + }, + { + "epoch": 0.14332, + "grad_norm": 1.3788925110863324, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 14332 + }, + { + "epoch": 0.14333, + "grad_norm": 1.068258870600438, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 14333 + }, + { + "epoch": 0.14334, + "grad_norm": 1.4019931902382263, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 14334 + }, + { + "epoch": 0.14335, + "grad_norm": 1.0535736941749134, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 14335 + }, + { + "epoch": 0.14336, + "grad_norm": 1.306283302232701, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14336 + }, + { + "epoch": 0.14337, + "grad_norm": 1.111328959070111, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 14337 + }, + { + "epoch": 0.14338, + "grad_norm": 1.1975664423135413, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14338 + }, + { + "epoch": 0.14339, + "grad_norm": 1.2191116259359973, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 14339 + }, + { + "epoch": 0.1434, + "grad_norm": 1.1840868502775341, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 14340 + }, + { + "epoch": 0.14341, + "grad_norm": 1.137478876561676, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 14341 + }, + { + "epoch": 0.14342, + "grad_norm": 1.3700259954550276, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14342 + }, + { + "epoch": 0.14343, + "grad_norm": 1.170377210981172, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 14343 + }, + { + "epoch": 0.14344, + "grad_norm": 1.2818784705552841, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 14344 + }, + { + "epoch": 0.14345, + "grad_norm": 1.4317888747741843, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 14345 + }, + { + "epoch": 0.14346, + "grad_norm": 0.9983553444856835, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14346 + }, + { + "epoch": 0.14347, + "grad_norm": 1.356555777043041, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14347 + }, + { + "epoch": 0.14348, + "grad_norm": 1.143216391544617, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 14348 + }, + { + "epoch": 0.14349, + "grad_norm": 1.3525650486340823, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14349 + }, + { + "epoch": 0.1435, + "grad_norm": 1.0057561246421187, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14350 + }, + { + "epoch": 0.14351, + "grad_norm": 1.3963353599027775, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14351 + }, + { + "epoch": 0.14352, + "grad_norm": 1.330890632598169, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 14352 + }, + { + "epoch": 0.14353, + "grad_norm": 1.1516104014607385, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 14353 + }, + { + "epoch": 0.14354, + "grad_norm": 1.1232317363122273, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14354 + }, + { + "epoch": 0.14355, + "grad_norm": 1.1755740122698235, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 14355 + }, + { + "epoch": 0.14356, + "grad_norm": 1.3381647039895606, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14356 + }, + { + "epoch": 0.14357, + "grad_norm": 1.0638132963036395, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 14357 + }, + { + "epoch": 0.14358, + "grad_norm": 1.1919994563744336, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 14358 + }, + { + "epoch": 0.14359, + "grad_norm": 1.1802707096333935, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 14359 + }, + { + "epoch": 0.1436, + "grad_norm": 1.196571939548632, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14360 + }, + { + "epoch": 0.14361, + "grad_norm": 1.2288744185635048, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 14361 + }, + { + "epoch": 0.14362, + "grad_norm": 1.2168584885220672, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 14362 + }, + { + "epoch": 0.14363, + "grad_norm": 1.5046380694248807, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 14363 + }, + { + "epoch": 0.14364, + "grad_norm": 1.1395198564200926, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 14364 + }, + { + "epoch": 0.14365, + "grad_norm": 1.2341073392599913, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14365 + }, + { + "epoch": 0.14366, + "grad_norm": 1.3233230350475131, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14366 + }, + { + "epoch": 0.14367, + "grad_norm": 1.1290748432048177, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 14367 + }, + { + "epoch": 0.14368, + "grad_norm": 1.259093691839906, + "learning_rate": 0.003, + "loss": 4.062, + "step": 14368 + }, + { + "epoch": 0.14369, + "grad_norm": 1.3434206128946158, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14369 + }, + { + "epoch": 0.1437, + "grad_norm": 1.143664186598176, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 14370 + }, + { + "epoch": 0.14371, + "grad_norm": 1.2119997531247515, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14371 + }, + { + "epoch": 0.14372, + "grad_norm": 1.068786400245214, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 14372 + }, + { + "epoch": 0.14373, + "grad_norm": 1.2530697642056654, + "learning_rate": 0.003, + "loss": 4.085, + "step": 14373 + }, + { + "epoch": 0.14374, + "grad_norm": 1.065215088325605, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14374 + }, + { + "epoch": 0.14375, + "grad_norm": 1.2621075398736497, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 14375 + }, + { + "epoch": 0.14376, + "grad_norm": 1.1434925159815088, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14376 + }, + { + "epoch": 0.14377, + "grad_norm": 1.164602629665852, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14377 + }, + { + "epoch": 0.14378, + "grad_norm": 1.2341596077755344, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14378 + }, + { + "epoch": 0.14379, + "grad_norm": 1.4319774842479158, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 14379 + }, + { + "epoch": 0.1438, + "grad_norm": 1.181278783262906, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 14380 + }, + { + "epoch": 0.14381, + "grad_norm": 1.4463379635472426, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 14381 + }, + { + "epoch": 0.14382, + "grad_norm": 1.0250385679941718, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14382 + }, + { + "epoch": 0.14383, + "grad_norm": 1.3341913337356177, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14383 + }, + { + "epoch": 0.14384, + "grad_norm": 1.0578582594846084, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 14384 + }, + { + "epoch": 0.14385, + "grad_norm": 1.2006510088537283, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14385 + }, + { + "epoch": 0.14386, + "grad_norm": 1.1735284191020567, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 14386 + }, + { + "epoch": 0.14387, + "grad_norm": 1.3144827104637613, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 14387 + }, + { + "epoch": 0.14388, + "grad_norm": 1.234010698715963, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14388 + }, + { + "epoch": 0.14389, + "grad_norm": 1.3112248415874672, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 14389 + }, + { + "epoch": 0.1439, + "grad_norm": 0.9695468424280841, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 14390 + }, + { + "epoch": 0.14391, + "grad_norm": 1.4268333526903854, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 14391 + }, + { + "epoch": 0.14392, + "grad_norm": 1.079939896224488, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14392 + }, + { + "epoch": 0.14393, + "grad_norm": 1.2101070030796783, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 14393 + }, + { + "epoch": 0.14394, + "grad_norm": 1.2107748764865611, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 14394 + }, + { + "epoch": 0.14395, + "grad_norm": 1.1654521766245456, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 14395 + }, + { + "epoch": 0.14396, + "grad_norm": 1.2679245581546836, + "learning_rate": 0.003, + "loss": 4.082, + "step": 14396 + }, + { + "epoch": 0.14397, + "grad_norm": 1.2035374322115195, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14397 + }, + { + "epoch": 0.14398, + "grad_norm": 1.35203795746024, + "learning_rate": 0.003, + "loss": 4.06, + "step": 14398 + }, + { + "epoch": 0.14399, + "grad_norm": 1.095843910890968, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 14399 + }, + { + "epoch": 0.144, + "grad_norm": 1.1826386393611024, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 14400 + }, + { + "epoch": 0.14401, + "grad_norm": 1.3204081501624294, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 14401 + }, + { + "epoch": 0.14402, + "grad_norm": 1.1316878090286906, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 14402 + }, + { + "epoch": 0.14403, + "grad_norm": 1.27626622488073, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 14403 + }, + { + "epoch": 0.14404, + "grad_norm": 1.1663925536898734, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 14404 + }, + { + "epoch": 0.14405, + "grad_norm": 1.4300762716857476, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14405 + }, + { + "epoch": 0.14406, + "grad_norm": 0.9779121704087891, + "learning_rate": 0.003, + "loss": 4.091, + "step": 14406 + }, + { + "epoch": 0.14407, + "grad_norm": 1.1832335778071723, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14407 + }, + { + "epoch": 0.14408, + "grad_norm": 1.2603683727876147, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 14408 + }, + { + "epoch": 0.14409, + "grad_norm": 1.5610546074596634, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 14409 + }, + { + "epoch": 0.1441, + "grad_norm": 1.0511442118032344, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14410 + }, + { + "epoch": 0.14411, + "grad_norm": 1.3488788906343954, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 14411 + }, + { + "epoch": 0.14412, + "grad_norm": 1.3284511507449208, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14412 + }, + { + "epoch": 0.14413, + "grad_norm": 0.9772219897331645, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14413 + }, + { + "epoch": 0.14414, + "grad_norm": 1.2426001842398948, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14414 + }, + { + "epoch": 0.14415, + "grad_norm": 1.0887974596263341, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14415 + }, + { + "epoch": 0.14416, + "grad_norm": 1.16577074086037, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 14416 + }, + { + "epoch": 0.14417, + "grad_norm": 1.0636381968227409, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 14417 + }, + { + "epoch": 0.14418, + "grad_norm": 1.3337924665293321, + "learning_rate": 0.003, + "loss": 4.082, + "step": 14418 + }, + { + "epoch": 0.14419, + "grad_norm": 1.3674897024573358, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14419 + }, + { + "epoch": 0.1442, + "grad_norm": 1.0381744655777854, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 14420 + }, + { + "epoch": 0.14421, + "grad_norm": 1.229329190596495, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 14421 + }, + { + "epoch": 0.14422, + "grad_norm": 1.1932692329223549, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14422 + }, + { + "epoch": 0.14423, + "grad_norm": 1.2891152046270844, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 14423 + }, + { + "epoch": 0.14424, + "grad_norm": 1.2381292606460514, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14424 + }, + { + "epoch": 0.14425, + "grad_norm": 1.0439278430058745, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14425 + }, + { + "epoch": 0.14426, + "grad_norm": 1.297336142612756, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14426 + }, + { + "epoch": 0.14427, + "grad_norm": 1.0550985053749837, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 14427 + }, + { + "epoch": 0.14428, + "grad_norm": 1.2899119998999717, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 14428 + }, + { + "epoch": 0.14429, + "grad_norm": 1.1821243947260043, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 14429 + }, + { + "epoch": 0.1443, + "grad_norm": 1.30751709397349, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 14430 + }, + { + "epoch": 0.14431, + "grad_norm": 1.093150059274714, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 14431 + }, + { + "epoch": 0.14432, + "grad_norm": 1.3784750328904125, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 14432 + }, + { + "epoch": 0.14433, + "grad_norm": 1.08182832775297, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 14433 + }, + { + "epoch": 0.14434, + "grad_norm": 1.5049206866725384, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14434 + }, + { + "epoch": 0.14435, + "grad_norm": 1.317178243386282, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 14435 + }, + { + "epoch": 0.14436, + "grad_norm": 1.2227613692858288, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 14436 + }, + { + "epoch": 0.14437, + "grad_norm": 1.304179770019113, + "learning_rate": 0.003, + "loss": 4.072, + "step": 14437 + }, + { + "epoch": 0.14438, + "grad_norm": 1.141831161472869, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 14438 + }, + { + "epoch": 0.14439, + "grad_norm": 1.0747145230342643, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 14439 + }, + { + "epoch": 0.1444, + "grad_norm": 1.299113550200919, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14440 + }, + { + "epoch": 0.14441, + "grad_norm": 0.9820380080128326, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 14441 + }, + { + "epoch": 0.14442, + "grad_norm": 1.3189405568861905, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14442 + }, + { + "epoch": 0.14443, + "grad_norm": 1.1460870888331007, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 14443 + }, + { + "epoch": 0.14444, + "grad_norm": 1.3368224171260885, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 14444 + }, + { + "epoch": 0.14445, + "grad_norm": 1.291908146511478, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 14445 + }, + { + "epoch": 0.14446, + "grad_norm": 1.1620244454533144, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 14446 + }, + { + "epoch": 0.14447, + "grad_norm": 1.4377785575175646, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 14447 + }, + { + "epoch": 0.14448, + "grad_norm": 1.0584026951747234, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 14448 + }, + { + "epoch": 0.14449, + "grad_norm": 1.1819828467010947, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 14449 + }, + { + "epoch": 0.1445, + "grad_norm": 1.0794523137185799, + "learning_rate": 0.003, + "loss": 4.087, + "step": 14450 + }, + { + "epoch": 0.14451, + "grad_norm": 1.1818184068695379, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14451 + }, + { + "epoch": 0.14452, + "grad_norm": 1.2710421760018722, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 14452 + }, + { + "epoch": 0.14453, + "grad_norm": 1.1622127964996545, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 14453 + }, + { + "epoch": 0.14454, + "grad_norm": 1.4394457125821694, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 14454 + }, + { + "epoch": 0.14455, + "grad_norm": 1.4153297904619657, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 14455 + }, + { + "epoch": 0.14456, + "grad_norm": 1.322307430442675, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 14456 + }, + { + "epoch": 0.14457, + "grad_norm": 1.1002488156671828, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 14457 + }, + { + "epoch": 0.14458, + "grad_norm": 1.2714201699644943, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14458 + }, + { + "epoch": 0.14459, + "grad_norm": 1.0095749116294088, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14459 + }, + { + "epoch": 0.1446, + "grad_norm": 1.4375780380540866, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 14460 + }, + { + "epoch": 0.14461, + "grad_norm": 1.14707387495912, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 14461 + }, + { + "epoch": 0.14462, + "grad_norm": 1.0885637677444406, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 14462 + }, + { + "epoch": 0.14463, + "grad_norm": 1.1986989329327669, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 14463 + }, + { + "epoch": 0.14464, + "grad_norm": 1.0866883339116706, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14464 + }, + { + "epoch": 0.14465, + "grad_norm": 1.1770390100324601, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 14465 + }, + { + "epoch": 0.14466, + "grad_norm": 0.9286184501251965, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 14466 + }, + { + "epoch": 0.14467, + "grad_norm": 1.3163757201728299, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 14467 + }, + { + "epoch": 0.14468, + "grad_norm": 1.17131106150134, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14468 + }, + { + "epoch": 0.14469, + "grad_norm": 1.2925242105191204, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 14469 + }, + { + "epoch": 0.1447, + "grad_norm": 1.1042283407728772, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 14470 + }, + { + "epoch": 0.14471, + "grad_norm": 1.36741571361758, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14471 + }, + { + "epoch": 0.14472, + "grad_norm": 1.4041101426859173, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 14472 + }, + { + "epoch": 0.14473, + "grad_norm": 1.4291062487502415, + "learning_rate": 0.003, + "loss": 4.049, + "step": 14473 + }, + { + "epoch": 0.14474, + "grad_norm": 1.280473028989063, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 14474 + }, + { + "epoch": 0.14475, + "grad_norm": 1.0625179105244094, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 14475 + }, + { + "epoch": 0.14476, + "grad_norm": 1.328498228319439, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 14476 + }, + { + "epoch": 0.14477, + "grad_norm": 1.2486546627827975, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 14477 + }, + { + "epoch": 0.14478, + "grad_norm": 1.0105945330782855, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 14478 + }, + { + "epoch": 0.14479, + "grad_norm": 1.144118833057518, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 14479 + }, + { + "epoch": 0.1448, + "grad_norm": 1.2580688610346515, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14480 + }, + { + "epoch": 0.14481, + "grad_norm": 1.1638733787914244, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 14481 + }, + { + "epoch": 0.14482, + "grad_norm": 1.2923456312341166, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 14482 + }, + { + "epoch": 0.14483, + "grad_norm": 1.3403808718963135, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 14483 + }, + { + "epoch": 0.14484, + "grad_norm": 1.2968967075420859, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14484 + }, + { + "epoch": 0.14485, + "grad_norm": 1.5068711403378459, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14485 + }, + { + "epoch": 0.14486, + "grad_norm": 0.8011818290123951, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14486 + }, + { + "epoch": 0.14487, + "grad_norm": 0.9656806963890344, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14487 + }, + { + "epoch": 0.14488, + "grad_norm": 1.1437377945148044, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 14488 + }, + { + "epoch": 0.14489, + "grad_norm": 1.2191665583527924, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14489 + }, + { + "epoch": 0.1449, + "grad_norm": 1.0879306524530827, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14490 + }, + { + "epoch": 0.14491, + "grad_norm": 1.2127406424972869, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 14491 + }, + { + "epoch": 0.14492, + "grad_norm": 1.1569332906839882, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14492 + }, + { + "epoch": 0.14493, + "grad_norm": 1.2566899839962806, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14493 + }, + { + "epoch": 0.14494, + "grad_norm": 1.2080703694035553, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 14494 + }, + { + "epoch": 0.14495, + "grad_norm": 1.2112001686441412, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14495 + }, + { + "epoch": 0.14496, + "grad_norm": 1.2551719179102072, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 14496 + }, + { + "epoch": 0.14497, + "grad_norm": 1.1267006066327367, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14497 + }, + { + "epoch": 0.14498, + "grad_norm": 1.173983188727913, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 14498 + }, + { + "epoch": 0.14499, + "grad_norm": 1.299307892459785, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 14499 + }, + { + "epoch": 0.145, + "grad_norm": 1.2623974500837152, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 14500 + }, + { + "epoch": 0.14501, + "grad_norm": 1.2833983713516892, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14501 + }, + { + "epoch": 0.14502, + "grad_norm": 1.1527850471661591, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14502 + }, + { + "epoch": 0.14503, + "grad_norm": 1.3253761895311922, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 14503 + }, + { + "epoch": 0.14504, + "grad_norm": 1.302871022894076, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 14504 + }, + { + "epoch": 0.14505, + "grad_norm": 1.7242406969598858, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14505 + }, + { + "epoch": 0.14506, + "grad_norm": 1.0678467046870714, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 14506 + }, + { + "epoch": 0.14507, + "grad_norm": 1.3537159549441753, + "learning_rate": 0.003, + "loss": 4.046, + "step": 14507 + }, + { + "epoch": 0.14508, + "grad_norm": 1.3472589423143078, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 14508 + }, + { + "epoch": 0.14509, + "grad_norm": 1.1619264215472223, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 14509 + }, + { + "epoch": 0.1451, + "grad_norm": 1.1637797261869363, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 14510 + }, + { + "epoch": 0.14511, + "grad_norm": 1.1056960026610345, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 14511 + }, + { + "epoch": 0.14512, + "grad_norm": 1.3473753956609713, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 14512 + }, + { + "epoch": 0.14513, + "grad_norm": 1.3543840235294828, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14513 + }, + { + "epoch": 0.14514, + "grad_norm": 1.2032831118156042, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 14514 + }, + { + "epoch": 0.14515, + "grad_norm": 1.1997784406249512, + "learning_rate": 0.003, + "loss": 4.077, + "step": 14515 + }, + { + "epoch": 0.14516, + "grad_norm": 1.2434269634164505, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14516 + }, + { + "epoch": 0.14517, + "grad_norm": 1.1062979063611453, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 14517 + }, + { + "epoch": 0.14518, + "grad_norm": 1.1578608691856607, + "learning_rate": 0.003, + "loss": 4.081, + "step": 14518 + }, + { + "epoch": 0.14519, + "grad_norm": 1.0622328252719972, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14519 + }, + { + "epoch": 0.1452, + "grad_norm": 1.2651723404892912, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 14520 + }, + { + "epoch": 0.14521, + "grad_norm": 1.0940269947093042, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 14521 + }, + { + "epoch": 0.14522, + "grad_norm": 1.3365778696794945, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 14522 + }, + { + "epoch": 0.14523, + "grad_norm": 1.1536610931751508, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14523 + }, + { + "epoch": 0.14524, + "grad_norm": 1.1571296642834157, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 14524 + }, + { + "epoch": 0.14525, + "grad_norm": 1.181679668807532, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 14525 + }, + { + "epoch": 0.14526, + "grad_norm": 1.0646818438185597, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 14526 + }, + { + "epoch": 0.14527, + "grad_norm": 1.5444754672321084, + "learning_rate": 0.003, + "loss": 4.072, + "step": 14527 + }, + { + "epoch": 0.14528, + "grad_norm": 1.110119502820521, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 14528 + }, + { + "epoch": 0.14529, + "grad_norm": 1.3202879624253951, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14529 + }, + { + "epoch": 0.1453, + "grad_norm": 1.0652238096811364, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 14530 + }, + { + "epoch": 0.14531, + "grad_norm": 1.3146546875809235, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 14531 + }, + { + "epoch": 0.14532, + "grad_norm": 1.316481791494692, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 14532 + }, + { + "epoch": 0.14533, + "grad_norm": 1.195736762059849, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14533 + }, + { + "epoch": 0.14534, + "grad_norm": 1.3153453343964692, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 14534 + }, + { + "epoch": 0.14535, + "grad_norm": 1.1661697160363889, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 14535 + }, + { + "epoch": 0.14536, + "grad_norm": 1.2644956968313958, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 14536 + }, + { + "epoch": 0.14537, + "grad_norm": 1.1889914671754607, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 14537 + }, + { + "epoch": 0.14538, + "grad_norm": 1.2755526897119813, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14538 + }, + { + "epoch": 0.14539, + "grad_norm": 1.0977307408454553, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 14539 + }, + { + "epoch": 0.1454, + "grad_norm": 1.1425193492569852, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 14540 + }, + { + "epoch": 0.14541, + "grad_norm": 1.3654741018509804, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 14541 + }, + { + "epoch": 0.14542, + "grad_norm": 1.3513178652433542, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 14542 + }, + { + "epoch": 0.14543, + "grad_norm": 1.2139546213180623, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 14543 + }, + { + "epoch": 0.14544, + "grad_norm": 1.1412586493058874, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 14544 + }, + { + "epoch": 0.14545, + "grad_norm": 1.2505032723432452, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 14545 + }, + { + "epoch": 0.14546, + "grad_norm": 1.1090289144477474, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 14546 + }, + { + "epoch": 0.14547, + "grad_norm": 1.1512395586013198, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 14547 + }, + { + "epoch": 0.14548, + "grad_norm": 1.1796163733534573, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14548 + }, + { + "epoch": 0.14549, + "grad_norm": 1.171288245791135, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14549 + }, + { + "epoch": 0.1455, + "grad_norm": 1.0909536639205564, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14550 + }, + { + "epoch": 0.14551, + "grad_norm": 1.2379349235427564, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 14551 + }, + { + "epoch": 0.14552, + "grad_norm": 1.1946767662358404, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 14552 + }, + { + "epoch": 0.14553, + "grad_norm": 1.3425360041150831, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 14553 + }, + { + "epoch": 0.14554, + "grad_norm": 1.2701523855721424, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 14554 + }, + { + "epoch": 0.14555, + "grad_norm": 1.4560475136856261, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14555 + }, + { + "epoch": 0.14556, + "grad_norm": 1.0558293006911164, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 14556 + }, + { + "epoch": 0.14557, + "grad_norm": 1.277053527905692, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14557 + }, + { + "epoch": 0.14558, + "grad_norm": 1.0996275979016732, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14558 + }, + { + "epoch": 0.14559, + "grad_norm": 1.6489839246077922, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14559 + }, + { + "epoch": 0.1456, + "grad_norm": 0.8985364962976041, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 14560 + }, + { + "epoch": 0.14561, + "grad_norm": 1.0939749863890005, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 14561 + }, + { + "epoch": 0.14562, + "grad_norm": 1.2240643266714866, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 14562 + }, + { + "epoch": 0.14563, + "grad_norm": 1.2048737550612332, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14563 + }, + { + "epoch": 0.14564, + "grad_norm": 1.1849894366036393, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 14564 + }, + { + "epoch": 0.14565, + "grad_norm": 1.2868216966661061, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 14565 + }, + { + "epoch": 0.14566, + "grad_norm": 1.1401900748690066, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 14566 + }, + { + "epoch": 0.14567, + "grad_norm": 1.17255534523372, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 14567 + }, + { + "epoch": 0.14568, + "grad_norm": 1.3814064011696163, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14568 + }, + { + "epoch": 0.14569, + "grad_norm": 1.4975264483666402, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 14569 + }, + { + "epoch": 0.1457, + "grad_norm": 1.161954516152272, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14570 + }, + { + "epoch": 0.14571, + "grad_norm": 1.3083161772445326, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 14571 + }, + { + "epoch": 0.14572, + "grad_norm": 1.0661306176980925, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 14572 + }, + { + "epoch": 0.14573, + "grad_norm": 1.168348396939506, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 14573 + }, + { + "epoch": 0.14574, + "grad_norm": 1.2995611491474106, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14574 + }, + { + "epoch": 0.14575, + "grad_norm": 1.1051839229605225, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 14575 + }, + { + "epoch": 0.14576, + "grad_norm": 1.426897734370533, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14576 + }, + { + "epoch": 0.14577, + "grad_norm": 1.3561950697517535, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 14577 + }, + { + "epoch": 0.14578, + "grad_norm": 1.1606752158833817, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14578 + }, + { + "epoch": 0.14579, + "grad_norm": 1.0561133106875344, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14579 + }, + { + "epoch": 0.1458, + "grad_norm": 1.313880559739441, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14580 + }, + { + "epoch": 0.14581, + "grad_norm": 1.175241277341145, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14581 + }, + { + "epoch": 0.14582, + "grad_norm": 1.0354369918458097, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 14582 + }, + { + "epoch": 0.14583, + "grad_norm": 1.3564197408932643, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 14583 + }, + { + "epoch": 0.14584, + "grad_norm": 1.2592562167383652, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 14584 + }, + { + "epoch": 0.14585, + "grad_norm": 1.1655502695806124, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 14585 + }, + { + "epoch": 0.14586, + "grad_norm": 1.1270144122520829, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 14586 + }, + { + "epoch": 0.14587, + "grad_norm": 1.3758566632654936, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14587 + }, + { + "epoch": 0.14588, + "grad_norm": 1.2272085490990263, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 14588 + }, + { + "epoch": 0.14589, + "grad_norm": 1.2677737579782817, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 14589 + }, + { + "epoch": 0.1459, + "grad_norm": 1.1360459914972227, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 14590 + }, + { + "epoch": 0.14591, + "grad_norm": 1.651258332471172, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 14591 + }, + { + "epoch": 0.14592, + "grad_norm": 0.9382332788255714, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14592 + }, + { + "epoch": 0.14593, + "grad_norm": 1.3978775775037138, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 14593 + }, + { + "epoch": 0.14594, + "grad_norm": 1.2423191227807584, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 14594 + }, + { + "epoch": 0.14595, + "grad_norm": 1.2538985231476125, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14595 + }, + { + "epoch": 0.14596, + "grad_norm": 1.1603969821652953, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 14596 + }, + { + "epoch": 0.14597, + "grad_norm": 1.2309084376760162, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 14597 + }, + { + "epoch": 0.14598, + "grad_norm": 1.4019712882539683, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 14598 + }, + { + "epoch": 0.14599, + "grad_norm": 1.0781586012504811, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 14599 + }, + { + "epoch": 0.146, + "grad_norm": 1.2463531132231722, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 14600 + }, + { + "epoch": 0.14601, + "grad_norm": 1.095285384160531, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 14601 + }, + { + "epoch": 0.14602, + "grad_norm": 1.5438704468250124, + "learning_rate": 0.003, + "loss": 4.066, + "step": 14602 + }, + { + "epoch": 0.14603, + "grad_norm": 1.008786932453542, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 14603 + }, + { + "epoch": 0.14604, + "grad_norm": 1.3263740335801186, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14604 + }, + { + "epoch": 0.14605, + "grad_norm": 1.2035455321360693, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14605 + }, + { + "epoch": 0.14606, + "grad_norm": 1.1297030002565212, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 14606 + }, + { + "epoch": 0.14607, + "grad_norm": 1.1514523890993336, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 14607 + }, + { + "epoch": 0.14608, + "grad_norm": 1.1003428973761358, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 14608 + }, + { + "epoch": 0.14609, + "grad_norm": 1.3220980722298383, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 14609 + }, + { + "epoch": 0.1461, + "grad_norm": 0.9078562364085505, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 14610 + }, + { + "epoch": 0.14611, + "grad_norm": 1.4187892743750314, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 14611 + }, + { + "epoch": 0.14612, + "grad_norm": 1.1193931114685458, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 14612 + }, + { + "epoch": 0.14613, + "grad_norm": 1.3084042564846994, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 14613 + }, + { + "epoch": 0.14614, + "grad_norm": 1.217749523812658, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 14614 + }, + { + "epoch": 0.14615, + "grad_norm": 1.1171054826532747, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14615 + }, + { + "epoch": 0.14616, + "grad_norm": 1.1533725317270105, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 14616 + }, + { + "epoch": 0.14617, + "grad_norm": 1.2648670840148972, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14617 + }, + { + "epoch": 0.14618, + "grad_norm": 1.207641924938561, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 14618 + }, + { + "epoch": 0.14619, + "grad_norm": 1.2927644215866563, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 14619 + }, + { + "epoch": 0.1462, + "grad_norm": 1.2135111822809954, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 14620 + }, + { + "epoch": 0.14621, + "grad_norm": 1.3374066515976388, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 14621 + }, + { + "epoch": 0.14622, + "grad_norm": 1.1798176510548282, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14622 + }, + { + "epoch": 0.14623, + "grad_norm": 1.101241817333466, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 14623 + }, + { + "epoch": 0.14624, + "grad_norm": 1.3165157136298984, + "learning_rate": 0.003, + "loss": 4.086, + "step": 14624 + }, + { + "epoch": 0.14625, + "grad_norm": 1.204010145714381, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 14625 + }, + { + "epoch": 0.14626, + "grad_norm": 1.1970570695525438, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 14626 + }, + { + "epoch": 0.14627, + "grad_norm": 1.3101272713075305, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 14627 + }, + { + "epoch": 0.14628, + "grad_norm": 1.2438733569226708, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 14628 + }, + { + "epoch": 0.14629, + "grad_norm": 1.1144384917204837, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 14629 + }, + { + "epoch": 0.1463, + "grad_norm": 1.3231476742228823, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 14630 + }, + { + "epoch": 0.14631, + "grad_norm": 1.2251934200995147, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 14631 + }, + { + "epoch": 0.14632, + "grad_norm": 1.4149018197203078, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 14632 + }, + { + "epoch": 0.14633, + "grad_norm": 1.111321909938563, + "learning_rate": 0.003, + "loss": 4.06, + "step": 14633 + }, + { + "epoch": 0.14634, + "grad_norm": 1.4009877478969213, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 14634 + }, + { + "epoch": 0.14635, + "grad_norm": 0.9757433393240569, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14635 + }, + { + "epoch": 0.14636, + "grad_norm": 1.5448628476612054, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14636 + }, + { + "epoch": 0.14637, + "grad_norm": 1.0681842219058078, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 14637 + }, + { + "epoch": 0.14638, + "grad_norm": 1.270278890838018, + "learning_rate": 0.003, + "loss": 4.049, + "step": 14638 + }, + { + "epoch": 0.14639, + "grad_norm": 1.1293518586686984, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 14639 + }, + { + "epoch": 0.1464, + "grad_norm": 1.2230323646100287, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 14640 + }, + { + "epoch": 0.14641, + "grad_norm": 0.9605823373334033, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 14641 + }, + { + "epoch": 0.14642, + "grad_norm": 1.395513703350284, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14642 + }, + { + "epoch": 0.14643, + "grad_norm": 1.111260789455305, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14643 + }, + { + "epoch": 0.14644, + "grad_norm": 1.3088383951677751, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 14644 + }, + { + "epoch": 0.14645, + "grad_norm": 1.1924587194471679, + "learning_rate": 0.003, + "loss": 4.085, + "step": 14645 + }, + { + "epoch": 0.14646, + "grad_norm": 1.0230377961179655, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14646 + }, + { + "epoch": 0.14647, + "grad_norm": 1.3166502301604532, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 14647 + }, + { + "epoch": 0.14648, + "grad_norm": 1.108098794298806, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 14648 + }, + { + "epoch": 0.14649, + "grad_norm": 1.2678976831802262, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 14649 + }, + { + "epoch": 0.1465, + "grad_norm": 1.0779314944871512, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14650 + }, + { + "epoch": 0.14651, + "grad_norm": 1.3070108905573998, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 14651 + }, + { + "epoch": 0.14652, + "grad_norm": 1.2173180299324342, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 14652 + }, + { + "epoch": 0.14653, + "grad_norm": 1.3059875979156272, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 14653 + }, + { + "epoch": 0.14654, + "grad_norm": 1.396883465923019, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 14654 + }, + { + "epoch": 0.14655, + "grad_norm": 1.0141344618956727, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 14655 + }, + { + "epoch": 0.14656, + "grad_norm": 1.4713846091980216, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14656 + }, + { + "epoch": 0.14657, + "grad_norm": 1.083585755943934, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 14657 + }, + { + "epoch": 0.14658, + "grad_norm": 1.2032480378616042, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14658 + }, + { + "epoch": 0.14659, + "grad_norm": 1.1574205075152437, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14659 + }, + { + "epoch": 0.1466, + "grad_norm": 1.1533865134835117, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 14660 + }, + { + "epoch": 0.14661, + "grad_norm": 1.1557425319475578, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14661 + }, + { + "epoch": 0.14662, + "grad_norm": 1.2140824735376958, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 14662 + }, + { + "epoch": 0.14663, + "grad_norm": 1.3119847972636143, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 14663 + }, + { + "epoch": 0.14664, + "grad_norm": 1.278637541443759, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 14664 + }, + { + "epoch": 0.14665, + "grad_norm": 1.5537971245262678, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 14665 + }, + { + "epoch": 0.14666, + "grad_norm": 1.0543789191944573, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 14666 + }, + { + "epoch": 0.14667, + "grad_norm": 1.321352396499266, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14667 + }, + { + "epoch": 0.14668, + "grad_norm": 1.452374422058879, + "learning_rate": 0.003, + "loss": 4.104, + "step": 14668 + }, + { + "epoch": 0.14669, + "grad_norm": 0.9483661975026233, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 14669 + }, + { + "epoch": 0.1467, + "grad_norm": 1.2457290239459582, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14670 + }, + { + "epoch": 0.14671, + "grad_norm": 1.166673424790351, + "learning_rate": 0.003, + "loss": 4.092, + "step": 14671 + }, + { + "epoch": 0.14672, + "grad_norm": 1.134676582970826, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 14672 + }, + { + "epoch": 0.14673, + "grad_norm": 1.4049530615079515, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 14673 + }, + { + "epoch": 0.14674, + "grad_norm": 1.0791538339719724, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14674 + }, + { + "epoch": 0.14675, + "grad_norm": 1.3387749229849049, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 14675 + }, + { + "epoch": 0.14676, + "grad_norm": 1.0816776734773716, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 14676 + }, + { + "epoch": 0.14677, + "grad_norm": 1.3010340266823335, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 14677 + }, + { + "epoch": 0.14678, + "grad_norm": 1.273904467089526, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14678 + }, + { + "epoch": 0.14679, + "grad_norm": 1.2050737400060467, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 14679 + }, + { + "epoch": 0.1468, + "grad_norm": 1.1782396903671484, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 14680 + }, + { + "epoch": 0.14681, + "grad_norm": 1.2641672769658576, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14681 + }, + { + "epoch": 0.14682, + "grad_norm": 1.3683531524304238, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 14682 + }, + { + "epoch": 0.14683, + "grad_norm": 1.1681281783864517, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 14683 + }, + { + "epoch": 0.14684, + "grad_norm": 1.2147665860320387, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 14684 + }, + { + "epoch": 0.14685, + "grad_norm": 1.1383565292447009, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 14685 + }, + { + "epoch": 0.14686, + "grad_norm": 1.2606728631128676, + "learning_rate": 0.003, + "loss": 4.045, + "step": 14686 + }, + { + "epoch": 0.14687, + "grad_norm": 1.2339376261541728, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 14687 + }, + { + "epoch": 0.14688, + "grad_norm": 1.1587680031981418, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 14688 + }, + { + "epoch": 0.14689, + "grad_norm": 1.1302148406483674, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 14689 + }, + { + "epoch": 0.1469, + "grad_norm": 1.160721415065776, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 14690 + }, + { + "epoch": 0.14691, + "grad_norm": 1.2096541604524635, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 14691 + }, + { + "epoch": 0.14692, + "grad_norm": 1.098711790790942, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14692 + }, + { + "epoch": 0.14693, + "grad_norm": 1.5608335345805666, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 14693 + }, + { + "epoch": 0.14694, + "grad_norm": 1.3805529775587744, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 14694 + }, + { + "epoch": 0.14695, + "grad_norm": 1.1496335459014944, + "learning_rate": 0.003, + "loss": 4.093, + "step": 14695 + }, + { + "epoch": 0.14696, + "grad_norm": 1.2934564888383862, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 14696 + }, + { + "epoch": 0.14697, + "grad_norm": 1.0298392730293933, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14697 + }, + { + "epoch": 0.14698, + "grad_norm": 1.2493236769696854, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 14698 + }, + { + "epoch": 0.14699, + "grad_norm": 1.1966206064149618, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 14699 + }, + { + "epoch": 0.147, + "grad_norm": 1.2008163126396347, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 14700 + }, + { + "epoch": 0.14701, + "grad_norm": 1.307359185613977, + "learning_rate": 0.003, + "loss": 4.062, + "step": 14701 + }, + { + "epoch": 0.14702, + "grad_norm": 1.17130778803398, + "learning_rate": 0.003, + "loss": 4.064, + "step": 14702 + }, + { + "epoch": 0.14703, + "grad_norm": 0.9278549741557135, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 14703 + }, + { + "epoch": 0.14704, + "grad_norm": 1.2416731233037417, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 14704 + }, + { + "epoch": 0.14705, + "grad_norm": 1.443478884981792, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 14705 + }, + { + "epoch": 0.14706, + "grad_norm": 1.169988258938255, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 14706 + }, + { + "epoch": 0.14707, + "grad_norm": 1.3884268894595029, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 14707 + }, + { + "epoch": 0.14708, + "grad_norm": 1.2376664296974007, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 14708 + }, + { + "epoch": 0.14709, + "grad_norm": 1.3585251644666752, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14709 + }, + { + "epoch": 0.1471, + "grad_norm": 1.1070441241571778, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 14710 + }, + { + "epoch": 0.14711, + "grad_norm": 1.209342609620908, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 14711 + }, + { + "epoch": 0.14712, + "grad_norm": 1.1481596216904917, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14712 + }, + { + "epoch": 0.14713, + "grad_norm": 1.362689261051619, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 14713 + }, + { + "epoch": 0.14714, + "grad_norm": 1.1744847375395815, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 14714 + }, + { + "epoch": 0.14715, + "grad_norm": 1.2132539753611855, + "learning_rate": 0.003, + "loss": 4.091, + "step": 14715 + }, + { + "epoch": 0.14716, + "grad_norm": 1.3284941942614448, + "learning_rate": 0.003, + "loss": 4.102, + "step": 14716 + }, + { + "epoch": 0.14717, + "grad_norm": 0.9466611944947958, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 14717 + }, + { + "epoch": 0.14718, + "grad_norm": 1.2938546324781965, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 14718 + }, + { + "epoch": 0.14719, + "grad_norm": 1.1099514972231117, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 14719 + }, + { + "epoch": 0.1472, + "grad_norm": 1.3147842488951789, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 14720 + }, + { + "epoch": 0.14721, + "grad_norm": 1.3027693784383552, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 14721 + }, + { + "epoch": 0.14722, + "grad_norm": 1.1463487338501712, + "learning_rate": 0.003, + "loss": 4.08, + "step": 14722 + }, + { + "epoch": 0.14723, + "grad_norm": 1.244333943250865, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 14723 + }, + { + "epoch": 0.14724, + "grad_norm": 1.1755040875323115, + "learning_rate": 0.003, + "loss": 4.07, + "step": 14724 + }, + { + "epoch": 0.14725, + "grad_norm": 1.3023666248547623, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 14725 + }, + { + "epoch": 0.14726, + "grad_norm": 1.1758954249843652, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14726 + }, + { + "epoch": 0.14727, + "grad_norm": 1.4709314206982071, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14727 + }, + { + "epoch": 0.14728, + "grad_norm": 0.9819314270248706, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 14728 + }, + { + "epoch": 0.14729, + "grad_norm": 1.2375661512872944, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 14729 + }, + { + "epoch": 0.1473, + "grad_norm": 1.2885113061978055, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 14730 + }, + { + "epoch": 0.14731, + "grad_norm": 1.0505392875289823, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 14731 + }, + { + "epoch": 0.14732, + "grad_norm": 1.5160304880082762, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 14732 + }, + { + "epoch": 0.14733, + "grad_norm": 1.0131695855765106, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14733 + }, + { + "epoch": 0.14734, + "grad_norm": 1.2967729921181999, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 14734 + }, + { + "epoch": 0.14735, + "grad_norm": 1.154259022839686, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14735 + }, + { + "epoch": 0.14736, + "grad_norm": 1.1909045056718182, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14736 + }, + { + "epoch": 0.14737, + "grad_norm": 1.2765410017020464, + "learning_rate": 0.003, + "loss": 4.064, + "step": 14737 + }, + { + "epoch": 0.14738, + "grad_norm": 1.1805332411128984, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 14738 + }, + { + "epoch": 0.14739, + "grad_norm": 1.2642126914844365, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 14739 + }, + { + "epoch": 0.1474, + "grad_norm": 1.0699407445012783, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 14740 + }, + { + "epoch": 0.14741, + "grad_norm": 1.1499692761506615, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 14741 + }, + { + "epoch": 0.14742, + "grad_norm": 1.0295183618628216, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 14742 + }, + { + "epoch": 0.14743, + "grad_norm": 1.3084781291711867, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 14743 + }, + { + "epoch": 0.14744, + "grad_norm": 1.0996251198527003, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 14744 + }, + { + "epoch": 0.14745, + "grad_norm": 1.134401093824059, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 14745 + }, + { + "epoch": 0.14746, + "grad_norm": 1.2803354774841316, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 14746 + }, + { + "epoch": 0.14747, + "grad_norm": 1.3988377679385435, + "learning_rate": 0.003, + "loss": 4.077, + "step": 14747 + }, + { + "epoch": 0.14748, + "grad_norm": 1.0568472895986643, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14748 + }, + { + "epoch": 0.14749, + "grad_norm": 1.4024469804051825, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14749 + }, + { + "epoch": 0.1475, + "grad_norm": 1.1838146800837628, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 14750 + }, + { + "epoch": 0.14751, + "grad_norm": 1.1286320877224711, + "learning_rate": 0.003, + "loss": 4.044, + "step": 14751 + }, + { + "epoch": 0.14752, + "grad_norm": 1.3952019890043283, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 14752 + }, + { + "epoch": 0.14753, + "grad_norm": 1.3268177274947321, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 14753 + }, + { + "epoch": 0.14754, + "grad_norm": 1.3229095606049164, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14754 + }, + { + "epoch": 0.14755, + "grad_norm": 1.1673253364225489, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 14755 + }, + { + "epoch": 0.14756, + "grad_norm": 1.382402758451, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 14756 + }, + { + "epoch": 0.14757, + "grad_norm": 1.1441186256744826, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 14757 + }, + { + "epoch": 0.14758, + "grad_norm": 1.1621501001512715, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 14758 + }, + { + "epoch": 0.14759, + "grad_norm": 1.3877917651303693, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 14759 + }, + { + "epoch": 0.1476, + "grad_norm": 1.3096270746637773, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14760 + }, + { + "epoch": 0.14761, + "grad_norm": 1.107020187704195, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 14761 + }, + { + "epoch": 0.14762, + "grad_norm": 1.1961993881582595, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 14762 + }, + { + "epoch": 0.14763, + "grad_norm": 1.303135898447819, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 14763 + }, + { + "epoch": 0.14764, + "grad_norm": 1.2515419072382497, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14764 + }, + { + "epoch": 0.14765, + "grad_norm": 1.2859986321936037, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14765 + }, + { + "epoch": 0.14766, + "grad_norm": 1.021156011483574, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 14766 + }, + { + "epoch": 0.14767, + "grad_norm": 1.3546518718683798, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 14767 + }, + { + "epoch": 0.14768, + "grad_norm": 1.1048674849286966, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14768 + }, + { + "epoch": 0.14769, + "grad_norm": 1.213721024463415, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 14769 + }, + { + "epoch": 0.1477, + "grad_norm": 1.181611806859025, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 14770 + }, + { + "epoch": 0.14771, + "grad_norm": 1.4015824040772773, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14771 + }, + { + "epoch": 0.14772, + "grad_norm": 1.197184787661011, + "learning_rate": 0.003, + "loss": 4.055, + "step": 14772 + }, + { + "epoch": 0.14773, + "grad_norm": 1.1226164282132731, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 14773 + }, + { + "epoch": 0.14774, + "grad_norm": 1.3344600923991445, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 14774 + }, + { + "epoch": 0.14775, + "grad_norm": 1.1506419423756051, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 14775 + }, + { + "epoch": 0.14776, + "grad_norm": 1.293786746697097, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 14776 + }, + { + "epoch": 0.14777, + "grad_norm": 1.1234974672375049, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 14777 + }, + { + "epoch": 0.14778, + "grad_norm": 1.3024728459130335, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14778 + }, + { + "epoch": 0.14779, + "grad_norm": 0.9917384616067323, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 14779 + }, + { + "epoch": 0.1478, + "grad_norm": 1.4676339518358084, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14780 + }, + { + "epoch": 0.14781, + "grad_norm": 1.034039632520674, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14781 + }, + { + "epoch": 0.14782, + "grad_norm": 1.407785614525778, + "learning_rate": 0.003, + "loss": 4.091, + "step": 14782 + }, + { + "epoch": 0.14783, + "grad_norm": 1.13730215414004, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14783 + }, + { + "epoch": 0.14784, + "grad_norm": 1.2776348796678363, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 14784 + }, + { + "epoch": 0.14785, + "grad_norm": 1.2582277140130618, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 14785 + }, + { + "epoch": 0.14786, + "grad_norm": 1.1181549682198384, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14786 + }, + { + "epoch": 0.14787, + "grad_norm": 1.3305483061294012, + "learning_rate": 0.003, + "loss": 4.064, + "step": 14787 + }, + { + "epoch": 0.14788, + "grad_norm": 1.3643327116171096, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 14788 + }, + { + "epoch": 0.14789, + "grad_norm": 1.2123896710327138, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 14789 + }, + { + "epoch": 0.1479, + "grad_norm": 1.257666517907204, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 14790 + }, + { + "epoch": 0.14791, + "grad_norm": 1.2042170863016024, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14791 + }, + { + "epoch": 0.14792, + "grad_norm": 1.2417665699180211, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14792 + }, + { + "epoch": 0.14793, + "grad_norm": 1.2227523296617067, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 14793 + }, + { + "epoch": 0.14794, + "grad_norm": 1.224635598347942, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14794 + }, + { + "epoch": 0.14795, + "grad_norm": 1.068073493344534, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 14795 + }, + { + "epoch": 0.14796, + "grad_norm": 1.2125563865221247, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 14796 + }, + { + "epoch": 0.14797, + "grad_norm": 1.1118542142201944, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 14797 + }, + { + "epoch": 0.14798, + "grad_norm": 1.3720803310657834, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 14798 + }, + { + "epoch": 0.14799, + "grad_norm": 1.0967274438274146, + "learning_rate": 0.003, + "loss": 4.071, + "step": 14799 + }, + { + "epoch": 0.148, + "grad_norm": 1.4347201117761799, + "learning_rate": 0.003, + "loss": 4.059, + "step": 14800 + }, + { + "epoch": 0.14801, + "grad_norm": 1.0571807698765756, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 14801 + }, + { + "epoch": 0.14802, + "grad_norm": 1.33974434466865, + "learning_rate": 0.003, + "loss": 4.075, + "step": 14802 + }, + { + "epoch": 0.14803, + "grad_norm": 1.0461335273789005, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 14803 + }, + { + "epoch": 0.14804, + "grad_norm": 1.5678993159546522, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 14804 + }, + { + "epoch": 0.14805, + "grad_norm": 1.5017874509817701, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14805 + }, + { + "epoch": 0.14806, + "grad_norm": 1.3135443467584271, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 14806 + }, + { + "epoch": 0.14807, + "grad_norm": 1.2573607343279003, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 14807 + }, + { + "epoch": 0.14808, + "grad_norm": 1.340007504898721, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 14808 + }, + { + "epoch": 0.14809, + "grad_norm": 1.2737817685965107, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14809 + }, + { + "epoch": 0.1481, + "grad_norm": 1.231657153429162, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 14810 + }, + { + "epoch": 0.14811, + "grad_norm": 0.9652700786951047, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 14811 + }, + { + "epoch": 0.14812, + "grad_norm": 1.134367013325509, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14812 + }, + { + "epoch": 0.14813, + "grad_norm": 1.2739003479310576, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 14813 + }, + { + "epoch": 0.14814, + "grad_norm": 1.2158291141016366, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14814 + }, + { + "epoch": 0.14815, + "grad_norm": 1.2046555948864488, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 14815 + }, + { + "epoch": 0.14816, + "grad_norm": 0.9447172429128873, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 14816 + }, + { + "epoch": 0.14817, + "grad_norm": 1.2092539474962263, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14817 + }, + { + "epoch": 0.14818, + "grad_norm": 1.2998028016770053, + "learning_rate": 0.003, + "loss": 4.034, + "step": 14818 + }, + { + "epoch": 0.14819, + "grad_norm": 1.5738605231395526, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 14819 + }, + { + "epoch": 0.1482, + "grad_norm": 1.1981548986992587, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14820 + }, + { + "epoch": 0.14821, + "grad_norm": 1.2884537538247403, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 14821 + }, + { + "epoch": 0.14822, + "grad_norm": 1.2873606531452093, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 14822 + }, + { + "epoch": 0.14823, + "grad_norm": 1.2819850551690075, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 14823 + }, + { + "epoch": 0.14824, + "grad_norm": 1.1941581998763113, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 14824 + }, + { + "epoch": 0.14825, + "grad_norm": 1.3312177938775116, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 14825 + }, + { + "epoch": 0.14826, + "grad_norm": 1.133401460840803, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 14826 + }, + { + "epoch": 0.14827, + "grad_norm": 1.2103649040830275, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 14827 + }, + { + "epoch": 0.14828, + "grad_norm": 1.2024206746365804, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14828 + }, + { + "epoch": 0.14829, + "grad_norm": 1.049474499056886, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14829 + }, + { + "epoch": 0.1483, + "grad_norm": 1.1404466141485063, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14830 + }, + { + "epoch": 0.14831, + "grad_norm": 1.2028619755384187, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 14831 + }, + { + "epoch": 0.14832, + "grad_norm": 1.274719239737427, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14832 + }, + { + "epoch": 0.14833, + "grad_norm": 1.168398802387037, + "learning_rate": 0.003, + "loss": 4.078, + "step": 14833 + }, + { + "epoch": 0.14834, + "grad_norm": 1.2480203646129735, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 14834 + }, + { + "epoch": 0.14835, + "grad_norm": 1.3441893456511147, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14835 + }, + { + "epoch": 0.14836, + "grad_norm": 1.263215019256534, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14836 + }, + { + "epoch": 0.14837, + "grad_norm": 1.0878443963119684, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 14837 + }, + { + "epoch": 0.14838, + "grad_norm": 1.4808670361142087, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 14838 + }, + { + "epoch": 0.14839, + "grad_norm": 1.026590985882426, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 14839 + }, + { + "epoch": 0.1484, + "grad_norm": 1.4501261403621308, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 14840 + }, + { + "epoch": 0.14841, + "grad_norm": 1.0557531101030002, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14841 + }, + { + "epoch": 0.14842, + "grad_norm": 1.3984652412164706, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 14842 + }, + { + "epoch": 0.14843, + "grad_norm": 1.2635315636832918, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 14843 + }, + { + "epoch": 0.14844, + "grad_norm": 1.3736676538535175, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 14844 + }, + { + "epoch": 0.14845, + "grad_norm": 1.351947964711682, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 14845 + }, + { + "epoch": 0.14846, + "grad_norm": 1.2076037307068557, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 14846 + }, + { + "epoch": 0.14847, + "grad_norm": 1.2853195481414905, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 14847 + }, + { + "epoch": 0.14848, + "grad_norm": 1.189469804334418, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 14848 + }, + { + "epoch": 0.14849, + "grad_norm": 1.3956045687534402, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14849 + }, + { + "epoch": 0.1485, + "grad_norm": 0.9904765968057987, + "learning_rate": 0.003, + "loss": 4.052, + "step": 14850 + }, + { + "epoch": 0.14851, + "grad_norm": 1.4490773260474896, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 14851 + }, + { + "epoch": 0.14852, + "grad_norm": 1.179826635925976, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 14852 + }, + { + "epoch": 0.14853, + "grad_norm": 1.3174642687265041, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 14853 + }, + { + "epoch": 0.14854, + "grad_norm": 1.2344900327272792, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 14854 + }, + { + "epoch": 0.14855, + "grad_norm": 1.2097266505343784, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 14855 + }, + { + "epoch": 0.14856, + "grad_norm": 1.3055087196965998, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 14856 + }, + { + "epoch": 0.14857, + "grad_norm": 1.107104085111295, + "learning_rate": 0.003, + "loss": 4.088, + "step": 14857 + }, + { + "epoch": 0.14858, + "grad_norm": 1.2705730843296212, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14858 + }, + { + "epoch": 0.14859, + "grad_norm": 0.9637858708447968, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 14859 + }, + { + "epoch": 0.1486, + "grad_norm": 1.1497883211281064, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14860 + }, + { + "epoch": 0.14861, + "grad_norm": 1.2953516340935691, + "learning_rate": 0.003, + "loss": 4.08, + "step": 14861 + }, + { + "epoch": 0.14862, + "grad_norm": 1.054124052608267, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 14862 + }, + { + "epoch": 0.14863, + "grad_norm": 1.3789660702493187, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14863 + }, + { + "epoch": 0.14864, + "grad_norm": 1.042024050437487, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14864 + }, + { + "epoch": 0.14865, + "grad_norm": 1.4555256458496995, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 14865 + }, + { + "epoch": 0.14866, + "grad_norm": 1.1897332584183165, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 14866 + }, + { + "epoch": 0.14867, + "grad_norm": 1.490532399523928, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 14867 + }, + { + "epoch": 0.14868, + "grad_norm": 1.441816676541135, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 14868 + }, + { + "epoch": 0.14869, + "grad_norm": 1.0043574852900115, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 14869 + }, + { + "epoch": 0.1487, + "grad_norm": 1.4047636039319338, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 14870 + }, + { + "epoch": 0.14871, + "grad_norm": 1.3884438161536075, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 14871 + }, + { + "epoch": 0.14872, + "grad_norm": 1.1132691376640853, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 14872 + }, + { + "epoch": 0.14873, + "grad_norm": 1.1977345192860807, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 14873 + }, + { + "epoch": 0.14874, + "grad_norm": 1.245883895286039, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 14874 + }, + { + "epoch": 0.14875, + "grad_norm": 1.207678386998773, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14875 + }, + { + "epoch": 0.14876, + "grad_norm": 1.2774366088212148, + "learning_rate": 0.003, + "loss": 4.083, + "step": 14876 + }, + { + "epoch": 0.14877, + "grad_norm": 1.6502687525664734, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 14877 + }, + { + "epoch": 0.14878, + "grad_norm": 0.9857659560796367, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14878 + }, + { + "epoch": 0.14879, + "grad_norm": 1.4082595417112858, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 14879 + }, + { + "epoch": 0.1488, + "grad_norm": 1.0771591045335969, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 14880 + }, + { + "epoch": 0.14881, + "grad_norm": 1.3166686548948419, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 14881 + }, + { + "epoch": 0.14882, + "grad_norm": 1.233792941159962, + "learning_rate": 0.003, + "loss": 4.093, + "step": 14882 + }, + { + "epoch": 0.14883, + "grad_norm": 1.2791038993936084, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14883 + }, + { + "epoch": 0.14884, + "grad_norm": 1.2305968988624012, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14884 + }, + { + "epoch": 0.14885, + "grad_norm": 1.0241945735482276, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 14885 + }, + { + "epoch": 0.14886, + "grad_norm": 1.3974200331093105, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 14886 + }, + { + "epoch": 0.14887, + "grad_norm": 1.0833386368940932, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14887 + }, + { + "epoch": 0.14888, + "grad_norm": 1.1964013561459441, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 14888 + }, + { + "epoch": 0.14889, + "grad_norm": 1.1799582435469052, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 14889 + }, + { + "epoch": 0.1489, + "grad_norm": 1.390973260807786, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 14890 + }, + { + "epoch": 0.14891, + "grad_norm": 0.9525587988948696, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 14891 + }, + { + "epoch": 0.14892, + "grad_norm": 1.204310502728258, + "learning_rate": 0.003, + "loss": 4.074, + "step": 14892 + }, + { + "epoch": 0.14893, + "grad_norm": 1.0966924371763758, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 14893 + }, + { + "epoch": 0.14894, + "grad_norm": 1.2787229752399252, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 14894 + }, + { + "epoch": 0.14895, + "grad_norm": 1.2750937206101463, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 14895 + }, + { + "epoch": 0.14896, + "grad_norm": 1.2756110885462355, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 14896 + }, + { + "epoch": 0.14897, + "grad_norm": 1.0988749320591142, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 14897 + }, + { + "epoch": 0.14898, + "grad_norm": 1.1670682622351207, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 14898 + }, + { + "epoch": 0.14899, + "grad_norm": 1.1043180262888312, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14899 + }, + { + "epoch": 0.149, + "grad_norm": 1.4613050465833892, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 14900 + }, + { + "epoch": 0.14901, + "grad_norm": 1.4317226682036521, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 14901 + }, + { + "epoch": 0.14902, + "grad_norm": 1.3030218748593272, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 14902 + }, + { + "epoch": 0.14903, + "grad_norm": 1.2349257456957607, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 14903 + }, + { + "epoch": 0.14904, + "grad_norm": 1.222522424979833, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 14904 + }, + { + "epoch": 0.14905, + "grad_norm": 1.3452550190203154, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 14905 + }, + { + "epoch": 0.14906, + "grad_norm": 0.9832696376167528, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 14906 + }, + { + "epoch": 0.14907, + "grad_norm": 1.3907971745147707, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 14907 + }, + { + "epoch": 0.14908, + "grad_norm": 1.1372164335048494, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 14908 + }, + { + "epoch": 0.14909, + "grad_norm": 1.1219052450795157, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 14909 + }, + { + "epoch": 0.1491, + "grad_norm": 1.1825502432967192, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 14910 + }, + { + "epoch": 0.14911, + "grad_norm": 1.2010941132677941, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 14911 + }, + { + "epoch": 0.14912, + "grad_norm": 1.3070054434027416, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 14912 + }, + { + "epoch": 0.14913, + "grad_norm": 1.2596303439594332, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 14913 + }, + { + "epoch": 0.14914, + "grad_norm": 1.1702506535315904, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 14914 + }, + { + "epoch": 0.14915, + "grad_norm": 1.3948642013157526, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 14915 + }, + { + "epoch": 0.14916, + "grad_norm": 1.0853748743773275, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14916 + }, + { + "epoch": 0.14917, + "grad_norm": 1.4939579311245224, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 14917 + }, + { + "epoch": 0.14918, + "grad_norm": 1.1587444653275318, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14918 + }, + { + "epoch": 0.14919, + "grad_norm": 1.2655904622340168, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 14919 + }, + { + "epoch": 0.1492, + "grad_norm": 1.0995955087144285, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 14920 + }, + { + "epoch": 0.14921, + "grad_norm": 1.3928205441425652, + "learning_rate": 0.003, + "loss": 4.083, + "step": 14921 + }, + { + "epoch": 0.14922, + "grad_norm": 1.0865550751685151, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 14922 + }, + { + "epoch": 0.14923, + "grad_norm": 1.2093099353750922, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14923 + }, + { + "epoch": 0.14924, + "grad_norm": 1.2166173806791813, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 14924 + }, + { + "epoch": 0.14925, + "grad_norm": 1.1776264155167722, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 14925 + }, + { + "epoch": 0.14926, + "grad_norm": 1.3314021975619488, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 14926 + }, + { + "epoch": 0.14927, + "grad_norm": 1.060505965970766, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 14927 + }, + { + "epoch": 0.14928, + "grad_norm": 1.4406791970002304, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 14928 + }, + { + "epoch": 0.14929, + "grad_norm": 1.082162795184589, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 14929 + }, + { + "epoch": 0.1493, + "grad_norm": 1.2086048080899723, + "learning_rate": 0.003, + "loss": 4.066, + "step": 14930 + }, + { + "epoch": 0.14931, + "grad_norm": 1.2695538610859594, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 14931 + }, + { + "epoch": 0.14932, + "grad_norm": 1.1318361431289865, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 14932 + }, + { + "epoch": 0.14933, + "grad_norm": 1.446875788888267, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 14933 + }, + { + "epoch": 0.14934, + "grad_norm": 1.1462224150787974, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 14934 + }, + { + "epoch": 0.14935, + "grad_norm": 1.4912216225843362, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 14935 + }, + { + "epoch": 0.14936, + "grad_norm": 1.0533209518682838, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 14936 + }, + { + "epoch": 0.14937, + "grad_norm": 1.1819293779876494, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 14937 + }, + { + "epoch": 0.14938, + "grad_norm": 1.5128539093937539, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 14938 + }, + { + "epoch": 0.14939, + "grad_norm": 0.9915716999665365, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 14939 + }, + { + "epoch": 0.1494, + "grad_norm": 1.187533299776154, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 14940 + }, + { + "epoch": 0.14941, + "grad_norm": 1.2326825324147908, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 14941 + }, + { + "epoch": 0.14942, + "grad_norm": 1.1264556500091154, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 14942 + }, + { + "epoch": 0.14943, + "grad_norm": 1.411943555864094, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 14943 + }, + { + "epoch": 0.14944, + "grad_norm": 0.998547571587907, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 14944 + }, + { + "epoch": 0.14945, + "grad_norm": 1.3654248880724962, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14945 + }, + { + "epoch": 0.14946, + "grad_norm": 1.1308853175079046, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 14946 + }, + { + "epoch": 0.14947, + "grad_norm": 1.1934755301548037, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 14947 + }, + { + "epoch": 0.14948, + "grad_norm": 1.2485781004994043, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 14948 + }, + { + "epoch": 0.14949, + "grad_norm": 1.2913617741593948, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 14949 + }, + { + "epoch": 0.1495, + "grad_norm": 1.487539410018331, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 14950 + }, + { + "epoch": 0.14951, + "grad_norm": 1.1026109116379552, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 14951 + }, + { + "epoch": 0.14952, + "grad_norm": 1.2953541884433633, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 14952 + }, + { + "epoch": 0.14953, + "grad_norm": 1.3740274670281512, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 14953 + }, + { + "epoch": 0.14954, + "grad_norm": 1.0925317694287173, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 14954 + }, + { + "epoch": 0.14955, + "grad_norm": 1.44005155343115, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 14955 + }, + { + "epoch": 0.14956, + "grad_norm": 1.1721710223911144, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 14956 + }, + { + "epoch": 0.14957, + "grad_norm": 1.2069504172841292, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 14957 + }, + { + "epoch": 0.14958, + "grad_norm": 1.2642042169712921, + "learning_rate": 0.003, + "loss": 4.1, + "step": 14958 + }, + { + "epoch": 0.14959, + "grad_norm": 1.3063621719604683, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14959 + }, + { + "epoch": 0.1496, + "grad_norm": 1.3659620018831007, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 14960 + }, + { + "epoch": 0.14961, + "grad_norm": 1.1731188763545322, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 14961 + }, + { + "epoch": 0.14962, + "grad_norm": 1.0893103154843458, + "learning_rate": 0.003, + "loss": 4.057, + "step": 14962 + }, + { + "epoch": 0.14963, + "grad_norm": 1.1524601757780435, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 14963 + }, + { + "epoch": 0.14964, + "grad_norm": 1.208167351977869, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 14964 + }, + { + "epoch": 0.14965, + "grad_norm": 1.23294526359167, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14965 + }, + { + "epoch": 0.14966, + "grad_norm": 1.1538788312035761, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 14966 + }, + { + "epoch": 0.14967, + "grad_norm": 1.1682462372892446, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 14967 + }, + { + "epoch": 0.14968, + "grad_norm": 1.182504116597818, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 14968 + }, + { + "epoch": 0.14969, + "grad_norm": 1.223622558305158, + "learning_rate": 0.003, + "loss": 4.079, + "step": 14969 + }, + { + "epoch": 0.1497, + "grad_norm": 1.077729223328631, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14970 + }, + { + "epoch": 0.14971, + "grad_norm": 1.146948410877107, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 14971 + }, + { + "epoch": 0.14972, + "grad_norm": 1.2407325286138748, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 14972 + }, + { + "epoch": 0.14973, + "grad_norm": 1.2644632188687903, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 14973 + }, + { + "epoch": 0.14974, + "grad_norm": 1.15196408084641, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 14974 + }, + { + "epoch": 0.14975, + "grad_norm": 1.3690869462360291, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 14975 + }, + { + "epoch": 0.14976, + "grad_norm": 1.146252532722954, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 14976 + }, + { + "epoch": 0.14977, + "grad_norm": 1.2893203491323464, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 14977 + }, + { + "epoch": 0.14978, + "grad_norm": 1.5339944589904086, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 14978 + }, + { + "epoch": 0.14979, + "grad_norm": 0.9614418891020783, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14979 + }, + { + "epoch": 0.1498, + "grad_norm": 1.000608635648666, + "learning_rate": 0.003, + "loss": 4.06, + "step": 14980 + }, + { + "epoch": 0.14981, + "grad_norm": 1.3009567765546934, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 14981 + }, + { + "epoch": 0.14982, + "grad_norm": 1.3233483659166978, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 14982 + }, + { + "epoch": 0.14983, + "grad_norm": 1.2615111957089797, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 14983 + }, + { + "epoch": 0.14984, + "grad_norm": 1.3554690619827399, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 14984 + }, + { + "epoch": 0.14985, + "grad_norm": 1.1924340871785533, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 14985 + }, + { + "epoch": 0.14986, + "grad_norm": 1.2318918468737168, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 14986 + }, + { + "epoch": 0.14987, + "grad_norm": 1.0930518175296557, + "learning_rate": 0.003, + "loss": 4.058, + "step": 14987 + }, + { + "epoch": 0.14988, + "grad_norm": 1.1663896180400055, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 14988 + }, + { + "epoch": 0.14989, + "grad_norm": 1.2063698055393761, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 14989 + }, + { + "epoch": 0.1499, + "grad_norm": 1.2874535734867687, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 14990 + }, + { + "epoch": 0.14991, + "grad_norm": 1.1759325809215206, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 14991 + }, + { + "epoch": 0.14992, + "grad_norm": 1.4122874229515472, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 14992 + }, + { + "epoch": 0.14993, + "grad_norm": 1.2843662582190924, + "learning_rate": 0.003, + "loss": 4.056, + "step": 14993 + }, + { + "epoch": 0.14994, + "grad_norm": 1.5176369330876351, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 14994 + }, + { + "epoch": 0.14995, + "grad_norm": 1.1588560544407271, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 14995 + }, + { + "epoch": 0.14996, + "grad_norm": 1.1271968264084253, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 14996 + }, + { + "epoch": 0.14997, + "grad_norm": 1.2102659856566516, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 14997 + }, + { + "epoch": 0.14998, + "grad_norm": 1.275108957785287, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 14998 + }, + { + "epoch": 0.14999, + "grad_norm": 1.265708011373228, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 14999 + }, + { + "epoch": 0.15, + "grad_norm": 1.1439138688198385, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 15000 + }, + { + "epoch": 0.15001, + "grad_norm": 1.2421886806834463, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 15001 + }, + { + "epoch": 0.15002, + "grad_norm": 1.1004519397192334, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15002 + }, + { + "epoch": 0.15003, + "grad_norm": 1.158455354790534, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 15003 + }, + { + "epoch": 0.15004, + "grad_norm": 1.2229740132210145, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 15004 + }, + { + "epoch": 0.15005, + "grad_norm": 1.2085418202639537, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 15005 + }, + { + "epoch": 0.15006, + "grad_norm": 1.3294954845260496, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 15006 + }, + { + "epoch": 0.15007, + "grad_norm": 1.3285655789422341, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 15007 + }, + { + "epoch": 0.15008, + "grad_norm": 1.2991360798789722, + "learning_rate": 0.003, + "loss": 4.055, + "step": 15008 + }, + { + "epoch": 0.15009, + "grad_norm": 1.3008766761442365, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15009 + }, + { + "epoch": 0.1501, + "grad_norm": 1.157389568496755, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 15010 + }, + { + "epoch": 0.15011, + "grad_norm": 1.309388746363442, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 15011 + }, + { + "epoch": 0.15012, + "grad_norm": 1.1518291599840276, + "learning_rate": 0.003, + "loss": 4.046, + "step": 15012 + }, + { + "epoch": 0.15013, + "grad_norm": 1.130969825922695, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 15013 + }, + { + "epoch": 0.15014, + "grad_norm": 1.2675786794096975, + "learning_rate": 0.003, + "loss": 4.046, + "step": 15014 + }, + { + "epoch": 0.15015, + "grad_norm": 1.254882157454107, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15015 + }, + { + "epoch": 0.15016, + "grad_norm": 1.2056073937925937, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 15016 + }, + { + "epoch": 0.15017, + "grad_norm": 1.2294431401855699, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15017 + }, + { + "epoch": 0.15018, + "grad_norm": 1.1514030329240634, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15018 + }, + { + "epoch": 0.15019, + "grad_norm": 1.0910474852609486, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15019 + }, + { + "epoch": 0.1502, + "grad_norm": 1.3194572190844807, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15020 + }, + { + "epoch": 0.15021, + "grad_norm": 1.0878947290892265, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15021 + }, + { + "epoch": 0.15022, + "grad_norm": 1.359752959173593, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15022 + }, + { + "epoch": 0.15023, + "grad_norm": 1.3272200351660286, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 15023 + }, + { + "epoch": 0.15024, + "grad_norm": 1.3650082466937745, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 15024 + }, + { + "epoch": 0.15025, + "grad_norm": 1.0567603693339083, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 15025 + }, + { + "epoch": 0.15026, + "grad_norm": 1.2385010885446035, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15026 + }, + { + "epoch": 0.15027, + "grad_norm": 0.9679446810635529, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 15027 + }, + { + "epoch": 0.15028, + "grad_norm": 1.503314405318618, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 15028 + }, + { + "epoch": 0.15029, + "grad_norm": 1.0213436660067043, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 15029 + }, + { + "epoch": 0.1503, + "grad_norm": 1.6028553178422178, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 15030 + }, + { + "epoch": 0.15031, + "grad_norm": 0.9355461869193742, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 15031 + }, + { + "epoch": 0.15032, + "grad_norm": 1.3076981026785015, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 15032 + }, + { + "epoch": 0.15033, + "grad_norm": 1.3188057675135088, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15033 + }, + { + "epoch": 0.15034, + "grad_norm": 1.4290179136677195, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 15034 + }, + { + "epoch": 0.15035, + "grad_norm": 1.1285669561399247, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 15035 + }, + { + "epoch": 0.15036, + "grad_norm": 1.3493413030420565, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 15036 + }, + { + "epoch": 0.15037, + "grad_norm": 1.1927330301156278, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 15037 + }, + { + "epoch": 0.15038, + "grad_norm": 1.4071175189669094, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 15038 + }, + { + "epoch": 0.15039, + "grad_norm": 1.1432060701143985, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 15039 + }, + { + "epoch": 0.1504, + "grad_norm": 1.3626859879510167, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 15040 + }, + { + "epoch": 0.15041, + "grad_norm": 1.1375881639804108, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 15041 + }, + { + "epoch": 0.15042, + "grad_norm": 1.5718864410411195, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 15042 + }, + { + "epoch": 0.15043, + "grad_norm": 1.0101162506936627, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 15043 + }, + { + "epoch": 0.15044, + "grad_norm": 1.4177728319729006, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15044 + }, + { + "epoch": 0.15045, + "grad_norm": 1.2584862320579329, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15045 + }, + { + "epoch": 0.15046, + "grad_norm": 1.0371645732582666, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 15046 + }, + { + "epoch": 0.15047, + "grad_norm": 1.4680775466139233, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 15047 + }, + { + "epoch": 0.15048, + "grad_norm": 1.240976206826836, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15048 + }, + { + "epoch": 0.15049, + "grad_norm": 1.4725510420309051, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 15049 + }, + { + "epoch": 0.1505, + "grad_norm": 1.2503013556073201, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 15050 + }, + { + "epoch": 0.15051, + "grad_norm": 1.3187053136228997, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15051 + }, + { + "epoch": 0.15052, + "grad_norm": 1.1261744199142463, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 15052 + }, + { + "epoch": 0.15053, + "grad_norm": 1.1922341562009968, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 15053 + }, + { + "epoch": 0.15054, + "grad_norm": 1.280375040618315, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15054 + }, + { + "epoch": 0.15055, + "grad_norm": 1.2626409514997128, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 15055 + }, + { + "epoch": 0.15056, + "grad_norm": 1.3031502941906068, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15056 + }, + { + "epoch": 0.15057, + "grad_norm": 1.146443655378379, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 15057 + }, + { + "epoch": 0.15058, + "grad_norm": 1.234747124111013, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 15058 + }, + { + "epoch": 0.15059, + "grad_norm": 1.1111204119286546, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 15059 + }, + { + "epoch": 0.1506, + "grad_norm": 1.2782788932590696, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 15060 + }, + { + "epoch": 0.15061, + "grad_norm": 1.0913728828104843, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15061 + }, + { + "epoch": 0.15062, + "grad_norm": 1.273029391481596, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15062 + }, + { + "epoch": 0.15063, + "grad_norm": 1.136396228675531, + "learning_rate": 0.003, + "loss": 4.085, + "step": 15063 + }, + { + "epoch": 0.15064, + "grad_norm": 1.1737411571075842, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15064 + }, + { + "epoch": 0.15065, + "grad_norm": 1.2904500760837567, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 15065 + }, + { + "epoch": 0.15066, + "grad_norm": 1.4351156073359232, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 15066 + }, + { + "epoch": 0.15067, + "grad_norm": 1.1296351993384457, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 15067 + }, + { + "epoch": 0.15068, + "grad_norm": 1.4385586669901966, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15068 + }, + { + "epoch": 0.15069, + "grad_norm": 1.104913445664609, + "learning_rate": 0.003, + "loss": 4.071, + "step": 15069 + }, + { + "epoch": 0.1507, + "grad_norm": 1.0053274139043227, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 15070 + }, + { + "epoch": 0.15071, + "grad_norm": 1.2780511731178645, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 15071 + }, + { + "epoch": 0.15072, + "grad_norm": 1.4261993717400032, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 15072 + }, + { + "epoch": 0.15073, + "grad_norm": 1.2855849210568024, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 15073 + }, + { + "epoch": 0.15074, + "grad_norm": 1.5074041648077694, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 15074 + }, + { + "epoch": 0.15075, + "grad_norm": 0.8953110771587716, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 15075 + }, + { + "epoch": 0.15076, + "grad_norm": 1.1679691805137673, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 15076 + }, + { + "epoch": 0.15077, + "grad_norm": 1.2715640089224092, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 15077 + }, + { + "epoch": 0.15078, + "grad_norm": 1.3102155377085043, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 15078 + }, + { + "epoch": 0.15079, + "grad_norm": 1.197144269451487, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 15079 + }, + { + "epoch": 0.1508, + "grad_norm": 1.2880148382239833, + "learning_rate": 0.003, + "loss": 4.094, + "step": 15080 + }, + { + "epoch": 0.15081, + "grad_norm": 1.2900635608171667, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 15081 + }, + { + "epoch": 0.15082, + "grad_norm": 1.2647862818670748, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 15082 + }, + { + "epoch": 0.15083, + "grad_norm": 1.2757016919839912, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 15083 + }, + { + "epoch": 0.15084, + "grad_norm": 1.4092964104632957, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15084 + }, + { + "epoch": 0.15085, + "grad_norm": 1.2947858384427136, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 15085 + }, + { + "epoch": 0.15086, + "grad_norm": 1.3530504623742912, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 15086 + }, + { + "epoch": 0.15087, + "grad_norm": 1.2112907542851699, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15087 + }, + { + "epoch": 0.15088, + "grad_norm": 1.3015483423935452, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15088 + }, + { + "epoch": 0.15089, + "grad_norm": 1.1048537637635403, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15089 + }, + { + "epoch": 0.1509, + "grad_norm": 1.437913899117062, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 15090 + }, + { + "epoch": 0.15091, + "grad_norm": 0.9485030774003165, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 15091 + }, + { + "epoch": 0.15092, + "grad_norm": 1.2492208854554452, + "learning_rate": 0.003, + "loss": 4.09, + "step": 15092 + }, + { + "epoch": 0.15093, + "grad_norm": 1.233147518537233, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15093 + }, + { + "epoch": 0.15094, + "grad_norm": 1.254102187917705, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15094 + }, + { + "epoch": 0.15095, + "grad_norm": 1.0789350378474252, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 15095 + }, + { + "epoch": 0.15096, + "grad_norm": 1.2446487998554652, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 15096 + }, + { + "epoch": 0.15097, + "grad_norm": 1.3909024001347274, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15097 + }, + { + "epoch": 0.15098, + "grad_norm": 1.6132558578797918, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 15098 + }, + { + "epoch": 0.15099, + "grad_norm": 1.3620799585695513, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15099 + }, + { + "epoch": 0.151, + "grad_norm": 1.177533687643014, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15100 + }, + { + "epoch": 0.15101, + "grad_norm": 1.2328881779449214, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 15101 + }, + { + "epoch": 0.15102, + "grad_norm": 1.128148852596184, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15102 + }, + { + "epoch": 0.15103, + "grad_norm": 1.3633640245779437, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 15103 + }, + { + "epoch": 0.15104, + "grad_norm": 1.0922438358343403, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 15104 + }, + { + "epoch": 0.15105, + "grad_norm": 1.2945037789067804, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15105 + }, + { + "epoch": 0.15106, + "grad_norm": 1.3206419658757098, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15106 + }, + { + "epoch": 0.15107, + "grad_norm": 1.1755714323357802, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15107 + }, + { + "epoch": 0.15108, + "grad_norm": 1.2204786349814887, + "learning_rate": 0.003, + "loss": 4.08, + "step": 15108 + }, + { + "epoch": 0.15109, + "grad_norm": 1.1234280468452336, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 15109 + }, + { + "epoch": 0.1511, + "grad_norm": 1.128668822196183, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 15110 + }, + { + "epoch": 0.15111, + "grad_norm": 0.9421908290843221, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 15111 + }, + { + "epoch": 0.15112, + "grad_norm": 1.2068180002027014, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 15112 + }, + { + "epoch": 0.15113, + "grad_norm": 1.2186237898118937, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 15113 + }, + { + "epoch": 0.15114, + "grad_norm": 1.4641406766827232, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 15114 + }, + { + "epoch": 0.15115, + "grad_norm": 1.1919232290832495, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15115 + }, + { + "epoch": 0.15116, + "grad_norm": 1.4728097731686103, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 15116 + }, + { + "epoch": 0.15117, + "grad_norm": 1.3330637176888094, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15117 + }, + { + "epoch": 0.15118, + "grad_norm": 1.2385592706364272, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 15118 + }, + { + "epoch": 0.15119, + "grad_norm": 1.4067574442613973, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 15119 + }, + { + "epoch": 0.1512, + "grad_norm": 1.0519760618581901, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 15120 + }, + { + "epoch": 0.15121, + "grad_norm": 1.2466957115531099, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15121 + }, + { + "epoch": 0.15122, + "grad_norm": 0.9444434919056375, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15122 + }, + { + "epoch": 0.15123, + "grad_norm": 1.1491075553362577, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 15123 + }, + { + "epoch": 0.15124, + "grad_norm": 1.1463798435149848, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15124 + }, + { + "epoch": 0.15125, + "grad_norm": 1.2427319169671964, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 15125 + }, + { + "epoch": 0.15126, + "grad_norm": 1.2947749681988245, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15126 + }, + { + "epoch": 0.15127, + "grad_norm": 1.2967915386781237, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 15127 + }, + { + "epoch": 0.15128, + "grad_norm": 1.4545692904296599, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 15128 + }, + { + "epoch": 0.15129, + "grad_norm": 1.2941914586636714, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 15129 + }, + { + "epoch": 0.1513, + "grad_norm": 1.472383577636909, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 15130 + }, + { + "epoch": 0.15131, + "grad_norm": 1.0286634832854549, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15131 + }, + { + "epoch": 0.15132, + "grad_norm": 1.2498407259586195, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 15132 + }, + { + "epoch": 0.15133, + "grad_norm": 1.0802144509705898, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15133 + }, + { + "epoch": 0.15134, + "grad_norm": 1.246959344241732, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 15134 + }, + { + "epoch": 0.15135, + "grad_norm": 1.0771065009220033, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 15135 + }, + { + "epoch": 0.15136, + "grad_norm": 1.1340540660269705, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15136 + }, + { + "epoch": 0.15137, + "grad_norm": 1.1862769462900704, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 15137 + }, + { + "epoch": 0.15138, + "grad_norm": 1.6039770521958727, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 15138 + }, + { + "epoch": 0.15139, + "grad_norm": 1.550446799994394, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 15139 + }, + { + "epoch": 0.1514, + "grad_norm": 1.087988808876127, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 15140 + }, + { + "epoch": 0.15141, + "grad_norm": 1.2878979634976375, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 15141 + }, + { + "epoch": 0.15142, + "grad_norm": 1.4691629254672802, + "learning_rate": 0.003, + "loss": 4.084, + "step": 15142 + }, + { + "epoch": 0.15143, + "grad_norm": 1.249094550079718, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 15143 + }, + { + "epoch": 0.15144, + "grad_norm": 1.2074184009542037, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 15144 + }, + { + "epoch": 0.15145, + "grad_norm": 1.1486188631745466, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 15145 + }, + { + "epoch": 0.15146, + "grad_norm": 1.3538820591910932, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 15146 + }, + { + "epoch": 0.15147, + "grad_norm": 1.2287767445354425, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15147 + }, + { + "epoch": 0.15148, + "grad_norm": 1.0631822777129871, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 15148 + }, + { + "epoch": 0.15149, + "grad_norm": 1.3665721519318366, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 15149 + }, + { + "epoch": 0.1515, + "grad_norm": 1.0531931029076098, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15150 + }, + { + "epoch": 0.15151, + "grad_norm": 1.3861834427558128, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 15151 + }, + { + "epoch": 0.15152, + "grad_norm": 1.110165551093399, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 15152 + }, + { + "epoch": 0.15153, + "grad_norm": 1.373577010685897, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 15153 + }, + { + "epoch": 0.15154, + "grad_norm": 1.0768957629492795, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 15154 + }, + { + "epoch": 0.15155, + "grad_norm": 1.3036681944200013, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15155 + }, + { + "epoch": 0.15156, + "grad_norm": 1.0134078913275204, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 15156 + }, + { + "epoch": 0.15157, + "grad_norm": 1.3149188647412473, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 15157 + }, + { + "epoch": 0.15158, + "grad_norm": 1.0714720525448034, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 15158 + }, + { + "epoch": 0.15159, + "grad_norm": 1.4667820080180005, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 15159 + }, + { + "epoch": 0.1516, + "grad_norm": 1.0332003953815936, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 15160 + }, + { + "epoch": 0.15161, + "grad_norm": 1.5389260212058098, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 15161 + }, + { + "epoch": 0.15162, + "grad_norm": 1.3362982356832491, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 15162 + }, + { + "epoch": 0.15163, + "grad_norm": 1.3142600675432459, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 15163 + }, + { + "epoch": 0.15164, + "grad_norm": 1.0328530820609112, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 15164 + }, + { + "epoch": 0.15165, + "grad_norm": 1.3797164887047848, + "learning_rate": 0.003, + "loss": 4.055, + "step": 15165 + }, + { + "epoch": 0.15166, + "grad_norm": 1.4007279063856606, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 15166 + }, + { + "epoch": 0.15167, + "grad_norm": 1.0361904823528456, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 15167 + }, + { + "epoch": 0.15168, + "grad_norm": 1.3243661623460543, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 15168 + }, + { + "epoch": 0.15169, + "grad_norm": 1.3822565238845799, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15169 + }, + { + "epoch": 0.1517, + "grad_norm": 1.2175661961955546, + "learning_rate": 0.003, + "loss": 4.056, + "step": 15170 + }, + { + "epoch": 0.15171, + "grad_norm": 1.1780681534154256, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15171 + }, + { + "epoch": 0.15172, + "grad_norm": 1.2003824494889688, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15172 + }, + { + "epoch": 0.15173, + "grad_norm": 1.1151207640916274, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15173 + }, + { + "epoch": 0.15174, + "grad_norm": 1.1997119339423539, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15174 + }, + { + "epoch": 0.15175, + "grad_norm": 1.0815699925379336, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15175 + }, + { + "epoch": 0.15176, + "grad_norm": 1.6680875842354892, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15176 + }, + { + "epoch": 0.15177, + "grad_norm": 1.177468208006591, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15177 + }, + { + "epoch": 0.15178, + "grad_norm": 1.4624994320532954, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 15178 + }, + { + "epoch": 0.15179, + "grad_norm": 1.246454152068424, + "learning_rate": 0.003, + "loss": 4.073, + "step": 15179 + }, + { + "epoch": 0.1518, + "grad_norm": 1.1371684814178287, + "learning_rate": 0.003, + "loss": 4.053, + "step": 15180 + }, + { + "epoch": 0.15181, + "grad_norm": 1.4943192277886241, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 15181 + }, + { + "epoch": 0.15182, + "grad_norm": 1.2107772520176932, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 15182 + }, + { + "epoch": 0.15183, + "grad_norm": 1.1054710008402246, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 15183 + }, + { + "epoch": 0.15184, + "grad_norm": 1.1351215569960855, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 15184 + }, + { + "epoch": 0.15185, + "grad_norm": 1.4329950415901205, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15185 + }, + { + "epoch": 0.15186, + "grad_norm": 1.149140906166895, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 15186 + }, + { + "epoch": 0.15187, + "grad_norm": 1.2625000256864467, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 15187 + }, + { + "epoch": 0.15188, + "grad_norm": 0.9386358846680741, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15188 + }, + { + "epoch": 0.15189, + "grad_norm": 1.1714196410569182, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 15189 + }, + { + "epoch": 0.1519, + "grad_norm": 1.2510415235166548, + "learning_rate": 0.003, + "loss": 4.087, + "step": 15190 + }, + { + "epoch": 0.15191, + "grad_norm": 1.2872169549150057, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 15191 + }, + { + "epoch": 0.15192, + "grad_norm": 1.3119991182040003, + "learning_rate": 0.003, + "loss": 4.068, + "step": 15192 + }, + { + "epoch": 0.15193, + "grad_norm": 1.0603377530914684, + "learning_rate": 0.003, + "loss": 4.065, + "step": 15193 + }, + { + "epoch": 0.15194, + "grad_norm": 1.2822519431801203, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 15194 + }, + { + "epoch": 0.15195, + "grad_norm": 1.004259428252377, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 15195 + }, + { + "epoch": 0.15196, + "grad_norm": 1.4231490250930479, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 15196 + }, + { + "epoch": 0.15197, + "grad_norm": 1.420169991810222, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15197 + }, + { + "epoch": 0.15198, + "grad_norm": 1.1784357099586809, + "learning_rate": 0.003, + "loss": 4.082, + "step": 15198 + }, + { + "epoch": 0.15199, + "grad_norm": 1.392187510395877, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15199 + }, + { + "epoch": 0.152, + "grad_norm": 1.116211059199764, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 15200 + }, + { + "epoch": 0.15201, + "grad_norm": 1.4263333656125654, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15201 + }, + { + "epoch": 0.15202, + "grad_norm": 1.2286807081200959, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 15202 + }, + { + "epoch": 0.15203, + "grad_norm": 1.0920509862065366, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 15203 + }, + { + "epoch": 0.15204, + "grad_norm": 1.2614998089297877, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 15204 + }, + { + "epoch": 0.15205, + "grad_norm": 1.1692633950773663, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15205 + }, + { + "epoch": 0.15206, + "grad_norm": 1.409550414533334, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15206 + }, + { + "epoch": 0.15207, + "grad_norm": 1.3113605146737568, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15207 + }, + { + "epoch": 0.15208, + "grad_norm": 1.309460276567801, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 15208 + }, + { + "epoch": 0.15209, + "grad_norm": 1.1894917430330245, + "learning_rate": 0.003, + "loss": 4.087, + "step": 15209 + }, + { + "epoch": 0.1521, + "grad_norm": 1.1703660954370134, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 15210 + }, + { + "epoch": 0.15211, + "grad_norm": 1.2989528506275134, + "learning_rate": 0.003, + "loss": 4.079, + "step": 15211 + }, + { + "epoch": 0.15212, + "grad_norm": 1.0765810928628827, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 15212 + }, + { + "epoch": 0.15213, + "grad_norm": 1.4756608926855814, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15213 + }, + { + "epoch": 0.15214, + "grad_norm": 1.1912439313479937, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15214 + }, + { + "epoch": 0.15215, + "grad_norm": 1.006524948376339, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15215 + }, + { + "epoch": 0.15216, + "grad_norm": 1.261103228412681, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15216 + }, + { + "epoch": 0.15217, + "grad_norm": 1.1249186122517645, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 15217 + }, + { + "epoch": 0.15218, + "grad_norm": 1.278349145361421, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 15218 + }, + { + "epoch": 0.15219, + "grad_norm": 1.0713858095086553, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 15219 + }, + { + "epoch": 0.1522, + "grad_norm": 1.6026457850936104, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15220 + }, + { + "epoch": 0.15221, + "grad_norm": 1.0833468242383575, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 15221 + }, + { + "epoch": 0.15222, + "grad_norm": 1.3253246748494185, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 15222 + }, + { + "epoch": 0.15223, + "grad_norm": 1.3226742938326956, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15223 + }, + { + "epoch": 0.15224, + "grad_norm": 1.0668314677110395, + "learning_rate": 0.003, + "loss": 4.085, + "step": 15224 + }, + { + "epoch": 0.15225, + "grad_norm": 1.339893976702488, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 15225 + }, + { + "epoch": 0.15226, + "grad_norm": 0.9909076328643073, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15226 + }, + { + "epoch": 0.15227, + "grad_norm": 1.3457475132936831, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 15227 + }, + { + "epoch": 0.15228, + "grad_norm": 1.223750730023511, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 15228 + }, + { + "epoch": 0.15229, + "grad_norm": 1.256632194523041, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15229 + }, + { + "epoch": 0.1523, + "grad_norm": 1.059870090530304, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 15230 + }, + { + "epoch": 0.15231, + "grad_norm": 1.689885438587469, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 15231 + }, + { + "epoch": 0.15232, + "grad_norm": 1.181258046402351, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15232 + }, + { + "epoch": 0.15233, + "grad_norm": 1.5769078838020736, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15233 + }, + { + "epoch": 0.15234, + "grad_norm": 0.9274230056976925, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 15234 + }, + { + "epoch": 0.15235, + "grad_norm": 1.2926060469909377, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15235 + }, + { + "epoch": 0.15236, + "grad_norm": 1.240377557258737, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 15236 + }, + { + "epoch": 0.15237, + "grad_norm": 1.298389514130426, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15237 + }, + { + "epoch": 0.15238, + "grad_norm": 1.2163521689730095, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 15238 + }, + { + "epoch": 0.15239, + "grad_norm": 1.0937131665994184, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15239 + }, + { + "epoch": 0.1524, + "grad_norm": 1.3494526264556403, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15240 + }, + { + "epoch": 0.15241, + "grad_norm": 1.2086076171424276, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 15241 + }, + { + "epoch": 0.15242, + "grad_norm": 1.2275639454465173, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 15242 + }, + { + "epoch": 0.15243, + "grad_norm": 1.1768041134744753, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15243 + }, + { + "epoch": 0.15244, + "grad_norm": 1.1460609617592146, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 15244 + }, + { + "epoch": 0.15245, + "grad_norm": 1.0362631941590839, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 15245 + }, + { + "epoch": 0.15246, + "grad_norm": 1.2221248760518095, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15246 + }, + { + "epoch": 0.15247, + "grad_norm": 1.095691988113859, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 15247 + }, + { + "epoch": 0.15248, + "grad_norm": 1.4252171113365828, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 15248 + }, + { + "epoch": 0.15249, + "grad_norm": 0.9907212320633807, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 15249 + }, + { + "epoch": 0.1525, + "grad_norm": 1.1648163444936928, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 15250 + }, + { + "epoch": 0.15251, + "grad_norm": 1.3792381729375842, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 15251 + }, + { + "epoch": 0.15252, + "grad_norm": 1.0966643893758867, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 15252 + }, + { + "epoch": 0.15253, + "grad_norm": 1.352568157703781, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 15253 + }, + { + "epoch": 0.15254, + "grad_norm": 1.162965904860133, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15254 + }, + { + "epoch": 0.15255, + "grad_norm": 1.230108589317307, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 15255 + }, + { + "epoch": 0.15256, + "grad_norm": 1.306415003864345, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15256 + }, + { + "epoch": 0.15257, + "grad_norm": 1.2698699716294912, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15257 + }, + { + "epoch": 0.15258, + "grad_norm": 1.2202436427791954, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 15258 + }, + { + "epoch": 0.15259, + "grad_norm": 1.3699491692866101, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15259 + }, + { + "epoch": 0.1526, + "grad_norm": 1.2865829303939478, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15260 + }, + { + "epoch": 0.15261, + "grad_norm": 1.1907326806337453, + "learning_rate": 0.003, + "loss": 4.075, + "step": 15261 + }, + { + "epoch": 0.15262, + "grad_norm": 1.3741574656043807, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15262 + }, + { + "epoch": 0.15263, + "grad_norm": 1.2461198602893133, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 15263 + }, + { + "epoch": 0.15264, + "grad_norm": 1.3037038166580812, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 15264 + }, + { + "epoch": 0.15265, + "grad_norm": 1.0973837487129527, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 15265 + }, + { + "epoch": 0.15266, + "grad_norm": 1.17550255547033, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 15266 + }, + { + "epoch": 0.15267, + "grad_norm": 1.350055319387001, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 15267 + }, + { + "epoch": 0.15268, + "grad_norm": 1.2135196925321778, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15268 + }, + { + "epoch": 0.15269, + "grad_norm": 1.052807123240445, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15269 + }, + { + "epoch": 0.1527, + "grad_norm": 1.5162917289975957, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 15270 + }, + { + "epoch": 0.15271, + "grad_norm": 0.9087099666900801, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15271 + }, + { + "epoch": 0.15272, + "grad_norm": 1.2080481623654342, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 15272 + }, + { + "epoch": 0.15273, + "grad_norm": 1.205366680073686, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 15273 + }, + { + "epoch": 0.15274, + "grad_norm": 1.2646904527024778, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15274 + }, + { + "epoch": 0.15275, + "grad_norm": 1.5286956874796136, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 15275 + }, + { + "epoch": 0.15276, + "grad_norm": 1.0009018904534552, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 15276 + }, + { + "epoch": 0.15277, + "grad_norm": 1.2494785309856136, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15277 + }, + { + "epoch": 0.15278, + "grad_norm": 1.2861304222588785, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 15278 + }, + { + "epoch": 0.15279, + "grad_norm": 1.1892301729765877, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 15279 + }, + { + "epoch": 0.1528, + "grad_norm": 1.1204842553347314, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 15280 + }, + { + "epoch": 0.15281, + "grad_norm": 1.379822288030074, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 15281 + }, + { + "epoch": 0.15282, + "grad_norm": 1.183577530960742, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 15282 + }, + { + "epoch": 0.15283, + "grad_norm": 1.2435364515876217, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 15283 + }, + { + "epoch": 0.15284, + "grad_norm": 1.1296979253872497, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 15284 + }, + { + "epoch": 0.15285, + "grad_norm": 1.3123458983768324, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 15285 + }, + { + "epoch": 0.15286, + "grad_norm": 1.3325798748619977, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 15286 + }, + { + "epoch": 0.15287, + "grad_norm": 1.250202260518972, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15287 + }, + { + "epoch": 0.15288, + "grad_norm": 1.322467529630218, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15288 + }, + { + "epoch": 0.15289, + "grad_norm": 1.294869647610876, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15289 + }, + { + "epoch": 0.1529, + "grad_norm": 1.0919572359477165, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 15290 + }, + { + "epoch": 0.15291, + "grad_norm": 1.2314711363464752, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 15291 + }, + { + "epoch": 0.15292, + "grad_norm": 1.1864497583011213, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15292 + }, + { + "epoch": 0.15293, + "grad_norm": 1.273989288109886, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 15293 + }, + { + "epoch": 0.15294, + "grad_norm": 1.1428627955755606, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 15294 + }, + { + "epoch": 0.15295, + "grad_norm": 1.3318276639584221, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 15295 + }, + { + "epoch": 0.15296, + "grad_norm": 0.9996627595976725, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 15296 + }, + { + "epoch": 0.15297, + "grad_norm": 1.3567545643645589, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 15297 + }, + { + "epoch": 0.15298, + "grad_norm": 1.0888841094184967, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15298 + }, + { + "epoch": 0.15299, + "grad_norm": 1.3143604999336704, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 15299 + }, + { + "epoch": 0.153, + "grad_norm": 1.10478291680258, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 15300 + }, + { + "epoch": 0.15301, + "grad_norm": 1.332542631690166, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 15301 + }, + { + "epoch": 0.15302, + "grad_norm": 1.3610143355801558, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15302 + }, + { + "epoch": 0.15303, + "grad_norm": 1.6499931279519562, + "learning_rate": 0.003, + "loss": 4.101, + "step": 15303 + }, + { + "epoch": 0.15304, + "grad_norm": 0.9729268294633743, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 15304 + }, + { + "epoch": 0.15305, + "grad_norm": 1.2812268645301028, + "learning_rate": 0.003, + "loss": 4.064, + "step": 15305 + }, + { + "epoch": 0.15306, + "grad_norm": 1.3219200685292423, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 15306 + }, + { + "epoch": 0.15307, + "grad_norm": 1.161016281668427, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 15307 + }, + { + "epoch": 0.15308, + "grad_norm": 1.117450920302711, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15308 + }, + { + "epoch": 0.15309, + "grad_norm": 1.197542675446433, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 15309 + }, + { + "epoch": 0.1531, + "grad_norm": 1.3599553040681978, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 15310 + }, + { + "epoch": 0.15311, + "grad_norm": 1.3874569351603045, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 15311 + }, + { + "epoch": 0.15312, + "grad_norm": 1.2480329082893813, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15312 + }, + { + "epoch": 0.15313, + "grad_norm": 1.1836604804725042, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 15313 + }, + { + "epoch": 0.15314, + "grad_norm": 1.3363089275512905, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15314 + }, + { + "epoch": 0.15315, + "grad_norm": 1.2005744273055012, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 15315 + }, + { + "epoch": 0.15316, + "grad_norm": 1.272784295793868, + "learning_rate": 0.003, + "loss": 4.093, + "step": 15316 + }, + { + "epoch": 0.15317, + "grad_norm": 1.028788498800255, + "learning_rate": 0.003, + "loss": 4.088, + "step": 15317 + }, + { + "epoch": 0.15318, + "grad_norm": 1.5934021220240417, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15318 + }, + { + "epoch": 0.15319, + "grad_norm": 1.0956734498965002, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15319 + }, + { + "epoch": 0.1532, + "grad_norm": 1.6551031206932727, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 15320 + }, + { + "epoch": 0.15321, + "grad_norm": 1.2102759249965225, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15321 + }, + { + "epoch": 0.15322, + "grad_norm": 1.1370647599775683, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 15322 + }, + { + "epoch": 0.15323, + "grad_norm": 1.3220558450236222, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 15323 + }, + { + "epoch": 0.15324, + "grad_norm": 1.2840062337198395, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 15324 + }, + { + "epoch": 0.15325, + "grad_norm": 1.3013782425747533, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 15325 + }, + { + "epoch": 0.15326, + "grad_norm": 1.1975442037482786, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 15326 + }, + { + "epoch": 0.15327, + "grad_norm": 1.1269489485313762, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15327 + }, + { + "epoch": 0.15328, + "grad_norm": 1.1410612864581509, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15328 + }, + { + "epoch": 0.15329, + "grad_norm": 1.3460362322188562, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15329 + }, + { + "epoch": 0.1533, + "grad_norm": 1.3637334924767632, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 15330 + }, + { + "epoch": 0.15331, + "grad_norm": 1.0996693942208156, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 15331 + }, + { + "epoch": 0.15332, + "grad_norm": 1.3134945624775438, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 15332 + }, + { + "epoch": 0.15333, + "grad_norm": 0.9603032493816029, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 15333 + }, + { + "epoch": 0.15334, + "grad_norm": 1.4163721083411738, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15334 + }, + { + "epoch": 0.15335, + "grad_norm": 1.0705217711953943, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 15335 + }, + { + "epoch": 0.15336, + "grad_norm": 1.2004877087986423, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15336 + }, + { + "epoch": 0.15337, + "grad_norm": 1.2784195202324355, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15337 + }, + { + "epoch": 0.15338, + "grad_norm": 0.9428270674974255, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 15338 + }, + { + "epoch": 0.15339, + "grad_norm": 1.232716499678335, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15339 + }, + { + "epoch": 0.1534, + "grad_norm": 1.3422285090284969, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 15340 + }, + { + "epoch": 0.15341, + "grad_norm": 1.0887037346989066, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 15341 + }, + { + "epoch": 0.15342, + "grad_norm": 1.40533166454307, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 15342 + }, + { + "epoch": 0.15343, + "grad_norm": 1.2808704916445435, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 15343 + }, + { + "epoch": 0.15344, + "grad_norm": 1.7381559003301226, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 15344 + }, + { + "epoch": 0.15345, + "grad_norm": 0.8981993766105962, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 15345 + }, + { + "epoch": 0.15346, + "grad_norm": 1.021591731175303, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15346 + }, + { + "epoch": 0.15347, + "grad_norm": 1.342833780617317, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 15347 + }, + { + "epoch": 0.15348, + "grad_norm": 1.055504084451775, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 15348 + }, + { + "epoch": 0.15349, + "grad_norm": 1.4292975403438253, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 15349 + }, + { + "epoch": 0.1535, + "grad_norm": 1.242457260105871, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 15350 + }, + { + "epoch": 0.15351, + "grad_norm": 1.607477705777395, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 15351 + }, + { + "epoch": 0.15352, + "grad_norm": 1.0042952049385223, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15352 + }, + { + "epoch": 0.15353, + "grad_norm": 1.5215221222668274, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15353 + }, + { + "epoch": 0.15354, + "grad_norm": 1.1167885093181256, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 15354 + }, + { + "epoch": 0.15355, + "grad_norm": 1.337051204486555, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 15355 + }, + { + "epoch": 0.15356, + "grad_norm": 1.2014364989112793, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 15356 + }, + { + "epoch": 0.15357, + "grad_norm": 1.0611147472963682, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15357 + }, + { + "epoch": 0.15358, + "grad_norm": 1.4014457304898078, + "learning_rate": 0.003, + "loss": 4.067, + "step": 15358 + }, + { + "epoch": 0.15359, + "grad_norm": 1.0441717054483164, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 15359 + }, + { + "epoch": 0.1536, + "grad_norm": 1.2532453030452138, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 15360 + }, + { + "epoch": 0.15361, + "grad_norm": 1.0668416216593732, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 15361 + }, + { + "epoch": 0.15362, + "grad_norm": 1.44348968864017, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 15362 + }, + { + "epoch": 0.15363, + "grad_norm": 1.0324397895486972, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 15363 + }, + { + "epoch": 0.15364, + "grad_norm": 1.4302096146858072, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 15364 + }, + { + "epoch": 0.15365, + "grad_norm": 1.445770752562098, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 15365 + }, + { + "epoch": 0.15366, + "grad_norm": 1.4021681092834855, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 15366 + }, + { + "epoch": 0.15367, + "grad_norm": 1.1844554825553057, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 15367 + }, + { + "epoch": 0.15368, + "grad_norm": 1.3157634932169993, + "learning_rate": 0.003, + "loss": 4.069, + "step": 15368 + }, + { + "epoch": 0.15369, + "grad_norm": 1.5005684371665555, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 15369 + }, + { + "epoch": 0.1537, + "grad_norm": 1.1268061902873954, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 15370 + }, + { + "epoch": 0.15371, + "grad_norm": 1.2839413286799333, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15371 + }, + { + "epoch": 0.15372, + "grad_norm": 1.401179204053029, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15372 + }, + { + "epoch": 0.15373, + "grad_norm": 1.1167028590635872, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15373 + }, + { + "epoch": 0.15374, + "grad_norm": 1.4268777343963048, + "learning_rate": 0.003, + "loss": 4.081, + "step": 15374 + }, + { + "epoch": 0.15375, + "grad_norm": 1.2064984743607405, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 15375 + }, + { + "epoch": 0.15376, + "grad_norm": 1.0546248966109226, + "learning_rate": 0.003, + "loss": 4.08, + "step": 15376 + }, + { + "epoch": 0.15377, + "grad_norm": 1.157082907867689, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 15377 + }, + { + "epoch": 0.15378, + "grad_norm": 1.0936352419546245, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 15378 + }, + { + "epoch": 0.15379, + "grad_norm": 1.2548739883846005, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 15379 + }, + { + "epoch": 0.1538, + "grad_norm": 1.4031950487192069, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 15380 + }, + { + "epoch": 0.15381, + "grad_norm": 1.2224787943842523, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 15381 + }, + { + "epoch": 0.15382, + "grad_norm": 1.2236705526083234, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 15382 + }, + { + "epoch": 0.15383, + "grad_norm": 1.0648812817113953, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 15383 + }, + { + "epoch": 0.15384, + "grad_norm": 1.2879123629622429, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15384 + }, + { + "epoch": 0.15385, + "grad_norm": 1.012268736905604, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 15385 + }, + { + "epoch": 0.15386, + "grad_norm": 1.2187582005774895, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15386 + }, + { + "epoch": 0.15387, + "grad_norm": 1.1884376651243376, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 15387 + }, + { + "epoch": 0.15388, + "grad_norm": 1.3226916097039858, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15388 + }, + { + "epoch": 0.15389, + "grad_norm": 1.3303186594915881, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 15389 + }, + { + "epoch": 0.1539, + "grad_norm": 1.3866925840894608, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 15390 + }, + { + "epoch": 0.15391, + "grad_norm": 1.3728111518143102, + "learning_rate": 0.003, + "loss": 4.073, + "step": 15391 + }, + { + "epoch": 0.15392, + "grad_norm": 1.3402808808497404, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 15392 + }, + { + "epoch": 0.15393, + "grad_norm": 1.1579581855029002, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 15393 + }, + { + "epoch": 0.15394, + "grad_norm": 1.2034269596203029, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 15394 + }, + { + "epoch": 0.15395, + "grad_norm": 1.1618115125409054, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 15395 + }, + { + "epoch": 0.15396, + "grad_norm": 1.2863291450699967, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 15396 + }, + { + "epoch": 0.15397, + "grad_norm": 1.2100596132475223, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 15397 + }, + { + "epoch": 0.15398, + "grad_norm": 1.3418803257466523, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15398 + }, + { + "epoch": 0.15399, + "grad_norm": 0.8759960471617386, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15399 + }, + { + "epoch": 0.154, + "grad_norm": 0.9909269155497589, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 15400 + }, + { + "epoch": 0.15401, + "grad_norm": 1.280760201565616, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 15401 + }, + { + "epoch": 0.15402, + "grad_norm": 1.3039485153919603, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 15402 + }, + { + "epoch": 0.15403, + "grad_norm": 1.1319622563485687, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15403 + }, + { + "epoch": 0.15404, + "grad_norm": 1.5872253219858796, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 15404 + }, + { + "epoch": 0.15405, + "grad_norm": 1.1610910153502412, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 15405 + }, + { + "epoch": 0.15406, + "grad_norm": 1.3946380911331175, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 15406 + }, + { + "epoch": 0.15407, + "grad_norm": 1.1320665509097738, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 15407 + }, + { + "epoch": 0.15408, + "grad_norm": 1.3176567139986404, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15408 + }, + { + "epoch": 0.15409, + "grad_norm": 1.3445347155612486, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 15409 + }, + { + "epoch": 0.1541, + "grad_norm": 1.2194685040205706, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 15410 + }, + { + "epoch": 0.15411, + "grad_norm": 1.217023193541309, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 15411 + }, + { + "epoch": 0.15412, + "grad_norm": 1.1999213085696572, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 15412 + }, + { + "epoch": 0.15413, + "grad_norm": 1.2812853286852148, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 15413 + }, + { + "epoch": 0.15414, + "grad_norm": 1.3143368525425625, + "learning_rate": 0.003, + "loss": 4.064, + "step": 15414 + }, + { + "epoch": 0.15415, + "grad_norm": 1.3786879212302876, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15415 + }, + { + "epoch": 0.15416, + "grad_norm": 1.1378575541792832, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15416 + }, + { + "epoch": 0.15417, + "grad_norm": 1.4468712668422241, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15417 + }, + { + "epoch": 0.15418, + "grad_norm": 1.1350194620282688, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 15418 + }, + { + "epoch": 0.15419, + "grad_norm": 1.3058088501430103, + "learning_rate": 0.003, + "loss": 4.056, + "step": 15419 + }, + { + "epoch": 0.1542, + "grad_norm": 1.2290528501643871, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 15420 + }, + { + "epoch": 0.15421, + "grad_norm": 1.34099276740831, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 15421 + }, + { + "epoch": 0.15422, + "grad_norm": 1.2689686644532474, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 15422 + }, + { + "epoch": 0.15423, + "grad_norm": 1.1798623382145859, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15423 + }, + { + "epoch": 0.15424, + "grad_norm": 1.1414205811208402, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 15424 + }, + { + "epoch": 0.15425, + "grad_norm": 1.382624074165894, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15425 + }, + { + "epoch": 0.15426, + "grad_norm": 1.0098622350888455, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 15426 + }, + { + "epoch": 0.15427, + "grad_norm": 1.2158447740165523, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 15427 + }, + { + "epoch": 0.15428, + "grad_norm": 1.0800505764651744, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15428 + }, + { + "epoch": 0.15429, + "grad_norm": 1.4389797603354793, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15429 + }, + { + "epoch": 0.1543, + "grad_norm": 0.9131784264756577, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 15430 + }, + { + "epoch": 0.15431, + "grad_norm": 1.1406774347349673, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 15431 + }, + { + "epoch": 0.15432, + "grad_norm": 1.4160712255500911, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15432 + }, + { + "epoch": 0.15433, + "grad_norm": 1.0321017123231477, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 15433 + }, + { + "epoch": 0.15434, + "grad_norm": 1.3129401837773598, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15434 + }, + { + "epoch": 0.15435, + "grad_norm": 1.295097177277385, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 15435 + }, + { + "epoch": 0.15436, + "grad_norm": 1.3445014318375115, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 15436 + }, + { + "epoch": 0.15437, + "grad_norm": 1.2037512242094492, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 15437 + }, + { + "epoch": 0.15438, + "grad_norm": 1.2204961008983182, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 15438 + }, + { + "epoch": 0.15439, + "grad_norm": 1.1936893451355557, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 15439 + }, + { + "epoch": 0.1544, + "grad_norm": 1.110082884413141, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 15440 + }, + { + "epoch": 0.15441, + "grad_norm": 1.17060566072975, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15441 + }, + { + "epoch": 0.15442, + "grad_norm": 1.2001477195120134, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 15442 + }, + { + "epoch": 0.15443, + "grad_norm": 1.2014003725220923, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15443 + }, + { + "epoch": 0.15444, + "grad_norm": 1.2729262505707326, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15444 + }, + { + "epoch": 0.15445, + "grad_norm": 1.3667928177643467, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15445 + }, + { + "epoch": 0.15446, + "grad_norm": 1.1717222335822768, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 15446 + }, + { + "epoch": 0.15447, + "grad_norm": 1.4097833628228227, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15447 + }, + { + "epoch": 0.15448, + "grad_norm": 1.4843293064081833, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 15448 + }, + { + "epoch": 0.15449, + "grad_norm": 1.073061414519865, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 15449 + }, + { + "epoch": 0.1545, + "grad_norm": 1.4751810338681888, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15450 + }, + { + "epoch": 0.15451, + "grad_norm": 1.1619538323488983, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15451 + }, + { + "epoch": 0.15452, + "grad_norm": 1.4427250183680014, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 15452 + }, + { + "epoch": 0.15453, + "grad_norm": 1.0283716662610662, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 15453 + }, + { + "epoch": 0.15454, + "grad_norm": 1.2388449133718995, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 15454 + }, + { + "epoch": 0.15455, + "grad_norm": 1.2397170781988538, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 15455 + }, + { + "epoch": 0.15456, + "grad_norm": 1.335769548973706, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 15456 + }, + { + "epoch": 0.15457, + "grad_norm": 1.2550274367262313, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15457 + }, + { + "epoch": 0.15458, + "grad_norm": 1.6013706385545896, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 15458 + }, + { + "epoch": 0.15459, + "grad_norm": 1.4113908926528906, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 15459 + }, + { + "epoch": 0.1546, + "grad_norm": 1.1872880676167363, + "learning_rate": 0.003, + "loss": 4.052, + "step": 15460 + }, + { + "epoch": 0.15461, + "grad_norm": 1.1602721882141036, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 15461 + }, + { + "epoch": 0.15462, + "grad_norm": 1.2657351307427533, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 15462 + }, + { + "epoch": 0.15463, + "grad_norm": 1.017266066341593, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15463 + }, + { + "epoch": 0.15464, + "grad_norm": 1.22408385113059, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 15464 + }, + { + "epoch": 0.15465, + "grad_norm": 1.2388371752459086, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15465 + }, + { + "epoch": 0.15466, + "grad_norm": 1.3378166995203025, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 15466 + }, + { + "epoch": 0.15467, + "grad_norm": 1.2673971792020895, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15467 + }, + { + "epoch": 0.15468, + "grad_norm": 1.2670871657029013, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 15468 + }, + { + "epoch": 0.15469, + "grad_norm": 1.4855845526650822, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 15469 + }, + { + "epoch": 0.1547, + "grad_norm": 1.1664566423581224, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 15470 + }, + { + "epoch": 0.15471, + "grad_norm": 1.43333578071656, + "learning_rate": 0.003, + "loss": 4.07, + "step": 15471 + }, + { + "epoch": 0.15472, + "grad_norm": 0.8916827173933783, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 15472 + }, + { + "epoch": 0.15473, + "grad_norm": 1.2228661035293775, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 15473 + }, + { + "epoch": 0.15474, + "grad_norm": 1.3045348002003534, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 15474 + }, + { + "epoch": 0.15475, + "grad_norm": 0.9752922594250674, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15475 + }, + { + "epoch": 0.15476, + "grad_norm": 1.399714944150617, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 15476 + }, + { + "epoch": 0.15477, + "grad_norm": 0.967828553650773, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15477 + }, + { + "epoch": 0.15478, + "grad_norm": 1.54376698998061, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 15478 + }, + { + "epoch": 0.15479, + "grad_norm": 1.111321915290337, + "learning_rate": 0.003, + "loss": 4.056, + "step": 15479 + }, + { + "epoch": 0.1548, + "grad_norm": 1.3432186028132298, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 15480 + }, + { + "epoch": 0.15481, + "grad_norm": 1.2045266952666287, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 15481 + }, + { + "epoch": 0.15482, + "grad_norm": 1.5509962795872299, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15482 + }, + { + "epoch": 0.15483, + "grad_norm": 1.038507871352347, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 15483 + }, + { + "epoch": 0.15484, + "grad_norm": 1.3999752602376994, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 15484 + }, + { + "epoch": 0.15485, + "grad_norm": 1.2488920313017315, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 15485 + }, + { + "epoch": 0.15486, + "grad_norm": 1.2928076344043185, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 15486 + }, + { + "epoch": 0.15487, + "grad_norm": 1.1813696870402486, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 15487 + }, + { + "epoch": 0.15488, + "grad_norm": 1.4851073401211132, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 15488 + }, + { + "epoch": 0.15489, + "grad_norm": 1.005290261773753, + "learning_rate": 0.003, + "loss": 4.063, + "step": 15489 + }, + { + "epoch": 0.1549, + "grad_norm": 1.4472636495984204, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 15490 + }, + { + "epoch": 0.15491, + "grad_norm": 1.1113667748653209, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 15491 + }, + { + "epoch": 0.15492, + "grad_norm": 1.1923215483799454, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 15492 + }, + { + "epoch": 0.15493, + "grad_norm": 1.3557008672418909, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15493 + }, + { + "epoch": 0.15494, + "grad_norm": 1.2625023638728134, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 15494 + }, + { + "epoch": 0.15495, + "grad_norm": 1.486884034569414, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 15495 + }, + { + "epoch": 0.15496, + "grad_norm": 1.0947414987682853, + "learning_rate": 0.003, + "loss": 4.071, + "step": 15496 + }, + { + "epoch": 0.15497, + "grad_norm": 1.3444649580568884, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15497 + }, + { + "epoch": 0.15498, + "grad_norm": 1.2517682754123742, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15498 + }, + { + "epoch": 0.15499, + "grad_norm": 1.2939405259023302, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 15499 + }, + { + "epoch": 0.155, + "grad_norm": 1.2833960560786328, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 15500 + }, + { + "epoch": 0.15501, + "grad_norm": 1.1128859882325806, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15501 + }, + { + "epoch": 0.15502, + "grad_norm": 1.1979063438901953, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 15502 + }, + { + "epoch": 0.15503, + "grad_norm": 1.3011241842981904, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 15503 + }, + { + "epoch": 0.15504, + "grad_norm": 1.327429675523034, + "learning_rate": 0.003, + "loss": 4.093, + "step": 15504 + }, + { + "epoch": 0.15505, + "grad_norm": 1.037004976848662, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 15505 + }, + { + "epoch": 0.15506, + "grad_norm": 1.358924356732899, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 15506 + }, + { + "epoch": 0.15507, + "grad_norm": 1.0333510095833958, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 15507 + }, + { + "epoch": 0.15508, + "grad_norm": 1.6523371746178406, + "learning_rate": 0.003, + "loss": 4.074, + "step": 15508 + }, + { + "epoch": 0.15509, + "grad_norm": 1.151636731367369, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15509 + }, + { + "epoch": 0.1551, + "grad_norm": 1.5218639023836638, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 15510 + }, + { + "epoch": 0.15511, + "grad_norm": 0.9612598194560834, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 15511 + }, + { + "epoch": 0.15512, + "grad_norm": 1.1547449258340936, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 15512 + }, + { + "epoch": 0.15513, + "grad_norm": 0.9827899496468655, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15513 + }, + { + "epoch": 0.15514, + "grad_norm": 1.3755150660300457, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 15514 + }, + { + "epoch": 0.15515, + "grad_norm": 1.1498310420367837, + "learning_rate": 0.003, + "loss": 4.065, + "step": 15515 + }, + { + "epoch": 0.15516, + "grad_norm": 1.26169842384535, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15516 + }, + { + "epoch": 0.15517, + "grad_norm": 1.2913450985149533, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15517 + }, + { + "epoch": 0.15518, + "grad_norm": 1.3689027325340566, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 15518 + }, + { + "epoch": 0.15519, + "grad_norm": 1.11909032130159, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15519 + }, + { + "epoch": 0.1552, + "grad_norm": 1.3565836554568418, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 15520 + }, + { + "epoch": 0.15521, + "grad_norm": 1.1839474669898502, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 15521 + }, + { + "epoch": 0.15522, + "grad_norm": 1.2984200170694058, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 15522 + }, + { + "epoch": 0.15523, + "grad_norm": 1.0926685665559404, + "learning_rate": 0.003, + "loss": 4.084, + "step": 15523 + }, + { + "epoch": 0.15524, + "grad_norm": 1.2953543711923454, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 15524 + }, + { + "epoch": 0.15525, + "grad_norm": 1.1528653541075322, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 15525 + }, + { + "epoch": 0.15526, + "grad_norm": 1.5481099803544527, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 15526 + }, + { + "epoch": 0.15527, + "grad_norm": 1.1766676107693896, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 15527 + }, + { + "epoch": 0.15528, + "grad_norm": 1.323063424820846, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15528 + }, + { + "epoch": 0.15529, + "grad_norm": 1.2147770426421365, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15529 + }, + { + "epoch": 0.1553, + "grad_norm": 1.4566094408042551, + "learning_rate": 0.003, + "loss": 4.098, + "step": 15530 + }, + { + "epoch": 0.15531, + "grad_norm": 1.0736352163009661, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 15531 + }, + { + "epoch": 0.15532, + "grad_norm": 1.2415821433548437, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 15532 + }, + { + "epoch": 0.15533, + "grad_norm": 1.2764666863250977, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 15533 + }, + { + "epoch": 0.15534, + "grad_norm": 1.2784009664834688, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 15534 + }, + { + "epoch": 0.15535, + "grad_norm": 1.2456733455217581, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 15535 + }, + { + "epoch": 0.15536, + "grad_norm": 1.4517271950516681, + "learning_rate": 0.003, + "loss": 4.064, + "step": 15536 + }, + { + "epoch": 0.15537, + "grad_norm": 1.054271128214894, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 15537 + }, + { + "epoch": 0.15538, + "grad_norm": 1.3506048182368122, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15538 + }, + { + "epoch": 0.15539, + "grad_norm": 1.1058801360274975, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 15539 + }, + { + "epoch": 0.1554, + "grad_norm": 1.1772540177736688, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 15540 + }, + { + "epoch": 0.15541, + "grad_norm": 1.148712931209117, + "learning_rate": 0.003, + "loss": 4.09, + "step": 15541 + }, + { + "epoch": 0.15542, + "grad_norm": 1.2879276566085887, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15542 + }, + { + "epoch": 0.15543, + "grad_norm": 1.276845100107273, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 15543 + }, + { + "epoch": 0.15544, + "grad_norm": 1.3452703390926941, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 15544 + }, + { + "epoch": 0.15545, + "grad_norm": 1.472933607340616, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 15545 + }, + { + "epoch": 0.15546, + "grad_norm": 1.119633066296943, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 15546 + }, + { + "epoch": 0.15547, + "grad_norm": 1.370763514248208, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15547 + }, + { + "epoch": 0.15548, + "grad_norm": 1.112782868303845, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 15548 + }, + { + "epoch": 0.15549, + "grad_norm": 1.3146426369934066, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 15549 + }, + { + "epoch": 0.1555, + "grad_norm": 1.1262815670981912, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 15550 + }, + { + "epoch": 0.15551, + "grad_norm": 1.4412305467903548, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 15551 + }, + { + "epoch": 0.15552, + "grad_norm": 1.0376583191100974, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 15552 + }, + { + "epoch": 0.15553, + "grad_norm": 1.4330994659926226, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 15553 + }, + { + "epoch": 0.15554, + "grad_norm": 1.0885961883362105, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15554 + }, + { + "epoch": 0.15555, + "grad_norm": 1.6627063289448987, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 15555 + }, + { + "epoch": 0.15556, + "grad_norm": 1.1627993597262296, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 15556 + }, + { + "epoch": 0.15557, + "grad_norm": 1.181147045209965, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 15557 + }, + { + "epoch": 0.15558, + "grad_norm": 1.2773965711076947, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15558 + }, + { + "epoch": 0.15559, + "grad_norm": 1.3169844120615914, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 15559 + }, + { + "epoch": 0.1556, + "grad_norm": 1.141942025857773, + "learning_rate": 0.003, + "loss": 4.079, + "step": 15560 + }, + { + "epoch": 0.15561, + "grad_norm": 1.1580066271965455, + "learning_rate": 0.003, + "loss": 4.048, + "step": 15561 + }, + { + "epoch": 0.15562, + "grad_norm": 1.2786956245622974, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 15562 + }, + { + "epoch": 0.15563, + "grad_norm": 1.1445034130736649, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 15563 + }, + { + "epoch": 0.15564, + "grad_norm": 1.2313003477981035, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 15564 + }, + { + "epoch": 0.15565, + "grad_norm": 1.0935705973884606, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 15565 + }, + { + "epoch": 0.15566, + "grad_norm": 1.324605016905411, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15566 + }, + { + "epoch": 0.15567, + "grad_norm": 1.115542163926082, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15567 + }, + { + "epoch": 0.15568, + "grad_norm": 1.2483062696111795, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 15568 + }, + { + "epoch": 0.15569, + "grad_norm": 1.5644969184712574, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 15569 + }, + { + "epoch": 0.1557, + "grad_norm": 1.2953495515696785, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 15570 + }, + { + "epoch": 0.15571, + "grad_norm": 1.4506731944335949, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 15571 + }, + { + "epoch": 0.15572, + "grad_norm": 1.1401530574177883, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 15572 + }, + { + "epoch": 0.15573, + "grad_norm": 1.4880401906606766, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 15573 + }, + { + "epoch": 0.15574, + "grad_norm": 0.9606893059570549, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 15574 + }, + { + "epoch": 0.15575, + "grad_norm": 1.364810777937475, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 15575 + }, + { + "epoch": 0.15576, + "grad_norm": 1.2727795282851142, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 15576 + }, + { + "epoch": 0.15577, + "grad_norm": 1.1629652085594882, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 15577 + }, + { + "epoch": 0.15578, + "grad_norm": 1.512206308929831, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15578 + }, + { + "epoch": 0.15579, + "grad_norm": 1.2138603565111312, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 15579 + }, + { + "epoch": 0.1558, + "grad_norm": 1.1349088167246613, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 15580 + }, + { + "epoch": 0.15581, + "grad_norm": 1.3414979120902917, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 15581 + }, + { + "epoch": 0.15582, + "grad_norm": 1.201843636397302, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 15582 + }, + { + "epoch": 0.15583, + "grad_norm": 1.1345880097815193, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15583 + }, + { + "epoch": 0.15584, + "grad_norm": 1.0766859641248807, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 15584 + }, + { + "epoch": 0.15585, + "grad_norm": 1.5509132059441428, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 15585 + }, + { + "epoch": 0.15586, + "grad_norm": 1.2169131996422387, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15586 + }, + { + "epoch": 0.15587, + "grad_norm": 1.097994332980497, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 15587 + }, + { + "epoch": 0.15588, + "grad_norm": 1.2743281404250246, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 15588 + }, + { + "epoch": 0.15589, + "grad_norm": 1.1074285724019182, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 15589 + }, + { + "epoch": 0.1559, + "grad_norm": 1.1906649516913077, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 15590 + }, + { + "epoch": 0.15591, + "grad_norm": 1.560290123206607, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 15591 + }, + { + "epoch": 0.15592, + "grad_norm": 1.2565321454769511, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 15592 + }, + { + "epoch": 0.15593, + "grad_norm": 1.4275719733904104, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 15593 + }, + { + "epoch": 0.15594, + "grad_norm": 0.9907865538415417, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15594 + }, + { + "epoch": 0.15595, + "grad_norm": 1.4361657276878441, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 15595 + }, + { + "epoch": 0.15596, + "grad_norm": 0.9457249881724901, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 15596 + }, + { + "epoch": 0.15597, + "grad_norm": 1.4679972367121594, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15597 + }, + { + "epoch": 0.15598, + "grad_norm": 1.261988434296396, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 15598 + }, + { + "epoch": 0.15599, + "grad_norm": 1.1806554796870041, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 15599 + }, + { + "epoch": 0.156, + "grad_norm": 1.3879861431774005, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15600 + }, + { + "epoch": 0.15601, + "grad_norm": 0.9985869460398353, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 15601 + }, + { + "epoch": 0.15602, + "grad_norm": 1.2678765858709244, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 15602 + }, + { + "epoch": 0.15603, + "grad_norm": 0.9736360743365328, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 15603 + }, + { + "epoch": 0.15604, + "grad_norm": 1.2809955932621875, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 15604 + }, + { + "epoch": 0.15605, + "grad_norm": 1.148291316111474, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 15605 + }, + { + "epoch": 0.15606, + "grad_norm": 1.3178519897612266, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15606 + }, + { + "epoch": 0.15607, + "grad_norm": 1.2750891947071228, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 15607 + }, + { + "epoch": 0.15608, + "grad_norm": 1.4865355424381292, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 15608 + }, + { + "epoch": 0.15609, + "grad_norm": 1.2259715257260813, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15609 + }, + { + "epoch": 0.1561, + "grad_norm": 1.2060706359677797, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 15610 + }, + { + "epoch": 0.15611, + "grad_norm": 1.2223137806997504, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 15611 + }, + { + "epoch": 0.15612, + "grad_norm": 1.2042145879676174, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 15612 + }, + { + "epoch": 0.15613, + "grad_norm": 1.3304156227239157, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 15613 + }, + { + "epoch": 0.15614, + "grad_norm": 1.1535381824815405, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 15614 + }, + { + "epoch": 0.15615, + "grad_norm": 1.2497216772094886, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 15615 + }, + { + "epoch": 0.15616, + "grad_norm": 1.1589575226176436, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 15616 + }, + { + "epoch": 0.15617, + "grad_norm": 1.214026917629878, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 15617 + }, + { + "epoch": 0.15618, + "grad_norm": 1.3279395508300051, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15618 + }, + { + "epoch": 0.15619, + "grad_norm": 1.1274343927832446, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15619 + }, + { + "epoch": 0.1562, + "grad_norm": 1.4311028687528993, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 15620 + }, + { + "epoch": 0.15621, + "grad_norm": 1.0211279971828127, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 15621 + }, + { + "epoch": 0.15622, + "grad_norm": 1.6800417751483885, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 15622 + }, + { + "epoch": 0.15623, + "grad_norm": 1.1727981745287182, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 15623 + }, + { + "epoch": 0.15624, + "grad_norm": 1.117591637939842, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 15624 + }, + { + "epoch": 0.15625, + "grad_norm": 1.2896942605248825, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 15625 + }, + { + "epoch": 0.15626, + "grad_norm": 1.3008859506847203, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 15626 + }, + { + "epoch": 0.15627, + "grad_norm": 1.3904328415175207, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 15627 + }, + { + "epoch": 0.15628, + "grad_norm": 1.234645460229858, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 15628 + }, + { + "epoch": 0.15629, + "grad_norm": 1.1809067210150166, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 15629 + }, + { + "epoch": 0.1563, + "grad_norm": 1.3195640443236507, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 15630 + }, + { + "epoch": 0.15631, + "grad_norm": 1.1508156778152754, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 15631 + }, + { + "epoch": 0.15632, + "grad_norm": 1.4927637092891726, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 15632 + }, + { + "epoch": 0.15633, + "grad_norm": 1.105076574968614, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 15633 + }, + { + "epoch": 0.15634, + "grad_norm": 1.4510587816784541, + "learning_rate": 0.003, + "loss": 4.088, + "step": 15634 + }, + { + "epoch": 0.15635, + "grad_norm": 1.183188151744009, + "learning_rate": 0.003, + "loss": 4.067, + "step": 15635 + }, + { + "epoch": 0.15636, + "grad_norm": 1.2823032120205844, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15636 + }, + { + "epoch": 0.15637, + "grad_norm": 1.3473647772899666, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 15637 + }, + { + "epoch": 0.15638, + "grad_norm": 1.1079315472829223, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 15638 + }, + { + "epoch": 0.15639, + "grad_norm": 1.2366179096289918, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 15639 + }, + { + "epoch": 0.1564, + "grad_norm": 1.1307362871644169, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 15640 + }, + { + "epoch": 0.15641, + "grad_norm": 1.1955411780508391, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 15641 + }, + { + "epoch": 0.15642, + "grad_norm": 1.1580441924185543, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 15642 + }, + { + "epoch": 0.15643, + "grad_norm": 1.2320659842797703, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 15643 + }, + { + "epoch": 0.15644, + "grad_norm": 1.3220035146985878, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 15644 + }, + { + "epoch": 0.15645, + "grad_norm": 1.2045901852908378, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 15645 + }, + { + "epoch": 0.15646, + "grad_norm": 1.1124091615092142, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 15646 + }, + { + "epoch": 0.15647, + "grad_norm": 1.271170063053241, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 15647 + }, + { + "epoch": 0.15648, + "grad_norm": 1.3214408727366123, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 15648 + }, + { + "epoch": 0.15649, + "grad_norm": 1.187372164818155, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 15649 + }, + { + "epoch": 0.1565, + "grad_norm": 1.6176252258318626, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 15650 + }, + { + "epoch": 0.15651, + "grad_norm": 1.1386870708317176, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15651 + }, + { + "epoch": 0.15652, + "grad_norm": 1.3083055135317863, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 15652 + }, + { + "epoch": 0.15653, + "grad_norm": 1.40915147872283, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15653 + }, + { + "epoch": 0.15654, + "grad_norm": 0.8538626665931397, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 15654 + }, + { + "epoch": 0.15655, + "grad_norm": 0.9678335918623943, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 15655 + }, + { + "epoch": 0.15656, + "grad_norm": 1.5401136030257625, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 15656 + }, + { + "epoch": 0.15657, + "grad_norm": 0.9994550280432748, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 15657 + }, + { + "epoch": 0.15658, + "grad_norm": 1.4934189160213556, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 15658 + }, + { + "epoch": 0.15659, + "grad_norm": 1.0734913281470804, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 15659 + }, + { + "epoch": 0.1566, + "grad_norm": 1.3201131927167937, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 15660 + }, + { + "epoch": 0.15661, + "grad_norm": 1.1840754960357602, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15661 + }, + { + "epoch": 0.15662, + "grad_norm": 1.3540471955572964, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 15662 + }, + { + "epoch": 0.15663, + "grad_norm": 1.1679447196377668, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15663 + }, + { + "epoch": 0.15664, + "grad_norm": 1.2586333202804696, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 15664 + }, + { + "epoch": 0.15665, + "grad_norm": 1.3458507787994516, + "learning_rate": 0.003, + "loss": 4.081, + "step": 15665 + }, + { + "epoch": 0.15666, + "grad_norm": 1.2384653285613236, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 15666 + }, + { + "epoch": 0.15667, + "grad_norm": 1.2330148927447746, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 15667 + }, + { + "epoch": 0.15668, + "grad_norm": 1.3681828569052863, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 15668 + }, + { + "epoch": 0.15669, + "grad_norm": 1.2420398174536744, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 15669 + }, + { + "epoch": 0.1567, + "grad_norm": 1.0635111815096139, + "learning_rate": 0.003, + "loss": 4.046, + "step": 15670 + }, + { + "epoch": 0.15671, + "grad_norm": 1.4294078233540295, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15671 + }, + { + "epoch": 0.15672, + "grad_norm": 1.235341350191663, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 15672 + }, + { + "epoch": 0.15673, + "grad_norm": 1.2249451316293496, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 15673 + }, + { + "epoch": 0.15674, + "grad_norm": 1.3152866164886607, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 15674 + }, + { + "epoch": 0.15675, + "grad_norm": 1.2188115890433888, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15675 + }, + { + "epoch": 0.15676, + "grad_norm": 1.2487670894851186, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 15676 + }, + { + "epoch": 0.15677, + "grad_norm": 1.1509127594545203, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 15677 + }, + { + "epoch": 0.15678, + "grad_norm": 1.2295049671565317, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 15678 + }, + { + "epoch": 0.15679, + "grad_norm": 1.2564199180091304, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15679 + }, + { + "epoch": 0.1568, + "grad_norm": 1.2011586651713093, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 15680 + }, + { + "epoch": 0.15681, + "grad_norm": 1.4087063015759582, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 15681 + }, + { + "epoch": 0.15682, + "grad_norm": 1.191937364455209, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15682 + }, + { + "epoch": 0.15683, + "grad_norm": 1.4745662297874194, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 15683 + }, + { + "epoch": 0.15684, + "grad_norm": 1.001726375746545, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15684 + }, + { + "epoch": 0.15685, + "grad_norm": 1.29555761953909, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 15685 + }, + { + "epoch": 0.15686, + "grad_norm": 1.2168139453733735, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 15686 + }, + { + "epoch": 0.15687, + "grad_norm": 1.1113718267499884, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 15687 + }, + { + "epoch": 0.15688, + "grad_norm": 1.2747869571151393, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 15688 + }, + { + "epoch": 0.15689, + "grad_norm": 1.0391721449384115, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 15689 + }, + { + "epoch": 0.1569, + "grad_norm": 1.329338261019493, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 15690 + }, + { + "epoch": 0.15691, + "grad_norm": 1.038286920678713, + "learning_rate": 0.003, + "loss": 4.065, + "step": 15691 + }, + { + "epoch": 0.15692, + "grad_norm": 1.5112599084156186, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 15692 + }, + { + "epoch": 0.15693, + "grad_norm": 1.0740745559288278, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15693 + }, + { + "epoch": 0.15694, + "grad_norm": 1.6952246256536485, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 15694 + }, + { + "epoch": 0.15695, + "grad_norm": 1.1356207728741108, + "learning_rate": 0.003, + "loss": 4.074, + "step": 15695 + }, + { + "epoch": 0.15696, + "grad_norm": 1.600737678873874, + "learning_rate": 0.003, + "loss": 4.078, + "step": 15696 + }, + { + "epoch": 0.15697, + "grad_norm": 1.3798245399881897, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 15697 + }, + { + "epoch": 0.15698, + "grad_norm": 1.2319880529302756, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 15698 + }, + { + "epoch": 0.15699, + "grad_norm": 1.162085383918692, + "learning_rate": 0.003, + "loss": 4.11, + "step": 15699 + }, + { + "epoch": 0.157, + "grad_norm": 1.1861369428968611, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15700 + }, + { + "epoch": 0.15701, + "grad_norm": 1.1022028086837685, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 15701 + }, + { + "epoch": 0.15702, + "grad_norm": 1.3519041828221627, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 15702 + }, + { + "epoch": 0.15703, + "grad_norm": 1.2546178713888105, + "learning_rate": 0.003, + "loss": 4.077, + "step": 15703 + }, + { + "epoch": 0.15704, + "grad_norm": 1.3885754744550656, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15704 + }, + { + "epoch": 0.15705, + "grad_norm": 1.0136676281040595, + "learning_rate": 0.003, + "loss": 4.053, + "step": 15705 + }, + { + "epoch": 0.15706, + "grad_norm": 1.479423464466427, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15706 + }, + { + "epoch": 0.15707, + "grad_norm": 1.0692242216649765, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 15707 + }, + { + "epoch": 0.15708, + "grad_norm": 1.6514457947652423, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 15708 + }, + { + "epoch": 0.15709, + "grad_norm": 1.0980061962964487, + "learning_rate": 0.003, + "loss": 4.076, + "step": 15709 + }, + { + "epoch": 0.1571, + "grad_norm": 1.1629910875995582, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 15710 + }, + { + "epoch": 0.15711, + "grad_norm": 1.3835072420169645, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15711 + }, + { + "epoch": 0.15712, + "grad_norm": 0.9814395529013048, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 15712 + }, + { + "epoch": 0.15713, + "grad_norm": 1.3044701713468254, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15713 + }, + { + "epoch": 0.15714, + "grad_norm": 1.0128719798895127, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15714 + }, + { + "epoch": 0.15715, + "grad_norm": 1.4611873649306613, + "learning_rate": 0.003, + "loss": 4.059, + "step": 15715 + }, + { + "epoch": 0.15716, + "grad_norm": 1.1673127264656047, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 15716 + }, + { + "epoch": 0.15717, + "grad_norm": 1.6321591972869431, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 15717 + }, + { + "epoch": 0.15718, + "grad_norm": 1.194441152695339, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15718 + }, + { + "epoch": 0.15719, + "grad_norm": 1.4729419167542548, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15719 + }, + { + "epoch": 0.1572, + "grad_norm": 1.153977332613549, + "learning_rate": 0.003, + "loss": 4.077, + "step": 15720 + }, + { + "epoch": 0.15721, + "grad_norm": 1.3469995913475945, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 15721 + }, + { + "epoch": 0.15722, + "grad_norm": 1.1257746335495376, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 15722 + }, + { + "epoch": 0.15723, + "grad_norm": 1.273311840186899, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 15723 + }, + { + "epoch": 0.15724, + "grad_norm": 1.2425020612828945, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 15724 + }, + { + "epoch": 0.15725, + "grad_norm": 1.0998561464008263, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15725 + }, + { + "epoch": 0.15726, + "grad_norm": 1.2074450589504484, + "learning_rate": 0.003, + "loss": 4.044, + "step": 15726 + }, + { + "epoch": 0.15727, + "grad_norm": 1.2465016405631146, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 15727 + }, + { + "epoch": 0.15728, + "grad_norm": 1.2308298672569458, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15728 + }, + { + "epoch": 0.15729, + "grad_norm": 1.2353477693601442, + "learning_rate": 0.003, + "loss": 4.076, + "step": 15729 + }, + { + "epoch": 0.1573, + "grad_norm": 1.1997809145637395, + "learning_rate": 0.003, + "loss": 4.075, + "step": 15730 + }, + { + "epoch": 0.15731, + "grad_norm": 1.055664502187246, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 15731 + }, + { + "epoch": 0.15732, + "grad_norm": 1.1251817771876742, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 15732 + }, + { + "epoch": 0.15733, + "grad_norm": 1.2684826751675962, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 15733 + }, + { + "epoch": 0.15734, + "grad_norm": 1.176261122634088, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 15734 + }, + { + "epoch": 0.15735, + "grad_norm": 1.3011228658316907, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 15735 + }, + { + "epoch": 0.15736, + "grad_norm": 1.4022393363425198, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 15736 + }, + { + "epoch": 0.15737, + "grad_norm": 1.4943081663486804, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 15737 + }, + { + "epoch": 0.15738, + "grad_norm": 1.2095662083063563, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 15738 + }, + { + "epoch": 0.15739, + "grad_norm": 1.1900215263482639, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 15739 + }, + { + "epoch": 0.1574, + "grad_norm": 1.498445894217828, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15740 + }, + { + "epoch": 0.15741, + "grad_norm": 1.247342175522623, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 15741 + }, + { + "epoch": 0.15742, + "grad_norm": 1.281881136558292, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15742 + }, + { + "epoch": 0.15743, + "grad_norm": 1.1637487580323629, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 15743 + }, + { + "epoch": 0.15744, + "grad_norm": 1.6250204266507926, + "learning_rate": 0.003, + "loss": 4.051, + "step": 15744 + }, + { + "epoch": 0.15745, + "grad_norm": 1.1120053816868876, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 15745 + }, + { + "epoch": 0.15746, + "grad_norm": 1.1806132697403393, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 15746 + }, + { + "epoch": 0.15747, + "grad_norm": 1.3753350852825859, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 15747 + }, + { + "epoch": 0.15748, + "grad_norm": 1.3617899171508971, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 15748 + }, + { + "epoch": 0.15749, + "grad_norm": 1.224551816489951, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 15749 + }, + { + "epoch": 0.1575, + "grad_norm": 1.2329503063424165, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15750 + }, + { + "epoch": 0.15751, + "grad_norm": 1.3487282436055932, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15751 + }, + { + "epoch": 0.15752, + "grad_norm": 0.8983113736476528, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 15752 + }, + { + "epoch": 0.15753, + "grad_norm": 1.1205134642362236, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 15753 + }, + { + "epoch": 0.15754, + "grad_norm": 1.1108146780065227, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 15754 + }, + { + "epoch": 0.15755, + "grad_norm": 1.3179332530517076, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 15755 + }, + { + "epoch": 0.15756, + "grad_norm": 1.2340293618335434, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 15756 + }, + { + "epoch": 0.15757, + "grad_norm": 1.2067630444776245, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 15757 + }, + { + "epoch": 0.15758, + "grad_norm": 1.358633075141287, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 15758 + }, + { + "epoch": 0.15759, + "grad_norm": 1.2713680632283646, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 15759 + }, + { + "epoch": 0.1576, + "grad_norm": 1.429606807726027, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 15760 + }, + { + "epoch": 0.15761, + "grad_norm": 1.1698066698894496, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 15761 + }, + { + "epoch": 0.15762, + "grad_norm": 1.3301859759951484, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 15762 + }, + { + "epoch": 0.15763, + "grad_norm": 1.0043360109228892, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 15763 + }, + { + "epoch": 0.15764, + "grad_norm": 1.2805820162353942, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15764 + }, + { + "epoch": 0.15765, + "grad_norm": 1.084130383097619, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15765 + }, + { + "epoch": 0.15766, + "grad_norm": 1.4455323926932548, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 15766 + }, + { + "epoch": 0.15767, + "grad_norm": 1.2458212253441476, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 15767 + }, + { + "epoch": 0.15768, + "grad_norm": 1.4079166675160757, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15768 + }, + { + "epoch": 0.15769, + "grad_norm": 1.0886689940232694, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 15769 + }, + { + "epoch": 0.1577, + "grad_norm": 1.33669784595432, + "learning_rate": 0.003, + "loss": 4.073, + "step": 15770 + }, + { + "epoch": 0.15771, + "grad_norm": 1.2659884771872834, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 15771 + }, + { + "epoch": 0.15772, + "grad_norm": 1.2676175178618903, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 15772 + }, + { + "epoch": 0.15773, + "grad_norm": 1.3247024497021374, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 15773 + }, + { + "epoch": 0.15774, + "grad_norm": 1.1798670009014236, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15774 + }, + { + "epoch": 0.15775, + "grad_norm": 1.2745844735563765, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 15775 + }, + { + "epoch": 0.15776, + "grad_norm": 1.2586926574914314, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 15776 + }, + { + "epoch": 0.15777, + "grad_norm": 1.13855349226103, + "learning_rate": 0.003, + "loss": 4.047, + "step": 15777 + }, + { + "epoch": 0.15778, + "grad_norm": 1.1802061114019076, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 15778 + }, + { + "epoch": 0.15779, + "grad_norm": 1.2664395752802722, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 15779 + }, + { + "epoch": 0.1578, + "grad_norm": 1.3060934404648121, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15780 + }, + { + "epoch": 0.15781, + "grad_norm": 1.380042239870967, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15781 + }, + { + "epoch": 0.15782, + "grad_norm": 1.087924226172932, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 15782 + }, + { + "epoch": 0.15783, + "grad_norm": 1.3430527522168907, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 15783 + }, + { + "epoch": 0.15784, + "grad_norm": 1.0104217717628523, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 15784 + }, + { + "epoch": 0.15785, + "grad_norm": 1.4267809040482224, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 15785 + }, + { + "epoch": 0.15786, + "grad_norm": 1.216574584344044, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 15786 + }, + { + "epoch": 0.15787, + "grad_norm": 1.212189380409753, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15787 + }, + { + "epoch": 0.15788, + "grad_norm": 1.2650434917429774, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 15788 + }, + { + "epoch": 0.15789, + "grad_norm": 1.3134562170382684, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 15789 + }, + { + "epoch": 0.1579, + "grad_norm": 1.0200074018220817, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15790 + }, + { + "epoch": 0.15791, + "grad_norm": 1.508952445135445, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 15791 + }, + { + "epoch": 0.15792, + "grad_norm": 1.0099881102959938, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15792 + }, + { + "epoch": 0.15793, + "grad_norm": 1.5182111830652727, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 15793 + }, + { + "epoch": 0.15794, + "grad_norm": 1.0140123650871258, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15794 + }, + { + "epoch": 0.15795, + "grad_norm": 1.630216472140459, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 15795 + }, + { + "epoch": 0.15796, + "grad_norm": 1.2356276021322978, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 15796 + }, + { + "epoch": 0.15797, + "grad_norm": 1.1195009082517002, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15797 + }, + { + "epoch": 0.15798, + "grad_norm": 1.2554496879615797, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 15798 + }, + { + "epoch": 0.15799, + "grad_norm": 1.1882650169054183, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 15799 + }, + { + "epoch": 0.158, + "grad_norm": 1.025180895775637, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 15800 + }, + { + "epoch": 0.15801, + "grad_norm": 1.3323650505869045, + "learning_rate": 0.003, + "loss": 4.07, + "step": 15801 + }, + { + "epoch": 0.15802, + "grad_norm": 1.1648391893077823, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15802 + }, + { + "epoch": 0.15803, + "grad_norm": 1.3263065106366811, + "learning_rate": 0.003, + "loss": 4.051, + "step": 15803 + }, + { + "epoch": 0.15804, + "grad_norm": 1.019953277288765, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 15804 + }, + { + "epoch": 0.15805, + "grad_norm": 1.3199518646915922, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 15805 + }, + { + "epoch": 0.15806, + "grad_norm": 1.1567590802816259, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 15806 + }, + { + "epoch": 0.15807, + "grad_norm": 1.2112833131281036, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 15807 + }, + { + "epoch": 0.15808, + "grad_norm": 1.543700800987061, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 15808 + }, + { + "epoch": 0.15809, + "grad_norm": 1.2148372812557238, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 15809 + }, + { + "epoch": 0.1581, + "grad_norm": 1.64595741304971, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15810 + }, + { + "epoch": 0.15811, + "grad_norm": 0.9473241818368024, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 15811 + }, + { + "epoch": 0.15812, + "grad_norm": 1.2074555335242219, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 15812 + }, + { + "epoch": 0.15813, + "grad_norm": 1.7567532028410888, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15813 + }, + { + "epoch": 0.15814, + "grad_norm": 1.1609304991514573, + "learning_rate": 0.003, + "loss": 4.083, + "step": 15814 + }, + { + "epoch": 0.15815, + "grad_norm": 1.315570986779992, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 15815 + }, + { + "epoch": 0.15816, + "grad_norm": 1.0790403099689905, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 15816 + }, + { + "epoch": 0.15817, + "grad_norm": 1.4012912339396493, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 15817 + }, + { + "epoch": 0.15818, + "grad_norm": 0.9990085088007167, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 15818 + }, + { + "epoch": 0.15819, + "grad_norm": 1.3108872436591201, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 15819 + }, + { + "epoch": 0.1582, + "grad_norm": 1.2860627087013567, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 15820 + }, + { + "epoch": 0.15821, + "grad_norm": 1.313384160400915, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 15821 + }, + { + "epoch": 0.15822, + "grad_norm": 1.2846586054114109, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 15822 + }, + { + "epoch": 0.15823, + "grad_norm": 1.1358164618290276, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15823 + }, + { + "epoch": 0.15824, + "grad_norm": 1.4458733395957897, + "learning_rate": 0.003, + "loss": 4.09, + "step": 15824 + }, + { + "epoch": 0.15825, + "grad_norm": 1.1737861594039298, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 15825 + }, + { + "epoch": 0.15826, + "grad_norm": 1.2185578330836349, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 15826 + }, + { + "epoch": 0.15827, + "grad_norm": 1.1982989834195201, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 15827 + }, + { + "epoch": 0.15828, + "grad_norm": 1.3110406043806941, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 15828 + }, + { + "epoch": 0.15829, + "grad_norm": 1.466595140772327, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 15829 + }, + { + "epoch": 0.1583, + "grad_norm": 1.1028802305304912, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 15830 + }, + { + "epoch": 0.15831, + "grad_norm": 1.4401414313447025, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 15831 + }, + { + "epoch": 0.15832, + "grad_norm": 1.0121391822966723, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 15832 + }, + { + "epoch": 0.15833, + "grad_norm": 1.3701569406549603, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 15833 + }, + { + "epoch": 0.15834, + "grad_norm": 1.0467743466126278, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 15834 + }, + { + "epoch": 0.15835, + "grad_norm": 1.4350110033054864, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15835 + }, + { + "epoch": 0.15836, + "grad_norm": 1.1655596804028503, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 15836 + }, + { + "epoch": 0.15837, + "grad_norm": 1.366809734456757, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 15837 + }, + { + "epoch": 0.15838, + "grad_norm": 1.1015343654202836, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 15838 + }, + { + "epoch": 0.15839, + "grad_norm": 1.497404289489368, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 15839 + }, + { + "epoch": 0.1584, + "grad_norm": 1.162686134233669, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 15840 + }, + { + "epoch": 0.15841, + "grad_norm": 1.4122265105717202, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 15841 + }, + { + "epoch": 0.15842, + "grad_norm": 1.0646479436981007, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 15842 + }, + { + "epoch": 0.15843, + "grad_norm": 1.4378996909096384, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 15843 + }, + { + "epoch": 0.15844, + "grad_norm": 1.1335472518084901, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 15844 + }, + { + "epoch": 0.15845, + "grad_norm": 1.202254551336859, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 15845 + }, + { + "epoch": 0.15846, + "grad_norm": 1.1023046241915795, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 15846 + }, + { + "epoch": 0.15847, + "grad_norm": 1.2241215062872985, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 15847 + }, + { + "epoch": 0.15848, + "grad_norm": 1.2360989485035663, + "learning_rate": 0.003, + "loss": 4.061, + "step": 15848 + }, + { + "epoch": 0.15849, + "grad_norm": 1.2851274263498322, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 15849 + }, + { + "epoch": 0.1585, + "grad_norm": 1.5063698866099746, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 15850 + }, + { + "epoch": 0.15851, + "grad_norm": 1.3370699560578674, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15851 + }, + { + "epoch": 0.15852, + "grad_norm": 1.0649414968565, + "learning_rate": 0.003, + "loss": 4.044, + "step": 15852 + }, + { + "epoch": 0.15853, + "grad_norm": 1.2923127203605576, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 15853 + }, + { + "epoch": 0.15854, + "grad_norm": 1.17886527455772, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 15854 + }, + { + "epoch": 0.15855, + "grad_norm": 1.110334224682506, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 15855 + }, + { + "epoch": 0.15856, + "grad_norm": 1.014714199014075, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15856 + }, + { + "epoch": 0.15857, + "grad_norm": 1.3186310147508855, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 15857 + }, + { + "epoch": 0.15858, + "grad_norm": 1.0082773124492816, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 15858 + }, + { + "epoch": 0.15859, + "grad_norm": 1.341166900129246, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 15859 + }, + { + "epoch": 0.1586, + "grad_norm": 0.9757690221429678, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15860 + }, + { + "epoch": 0.15861, + "grad_norm": 1.5181678524071738, + "learning_rate": 0.003, + "loss": 4.075, + "step": 15861 + }, + { + "epoch": 0.15862, + "grad_norm": 1.4198118300073572, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 15862 + }, + { + "epoch": 0.15863, + "grad_norm": 1.3250680079462993, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 15863 + }, + { + "epoch": 0.15864, + "grad_norm": 1.1219890722185781, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 15864 + }, + { + "epoch": 0.15865, + "grad_norm": 1.3773526564893919, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15865 + }, + { + "epoch": 0.15866, + "grad_norm": 1.1187512363140182, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15866 + }, + { + "epoch": 0.15867, + "grad_norm": 1.3873696864503684, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 15867 + }, + { + "epoch": 0.15868, + "grad_norm": 1.05050287732908, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 15868 + }, + { + "epoch": 0.15869, + "grad_norm": 1.3226400567507672, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 15869 + }, + { + "epoch": 0.1587, + "grad_norm": 1.112817535635143, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 15870 + }, + { + "epoch": 0.15871, + "grad_norm": 1.1031885680213505, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15871 + }, + { + "epoch": 0.15872, + "grad_norm": 1.3063976521351819, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 15872 + }, + { + "epoch": 0.15873, + "grad_norm": 1.3075266758431832, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 15873 + }, + { + "epoch": 0.15874, + "grad_norm": 1.4796353974658847, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15874 + }, + { + "epoch": 0.15875, + "grad_norm": 1.1544673497341547, + "learning_rate": 0.003, + "loss": 4.044, + "step": 15875 + }, + { + "epoch": 0.15876, + "grad_norm": 1.2104504771981357, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 15876 + }, + { + "epoch": 0.15877, + "grad_norm": 1.3874317918786958, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 15877 + }, + { + "epoch": 0.15878, + "grad_norm": 0.8694977231904792, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 15878 + }, + { + "epoch": 0.15879, + "grad_norm": 0.9167007165846688, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15879 + }, + { + "epoch": 0.1588, + "grad_norm": 1.3094180982937673, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 15880 + }, + { + "epoch": 0.15881, + "grad_norm": 1.0940458833334197, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 15881 + }, + { + "epoch": 0.15882, + "grad_norm": 1.6272294752909982, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 15882 + }, + { + "epoch": 0.15883, + "grad_norm": 0.953058667318075, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 15883 + }, + { + "epoch": 0.15884, + "grad_norm": 1.3800984250047426, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 15884 + }, + { + "epoch": 0.15885, + "grad_norm": 1.0469522999452336, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 15885 + }, + { + "epoch": 0.15886, + "grad_norm": 1.3807583853414365, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 15886 + }, + { + "epoch": 0.15887, + "grad_norm": 1.1689417177816546, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 15887 + }, + { + "epoch": 0.15888, + "grad_norm": 1.166352545373565, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 15888 + }, + { + "epoch": 0.15889, + "grad_norm": 1.1497410035203872, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 15889 + }, + { + "epoch": 0.1589, + "grad_norm": 1.1474521902301942, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 15890 + }, + { + "epoch": 0.15891, + "grad_norm": 1.378472287993341, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 15891 + }, + { + "epoch": 0.15892, + "grad_norm": 1.0524190751041174, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 15892 + }, + { + "epoch": 0.15893, + "grad_norm": 1.4838436770972758, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 15893 + }, + { + "epoch": 0.15894, + "grad_norm": 1.4774807045801406, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 15894 + }, + { + "epoch": 0.15895, + "grad_norm": 1.6066140641730953, + "learning_rate": 0.003, + "loss": 4.084, + "step": 15895 + }, + { + "epoch": 0.15896, + "grad_norm": 1.1622148299667838, + "learning_rate": 0.003, + "loss": 4.046, + "step": 15896 + }, + { + "epoch": 0.15897, + "grad_norm": 1.1931281726448333, + "learning_rate": 0.003, + "loss": 4.036, + "step": 15897 + }, + { + "epoch": 0.15898, + "grad_norm": 1.302918768938021, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 15898 + }, + { + "epoch": 0.15899, + "grad_norm": 1.2655608973386931, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 15899 + }, + { + "epoch": 0.159, + "grad_norm": 1.3007264409876362, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15900 + }, + { + "epoch": 0.15901, + "grad_norm": 1.4017728789526172, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 15901 + }, + { + "epoch": 0.15902, + "grad_norm": 1.096907458183354, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 15902 + }, + { + "epoch": 0.15903, + "grad_norm": 1.148682715983762, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 15903 + }, + { + "epoch": 0.15904, + "grad_norm": 1.2946860163634188, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 15904 + }, + { + "epoch": 0.15905, + "grad_norm": 1.122481448668236, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 15905 + }, + { + "epoch": 0.15906, + "grad_norm": 1.5089008973010962, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 15906 + }, + { + "epoch": 0.15907, + "grad_norm": 1.2836271686701424, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 15907 + }, + { + "epoch": 0.15908, + "grad_norm": 1.05237679745181, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 15908 + }, + { + "epoch": 0.15909, + "grad_norm": 1.2775159608553672, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 15909 + }, + { + "epoch": 0.1591, + "grad_norm": 1.2042252711867507, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 15910 + }, + { + "epoch": 0.15911, + "grad_norm": 1.2424910856410036, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 15911 + }, + { + "epoch": 0.15912, + "grad_norm": 1.004066165602063, + "learning_rate": 0.003, + "loss": 4.067, + "step": 15912 + }, + { + "epoch": 0.15913, + "grad_norm": 1.2899718102539786, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 15913 + }, + { + "epoch": 0.15914, + "grad_norm": 1.265411655883751, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 15914 + }, + { + "epoch": 0.15915, + "grad_norm": 1.3841302073176251, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 15915 + }, + { + "epoch": 0.15916, + "grad_norm": 1.0780930650512748, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 15916 + }, + { + "epoch": 0.15917, + "grad_norm": 1.3632698488776467, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 15917 + }, + { + "epoch": 0.15918, + "grad_norm": 1.3193444004494173, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 15918 + }, + { + "epoch": 0.15919, + "grad_norm": 1.2523241950513657, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 15919 + }, + { + "epoch": 0.1592, + "grad_norm": 1.1129624685015624, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 15920 + }, + { + "epoch": 0.15921, + "grad_norm": 1.2390879797094592, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 15921 + }, + { + "epoch": 0.15922, + "grad_norm": 1.018065075102559, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 15922 + }, + { + "epoch": 0.15923, + "grad_norm": 1.3992497510637925, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 15923 + }, + { + "epoch": 0.15924, + "grad_norm": 0.9807181832191064, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 15924 + }, + { + "epoch": 0.15925, + "grad_norm": 1.443845094545865, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 15925 + }, + { + "epoch": 0.15926, + "grad_norm": 1.3191998073770685, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15926 + }, + { + "epoch": 0.15927, + "grad_norm": 1.4318050152423651, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 15927 + }, + { + "epoch": 0.15928, + "grad_norm": 1.5445127896410225, + "learning_rate": 0.003, + "loss": 4.071, + "step": 15928 + }, + { + "epoch": 0.15929, + "grad_norm": 1.646914315146204, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 15929 + }, + { + "epoch": 0.1593, + "grad_norm": 0.9881917367102422, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15930 + }, + { + "epoch": 0.15931, + "grad_norm": 1.4051895948329052, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 15931 + }, + { + "epoch": 0.15932, + "grad_norm": 1.3037727964692034, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 15932 + }, + { + "epoch": 0.15933, + "grad_norm": 1.4378668261881007, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 15933 + }, + { + "epoch": 0.15934, + "grad_norm": 1.1431656766108897, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 15934 + }, + { + "epoch": 0.15935, + "grad_norm": 1.124566753972068, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 15935 + }, + { + "epoch": 0.15936, + "grad_norm": 1.16275651830534, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 15936 + }, + { + "epoch": 0.15937, + "grad_norm": 1.2342222553688793, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 15937 + }, + { + "epoch": 0.15938, + "grad_norm": 1.19713324059584, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 15938 + }, + { + "epoch": 0.15939, + "grad_norm": 1.1509074055799497, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15939 + }, + { + "epoch": 0.1594, + "grad_norm": 1.3990238341878423, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15940 + }, + { + "epoch": 0.15941, + "grad_norm": 1.12517011632491, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 15941 + }, + { + "epoch": 0.15942, + "grad_norm": 1.4062696973627182, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15942 + }, + { + "epoch": 0.15943, + "grad_norm": 1.2298148700416418, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 15943 + }, + { + "epoch": 0.15944, + "grad_norm": 1.0423298803055772, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 15944 + }, + { + "epoch": 0.15945, + "grad_norm": 1.5119355212233925, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 15945 + }, + { + "epoch": 0.15946, + "grad_norm": 1.1043321719037542, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 15946 + }, + { + "epoch": 0.15947, + "grad_norm": 1.5754318100897307, + "learning_rate": 0.003, + "loss": 4.062, + "step": 15947 + }, + { + "epoch": 0.15948, + "grad_norm": 1.2247612008785362, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 15948 + }, + { + "epoch": 0.15949, + "grad_norm": 1.2780057860668521, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 15949 + }, + { + "epoch": 0.1595, + "grad_norm": 1.0935819702783598, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 15950 + }, + { + "epoch": 0.15951, + "grad_norm": 1.4264660530503124, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 15951 + }, + { + "epoch": 0.15952, + "grad_norm": 1.4432776864714585, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 15952 + }, + { + "epoch": 0.15953, + "grad_norm": 1.0936306035911116, + "learning_rate": 0.003, + "loss": 4.082, + "step": 15953 + }, + { + "epoch": 0.15954, + "grad_norm": 1.3093780200539074, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 15954 + }, + { + "epoch": 0.15955, + "grad_norm": 1.1337523344829086, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 15955 + }, + { + "epoch": 0.15956, + "grad_norm": 1.2976729310465538, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 15956 + }, + { + "epoch": 0.15957, + "grad_norm": 1.1486881406964997, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 15957 + }, + { + "epoch": 0.15958, + "grad_norm": 1.1649483196714951, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 15958 + }, + { + "epoch": 0.15959, + "grad_norm": 1.0751428111362795, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 15959 + }, + { + "epoch": 0.1596, + "grad_norm": 1.525072047246104, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 15960 + }, + { + "epoch": 0.15961, + "grad_norm": 1.2430663142139295, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 15961 + }, + { + "epoch": 0.15962, + "grad_norm": 1.3229566512908846, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 15962 + }, + { + "epoch": 0.15963, + "grad_norm": 1.2326710037686879, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 15963 + }, + { + "epoch": 0.15964, + "grad_norm": 1.178974318726838, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15964 + }, + { + "epoch": 0.15965, + "grad_norm": 1.4016731036877232, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 15965 + }, + { + "epoch": 0.15966, + "grad_norm": 1.346665660360529, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 15966 + }, + { + "epoch": 0.15967, + "grad_norm": 1.5714256770776949, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 15967 + }, + { + "epoch": 0.15968, + "grad_norm": 1.174841832382644, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 15968 + }, + { + "epoch": 0.15969, + "grad_norm": 1.1664948385353457, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 15969 + }, + { + "epoch": 0.1597, + "grad_norm": 1.1363541719355623, + "learning_rate": 0.003, + "loss": 4.088, + "step": 15970 + }, + { + "epoch": 0.15971, + "grad_norm": 1.2671228496799418, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 15971 + }, + { + "epoch": 0.15972, + "grad_norm": 1.3127969455545418, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 15972 + }, + { + "epoch": 0.15973, + "grad_norm": 1.353167208597876, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15973 + }, + { + "epoch": 0.15974, + "grad_norm": 0.9988635512343115, + "learning_rate": 0.003, + "loss": 4.079, + "step": 15974 + }, + { + "epoch": 0.15975, + "grad_norm": 1.4102942267563905, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 15975 + }, + { + "epoch": 0.15976, + "grad_norm": 1.0329889844965574, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 15976 + }, + { + "epoch": 0.15977, + "grad_norm": 1.4709554695572877, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 15977 + }, + { + "epoch": 0.15978, + "grad_norm": 1.0835729764776088, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 15978 + }, + { + "epoch": 0.15979, + "grad_norm": 1.3346053161633171, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15979 + }, + { + "epoch": 0.1598, + "grad_norm": 1.4367864476983094, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 15980 + }, + { + "epoch": 0.15981, + "grad_norm": 1.211044054434938, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 15981 + }, + { + "epoch": 0.15982, + "grad_norm": 1.130377839272506, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 15982 + }, + { + "epoch": 0.15983, + "grad_norm": 1.199590883702763, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 15983 + }, + { + "epoch": 0.15984, + "grad_norm": 1.2234798884021005, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 15984 + }, + { + "epoch": 0.15985, + "grad_norm": 1.0621335951197872, + "learning_rate": 0.003, + "loss": 4.075, + "step": 15985 + }, + { + "epoch": 0.15986, + "grad_norm": 1.4802870810804452, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 15986 + }, + { + "epoch": 0.15987, + "grad_norm": 1.3745377945452901, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 15987 + }, + { + "epoch": 0.15988, + "grad_norm": 1.2800826333131476, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 15988 + }, + { + "epoch": 0.15989, + "grad_norm": 1.249017924347872, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 15989 + }, + { + "epoch": 0.1599, + "grad_norm": 1.087436131985846, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 15990 + }, + { + "epoch": 0.15991, + "grad_norm": 1.6208015102893236, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 15991 + }, + { + "epoch": 0.15992, + "grad_norm": 0.9553147533503118, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 15992 + }, + { + "epoch": 0.15993, + "grad_norm": 1.2399004682165844, + "learning_rate": 0.003, + "loss": 4.08, + "step": 15993 + }, + { + "epoch": 0.15994, + "grad_norm": 1.151216794905892, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 15994 + }, + { + "epoch": 0.15995, + "grad_norm": 1.209520667594038, + "learning_rate": 0.003, + "loss": 4.058, + "step": 15995 + }, + { + "epoch": 0.15996, + "grad_norm": 1.2654515123798205, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 15996 + }, + { + "epoch": 0.15997, + "grad_norm": 1.090738242669251, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 15997 + }, + { + "epoch": 0.15998, + "grad_norm": 1.6241870427660114, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 15998 + }, + { + "epoch": 0.15999, + "grad_norm": 1.037839226991951, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 15999 + }, + { + "epoch": 0.16, + "grad_norm": 1.468254338539844, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 16000 + }, + { + "epoch": 0.16001, + "grad_norm": 1.0569871997225846, + "learning_rate": 0.003, + "loss": 4.065, + "step": 16001 + }, + { + "epoch": 0.16002, + "grad_norm": 1.3543454173351195, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16002 + }, + { + "epoch": 0.16003, + "grad_norm": 1.219238499758112, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 16003 + }, + { + "epoch": 0.16004, + "grad_norm": 1.2148919686004898, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16004 + }, + { + "epoch": 0.16005, + "grad_norm": 1.3208705578435487, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 16005 + }, + { + "epoch": 0.16006, + "grad_norm": 1.3542892074931192, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16006 + }, + { + "epoch": 0.16007, + "grad_norm": 1.3671824625496356, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16007 + }, + { + "epoch": 0.16008, + "grad_norm": 1.0668311187906165, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 16008 + }, + { + "epoch": 0.16009, + "grad_norm": 1.3470746547863566, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 16009 + }, + { + "epoch": 0.1601, + "grad_norm": 1.0772154800713234, + "learning_rate": 0.003, + "loss": 4.049, + "step": 16010 + }, + { + "epoch": 0.16011, + "grad_norm": 1.3510199835884709, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 16011 + }, + { + "epoch": 0.16012, + "grad_norm": 1.1924190650854338, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 16012 + }, + { + "epoch": 0.16013, + "grad_norm": 1.1614570082005813, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 16013 + }, + { + "epoch": 0.16014, + "grad_norm": 1.2943798068788903, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16014 + }, + { + "epoch": 0.16015, + "grad_norm": 1.3157424343117976, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16015 + }, + { + "epoch": 0.16016, + "grad_norm": 1.2977011987523548, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 16016 + }, + { + "epoch": 0.16017, + "grad_norm": 1.2413929690186232, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 16017 + }, + { + "epoch": 0.16018, + "grad_norm": 1.6940441644090516, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 16018 + }, + { + "epoch": 0.16019, + "grad_norm": 1.3383232864298142, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 16019 + }, + { + "epoch": 0.1602, + "grad_norm": 1.214208237044485, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16020 + }, + { + "epoch": 0.16021, + "grad_norm": 1.385184359152161, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 16021 + }, + { + "epoch": 0.16022, + "grad_norm": 1.1521194968074087, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16022 + }, + { + "epoch": 0.16023, + "grad_norm": 1.2034260544674427, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 16023 + }, + { + "epoch": 0.16024, + "grad_norm": 1.024534410119776, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16024 + }, + { + "epoch": 0.16025, + "grad_norm": 1.535277051756849, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 16025 + }, + { + "epoch": 0.16026, + "grad_norm": 0.9833674829901832, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 16026 + }, + { + "epoch": 0.16027, + "grad_norm": 1.4822598816434278, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 16027 + }, + { + "epoch": 0.16028, + "grad_norm": 1.2765333691830032, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 16028 + }, + { + "epoch": 0.16029, + "grad_norm": 1.1821927698993437, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 16029 + }, + { + "epoch": 0.1603, + "grad_norm": 1.3831232505540298, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16030 + }, + { + "epoch": 0.16031, + "grad_norm": 1.0981332122382825, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 16031 + }, + { + "epoch": 0.16032, + "grad_norm": 1.184711305254577, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16032 + }, + { + "epoch": 0.16033, + "grad_norm": 1.1372011397446296, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 16033 + }, + { + "epoch": 0.16034, + "grad_norm": 1.5536688790305972, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16034 + }, + { + "epoch": 0.16035, + "grad_norm": 1.2204597945609714, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16035 + }, + { + "epoch": 0.16036, + "grad_norm": 1.2896458181398158, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 16036 + }, + { + "epoch": 0.16037, + "grad_norm": 1.169264447255346, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16037 + }, + { + "epoch": 0.16038, + "grad_norm": 1.2634712517724833, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 16038 + }, + { + "epoch": 0.16039, + "grad_norm": 1.0855605432845257, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16039 + }, + { + "epoch": 0.1604, + "grad_norm": 1.224483811509826, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16040 + }, + { + "epoch": 0.16041, + "grad_norm": 1.3331507875882669, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 16041 + }, + { + "epoch": 0.16042, + "grad_norm": 1.3891453967678526, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 16042 + }, + { + "epoch": 0.16043, + "grad_norm": 1.0795881114598527, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16043 + }, + { + "epoch": 0.16044, + "grad_norm": 1.3023423697862442, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 16044 + }, + { + "epoch": 0.16045, + "grad_norm": 1.3246120268063921, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 16045 + }, + { + "epoch": 0.16046, + "grad_norm": 1.1689040553881331, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 16046 + }, + { + "epoch": 0.16047, + "grad_norm": 1.3412929735739219, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16047 + }, + { + "epoch": 0.16048, + "grad_norm": 1.079196558883636, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 16048 + }, + { + "epoch": 0.16049, + "grad_norm": 1.203615237844193, + "learning_rate": 0.003, + "loss": 4.059, + "step": 16049 + }, + { + "epoch": 0.1605, + "grad_norm": 1.2286539119716313, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 16050 + }, + { + "epoch": 0.16051, + "grad_norm": 1.1185597063613093, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 16051 + }, + { + "epoch": 0.16052, + "grad_norm": 1.163864268386705, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16052 + }, + { + "epoch": 0.16053, + "grad_norm": 1.1613314754989137, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 16053 + }, + { + "epoch": 0.16054, + "grad_norm": 1.4948819073330097, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 16054 + }, + { + "epoch": 0.16055, + "grad_norm": 1.3371682943048482, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16055 + }, + { + "epoch": 0.16056, + "grad_norm": 1.3145085674969466, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16056 + }, + { + "epoch": 0.16057, + "grad_norm": 1.2269871285738352, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 16057 + }, + { + "epoch": 0.16058, + "grad_norm": 1.0244775848847234, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16058 + }, + { + "epoch": 0.16059, + "grad_norm": 1.4063519288689355, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 16059 + }, + { + "epoch": 0.1606, + "grad_norm": 0.9709192756974219, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16060 + }, + { + "epoch": 0.16061, + "grad_norm": 1.411425188042891, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 16061 + }, + { + "epoch": 0.16062, + "grad_norm": 1.1177862472664484, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16062 + }, + { + "epoch": 0.16063, + "grad_norm": 1.536173225451104, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16063 + }, + { + "epoch": 0.16064, + "grad_norm": 0.9707829048308666, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 16064 + }, + { + "epoch": 0.16065, + "grad_norm": 1.2900190071381814, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 16065 + }, + { + "epoch": 0.16066, + "grad_norm": 1.1189790366172518, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 16066 + }, + { + "epoch": 0.16067, + "grad_norm": 1.3640985307959768, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 16067 + }, + { + "epoch": 0.16068, + "grad_norm": 1.296619773757928, + "learning_rate": 0.003, + "loss": 4.054, + "step": 16068 + }, + { + "epoch": 0.16069, + "grad_norm": 1.2669442982556671, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 16069 + }, + { + "epoch": 0.1607, + "grad_norm": 1.4834545045440388, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 16070 + }, + { + "epoch": 0.16071, + "grad_norm": 1.02338157717909, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 16071 + }, + { + "epoch": 0.16072, + "grad_norm": 1.1439734656409877, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 16072 + }, + { + "epoch": 0.16073, + "grad_norm": 1.5239911455146988, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 16073 + }, + { + "epoch": 0.16074, + "grad_norm": 1.0346397079451986, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16074 + }, + { + "epoch": 0.16075, + "grad_norm": 1.497834128932681, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 16075 + }, + { + "epoch": 0.16076, + "grad_norm": 0.9777741720449256, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 16076 + }, + { + "epoch": 0.16077, + "grad_norm": 1.301822090815537, + "learning_rate": 0.003, + "loss": 4.058, + "step": 16077 + }, + { + "epoch": 0.16078, + "grad_norm": 1.1638833095483208, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 16078 + }, + { + "epoch": 0.16079, + "grad_norm": 1.421476786130954, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 16079 + }, + { + "epoch": 0.1608, + "grad_norm": 1.42572680340593, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 16080 + }, + { + "epoch": 0.16081, + "grad_norm": 1.6065065791865625, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 16081 + }, + { + "epoch": 0.16082, + "grad_norm": 1.0614834918617888, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 16082 + }, + { + "epoch": 0.16083, + "grad_norm": 1.0720843631531674, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 16083 + }, + { + "epoch": 0.16084, + "grad_norm": 1.2596062426915247, + "learning_rate": 0.003, + "loss": 4.091, + "step": 16084 + }, + { + "epoch": 0.16085, + "grad_norm": 1.3197603521255723, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16085 + }, + { + "epoch": 0.16086, + "grad_norm": 1.209075789456963, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16086 + }, + { + "epoch": 0.16087, + "grad_norm": 1.3986140934480322, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 16087 + }, + { + "epoch": 0.16088, + "grad_norm": 1.29174831939878, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16088 + }, + { + "epoch": 0.16089, + "grad_norm": 1.1118705050907147, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 16089 + }, + { + "epoch": 0.1609, + "grad_norm": 1.3356641655821764, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 16090 + }, + { + "epoch": 0.16091, + "grad_norm": 1.199188973900642, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 16091 + }, + { + "epoch": 0.16092, + "grad_norm": 1.793560222834461, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 16092 + }, + { + "epoch": 0.16093, + "grad_norm": 1.1309312644698293, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 16093 + }, + { + "epoch": 0.16094, + "grad_norm": 1.3730381439115702, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 16094 + }, + { + "epoch": 0.16095, + "grad_norm": 1.4378123039792088, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 16095 + }, + { + "epoch": 0.16096, + "grad_norm": 1.3389759471446794, + "learning_rate": 0.003, + "loss": 4.072, + "step": 16096 + }, + { + "epoch": 0.16097, + "grad_norm": 1.3158504027456166, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 16097 + }, + { + "epoch": 0.16098, + "grad_norm": 0.9877951756956128, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16098 + }, + { + "epoch": 0.16099, + "grad_norm": 1.4479428336464673, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16099 + }, + { + "epoch": 0.161, + "grad_norm": 1.0690018663453107, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16100 + }, + { + "epoch": 0.16101, + "grad_norm": 1.1087175494936976, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16101 + }, + { + "epoch": 0.16102, + "grad_norm": 1.2190483165806931, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16102 + }, + { + "epoch": 0.16103, + "grad_norm": 1.2464640512391514, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16103 + }, + { + "epoch": 0.16104, + "grad_norm": 1.2605449243517843, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 16104 + }, + { + "epoch": 0.16105, + "grad_norm": 1.3609265610053145, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 16105 + }, + { + "epoch": 0.16106, + "grad_norm": 1.1561289517191853, + "learning_rate": 0.003, + "loss": 4.06, + "step": 16106 + }, + { + "epoch": 0.16107, + "grad_norm": 1.2108900446658217, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 16107 + }, + { + "epoch": 0.16108, + "grad_norm": 1.3871226865725534, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 16108 + }, + { + "epoch": 0.16109, + "grad_norm": 1.2416300741278905, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16109 + }, + { + "epoch": 0.1611, + "grad_norm": 1.3549279352566184, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 16110 + }, + { + "epoch": 0.16111, + "grad_norm": 1.3554516222304485, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16111 + }, + { + "epoch": 0.16112, + "grad_norm": 1.1836975651978479, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16112 + }, + { + "epoch": 0.16113, + "grad_norm": 1.215756601682861, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16113 + }, + { + "epoch": 0.16114, + "grad_norm": 1.077720507183036, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 16114 + }, + { + "epoch": 0.16115, + "grad_norm": 1.2836790769497726, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16115 + }, + { + "epoch": 0.16116, + "grad_norm": 1.2067943231225828, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 16116 + }, + { + "epoch": 0.16117, + "grad_norm": 1.2529076106920993, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16117 + }, + { + "epoch": 0.16118, + "grad_norm": 1.261711886375417, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16118 + }, + { + "epoch": 0.16119, + "grad_norm": 1.1898514059835839, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 16119 + }, + { + "epoch": 0.1612, + "grad_norm": 1.042256094147418, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16120 + }, + { + "epoch": 0.16121, + "grad_norm": 1.2608642545166504, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 16121 + }, + { + "epoch": 0.16122, + "grad_norm": 1.2194499849594356, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 16122 + }, + { + "epoch": 0.16123, + "grad_norm": 1.3722797412364909, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 16123 + }, + { + "epoch": 0.16124, + "grad_norm": 1.1532481352638193, + "learning_rate": 0.003, + "loss": 4.056, + "step": 16124 + }, + { + "epoch": 0.16125, + "grad_norm": 1.3444693723979186, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 16125 + }, + { + "epoch": 0.16126, + "grad_norm": 1.3587021060251283, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 16126 + }, + { + "epoch": 0.16127, + "grad_norm": 1.4680093616394054, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16127 + }, + { + "epoch": 0.16128, + "grad_norm": 1.3708040186891726, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 16128 + }, + { + "epoch": 0.16129, + "grad_norm": 1.265944768899528, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16129 + }, + { + "epoch": 0.1613, + "grad_norm": 1.2147761587916155, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16130 + }, + { + "epoch": 0.16131, + "grad_norm": 1.2783421175922953, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 16131 + }, + { + "epoch": 0.16132, + "grad_norm": 1.2751345573542021, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16132 + }, + { + "epoch": 0.16133, + "grad_norm": 1.2163991871109832, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 16133 + }, + { + "epoch": 0.16134, + "grad_norm": 1.2495786526427402, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 16134 + }, + { + "epoch": 0.16135, + "grad_norm": 1.2552880181808581, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 16135 + }, + { + "epoch": 0.16136, + "grad_norm": 1.4879017034123712, + "learning_rate": 0.003, + "loss": 4.048, + "step": 16136 + }, + { + "epoch": 0.16137, + "grad_norm": 1.1368841189315837, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16137 + }, + { + "epoch": 0.16138, + "grad_norm": 1.4063956846527264, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 16138 + }, + { + "epoch": 0.16139, + "grad_norm": 1.2446175451255377, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16139 + }, + { + "epoch": 0.1614, + "grad_norm": 1.0876939463342667, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 16140 + }, + { + "epoch": 0.16141, + "grad_norm": 1.3324680737003314, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 16141 + }, + { + "epoch": 0.16142, + "grad_norm": 1.341488522037896, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 16142 + }, + { + "epoch": 0.16143, + "grad_norm": 1.0284822379365979, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16143 + }, + { + "epoch": 0.16144, + "grad_norm": 1.446545691671991, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 16144 + }, + { + "epoch": 0.16145, + "grad_norm": 1.198007631522013, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 16145 + }, + { + "epoch": 0.16146, + "grad_norm": 1.068927102975408, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 16146 + }, + { + "epoch": 0.16147, + "grad_norm": 1.5003008190770473, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 16147 + }, + { + "epoch": 0.16148, + "grad_norm": 1.0764359861587243, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 16148 + }, + { + "epoch": 0.16149, + "grad_norm": 1.2309420090166217, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16149 + }, + { + "epoch": 0.1615, + "grad_norm": 1.3078507897419687, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 16150 + }, + { + "epoch": 0.16151, + "grad_norm": 1.462687140006635, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16151 + }, + { + "epoch": 0.16152, + "grad_norm": 1.1113703241677952, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 16152 + }, + { + "epoch": 0.16153, + "grad_norm": 1.2475908893404852, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16153 + }, + { + "epoch": 0.16154, + "grad_norm": 1.6105637724327737, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 16154 + }, + { + "epoch": 0.16155, + "grad_norm": 1.0413256463913592, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 16155 + }, + { + "epoch": 0.16156, + "grad_norm": 1.3574353999341298, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 16156 + }, + { + "epoch": 0.16157, + "grad_norm": 1.2846752995761646, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16157 + }, + { + "epoch": 0.16158, + "grad_norm": 1.3225066680499187, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16158 + }, + { + "epoch": 0.16159, + "grad_norm": 1.3116346575986928, + "learning_rate": 0.003, + "loss": 4.093, + "step": 16159 + }, + { + "epoch": 0.1616, + "grad_norm": 1.0986634848005652, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 16160 + }, + { + "epoch": 0.16161, + "grad_norm": 1.489009004397367, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 16161 + }, + { + "epoch": 0.16162, + "grad_norm": 1.0671443540009917, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16162 + }, + { + "epoch": 0.16163, + "grad_norm": 1.3852087044343444, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 16163 + }, + { + "epoch": 0.16164, + "grad_norm": 1.305599729039529, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16164 + }, + { + "epoch": 0.16165, + "grad_norm": 1.151567243082139, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16165 + }, + { + "epoch": 0.16166, + "grad_norm": 1.501787296654248, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 16166 + }, + { + "epoch": 0.16167, + "grad_norm": 1.1041427692019146, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 16167 + }, + { + "epoch": 0.16168, + "grad_norm": 1.4258051081646463, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 16168 + }, + { + "epoch": 0.16169, + "grad_norm": 1.1599263469979852, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 16169 + }, + { + "epoch": 0.1617, + "grad_norm": 1.1963195979131709, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 16170 + }, + { + "epoch": 0.16171, + "grad_norm": 1.2205864585786506, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16171 + }, + { + "epoch": 0.16172, + "grad_norm": 1.0770953868565933, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16172 + }, + { + "epoch": 0.16173, + "grad_norm": 1.3909570250524284, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 16173 + }, + { + "epoch": 0.16174, + "grad_norm": 0.9878212489967368, + "learning_rate": 0.003, + "loss": 4.082, + "step": 16174 + }, + { + "epoch": 0.16175, + "grad_norm": 1.3457320433704199, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 16175 + }, + { + "epoch": 0.16176, + "grad_norm": 0.9850536798331347, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16176 + }, + { + "epoch": 0.16177, + "grad_norm": 1.4579496653506565, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 16177 + }, + { + "epoch": 0.16178, + "grad_norm": 1.3846670682448412, + "learning_rate": 0.003, + "loss": 4.082, + "step": 16178 + }, + { + "epoch": 0.16179, + "grad_norm": 1.559565431975786, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 16179 + }, + { + "epoch": 0.1618, + "grad_norm": 0.9908894518485826, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 16180 + }, + { + "epoch": 0.16181, + "grad_norm": 1.348064109065105, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 16181 + }, + { + "epoch": 0.16182, + "grad_norm": 1.399950817707883, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 16182 + }, + { + "epoch": 0.16183, + "grad_norm": 1.118259926151866, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16183 + }, + { + "epoch": 0.16184, + "grad_norm": 1.394829148863611, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 16184 + }, + { + "epoch": 0.16185, + "grad_norm": 1.1745608278679487, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 16185 + }, + { + "epoch": 0.16186, + "grad_norm": 1.251648889195004, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16186 + }, + { + "epoch": 0.16187, + "grad_norm": 1.2524277254239389, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 16187 + }, + { + "epoch": 0.16188, + "grad_norm": 1.2350728005269838, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 16188 + }, + { + "epoch": 0.16189, + "grad_norm": 1.291575283921151, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 16189 + }, + { + "epoch": 0.1619, + "grad_norm": 1.36391824636541, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 16190 + }, + { + "epoch": 0.16191, + "grad_norm": 1.5770002178564555, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 16191 + }, + { + "epoch": 0.16192, + "grad_norm": 1.2634525353872517, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16192 + }, + { + "epoch": 0.16193, + "grad_norm": 1.28488165744424, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 16193 + }, + { + "epoch": 0.16194, + "grad_norm": 1.3465642431600706, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 16194 + }, + { + "epoch": 0.16195, + "grad_norm": 1.2280001595506032, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 16195 + }, + { + "epoch": 0.16196, + "grad_norm": 1.3282442151996745, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 16196 + }, + { + "epoch": 0.16197, + "grad_norm": 0.9150787690474047, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16197 + }, + { + "epoch": 0.16198, + "grad_norm": 1.1928188777396278, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 16198 + }, + { + "epoch": 0.16199, + "grad_norm": 1.20056096196458, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16199 + }, + { + "epoch": 0.162, + "grad_norm": 1.2854712720823054, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 16200 + }, + { + "epoch": 0.16201, + "grad_norm": 1.0471238400168474, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 16201 + }, + { + "epoch": 0.16202, + "grad_norm": 1.2771481878611228, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 16202 + }, + { + "epoch": 0.16203, + "grad_norm": 1.2991642908445467, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 16203 + }, + { + "epoch": 0.16204, + "grad_norm": 1.391096078521732, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 16204 + }, + { + "epoch": 0.16205, + "grad_norm": 1.2138677594564928, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16205 + }, + { + "epoch": 0.16206, + "grad_norm": 1.5280425359595857, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16206 + }, + { + "epoch": 0.16207, + "grad_norm": 1.2393184803608204, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 16207 + }, + { + "epoch": 0.16208, + "grad_norm": 1.1685867078437393, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 16208 + }, + { + "epoch": 0.16209, + "grad_norm": 1.313631968426678, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 16209 + }, + { + "epoch": 0.1621, + "grad_norm": 1.183468025996727, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 16210 + }, + { + "epoch": 0.16211, + "grad_norm": 1.3508023550431656, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16211 + }, + { + "epoch": 0.16212, + "grad_norm": 1.2543446443891058, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16212 + }, + { + "epoch": 0.16213, + "grad_norm": 1.07806009180168, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 16213 + }, + { + "epoch": 0.16214, + "grad_norm": 1.2482642071439212, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 16214 + }, + { + "epoch": 0.16215, + "grad_norm": 1.12070374858544, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 16215 + }, + { + "epoch": 0.16216, + "grad_norm": 1.2417339257100313, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16216 + }, + { + "epoch": 0.16217, + "grad_norm": 1.021932077670497, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 16217 + }, + { + "epoch": 0.16218, + "grad_norm": 1.4476588358430686, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 16218 + }, + { + "epoch": 0.16219, + "grad_norm": 1.2801147532941688, + "learning_rate": 0.003, + "loss": 4.068, + "step": 16219 + }, + { + "epoch": 0.1622, + "grad_norm": 1.3666638771024235, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16220 + }, + { + "epoch": 0.16221, + "grad_norm": 1.6919655916293102, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 16221 + }, + { + "epoch": 0.16222, + "grad_norm": 1.133020693571414, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 16222 + }, + { + "epoch": 0.16223, + "grad_norm": 1.2765201950142349, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 16223 + }, + { + "epoch": 0.16224, + "grad_norm": 1.292954656723287, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16224 + }, + { + "epoch": 0.16225, + "grad_norm": 1.4715150749213328, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 16225 + }, + { + "epoch": 0.16226, + "grad_norm": 0.9149378601180623, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 16226 + }, + { + "epoch": 0.16227, + "grad_norm": 1.230883163804639, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 16227 + }, + { + "epoch": 0.16228, + "grad_norm": 1.3347318552909315, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 16228 + }, + { + "epoch": 0.16229, + "grad_norm": 1.0674427620211304, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 16229 + }, + { + "epoch": 0.1623, + "grad_norm": 1.338595938504359, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16230 + }, + { + "epoch": 0.16231, + "grad_norm": 1.101809798055724, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16231 + }, + { + "epoch": 0.16232, + "grad_norm": 1.2083469494542567, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16232 + }, + { + "epoch": 0.16233, + "grad_norm": 1.1788584236976805, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 16233 + }, + { + "epoch": 0.16234, + "grad_norm": 1.438190409187352, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 16234 + }, + { + "epoch": 0.16235, + "grad_norm": 1.3217102563848495, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 16235 + }, + { + "epoch": 0.16236, + "grad_norm": 1.1574464022417847, + "learning_rate": 0.003, + "loss": 4.058, + "step": 16236 + }, + { + "epoch": 0.16237, + "grad_norm": 1.6586220373587746, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 16237 + }, + { + "epoch": 0.16238, + "grad_norm": 1.1904629378808893, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16238 + }, + { + "epoch": 0.16239, + "grad_norm": 1.4029682136773813, + "learning_rate": 0.003, + "loss": 4.092, + "step": 16239 + }, + { + "epoch": 0.1624, + "grad_norm": 1.309501521771856, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 16240 + }, + { + "epoch": 0.16241, + "grad_norm": 1.1912634851965251, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 16241 + }, + { + "epoch": 0.16242, + "grad_norm": 1.1957762603851831, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 16242 + }, + { + "epoch": 0.16243, + "grad_norm": 0.9978203017900059, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 16243 + }, + { + "epoch": 0.16244, + "grad_norm": 1.3631017302455204, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 16244 + }, + { + "epoch": 0.16245, + "grad_norm": 1.1172482592038133, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 16245 + }, + { + "epoch": 0.16246, + "grad_norm": 1.291457136090636, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 16246 + }, + { + "epoch": 0.16247, + "grad_norm": 1.241648838090917, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16247 + }, + { + "epoch": 0.16248, + "grad_norm": 1.1912627487580116, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16248 + }, + { + "epoch": 0.16249, + "grad_norm": 1.4376949678364168, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 16249 + }, + { + "epoch": 0.1625, + "grad_norm": 1.332877486817461, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 16250 + }, + { + "epoch": 0.16251, + "grad_norm": 1.450341644372587, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16251 + }, + { + "epoch": 0.16252, + "grad_norm": 1.2216004688391267, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 16252 + }, + { + "epoch": 0.16253, + "grad_norm": 1.38458126587778, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16253 + }, + { + "epoch": 0.16254, + "grad_norm": 1.1578657329660573, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 16254 + }, + { + "epoch": 0.16255, + "grad_norm": 1.3584737020831286, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16255 + }, + { + "epoch": 0.16256, + "grad_norm": 1.0626796569996118, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 16256 + }, + { + "epoch": 0.16257, + "grad_norm": 1.1854307824297667, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 16257 + }, + { + "epoch": 0.16258, + "grad_norm": 1.0881261326227933, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 16258 + }, + { + "epoch": 0.16259, + "grad_norm": 1.1159855556877203, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 16259 + }, + { + "epoch": 0.1626, + "grad_norm": 1.1850051078218073, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16260 + }, + { + "epoch": 0.16261, + "grad_norm": 1.2954838458818088, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 16261 + }, + { + "epoch": 0.16262, + "grad_norm": 1.3844454619665678, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 16262 + }, + { + "epoch": 0.16263, + "grad_norm": 1.4326657049181495, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16263 + }, + { + "epoch": 0.16264, + "grad_norm": 1.0972597975116023, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 16264 + }, + { + "epoch": 0.16265, + "grad_norm": 1.456728688815861, + "learning_rate": 0.003, + "loss": 4.058, + "step": 16265 + }, + { + "epoch": 0.16266, + "grad_norm": 1.0893231721970618, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 16266 + }, + { + "epoch": 0.16267, + "grad_norm": 1.4265836466647381, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 16267 + }, + { + "epoch": 0.16268, + "grad_norm": 1.026908516804711, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 16268 + }, + { + "epoch": 0.16269, + "grad_norm": 1.2867505437326605, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 16269 + }, + { + "epoch": 0.1627, + "grad_norm": 1.3769738983162028, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 16270 + }, + { + "epoch": 0.16271, + "grad_norm": 1.1512571132733376, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 16271 + }, + { + "epoch": 0.16272, + "grad_norm": 1.2946312032229343, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 16272 + }, + { + "epoch": 0.16273, + "grad_norm": 1.160598306973601, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16273 + }, + { + "epoch": 0.16274, + "grad_norm": 1.190913322895234, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 16274 + }, + { + "epoch": 0.16275, + "grad_norm": 1.3565131412588454, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 16275 + }, + { + "epoch": 0.16276, + "grad_norm": 1.2228374178624037, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 16276 + }, + { + "epoch": 0.16277, + "grad_norm": 1.157740305349302, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 16277 + }, + { + "epoch": 0.16278, + "grad_norm": 1.1589390738190832, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16278 + }, + { + "epoch": 0.16279, + "grad_norm": 1.1427874625875416, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16279 + }, + { + "epoch": 0.1628, + "grad_norm": 1.4394281861045208, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 16280 + }, + { + "epoch": 0.16281, + "grad_norm": 1.2479386614109227, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16281 + }, + { + "epoch": 0.16282, + "grad_norm": 1.4435228993586355, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 16282 + }, + { + "epoch": 0.16283, + "grad_norm": 1.2367412237166255, + "learning_rate": 0.003, + "loss": 4.025, + "step": 16283 + }, + { + "epoch": 0.16284, + "grad_norm": 1.630935112607969, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 16284 + }, + { + "epoch": 0.16285, + "grad_norm": 1.2151600151267754, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16285 + }, + { + "epoch": 0.16286, + "grad_norm": 0.99844286924819, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 16286 + }, + { + "epoch": 0.16287, + "grad_norm": 1.3311858498500226, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 16287 + }, + { + "epoch": 0.16288, + "grad_norm": 1.1277454830696854, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 16288 + }, + { + "epoch": 0.16289, + "grad_norm": 1.3751420588987922, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 16289 + }, + { + "epoch": 0.1629, + "grad_norm": 1.2574333033294496, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 16290 + }, + { + "epoch": 0.16291, + "grad_norm": 1.2709011614362218, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16291 + }, + { + "epoch": 0.16292, + "grad_norm": 1.1731475105596485, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 16292 + }, + { + "epoch": 0.16293, + "grad_norm": 1.362351505374211, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16293 + }, + { + "epoch": 0.16294, + "grad_norm": 1.2153049142800547, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 16294 + }, + { + "epoch": 0.16295, + "grad_norm": 1.2855140261647056, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16295 + }, + { + "epoch": 0.16296, + "grad_norm": 1.2451835918537577, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 16296 + }, + { + "epoch": 0.16297, + "grad_norm": 1.0899963185254051, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16297 + }, + { + "epoch": 0.16298, + "grad_norm": 1.5368176411159515, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 16298 + }, + { + "epoch": 0.16299, + "grad_norm": 1.1951570026687528, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 16299 + }, + { + "epoch": 0.163, + "grad_norm": 1.677062653199514, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 16300 + }, + { + "epoch": 0.16301, + "grad_norm": 1.3448873086832585, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 16301 + }, + { + "epoch": 0.16302, + "grad_norm": 1.1602759708115664, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16302 + }, + { + "epoch": 0.16303, + "grad_norm": 1.3811138715383926, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 16303 + }, + { + "epoch": 0.16304, + "grad_norm": 1.1069545503836433, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16304 + }, + { + "epoch": 0.16305, + "grad_norm": 1.570100339629662, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 16305 + }, + { + "epoch": 0.16306, + "grad_norm": 1.232831755226583, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 16306 + }, + { + "epoch": 0.16307, + "grad_norm": 1.2644840112025777, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 16307 + }, + { + "epoch": 0.16308, + "grad_norm": 1.1619975932512598, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 16308 + }, + { + "epoch": 0.16309, + "grad_norm": 1.3731086711484763, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16309 + }, + { + "epoch": 0.1631, + "grad_norm": 1.4049245801410037, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 16310 + }, + { + "epoch": 0.16311, + "grad_norm": 1.064243387322423, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16311 + }, + { + "epoch": 0.16312, + "grad_norm": 1.3639565242761726, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 16312 + }, + { + "epoch": 0.16313, + "grad_norm": 1.1172881919093074, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 16313 + }, + { + "epoch": 0.16314, + "grad_norm": 1.165383167634807, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16314 + }, + { + "epoch": 0.16315, + "grad_norm": 1.205942669437338, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 16315 + }, + { + "epoch": 0.16316, + "grad_norm": 1.1958451581976022, + "learning_rate": 0.003, + "loss": 4.069, + "step": 16316 + }, + { + "epoch": 0.16317, + "grad_norm": 1.1744914363089543, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 16317 + }, + { + "epoch": 0.16318, + "grad_norm": 1.466582108406998, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 16318 + }, + { + "epoch": 0.16319, + "grad_norm": 1.0385254669559079, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16319 + }, + { + "epoch": 0.1632, + "grad_norm": 1.619029579687159, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 16320 + }, + { + "epoch": 0.16321, + "grad_norm": 1.1827478809755751, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 16321 + }, + { + "epoch": 0.16322, + "grad_norm": 1.4517056777140702, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 16322 + }, + { + "epoch": 0.16323, + "grad_norm": 1.3651946112406492, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 16323 + }, + { + "epoch": 0.16324, + "grad_norm": 1.661911846035659, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 16324 + }, + { + "epoch": 0.16325, + "grad_norm": 1.0679565552916765, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 16325 + }, + { + "epoch": 0.16326, + "grad_norm": 1.3346690818098346, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 16326 + }, + { + "epoch": 0.16327, + "grad_norm": 1.389890495562059, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 16327 + }, + { + "epoch": 0.16328, + "grad_norm": 1.0516321084279026, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 16328 + }, + { + "epoch": 0.16329, + "grad_norm": 1.3894814680474321, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 16329 + }, + { + "epoch": 0.1633, + "grad_norm": 1.1799369245058706, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16330 + }, + { + "epoch": 0.16331, + "grad_norm": 1.2172999397129118, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16331 + }, + { + "epoch": 0.16332, + "grad_norm": 1.4402792548750516, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16332 + }, + { + "epoch": 0.16333, + "grad_norm": 1.0962335816678261, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 16333 + }, + { + "epoch": 0.16334, + "grad_norm": 1.0535394881053008, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 16334 + }, + { + "epoch": 0.16335, + "grad_norm": 1.3723287192785087, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 16335 + }, + { + "epoch": 0.16336, + "grad_norm": 1.2712505974707535, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 16336 + }, + { + "epoch": 0.16337, + "grad_norm": 1.4053550627250662, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 16337 + }, + { + "epoch": 0.16338, + "grad_norm": 1.3102522704393156, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16338 + }, + { + "epoch": 0.16339, + "grad_norm": 1.0884281320779658, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 16339 + }, + { + "epoch": 0.1634, + "grad_norm": 1.369979709062412, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16340 + }, + { + "epoch": 0.16341, + "grad_norm": 1.2454029918205807, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 16341 + }, + { + "epoch": 0.16342, + "grad_norm": 1.1386512829299478, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16342 + }, + { + "epoch": 0.16343, + "grad_norm": 1.4316064097597532, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 16343 + }, + { + "epoch": 0.16344, + "grad_norm": 1.3025839922559512, + "learning_rate": 0.003, + "loss": 4.048, + "step": 16344 + }, + { + "epoch": 0.16345, + "grad_norm": 1.4062081379943796, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 16345 + }, + { + "epoch": 0.16346, + "grad_norm": 1.380426926743026, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16346 + }, + { + "epoch": 0.16347, + "grad_norm": 1.048439154887281, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 16347 + }, + { + "epoch": 0.16348, + "grad_norm": 1.4187306660020738, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 16348 + }, + { + "epoch": 0.16349, + "grad_norm": 1.133000049884522, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 16349 + }, + { + "epoch": 0.1635, + "grad_norm": 1.152077902344025, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16350 + }, + { + "epoch": 0.16351, + "grad_norm": 1.2543029760980908, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 16351 + }, + { + "epoch": 0.16352, + "grad_norm": 1.1991339006489674, + "learning_rate": 0.003, + "loss": 4.035, + "step": 16352 + }, + { + "epoch": 0.16353, + "grad_norm": 1.3279973071452171, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 16353 + }, + { + "epoch": 0.16354, + "grad_norm": 1.3209555143437766, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 16354 + }, + { + "epoch": 0.16355, + "grad_norm": 1.6069005203635782, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 16355 + }, + { + "epoch": 0.16356, + "grad_norm": 1.3874060586460157, + "learning_rate": 0.003, + "loss": 4.044, + "step": 16356 + }, + { + "epoch": 0.16357, + "grad_norm": 1.2573377261700582, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16357 + }, + { + "epoch": 0.16358, + "grad_norm": 1.5693088348261484, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 16358 + }, + { + "epoch": 0.16359, + "grad_norm": 1.2199974621945564, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 16359 + }, + { + "epoch": 0.1636, + "grad_norm": 1.198373588602752, + "learning_rate": 0.003, + "loss": 4.079, + "step": 16360 + }, + { + "epoch": 0.16361, + "grad_norm": 1.2415033730433311, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 16361 + }, + { + "epoch": 0.16362, + "grad_norm": 1.300340553941177, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 16362 + }, + { + "epoch": 0.16363, + "grad_norm": 1.3048339173505643, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16363 + }, + { + "epoch": 0.16364, + "grad_norm": 1.2023696939439263, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16364 + }, + { + "epoch": 0.16365, + "grad_norm": 1.4217090836305526, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 16365 + }, + { + "epoch": 0.16366, + "grad_norm": 1.1233430770093011, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16366 + }, + { + "epoch": 0.16367, + "grad_norm": 1.4427845967168846, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 16367 + }, + { + "epoch": 0.16368, + "grad_norm": 1.1155905332168412, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 16368 + }, + { + "epoch": 0.16369, + "grad_norm": 1.3464489557382935, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16369 + }, + { + "epoch": 0.1637, + "grad_norm": 1.0578550159183047, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 16370 + }, + { + "epoch": 0.16371, + "grad_norm": 1.511918444515664, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 16371 + }, + { + "epoch": 0.16372, + "grad_norm": 1.129299153317865, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 16372 + }, + { + "epoch": 0.16373, + "grad_norm": 1.587989632774138, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 16373 + }, + { + "epoch": 0.16374, + "grad_norm": 1.1700883074011486, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 16374 + }, + { + "epoch": 0.16375, + "grad_norm": 1.3040537631563571, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16375 + }, + { + "epoch": 0.16376, + "grad_norm": 1.0856008882922459, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 16376 + }, + { + "epoch": 0.16377, + "grad_norm": 1.3168505447124415, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16377 + }, + { + "epoch": 0.16378, + "grad_norm": 1.409911310129991, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16378 + }, + { + "epoch": 0.16379, + "grad_norm": 1.1578313905653308, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16379 + }, + { + "epoch": 0.1638, + "grad_norm": 1.2243553009162318, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16380 + }, + { + "epoch": 0.16381, + "grad_norm": 1.3211151537680381, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 16381 + }, + { + "epoch": 0.16382, + "grad_norm": 1.0003765638365243, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 16382 + }, + { + "epoch": 0.16383, + "grad_norm": 1.4431212791886239, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 16383 + }, + { + "epoch": 0.16384, + "grad_norm": 1.1418299209144636, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16384 + }, + { + "epoch": 0.16385, + "grad_norm": 1.2792884718943385, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 16385 + }, + { + "epoch": 0.16386, + "grad_norm": 1.1404948764661138, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 16386 + }, + { + "epoch": 0.16387, + "grad_norm": 1.284319680515499, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16387 + }, + { + "epoch": 0.16388, + "grad_norm": 1.4158203236465858, + "learning_rate": 0.003, + "loss": 4.071, + "step": 16388 + }, + { + "epoch": 0.16389, + "grad_norm": 1.4336931757387177, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16389 + }, + { + "epoch": 0.1639, + "grad_norm": 1.2536201072137658, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 16390 + }, + { + "epoch": 0.16391, + "grad_norm": 1.2836875225183824, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 16391 + }, + { + "epoch": 0.16392, + "grad_norm": 1.2176262982047852, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 16392 + }, + { + "epoch": 0.16393, + "grad_norm": 1.4483108282285946, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 16393 + }, + { + "epoch": 0.16394, + "grad_norm": 1.246150339402114, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 16394 + }, + { + "epoch": 0.16395, + "grad_norm": 1.381316017329402, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16395 + }, + { + "epoch": 0.16396, + "grad_norm": 1.3577173637161588, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 16396 + }, + { + "epoch": 0.16397, + "grad_norm": 1.1911270331753483, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 16397 + }, + { + "epoch": 0.16398, + "grad_norm": 1.420612509964672, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 16398 + }, + { + "epoch": 0.16399, + "grad_norm": 1.1716767922518303, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16399 + }, + { + "epoch": 0.164, + "grad_norm": 1.380990633593178, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 16400 + }, + { + "epoch": 0.16401, + "grad_norm": 1.2059872705480625, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 16401 + }, + { + "epoch": 0.16402, + "grad_norm": 1.2082775563370716, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 16402 + }, + { + "epoch": 0.16403, + "grad_norm": 1.3569918868573316, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 16403 + }, + { + "epoch": 0.16404, + "grad_norm": 1.2132090912442626, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 16404 + }, + { + "epoch": 0.16405, + "grad_norm": 1.3219062034164897, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 16405 + }, + { + "epoch": 0.16406, + "grad_norm": 1.2571787124909675, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16406 + }, + { + "epoch": 0.16407, + "grad_norm": 1.090618551506124, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16407 + }, + { + "epoch": 0.16408, + "grad_norm": 1.4297594624104537, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16408 + }, + { + "epoch": 0.16409, + "grad_norm": 1.1132480973468442, + "learning_rate": 0.003, + "loss": 4.035, + "step": 16409 + }, + { + "epoch": 0.1641, + "grad_norm": 1.5933083288070427, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 16410 + }, + { + "epoch": 0.16411, + "grad_norm": 1.1376891610757405, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 16411 + }, + { + "epoch": 0.16412, + "grad_norm": 1.114888268375818, + "learning_rate": 0.003, + "loss": 4.059, + "step": 16412 + }, + { + "epoch": 0.16413, + "grad_norm": 1.2345716566950218, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16413 + }, + { + "epoch": 0.16414, + "grad_norm": 1.246216303691841, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 16414 + }, + { + "epoch": 0.16415, + "grad_norm": 1.1663422313702552, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16415 + }, + { + "epoch": 0.16416, + "grad_norm": 1.3673709590515428, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16416 + }, + { + "epoch": 0.16417, + "grad_norm": 1.4693437831528886, + "learning_rate": 0.003, + "loss": 4.061, + "step": 16417 + }, + { + "epoch": 0.16418, + "grad_norm": 1.1785751760532168, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 16418 + }, + { + "epoch": 0.16419, + "grad_norm": 1.2877570452345621, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 16419 + }, + { + "epoch": 0.1642, + "grad_norm": 1.3446446582739822, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 16420 + }, + { + "epoch": 0.16421, + "grad_norm": 1.0676062088698817, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 16421 + }, + { + "epoch": 0.16422, + "grad_norm": 1.1319626846229354, + "learning_rate": 0.003, + "loss": 4.03, + "step": 16422 + }, + { + "epoch": 0.16423, + "grad_norm": 1.5613091015565685, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 16423 + }, + { + "epoch": 0.16424, + "grad_norm": 1.4368794722217562, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16424 + }, + { + "epoch": 0.16425, + "grad_norm": 1.139300263428175, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16425 + }, + { + "epoch": 0.16426, + "grad_norm": 1.2165822346927826, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 16426 + }, + { + "epoch": 0.16427, + "grad_norm": 1.32474169950318, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 16427 + }, + { + "epoch": 0.16428, + "grad_norm": 1.2124515065784107, + "learning_rate": 0.003, + "loss": 4.068, + "step": 16428 + }, + { + "epoch": 0.16429, + "grad_norm": 1.2099272814720674, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 16429 + }, + { + "epoch": 0.1643, + "grad_norm": 1.3165838134959365, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16430 + }, + { + "epoch": 0.16431, + "grad_norm": 1.3124340867269597, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16431 + }, + { + "epoch": 0.16432, + "grad_norm": 1.3083488161900734, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16432 + }, + { + "epoch": 0.16433, + "grad_norm": 1.2212582053693837, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 16433 + }, + { + "epoch": 0.16434, + "grad_norm": 1.248712178980779, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 16434 + }, + { + "epoch": 0.16435, + "grad_norm": 1.10247019771398, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16435 + }, + { + "epoch": 0.16436, + "grad_norm": 1.3138991041673305, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 16436 + }, + { + "epoch": 0.16437, + "grad_norm": 1.1426017630552887, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16437 + }, + { + "epoch": 0.16438, + "grad_norm": 1.3516739937256734, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 16438 + }, + { + "epoch": 0.16439, + "grad_norm": 1.1325632245553459, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 16439 + }, + { + "epoch": 0.1644, + "grad_norm": 1.5530152726734836, + "learning_rate": 0.003, + "loss": 4.061, + "step": 16440 + }, + { + "epoch": 0.16441, + "grad_norm": 1.0258403136890335, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 16441 + }, + { + "epoch": 0.16442, + "grad_norm": 1.6038201349055385, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 16442 + }, + { + "epoch": 0.16443, + "grad_norm": 1.1902209962381898, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 16443 + }, + { + "epoch": 0.16444, + "grad_norm": 1.2619373346371396, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 16444 + }, + { + "epoch": 0.16445, + "grad_norm": 1.0269781849552861, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16445 + }, + { + "epoch": 0.16446, + "grad_norm": 1.3911572338821512, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16446 + }, + { + "epoch": 0.16447, + "grad_norm": 1.0130236852413237, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 16447 + }, + { + "epoch": 0.16448, + "grad_norm": 1.3752196679624056, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 16448 + }, + { + "epoch": 0.16449, + "grad_norm": 1.2228405074153301, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 16449 + }, + { + "epoch": 0.1645, + "grad_norm": 1.5814816887535803, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 16450 + }, + { + "epoch": 0.16451, + "grad_norm": 1.315260931427507, + "learning_rate": 0.003, + "loss": 4.08, + "step": 16451 + }, + { + "epoch": 0.16452, + "grad_norm": 1.4668306986859874, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 16452 + }, + { + "epoch": 0.16453, + "grad_norm": 1.4764637017528175, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 16453 + }, + { + "epoch": 0.16454, + "grad_norm": 1.1302037993586078, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 16454 + }, + { + "epoch": 0.16455, + "grad_norm": 1.2718503840267625, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 16455 + }, + { + "epoch": 0.16456, + "grad_norm": 1.1737062799099678, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16456 + }, + { + "epoch": 0.16457, + "grad_norm": 1.1129495146380497, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 16457 + }, + { + "epoch": 0.16458, + "grad_norm": 1.0785969337090933, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 16458 + }, + { + "epoch": 0.16459, + "grad_norm": 1.271296796031856, + "learning_rate": 0.003, + "loss": 4.074, + "step": 16459 + }, + { + "epoch": 0.1646, + "grad_norm": 1.0523375357333073, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 16460 + }, + { + "epoch": 0.16461, + "grad_norm": 1.368714109205373, + "learning_rate": 0.003, + "loss": 4.061, + "step": 16461 + }, + { + "epoch": 0.16462, + "grad_norm": 1.0702950855721094, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 16462 + }, + { + "epoch": 0.16463, + "grad_norm": 1.57216920795091, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 16463 + }, + { + "epoch": 0.16464, + "grad_norm": 1.137099088097417, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 16464 + }, + { + "epoch": 0.16465, + "grad_norm": 1.1996377255304156, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 16465 + }, + { + "epoch": 0.16466, + "grad_norm": 1.6724182629270892, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 16466 + }, + { + "epoch": 0.16467, + "grad_norm": 1.5553433654665862, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 16467 + }, + { + "epoch": 0.16468, + "grad_norm": 1.1064706461046685, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 16468 + }, + { + "epoch": 0.16469, + "grad_norm": 1.3002040341038863, + "learning_rate": 0.003, + "loss": 4.09, + "step": 16469 + }, + { + "epoch": 0.1647, + "grad_norm": 1.2451762501791963, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 16470 + }, + { + "epoch": 0.16471, + "grad_norm": 1.4559204829357344, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 16471 + }, + { + "epoch": 0.16472, + "grad_norm": 1.2892392909664752, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 16472 + }, + { + "epoch": 0.16473, + "grad_norm": 1.4956841084217132, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 16473 + }, + { + "epoch": 0.16474, + "grad_norm": 1.0187601518228537, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 16474 + }, + { + "epoch": 0.16475, + "grad_norm": 1.2428533920598925, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 16475 + }, + { + "epoch": 0.16476, + "grad_norm": 1.1668821201355952, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 16476 + }, + { + "epoch": 0.16477, + "grad_norm": 1.1982463835413693, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 16477 + }, + { + "epoch": 0.16478, + "grad_norm": 1.2485534588013527, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 16478 + }, + { + "epoch": 0.16479, + "grad_norm": 1.3181502788303614, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16479 + }, + { + "epoch": 0.1648, + "grad_norm": 1.1974651532537637, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 16480 + }, + { + "epoch": 0.16481, + "grad_norm": 1.5443286790141675, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 16481 + }, + { + "epoch": 0.16482, + "grad_norm": 1.0580173964338795, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 16482 + }, + { + "epoch": 0.16483, + "grad_norm": 1.4011464567334797, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 16483 + }, + { + "epoch": 0.16484, + "grad_norm": 1.2466234410214732, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16484 + }, + { + "epoch": 0.16485, + "grad_norm": 1.4197020019380282, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 16485 + }, + { + "epoch": 0.16486, + "grad_norm": 1.3467791156955868, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 16486 + }, + { + "epoch": 0.16487, + "grad_norm": 1.1286023331651984, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 16487 + }, + { + "epoch": 0.16488, + "grad_norm": 1.3406858944819247, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 16488 + }, + { + "epoch": 0.16489, + "grad_norm": 1.3346421533134032, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 16489 + }, + { + "epoch": 0.1649, + "grad_norm": 1.299839265832347, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 16490 + }, + { + "epoch": 0.16491, + "grad_norm": 1.3447492313470124, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 16491 + }, + { + "epoch": 0.16492, + "grad_norm": 1.480927716158016, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 16492 + }, + { + "epoch": 0.16493, + "grad_norm": 1.15363763860873, + "learning_rate": 0.003, + "loss": 4.075, + "step": 16493 + }, + { + "epoch": 0.16494, + "grad_norm": 1.088843248716626, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 16494 + }, + { + "epoch": 0.16495, + "grad_norm": 1.3271018110522426, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 16495 + }, + { + "epoch": 0.16496, + "grad_norm": 1.3579986770082573, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16496 + }, + { + "epoch": 0.16497, + "grad_norm": 1.246765843625915, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 16497 + }, + { + "epoch": 0.16498, + "grad_norm": 1.3336299537332825, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16498 + }, + { + "epoch": 0.16499, + "grad_norm": 1.1407031711918372, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16499 + }, + { + "epoch": 0.165, + "grad_norm": 1.3208808146452224, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 16500 + }, + { + "epoch": 0.16501, + "grad_norm": 1.4543504750256484, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 16501 + }, + { + "epoch": 0.16502, + "grad_norm": 1.3363482030749658, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 16502 + }, + { + "epoch": 0.16503, + "grad_norm": 1.329011551322635, + "learning_rate": 0.003, + "loss": 4.056, + "step": 16503 + }, + { + "epoch": 0.16504, + "grad_norm": 1.3054054631462377, + "learning_rate": 0.003, + "loss": 4.082, + "step": 16504 + }, + { + "epoch": 0.16505, + "grad_norm": 1.1677461912364604, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 16505 + }, + { + "epoch": 0.16506, + "grad_norm": 1.1645870622599563, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 16506 + }, + { + "epoch": 0.16507, + "grad_norm": 1.2277219991165256, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 16507 + }, + { + "epoch": 0.16508, + "grad_norm": 1.1194755996128067, + "learning_rate": 0.003, + "loss": 4.051, + "step": 16508 + }, + { + "epoch": 0.16509, + "grad_norm": 1.5244986955115396, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 16509 + }, + { + "epoch": 0.1651, + "grad_norm": 0.8323002691807789, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 16510 + }, + { + "epoch": 0.16511, + "grad_norm": 1.1334584499206646, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 16511 + }, + { + "epoch": 0.16512, + "grad_norm": 1.2896129322877243, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 16512 + }, + { + "epoch": 0.16513, + "grad_norm": 1.0856970973611928, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16513 + }, + { + "epoch": 0.16514, + "grad_norm": 1.1817625275124146, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 16514 + }, + { + "epoch": 0.16515, + "grad_norm": 1.3875030521708789, + "learning_rate": 0.003, + "loss": 4.07, + "step": 16515 + }, + { + "epoch": 0.16516, + "grad_norm": 1.5322349487881628, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16516 + }, + { + "epoch": 0.16517, + "grad_norm": 1.214093787219609, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16517 + }, + { + "epoch": 0.16518, + "grad_norm": 1.3264218281675586, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 16518 + }, + { + "epoch": 0.16519, + "grad_norm": 1.3037814719409515, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 16519 + }, + { + "epoch": 0.1652, + "grad_norm": 1.347256831948668, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 16520 + }, + { + "epoch": 0.16521, + "grad_norm": 1.135833220945594, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 16521 + }, + { + "epoch": 0.16522, + "grad_norm": 1.1635330169048987, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 16522 + }, + { + "epoch": 0.16523, + "grad_norm": 1.0707791474315451, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16523 + }, + { + "epoch": 0.16524, + "grad_norm": 1.25384814518774, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 16524 + }, + { + "epoch": 0.16525, + "grad_norm": 1.2132133705864718, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16525 + }, + { + "epoch": 0.16526, + "grad_norm": 1.7649845481238309, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 16526 + }, + { + "epoch": 0.16527, + "grad_norm": 1.1197744287796618, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 16527 + }, + { + "epoch": 0.16528, + "grad_norm": 1.5118003809505407, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16528 + }, + { + "epoch": 0.16529, + "grad_norm": 1.4730620724374555, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 16529 + }, + { + "epoch": 0.1653, + "grad_norm": 0.90921825476028, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 16530 + }, + { + "epoch": 0.16531, + "grad_norm": 1.2822372958361714, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16531 + }, + { + "epoch": 0.16532, + "grad_norm": 1.2598019825381823, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 16532 + }, + { + "epoch": 0.16533, + "grad_norm": 1.3640189473188689, + "learning_rate": 0.003, + "loss": 4.067, + "step": 16533 + }, + { + "epoch": 0.16534, + "grad_norm": 1.250697213719519, + "learning_rate": 0.003, + "loss": 4.081, + "step": 16534 + }, + { + "epoch": 0.16535, + "grad_norm": 1.1863462394733493, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16535 + }, + { + "epoch": 0.16536, + "grad_norm": 1.193772061769314, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 16536 + }, + { + "epoch": 0.16537, + "grad_norm": 1.3875022671160369, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 16537 + }, + { + "epoch": 0.16538, + "grad_norm": 1.1032716696200444, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16538 + }, + { + "epoch": 0.16539, + "grad_norm": 1.2085962067759086, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 16539 + }, + { + "epoch": 0.1654, + "grad_norm": 1.4437133351830067, + "learning_rate": 0.003, + "loss": 4.064, + "step": 16540 + }, + { + "epoch": 0.16541, + "grad_norm": 1.1113670923330192, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16541 + }, + { + "epoch": 0.16542, + "grad_norm": 1.2923329685370313, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16542 + }, + { + "epoch": 0.16543, + "grad_norm": 1.2265620690901557, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 16543 + }, + { + "epoch": 0.16544, + "grad_norm": 1.2889039235514912, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 16544 + }, + { + "epoch": 0.16545, + "grad_norm": 1.0361539017314416, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16545 + }, + { + "epoch": 0.16546, + "grad_norm": 1.5710515904146527, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16546 + }, + { + "epoch": 0.16547, + "grad_norm": 0.9675541228764893, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 16547 + }, + { + "epoch": 0.16548, + "grad_norm": 1.3690807668603275, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16548 + }, + { + "epoch": 0.16549, + "grad_norm": 1.2238178932528583, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 16549 + }, + { + "epoch": 0.1655, + "grad_norm": 1.3624593844958657, + "learning_rate": 0.003, + "loss": 4.057, + "step": 16550 + }, + { + "epoch": 0.16551, + "grad_norm": 1.1864513195867061, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16551 + }, + { + "epoch": 0.16552, + "grad_norm": 1.516787527114139, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 16552 + }, + { + "epoch": 0.16553, + "grad_norm": 0.9501015686654448, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 16553 + }, + { + "epoch": 0.16554, + "grad_norm": 1.3711422680821514, + "learning_rate": 0.003, + "loss": 4.062, + "step": 16554 + }, + { + "epoch": 0.16555, + "grad_norm": 1.3117752994242997, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 16555 + }, + { + "epoch": 0.16556, + "grad_norm": 1.6255604613126764, + "learning_rate": 0.003, + "loss": 4.066, + "step": 16556 + }, + { + "epoch": 0.16557, + "grad_norm": 1.0632114951015792, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 16557 + }, + { + "epoch": 0.16558, + "grad_norm": 1.59166839672321, + "learning_rate": 0.003, + "loss": 4.073, + "step": 16558 + }, + { + "epoch": 0.16559, + "grad_norm": 1.0469939320043071, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16559 + }, + { + "epoch": 0.1656, + "grad_norm": 1.4773552474017304, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 16560 + }, + { + "epoch": 0.16561, + "grad_norm": 1.130145898065509, + "learning_rate": 0.003, + "loss": 4.095, + "step": 16561 + }, + { + "epoch": 0.16562, + "grad_norm": 1.205195347039316, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 16562 + }, + { + "epoch": 0.16563, + "grad_norm": 1.276745680601048, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16563 + }, + { + "epoch": 0.16564, + "grad_norm": 1.340516278687605, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 16564 + }, + { + "epoch": 0.16565, + "grad_norm": 1.2179714282624683, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 16565 + }, + { + "epoch": 0.16566, + "grad_norm": 1.5093543786751196, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 16566 + }, + { + "epoch": 0.16567, + "grad_norm": 1.4059023012697849, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16567 + }, + { + "epoch": 0.16568, + "grad_norm": 1.1544566308338915, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16568 + }, + { + "epoch": 0.16569, + "grad_norm": 1.4033121208928, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 16569 + }, + { + "epoch": 0.1657, + "grad_norm": 1.2238377475927165, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 16570 + }, + { + "epoch": 0.16571, + "grad_norm": 1.219524154636383, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 16571 + }, + { + "epoch": 0.16572, + "grad_norm": 1.308759384697221, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16572 + }, + { + "epoch": 0.16573, + "grad_norm": 1.0418892214396567, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 16573 + }, + { + "epoch": 0.16574, + "grad_norm": 1.2344157645941867, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 16574 + }, + { + "epoch": 0.16575, + "grad_norm": 1.186935435165905, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 16575 + }, + { + "epoch": 0.16576, + "grad_norm": 1.3678784952055867, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 16576 + }, + { + "epoch": 0.16577, + "grad_norm": 1.0878995328997643, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 16577 + }, + { + "epoch": 0.16578, + "grad_norm": 1.3512824724156975, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 16578 + }, + { + "epoch": 0.16579, + "grad_norm": 1.1906761604147342, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16579 + }, + { + "epoch": 0.1658, + "grad_norm": 1.1456813687747813, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16580 + }, + { + "epoch": 0.16581, + "grad_norm": 1.2775782086054193, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 16581 + }, + { + "epoch": 0.16582, + "grad_norm": 1.6744773766700716, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 16582 + }, + { + "epoch": 0.16583, + "grad_norm": 1.2677766208582708, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16583 + }, + { + "epoch": 0.16584, + "grad_norm": 1.4270773934465257, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 16584 + }, + { + "epoch": 0.16585, + "grad_norm": 1.3975056354370472, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16585 + }, + { + "epoch": 0.16586, + "grad_norm": 1.2458642694686481, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16586 + }, + { + "epoch": 0.16587, + "grad_norm": 1.2736822759593291, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 16587 + }, + { + "epoch": 0.16588, + "grad_norm": 1.236275197540774, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 16588 + }, + { + "epoch": 0.16589, + "grad_norm": 1.1066772994727698, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 16589 + }, + { + "epoch": 0.1659, + "grad_norm": 1.1232249599314457, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 16590 + }, + { + "epoch": 0.16591, + "grad_norm": 1.4093779362858834, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 16591 + }, + { + "epoch": 0.16592, + "grad_norm": 1.429195971694986, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 16592 + }, + { + "epoch": 0.16593, + "grad_norm": 1.1240960247071905, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 16593 + }, + { + "epoch": 0.16594, + "grad_norm": 1.2425919778673997, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 16594 + }, + { + "epoch": 0.16595, + "grad_norm": 1.397375625252414, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 16595 + }, + { + "epoch": 0.16596, + "grad_norm": 1.2009407512656327, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16596 + }, + { + "epoch": 0.16597, + "grad_norm": 1.276527633707103, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 16597 + }, + { + "epoch": 0.16598, + "grad_norm": 1.533636505639626, + "learning_rate": 0.003, + "loss": 4.032, + "step": 16598 + }, + { + "epoch": 0.16599, + "grad_norm": 0.9518186678655945, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 16599 + }, + { + "epoch": 0.166, + "grad_norm": 1.2160159835882107, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 16600 + }, + { + "epoch": 0.16601, + "grad_norm": 1.2152796106285066, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 16601 + }, + { + "epoch": 0.16602, + "grad_norm": 1.2746660739876183, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 16602 + }, + { + "epoch": 0.16603, + "grad_norm": 1.4252702373047241, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 16603 + }, + { + "epoch": 0.16604, + "grad_norm": 1.0635959392211614, + "learning_rate": 0.003, + "loss": 4.053, + "step": 16604 + }, + { + "epoch": 0.16605, + "grad_norm": 1.552258145640101, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16605 + }, + { + "epoch": 0.16606, + "grad_norm": 1.204222339131148, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 16606 + }, + { + "epoch": 0.16607, + "grad_norm": 1.5978386275991108, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 16607 + }, + { + "epoch": 0.16608, + "grad_norm": 1.1061872445758147, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16608 + }, + { + "epoch": 0.16609, + "grad_norm": 1.2701346668331372, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 16609 + }, + { + "epoch": 0.1661, + "grad_norm": 1.3289461078989357, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16610 + }, + { + "epoch": 0.16611, + "grad_norm": 1.1168024268611252, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16611 + }, + { + "epoch": 0.16612, + "grad_norm": 1.3048247232069652, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16612 + }, + { + "epoch": 0.16613, + "grad_norm": 1.018674315159564, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 16613 + }, + { + "epoch": 0.16614, + "grad_norm": 1.2739071182800958, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 16614 + }, + { + "epoch": 0.16615, + "grad_norm": 1.448527645704851, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 16615 + }, + { + "epoch": 0.16616, + "grad_norm": 1.2618724216619026, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 16616 + }, + { + "epoch": 0.16617, + "grad_norm": 1.4716514933801228, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 16617 + }, + { + "epoch": 0.16618, + "grad_norm": 1.000641246318044, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16618 + }, + { + "epoch": 0.16619, + "grad_norm": 1.5442007930216206, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16619 + }, + { + "epoch": 0.1662, + "grad_norm": 1.0895470650253285, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 16620 + }, + { + "epoch": 0.16621, + "grad_norm": 1.4581178360506257, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 16621 + }, + { + "epoch": 0.16622, + "grad_norm": 1.2356933981899747, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 16622 + }, + { + "epoch": 0.16623, + "grad_norm": 1.176423111853083, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 16623 + }, + { + "epoch": 0.16624, + "grad_norm": 1.3403710983985258, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 16624 + }, + { + "epoch": 0.16625, + "grad_norm": 1.167415870709617, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 16625 + }, + { + "epoch": 0.16626, + "grad_norm": 1.3870341030724083, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 16626 + }, + { + "epoch": 0.16627, + "grad_norm": 1.207894431592241, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16627 + }, + { + "epoch": 0.16628, + "grad_norm": 1.4728929292785442, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 16628 + }, + { + "epoch": 0.16629, + "grad_norm": 1.1112205171923735, + "learning_rate": 0.003, + "loss": 4.101, + "step": 16629 + }, + { + "epoch": 0.1663, + "grad_norm": 1.1590305896874513, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 16630 + }, + { + "epoch": 0.16631, + "grad_norm": 1.482380357341026, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 16631 + }, + { + "epoch": 0.16632, + "grad_norm": 1.0773970821438095, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 16632 + }, + { + "epoch": 0.16633, + "grad_norm": 1.3817570315473837, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 16633 + }, + { + "epoch": 0.16634, + "grad_norm": 1.1575282789470405, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 16634 + }, + { + "epoch": 0.16635, + "grad_norm": 1.415087075334309, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 16635 + }, + { + "epoch": 0.16636, + "grad_norm": 0.9255282579734383, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16636 + }, + { + "epoch": 0.16637, + "grad_norm": 1.2282673114747942, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 16637 + }, + { + "epoch": 0.16638, + "grad_norm": 1.6337941574914259, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 16638 + }, + { + "epoch": 0.16639, + "grad_norm": 1.1293362893753414, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 16639 + }, + { + "epoch": 0.1664, + "grad_norm": 1.2601235958197192, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 16640 + }, + { + "epoch": 0.16641, + "grad_norm": 1.2557510037069466, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 16641 + }, + { + "epoch": 0.16642, + "grad_norm": 1.1529014840095058, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16642 + }, + { + "epoch": 0.16643, + "grad_norm": 1.4280815118907935, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 16643 + }, + { + "epoch": 0.16644, + "grad_norm": 1.1052315863725835, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16644 + }, + { + "epoch": 0.16645, + "grad_norm": 1.2503694285321836, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 16645 + }, + { + "epoch": 0.16646, + "grad_norm": 1.2495900248899456, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 16646 + }, + { + "epoch": 0.16647, + "grad_norm": 1.375768919260447, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 16647 + }, + { + "epoch": 0.16648, + "grad_norm": 1.2305083673039703, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 16648 + }, + { + "epoch": 0.16649, + "grad_norm": 1.564214022867252, + "learning_rate": 0.003, + "loss": 4.102, + "step": 16649 + }, + { + "epoch": 0.1665, + "grad_norm": 1.2337311610277657, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 16650 + }, + { + "epoch": 0.16651, + "grad_norm": 1.3018102841398165, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16651 + }, + { + "epoch": 0.16652, + "grad_norm": 1.4591675818303105, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 16652 + }, + { + "epoch": 0.16653, + "grad_norm": 1.3195737330554282, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16653 + }, + { + "epoch": 0.16654, + "grad_norm": 1.3569877794562422, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 16654 + }, + { + "epoch": 0.16655, + "grad_norm": 1.1633207592710884, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16655 + }, + { + "epoch": 0.16656, + "grad_norm": 1.2950788549807473, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 16656 + }, + { + "epoch": 0.16657, + "grad_norm": 1.2092542748592239, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 16657 + }, + { + "epoch": 0.16658, + "grad_norm": 1.0939329273704779, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 16658 + }, + { + "epoch": 0.16659, + "grad_norm": 1.123124855614399, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 16659 + }, + { + "epoch": 0.1666, + "grad_norm": 1.25646383028317, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 16660 + }, + { + "epoch": 0.16661, + "grad_norm": 1.2927711530505899, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 16661 + }, + { + "epoch": 0.16662, + "grad_norm": 1.5065359649910308, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16662 + }, + { + "epoch": 0.16663, + "grad_norm": 1.2217810993901903, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 16663 + }, + { + "epoch": 0.16664, + "grad_norm": 1.2506747027717897, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16664 + }, + { + "epoch": 0.16665, + "grad_norm": 1.3679567019358465, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 16665 + }, + { + "epoch": 0.16666, + "grad_norm": 1.0847656585217225, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 16666 + }, + { + "epoch": 0.16667, + "grad_norm": 1.2440115678312678, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 16667 + }, + { + "epoch": 0.16668, + "grad_norm": 1.2986418728887241, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16668 + }, + { + "epoch": 0.16669, + "grad_norm": 1.3937775860243589, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 16669 + }, + { + "epoch": 0.1667, + "grad_norm": 1.1444490580546132, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 16670 + }, + { + "epoch": 0.16671, + "grad_norm": 1.2547154137614827, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 16671 + }, + { + "epoch": 0.16672, + "grad_norm": 1.4686589795806708, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16672 + }, + { + "epoch": 0.16673, + "grad_norm": 1.2966678797714233, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 16673 + }, + { + "epoch": 0.16674, + "grad_norm": 1.4008422920220023, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 16674 + }, + { + "epoch": 0.16675, + "grad_norm": 1.4374214591850536, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 16675 + }, + { + "epoch": 0.16676, + "grad_norm": 1.333975506995556, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 16676 + }, + { + "epoch": 0.16677, + "grad_norm": 1.3659889263727105, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 16677 + }, + { + "epoch": 0.16678, + "grad_norm": 1.3232398475591127, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16678 + }, + { + "epoch": 0.16679, + "grad_norm": 1.372056133834555, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 16679 + }, + { + "epoch": 0.1668, + "grad_norm": 1.115400796578888, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16680 + }, + { + "epoch": 0.16681, + "grad_norm": 1.1462923783523356, + "learning_rate": 0.003, + "loss": 4.046, + "step": 16681 + }, + { + "epoch": 0.16682, + "grad_norm": 1.2822705786345197, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 16682 + }, + { + "epoch": 0.16683, + "grad_norm": 1.033317131360321, + "learning_rate": 0.003, + "loss": 4.055, + "step": 16683 + }, + { + "epoch": 0.16684, + "grad_norm": 1.5289135933913132, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16684 + }, + { + "epoch": 0.16685, + "grad_norm": 1.3185212238437796, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16685 + }, + { + "epoch": 0.16686, + "grad_norm": 1.2775429286296354, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 16686 + }, + { + "epoch": 0.16687, + "grad_norm": 1.310704859490125, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16687 + }, + { + "epoch": 0.16688, + "grad_norm": 1.1067398773051462, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 16688 + }, + { + "epoch": 0.16689, + "grad_norm": 1.4483876947604948, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16689 + }, + { + "epoch": 0.1669, + "grad_norm": 1.1567384752757102, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 16690 + }, + { + "epoch": 0.16691, + "grad_norm": 1.278726343385809, + "learning_rate": 0.003, + "loss": 4.077, + "step": 16691 + }, + { + "epoch": 0.16692, + "grad_norm": 1.0648070520419866, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 16692 + }, + { + "epoch": 0.16693, + "grad_norm": 1.4081478664976732, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 16693 + }, + { + "epoch": 0.16694, + "grad_norm": 1.1594991371369678, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16694 + }, + { + "epoch": 0.16695, + "grad_norm": 1.4636847314266446, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 16695 + }, + { + "epoch": 0.16696, + "grad_norm": 1.3951427322084577, + "learning_rate": 0.003, + "loss": 4.045, + "step": 16696 + }, + { + "epoch": 0.16697, + "grad_norm": 1.0347629487687562, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 16697 + }, + { + "epoch": 0.16698, + "grad_norm": 1.684232743457682, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 16698 + }, + { + "epoch": 0.16699, + "grad_norm": 0.8511245416361911, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 16699 + }, + { + "epoch": 0.167, + "grad_norm": 1.1386843043359771, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16700 + }, + { + "epoch": 0.16701, + "grad_norm": 1.7512591591733795, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 16701 + }, + { + "epoch": 0.16702, + "grad_norm": 0.9962913652943098, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 16702 + }, + { + "epoch": 0.16703, + "grad_norm": 1.2565831934183758, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16703 + }, + { + "epoch": 0.16704, + "grad_norm": 1.3433961964467531, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 16704 + }, + { + "epoch": 0.16705, + "grad_norm": 1.3910816757951838, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16705 + }, + { + "epoch": 0.16706, + "grad_norm": 1.3316360416361006, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 16706 + }, + { + "epoch": 0.16707, + "grad_norm": 0.991523558879406, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 16707 + }, + { + "epoch": 0.16708, + "grad_norm": 1.4048042006132784, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 16708 + }, + { + "epoch": 0.16709, + "grad_norm": 1.3886487992408054, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 16709 + }, + { + "epoch": 0.1671, + "grad_norm": 1.3016910098704964, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 16710 + }, + { + "epoch": 0.16711, + "grad_norm": 1.3477782241628462, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 16711 + }, + { + "epoch": 0.16712, + "grad_norm": 1.134131580682221, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 16712 + }, + { + "epoch": 0.16713, + "grad_norm": 1.3070182485559603, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 16713 + }, + { + "epoch": 0.16714, + "grad_norm": 1.1352574486120421, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 16714 + }, + { + "epoch": 0.16715, + "grad_norm": 1.3861551298051138, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 16715 + }, + { + "epoch": 0.16716, + "grad_norm": 1.221713531815258, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 16716 + }, + { + "epoch": 0.16717, + "grad_norm": 1.126946267158167, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16717 + }, + { + "epoch": 0.16718, + "grad_norm": 1.391457071408226, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16718 + }, + { + "epoch": 0.16719, + "grad_norm": 1.259439961310894, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 16719 + }, + { + "epoch": 0.1672, + "grad_norm": 1.3993421115373907, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16720 + }, + { + "epoch": 0.16721, + "grad_norm": 1.080223645130753, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 16721 + }, + { + "epoch": 0.16722, + "grad_norm": 1.5730549980475002, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 16722 + }, + { + "epoch": 0.16723, + "grad_norm": 1.1297822430243398, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16723 + }, + { + "epoch": 0.16724, + "grad_norm": 1.186892184655547, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 16724 + }, + { + "epoch": 0.16725, + "grad_norm": 1.2311396580283709, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 16725 + }, + { + "epoch": 0.16726, + "grad_norm": 1.140182561961307, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 16726 + }, + { + "epoch": 0.16727, + "grad_norm": 1.32465115746718, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 16727 + }, + { + "epoch": 0.16728, + "grad_norm": 1.2796869537548115, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 16728 + }, + { + "epoch": 0.16729, + "grad_norm": 1.181847879728677, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 16729 + }, + { + "epoch": 0.1673, + "grad_norm": 1.3776713468609083, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 16730 + }, + { + "epoch": 0.16731, + "grad_norm": 1.2565759096060845, + "learning_rate": 0.003, + "loss": 4.057, + "step": 16731 + }, + { + "epoch": 0.16732, + "grad_norm": 1.2550901339996967, + "learning_rate": 0.003, + "loss": 4.067, + "step": 16732 + }, + { + "epoch": 0.16733, + "grad_norm": 1.1175831388571573, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 16733 + }, + { + "epoch": 0.16734, + "grad_norm": 1.3450730195508063, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 16734 + }, + { + "epoch": 0.16735, + "grad_norm": 1.116004009871605, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 16735 + }, + { + "epoch": 0.16736, + "grad_norm": 1.621623953487288, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 16736 + }, + { + "epoch": 0.16737, + "grad_norm": 1.2525261016436242, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 16737 + }, + { + "epoch": 0.16738, + "grad_norm": 1.4305063793483448, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16738 + }, + { + "epoch": 0.16739, + "grad_norm": 1.662838940432477, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 16739 + }, + { + "epoch": 0.1674, + "grad_norm": 0.9891646601073584, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 16740 + }, + { + "epoch": 0.16741, + "grad_norm": 1.1384552528164928, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 16741 + }, + { + "epoch": 0.16742, + "grad_norm": 1.3547129028711147, + "learning_rate": 0.003, + "loss": 4.051, + "step": 16742 + }, + { + "epoch": 0.16743, + "grad_norm": 1.6243793382905138, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 16743 + }, + { + "epoch": 0.16744, + "grad_norm": 1.0975972967796066, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 16744 + }, + { + "epoch": 0.16745, + "grad_norm": 1.3074619617397136, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 16745 + }, + { + "epoch": 0.16746, + "grad_norm": 1.230182685581038, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 16746 + }, + { + "epoch": 0.16747, + "grad_norm": 1.1312420920964736, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 16747 + }, + { + "epoch": 0.16748, + "grad_norm": 1.2507249543126753, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 16748 + }, + { + "epoch": 0.16749, + "grad_norm": 1.1980369151237098, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 16749 + }, + { + "epoch": 0.1675, + "grad_norm": 1.4655643665371967, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 16750 + }, + { + "epoch": 0.16751, + "grad_norm": 1.286853091051945, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 16751 + }, + { + "epoch": 0.16752, + "grad_norm": 1.415363481460103, + "learning_rate": 0.003, + "loss": 4.063, + "step": 16752 + }, + { + "epoch": 0.16753, + "grad_norm": 1.377393120294197, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 16753 + }, + { + "epoch": 0.16754, + "grad_norm": 1.148828273784319, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 16754 + }, + { + "epoch": 0.16755, + "grad_norm": 1.2696627055678036, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 16755 + }, + { + "epoch": 0.16756, + "grad_norm": 1.3370050291485667, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 16756 + }, + { + "epoch": 0.16757, + "grad_norm": 1.1439509199832336, + "learning_rate": 0.003, + "loss": 4.066, + "step": 16757 + }, + { + "epoch": 0.16758, + "grad_norm": 1.2863755871018208, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16758 + }, + { + "epoch": 0.16759, + "grad_norm": 1.0127227549338795, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 16759 + }, + { + "epoch": 0.1676, + "grad_norm": 1.9709684350556145, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 16760 + }, + { + "epoch": 0.16761, + "grad_norm": 0.8838581402290888, + "learning_rate": 0.003, + "loss": 4.085, + "step": 16761 + }, + { + "epoch": 0.16762, + "grad_norm": 1.327185229227163, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 16762 + }, + { + "epoch": 0.16763, + "grad_norm": 1.0901753046638558, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 16763 + }, + { + "epoch": 0.16764, + "grad_norm": 1.1942403829502464, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 16764 + }, + { + "epoch": 0.16765, + "grad_norm": 1.4232219300432742, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 16765 + }, + { + "epoch": 0.16766, + "grad_norm": 1.1262783920801236, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 16766 + }, + { + "epoch": 0.16767, + "grad_norm": 1.4223258265073562, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 16767 + }, + { + "epoch": 0.16768, + "grad_norm": 1.4295881431617428, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 16768 + }, + { + "epoch": 0.16769, + "grad_norm": 1.0963666305613717, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16769 + }, + { + "epoch": 0.1677, + "grad_norm": 1.1414442762819592, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16770 + }, + { + "epoch": 0.16771, + "grad_norm": 1.2238226630858162, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 16771 + }, + { + "epoch": 0.16772, + "grad_norm": 1.6783808507277331, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 16772 + }, + { + "epoch": 0.16773, + "grad_norm": 1.095242159359064, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 16773 + }, + { + "epoch": 0.16774, + "grad_norm": 1.332248501252304, + "learning_rate": 0.003, + "loss": 4.083, + "step": 16774 + }, + { + "epoch": 0.16775, + "grad_norm": 1.155182945765907, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 16775 + }, + { + "epoch": 0.16776, + "grad_norm": 1.3650135785233826, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16776 + }, + { + "epoch": 0.16777, + "grad_norm": 1.4514998796554472, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16777 + }, + { + "epoch": 0.16778, + "grad_norm": 1.287162740762738, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 16778 + }, + { + "epoch": 0.16779, + "grad_norm": 1.3125614663209144, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 16779 + }, + { + "epoch": 0.1678, + "grad_norm": 1.0247349687445304, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 16780 + }, + { + "epoch": 0.16781, + "grad_norm": 1.454803588060753, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 16781 + }, + { + "epoch": 0.16782, + "grad_norm": 1.1087921509488263, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 16782 + }, + { + "epoch": 0.16783, + "grad_norm": 1.3142906154006264, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 16783 + }, + { + "epoch": 0.16784, + "grad_norm": 1.1099279060836573, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16784 + }, + { + "epoch": 0.16785, + "grad_norm": 1.220567366028567, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 16785 + }, + { + "epoch": 0.16786, + "grad_norm": 1.2919404227923874, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 16786 + }, + { + "epoch": 0.16787, + "grad_norm": 1.4363324714839323, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 16787 + }, + { + "epoch": 0.16788, + "grad_norm": 1.170231924201678, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 16788 + }, + { + "epoch": 0.16789, + "grad_norm": 1.393351981675191, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 16789 + }, + { + "epoch": 0.1679, + "grad_norm": 1.1336535552565001, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 16790 + }, + { + "epoch": 0.16791, + "grad_norm": 1.1474977162170739, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 16791 + }, + { + "epoch": 0.16792, + "grad_norm": 1.3920508668947844, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 16792 + }, + { + "epoch": 0.16793, + "grad_norm": 1.4227528877948599, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 16793 + }, + { + "epoch": 0.16794, + "grad_norm": 1.2060257722629177, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 16794 + }, + { + "epoch": 0.16795, + "grad_norm": 1.3600827230748966, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 16795 + }, + { + "epoch": 0.16796, + "grad_norm": 1.1770420210337125, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 16796 + }, + { + "epoch": 0.16797, + "grad_norm": 1.4393574934891733, + "learning_rate": 0.003, + "loss": 4.077, + "step": 16797 + }, + { + "epoch": 0.16798, + "grad_norm": 1.0901274871323479, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16798 + }, + { + "epoch": 0.16799, + "grad_norm": 1.5252149919240858, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 16799 + }, + { + "epoch": 0.168, + "grad_norm": 1.14032647596702, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 16800 + }, + { + "epoch": 0.16801, + "grad_norm": 1.327648166260483, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 16801 + }, + { + "epoch": 0.16802, + "grad_norm": 1.1002849680966285, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 16802 + }, + { + "epoch": 0.16803, + "grad_norm": 1.5068799667060553, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 16803 + }, + { + "epoch": 0.16804, + "grad_norm": 1.3470880046221414, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16804 + }, + { + "epoch": 0.16805, + "grad_norm": 1.2200143859080324, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 16805 + }, + { + "epoch": 0.16806, + "grad_norm": 1.307463612754022, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 16806 + }, + { + "epoch": 0.16807, + "grad_norm": 1.266542756627587, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 16807 + }, + { + "epoch": 0.16808, + "grad_norm": 1.123860411350716, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 16808 + }, + { + "epoch": 0.16809, + "grad_norm": 1.2527402200254805, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 16809 + }, + { + "epoch": 0.1681, + "grad_norm": 1.2286560649219227, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 16810 + }, + { + "epoch": 0.16811, + "grad_norm": 1.2794688476481573, + "learning_rate": 0.003, + "loss": 4.091, + "step": 16811 + }, + { + "epoch": 0.16812, + "grad_norm": 1.0624837677134609, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16812 + }, + { + "epoch": 0.16813, + "grad_norm": 1.3450544802064075, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 16813 + }, + { + "epoch": 0.16814, + "grad_norm": 1.1588218630193072, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16814 + }, + { + "epoch": 0.16815, + "grad_norm": 1.276754195656693, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 16815 + }, + { + "epoch": 0.16816, + "grad_norm": 1.2745892758250235, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 16816 + }, + { + "epoch": 0.16817, + "grad_norm": 1.3571051184806497, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 16817 + }, + { + "epoch": 0.16818, + "grad_norm": 1.267336203960523, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 16818 + }, + { + "epoch": 0.16819, + "grad_norm": 1.1924544666810764, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 16819 + }, + { + "epoch": 0.1682, + "grad_norm": 1.207256301555306, + "learning_rate": 0.003, + "loss": 4.054, + "step": 16820 + }, + { + "epoch": 0.16821, + "grad_norm": 1.370854375794414, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 16821 + }, + { + "epoch": 0.16822, + "grad_norm": 1.3663088529821639, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 16822 + }, + { + "epoch": 0.16823, + "grad_norm": 1.1774873108786654, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 16823 + }, + { + "epoch": 0.16824, + "grad_norm": 1.4194242487443058, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 16824 + }, + { + "epoch": 0.16825, + "grad_norm": 1.3822584662925912, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 16825 + }, + { + "epoch": 0.16826, + "grad_norm": 1.1102142883788981, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 16826 + }, + { + "epoch": 0.16827, + "grad_norm": 1.38003821159165, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16827 + }, + { + "epoch": 0.16828, + "grad_norm": 1.1851290243500439, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16828 + }, + { + "epoch": 0.16829, + "grad_norm": 1.6430352870086788, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 16829 + }, + { + "epoch": 0.1683, + "grad_norm": 1.0075497669286884, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 16830 + }, + { + "epoch": 0.16831, + "grad_norm": 1.5054189554987507, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 16831 + }, + { + "epoch": 0.16832, + "grad_norm": 1.3496529029391273, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 16832 + }, + { + "epoch": 0.16833, + "grad_norm": 1.4899614656492906, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 16833 + }, + { + "epoch": 0.16834, + "grad_norm": 1.2239859880131512, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 16834 + }, + { + "epoch": 0.16835, + "grad_norm": 1.0447703047011303, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16835 + }, + { + "epoch": 0.16836, + "grad_norm": 1.2672507917954132, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 16836 + }, + { + "epoch": 0.16837, + "grad_norm": 1.2027880989996271, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 16837 + }, + { + "epoch": 0.16838, + "grad_norm": 1.4245659086392621, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 16838 + }, + { + "epoch": 0.16839, + "grad_norm": 1.2037598344828246, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 16839 + }, + { + "epoch": 0.1684, + "grad_norm": 1.0951149498838402, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 16840 + }, + { + "epoch": 0.16841, + "grad_norm": 1.4364129329916702, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 16841 + }, + { + "epoch": 0.16842, + "grad_norm": 1.068806632324706, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 16842 + }, + { + "epoch": 0.16843, + "grad_norm": 1.4819403695542883, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 16843 + }, + { + "epoch": 0.16844, + "grad_norm": 1.1490258055091571, + "learning_rate": 0.003, + "loss": 4.093, + "step": 16844 + }, + { + "epoch": 0.16845, + "grad_norm": 1.3541141006956319, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 16845 + }, + { + "epoch": 0.16846, + "grad_norm": 1.1206996339264277, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 16846 + }, + { + "epoch": 0.16847, + "grad_norm": 1.5244419982059054, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16847 + }, + { + "epoch": 0.16848, + "grad_norm": 1.09708306061788, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16848 + }, + { + "epoch": 0.16849, + "grad_norm": 1.470418708707691, + "learning_rate": 0.003, + "loss": 4.054, + "step": 16849 + }, + { + "epoch": 0.1685, + "grad_norm": 1.2319714683828253, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 16850 + }, + { + "epoch": 0.16851, + "grad_norm": 1.4095576657043976, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 16851 + }, + { + "epoch": 0.16852, + "grad_norm": 1.1033045827696277, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 16852 + }, + { + "epoch": 0.16853, + "grad_norm": 1.3847096334809648, + "learning_rate": 0.003, + "loss": 4.088, + "step": 16853 + }, + { + "epoch": 0.16854, + "grad_norm": 1.1113328080159366, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 16854 + }, + { + "epoch": 0.16855, + "grad_norm": 1.224386372223287, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 16855 + }, + { + "epoch": 0.16856, + "grad_norm": 1.1916273075115396, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 16856 + }, + { + "epoch": 0.16857, + "grad_norm": 1.1481876709715975, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 16857 + }, + { + "epoch": 0.16858, + "grad_norm": 1.3593802424987913, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 16858 + }, + { + "epoch": 0.16859, + "grad_norm": 1.273283610859603, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 16859 + }, + { + "epoch": 0.1686, + "grad_norm": 1.496853202283384, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 16860 + }, + { + "epoch": 0.16861, + "grad_norm": 1.40736923510857, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 16861 + }, + { + "epoch": 0.16862, + "grad_norm": 1.490879432399349, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16862 + }, + { + "epoch": 0.16863, + "grad_norm": 1.31948245652066, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 16863 + }, + { + "epoch": 0.16864, + "grad_norm": 1.301382904637517, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 16864 + }, + { + "epoch": 0.16865, + "grad_norm": 1.5323866546447675, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 16865 + }, + { + "epoch": 0.16866, + "grad_norm": 1.1631458714171514, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 16866 + }, + { + "epoch": 0.16867, + "grad_norm": 1.158729013732875, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 16867 + }, + { + "epoch": 0.16868, + "grad_norm": 1.1228514993443777, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 16868 + }, + { + "epoch": 0.16869, + "grad_norm": 1.393778205896287, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 16869 + }, + { + "epoch": 0.1687, + "grad_norm": 1.3243045374960487, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 16870 + }, + { + "epoch": 0.16871, + "grad_norm": 1.2010567934501997, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16871 + }, + { + "epoch": 0.16872, + "grad_norm": 1.440514056954753, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 16872 + }, + { + "epoch": 0.16873, + "grad_norm": 1.2139601121876216, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 16873 + }, + { + "epoch": 0.16874, + "grad_norm": 1.4099829826682746, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 16874 + }, + { + "epoch": 0.16875, + "grad_norm": 1.146165363835504, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 16875 + }, + { + "epoch": 0.16876, + "grad_norm": 1.241179998010361, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 16876 + }, + { + "epoch": 0.16877, + "grad_norm": 1.1944182913192236, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 16877 + }, + { + "epoch": 0.16878, + "grad_norm": 1.2242141822087804, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 16878 + }, + { + "epoch": 0.16879, + "grad_norm": 1.115057333475243, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16879 + }, + { + "epoch": 0.1688, + "grad_norm": 1.270200199752863, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 16880 + }, + { + "epoch": 0.16881, + "grad_norm": 1.2019256836950207, + "learning_rate": 0.003, + "loss": 4.047, + "step": 16881 + }, + { + "epoch": 0.16882, + "grad_norm": 1.4479286266212343, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 16882 + }, + { + "epoch": 0.16883, + "grad_norm": 1.062123034185907, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 16883 + }, + { + "epoch": 0.16884, + "grad_norm": 1.3882209746942173, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16884 + }, + { + "epoch": 0.16885, + "grad_norm": 1.109180520492913, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 16885 + }, + { + "epoch": 0.16886, + "grad_norm": 1.2836788033972242, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 16886 + }, + { + "epoch": 0.16887, + "grad_norm": 1.2521604536286681, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 16887 + }, + { + "epoch": 0.16888, + "grad_norm": 1.4136543025600252, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 16888 + }, + { + "epoch": 0.16889, + "grad_norm": 1.5782665241111962, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 16889 + }, + { + "epoch": 0.1689, + "grad_norm": 1.1959929754457943, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 16890 + }, + { + "epoch": 0.16891, + "grad_norm": 1.2352924174891171, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 16891 + }, + { + "epoch": 0.16892, + "grad_norm": 1.58818430273992, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 16892 + }, + { + "epoch": 0.16893, + "grad_norm": 1.1601970694296755, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16893 + }, + { + "epoch": 0.16894, + "grad_norm": 1.4143245558241038, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 16894 + }, + { + "epoch": 0.16895, + "grad_norm": 1.1094466555637557, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 16895 + }, + { + "epoch": 0.16896, + "grad_norm": 1.2896671978581782, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 16896 + }, + { + "epoch": 0.16897, + "grad_norm": 1.2833778034226433, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 16897 + }, + { + "epoch": 0.16898, + "grad_norm": 1.2460864940528413, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 16898 + }, + { + "epoch": 0.16899, + "grad_norm": 1.2819890523227007, + "learning_rate": 0.003, + "loss": 4.076, + "step": 16899 + }, + { + "epoch": 0.169, + "grad_norm": 1.2147009440965972, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 16900 + }, + { + "epoch": 0.16901, + "grad_norm": 1.544973459892542, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 16901 + }, + { + "epoch": 0.16902, + "grad_norm": 1.3220660850852575, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 16902 + }, + { + "epoch": 0.16903, + "grad_norm": 1.399095635772759, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 16903 + }, + { + "epoch": 0.16904, + "grad_norm": 1.1102399477345783, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 16904 + }, + { + "epoch": 0.16905, + "grad_norm": 1.4593558716593782, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 16905 + }, + { + "epoch": 0.16906, + "grad_norm": 1.3101786260146533, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 16906 + }, + { + "epoch": 0.16907, + "grad_norm": 1.0702024985287735, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 16907 + }, + { + "epoch": 0.16908, + "grad_norm": 1.3887646224909347, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16908 + }, + { + "epoch": 0.16909, + "grad_norm": 1.1828192580708548, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 16909 + }, + { + "epoch": 0.1691, + "grad_norm": 1.3176191915351163, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16910 + }, + { + "epoch": 0.16911, + "grad_norm": 1.4570497416853039, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 16911 + }, + { + "epoch": 0.16912, + "grad_norm": 1.347768060910466, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 16912 + }, + { + "epoch": 0.16913, + "grad_norm": 1.4406348710462704, + "learning_rate": 0.003, + "loss": 4.059, + "step": 16913 + }, + { + "epoch": 0.16914, + "grad_norm": 1.1302934221198233, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 16914 + }, + { + "epoch": 0.16915, + "grad_norm": 1.2401024203800077, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 16915 + }, + { + "epoch": 0.16916, + "grad_norm": 1.2956588105084923, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 16916 + }, + { + "epoch": 0.16917, + "grad_norm": 1.24168297084552, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 16917 + }, + { + "epoch": 0.16918, + "grad_norm": 1.2961998472371414, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 16918 + }, + { + "epoch": 0.16919, + "grad_norm": 1.2452220446431184, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 16919 + }, + { + "epoch": 0.1692, + "grad_norm": 1.2714969308512136, + "learning_rate": 0.003, + "loss": 4.038, + "step": 16920 + }, + { + "epoch": 0.16921, + "grad_norm": 1.0320651012042785, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 16921 + }, + { + "epoch": 0.16922, + "grad_norm": 1.3198445938865049, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 16922 + }, + { + "epoch": 0.16923, + "grad_norm": 1.1314290964153324, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16923 + }, + { + "epoch": 0.16924, + "grad_norm": 1.4219952718125404, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 16924 + }, + { + "epoch": 0.16925, + "grad_norm": 1.1234743637945024, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 16925 + }, + { + "epoch": 0.16926, + "grad_norm": 1.4123328622018225, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16926 + }, + { + "epoch": 0.16927, + "grad_norm": 1.4473500817037783, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 16927 + }, + { + "epoch": 0.16928, + "grad_norm": 1.2799292837639202, + "learning_rate": 0.003, + "loss": 4.07, + "step": 16928 + }, + { + "epoch": 0.16929, + "grad_norm": 1.4288548030602994, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 16929 + }, + { + "epoch": 0.1693, + "grad_norm": 1.3195056052785827, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16930 + }, + { + "epoch": 0.16931, + "grad_norm": 1.1625917507435435, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 16931 + }, + { + "epoch": 0.16932, + "grad_norm": 1.4378633883226684, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 16932 + }, + { + "epoch": 0.16933, + "grad_norm": 1.1256401510921825, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16933 + }, + { + "epoch": 0.16934, + "grad_norm": 1.3449069483606546, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 16934 + }, + { + "epoch": 0.16935, + "grad_norm": 1.1393050219429626, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 16935 + }, + { + "epoch": 0.16936, + "grad_norm": 1.2662484153953546, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16936 + }, + { + "epoch": 0.16937, + "grad_norm": 1.2963480614111844, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 16937 + }, + { + "epoch": 0.16938, + "grad_norm": 1.1074316734336083, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 16938 + }, + { + "epoch": 0.16939, + "grad_norm": 1.3065192526603162, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 16939 + }, + { + "epoch": 0.1694, + "grad_norm": 1.225136446838548, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 16940 + }, + { + "epoch": 0.16941, + "grad_norm": 1.5104008620908072, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 16941 + }, + { + "epoch": 0.16942, + "grad_norm": 1.1440377001531545, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 16942 + }, + { + "epoch": 0.16943, + "grad_norm": 1.3343754663051575, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16943 + }, + { + "epoch": 0.16944, + "grad_norm": 1.278040988184129, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16944 + }, + { + "epoch": 0.16945, + "grad_norm": 1.3936716350946827, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 16945 + }, + { + "epoch": 0.16946, + "grad_norm": 0.9878081793633148, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 16946 + }, + { + "epoch": 0.16947, + "grad_norm": 1.4320519568494658, + "learning_rate": 0.003, + "loss": 4.069, + "step": 16947 + }, + { + "epoch": 0.16948, + "grad_norm": 1.244286346836353, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 16948 + }, + { + "epoch": 0.16949, + "grad_norm": 1.2299209004693132, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16949 + }, + { + "epoch": 0.1695, + "grad_norm": 1.0409565052429766, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 16950 + }, + { + "epoch": 0.16951, + "grad_norm": 1.2508836847989877, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 16951 + }, + { + "epoch": 0.16952, + "grad_norm": 1.2967029164784285, + "learning_rate": 0.003, + "loss": 4.056, + "step": 16952 + }, + { + "epoch": 0.16953, + "grad_norm": 1.4077178182440349, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 16953 + }, + { + "epoch": 0.16954, + "grad_norm": 1.0826038727213632, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 16954 + }, + { + "epoch": 0.16955, + "grad_norm": 1.2115900177002683, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 16955 + }, + { + "epoch": 0.16956, + "grad_norm": 1.4481540542492375, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 16956 + }, + { + "epoch": 0.16957, + "grad_norm": 1.54688189499958, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 16957 + }, + { + "epoch": 0.16958, + "grad_norm": 1.3514547824099727, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 16958 + }, + { + "epoch": 0.16959, + "grad_norm": 1.3902847878239193, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 16959 + }, + { + "epoch": 0.1696, + "grad_norm": 1.0485211572345803, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 16960 + }, + { + "epoch": 0.16961, + "grad_norm": 1.4504868829279804, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 16961 + }, + { + "epoch": 0.16962, + "grad_norm": 1.2840105861582294, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 16962 + }, + { + "epoch": 0.16963, + "grad_norm": 1.2296728562735413, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 16963 + }, + { + "epoch": 0.16964, + "grad_norm": 1.1315041822083285, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 16964 + }, + { + "epoch": 0.16965, + "grad_norm": 1.3289468427844238, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16965 + }, + { + "epoch": 0.16966, + "grad_norm": 1.0464856718548334, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 16966 + }, + { + "epoch": 0.16967, + "grad_norm": 1.3937570112756352, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 16967 + }, + { + "epoch": 0.16968, + "grad_norm": 1.3874378247783272, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 16968 + }, + { + "epoch": 0.16969, + "grad_norm": 1.1011033569991182, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16969 + }, + { + "epoch": 0.1697, + "grad_norm": 1.5819831429822668, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 16970 + }, + { + "epoch": 0.16971, + "grad_norm": 0.9604786741124874, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 16971 + }, + { + "epoch": 0.16972, + "grad_norm": 1.4722719123412784, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 16972 + }, + { + "epoch": 0.16973, + "grad_norm": 1.0411160539916922, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 16973 + }, + { + "epoch": 0.16974, + "grad_norm": 1.3526490148604315, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 16974 + }, + { + "epoch": 0.16975, + "grad_norm": 1.2802469104047067, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16975 + }, + { + "epoch": 0.16976, + "grad_norm": 1.2589720717564747, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 16976 + }, + { + "epoch": 0.16977, + "grad_norm": 1.2634228865755068, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 16977 + }, + { + "epoch": 0.16978, + "grad_norm": 1.1191953796902516, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 16978 + }, + { + "epoch": 0.16979, + "grad_norm": 1.2790069978179737, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 16979 + }, + { + "epoch": 0.1698, + "grad_norm": 1.5249885874241196, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 16980 + }, + { + "epoch": 0.16981, + "grad_norm": 1.1204813867109078, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 16981 + }, + { + "epoch": 0.16982, + "grad_norm": 1.6120602744889545, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 16982 + }, + { + "epoch": 0.16983, + "grad_norm": 1.2142625856634415, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16983 + }, + { + "epoch": 0.16984, + "grad_norm": 1.5650189553425617, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 16984 + }, + { + "epoch": 0.16985, + "grad_norm": 1.2220873866641324, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 16985 + }, + { + "epoch": 0.16986, + "grad_norm": 1.251526545562485, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 16986 + }, + { + "epoch": 0.16987, + "grad_norm": 1.2099160891817955, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 16987 + }, + { + "epoch": 0.16988, + "grad_norm": 1.1561308162598547, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 16988 + }, + { + "epoch": 0.16989, + "grad_norm": 1.3716886639666412, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 16989 + }, + { + "epoch": 0.1699, + "grad_norm": 1.2276839460561324, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 16990 + }, + { + "epoch": 0.16991, + "grad_norm": 1.4220253366428155, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 16991 + }, + { + "epoch": 0.16992, + "grad_norm": 1.4182198336528233, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 16992 + }, + { + "epoch": 0.16993, + "grad_norm": 1.3889111493859785, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 16993 + }, + { + "epoch": 0.16994, + "grad_norm": 1.2130238797491997, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 16994 + }, + { + "epoch": 0.16995, + "grad_norm": 1.1780230433886487, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 16995 + }, + { + "epoch": 0.16996, + "grad_norm": 1.3603919110346885, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 16996 + }, + { + "epoch": 0.16997, + "grad_norm": 1.2760748429890882, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 16997 + }, + { + "epoch": 0.16998, + "grad_norm": 1.2502815178778632, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 16998 + }, + { + "epoch": 0.16999, + "grad_norm": 1.6165519910026174, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 16999 + }, + { + "epoch": 0.17, + "grad_norm": 1.0053911199760457, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 17000 + }, + { + "epoch": 0.17001, + "grad_norm": 1.2888523071738722, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 17001 + }, + { + "epoch": 0.17002, + "grad_norm": 1.0247128604707927, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 17002 + }, + { + "epoch": 0.17003, + "grad_norm": 1.4999931056125426, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 17003 + }, + { + "epoch": 0.17004, + "grad_norm": 1.3596282609549202, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17004 + }, + { + "epoch": 0.17005, + "grad_norm": 1.1514114776069913, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17005 + }, + { + "epoch": 0.17006, + "grad_norm": 1.3312429685018117, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 17006 + }, + { + "epoch": 0.17007, + "grad_norm": 1.218792080089117, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 17007 + }, + { + "epoch": 0.17008, + "grad_norm": 1.197802950001522, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 17008 + }, + { + "epoch": 0.17009, + "grad_norm": 1.3763630217530811, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17009 + }, + { + "epoch": 0.1701, + "grad_norm": 1.4212308500084787, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17010 + }, + { + "epoch": 0.17011, + "grad_norm": 0.953128092845532, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 17011 + }, + { + "epoch": 0.17012, + "grad_norm": 1.324057361967305, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 17012 + }, + { + "epoch": 0.17013, + "grad_norm": 1.4539562806509976, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 17013 + }, + { + "epoch": 0.17014, + "grad_norm": 1.3728094854799713, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 17014 + }, + { + "epoch": 0.17015, + "grad_norm": 1.7138404683265989, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 17015 + }, + { + "epoch": 0.17016, + "grad_norm": 1.0001556175972248, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17016 + }, + { + "epoch": 0.17017, + "grad_norm": 1.3144709201023723, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17017 + }, + { + "epoch": 0.17018, + "grad_norm": 1.183298451331619, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 17018 + }, + { + "epoch": 0.17019, + "grad_norm": 1.2976131957707802, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17019 + }, + { + "epoch": 0.1702, + "grad_norm": 1.4441694521919246, + "learning_rate": 0.003, + "loss": 4.079, + "step": 17020 + }, + { + "epoch": 0.17021, + "grad_norm": 1.0231368351229009, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 17021 + }, + { + "epoch": 0.17022, + "grad_norm": 1.4522769663622839, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 17022 + }, + { + "epoch": 0.17023, + "grad_norm": 1.2115552248199124, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 17023 + }, + { + "epoch": 0.17024, + "grad_norm": 1.3955549248970327, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 17024 + }, + { + "epoch": 0.17025, + "grad_norm": 1.767413058772527, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 17025 + }, + { + "epoch": 0.17026, + "grad_norm": 1.1257427274255005, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 17026 + }, + { + "epoch": 0.17027, + "grad_norm": 1.3977854342088645, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 17027 + }, + { + "epoch": 0.17028, + "grad_norm": 1.4133438305646633, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17028 + }, + { + "epoch": 0.17029, + "grad_norm": 0.9857828699963934, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 17029 + }, + { + "epoch": 0.1703, + "grad_norm": 1.2909867094838485, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17030 + }, + { + "epoch": 0.17031, + "grad_norm": 1.2136917950238058, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 17031 + }, + { + "epoch": 0.17032, + "grad_norm": 1.3308725833491453, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17032 + }, + { + "epoch": 0.17033, + "grad_norm": 1.1916228887328304, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 17033 + }, + { + "epoch": 0.17034, + "grad_norm": 1.3075246141432557, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17034 + }, + { + "epoch": 0.17035, + "grad_norm": 1.2721657649715863, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17035 + }, + { + "epoch": 0.17036, + "grad_norm": 1.2104509544352788, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17036 + }, + { + "epoch": 0.17037, + "grad_norm": 1.1525415633438325, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 17037 + }, + { + "epoch": 0.17038, + "grad_norm": 1.1475974463818974, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 17038 + }, + { + "epoch": 0.17039, + "grad_norm": 1.5018976469381597, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 17039 + }, + { + "epoch": 0.1704, + "grad_norm": 1.0270191604314975, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 17040 + }, + { + "epoch": 0.17041, + "grad_norm": 1.3971266040112438, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 17041 + }, + { + "epoch": 0.17042, + "grad_norm": 1.3538705145785344, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17042 + }, + { + "epoch": 0.17043, + "grad_norm": 1.2683846167389807, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 17043 + }, + { + "epoch": 0.17044, + "grad_norm": 1.4541193508004426, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 17044 + }, + { + "epoch": 0.17045, + "grad_norm": 1.0988360012839928, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17045 + }, + { + "epoch": 0.17046, + "grad_norm": 1.6000719523297462, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 17046 + }, + { + "epoch": 0.17047, + "grad_norm": 1.2817610928722536, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 17047 + }, + { + "epoch": 0.17048, + "grad_norm": 1.2292969932309563, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 17048 + }, + { + "epoch": 0.17049, + "grad_norm": 1.2630107557417314, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17049 + }, + { + "epoch": 0.1705, + "grad_norm": 1.276507448279547, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17050 + }, + { + "epoch": 0.17051, + "grad_norm": 1.3492516633248473, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17051 + }, + { + "epoch": 0.17052, + "grad_norm": 1.1940749532981079, + "learning_rate": 0.003, + "loss": 4.088, + "step": 17052 + }, + { + "epoch": 0.17053, + "grad_norm": 1.573189456458545, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 17053 + }, + { + "epoch": 0.17054, + "grad_norm": 1.210068121530533, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 17054 + }, + { + "epoch": 0.17055, + "grad_norm": 1.169180023801874, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 17055 + }, + { + "epoch": 0.17056, + "grad_norm": 1.2195067811040499, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 17056 + }, + { + "epoch": 0.17057, + "grad_norm": 1.360570643958307, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 17057 + }, + { + "epoch": 0.17058, + "grad_norm": 1.0253700488945972, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 17058 + }, + { + "epoch": 0.17059, + "grad_norm": 1.5336780724536332, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 17059 + }, + { + "epoch": 0.1706, + "grad_norm": 1.178011295450224, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 17060 + }, + { + "epoch": 0.17061, + "grad_norm": 1.3498779313088982, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 17061 + }, + { + "epoch": 0.17062, + "grad_norm": 1.1502540681934281, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17062 + }, + { + "epoch": 0.17063, + "grad_norm": 1.384156634041985, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 17063 + }, + { + "epoch": 0.17064, + "grad_norm": 1.2550908972261812, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 17064 + }, + { + "epoch": 0.17065, + "grad_norm": 1.3201841544989974, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17065 + }, + { + "epoch": 0.17066, + "grad_norm": 1.3370917019906838, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17066 + }, + { + "epoch": 0.17067, + "grad_norm": 1.2869269678962787, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 17067 + }, + { + "epoch": 0.17068, + "grad_norm": 1.219233000989476, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 17068 + }, + { + "epoch": 0.17069, + "grad_norm": 1.4603560265924007, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 17069 + }, + { + "epoch": 0.1707, + "grad_norm": 1.313142605798217, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17070 + }, + { + "epoch": 0.17071, + "grad_norm": 1.354739867548382, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 17071 + }, + { + "epoch": 0.17072, + "grad_norm": 1.322882693187208, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17072 + }, + { + "epoch": 0.17073, + "grad_norm": 1.324393461801636, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 17073 + }, + { + "epoch": 0.17074, + "grad_norm": 1.1466469394594943, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17074 + }, + { + "epoch": 0.17075, + "grad_norm": 1.184709216749383, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17075 + }, + { + "epoch": 0.17076, + "grad_norm": 1.380052308242218, + "learning_rate": 0.003, + "loss": 4.047, + "step": 17076 + }, + { + "epoch": 0.17077, + "grad_norm": 1.273079251798578, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17077 + }, + { + "epoch": 0.17078, + "grad_norm": 1.3272735724648692, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 17078 + }, + { + "epoch": 0.17079, + "grad_norm": 1.0803079847919785, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17079 + }, + { + "epoch": 0.1708, + "grad_norm": 1.1239999521431894, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17080 + }, + { + "epoch": 0.17081, + "grad_norm": 1.430104041450772, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17081 + }, + { + "epoch": 0.17082, + "grad_norm": 1.4949733715992284, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 17082 + }, + { + "epoch": 0.17083, + "grad_norm": 1.2917046206070553, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17083 + }, + { + "epoch": 0.17084, + "grad_norm": 1.2323312232063808, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 17084 + }, + { + "epoch": 0.17085, + "grad_norm": 1.4111662613496083, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 17085 + }, + { + "epoch": 0.17086, + "grad_norm": 1.3477746976894227, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17086 + }, + { + "epoch": 0.17087, + "grad_norm": 1.233966993941767, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17087 + }, + { + "epoch": 0.17088, + "grad_norm": 1.3116288329107746, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17088 + }, + { + "epoch": 0.17089, + "grad_norm": 1.273728819976221, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17089 + }, + { + "epoch": 0.1709, + "grad_norm": 1.505682932680468, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17090 + }, + { + "epoch": 0.17091, + "grad_norm": 1.337474022173787, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 17091 + }, + { + "epoch": 0.17092, + "grad_norm": 1.4887830578497672, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 17092 + }, + { + "epoch": 0.17093, + "grad_norm": 1.315462724256315, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 17093 + }, + { + "epoch": 0.17094, + "grad_norm": 1.2402138789365362, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 17094 + }, + { + "epoch": 0.17095, + "grad_norm": 1.2766321856539715, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 17095 + }, + { + "epoch": 0.17096, + "grad_norm": 1.3477141194221316, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17096 + }, + { + "epoch": 0.17097, + "grad_norm": 1.0294500312951556, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 17097 + }, + { + "epoch": 0.17098, + "grad_norm": 1.2372975873508332, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 17098 + }, + { + "epoch": 0.17099, + "grad_norm": 1.2999064965554796, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 17099 + }, + { + "epoch": 0.171, + "grad_norm": 1.2421551693192254, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 17100 + }, + { + "epoch": 0.17101, + "grad_norm": 1.3029925757656373, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 17101 + }, + { + "epoch": 0.17102, + "grad_norm": 1.413628790406156, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17102 + }, + { + "epoch": 0.17103, + "grad_norm": 1.270813481606152, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17103 + }, + { + "epoch": 0.17104, + "grad_norm": 1.4491769084696768, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 17104 + }, + { + "epoch": 0.17105, + "grad_norm": 1.2999105264354298, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 17105 + }, + { + "epoch": 0.17106, + "grad_norm": 1.493524806336209, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17106 + }, + { + "epoch": 0.17107, + "grad_norm": 1.1253355514126384, + "learning_rate": 0.003, + "loss": 4.082, + "step": 17107 + }, + { + "epoch": 0.17108, + "grad_norm": 1.211621058987755, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17108 + }, + { + "epoch": 0.17109, + "grad_norm": 1.2953679514880798, + "learning_rate": 0.003, + "loss": 4.053, + "step": 17109 + }, + { + "epoch": 0.1711, + "grad_norm": 1.2760926805103348, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17110 + }, + { + "epoch": 0.17111, + "grad_norm": 1.2362842209560756, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 17111 + }, + { + "epoch": 0.17112, + "grad_norm": 1.2360788813061327, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17112 + }, + { + "epoch": 0.17113, + "grad_norm": 1.339630283544096, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 17113 + }, + { + "epoch": 0.17114, + "grad_norm": 1.2114235919349148, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 17114 + }, + { + "epoch": 0.17115, + "grad_norm": 1.4198339218631812, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17115 + }, + { + "epoch": 0.17116, + "grad_norm": 1.1717499098535376, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 17116 + }, + { + "epoch": 0.17117, + "grad_norm": 1.2909379254052165, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 17117 + }, + { + "epoch": 0.17118, + "grad_norm": 1.283910040750877, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 17118 + }, + { + "epoch": 0.17119, + "grad_norm": 1.1893986812955086, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17119 + }, + { + "epoch": 0.1712, + "grad_norm": 1.265651088184533, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 17120 + }, + { + "epoch": 0.17121, + "grad_norm": 1.071795774668601, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 17121 + }, + { + "epoch": 0.17122, + "grad_norm": 1.648050976357266, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 17122 + }, + { + "epoch": 0.17123, + "grad_norm": 0.9528963605296452, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17123 + }, + { + "epoch": 0.17124, + "grad_norm": 1.5673856382501787, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17124 + }, + { + "epoch": 0.17125, + "grad_norm": 0.9182606986835793, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 17125 + }, + { + "epoch": 0.17126, + "grad_norm": 1.5321951332412338, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 17126 + }, + { + "epoch": 0.17127, + "grad_norm": 1.1578440872666076, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 17127 + }, + { + "epoch": 0.17128, + "grad_norm": 1.221124572752152, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 17128 + }, + { + "epoch": 0.17129, + "grad_norm": 1.3130787941979272, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 17129 + }, + { + "epoch": 0.1713, + "grad_norm": 1.1637149225438805, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17130 + }, + { + "epoch": 0.17131, + "grad_norm": 1.2916207885201532, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17131 + }, + { + "epoch": 0.17132, + "grad_norm": 1.4271216764913037, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 17132 + }, + { + "epoch": 0.17133, + "grad_norm": 1.3067500766517877, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 17133 + }, + { + "epoch": 0.17134, + "grad_norm": 1.32243194743986, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 17134 + }, + { + "epoch": 0.17135, + "grad_norm": 1.484578398902238, + "learning_rate": 0.003, + "loss": 4.077, + "step": 17135 + }, + { + "epoch": 0.17136, + "grad_norm": 1.2726072690525767, + "learning_rate": 0.003, + "loss": 4.055, + "step": 17136 + }, + { + "epoch": 0.17137, + "grad_norm": 1.393464284738465, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 17137 + }, + { + "epoch": 0.17138, + "grad_norm": 1.282219676916128, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17138 + }, + { + "epoch": 0.17139, + "grad_norm": 1.4666292925377877, + "learning_rate": 0.003, + "loss": 4.075, + "step": 17139 + }, + { + "epoch": 0.1714, + "grad_norm": 1.0967948384984816, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 17140 + }, + { + "epoch": 0.17141, + "grad_norm": 1.4127740532449033, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 17141 + }, + { + "epoch": 0.17142, + "grad_norm": 1.1499694824347622, + "learning_rate": 0.003, + "loss": 4.034, + "step": 17142 + }, + { + "epoch": 0.17143, + "grad_norm": 1.250961346843233, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 17143 + }, + { + "epoch": 0.17144, + "grad_norm": 1.4443254164579862, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17144 + }, + { + "epoch": 0.17145, + "grad_norm": 1.133039401868114, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17145 + }, + { + "epoch": 0.17146, + "grad_norm": 1.5210457604440224, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17146 + }, + { + "epoch": 0.17147, + "grad_norm": 1.3175377150941554, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 17147 + }, + { + "epoch": 0.17148, + "grad_norm": 1.2857199420481968, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17148 + }, + { + "epoch": 0.17149, + "grad_norm": 1.2421283281105395, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 17149 + }, + { + "epoch": 0.1715, + "grad_norm": 1.0859540141491995, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17150 + }, + { + "epoch": 0.17151, + "grad_norm": 1.3327157766348299, + "learning_rate": 0.003, + "loss": 4.073, + "step": 17151 + }, + { + "epoch": 0.17152, + "grad_norm": 1.2225879267594058, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 17152 + }, + { + "epoch": 0.17153, + "grad_norm": 1.4078252435542349, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17153 + }, + { + "epoch": 0.17154, + "grad_norm": 1.2346220383369924, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17154 + }, + { + "epoch": 0.17155, + "grad_norm": 1.3211567280413448, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 17155 + }, + { + "epoch": 0.17156, + "grad_norm": 1.6188785958146443, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 17156 + }, + { + "epoch": 0.17157, + "grad_norm": 1.1252843117347187, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17157 + }, + { + "epoch": 0.17158, + "grad_norm": 1.3454771111737065, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17158 + }, + { + "epoch": 0.17159, + "grad_norm": 1.3145208555465913, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 17159 + }, + { + "epoch": 0.1716, + "grad_norm": 1.3265742013322044, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17160 + }, + { + "epoch": 0.17161, + "grad_norm": 1.0264878314735832, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 17161 + }, + { + "epoch": 0.17162, + "grad_norm": 1.5524565318758603, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 17162 + }, + { + "epoch": 0.17163, + "grad_norm": 1.1832378641723125, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 17163 + }, + { + "epoch": 0.17164, + "grad_norm": 1.3867140127479662, + "learning_rate": 0.003, + "loss": 4.064, + "step": 17164 + }, + { + "epoch": 0.17165, + "grad_norm": 1.269827437724221, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17165 + }, + { + "epoch": 0.17166, + "grad_norm": 1.3508449911267215, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 17166 + }, + { + "epoch": 0.17167, + "grad_norm": 1.0951431787498167, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 17167 + }, + { + "epoch": 0.17168, + "grad_norm": 1.34342161127684, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 17168 + }, + { + "epoch": 0.17169, + "grad_norm": 1.2916502046230676, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 17169 + }, + { + "epoch": 0.1717, + "grad_norm": 1.1609694039144327, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 17170 + }, + { + "epoch": 0.17171, + "grad_norm": 1.320479587536383, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17171 + }, + { + "epoch": 0.17172, + "grad_norm": 1.4114748920774176, + "learning_rate": 0.003, + "loss": 4.072, + "step": 17172 + }, + { + "epoch": 0.17173, + "grad_norm": 1.1061921435867776, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 17173 + }, + { + "epoch": 0.17174, + "grad_norm": 1.63429457533596, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 17174 + }, + { + "epoch": 0.17175, + "grad_norm": 1.1211868271676955, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17175 + }, + { + "epoch": 0.17176, + "grad_norm": 1.347355707028804, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17176 + }, + { + "epoch": 0.17177, + "grad_norm": 1.3774178524334308, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17177 + }, + { + "epoch": 0.17178, + "grad_norm": 1.293544172783026, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 17178 + }, + { + "epoch": 0.17179, + "grad_norm": 1.4128510891275874, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 17179 + }, + { + "epoch": 0.1718, + "grad_norm": 1.1957458135373926, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 17180 + }, + { + "epoch": 0.17181, + "grad_norm": 1.3046029433494455, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 17181 + }, + { + "epoch": 0.17182, + "grad_norm": 1.1799759959603753, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 17182 + }, + { + "epoch": 0.17183, + "grad_norm": 1.1758628162170435, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17183 + }, + { + "epoch": 0.17184, + "grad_norm": 1.341975268725816, + "learning_rate": 0.003, + "loss": 4.072, + "step": 17184 + }, + { + "epoch": 0.17185, + "grad_norm": 1.1445576596269915, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17185 + }, + { + "epoch": 0.17186, + "grad_norm": 1.3224603315483399, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 17186 + }, + { + "epoch": 0.17187, + "grad_norm": 1.2395826243783463, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 17187 + }, + { + "epoch": 0.17188, + "grad_norm": 1.2690006660452913, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 17188 + }, + { + "epoch": 0.17189, + "grad_norm": 1.2573542334999037, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 17189 + }, + { + "epoch": 0.1719, + "grad_norm": 1.2335073132847372, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17190 + }, + { + "epoch": 0.17191, + "grad_norm": 1.1932810784949115, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 17191 + }, + { + "epoch": 0.17192, + "grad_norm": 1.4304829083222084, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 17192 + }, + { + "epoch": 0.17193, + "grad_norm": 1.1989208824157573, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 17193 + }, + { + "epoch": 0.17194, + "grad_norm": 1.3947380618125669, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 17194 + }, + { + "epoch": 0.17195, + "grad_norm": 0.9527703239185515, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17195 + }, + { + "epoch": 0.17196, + "grad_norm": 1.342668531453168, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 17196 + }, + { + "epoch": 0.17197, + "grad_norm": 1.421452676097791, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 17197 + }, + { + "epoch": 0.17198, + "grad_norm": 1.5677601226221192, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 17198 + }, + { + "epoch": 0.17199, + "grad_norm": 0.899927555636912, + "learning_rate": 0.003, + "loss": 4.066, + "step": 17199 + }, + { + "epoch": 0.172, + "grad_norm": 1.1429881140819786, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17200 + }, + { + "epoch": 0.17201, + "grad_norm": 1.3964666203361211, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17201 + }, + { + "epoch": 0.17202, + "grad_norm": 1.0194327815481596, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 17202 + }, + { + "epoch": 0.17203, + "grad_norm": 1.4128801485233948, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 17203 + }, + { + "epoch": 0.17204, + "grad_norm": 1.1642083930368834, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 17204 + }, + { + "epoch": 0.17205, + "grad_norm": 1.3433710344584748, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 17205 + }, + { + "epoch": 0.17206, + "grad_norm": 1.1485054644045822, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 17206 + }, + { + "epoch": 0.17207, + "grad_norm": 1.3944503319228418, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 17207 + }, + { + "epoch": 0.17208, + "grad_norm": 1.547361725635885, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 17208 + }, + { + "epoch": 0.17209, + "grad_norm": 1.5347040469120532, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17209 + }, + { + "epoch": 0.1721, + "grad_norm": 1.2648440376927654, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 17210 + }, + { + "epoch": 0.17211, + "grad_norm": 1.2446087587338501, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 17211 + }, + { + "epoch": 0.17212, + "grad_norm": 1.1910938246129608, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 17212 + }, + { + "epoch": 0.17213, + "grad_norm": 1.4170422029493879, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 17213 + }, + { + "epoch": 0.17214, + "grad_norm": 1.230923562426313, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17214 + }, + { + "epoch": 0.17215, + "grad_norm": 1.3215117269643324, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17215 + }, + { + "epoch": 0.17216, + "grad_norm": 1.332945031735043, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 17216 + }, + { + "epoch": 0.17217, + "grad_norm": 1.1035955522319827, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17217 + }, + { + "epoch": 0.17218, + "grad_norm": 1.428788610850809, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17218 + }, + { + "epoch": 0.17219, + "grad_norm": 1.3068129180189931, + "learning_rate": 0.003, + "loss": 4.04, + "step": 17219 + }, + { + "epoch": 0.1722, + "grad_norm": 1.0638078684259753, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17220 + }, + { + "epoch": 0.17221, + "grad_norm": 1.496968304741232, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17221 + }, + { + "epoch": 0.17222, + "grad_norm": 1.1011271093479307, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17222 + }, + { + "epoch": 0.17223, + "grad_norm": 1.3512456024019075, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 17223 + }, + { + "epoch": 0.17224, + "grad_norm": 1.1531275939399306, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 17224 + }, + { + "epoch": 0.17225, + "grad_norm": 1.4043274764932459, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17225 + }, + { + "epoch": 0.17226, + "grad_norm": 1.1798686365718773, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 17226 + }, + { + "epoch": 0.17227, + "grad_norm": 1.4155883410107954, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17227 + }, + { + "epoch": 0.17228, + "grad_norm": 1.261911184142582, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 17228 + }, + { + "epoch": 0.17229, + "grad_norm": 1.5439118749060368, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 17229 + }, + { + "epoch": 0.1723, + "grad_norm": 1.5607439747288108, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17230 + }, + { + "epoch": 0.17231, + "grad_norm": 1.1242778683582726, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 17231 + }, + { + "epoch": 0.17232, + "grad_norm": 1.5294082757902303, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 17232 + }, + { + "epoch": 0.17233, + "grad_norm": 1.244232824073628, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 17233 + }, + { + "epoch": 0.17234, + "grad_norm": 1.464765045433406, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 17234 + }, + { + "epoch": 0.17235, + "grad_norm": 1.212812122055719, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 17235 + }, + { + "epoch": 0.17236, + "grad_norm": 1.5073516799941888, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 17236 + }, + { + "epoch": 0.17237, + "grad_norm": 0.9852482225652126, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 17237 + }, + { + "epoch": 0.17238, + "grad_norm": 1.2345978313177626, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 17238 + }, + { + "epoch": 0.17239, + "grad_norm": 1.3266079466508103, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17239 + }, + { + "epoch": 0.1724, + "grad_norm": 1.2165704381666227, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17240 + }, + { + "epoch": 0.17241, + "grad_norm": 1.5382431048618486, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 17241 + }, + { + "epoch": 0.17242, + "grad_norm": 1.1256669061556397, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17242 + }, + { + "epoch": 0.17243, + "grad_norm": 1.3371283903340336, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17243 + }, + { + "epoch": 0.17244, + "grad_norm": 1.0856568234996853, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 17244 + }, + { + "epoch": 0.17245, + "grad_norm": 1.250163433629188, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 17245 + }, + { + "epoch": 0.17246, + "grad_norm": 1.4700541215967489, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 17246 + }, + { + "epoch": 0.17247, + "grad_norm": 1.2752347955260197, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17247 + }, + { + "epoch": 0.17248, + "grad_norm": 1.4364732398721989, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 17248 + }, + { + "epoch": 0.17249, + "grad_norm": 1.1495637776253087, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 17249 + }, + { + "epoch": 0.1725, + "grad_norm": 1.583303890796779, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 17250 + }, + { + "epoch": 0.17251, + "grad_norm": 1.1828008978179465, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 17251 + }, + { + "epoch": 0.17252, + "grad_norm": 1.1962184866477457, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 17252 + }, + { + "epoch": 0.17253, + "grad_norm": 1.2827090303428514, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 17253 + }, + { + "epoch": 0.17254, + "grad_norm": 1.3991216810491618, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 17254 + }, + { + "epoch": 0.17255, + "grad_norm": 1.5051394917977816, + "learning_rate": 0.003, + "loss": 4.061, + "step": 17255 + }, + { + "epoch": 0.17256, + "grad_norm": 1.1584665892286088, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 17256 + }, + { + "epoch": 0.17257, + "grad_norm": 1.300102958418158, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 17257 + }, + { + "epoch": 0.17258, + "grad_norm": 1.1476915110738537, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17258 + }, + { + "epoch": 0.17259, + "grad_norm": 1.314246452505438, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 17259 + }, + { + "epoch": 0.1726, + "grad_norm": 1.271661473550511, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 17260 + }, + { + "epoch": 0.17261, + "grad_norm": 1.3326567236325095, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 17261 + }, + { + "epoch": 0.17262, + "grad_norm": 1.1439022006444373, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 17262 + }, + { + "epoch": 0.17263, + "grad_norm": 1.2667710201126365, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 17263 + }, + { + "epoch": 0.17264, + "grad_norm": 1.2758170744510648, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 17264 + }, + { + "epoch": 0.17265, + "grad_norm": 1.0010394942517953, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 17265 + }, + { + "epoch": 0.17266, + "grad_norm": 1.571265103363544, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 17266 + }, + { + "epoch": 0.17267, + "grad_norm": 1.2650034695598296, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 17267 + }, + { + "epoch": 0.17268, + "grad_norm": 1.0913091722989166, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 17268 + }, + { + "epoch": 0.17269, + "grad_norm": 1.5213843839037597, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17269 + }, + { + "epoch": 0.1727, + "grad_norm": 1.3113409721606066, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 17270 + }, + { + "epoch": 0.17271, + "grad_norm": 1.3051990704978562, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 17271 + }, + { + "epoch": 0.17272, + "grad_norm": 1.205618793219367, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 17272 + }, + { + "epoch": 0.17273, + "grad_norm": 1.1300313604332728, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 17273 + }, + { + "epoch": 0.17274, + "grad_norm": 1.1243566938219207, + "learning_rate": 0.003, + "loss": 4.071, + "step": 17274 + }, + { + "epoch": 0.17275, + "grad_norm": 1.112047280737149, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 17275 + }, + { + "epoch": 0.17276, + "grad_norm": 1.8547947524290764, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17276 + }, + { + "epoch": 0.17277, + "grad_norm": 1.1514749606092396, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 17277 + }, + { + "epoch": 0.17278, + "grad_norm": 1.1727422654093491, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17278 + }, + { + "epoch": 0.17279, + "grad_norm": 1.3751968310876361, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17279 + }, + { + "epoch": 0.1728, + "grad_norm": 1.3185583131052647, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17280 + }, + { + "epoch": 0.17281, + "grad_norm": 1.175579160659979, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 17281 + }, + { + "epoch": 0.17282, + "grad_norm": 1.3651433359056662, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17282 + }, + { + "epoch": 0.17283, + "grad_norm": 1.202724664414615, + "learning_rate": 0.003, + "loss": 4.064, + "step": 17283 + }, + { + "epoch": 0.17284, + "grad_norm": 1.601234921452785, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17284 + }, + { + "epoch": 0.17285, + "grad_norm": 0.9875939541340876, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 17285 + }, + { + "epoch": 0.17286, + "grad_norm": 1.280664764636334, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 17286 + }, + { + "epoch": 0.17287, + "grad_norm": 1.3142995598924436, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 17287 + }, + { + "epoch": 0.17288, + "grad_norm": 1.3170755852035023, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 17288 + }, + { + "epoch": 0.17289, + "grad_norm": 1.3429507399872127, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17289 + }, + { + "epoch": 0.1729, + "grad_norm": 1.1784211039042498, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17290 + }, + { + "epoch": 0.17291, + "grad_norm": 1.4020553869097763, + "learning_rate": 0.003, + "loss": 4.061, + "step": 17291 + }, + { + "epoch": 0.17292, + "grad_norm": 1.3454413871152973, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17292 + }, + { + "epoch": 0.17293, + "grad_norm": 1.3043302180960887, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 17293 + }, + { + "epoch": 0.17294, + "grad_norm": 1.3963117103349811, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 17294 + }, + { + "epoch": 0.17295, + "grad_norm": 1.1466282181055694, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 17295 + }, + { + "epoch": 0.17296, + "grad_norm": 1.540582528810366, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 17296 + }, + { + "epoch": 0.17297, + "grad_norm": 1.4338787431328484, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 17297 + }, + { + "epoch": 0.17298, + "grad_norm": 1.1298772161127753, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 17298 + }, + { + "epoch": 0.17299, + "grad_norm": 1.364466493792987, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 17299 + }, + { + "epoch": 0.173, + "grad_norm": 1.2171701816028715, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 17300 + }, + { + "epoch": 0.17301, + "grad_norm": 1.1607674163362474, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 17301 + }, + { + "epoch": 0.17302, + "grad_norm": 1.325897453697002, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17302 + }, + { + "epoch": 0.17303, + "grad_norm": 1.15890586726132, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17303 + }, + { + "epoch": 0.17304, + "grad_norm": 1.2952615224255417, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 17304 + }, + { + "epoch": 0.17305, + "grad_norm": 1.140242946901405, + "learning_rate": 0.003, + "loss": 4.041, + "step": 17305 + }, + { + "epoch": 0.17306, + "grad_norm": 1.2200693633923876, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 17306 + }, + { + "epoch": 0.17307, + "grad_norm": 1.2683280442429594, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 17307 + }, + { + "epoch": 0.17308, + "grad_norm": 1.2786029341686145, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 17308 + }, + { + "epoch": 0.17309, + "grad_norm": 1.1671240380588164, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 17309 + }, + { + "epoch": 0.1731, + "grad_norm": 1.3699267211144008, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 17310 + }, + { + "epoch": 0.17311, + "grad_norm": 1.1161068521276887, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 17311 + }, + { + "epoch": 0.17312, + "grad_norm": 1.3884575625854214, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 17312 + }, + { + "epoch": 0.17313, + "grad_norm": 1.2709628688699923, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 17313 + }, + { + "epoch": 0.17314, + "grad_norm": 1.2587021165172843, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 17314 + }, + { + "epoch": 0.17315, + "grad_norm": 1.482888867163177, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 17315 + }, + { + "epoch": 0.17316, + "grad_norm": 1.314590397860297, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 17316 + }, + { + "epoch": 0.17317, + "grad_norm": 1.3136500728936051, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17317 + }, + { + "epoch": 0.17318, + "grad_norm": 1.3615270441333711, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 17318 + }, + { + "epoch": 0.17319, + "grad_norm": 1.3977904928288472, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 17319 + }, + { + "epoch": 0.1732, + "grad_norm": 1.4946716511230473, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17320 + }, + { + "epoch": 0.17321, + "grad_norm": 1.1554212838199085, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 17321 + }, + { + "epoch": 0.17322, + "grad_norm": 1.1850770640871187, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 17322 + }, + { + "epoch": 0.17323, + "grad_norm": 1.286672324486612, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 17323 + }, + { + "epoch": 0.17324, + "grad_norm": 1.233534924738266, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17324 + }, + { + "epoch": 0.17325, + "grad_norm": 1.3122657038020409, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 17325 + }, + { + "epoch": 0.17326, + "grad_norm": 1.3013893089950082, + "learning_rate": 0.003, + "loss": 4.047, + "step": 17326 + }, + { + "epoch": 0.17327, + "grad_norm": 1.7277467216767701, + "learning_rate": 0.003, + "loss": 4.097, + "step": 17327 + }, + { + "epoch": 0.17328, + "grad_norm": 1.1821787172337874, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17328 + }, + { + "epoch": 0.17329, + "grad_norm": 1.6527006005088747, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 17329 + }, + { + "epoch": 0.1733, + "grad_norm": 1.1597089388958801, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17330 + }, + { + "epoch": 0.17331, + "grad_norm": 1.160774057476352, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 17331 + }, + { + "epoch": 0.17332, + "grad_norm": 1.2581431200567026, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 17332 + }, + { + "epoch": 0.17333, + "grad_norm": 1.3622254185376, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 17333 + }, + { + "epoch": 0.17334, + "grad_norm": 1.359613278225029, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 17334 + }, + { + "epoch": 0.17335, + "grad_norm": 1.0871508787353932, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 17335 + }, + { + "epoch": 0.17336, + "grad_norm": 1.476010253662266, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 17336 + }, + { + "epoch": 0.17337, + "grad_norm": 1.1018459832652663, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 17337 + }, + { + "epoch": 0.17338, + "grad_norm": 1.6347777241101717, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 17338 + }, + { + "epoch": 0.17339, + "grad_norm": 1.3550539222364155, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 17339 + }, + { + "epoch": 0.1734, + "grad_norm": 1.2307412658433872, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 17340 + }, + { + "epoch": 0.17341, + "grad_norm": 1.3116085991874384, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 17341 + }, + { + "epoch": 0.17342, + "grad_norm": 1.1839396795080575, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17342 + }, + { + "epoch": 0.17343, + "grad_norm": 1.5878123562511406, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 17343 + }, + { + "epoch": 0.17344, + "grad_norm": 0.966311302374749, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17344 + }, + { + "epoch": 0.17345, + "grad_norm": 1.4614186302212233, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 17345 + }, + { + "epoch": 0.17346, + "grad_norm": 1.5076469202916865, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 17346 + }, + { + "epoch": 0.17347, + "grad_norm": 0.9372154908707657, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 17347 + }, + { + "epoch": 0.17348, + "grad_norm": 1.2906133943823377, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 17348 + }, + { + "epoch": 0.17349, + "grad_norm": 1.3538707872254896, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17349 + }, + { + "epoch": 0.1735, + "grad_norm": 1.3238868446804868, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17350 + }, + { + "epoch": 0.17351, + "grad_norm": 1.1704739066405059, + "learning_rate": 0.003, + "loss": 4.061, + "step": 17351 + }, + { + "epoch": 0.17352, + "grad_norm": 1.2941053022753086, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17352 + }, + { + "epoch": 0.17353, + "grad_norm": 1.213494266625952, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17353 + }, + { + "epoch": 0.17354, + "grad_norm": 1.161557789178661, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17354 + }, + { + "epoch": 0.17355, + "grad_norm": 1.3474693954183323, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 17355 + }, + { + "epoch": 0.17356, + "grad_norm": 1.1487157146444444, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17356 + }, + { + "epoch": 0.17357, + "grad_norm": 1.5917371809698078, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 17357 + }, + { + "epoch": 0.17358, + "grad_norm": 0.9987014860498662, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 17358 + }, + { + "epoch": 0.17359, + "grad_norm": 1.6232130332722863, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 17359 + }, + { + "epoch": 0.1736, + "grad_norm": 1.1533563501761357, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 17360 + }, + { + "epoch": 0.17361, + "grad_norm": 1.2258117354232874, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17361 + }, + { + "epoch": 0.17362, + "grad_norm": 1.3591228343764234, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17362 + }, + { + "epoch": 0.17363, + "grad_norm": 1.1575402182752415, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 17363 + }, + { + "epoch": 0.17364, + "grad_norm": 1.2584198090000192, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 17364 + }, + { + "epoch": 0.17365, + "grad_norm": 1.217932782500489, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17365 + }, + { + "epoch": 0.17366, + "grad_norm": 1.6319796303181, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17366 + }, + { + "epoch": 0.17367, + "grad_norm": 1.2496362335876412, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17367 + }, + { + "epoch": 0.17368, + "grad_norm": 1.3543288489384762, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 17368 + }, + { + "epoch": 0.17369, + "grad_norm": 1.227748208441705, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 17369 + }, + { + "epoch": 0.1737, + "grad_norm": 1.2128552652264637, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 17370 + }, + { + "epoch": 0.17371, + "grad_norm": 1.3151813991304573, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17371 + }, + { + "epoch": 0.17372, + "grad_norm": 1.2675731600786122, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 17372 + }, + { + "epoch": 0.17373, + "grad_norm": 1.1910753012986075, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 17373 + }, + { + "epoch": 0.17374, + "grad_norm": 1.6257728688556055, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 17374 + }, + { + "epoch": 0.17375, + "grad_norm": 1.0985533268317234, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17375 + }, + { + "epoch": 0.17376, + "grad_norm": 1.4441899532090265, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17376 + }, + { + "epoch": 0.17377, + "grad_norm": 1.1473629283807807, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17377 + }, + { + "epoch": 0.17378, + "grad_norm": 1.2819491385872488, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 17378 + }, + { + "epoch": 0.17379, + "grad_norm": 1.4537829518027603, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 17379 + }, + { + "epoch": 0.1738, + "grad_norm": 1.3293515545404049, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 17380 + }, + { + "epoch": 0.17381, + "grad_norm": 1.7194165571937199, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 17381 + }, + { + "epoch": 0.17382, + "grad_norm": 1.0295087960436615, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 17382 + }, + { + "epoch": 0.17383, + "grad_norm": 1.1044202128867775, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 17383 + }, + { + "epoch": 0.17384, + "grad_norm": 1.2696710055506293, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 17384 + }, + { + "epoch": 0.17385, + "grad_norm": 1.4853913519470598, + "learning_rate": 0.003, + "loss": 4.061, + "step": 17385 + }, + { + "epoch": 0.17386, + "grad_norm": 1.2943715520245132, + "learning_rate": 0.003, + "loss": 4.065, + "step": 17386 + }, + { + "epoch": 0.17387, + "grad_norm": 1.155838567800402, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 17387 + }, + { + "epoch": 0.17388, + "grad_norm": 1.5429276307218556, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 17388 + }, + { + "epoch": 0.17389, + "grad_norm": 1.2413990783254136, + "learning_rate": 0.003, + "loss": 4.065, + "step": 17389 + }, + { + "epoch": 0.1739, + "grad_norm": 1.3477300495649724, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 17390 + }, + { + "epoch": 0.17391, + "grad_norm": 1.2977172852945469, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 17391 + }, + { + "epoch": 0.17392, + "grad_norm": 1.1374129030541535, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 17392 + }, + { + "epoch": 0.17393, + "grad_norm": 1.1607938224887557, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17393 + }, + { + "epoch": 0.17394, + "grad_norm": 1.1673671826581336, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17394 + }, + { + "epoch": 0.17395, + "grad_norm": 1.1546886884764822, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 17395 + }, + { + "epoch": 0.17396, + "grad_norm": 1.3656784042418233, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 17396 + }, + { + "epoch": 0.17397, + "grad_norm": 0.8135727371113464, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 17397 + }, + { + "epoch": 0.17398, + "grad_norm": 1.137411890639474, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 17398 + }, + { + "epoch": 0.17399, + "grad_norm": 1.6386060306451693, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 17399 + }, + { + "epoch": 0.174, + "grad_norm": 1.021912059527306, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17400 + }, + { + "epoch": 0.17401, + "grad_norm": 1.6368228150526618, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 17401 + }, + { + "epoch": 0.17402, + "grad_norm": 1.3096375352568765, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17402 + }, + { + "epoch": 0.17403, + "grad_norm": 1.15501855236692, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 17403 + }, + { + "epoch": 0.17404, + "grad_norm": 1.5085837764292096, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 17404 + }, + { + "epoch": 0.17405, + "grad_norm": 1.143097257206991, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17405 + }, + { + "epoch": 0.17406, + "grad_norm": 1.3487692828366185, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 17406 + }, + { + "epoch": 0.17407, + "grad_norm": 1.0810910518682053, + "learning_rate": 0.003, + "loss": 4.038, + "step": 17407 + }, + { + "epoch": 0.17408, + "grad_norm": 1.2778119740052092, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 17408 + }, + { + "epoch": 0.17409, + "grad_norm": 1.1101934840980172, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 17409 + }, + { + "epoch": 0.1741, + "grad_norm": 1.1874662189268035, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17410 + }, + { + "epoch": 0.17411, + "grad_norm": 1.5161881786934777, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 17411 + }, + { + "epoch": 0.17412, + "grad_norm": 1.4253812994933417, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 17412 + }, + { + "epoch": 0.17413, + "grad_norm": 1.5970659377886456, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 17413 + }, + { + "epoch": 0.17414, + "grad_norm": 1.1977208048904462, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17414 + }, + { + "epoch": 0.17415, + "grad_norm": 1.2466475381012039, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 17415 + }, + { + "epoch": 0.17416, + "grad_norm": 1.4494269543884284, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17416 + }, + { + "epoch": 0.17417, + "grad_norm": 1.304335207124149, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17417 + }, + { + "epoch": 0.17418, + "grad_norm": 1.0728577346961645, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 17418 + }, + { + "epoch": 0.17419, + "grad_norm": 1.5547921561653366, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17419 + }, + { + "epoch": 0.1742, + "grad_norm": 1.1010696107279216, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 17420 + }, + { + "epoch": 0.17421, + "grad_norm": 1.1219877424315068, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 17421 + }, + { + "epoch": 0.17422, + "grad_norm": 1.241923058210318, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 17422 + }, + { + "epoch": 0.17423, + "grad_norm": 1.4826931444793456, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 17423 + }, + { + "epoch": 0.17424, + "grad_norm": 1.2285170108606516, + "learning_rate": 0.003, + "loss": 4.065, + "step": 17424 + }, + { + "epoch": 0.17425, + "grad_norm": 1.181199300063592, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 17425 + }, + { + "epoch": 0.17426, + "grad_norm": 1.3961143890285066, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17426 + }, + { + "epoch": 0.17427, + "grad_norm": 1.173246971106792, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17427 + }, + { + "epoch": 0.17428, + "grad_norm": 1.3700077212676505, + "learning_rate": 0.003, + "loss": 4.043, + "step": 17428 + }, + { + "epoch": 0.17429, + "grad_norm": 1.3895862890260624, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 17429 + }, + { + "epoch": 0.1743, + "grad_norm": 1.292937448029352, + "learning_rate": 0.003, + "loss": 4.036, + "step": 17430 + }, + { + "epoch": 0.17431, + "grad_norm": 1.3286655907704719, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17431 + }, + { + "epoch": 0.17432, + "grad_norm": 1.2366878735177564, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 17432 + }, + { + "epoch": 0.17433, + "grad_norm": 1.2354096916732125, + "learning_rate": 0.003, + "loss": 4.066, + "step": 17433 + }, + { + "epoch": 0.17434, + "grad_norm": 1.2531275227176735, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 17434 + }, + { + "epoch": 0.17435, + "grad_norm": 1.1650514653874386, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 17435 + }, + { + "epoch": 0.17436, + "grad_norm": 1.2550299354752033, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 17436 + }, + { + "epoch": 0.17437, + "grad_norm": 1.3582554391865993, + "learning_rate": 0.003, + "loss": 4.051, + "step": 17437 + }, + { + "epoch": 0.17438, + "grad_norm": 1.3617902568728197, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 17438 + }, + { + "epoch": 0.17439, + "grad_norm": 1.3004396587625653, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 17439 + }, + { + "epoch": 0.1744, + "grad_norm": 1.2032476227166182, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17440 + }, + { + "epoch": 0.17441, + "grad_norm": 1.2918767594577591, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 17441 + }, + { + "epoch": 0.17442, + "grad_norm": 1.2137929756564796, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 17442 + }, + { + "epoch": 0.17443, + "grad_norm": 1.401955872786344, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 17443 + }, + { + "epoch": 0.17444, + "grad_norm": 1.5254161669212798, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 17444 + }, + { + "epoch": 0.17445, + "grad_norm": 1.0197218411898092, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 17445 + }, + { + "epoch": 0.17446, + "grad_norm": 1.4127761657383924, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17446 + }, + { + "epoch": 0.17447, + "grad_norm": 1.3330025253011795, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17447 + }, + { + "epoch": 0.17448, + "grad_norm": 1.2495163571886592, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 17448 + }, + { + "epoch": 0.17449, + "grad_norm": 1.4294119441106288, + "learning_rate": 0.003, + "loss": 4.073, + "step": 17449 + }, + { + "epoch": 0.1745, + "grad_norm": 1.0491530867649907, + "learning_rate": 0.003, + "loss": 4.055, + "step": 17450 + }, + { + "epoch": 0.17451, + "grad_norm": 1.5565773989464897, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 17451 + }, + { + "epoch": 0.17452, + "grad_norm": 1.2571825790237903, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 17452 + }, + { + "epoch": 0.17453, + "grad_norm": 1.4828859884751822, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 17453 + }, + { + "epoch": 0.17454, + "grad_norm": 1.2956449750709225, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 17454 + }, + { + "epoch": 0.17455, + "grad_norm": 1.243516393411184, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 17455 + }, + { + "epoch": 0.17456, + "grad_norm": 1.319861010512668, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 17456 + }, + { + "epoch": 0.17457, + "grad_norm": 1.2945272589556633, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 17457 + }, + { + "epoch": 0.17458, + "grad_norm": 1.2834252737184408, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17458 + }, + { + "epoch": 0.17459, + "grad_norm": 1.517112472998707, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17459 + }, + { + "epoch": 0.1746, + "grad_norm": 1.0806886971250516, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 17460 + }, + { + "epoch": 0.17461, + "grad_norm": 1.3459936523173737, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17461 + }, + { + "epoch": 0.17462, + "grad_norm": 1.1351068594804266, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 17462 + }, + { + "epoch": 0.17463, + "grad_norm": 1.3315162335300437, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 17463 + }, + { + "epoch": 0.17464, + "grad_norm": 1.1421814018137226, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 17464 + }, + { + "epoch": 0.17465, + "grad_norm": 1.4720298962021945, + "learning_rate": 0.003, + "loss": 4.062, + "step": 17465 + }, + { + "epoch": 0.17466, + "grad_norm": 1.0970045489403935, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 17466 + }, + { + "epoch": 0.17467, + "grad_norm": 1.4051446872244608, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 17467 + }, + { + "epoch": 0.17468, + "grad_norm": 1.1788754409907096, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 17468 + }, + { + "epoch": 0.17469, + "grad_norm": 1.4085198117148778, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17469 + }, + { + "epoch": 0.1747, + "grad_norm": 1.0656980683755632, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 17470 + }, + { + "epoch": 0.17471, + "grad_norm": 1.3596266943541238, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 17471 + }, + { + "epoch": 0.17472, + "grad_norm": 1.4360775955395388, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 17472 + }, + { + "epoch": 0.17473, + "grad_norm": 1.1815663699829684, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17473 + }, + { + "epoch": 0.17474, + "grad_norm": 1.3773290285609925, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 17474 + }, + { + "epoch": 0.17475, + "grad_norm": 1.138640414247468, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 17475 + }, + { + "epoch": 0.17476, + "grad_norm": 1.4774328139223307, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17476 + }, + { + "epoch": 0.17477, + "grad_norm": 1.1258013640931122, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 17477 + }, + { + "epoch": 0.17478, + "grad_norm": 1.445371284232178, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 17478 + }, + { + "epoch": 0.17479, + "grad_norm": 1.0627794731721976, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17479 + }, + { + "epoch": 0.1748, + "grad_norm": 1.2629567251829825, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17480 + }, + { + "epoch": 0.17481, + "grad_norm": 1.162982937463057, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 17481 + }, + { + "epoch": 0.17482, + "grad_norm": 1.2109852386399709, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 17482 + }, + { + "epoch": 0.17483, + "grad_norm": 1.351926763850564, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17483 + }, + { + "epoch": 0.17484, + "grad_norm": 1.2675220236375762, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 17484 + }, + { + "epoch": 0.17485, + "grad_norm": 1.448801081514365, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17485 + }, + { + "epoch": 0.17486, + "grad_norm": 1.470879245043995, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 17486 + }, + { + "epoch": 0.17487, + "grad_norm": 1.5235474751119706, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 17487 + }, + { + "epoch": 0.17488, + "grad_norm": 1.2882001150246338, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17488 + }, + { + "epoch": 0.17489, + "grad_norm": 1.1262043232744685, + "learning_rate": 0.003, + "loss": 4.042, + "step": 17489 + }, + { + "epoch": 0.1749, + "grad_norm": 1.2529971851246242, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 17490 + }, + { + "epoch": 0.17491, + "grad_norm": 1.3515122585543957, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 17491 + }, + { + "epoch": 0.17492, + "grad_norm": 1.1888181326490819, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 17492 + }, + { + "epoch": 0.17493, + "grad_norm": 1.269549654881748, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 17493 + }, + { + "epoch": 0.17494, + "grad_norm": 1.399664853120194, + "learning_rate": 0.003, + "loss": 4.053, + "step": 17494 + }, + { + "epoch": 0.17495, + "grad_norm": 1.0547626195049762, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 17495 + }, + { + "epoch": 0.17496, + "grad_norm": 1.2589039501288364, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17496 + }, + { + "epoch": 0.17497, + "grad_norm": 1.2087009396582926, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 17497 + }, + { + "epoch": 0.17498, + "grad_norm": 1.5111210585816903, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 17498 + }, + { + "epoch": 0.17499, + "grad_norm": 1.1932866062692538, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 17499 + }, + { + "epoch": 0.175, + "grad_norm": 1.3685368613552835, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 17500 + }, + { + "epoch": 0.17501, + "grad_norm": 1.4259266355444051, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 17501 + }, + { + "epoch": 0.17502, + "grad_norm": 1.0811465802849207, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 17502 + }, + { + "epoch": 0.17503, + "grad_norm": 1.3631121796923622, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 17503 + }, + { + "epoch": 0.17504, + "grad_norm": 1.3601765579398404, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 17504 + }, + { + "epoch": 0.17505, + "grad_norm": 1.3204992282929688, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 17505 + }, + { + "epoch": 0.17506, + "grad_norm": 1.6043420025523707, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 17506 + }, + { + "epoch": 0.17507, + "grad_norm": 1.1102090085261342, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 17507 + }, + { + "epoch": 0.17508, + "grad_norm": 1.3589012557270086, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 17508 + }, + { + "epoch": 0.17509, + "grad_norm": 1.4570648815138552, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17509 + }, + { + "epoch": 0.1751, + "grad_norm": 1.2379406823564174, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17510 + }, + { + "epoch": 0.17511, + "grad_norm": 1.2745485508345238, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17511 + }, + { + "epoch": 0.17512, + "grad_norm": 1.458311323129371, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 17512 + }, + { + "epoch": 0.17513, + "grad_norm": 1.2523363294551801, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 17513 + }, + { + "epoch": 0.17514, + "grad_norm": 1.326353253061568, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 17514 + }, + { + "epoch": 0.17515, + "grad_norm": 1.2078497675630715, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 17515 + }, + { + "epoch": 0.17516, + "grad_norm": 1.332111315017004, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 17516 + }, + { + "epoch": 0.17517, + "grad_norm": 1.4681129568264164, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 17517 + }, + { + "epoch": 0.17518, + "grad_norm": 0.9898792924352027, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 17518 + }, + { + "epoch": 0.17519, + "grad_norm": 1.5452090908593286, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 17519 + }, + { + "epoch": 0.1752, + "grad_norm": 1.027115202082818, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 17520 + }, + { + "epoch": 0.17521, + "grad_norm": 1.5480272088455356, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 17521 + }, + { + "epoch": 0.17522, + "grad_norm": 1.0872498679033396, + "learning_rate": 0.003, + "loss": 4.049, + "step": 17522 + }, + { + "epoch": 0.17523, + "grad_norm": 1.4579650844831544, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 17523 + }, + { + "epoch": 0.17524, + "grad_norm": 1.3716290946186658, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17524 + }, + { + "epoch": 0.17525, + "grad_norm": 1.174774209157617, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17525 + }, + { + "epoch": 0.17526, + "grad_norm": 1.3772871583425932, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 17526 + }, + { + "epoch": 0.17527, + "grad_norm": 1.34133868900741, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 17527 + }, + { + "epoch": 0.17528, + "grad_norm": 1.089014538515785, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 17528 + }, + { + "epoch": 0.17529, + "grad_norm": 1.4197266927363068, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 17529 + }, + { + "epoch": 0.1753, + "grad_norm": 1.0939073308060439, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17530 + }, + { + "epoch": 0.17531, + "grad_norm": 1.2336453874907791, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 17531 + }, + { + "epoch": 0.17532, + "grad_norm": 1.1479276481511065, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17532 + }, + { + "epoch": 0.17533, + "grad_norm": 1.2202607669748253, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 17533 + }, + { + "epoch": 0.17534, + "grad_norm": 1.4945469580763797, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 17534 + }, + { + "epoch": 0.17535, + "grad_norm": 1.6356180536776386, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 17535 + }, + { + "epoch": 0.17536, + "grad_norm": 1.3521367342384314, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 17536 + }, + { + "epoch": 0.17537, + "grad_norm": 1.3497434493622424, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17537 + }, + { + "epoch": 0.17538, + "grad_norm": 1.1735820423921637, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 17538 + }, + { + "epoch": 0.17539, + "grad_norm": 1.3290898846115236, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 17539 + }, + { + "epoch": 0.1754, + "grad_norm": 1.1762529483557118, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 17540 + }, + { + "epoch": 0.17541, + "grad_norm": 1.2369001566462126, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 17541 + }, + { + "epoch": 0.17542, + "grad_norm": 1.253493920129149, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 17542 + }, + { + "epoch": 0.17543, + "grad_norm": 1.1991541022277417, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 17543 + }, + { + "epoch": 0.17544, + "grad_norm": 1.6048053140008003, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 17544 + }, + { + "epoch": 0.17545, + "grad_norm": 1.080637588554762, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17545 + }, + { + "epoch": 0.17546, + "grad_norm": 1.6465218961255268, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 17546 + }, + { + "epoch": 0.17547, + "grad_norm": 1.2059376789891145, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17547 + }, + { + "epoch": 0.17548, + "grad_norm": 1.5033238233306159, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 17548 + }, + { + "epoch": 0.17549, + "grad_norm": 1.2903528075058992, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 17549 + }, + { + "epoch": 0.1755, + "grad_norm": 1.1825193838204116, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 17550 + }, + { + "epoch": 0.17551, + "grad_norm": 1.2435071476516741, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 17551 + }, + { + "epoch": 0.17552, + "grad_norm": 1.2351634262921993, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 17552 + }, + { + "epoch": 0.17553, + "grad_norm": 1.1599814224863219, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 17553 + }, + { + "epoch": 0.17554, + "grad_norm": 1.209609419739831, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 17554 + }, + { + "epoch": 0.17555, + "grad_norm": 1.221752704114831, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17555 + }, + { + "epoch": 0.17556, + "grad_norm": 1.7212103012472932, + "learning_rate": 0.003, + "loss": 4.082, + "step": 17556 + }, + { + "epoch": 0.17557, + "grad_norm": 1.3017911530063735, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17557 + }, + { + "epoch": 0.17558, + "grad_norm": 1.485709091804075, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 17558 + }, + { + "epoch": 0.17559, + "grad_norm": 1.3049037188814676, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 17559 + }, + { + "epoch": 0.1756, + "grad_norm": 1.2577989391627555, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 17560 + }, + { + "epoch": 0.17561, + "grad_norm": 1.376837404561379, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 17561 + }, + { + "epoch": 0.17562, + "grad_norm": 1.1633686470895166, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 17562 + }, + { + "epoch": 0.17563, + "grad_norm": 1.182559838540788, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 17563 + }, + { + "epoch": 0.17564, + "grad_norm": 1.3425096746959984, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17564 + }, + { + "epoch": 0.17565, + "grad_norm": 1.133752253129469, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17565 + }, + { + "epoch": 0.17566, + "grad_norm": 1.5169381551979604, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 17566 + }, + { + "epoch": 0.17567, + "grad_norm": 1.341228303024416, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 17567 + }, + { + "epoch": 0.17568, + "grad_norm": 1.1788620587516216, + "learning_rate": 0.003, + "loss": 4.045, + "step": 17568 + }, + { + "epoch": 0.17569, + "grad_norm": 1.155754217235722, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 17569 + }, + { + "epoch": 0.1757, + "grad_norm": 1.2557432658871737, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17570 + }, + { + "epoch": 0.17571, + "grad_norm": 1.2620674018967353, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17571 + }, + { + "epoch": 0.17572, + "grad_norm": 1.2905583879720335, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17572 + }, + { + "epoch": 0.17573, + "grad_norm": 1.275273313168324, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 17573 + }, + { + "epoch": 0.17574, + "grad_norm": 1.5488047698359892, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 17574 + }, + { + "epoch": 0.17575, + "grad_norm": 1.2913531469165127, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 17575 + }, + { + "epoch": 0.17576, + "grad_norm": 1.259871684253289, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 17576 + }, + { + "epoch": 0.17577, + "grad_norm": 1.2102978803530382, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17577 + }, + { + "epoch": 0.17578, + "grad_norm": 1.5210035489703877, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 17578 + }, + { + "epoch": 0.17579, + "grad_norm": 1.285636618378266, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 17579 + }, + { + "epoch": 0.1758, + "grad_norm": 1.392919694020842, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 17580 + }, + { + "epoch": 0.17581, + "grad_norm": 1.3840133366277634, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 17581 + }, + { + "epoch": 0.17582, + "grad_norm": 1.0871155767026495, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 17582 + }, + { + "epoch": 0.17583, + "grad_norm": 1.5645413061585471, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 17583 + }, + { + "epoch": 0.17584, + "grad_norm": 1.17321561979706, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 17584 + }, + { + "epoch": 0.17585, + "grad_norm": 1.3604032628155835, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 17585 + }, + { + "epoch": 0.17586, + "grad_norm": 1.314642451894998, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 17586 + }, + { + "epoch": 0.17587, + "grad_norm": 1.3291505728511441, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 17587 + }, + { + "epoch": 0.17588, + "grad_norm": 1.4376103152540465, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 17588 + }, + { + "epoch": 0.17589, + "grad_norm": 1.3347456845034162, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 17589 + }, + { + "epoch": 0.1759, + "grad_norm": 1.363165539937303, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 17590 + }, + { + "epoch": 0.17591, + "grad_norm": 1.1446605795386675, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 17591 + }, + { + "epoch": 0.17592, + "grad_norm": 1.39653778260535, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17592 + }, + { + "epoch": 0.17593, + "grad_norm": 1.1912570373914466, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17593 + }, + { + "epoch": 0.17594, + "grad_norm": 1.2643293160733424, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 17594 + }, + { + "epoch": 0.17595, + "grad_norm": 0.9890806388280463, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 17595 + }, + { + "epoch": 0.17596, + "grad_norm": 1.3817387821035372, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17596 + }, + { + "epoch": 0.17597, + "grad_norm": 1.13316950929552, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17597 + }, + { + "epoch": 0.17598, + "grad_norm": 1.7346344042620478, + "learning_rate": 0.003, + "loss": 4.102, + "step": 17598 + }, + { + "epoch": 0.17599, + "grad_norm": 1.1569002155868353, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 17599 + }, + { + "epoch": 0.176, + "grad_norm": 1.5841135639006447, + "learning_rate": 0.003, + "loss": 4.057, + "step": 17600 + }, + { + "epoch": 0.17601, + "grad_norm": 1.1700214162493932, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 17601 + }, + { + "epoch": 0.17602, + "grad_norm": 1.395655377413905, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17602 + }, + { + "epoch": 0.17603, + "grad_norm": 1.2600613047588067, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17603 + }, + { + "epoch": 0.17604, + "grad_norm": 1.485621223796701, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17604 + }, + { + "epoch": 0.17605, + "grad_norm": 1.0934046197953102, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 17605 + }, + { + "epoch": 0.17606, + "grad_norm": 1.6813061462712295, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 17606 + }, + { + "epoch": 0.17607, + "grad_norm": 1.1496380657690461, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17607 + }, + { + "epoch": 0.17608, + "grad_norm": 1.1253490677517515, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17608 + }, + { + "epoch": 0.17609, + "grad_norm": 1.2947829489405915, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17609 + }, + { + "epoch": 0.1761, + "grad_norm": 1.2533389268660708, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 17610 + }, + { + "epoch": 0.17611, + "grad_norm": 1.221058807420232, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17611 + }, + { + "epoch": 0.17612, + "grad_norm": 1.211577910041733, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17612 + }, + { + "epoch": 0.17613, + "grad_norm": 1.4963671852486207, + "learning_rate": 0.003, + "loss": 4.059, + "step": 17613 + }, + { + "epoch": 0.17614, + "grad_norm": 1.3621629216998155, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 17614 + }, + { + "epoch": 0.17615, + "grad_norm": 1.1745568317301551, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 17615 + }, + { + "epoch": 0.17616, + "grad_norm": 1.4917818846232451, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 17616 + }, + { + "epoch": 0.17617, + "grad_norm": 1.4688369182348506, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 17617 + }, + { + "epoch": 0.17618, + "grad_norm": 1.037151684296233, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 17618 + }, + { + "epoch": 0.17619, + "grad_norm": 1.4524502216009345, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 17619 + }, + { + "epoch": 0.1762, + "grad_norm": 1.0697082776084865, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17620 + }, + { + "epoch": 0.17621, + "grad_norm": 1.1625501217376397, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 17621 + }, + { + "epoch": 0.17622, + "grad_norm": 1.1469754723606305, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 17622 + }, + { + "epoch": 0.17623, + "grad_norm": 1.16858664998892, + "learning_rate": 0.003, + "loss": 4.06, + "step": 17623 + }, + { + "epoch": 0.17624, + "grad_norm": 1.3135664062689731, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 17624 + }, + { + "epoch": 0.17625, + "grad_norm": 1.6383406481826261, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17625 + }, + { + "epoch": 0.17626, + "grad_norm": 1.2545188782831724, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17626 + }, + { + "epoch": 0.17627, + "grad_norm": 1.2478847855408735, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 17627 + }, + { + "epoch": 0.17628, + "grad_norm": 1.2154447103562211, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 17628 + }, + { + "epoch": 0.17629, + "grad_norm": 1.3128872316814053, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 17629 + }, + { + "epoch": 0.1763, + "grad_norm": 1.2558917905084745, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 17630 + }, + { + "epoch": 0.17631, + "grad_norm": 1.56447314552511, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 17631 + }, + { + "epoch": 0.17632, + "grad_norm": 1.2288214495890235, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 17632 + }, + { + "epoch": 0.17633, + "grad_norm": 1.461485377819148, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 17633 + }, + { + "epoch": 0.17634, + "grad_norm": 1.4210982268515342, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 17634 + }, + { + "epoch": 0.17635, + "grad_norm": 1.6796500872140807, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 17635 + }, + { + "epoch": 0.17636, + "grad_norm": 1.1161604941055283, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17636 + }, + { + "epoch": 0.17637, + "grad_norm": 1.2688593912531774, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17637 + }, + { + "epoch": 0.17638, + "grad_norm": 1.2239390108100066, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 17638 + }, + { + "epoch": 0.17639, + "grad_norm": 1.4543016085223777, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17639 + }, + { + "epoch": 0.1764, + "grad_norm": 1.1311921748288387, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 17640 + }, + { + "epoch": 0.17641, + "grad_norm": 1.6222284430220124, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17641 + }, + { + "epoch": 0.17642, + "grad_norm": 1.1514824491992248, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 17642 + }, + { + "epoch": 0.17643, + "grad_norm": 1.4233439574319937, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 17643 + }, + { + "epoch": 0.17644, + "grad_norm": 1.3460528345168452, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 17644 + }, + { + "epoch": 0.17645, + "grad_norm": 1.1280697162130136, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 17645 + }, + { + "epoch": 0.17646, + "grad_norm": 1.3235388297884105, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17646 + }, + { + "epoch": 0.17647, + "grad_norm": 1.1302885823025965, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 17647 + }, + { + "epoch": 0.17648, + "grad_norm": 1.3439873433984786, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 17648 + }, + { + "epoch": 0.17649, + "grad_norm": 1.290515056093695, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 17649 + }, + { + "epoch": 0.1765, + "grad_norm": 1.4803167997075957, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 17650 + }, + { + "epoch": 0.17651, + "grad_norm": 1.1718919994994168, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 17651 + }, + { + "epoch": 0.17652, + "grad_norm": 1.3025434998217125, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 17652 + }, + { + "epoch": 0.17653, + "grad_norm": 1.1461619738250126, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 17653 + }, + { + "epoch": 0.17654, + "grad_norm": 1.2045934682269839, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17654 + }, + { + "epoch": 0.17655, + "grad_norm": 1.2600544087934396, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 17655 + }, + { + "epoch": 0.17656, + "grad_norm": 1.2993441694569487, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 17656 + }, + { + "epoch": 0.17657, + "grad_norm": 1.58132482735925, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17657 + }, + { + "epoch": 0.17658, + "grad_norm": 1.2216840641376203, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 17658 + }, + { + "epoch": 0.17659, + "grad_norm": 1.3101905106329146, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 17659 + }, + { + "epoch": 0.1766, + "grad_norm": 1.4183979395592572, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 17660 + }, + { + "epoch": 0.17661, + "grad_norm": 1.2422597475901402, + "learning_rate": 0.003, + "loss": 4.059, + "step": 17661 + }, + { + "epoch": 0.17662, + "grad_norm": 1.5392641833011527, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17662 + }, + { + "epoch": 0.17663, + "grad_norm": 0.9302020789539822, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 17663 + }, + { + "epoch": 0.17664, + "grad_norm": 1.352504587734451, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 17664 + }, + { + "epoch": 0.17665, + "grad_norm": 1.1405003858502711, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 17665 + }, + { + "epoch": 0.17666, + "grad_norm": 1.5675115628172653, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 17666 + }, + { + "epoch": 0.17667, + "grad_norm": 1.6971132785861853, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 17667 + }, + { + "epoch": 0.17668, + "grad_norm": 0.9969641889111963, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 17668 + }, + { + "epoch": 0.17669, + "grad_norm": 1.2909894956823855, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17669 + }, + { + "epoch": 0.1767, + "grad_norm": 1.7379145287328, + "learning_rate": 0.003, + "loss": 4.055, + "step": 17670 + }, + { + "epoch": 0.17671, + "grad_norm": 0.9627833816379531, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 17671 + }, + { + "epoch": 0.17672, + "grad_norm": 1.0898450590560478, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 17672 + }, + { + "epoch": 0.17673, + "grad_norm": 1.248478035290169, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 17673 + }, + { + "epoch": 0.17674, + "grad_norm": 1.3593375285238194, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 17674 + }, + { + "epoch": 0.17675, + "grad_norm": 1.3523789795245187, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 17675 + }, + { + "epoch": 0.17676, + "grad_norm": 1.1310773657139135, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 17676 + }, + { + "epoch": 0.17677, + "grad_norm": 1.4442942484105357, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17677 + }, + { + "epoch": 0.17678, + "grad_norm": 1.1969913176518006, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 17678 + }, + { + "epoch": 0.17679, + "grad_norm": 1.163658778160288, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 17679 + }, + { + "epoch": 0.1768, + "grad_norm": 1.1784312238863694, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 17680 + }, + { + "epoch": 0.17681, + "grad_norm": 1.4807219084128371, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 17681 + }, + { + "epoch": 0.17682, + "grad_norm": 1.3152398566596446, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 17682 + }, + { + "epoch": 0.17683, + "grad_norm": 1.3587381566270007, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 17683 + }, + { + "epoch": 0.17684, + "grad_norm": 1.4291855672899187, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 17684 + }, + { + "epoch": 0.17685, + "grad_norm": 1.299069865520517, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 17685 + }, + { + "epoch": 0.17686, + "grad_norm": 1.423640480642554, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 17686 + }, + { + "epoch": 0.17687, + "grad_norm": 1.0631995763892585, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 17687 + }, + { + "epoch": 0.17688, + "grad_norm": 1.434347203828858, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17688 + }, + { + "epoch": 0.17689, + "grad_norm": 1.3244904318921638, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17689 + }, + { + "epoch": 0.1769, + "grad_norm": 1.3999105356202497, + "learning_rate": 0.003, + "loss": 4.086, + "step": 17690 + }, + { + "epoch": 0.17691, + "grad_norm": 1.2456988748227091, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 17691 + }, + { + "epoch": 0.17692, + "grad_norm": 1.4356245225259319, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 17692 + }, + { + "epoch": 0.17693, + "grad_norm": 1.2895322006881962, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 17693 + }, + { + "epoch": 0.17694, + "grad_norm": 1.5517959275513893, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 17694 + }, + { + "epoch": 0.17695, + "grad_norm": 1.1354139058927026, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 17695 + }, + { + "epoch": 0.17696, + "grad_norm": 1.6630191899167326, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17696 + }, + { + "epoch": 0.17697, + "grad_norm": 1.067344928888524, + "learning_rate": 0.003, + "loss": 4.064, + "step": 17697 + }, + { + "epoch": 0.17698, + "grad_norm": 1.272823799545181, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 17698 + }, + { + "epoch": 0.17699, + "grad_norm": 1.4013209199376062, + "learning_rate": 0.003, + "loss": 4.058, + "step": 17699 + }, + { + "epoch": 0.177, + "grad_norm": 1.1742128784613832, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17700 + }, + { + "epoch": 0.17701, + "grad_norm": 1.2639450997363644, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 17701 + }, + { + "epoch": 0.17702, + "grad_norm": 1.3879760637351526, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 17702 + }, + { + "epoch": 0.17703, + "grad_norm": 1.1270593312215273, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 17703 + }, + { + "epoch": 0.17704, + "grad_norm": 1.3320730940804644, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 17704 + }, + { + "epoch": 0.17705, + "grad_norm": 1.280742530407232, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 17705 + }, + { + "epoch": 0.17706, + "grad_norm": 1.242758770219724, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 17706 + }, + { + "epoch": 0.17707, + "grad_norm": 1.106987527962389, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 17707 + }, + { + "epoch": 0.17708, + "grad_norm": 1.4048896644701068, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 17708 + }, + { + "epoch": 0.17709, + "grad_norm": 1.4607143220839507, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 17709 + }, + { + "epoch": 0.1771, + "grad_norm": 1.2458144205557176, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 17710 + }, + { + "epoch": 0.17711, + "grad_norm": 1.2852096577321068, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 17711 + }, + { + "epoch": 0.17712, + "grad_norm": 1.6120701938973683, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17712 + }, + { + "epoch": 0.17713, + "grad_norm": 1.1268323134544185, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17713 + }, + { + "epoch": 0.17714, + "grad_norm": 1.3765118008965018, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 17714 + }, + { + "epoch": 0.17715, + "grad_norm": 1.2223185178695093, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 17715 + }, + { + "epoch": 0.17716, + "grad_norm": 1.1942437661947458, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 17716 + }, + { + "epoch": 0.17717, + "grad_norm": 1.348931352784379, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 17717 + }, + { + "epoch": 0.17718, + "grad_norm": 0.9868431322378017, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 17718 + }, + { + "epoch": 0.17719, + "grad_norm": 1.506214835504423, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 17719 + }, + { + "epoch": 0.1772, + "grad_norm": 1.2017136329561915, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 17720 + }, + { + "epoch": 0.17721, + "grad_norm": 1.3727715141245602, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 17721 + }, + { + "epoch": 0.17722, + "grad_norm": 1.3819746634324437, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 17722 + }, + { + "epoch": 0.17723, + "grad_norm": 1.4971091150145805, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 17723 + }, + { + "epoch": 0.17724, + "grad_norm": 1.1448220379209995, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17724 + }, + { + "epoch": 0.17725, + "grad_norm": 1.3090864820025123, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 17725 + }, + { + "epoch": 0.17726, + "grad_norm": 1.416815157664905, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17726 + }, + { + "epoch": 0.17727, + "grad_norm": 1.3592864538826421, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 17727 + }, + { + "epoch": 0.17728, + "grad_norm": 1.181017492502079, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 17728 + }, + { + "epoch": 0.17729, + "grad_norm": 1.416290189676951, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17729 + }, + { + "epoch": 0.1773, + "grad_norm": 1.2872245898560697, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 17730 + }, + { + "epoch": 0.17731, + "grad_norm": 1.1322378090274587, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17731 + }, + { + "epoch": 0.17732, + "grad_norm": 1.4674360771390844, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 17732 + }, + { + "epoch": 0.17733, + "grad_norm": 1.3912855207092103, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 17733 + }, + { + "epoch": 0.17734, + "grad_norm": 1.375396481262623, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 17734 + }, + { + "epoch": 0.17735, + "grad_norm": 1.0862521365725795, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 17735 + }, + { + "epoch": 0.17736, + "grad_norm": 1.3900815444433379, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 17736 + }, + { + "epoch": 0.17737, + "grad_norm": 1.1711284664780262, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 17737 + }, + { + "epoch": 0.17738, + "grad_norm": 1.2425651281137042, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 17738 + }, + { + "epoch": 0.17739, + "grad_norm": 1.3059184746608743, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 17739 + }, + { + "epoch": 0.1774, + "grad_norm": 1.3011870231355265, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 17740 + }, + { + "epoch": 0.17741, + "grad_norm": 1.2535436653602747, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17741 + }, + { + "epoch": 0.17742, + "grad_norm": 1.372881049562413, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 17742 + }, + { + "epoch": 0.17743, + "grad_norm": 1.737441794190293, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 17743 + }, + { + "epoch": 0.17744, + "grad_norm": 1.3626963374777032, + "learning_rate": 0.003, + "loss": 4.051, + "step": 17744 + }, + { + "epoch": 0.17745, + "grad_norm": 1.133048682635455, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17745 + }, + { + "epoch": 0.17746, + "grad_norm": 1.5952755971146844, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 17746 + }, + { + "epoch": 0.17747, + "grad_norm": 1.1208706148409617, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 17747 + }, + { + "epoch": 0.17748, + "grad_norm": 1.4713550224114142, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 17748 + }, + { + "epoch": 0.17749, + "grad_norm": 1.1095058750339268, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 17749 + }, + { + "epoch": 0.1775, + "grad_norm": 1.2984318209968637, + "learning_rate": 0.003, + "loss": 4.064, + "step": 17750 + }, + { + "epoch": 0.17751, + "grad_norm": 1.4630332546250533, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 17751 + }, + { + "epoch": 0.17752, + "grad_norm": 1.1080067959121638, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 17752 + }, + { + "epoch": 0.17753, + "grad_norm": 1.1593676261260744, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 17753 + }, + { + "epoch": 0.17754, + "grad_norm": 1.2372894872821043, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 17754 + }, + { + "epoch": 0.17755, + "grad_norm": 1.258396401792125, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17755 + }, + { + "epoch": 0.17756, + "grad_norm": 1.0485209910890672, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 17756 + }, + { + "epoch": 0.17757, + "grad_norm": 1.462847751511984, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 17757 + }, + { + "epoch": 0.17758, + "grad_norm": 1.145811857167575, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 17758 + }, + { + "epoch": 0.17759, + "grad_norm": 1.6684204600306993, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 17759 + }, + { + "epoch": 0.1776, + "grad_norm": 1.2550512015522435, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 17760 + }, + { + "epoch": 0.17761, + "grad_norm": 1.3213866534609735, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17761 + }, + { + "epoch": 0.17762, + "grad_norm": 1.34841225621269, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 17762 + }, + { + "epoch": 0.17763, + "grad_norm": 1.4116761442066719, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17763 + }, + { + "epoch": 0.17764, + "grad_norm": 1.2992687258706004, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 17764 + }, + { + "epoch": 0.17765, + "grad_norm": 1.2747375822536242, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 17765 + }, + { + "epoch": 0.17766, + "grad_norm": 1.1038618602071801, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17766 + }, + { + "epoch": 0.17767, + "grad_norm": 1.4109619706095022, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 17767 + }, + { + "epoch": 0.17768, + "grad_norm": 1.2175900409790084, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 17768 + }, + { + "epoch": 0.17769, + "grad_norm": 1.2926353182322496, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 17769 + }, + { + "epoch": 0.1777, + "grad_norm": 1.243801243216156, + "learning_rate": 0.003, + "loss": 4.037, + "step": 17770 + }, + { + "epoch": 0.17771, + "grad_norm": 1.2368272296033915, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 17771 + }, + { + "epoch": 0.17772, + "grad_norm": 1.2199582482353482, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 17772 + }, + { + "epoch": 0.17773, + "grad_norm": 1.1402358760709792, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 17773 + }, + { + "epoch": 0.17774, + "grad_norm": 1.6157640077166686, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 17774 + }, + { + "epoch": 0.17775, + "grad_norm": 1.0227908232124825, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17775 + }, + { + "epoch": 0.17776, + "grad_norm": 1.4092199269502843, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 17776 + }, + { + "epoch": 0.17777, + "grad_norm": 1.2584085745914688, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 17777 + }, + { + "epoch": 0.17778, + "grad_norm": 1.4731279422782921, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17778 + }, + { + "epoch": 0.17779, + "grad_norm": 1.4075459756405722, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 17779 + }, + { + "epoch": 0.1778, + "grad_norm": 1.1224644816175051, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 17780 + }, + { + "epoch": 0.17781, + "grad_norm": 1.4278592701354988, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 17781 + }, + { + "epoch": 0.17782, + "grad_norm": 1.4387534531644868, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 17782 + }, + { + "epoch": 0.17783, + "grad_norm": 1.2692227755173564, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 17783 + }, + { + "epoch": 0.17784, + "grad_norm": 1.3682848068132636, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 17784 + }, + { + "epoch": 0.17785, + "grad_norm": 1.169166375284477, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 17785 + }, + { + "epoch": 0.17786, + "grad_norm": 1.3169217438370486, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 17786 + }, + { + "epoch": 0.17787, + "grad_norm": 1.1376413048380454, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17787 + }, + { + "epoch": 0.17788, + "grad_norm": 1.7124592978411066, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 17788 + }, + { + "epoch": 0.17789, + "grad_norm": 1.2745510693179796, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17789 + }, + { + "epoch": 0.1779, + "grad_norm": 1.447901906431428, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 17790 + }, + { + "epoch": 0.17791, + "grad_norm": 1.3159462136060966, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 17791 + }, + { + "epoch": 0.17792, + "grad_norm": 1.1303918467644893, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17792 + }, + { + "epoch": 0.17793, + "grad_norm": 1.4418297651455436, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 17793 + }, + { + "epoch": 0.17794, + "grad_norm": 1.150072467165776, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17794 + }, + { + "epoch": 0.17795, + "grad_norm": 1.2882828435216742, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 17795 + }, + { + "epoch": 0.17796, + "grad_norm": 1.5300047434012467, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 17796 + }, + { + "epoch": 0.17797, + "grad_norm": 1.294882456997206, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 17797 + }, + { + "epoch": 0.17798, + "grad_norm": 1.1967947727127546, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 17798 + }, + { + "epoch": 0.17799, + "grad_norm": 1.3078744578101413, + "learning_rate": 0.003, + "loss": 4.081, + "step": 17799 + }, + { + "epoch": 0.178, + "grad_norm": 1.2823202298045382, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 17800 + }, + { + "epoch": 0.17801, + "grad_norm": 1.138629563832294, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 17801 + }, + { + "epoch": 0.17802, + "grad_norm": 1.2476615667142883, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 17802 + }, + { + "epoch": 0.17803, + "grad_norm": 1.3380044295600648, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 17803 + }, + { + "epoch": 0.17804, + "grad_norm": 1.3769474905486965, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 17804 + }, + { + "epoch": 0.17805, + "grad_norm": 1.144291069521152, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17805 + }, + { + "epoch": 0.17806, + "grad_norm": 1.0981636028409751, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 17806 + }, + { + "epoch": 0.17807, + "grad_norm": 1.3824357220871897, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 17807 + }, + { + "epoch": 0.17808, + "grad_norm": 1.2373087518469619, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 17808 + }, + { + "epoch": 0.17809, + "grad_norm": 1.389537481714246, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 17809 + }, + { + "epoch": 0.1781, + "grad_norm": 1.1752206957419395, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 17810 + }, + { + "epoch": 0.17811, + "grad_norm": 1.509089328615908, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 17811 + }, + { + "epoch": 0.17812, + "grad_norm": 1.1544782383342644, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17812 + }, + { + "epoch": 0.17813, + "grad_norm": 1.3880160019556476, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 17813 + }, + { + "epoch": 0.17814, + "grad_norm": 1.1144387814214616, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 17814 + }, + { + "epoch": 0.17815, + "grad_norm": 1.3763135851316985, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 17815 + }, + { + "epoch": 0.17816, + "grad_norm": 1.1987028557509924, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17816 + }, + { + "epoch": 0.17817, + "grad_norm": 1.3212547591704311, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 17817 + }, + { + "epoch": 0.17818, + "grad_norm": 1.030098384672589, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17818 + }, + { + "epoch": 0.17819, + "grad_norm": 1.4627642966562506, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 17819 + }, + { + "epoch": 0.1782, + "grad_norm": 1.2236767766115737, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 17820 + }, + { + "epoch": 0.17821, + "grad_norm": 1.2624109141416808, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 17821 + }, + { + "epoch": 0.17822, + "grad_norm": 1.1786888792865688, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 17822 + }, + { + "epoch": 0.17823, + "grad_norm": 1.0954014569065735, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 17823 + }, + { + "epoch": 0.17824, + "grad_norm": 1.406971076687692, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 17824 + }, + { + "epoch": 0.17825, + "grad_norm": 1.191252071858672, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17825 + }, + { + "epoch": 0.17826, + "grad_norm": 1.1932691508802327, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 17826 + }, + { + "epoch": 0.17827, + "grad_norm": 1.3445866454330102, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 17827 + }, + { + "epoch": 0.17828, + "grad_norm": 0.9395973104612575, + "learning_rate": 0.003, + "loss": 4.063, + "step": 17828 + }, + { + "epoch": 0.17829, + "grad_norm": 1.4964471425613408, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 17829 + }, + { + "epoch": 0.1783, + "grad_norm": 1.2770590995250022, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17830 + }, + { + "epoch": 0.17831, + "grad_norm": 1.7767237762098842, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 17831 + }, + { + "epoch": 0.17832, + "grad_norm": 1.560627635031259, + "learning_rate": 0.003, + "loss": 4.086, + "step": 17832 + }, + { + "epoch": 0.17833, + "grad_norm": 1.1862234180221531, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 17833 + }, + { + "epoch": 0.17834, + "grad_norm": 1.3094702246929275, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17834 + }, + { + "epoch": 0.17835, + "grad_norm": 1.5248087017840872, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 17835 + }, + { + "epoch": 0.17836, + "grad_norm": 1.3169564845416029, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17836 + }, + { + "epoch": 0.17837, + "grad_norm": 1.2675087087992958, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 17837 + }, + { + "epoch": 0.17838, + "grad_norm": 1.2709712053086606, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 17838 + }, + { + "epoch": 0.17839, + "grad_norm": 1.2522036966190369, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 17839 + }, + { + "epoch": 0.1784, + "grad_norm": 1.2620829344617028, + "learning_rate": 0.003, + "loss": 4.077, + "step": 17840 + }, + { + "epoch": 0.17841, + "grad_norm": 1.4012932938037403, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 17841 + }, + { + "epoch": 0.17842, + "grad_norm": 1.3422000839817314, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 17842 + }, + { + "epoch": 0.17843, + "grad_norm": 1.4006989988827396, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 17843 + }, + { + "epoch": 0.17844, + "grad_norm": 1.0190924565172015, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 17844 + }, + { + "epoch": 0.17845, + "grad_norm": 1.5054749591757886, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 17845 + }, + { + "epoch": 0.17846, + "grad_norm": 1.0716919117744008, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 17846 + }, + { + "epoch": 0.17847, + "grad_norm": 1.4316572708728408, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 17847 + }, + { + "epoch": 0.17848, + "grad_norm": 1.222068982447068, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 17848 + }, + { + "epoch": 0.17849, + "grad_norm": 1.7153759550487508, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 17849 + }, + { + "epoch": 0.1785, + "grad_norm": 1.2297861316703975, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 17850 + }, + { + "epoch": 0.17851, + "grad_norm": 1.4649983039897283, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 17851 + }, + { + "epoch": 0.17852, + "grad_norm": 1.4020148957301672, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 17852 + }, + { + "epoch": 0.17853, + "grad_norm": 1.2346682112425613, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 17853 + }, + { + "epoch": 0.17854, + "grad_norm": 1.106188497254593, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 17854 + }, + { + "epoch": 0.17855, + "grad_norm": 1.2493336816231602, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17855 + }, + { + "epoch": 0.17856, + "grad_norm": 1.179738117162747, + "learning_rate": 0.003, + "loss": 4.082, + "step": 17856 + }, + { + "epoch": 0.17857, + "grad_norm": 1.5935205634776302, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 17857 + }, + { + "epoch": 0.17858, + "grad_norm": 1.3713885979408045, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 17858 + }, + { + "epoch": 0.17859, + "grad_norm": 1.2860808614893842, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 17859 + }, + { + "epoch": 0.1786, + "grad_norm": 1.1659456901295933, + "learning_rate": 0.003, + "loss": 4.046, + "step": 17860 + }, + { + "epoch": 0.17861, + "grad_norm": 1.2412258951354143, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 17861 + }, + { + "epoch": 0.17862, + "grad_norm": 1.2206348088811467, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17862 + }, + { + "epoch": 0.17863, + "grad_norm": 1.2714996956301137, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 17863 + }, + { + "epoch": 0.17864, + "grad_norm": 1.2282390076542682, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 17864 + }, + { + "epoch": 0.17865, + "grad_norm": 1.2259074729816142, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17865 + }, + { + "epoch": 0.17866, + "grad_norm": 1.3227112430040862, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17866 + }, + { + "epoch": 0.17867, + "grad_norm": 1.2388192275713554, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 17867 + }, + { + "epoch": 0.17868, + "grad_norm": 1.36726409734838, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 17868 + }, + { + "epoch": 0.17869, + "grad_norm": 1.2070589878757991, + "learning_rate": 0.003, + "loss": 4.07, + "step": 17869 + }, + { + "epoch": 0.1787, + "grad_norm": 1.3158716539193773, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17870 + }, + { + "epoch": 0.17871, + "grad_norm": 1.1728661622004941, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 17871 + }, + { + "epoch": 0.17872, + "grad_norm": 1.2427782053645458, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17872 + }, + { + "epoch": 0.17873, + "grad_norm": 1.4042019950491103, + "learning_rate": 0.003, + "loss": 4.056, + "step": 17873 + }, + { + "epoch": 0.17874, + "grad_norm": 1.5812638478605645, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 17874 + }, + { + "epoch": 0.17875, + "grad_norm": 1.3564451180031354, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 17875 + }, + { + "epoch": 0.17876, + "grad_norm": 1.2726006566024621, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 17876 + }, + { + "epoch": 0.17877, + "grad_norm": 1.286756604906063, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17877 + }, + { + "epoch": 0.17878, + "grad_norm": 1.3630425772043697, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 17878 + }, + { + "epoch": 0.17879, + "grad_norm": 1.3858094253305993, + "learning_rate": 0.003, + "loss": 4.054, + "step": 17879 + }, + { + "epoch": 0.1788, + "grad_norm": 1.0266998705945807, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 17880 + }, + { + "epoch": 0.17881, + "grad_norm": 1.5182973033573133, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 17881 + }, + { + "epoch": 0.17882, + "grad_norm": 0.9844433945176064, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 17882 + }, + { + "epoch": 0.17883, + "grad_norm": 1.501822512948681, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 17883 + }, + { + "epoch": 0.17884, + "grad_norm": 1.3255113384277992, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 17884 + }, + { + "epoch": 0.17885, + "grad_norm": 1.198685716640698, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 17885 + }, + { + "epoch": 0.17886, + "grad_norm": 1.3021750977825552, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 17886 + }, + { + "epoch": 0.17887, + "grad_norm": 1.3747917696887602, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 17887 + }, + { + "epoch": 0.17888, + "grad_norm": 1.2841788620965964, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 17888 + }, + { + "epoch": 0.17889, + "grad_norm": 1.1959348659836255, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 17889 + }, + { + "epoch": 0.1789, + "grad_norm": 1.6015235031736559, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 17890 + }, + { + "epoch": 0.17891, + "grad_norm": 1.0814896382172312, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 17891 + }, + { + "epoch": 0.17892, + "grad_norm": 1.447675696751703, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 17892 + }, + { + "epoch": 0.17893, + "grad_norm": 1.223293511998367, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 17893 + }, + { + "epoch": 0.17894, + "grad_norm": 1.4438945880341931, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 17894 + }, + { + "epoch": 0.17895, + "grad_norm": 1.2821740281097793, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 17895 + }, + { + "epoch": 0.17896, + "grad_norm": 1.3215919068687485, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 17896 + }, + { + "epoch": 0.17897, + "grad_norm": 1.3140664733236573, + "learning_rate": 0.003, + "loss": 4.078, + "step": 17897 + }, + { + "epoch": 0.17898, + "grad_norm": 1.1907256660849348, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17898 + }, + { + "epoch": 0.17899, + "grad_norm": 1.3430593063476148, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 17899 + }, + { + "epoch": 0.179, + "grad_norm": 1.286488117104526, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 17900 + }, + { + "epoch": 0.17901, + "grad_norm": 1.0695589964441206, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 17901 + }, + { + "epoch": 0.17902, + "grad_norm": 1.3224104997066575, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 17902 + }, + { + "epoch": 0.17903, + "grad_norm": 1.0988145150369284, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 17903 + }, + { + "epoch": 0.17904, + "grad_norm": 1.5340834757275605, + "learning_rate": 0.003, + "loss": 4.072, + "step": 17904 + }, + { + "epoch": 0.17905, + "grad_norm": 1.0051242394701418, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 17905 + }, + { + "epoch": 0.17906, + "grad_norm": 1.7306028260938866, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 17906 + }, + { + "epoch": 0.17907, + "grad_norm": 1.3411396241899942, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17907 + }, + { + "epoch": 0.17908, + "grad_norm": 1.1706475143882584, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 17908 + }, + { + "epoch": 0.17909, + "grad_norm": 1.3571135714578095, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 17909 + }, + { + "epoch": 0.1791, + "grad_norm": 1.1324149438723727, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 17910 + }, + { + "epoch": 0.17911, + "grad_norm": 1.320902031192732, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17911 + }, + { + "epoch": 0.17912, + "grad_norm": 1.2994613905078916, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17912 + }, + { + "epoch": 0.17913, + "grad_norm": 1.3291750239753504, + "learning_rate": 0.003, + "loss": 4.086, + "step": 17913 + }, + { + "epoch": 0.17914, + "grad_norm": 1.172412070111314, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 17914 + }, + { + "epoch": 0.17915, + "grad_norm": 1.5141488612900191, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 17915 + }, + { + "epoch": 0.17916, + "grad_norm": 1.2874531966763487, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 17916 + }, + { + "epoch": 0.17917, + "grad_norm": 1.7777555042118187, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 17917 + }, + { + "epoch": 0.17918, + "grad_norm": 0.9335876254074895, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 17918 + }, + { + "epoch": 0.17919, + "grad_norm": 1.2730438456312825, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 17919 + }, + { + "epoch": 0.1792, + "grad_norm": 1.4749867341353853, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 17920 + }, + { + "epoch": 0.17921, + "grad_norm": 1.3918863739506826, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17921 + }, + { + "epoch": 0.17922, + "grad_norm": 1.3448413037065396, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 17922 + }, + { + "epoch": 0.17923, + "grad_norm": 1.4183796265762036, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 17923 + }, + { + "epoch": 0.17924, + "grad_norm": 0.9052872201481025, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 17924 + }, + { + "epoch": 0.17925, + "grad_norm": 1.3920930596029024, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 17925 + }, + { + "epoch": 0.17926, + "grad_norm": 1.2048862742257411, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17926 + }, + { + "epoch": 0.17927, + "grad_norm": 1.5643594677964943, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 17927 + }, + { + "epoch": 0.17928, + "grad_norm": 1.3310583454415035, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 17928 + }, + { + "epoch": 0.17929, + "grad_norm": 1.4820607233171916, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 17929 + }, + { + "epoch": 0.1793, + "grad_norm": 1.1100563858479453, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 17930 + }, + { + "epoch": 0.17931, + "grad_norm": 1.3002000108662704, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 17931 + }, + { + "epoch": 0.17932, + "grad_norm": 1.2290679867347463, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17932 + }, + { + "epoch": 0.17933, + "grad_norm": 1.251852641559448, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 17933 + }, + { + "epoch": 0.17934, + "grad_norm": 1.3677317247454568, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 17934 + }, + { + "epoch": 0.17935, + "grad_norm": 1.2576152143696935, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 17935 + }, + { + "epoch": 0.17936, + "grad_norm": 1.1526586197354032, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 17936 + }, + { + "epoch": 0.17937, + "grad_norm": 1.3564977878604803, + "learning_rate": 0.003, + "loss": 4.078, + "step": 17937 + }, + { + "epoch": 0.17938, + "grad_norm": 1.297570645938877, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 17938 + }, + { + "epoch": 0.17939, + "grad_norm": 1.2538252403226684, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 17939 + }, + { + "epoch": 0.1794, + "grad_norm": 1.4210528432483376, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 17940 + }, + { + "epoch": 0.17941, + "grad_norm": 1.29711148921775, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 17941 + }, + { + "epoch": 0.17942, + "grad_norm": 1.257774089270801, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 17942 + }, + { + "epoch": 0.17943, + "grad_norm": 1.3189931995828559, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 17943 + }, + { + "epoch": 0.17944, + "grad_norm": 1.5827618271151869, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 17944 + }, + { + "epoch": 0.17945, + "grad_norm": 1.221743354547214, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 17945 + }, + { + "epoch": 0.17946, + "grad_norm": 1.4848743350684563, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 17946 + }, + { + "epoch": 0.17947, + "grad_norm": 1.096557597019516, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17947 + }, + { + "epoch": 0.17948, + "grad_norm": 1.4094750081051672, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 17948 + }, + { + "epoch": 0.17949, + "grad_norm": 1.1307403812560006, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 17949 + }, + { + "epoch": 0.1795, + "grad_norm": 1.2544747953601025, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 17950 + }, + { + "epoch": 0.17951, + "grad_norm": 1.0712320237750856, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 17951 + }, + { + "epoch": 0.17952, + "grad_norm": 1.457389803142145, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 17952 + }, + { + "epoch": 0.17953, + "grad_norm": 1.2058223116981568, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 17953 + }, + { + "epoch": 0.17954, + "grad_norm": 1.3707871230133384, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 17954 + }, + { + "epoch": 0.17955, + "grad_norm": 1.4545497481766416, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 17955 + }, + { + "epoch": 0.17956, + "grad_norm": 1.5444783073178818, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17956 + }, + { + "epoch": 0.17957, + "grad_norm": 1.2718200438362388, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 17957 + }, + { + "epoch": 0.17958, + "grad_norm": 1.3896879812577554, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 17958 + }, + { + "epoch": 0.17959, + "grad_norm": 1.1996278596396401, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 17959 + }, + { + "epoch": 0.1796, + "grad_norm": 1.276239698502112, + "learning_rate": 0.003, + "loss": 4.072, + "step": 17960 + }, + { + "epoch": 0.17961, + "grad_norm": 1.1378747572368832, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 17961 + }, + { + "epoch": 0.17962, + "grad_norm": 1.2135863102985003, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 17962 + }, + { + "epoch": 0.17963, + "grad_norm": 1.3228436626369233, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17963 + }, + { + "epoch": 0.17964, + "grad_norm": 1.3254583377258655, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 17964 + }, + { + "epoch": 0.17965, + "grad_norm": 1.3328612987006105, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 17965 + }, + { + "epoch": 0.17966, + "grad_norm": 1.560116283233121, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 17966 + }, + { + "epoch": 0.17967, + "grad_norm": 1.315203801181062, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 17967 + }, + { + "epoch": 0.17968, + "grad_norm": 1.4446968555287374, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 17968 + }, + { + "epoch": 0.17969, + "grad_norm": 1.2675016047828012, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 17969 + }, + { + "epoch": 0.1797, + "grad_norm": 1.4689414019046123, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 17970 + }, + { + "epoch": 0.17971, + "grad_norm": 1.2425913769483934, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 17971 + }, + { + "epoch": 0.17972, + "grad_norm": 1.103654538103382, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 17972 + }, + { + "epoch": 0.17973, + "grad_norm": 1.3610423718319988, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 17973 + }, + { + "epoch": 0.17974, + "grad_norm": 1.2302631753766913, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 17974 + }, + { + "epoch": 0.17975, + "grad_norm": 1.608979570575041, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 17975 + }, + { + "epoch": 0.17976, + "grad_norm": 1.0055260741776726, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 17976 + }, + { + "epoch": 0.17977, + "grad_norm": 1.431267475756132, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 17977 + }, + { + "epoch": 0.17978, + "grad_norm": 1.2389833984104677, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 17978 + }, + { + "epoch": 0.17979, + "grad_norm": 1.228422133067524, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 17979 + }, + { + "epoch": 0.1798, + "grad_norm": 1.2010273619214957, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 17980 + }, + { + "epoch": 0.17981, + "grad_norm": 1.2308198217935225, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 17981 + }, + { + "epoch": 0.17982, + "grad_norm": 1.323924689196548, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 17982 + }, + { + "epoch": 0.17983, + "grad_norm": 1.1901092782957963, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 17983 + }, + { + "epoch": 0.17984, + "grad_norm": 1.2999788528968683, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 17984 + }, + { + "epoch": 0.17985, + "grad_norm": 1.2620420541392119, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 17985 + }, + { + "epoch": 0.17986, + "grad_norm": 1.1734711827379214, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 17986 + }, + { + "epoch": 0.17987, + "grad_norm": 1.3278895027861535, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 17987 + }, + { + "epoch": 0.17988, + "grad_norm": 1.3501424373966742, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 17988 + }, + { + "epoch": 0.17989, + "grad_norm": 1.1354312105079478, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 17989 + }, + { + "epoch": 0.1799, + "grad_norm": 1.329559896487889, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 17990 + }, + { + "epoch": 0.17991, + "grad_norm": 1.4349224235104836, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 17991 + }, + { + "epoch": 0.17992, + "grad_norm": 1.356964151271892, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 17992 + }, + { + "epoch": 0.17993, + "grad_norm": 1.5299973250772203, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 17993 + }, + { + "epoch": 0.17994, + "grad_norm": 1.3453109033511206, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 17994 + }, + { + "epoch": 0.17995, + "grad_norm": 1.4934737887896772, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 17995 + }, + { + "epoch": 0.17996, + "grad_norm": 1.2869454407391248, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 17996 + }, + { + "epoch": 0.17997, + "grad_norm": 1.1682327411401674, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 17997 + }, + { + "epoch": 0.17998, + "grad_norm": 1.5491421161896177, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 17998 + }, + { + "epoch": 0.17999, + "grad_norm": 1.0293246193267205, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 17999 + }, + { + "epoch": 0.18, + "grad_norm": 1.5678489562289901, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18000 + }, + { + "epoch": 0.18001, + "grad_norm": 1.1834753468106012, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 18001 + }, + { + "epoch": 0.18002, + "grad_norm": 1.6715971464408366, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 18002 + }, + { + "epoch": 0.18003, + "grad_norm": 1.4474001820058198, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18003 + }, + { + "epoch": 0.18004, + "grad_norm": 1.1168831694904024, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18004 + }, + { + "epoch": 0.18005, + "grad_norm": 1.3800372860497383, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18005 + }, + { + "epoch": 0.18006, + "grad_norm": 1.2425220773680559, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 18006 + }, + { + "epoch": 0.18007, + "grad_norm": 1.1991129047475015, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18007 + }, + { + "epoch": 0.18008, + "grad_norm": 1.391553308009128, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18008 + }, + { + "epoch": 0.18009, + "grad_norm": 1.3255645061052017, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 18009 + }, + { + "epoch": 0.1801, + "grad_norm": 1.2206631167682767, + "learning_rate": 0.003, + "loss": 4.053, + "step": 18010 + }, + { + "epoch": 0.18011, + "grad_norm": 1.1316284763305127, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18011 + }, + { + "epoch": 0.18012, + "grad_norm": 1.2504797489669202, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18012 + }, + { + "epoch": 0.18013, + "grad_norm": 1.2413159854996367, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 18013 + }, + { + "epoch": 0.18014, + "grad_norm": 1.2200518278211834, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18014 + }, + { + "epoch": 0.18015, + "grad_norm": 1.261611278686195, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 18015 + }, + { + "epoch": 0.18016, + "grad_norm": 1.2414894702918717, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18016 + }, + { + "epoch": 0.18017, + "grad_norm": 1.2697807333924054, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 18017 + }, + { + "epoch": 0.18018, + "grad_norm": 1.2761231347528381, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 18018 + }, + { + "epoch": 0.18019, + "grad_norm": 1.4608018721023053, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 18019 + }, + { + "epoch": 0.1802, + "grad_norm": 1.1142836402680227, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 18020 + }, + { + "epoch": 0.18021, + "grad_norm": 1.2716124604703851, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 18021 + }, + { + "epoch": 0.18022, + "grad_norm": 1.374731629938203, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18022 + }, + { + "epoch": 0.18023, + "grad_norm": 1.451811766902569, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 18023 + }, + { + "epoch": 0.18024, + "grad_norm": 1.4886827226216808, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 18024 + }, + { + "epoch": 0.18025, + "grad_norm": 1.195325145255202, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18025 + }, + { + "epoch": 0.18026, + "grad_norm": 1.5663069281027013, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 18026 + }, + { + "epoch": 0.18027, + "grad_norm": 1.1654133475363146, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18027 + }, + { + "epoch": 0.18028, + "grad_norm": 1.4276273643981463, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 18028 + }, + { + "epoch": 0.18029, + "grad_norm": 1.0649990139058456, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 18029 + }, + { + "epoch": 0.1803, + "grad_norm": 1.4174487375068698, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 18030 + }, + { + "epoch": 0.18031, + "grad_norm": 1.1292646849660204, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 18031 + }, + { + "epoch": 0.18032, + "grad_norm": 1.4413185788727525, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 18032 + }, + { + "epoch": 0.18033, + "grad_norm": 1.093605221382924, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 18033 + }, + { + "epoch": 0.18034, + "grad_norm": 1.7307527217592507, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 18034 + }, + { + "epoch": 0.18035, + "grad_norm": 1.2898895378608528, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18035 + }, + { + "epoch": 0.18036, + "grad_norm": 1.3536138462937466, + "learning_rate": 0.003, + "loss": 4.083, + "step": 18036 + }, + { + "epoch": 0.18037, + "grad_norm": 1.1937189446306715, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18037 + }, + { + "epoch": 0.18038, + "grad_norm": 1.656402603048207, + "learning_rate": 0.003, + "loss": 4.079, + "step": 18038 + }, + { + "epoch": 0.18039, + "grad_norm": 1.2231668587819087, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18039 + }, + { + "epoch": 0.1804, + "grad_norm": 1.3645169199537732, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 18040 + }, + { + "epoch": 0.18041, + "grad_norm": 1.4515925031452472, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18041 + }, + { + "epoch": 0.18042, + "grad_norm": 1.3464220796376452, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 18042 + }, + { + "epoch": 0.18043, + "grad_norm": 1.2258874359395866, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18043 + }, + { + "epoch": 0.18044, + "grad_norm": 1.4401874470380218, + "learning_rate": 0.003, + "loss": 4.109, + "step": 18044 + }, + { + "epoch": 0.18045, + "grad_norm": 1.1510208774316257, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 18045 + }, + { + "epoch": 0.18046, + "grad_norm": 1.1959196359638038, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 18046 + }, + { + "epoch": 0.18047, + "grad_norm": 1.276908581818575, + "learning_rate": 0.003, + "loss": 4.07, + "step": 18047 + }, + { + "epoch": 0.18048, + "grad_norm": 1.2791496037211918, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18048 + }, + { + "epoch": 0.18049, + "grad_norm": 1.481903853376919, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 18049 + }, + { + "epoch": 0.1805, + "grad_norm": 1.1366849161073074, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 18050 + }, + { + "epoch": 0.18051, + "grad_norm": 1.66043727102517, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 18051 + }, + { + "epoch": 0.18052, + "grad_norm": 1.1072219344214962, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 18052 + }, + { + "epoch": 0.18053, + "grad_norm": 1.384367018671037, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 18053 + }, + { + "epoch": 0.18054, + "grad_norm": 1.1999621726418512, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18054 + }, + { + "epoch": 0.18055, + "grad_norm": 1.2667827934525069, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 18055 + }, + { + "epoch": 0.18056, + "grad_norm": 1.301159822239713, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18056 + }, + { + "epoch": 0.18057, + "grad_norm": 1.3779888030283471, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 18057 + }, + { + "epoch": 0.18058, + "grad_norm": 1.3066880666788008, + "learning_rate": 0.003, + "loss": 4.054, + "step": 18058 + }, + { + "epoch": 0.18059, + "grad_norm": 1.0550051003911673, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 18059 + }, + { + "epoch": 0.1806, + "grad_norm": 1.4478437157575914, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 18060 + }, + { + "epoch": 0.18061, + "grad_norm": 1.1663387043785898, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18061 + }, + { + "epoch": 0.18062, + "grad_norm": 1.4054634477600398, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18062 + }, + { + "epoch": 0.18063, + "grad_norm": 1.2252286810544568, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 18063 + }, + { + "epoch": 0.18064, + "grad_norm": 1.517116308704269, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 18064 + }, + { + "epoch": 0.18065, + "grad_norm": 1.1897742736690238, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18065 + }, + { + "epoch": 0.18066, + "grad_norm": 1.7271267226074312, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18066 + }, + { + "epoch": 0.18067, + "grad_norm": 1.2051866040119912, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18067 + }, + { + "epoch": 0.18068, + "grad_norm": 1.1611631120066888, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 18068 + }, + { + "epoch": 0.18069, + "grad_norm": 1.527347211410476, + "learning_rate": 0.003, + "loss": 4.072, + "step": 18069 + }, + { + "epoch": 0.1807, + "grad_norm": 1.0483082392759384, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 18070 + }, + { + "epoch": 0.18071, + "grad_norm": 1.825337932151561, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18071 + }, + { + "epoch": 0.18072, + "grad_norm": 1.3795901853936237, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 18072 + }, + { + "epoch": 0.18073, + "grad_norm": 1.2409011222770525, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 18073 + }, + { + "epoch": 0.18074, + "grad_norm": 1.2368957540212302, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18074 + }, + { + "epoch": 0.18075, + "grad_norm": 1.2703828465389364, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 18075 + }, + { + "epoch": 0.18076, + "grad_norm": 1.2564309191832448, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18076 + }, + { + "epoch": 0.18077, + "grad_norm": 1.2206101110957257, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 18077 + }, + { + "epoch": 0.18078, + "grad_norm": 1.407713569678958, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18078 + }, + { + "epoch": 0.18079, + "grad_norm": 1.3661028245510936, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 18079 + }, + { + "epoch": 0.1808, + "grad_norm": 1.2258048402516895, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18080 + }, + { + "epoch": 0.18081, + "grad_norm": 1.5843330540638398, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 18081 + }, + { + "epoch": 0.18082, + "grad_norm": 0.6936825374337909, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 18082 + }, + { + "epoch": 0.18083, + "grad_norm": 1.0248203663148996, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 18083 + }, + { + "epoch": 0.18084, + "grad_norm": 1.716197730295827, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 18084 + }, + { + "epoch": 0.18085, + "grad_norm": 0.9585141219870723, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 18085 + }, + { + "epoch": 0.18086, + "grad_norm": 1.3379443208903115, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 18086 + }, + { + "epoch": 0.18087, + "grad_norm": 1.2366463569268664, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 18087 + }, + { + "epoch": 0.18088, + "grad_norm": 1.2636307127605448, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 18088 + }, + { + "epoch": 0.18089, + "grad_norm": 1.4224497920411832, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18089 + }, + { + "epoch": 0.1809, + "grad_norm": 1.1023749172205495, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 18090 + }, + { + "epoch": 0.18091, + "grad_norm": 1.2543315139083586, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 18091 + }, + { + "epoch": 0.18092, + "grad_norm": 1.1783772383280644, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 18092 + }, + { + "epoch": 0.18093, + "grad_norm": 1.5117848090462052, + "learning_rate": 0.003, + "loss": 4.095, + "step": 18093 + }, + { + "epoch": 0.18094, + "grad_norm": 1.4048702513973914, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 18094 + }, + { + "epoch": 0.18095, + "grad_norm": 1.2867382303272386, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 18095 + }, + { + "epoch": 0.18096, + "grad_norm": 1.0842882454502578, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 18096 + }, + { + "epoch": 0.18097, + "grad_norm": 1.402049280618989, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 18097 + }, + { + "epoch": 0.18098, + "grad_norm": 1.2740222906951524, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18098 + }, + { + "epoch": 0.18099, + "grad_norm": 1.4092021306239473, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18099 + }, + { + "epoch": 0.181, + "grad_norm": 1.212300283369773, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 18100 + }, + { + "epoch": 0.18101, + "grad_norm": 1.4631334539717482, + "learning_rate": 0.003, + "loss": 4.064, + "step": 18101 + }, + { + "epoch": 0.18102, + "grad_norm": 1.2044697301208913, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18102 + }, + { + "epoch": 0.18103, + "grad_norm": 1.3150892284595763, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 18103 + }, + { + "epoch": 0.18104, + "grad_norm": 1.2506959429607367, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 18104 + }, + { + "epoch": 0.18105, + "grad_norm": 1.2323593849004972, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 18105 + }, + { + "epoch": 0.18106, + "grad_norm": 1.1983625716479727, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 18106 + }, + { + "epoch": 0.18107, + "grad_norm": 1.380401888137709, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 18107 + }, + { + "epoch": 0.18108, + "grad_norm": 1.4616032112616257, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 18108 + }, + { + "epoch": 0.18109, + "grad_norm": 1.2819611007798433, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 18109 + }, + { + "epoch": 0.1811, + "grad_norm": 1.1511736124574043, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 18110 + }, + { + "epoch": 0.18111, + "grad_norm": 1.6064113458997202, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18111 + }, + { + "epoch": 0.18112, + "grad_norm": 1.0087387971819672, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 18112 + }, + { + "epoch": 0.18113, + "grad_norm": 1.6618416921990447, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 18113 + }, + { + "epoch": 0.18114, + "grad_norm": 1.2000846536824, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 18114 + }, + { + "epoch": 0.18115, + "grad_norm": 1.1682848970633601, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 18115 + }, + { + "epoch": 0.18116, + "grad_norm": 1.3772552377987377, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 18116 + }, + { + "epoch": 0.18117, + "grad_norm": 1.2928307550029645, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 18117 + }, + { + "epoch": 0.18118, + "grad_norm": 1.5027885344937797, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 18118 + }, + { + "epoch": 0.18119, + "grad_norm": 1.434871091486078, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 18119 + }, + { + "epoch": 0.1812, + "grad_norm": 1.2094159579471924, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18120 + }, + { + "epoch": 0.18121, + "grad_norm": 1.3542369206507303, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 18121 + }, + { + "epoch": 0.18122, + "grad_norm": 1.069341226497996, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 18122 + }, + { + "epoch": 0.18123, + "grad_norm": 1.270477721298167, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 18123 + }, + { + "epoch": 0.18124, + "grad_norm": 1.2497256681751179, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 18124 + }, + { + "epoch": 0.18125, + "grad_norm": 1.1922169911959235, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18125 + }, + { + "epoch": 0.18126, + "grad_norm": 1.4041714736757644, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 18126 + }, + { + "epoch": 0.18127, + "grad_norm": 1.4731383145523111, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 18127 + }, + { + "epoch": 0.18128, + "grad_norm": 1.2021229909504054, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 18128 + }, + { + "epoch": 0.18129, + "grad_norm": 1.5091910261960564, + "learning_rate": 0.003, + "loss": 4.047, + "step": 18129 + }, + { + "epoch": 0.1813, + "grad_norm": 1.3231377072043908, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18130 + }, + { + "epoch": 0.18131, + "grad_norm": 1.1523137535135268, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 18131 + }, + { + "epoch": 0.18132, + "grad_norm": 1.2015203464085935, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18132 + }, + { + "epoch": 0.18133, + "grad_norm": 1.2494390904610018, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 18133 + }, + { + "epoch": 0.18134, + "grad_norm": 1.2541054427205613, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 18134 + }, + { + "epoch": 0.18135, + "grad_norm": 1.2488046257842773, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 18135 + }, + { + "epoch": 0.18136, + "grad_norm": 1.2661888618752437, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 18136 + }, + { + "epoch": 0.18137, + "grad_norm": 1.3085164931949764, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 18137 + }, + { + "epoch": 0.18138, + "grad_norm": 1.1491090041473908, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 18138 + }, + { + "epoch": 0.18139, + "grad_norm": 1.491569568936032, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 18139 + }, + { + "epoch": 0.1814, + "grad_norm": 1.2807156445409174, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 18140 + }, + { + "epoch": 0.18141, + "grad_norm": 1.298896890123939, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18141 + }, + { + "epoch": 0.18142, + "grad_norm": 1.4534689097758247, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18142 + }, + { + "epoch": 0.18143, + "grad_norm": 1.2760736538512882, + "learning_rate": 0.003, + "loss": 4.041, + "step": 18143 + }, + { + "epoch": 0.18144, + "grad_norm": 1.2731093061022918, + "learning_rate": 0.003, + "loss": 4.074, + "step": 18144 + }, + { + "epoch": 0.18145, + "grad_norm": 1.397239557622711, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 18145 + }, + { + "epoch": 0.18146, + "grad_norm": 1.2370712894447977, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 18146 + }, + { + "epoch": 0.18147, + "grad_norm": 1.408516610656114, + "learning_rate": 0.003, + "loss": 4.059, + "step": 18147 + }, + { + "epoch": 0.18148, + "grad_norm": 1.1404495545854807, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 18148 + }, + { + "epoch": 0.18149, + "grad_norm": 1.4999510770973097, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18149 + }, + { + "epoch": 0.1815, + "grad_norm": 1.3412249640259764, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 18150 + }, + { + "epoch": 0.18151, + "grad_norm": 1.340655032055312, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 18151 + }, + { + "epoch": 0.18152, + "grad_norm": 1.3430861191611438, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18152 + }, + { + "epoch": 0.18153, + "grad_norm": 1.2229632330161644, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 18153 + }, + { + "epoch": 0.18154, + "grad_norm": 1.3686194210829088, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 18154 + }, + { + "epoch": 0.18155, + "grad_norm": 1.3695351862872984, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 18155 + }, + { + "epoch": 0.18156, + "grad_norm": 1.1962950950615079, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18156 + }, + { + "epoch": 0.18157, + "grad_norm": 1.3791216928398669, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18157 + }, + { + "epoch": 0.18158, + "grad_norm": 1.4249186981770756, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 18158 + }, + { + "epoch": 0.18159, + "grad_norm": 1.6859917691018185, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18159 + }, + { + "epoch": 0.1816, + "grad_norm": 1.0514003415436335, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 18160 + }, + { + "epoch": 0.18161, + "grad_norm": 1.3370164144149683, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 18161 + }, + { + "epoch": 0.18162, + "grad_norm": 1.4367597693541223, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 18162 + }, + { + "epoch": 0.18163, + "grad_norm": 1.1688501705480174, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 18163 + }, + { + "epoch": 0.18164, + "grad_norm": 1.4386363032319445, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18164 + }, + { + "epoch": 0.18165, + "grad_norm": 1.0009150657523624, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 18165 + }, + { + "epoch": 0.18166, + "grad_norm": 1.6000360556528737, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 18166 + }, + { + "epoch": 0.18167, + "grad_norm": 1.0258430483937688, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 18167 + }, + { + "epoch": 0.18168, + "grad_norm": 1.3458346581546479, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18168 + }, + { + "epoch": 0.18169, + "grad_norm": 1.6714686561058865, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 18169 + }, + { + "epoch": 0.1817, + "grad_norm": 1.4872920520624644, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 18170 + }, + { + "epoch": 0.18171, + "grad_norm": 1.0951495742612987, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 18171 + }, + { + "epoch": 0.18172, + "grad_norm": 1.1112247590858297, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18172 + }, + { + "epoch": 0.18173, + "grad_norm": 1.5872827478046938, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 18173 + }, + { + "epoch": 0.18174, + "grad_norm": 1.340632879143833, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 18174 + }, + { + "epoch": 0.18175, + "grad_norm": 1.12270157417685, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 18175 + }, + { + "epoch": 0.18176, + "grad_norm": 1.2548336496990686, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 18176 + }, + { + "epoch": 0.18177, + "grad_norm": 1.3348772544620668, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 18177 + }, + { + "epoch": 0.18178, + "grad_norm": 1.4976886591580911, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 18178 + }, + { + "epoch": 0.18179, + "grad_norm": 1.1434711682797878, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18179 + }, + { + "epoch": 0.1818, + "grad_norm": 1.4240373669379667, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 18180 + }, + { + "epoch": 0.18181, + "grad_norm": 1.2303593962497674, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 18181 + }, + { + "epoch": 0.18182, + "grad_norm": 1.3255765937167197, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18182 + }, + { + "epoch": 0.18183, + "grad_norm": 1.5157103448871325, + "learning_rate": 0.003, + "loss": 4.058, + "step": 18183 + }, + { + "epoch": 0.18184, + "grad_norm": 1.3625922244136983, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 18184 + }, + { + "epoch": 0.18185, + "grad_norm": 1.5395208661727047, + "learning_rate": 0.003, + "loss": 4.085, + "step": 18185 + }, + { + "epoch": 0.18186, + "grad_norm": 0.9853905024374693, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18186 + }, + { + "epoch": 0.18187, + "grad_norm": 1.5047072672416895, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 18187 + }, + { + "epoch": 0.18188, + "grad_norm": 1.1775613512995033, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18188 + }, + { + "epoch": 0.18189, + "grad_norm": 1.384245652769638, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 18189 + }, + { + "epoch": 0.1819, + "grad_norm": 1.333759677855863, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 18190 + }, + { + "epoch": 0.18191, + "grad_norm": 1.3372005644077896, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18191 + }, + { + "epoch": 0.18192, + "grad_norm": 1.3535946694206413, + "learning_rate": 0.003, + "loss": 4.074, + "step": 18192 + }, + { + "epoch": 0.18193, + "grad_norm": 1.2468378090575452, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 18193 + }, + { + "epoch": 0.18194, + "grad_norm": 1.4532766301040654, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 18194 + }, + { + "epoch": 0.18195, + "grad_norm": 1.1686842936325825, + "learning_rate": 0.003, + "loss": 4.066, + "step": 18195 + }, + { + "epoch": 0.18196, + "grad_norm": 1.4565038985790555, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 18196 + }, + { + "epoch": 0.18197, + "grad_norm": 1.0958727172186844, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 18197 + }, + { + "epoch": 0.18198, + "grad_norm": 1.4138226117913761, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 18198 + }, + { + "epoch": 0.18199, + "grad_norm": 1.162239856816722, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 18199 + }, + { + "epoch": 0.182, + "grad_norm": 1.2385407734307363, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18200 + }, + { + "epoch": 0.18201, + "grad_norm": 1.2635024331352518, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 18201 + }, + { + "epoch": 0.18202, + "grad_norm": 1.3092375521949944, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 18202 + }, + { + "epoch": 0.18203, + "grad_norm": 1.4449188880190564, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18203 + }, + { + "epoch": 0.18204, + "grad_norm": 1.3422035962759664, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 18204 + }, + { + "epoch": 0.18205, + "grad_norm": 1.4648471640630332, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 18205 + }, + { + "epoch": 0.18206, + "grad_norm": 1.2181134644769143, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18206 + }, + { + "epoch": 0.18207, + "grad_norm": 1.2960922502120704, + "learning_rate": 0.003, + "loss": 4.06, + "step": 18207 + }, + { + "epoch": 0.18208, + "grad_norm": 1.179304476592766, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 18208 + }, + { + "epoch": 0.18209, + "grad_norm": 1.1808590466178734, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 18209 + }, + { + "epoch": 0.1821, + "grad_norm": 1.2501246950972718, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 18210 + }, + { + "epoch": 0.18211, + "grad_norm": 1.175661923685183, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 18211 + }, + { + "epoch": 0.18212, + "grad_norm": 1.4023298714534649, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18212 + }, + { + "epoch": 0.18213, + "grad_norm": 1.3569060647030573, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 18213 + }, + { + "epoch": 0.18214, + "grad_norm": 1.5227566256911091, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18214 + }, + { + "epoch": 0.18215, + "grad_norm": 1.1338972886688918, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 18215 + }, + { + "epoch": 0.18216, + "grad_norm": 1.604034654117611, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 18216 + }, + { + "epoch": 0.18217, + "grad_norm": 1.1325825049736928, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18217 + }, + { + "epoch": 0.18218, + "grad_norm": 1.2664514497000146, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 18218 + }, + { + "epoch": 0.18219, + "grad_norm": 1.442785106896768, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18219 + }, + { + "epoch": 0.1822, + "grad_norm": 1.1107805701881037, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18220 + }, + { + "epoch": 0.18221, + "grad_norm": 1.4340328138758345, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 18221 + }, + { + "epoch": 0.18222, + "grad_norm": 1.135998506553484, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 18222 + }, + { + "epoch": 0.18223, + "grad_norm": 1.410665229499254, + "learning_rate": 0.003, + "loss": 4.081, + "step": 18223 + }, + { + "epoch": 0.18224, + "grad_norm": 1.2607020950425902, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 18224 + }, + { + "epoch": 0.18225, + "grad_norm": 1.4245556705655618, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 18225 + }, + { + "epoch": 0.18226, + "grad_norm": 1.3512642968003892, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 18226 + }, + { + "epoch": 0.18227, + "grad_norm": 1.1604604069289992, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 18227 + }, + { + "epoch": 0.18228, + "grad_norm": 1.4738669453150621, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18228 + }, + { + "epoch": 0.18229, + "grad_norm": 1.2406082791105169, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18229 + }, + { + "epoch": 0.1823, + "grad_norm": 1.5715096394050443, + "learning_rate": 0.003, + "loss": 4.082, + "step": 18230 + }, + { + "epoch": 0.18231, + "grad_norm": 1.1275615052222625, + "learning_rate": 0.003, + "loss": 4.048, + "step": 18231 + }, + { + "epoch": 0.18232, + "grad_norm": 1.3406386092189289, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 18232 + }, + { + "epoch": 0.18233, + "grad_norm": 1.451787714778182, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 18233 + }, + { + "epoch": 0.18234, + "grad_norm": 1.5453428043833808, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 18234 + }, + { + "epoch": 0.18235, + "grad_norm": 1.2602528537457918, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 18235 + }, + { + "epoch": 0.18236, + "grad_norm": 1.368527128648679, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18236 + }, + { + "epoch": 0.18237, + "grad_norm": 1.1857713306589048, + "learning_rate": 0.003, + "loss": 4.059, + "step": 18237 + }, + { + "epoch": 0.18238, + "grad_norm": 1.1543106412537665, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 18238 + }, + { + "epoch": 0.18239, + "grad_norm": 1.3016308243620187, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 18239 + }, + { + "epoch": 0.1824, + "grad_norm": 1.231102382837433, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 18240 + }, + { + "epoch": 0.18241, + "grad_norm": 1.1077947692409051, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 18241 + }, + { + "epoch": 0.18242, + "grad_norm": 1.4706854398724694, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 18242 + }, + { + "epoch": 0.18243, + "grad_norm": 1.325152785053419, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 18243 + }, + { + "epoch": 0.18244, + "grad_norm": 1.6457422299842572, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 18244 + }, + { + "epoch": 0.18245, + "grad_norm": 1.1406046734509585, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18245 + }, + { + "epoch": 0.18246, + "grad_norm": 1.5286619780425068, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18246 + }, + { + "epoch": 0.18247, + "grad_norm": 1.3525066251140536, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 18247 + }, + { + "epoch": 0.18248, + "grad_norm": 1.2439100214983063, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18248 + }, + { + "epoch": 0.18249, + "grad_norm": 1.154484097160147, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 18249 + }, + { + "epoch": 0.1825, + "grad_norm": 1.4881899126741536, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 18250 + }, + { + "epoch": 0.18251, + "grad_norm": 1.465279467451196, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 18251 + }, + { + "epoch": 0.18252, + "grad_norm": 1.1982241062906864, + "learning_rate": 0.003, + "loss": 4.095, + "step": 18252 + }, + { + "epoch": 0.18253, + "grad_norm": 1.185664148262174, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 18253 + }, + { + "epoch": 0.18254, + "grad_norm": 1.4467398087390484, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 18254 + }, + { + "epoch": 0.18255, + "grad_norm": 1.1997101263431185, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 18255 + }, + { + "epoch": 0.18256, + "grad_norm": 1.449567953387921, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 18256 + }, + { + "epoch": 0.18257, + "grad_norm": 1.0942210590155477, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 18257 + }, + { + "epoch": 0.18258, + "grad_norm": 1.452957264341418, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 18258 + }, + { + "epoch": 0.18259, + "grad_norm": 1.182946141147893, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18259 + }, + { + "epoch": 0.1826, + "grad_norm": 1.3462294868763975, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18260 + }, + { + "epoch": 0.18261, + "grad_norm": 1.3637873523197717, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 18261 + }, + { + "epoch": 0.18262, + "grad_norm": 1.16961441182416, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 18262 + }, + { + "epoch": 0.18263, + "grad_norm": 1.326741324794611, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18263 + }, + { + "epoch": 0.18264, + "grad_norm": 1.4158659471478507, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 18264 + }, + { + "epoch": 0.18265, + "grad_norm": 1.4652413297438658, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18265 + }, + { + "epoch": 0.18266, + "grad_norm": 1.2955548481668866, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18266 + }, + { + "epoch": 0.18267, + "grad_norm": 1.0863372608696646, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18267 + }, + { + "epoch": 0.18268, + "grad_norm": 1.4834843745030142, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18268 + }, + { + "epoch": 0.18269, + "grad_norm": 1.196583431259954, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 18269 + }, + { + "epoch": 0.1827, + "grad_norm": 1.7173668137332267, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 18270 + }, + { + "epoch": 0.18271, + "grad_norm": 0.9905754971041496, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 18271 + }, + { + "epoch": 0.18272, + "grad_norm": 1.3837736524671629, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18272 + }, + { + "epoch": 0.18273, + "grad_norm": 1.189224070759516, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 18273 + }, + { + "epoch": 0.18274, + "grad_norm": 1.163539976581659, + "learning_rate": 0.003, + "loss": 4.087, + "step": 18274 + }, + { + "epoch": 0.18275, + "grad_norm": 1.2124049633265934, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 18275 + }, + { + "epoch": 0.18276, + "grad_norm": 1.5260532893202114, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 18276 + }, + { + "epoch": 0.18277, + "grad_norm": 1.518910956257597, + "learning_rate": 0.003, + "loss": 4.08, + "step": 18277 + }, + { + "epoch": 0.18278, + "grad_norm": 1.400104167115685, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18278 + }, + { + "epoch": 0.18279, + "grad_norm": 1.1300618726814304, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 18279 + }, + { + "epoch": 0.1828, + "grad_norm": 1.2951842139848218, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18280 + }, + { + "epoch": 0.18281, + "grad_norm": 1.3267232344034476, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 18281 + }, + { + "epoch": 0.18282, + "grad_norm": 1.45139641040153, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18282 + }, + { + "epoch": 0.18283, + "grad_norm": 1.1427113404000755, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 18283 + }, + { + "epoch": 0.18284, + "grad_norm": 1.6198624702211222, + "learning_rate": 0.003, + "loss": 4.096, + "step": 18284 + }, + { + "epoch": 0.18285, + "grad_norm": 1.3551202175210069, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18285 + }, + { + "epoch": 0.18286, + "grad_norm": 1.1961224827929666, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 18286 + }, + { + "epoch": 0.18287, + "grad_norm": 1.3148573588163222, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 18287 + }, + { + "epoch": 0.18288, + "grad_norm": 1.3019818203493474, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 18288 + }, + { + "epoch": 0.18289, + "grad_norm": 1.1852351290061478, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18289 + }, + { + "epoch": 0.1829, + "grad_norm": 1.2162453347899769, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 18290 + }, + { + "epoch": 0.18291, + "grad_norm": 1.3228929662924833, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18291 + }, + { + "epoch": 0.18292, + "grad_norm": 1.2996681355619524, + "learning_rate": 0.003, + "loss": 4.085, + "step": 18292 + }, + { + "epoch": 0.18293, + "grad_norm": 1.4240841378139744, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18293 + }, + { + "epoch": 0.18294, + "grad_norm": 1.5406457155971125, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18294 + }, + { + "epoch": 0.18295, + "grad_norm": 1.2010498038914397, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18295 + }, + { + "epoch": 0.18296, + "grad_norm": 1.2747007627421347, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 18296 + }, + { + "epoch": 0.18297, + "grad_norm": 1.540524874943087, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 18297 + }, + { + "epoch": 0.18298, + "grad_norm": 1.199559660824532, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 18298 + }, + { + "epoch": 0.18299, + "grad_norm": 1.291984495675658, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 18299 + }, + { + "epoch": 0.183, + "grad_norm": 1.2274624100026723, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18300 + }, + { + "epoch": 0.18301, + "grad_norm": 1.3159384603583382, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 18301 + }, + { + "epoch": 0.18302, + "grad_norm": 1.151252609813012, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18302 + }, + { + "epoch": 0.18303, + "grad_norm": 1.3941105679982355, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18303 + }, + { + "epoch": 0.18304, + "grad_norm": 1.2919033233838624, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 18304 + }, + { + "epoch": 0.18305, + "grad_norm": 1.2959301435902835, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 18305 + }, + { + "epoch": 0.18306, + "grad_norm": 1.3290561075222391, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 18306 + }, + { + "epoch": 0.18307, + "grad_norm": 1.3431773936191693, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 18307 + }, + { + "epoch": 0.18308, + "grad_norm": 1.4850679199683352, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 18308 + }, + { + "epoch": 0.18309, + "grad_norm": 1.111470817529982, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 18309 + }, + { + "epoch": 0.1831, + "grad_norm": 1.2791218355500877, + "learning_rate": 0.003, + "loss": 4.065, + "step": 18310 + }, + { + "epoch": 0.18311, + "grad_norm": 1.2679741066880428, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 18311 + }, + { + "epoch": 0.18312, + "grad_norm": 1.2558317099155634, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 18312 + }, + { + "epoch": 0.18313, + "grad_norm": 1.3025506192884788, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 18313 + }, + { + "epoch": 0.18314, + "grad_norm": 1.1158220238267107, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 18314 + }, + { + "epoch": 0.18315, + "grad_norm": 1.5905909968246834, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 18315 + }, + { + "epoch": 0.18316, + "grad_norm": 1.1498620642458874, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 18316 + }, + { + "epoch": 0.18317, + "grad_norm": 1.484744737179069, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 18317 + }, + { + "epoch": 0.18318, + "grad_norm": 1.5264021135168917, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 18318 + }, + { + "epoch": 0.18319, + "grad_norm": 1.1805558057613221, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18319 + }, + { + "epoch": 0.1832, + "grad_norm": 1.5571258801998404, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 18320 + }, + { + "epoch": 0.18321, + "grad_norm": 1.1898018228480296, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 18321 + }, + { + "epoch": 0.18322, + "grad_norm": 1.4159144879762025, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 18322 + }, + { + "epoch": 0.18323, + "grad_norm": 1.1410631707793324, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 18323 + }, + { + "epoch": 0.18324, + "grad_norm": 1.4012920373946773, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18324 + }, + { + "epoch": 0.18325, + "grad_norm": 1.4299141559672417, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 18325 + }, + { + "epoch": 0.18326, + "grad_norm": 1.2930504859850316, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18326 + }, + { + "epoch": 0.18327, + "grad_norm": 1.3468924090170447, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18327 + }, + { + "epoch": 0.18328, + "grad_norm": 1.3180292036073251, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18328 + }, + { + "epoch": 0.18329, + "grad_norm": 1.320514777498404, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18329 + }, + { + "epoch": 0.1833, + "grad_norm": 1.4404597215645119, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 18330 + }, + { + "epoch": 0.18331, + "grad_norm": 1.0605172973346007, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 18331 + }, + { + "epoch": 0.18332, + "grad_norm": 1.4843437599893439, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 18332 + }, + { + "epoch": 0.18333, + "grad_norm": 1.1693918793582336, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 18333 + }, + { + "epoch": 0.18334, + "grad_norm": 1.5217297023948761, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 18334 + }, + { + "epoch": 0.18335, + "grad_norm": 1.1593004710167287, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18335 + }, + { + "epoch": 0.18336, + "grad_norm": 1.3327810915410703, + "learning_rate": 0.003, + "loss": 4.031, + "step": 18336 + }, + { + "epoch": 0.18337, + "grad_norm": 1.213999401958382, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18337 + }, + { + "epoch": 0.18338, + "grad_norm": 1.2807554556088505, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 18338 + }, + { + "epoch": 0.18339, + "grad_norm": 1.1927103956160532, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 18339 + }, + { + "epoch": 0.1834, + "grad_norm": 1.17816569156864, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 18340 + }, + { + "epoch": 0.18341, + "grad_norm": 1.3627814653918315, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 18341 + }, + { + "epoch": 0.18342, + "grad_norm": 1.2234631781988194, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18342 + }, + { + "epoch": 0.18343, + "grad_norm": 1.3372321269836562, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 18343 + }, + { + "epoch": 0.18344, + "grad_norm": 1.2661334492041663, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 18344 + }, + { + "epoch": 0.18345, + "grad_norm": 1.3393651557191961, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 18345 + }, + { + "epoch": 0.18346, + "grad_norm": 1.300894564841833, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 18346 + }, + { + "epoch": 0.18347, + "grad_norm": 1.3276680430228749, + "learning_rate": 0.003, + "loss": 4.06, + "step": 18347 + }, + { + "epoch": 0.18348, + "grad_norm": 1.3937152352869309, + "learning_rate": 0.003, + "loss": 4.075, + "step": 18348 + }, + { + "epoch": 0.18349, + "grad_norm": 1.2502815585549163, + "learning_rate": 0.003, + "loss": 4.055, + "step": 18349 + }, + { + "epoch": 0.1835, + "grad_norm": 1.3517760524979265, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 18350 + }, + { + "epoch": 0.18351, + "grad_norm": 1.1511040155399297, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18351 + }, + { + "epoch": 0.18352, + "grad_norm": 1.5658496591292388, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18352 + }, + { + "epoch": 0.18353, + "grad_norm": 1.1934301437246606, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18353 + }, + { + "epoch": 0.18354, + "grad_norm": 1.3428249291928236, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18354 + }, + { + "epoch": 0.18355, + "grad_norm": 1.262184086853639, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18355 + }, + { + "epoch": 0.18356, + "grad_norm": 1.1838306132491332, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 18356 + }, + { + "epoch": 0.18357, + "grad_norm": 1.3972135551754903, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 18357 + }, + { + "epoch": 0.18358, + "grad_norm": 1.2564781647962338, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18358 + }, + { + "epoch": 0.18359, + "grad_norm": 1.4244018335933282, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 18359 + }, + { + "epoch": 0.1836, + "grad_norm": 1.5316620076526126, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 18360 + }, + { + "epoch": 0.18361, + "grad_norm": 1.2795945610462196, + "learning_rate": 0.003, + "loss": 4.034, + "step": 18361 + }, + { + "epoch": 0.18362, + "grad_norm": 1.1518190277974463, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 18362 + }, + { + "epoch": 0.18363, + "grad_norm": 1.530199581234219, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 18363 + }, + { + "epoch": 0.18364, + "grad_norm": 1.231297952448577, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 18364 + }, + { + "epoch": 0.18365, + "grad_norm": 1.3570466325513808, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18365 + }, + { + "epoch": 0.18366, + "grad_norm": 1.108459820222994, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 18366 + }, + { + "epoch": 0.18367, + "grad_norm": 1.4759435295772985, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18367 + }, + { + "epoch": 0.18368, + "grad_norm": 1.2730892852219353, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18368 + }, + { + "epoch": 0.18369, + "grad_norm": 1.2393159631536057, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18369 + }, + { + "epoch": 0.1837, + "grad_norm": 1.5601089691527303, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18370 + }, + { + "epoch": 0.18371, + "grad_norm": 1.2743664874142249, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18371 + }, + { + "epoch": 0.18372, + "grad_norm": 1.2013851385814807, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18372 + }, + { + "epoch": 0.18373, + "grad_norm": 1.4056078532157135, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 18373 + }, + { + "epoch": 0.18374, + "grad_norm": 1.435369760830309, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18374 + }, + { + "epoch": 0.18375, + "grad_norm": 1.2704736230239198, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 18375 + }, + { + "epoch": 0.18376, + "grad_norm": 1.4623748469123223, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 18376 + }, + { + "epoch": 0.18377, + "grad_norm": 1.051727732073009, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 18377 + }, + { + "epoch": 0.18378, + "grad_norm": 1.3896777833282445, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 18378 + }, + { + "epoch": 0.18379, + "grad_norm": 1.3228853385275277, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18379 + }, + { + "epoch": 0.1838, + "grad_norm": 1.4608524000156693, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 18380 + }, + { + "epoch": 0.18381, + "grad_norm": 1.4029575412483193, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 18381 + }, + { + "epoch": 0.18382, + "grad_norm": 1.5158907956137597, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 18382 + }, + { + "epoch": 0.18383, + "grad_norm": 1.0716546311592323, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 18383 + }, + { + "epoch": 0.18384, + "grad_norm": 1.4981090955511132, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 18384 + }, + { + "epoch": 0.18385, + "grad_norm": 1.0268919429165768, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 18385 + }, + { + "epoch": 0.18386, + "grad_norm": 1.4229816452549553, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 18386 + }, + { + "epoch": 0.18387, + "grad_norm": 1.2366046979296248, + "learning_rate": 0.003, + "loss": 4.07, + "step": 18387 + }, + { + "epoch": 0.18388, + "grad_norm": 1.5639521377941439, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 18388 + }, + { + "epoch": 0.18389, + "grad_norm": 1.1573115439932287, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 18389 + }, + { + "epoch": 0.1839, + "grad_norm": 1.533474200163487, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 18390 + }, + { + "epoch": 0.18391, + "grad_norm": 1.3131162909328664, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 18391 + }, + { + "epoch": 0.18392, + "grad_norm": 1.4200948057489893, + "learning_rate": 0.003, + "loss": 4.041, + "step": 18392 + }, + { + "epoch": 0.18393, + "grad_norm": 1.4495701534876568, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 18393 + }, + { + "epoch": 0.18394, + "grad_norm": 1.0922084287919938, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 18394 + }, + { + "epoch": 0.18395, + "grad_norm": 1.3477790387176505, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18395 + }, + { + "epoch": 0.18396, + "grad_norm": 1.2410173223451075, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 18396 + }, + { + "epoch": 0.18397, + "grad_norm": 1.5249482865632253, + "learning_rate": 0.003, + "loss": 4.088, + "step": 18397 + }, + { + "epoch": 0.18398, + "grad_norm": 1.1072053712305903, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 18398 + }, + { + "epoch": 0.18399, + "grad_norm": 1.5476793058041969, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18399 + }, + { + "epoch": 0.184, + "grad_norm": 1.1607239907993379, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 18400 + }, + { + "epoch": 0.18401, + "grad_norm": 1.4989909112745787, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 18401 + }, + { + "epoch": 0.18402, + "grad_norm": 0.9660817032706551, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 18402 + }, + { + "epoch": 0.18403, + "grad_norm": 1.274004084475861, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 18403 + }, + { + "epoch": 0.18404, + "grad_norm": 1.2982193710720116, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 18404 + }, + { + "epoch": 0.18405, + "grad_norm": 1.3122029694472575, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 18405 + }, + { + "epoch": 0.18406, + "grad_norm": 1.3780064201413342, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 18406 + }, + { + "epoch": 0.18407, + "grad_norm": 1.3424378335998188, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 18407 + }, + { + "epoch": 0.18408, + "grad_norm": 1.0271180117058203, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 18408 + }, + { + "epoch": 0.18409, + "grad_norm": 1.5800784065971756, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 18409 + }, + { + "epoch": 0.1841, + "grad_norm": 1.0228906698663216, + "learning_rate": 0.003, + "loss": 4.06, + "step": 18410 + }, + { + "epoch": 0.18411, + "grad_norm": 1.156292790558681, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 18411 + }, + { + "epoch": 0.18412, + "grad_norm": 1.2843566367494446, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 18412 + }, + { + "epoch": 0.18413, + "grad_norm": 1.1838378450193543, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 18413 + }, + { + "epoch": 0.18414, + "grad_norm": 1.5250702367687015, + "learning_rate": 0.003, + "loss": 4.066, + "step": 18414 + }, + { + "epoch": 0.18415, + "grad_norm": 1.4222352386571093, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18415 + }, + { + "epoch": 0.18416, + "grad_norm": 1.435399268996228, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18416 + }, + { + "epoch": 0.18417, + "grad_norm": 1.0218622741098493, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 18417 + }, + { + "epoch": 0.18418, + "grad_norm": 1.3939864132572706, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18418 + }, + { + "epoch": 0.18419, + "grad_norm": 1.090328698455138, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 18419 + }, + { + "epoch": 0.1842, + "grad_norm": 1.3533141590388846, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 18420 + }, + { + "epoch": 0.18421, + "grad_norm": 1.5480003377160603, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 18421 + }, + { + "epoch": 0.18422, + "grad_norm": 1.3988223043662067, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 18422 + }, + { + "epoch": 0.18423, + "grad_norm": 1.389063700021048, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 18423 + }, + { + "epoch": 0.18424, + "grad_norm": 1.3197639854120087, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 18424 + }, + { + "epoch": 0.18425, + "grad_norm": 1.1066292086722558, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 18425 + }, + { + "epoch": 0.18426, + "grad_norm": 1.4317714961252748, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18426 + }, + { + "epoch": 0.18427, + "grad_norm": 1.1234393863334924, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 18427 + }, + { + "epoch": 0.18428, + "grad_norm": 1.3241461850043657, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18428 + }, + { + "epoch": 0.18429, + "grad_norm": 1.2278628309733401, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 18429 + }, + { + "epoch": 0.1843, + "grad_norm": 1.3376376277392692, + "learning_rate": 0.003, + "loss": 4.046, + "step": 18430 + }, + { + "epoch": 0.18431, + "grad_norm": 1.1514281077049928, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18431 + }, + { + "epoch": 0.18432, + "grad_norm": 1.4278693885956923, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18432 + }, + { + "epoch": 0.18433, + "grad_norm": 1.421603179534481, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 18433 + }, + { + "epoch": 0.18434, + "grad_norm": 1.7790992463663307, + "learning_rate": 0.003, + "loss": 4.069, + "step": 18434 + }, + { + "epoch": 0.18435, + "grad_norm": 1.281655460111802, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18435 + }, + { + "epoch": 0.18436, + "grad_norm": 1.447909389488687, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 18436 + }, + { + "epoch": 0.18437, + "grad_norm": 1.3215450363410834, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 18437 + }, + { + "epoch": 0.18438, + "grad_norm": 1.2994988643640621, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 18438 + }, + { + "epoch": 0.18439, + "grad_norm": 1.1530461997369483, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 18439 + }, + { + "epoch": 0.1844, + "grad_norm": 1.2723226530047036, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 18440 + }, + { + "epoch": 0.18441, + "grad_norm": 1.4190177482257076, + "learning_rate": 0.003, + "loss": 4.054, + "step": 18441 + }, + { + "epoch": 0.18442, + "grad_norm": 1.298334180626103, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18442 + }, + { + "epoch": 0.18443, + "grad_norm": 1.4213127263863718, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18443 + }, + { + "epoch": 0.18444, + "grad_norm": 1.282156090364548, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18444 + }, + { + "epoch": 0.18445, + "grad_norm": 1.2201086731789128, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 18445 + }, + { + "epoch": 0.18446, + "grad_norm": 1.4416977833642208, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 18446 + }, + { + "epoch": 0.18447, + "grad_norm": 1.1922971767874007, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18447 + }, + { + "epoch": 0.18448, + "grad_norm": 1.782110662174641, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 18448 + }, + { + "epoch": 0.18449, + "grad_norm": 1.042098559036565, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 18449 + }, + { + "epoch": 0.1845, + "grad_norm": 1.6867508076955657, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18450 + }, + { + "epoch": 0.18451, + "grad_norm": 1.171477974688329, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18451 + }, + { + "epoch": 0.18452, + "grad_norm": 1.6833010403008586, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18452 + }, + { + "epoch": 0.18453, + "grad_norm": 1.1108276265202692, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 18453 + }, + { + "epoch": 0.18454, + "grad_norm": 1.2513031676322688, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18454 + }, + { + "epoch": 0.18455, + "grad_norm": 1.2370626784842842, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 18455 + }, + { + "epoch": 0.18456, + "grad_norm": 1.3506690505066392, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 18456 + }, + { + "epoch": 0.18457, + "grad_norm": 1.514909151715256, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 18457 + }, + { + "epoch": 0.18458, + "grad_norm": 1.3048506316734978, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 18458 + }, + { + "epoch": 0.18459, + "grad_norm": 1.1903146478199034, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 18459 + }, + { + "epoch": 0.1846, + "grad_norm": 1.2739001027820747, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 18460 + }, + { + "epoch": 0.18461, + "grad_norm": 1.262827186533425, + "learning_rate": 0.003, + "loss": 4.055, + "step": 18461 + }, + { + "epoch": 0.18462, + "grad_norm": 1.4505564797482764, + "learning_rate": 0.003, + "loss": 4.062, + "step": 18462 + }, + { + "epoch": 0.18463, + "grad_norm": 1.2961279196648388, + "learning_rate": 0.003, + "loss": 4.073, + "step": 18463 + }, + { + "epoch": 0.18464, + "grad_norm": 1.10546721769061, + "learning_rate": 0.003, + "loss": 4.067, + "step": 18464 + }, + { + "epoch": 0.18465, + "grad_norm": 1.313201999660418, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18465 + }, + { + "epoch": 0.18466, + "grad_norm": 1.1669156833746952, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 18466 + }, + { + "epoch": 0.18467, + "grad_norm": 1.534547969859101, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 18467 + }, + { + "epoch": 0.18468, + "grad_norm": 1.1393395730416378, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 18468 + }, + { + "epoch": 0.18469, + "grad_norm": 1.5335779758703452, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 18469 + }, + { + "epoch": 0.1847, + "grad_norm": 1.4055567287610196, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 18470 + }, + { + "epoch": 0.18471, + "grad_norm": 1.3167880081135488, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 18471 + }, + { + "epoch": 0.18472, + "grad_norm": 1.2576717359061849, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 18472 + }, + { + "epoch": 0.18473, + "grad_norm": 1.3957171929089696, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 18473 + }, + { + "epoch": 0.18474, + "grad_norm": 1.400310244317996, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18474 + }, + { + "epoch": 0.18475, + "grad_norm": 1.4945282965314968, + "learning_rate": 0.003, + "loss": 4.066, + "step": 18475 + }, + { + "epoch": 0.18476, + "grad_norm": 1.145145941074851, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18476 + }, + { + "epoch": 0.18477, + "grad_norm": 1.455469946244908, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 18477 + }, + { + "epoch": 0.18478, + "grad_norm": 1.3604896246438607, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 18478 + }, + { + "epoch": 0.18479, + "grad_norm": 1.1812863009354766, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 18479 + }, + { + "epoch": 0.1848, + "grad_norm": 1.3283765497707292, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18480 + }, + { + "epoch": 0.18481, + "grad_norm": 1.428316991360103, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 18481 + }, + { + "epoch": 0.18482, + "grad_norm": 1.3072437728678792, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18482 + }, + { + "epoch": 0.18483, + "grad_norm": 1.0899128603018826, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 18483 + }, + { + "epoch": 0.18484, + "grad_norm": 1.3918421419823224, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 18484 + }, + { + "epoch": 0.18485, + "grad_norm": 1.288039295583636, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18485 + }, + { + "epoch": 0.18486, + "grad_norm": 1.3961753464614879, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 18486 + }, + { + "epoch": 0.18487, + "grad_norm": 1.416549086470837, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 18487 + }, + { + "epoch": 0.18488, + "grad_norm": 1.110424938616956, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 18488 + }, + { + "epoch": 0.18489, + "grad_norm": 1.3155622848496273, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18489 + }, + { + "epoch": 0.1849, + "grad_norm": 1.2342935085466868, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 18490 + }, + { + "epoch": 0.18491, + "grad_norm": 1.327027657773121, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18491 + }, + { + "epoch": 0.18492, + "grad_norm": 1.277790745038652, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 18492 + }, + { + "epoch": 0.18493, + "grad_norm": 1.252452693824744, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 18493 + }, + { + "epoch": 0.18494, + "grad_norm": 1.2126657596476098, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 18494 + }, + { + "epoch": 0.18495, + "grad_norm": 1.124775468288146, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 18495 + }, + { + "epoch": 0.18496, + "grad_norm": 1.3963422558678007, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18496 + }, + { + "epoch": 0.18497, + "grad_norm": 1.3112801504123868, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18497 + }, + { + "epoch": 0.18498, + "grad_norm": 1.7519857710163944, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 18498 + }, + { + "epoch": 0.18499, + "grad_norm": 1.194584371996459, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 18499 + }, + { + "epoch": 0.185, + "grad_norm": 1.3195094913957386, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 18500 + }, + { + "epoch": 0.18501, + "grad_norm": 1.3210369721715793, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18501 + }, + { + "epoch": 0.18502, + "grad_norm": 1.3695833326453506, + "learning_rate": 0.003, + "loss": 4.052, + "step": 18502 + }, + { + "epoch": 0.18503, + "grad_norm": 1.2099011060705465, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 18503 + }, + { + "epoch": 0.18504, + "grad_norm": 1.3185094202816177, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18504 + }, + { + "epoch": 0.18505, + "grad_norm": 1.2158771291760275, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 18505 + }, + { + "epoch": 0.18506, + "grad_norm": 1.429753701083564, + "learning_rate": 0.003, + "loss": 4.092, + "step": 18506 + }, + { + "epoch": 0.18507, + "grad_norm": 1.3421314391793797, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 18507 + }, + { + "epoch": 0.18508, + "grad_norm": 1.22920004751544, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 18508 + }, + { + "epoch": 0.18509, + "grad_norm": 1.7533299233045962, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 18509 + }, + { + "epoch": 0.1851, + "grad_norm": 1.1870119764968476, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 18510 + }, + { + "epoch": 0.18511, + "grad_norm": 1.7634954177335502, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 18511 + }, + { + "epoch": 0.18512, + "grad_norm": 1.1433215341938388, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 18512 + }, + { + "epoch": 0.18513, + "grad_norm": 1.1818368377792434, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 18513 + }, + { + "epoch": 0.18514, + "grad_norm": 1.3936303987837908, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 18514 + }, + { + "epoch": 0.18515, + "grad_norm": 1.4106979166568008, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18515 + }, + { + "epoch": 0.18516, + "grad_norm": 1.272932681366381, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18516 + }, + { + "epoch": 0.18517, + "grad_norm": 1.2137932700928198, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 18517 + }, + { + "epoch": 0.18518, + "grad_norm": 1.5125772549658068, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18518 + }, + { + "epoch": 0.18519, + "grad_norm": 1.1143423364661758, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 18519 + }, + { + "epoch": 0.1852, + "grad_norm": 1.3823758749490027, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18520 + }, + { + "epoch": 0.18521, + "grad_norm": 1.2381834575554078, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 18521 + }, + { + "epoch": 0.18522, + "grad_norm": 1.2819900778799418, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 18522 + }, + { + "epoch": 0.18523, + "grad_norm": 1.5474119030645124, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18523 + }, + { + "epoch": 0.18524, + "grad_norm": 1.5018361450590505, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 18524 + }, + { + "epoch": 0.18525, + "grad_norm": 1.2413311633679378, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 18525 + }, + { + "epoch": 0.18526, + "grad_norm": 1.1928737595391086, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 18526 + }, + { + "epoch": 0.18527, + "grad_norm": 1.131070084402993, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18527 + }, + { + "epoch": 0.18528, + "grad_norm": 1.3469415033017138, + "learning_rate": 0.003, + "loss": 4.085, + "step": 18528 + }, + { + "epoch": 0.18529, + "grad_norm": 1.136975179840116, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18529 + }, + { + "epoch": 0.1853, + "grad_norm": 1.4877184864763433, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18530 + }, + { + "epoch": 0.18531, + "grad_norm": 1.3011975456050204, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 18531 + }, + { + "epoch": 0.18532, + "grad_norm": 1.2775776488164456, + "learning_rate": 0.003, + "loss": 4.085, + "step": 18532 + }, + { + "epoch": 0.18533, + "grad_norm": 1.4722603050631387, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 18533 + }, + { + "epoch": 0.18534, + "grad_norm": 1.343931279868763, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 18534 + }, + { + "epoch": 0.18535, + "grad_norm": 1.1299003794550169, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 18535 + }, + { + "epoch": 0.18536, + "grad_norm": 1.7634204482223623, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18536 + }, + { + "epoch": 0.18537, + "grad_norm": 1.1402924195024549, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 18537 + }, + { + "epoch": 0.18538, + "grad_norm": 1.347476089176172, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18538 + }, + { + "epoch": 0.18539, + "grad_norm": 1.1904275311794337, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18539 + }, + { + "epoch": 0.1854, + "grad_norm": 1.2355404913474857, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 18540 + }, + { + "epoch": 0.18541, + "grad_norm": 1.3796686835983873, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18541 + }, + { + "epoch": 0.18542, + "grad_norm": 1.347006219996438, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 18542 + }, + { + "epoch": 0.18543, + "grad_norm": 1.3349654294153273, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18543 + }, + { + "epoch": 0.18544, + "grad_norm": 1.3986597306105117, + "learning_rate": 0.003, + "loss": 4.074, + "step": 18544 + }, + { + "epoch": 0.18545, + "grad_norm": 1.3750003043430603, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18545 + }, + { + "epoch": 0.18546, + "grad_norm": 1.300556246872616, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 18546 + }, + { + "epoch": 0.18547, + "grad_norm": 1.4428540809432855, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18547 + }, + { + "epoch": 0.18548, + "grad_norm": 1.2882429176637629, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 18548 + }, + { + "epoch": 0.18549, + "grad_norm": 1.159611227275313, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 18549 + }, + { + "epoch": 0.1855, + "grad_norm": 1.2285333893294648, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 18550 + }, + { + "epoch": 0.18551, + "grad_norm": 1.0752692148588605, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 18551 + }, + { + "epoch": 0.18552, + "grad_norm": 1.4897134426884537, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 18552 + }, + { + "epoch": 0.18553, + "grad_norm": 1.2080444956794303, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18553 + }, + { + "epoch": 0.18554, + "grad_norm": 1.3374987105954266, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 18554 + }, + { + "epoch": 0.18555, + "grad_norm": 1.155311641991209, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 18555 + }, + { + "epoch": 0.18556, + "grad_norm": 1.5116105928324368, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 18556 + }, + { + "epoch": 0.18557, + "grad_norm": 1.4892136206380056, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18557 + }, + { + "epoch": 0.18558, + "grad_norm": 1.6181035145542664, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 18558 + }, + { + "epoch": 0.18559, + "grad_norm": 1.0805838366203129, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 18559 + }, + { + "epoch": 0.1856, + "grad_norm": 1.2551320772028751, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18560 + }, + { + "epoch": 0.18561, + "grad_norm": 1.5086740960076293, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 18561 + }, + { + "epoch": 0.18562, + "grad_norm": 1.2445156477052786, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 18562 + }, + { + "epoch": 0.18563, + "grad_norm": 1.7085190677221334, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 18563 + }, + { + "epoch": 0.18564, + "grad_norm": 1.069580666816677, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 18564 + }, + { + "epoch": 0.18565, + "grad_norm": 1.5056137508436014, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 18565 + }, + { + "epoch": 0.18566, + "grad_norm": 1.375282844606368, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 18566 + }, + { + "epoch": 0.18567, + "grad_norm": 1.015088299986784, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 18567 + }, + { + "epoch": 0.18568, + "grad_norm": 1.4160477479350546, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 18568 + }, + { + "epoch": 0.18569, + "grad_norm": 1.1523439236916797, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 18569 + }, + { + "epoch": 0.1857, + "grad_norm": 1.288278721419816, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 18570 + }, + { + "epoch": 0.18571, + "grad_norm": 1.4018639203313916, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 18571 + }, + { + "epoch": 0.18572, + "grad_norm": 1.2945508993220207, + "learning_rate": 0.003, + "loss": 4.061, + "step": 18572 + }, + { + "epoch": 0.18573, + "grad_norm": 1.4638142568059527, + "learning_rate": 0.003, + "loss": 4.067, + "step": 18573 + }, + { + "epoch": 0.18574, + "grad_norm": 1.375968230740118, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 18574 + }, + { + "epoch": 0.18575, + "grad_norm": 1.2933957765850859, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 18575 + }, + { + "epoch": 0.18576, + "grad_norm": 1.1955482168446068, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 18576 + }, + { + "epoch": 0.18577, + "grad_norm": 1.2236811418253042, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18577 + }, + { + "epoch": 0.18578, + "grad_norm": 1.494000248261918, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 18578 + }, + { + "epoch": 0.18579, + "grad_norm": 1.2922934257170442, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 18579 + }, + { + "epoch": 0.1858, + "grad_norm": 1.2921553459066577, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 18580 + }, + { + "epoch": 0.18581, + "grad_norm": 1.3782924421301486, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 18581 + }, + { + "epoch": 0.18582, + "grad_norm": 1.1801858551405555, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 18582 + }, + { + "epoch": 0.18583, + "grad_norm": 1.2915391929830393, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 18583 + }, + { + "epoch": 0.18584, + "grad_norm": 1.3198335526433973, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18584 + }, + { + "epoch": 0.18585, + "grad_norm": 1.3653107136423173, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 18585 + }, + { + "epoch": 0.18586, + "grad_norm": 1.3187905420756678, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18586 + }, + { + "epoch": 0.18587, + "grad_norm": 1.192836291864276, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 18587 + }, + { + "epoch": 0.18588, + "grad_norm": 1.4228266801449778, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 18588 + }, + { + "epoch": 0.18589, + "grad_norm": 1.272825202985905, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18589 + }, + { + "epoch": 0.1859, + "grad_norm": 1.4866614615421272, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18590 + }, + { + "epoch": 0.18591, + "grad_norm": 1.1458157948550125, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18591 + }, + { + "epoch": 0.18592, + "grad_norm": 1.5460701241205963, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 18592 + }, + { + "epoch": 0.18593, + "grad_norm": 1.0751417400123864, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 18593 + }, + { + "epoch": 0.18594, + "grad_norm": 1.4584768764361886, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 18594 + }, + { + "epoch": 0.18595, + "grad_norm": 1.1815796913379357, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18595 + }, + { + "epoch": 0.18596, + "grad_norm": 1.3602598493256992, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 18596 + }, + { + "epoch": 0.18597, + "grad_norm": 1.2208729022936036, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 18597 + }, + { + "epoch": 0.18598, + "grad_norm": 1.3645593853664908, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 18598 + }, + { + "epoch": 0.18599, + "grad_norm": 1.1340468689586132, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 18599 + }, + { + "epoch": 0.186, + "grad_norm": 1.282165997408662, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 18600 + }, + { + "epoch": 0.18601, + "grad_norm": 1.2813171509229273, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 18601 + }, + { + "epoch": 0.18602, + "grad_norm": 1.4850002532379016, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 18602 + }, + { + "epoch": 0.18603, + "grad_norm": 1.626791547634331, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 18603 + }, + { + "epoch": 0.18604, + "grad_norm": 1.6681258479888328, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 18604 + }, + { + "epoch": 0.18605, + "grad_norm": 1.1555768713912762, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 18605 + }, + { + "epoch": 0.18606, + "grad_norm": 1.2016352473617748, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 18606 + }, + { + "epoch": 0.18607, + "grad_norm": 1.4890133528694773, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 18607 + }, + { + "epoch": 0.18608, + "grad_norm": 1.0252813859102472, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 18608 + }, + { + "epoch": 0.18609, + "grad_norm": 1.5067240132439041, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 18609 + }, + { + "epoch": 0.1861, + "grad_norm": 1.207825429153814, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 18610 + }, + { + "epoch": 0.18611, + "grad_norm": 1.6342571112613051, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 18611 + }, + { + "epoch": 0.18612, + "grad_norm": 1.1209411418605804, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18612 + }, + { + "epoch": 0.18613, + "grad_norm": 1.6259316754158002, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 18613 + }, + { + "epoch": 0.18614, + "grad_norm": 1.4766417740346707, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 18614 + }, + { + "epoch": 0.18615, + "grad_norm": 1.1620482832574794, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 18615 + }, + { + "epoch": 0.18616, + "grad_norm": 1.4598453729402767, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 18616 + }, + { + "epoch": 0.18617, + "grad_norm": 1.189093300061794, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 18617 + }, + { + "epoch": 0.18618, + "grad_norm": 1.2489389933885495, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 18618 + }, + { + "epoch": 0.18619, + "grad_norm": 1.1206001012677298, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18619 + }, + { + "epoch": 0.1862, + "grad_norm": 1.362284327011221, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 18620 + }, + { + "epoch": 0.18621, + "grad_norm": 1.109117796194937, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 18621 + }, + { + "epoch": 0.18622, + "grad_norm": 1.458631584685427, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 18622 + }, + { + "epoch": 0.18623, + "grad_norm": 1.2468584062896135, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 18623 + }, + { + "epoch": 0.18624, + "grad_norm": 1.5887506899891826, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18624 + }, + { + "epoch": 0.18625, + "grad_norm": 1.2933164217148045, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18625 + }, + { + "epoch": 0.18626, + "grad_norm": 1.2241022509890256, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 18626 + }, + { + "epoch": 0.18627, + "grad_norm": 1.4697934921800258, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 18627 + }, + { + "epoch": 0.18628, + "grad_norm": 1.0401361110722684, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18628 + }, + { + "epoch": 0.18629, + "grad_norm": 1.637268932483506, + "learning_rate": 0.003, + "loss": 4.058, + "step": 18629 + }, + { + "epoch": 0.1863, + "grad_norm": 0.9642316107689074, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 18630 + }, + { + "epoch": 0.18631, + "grad_norm": 1.676448425875956, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 18631 + }, + { + "epoch": 0.18632, + "grad_norm": 1.3233058532473072, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 18632 + }, + { + "epoch": 0.18633, + "grad_norm": 1.4546212115883517, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18633 + }, + { + "epoch": 0.18634, + "grad_norm": 1.1710395395526325, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18634 + }, + { + "epoch": 0.18635, + "grad_norm": 1.3333854885532885, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 18635 + }, + { + "epoch": 0.18636, + "grad_norm": 1.637573612476902, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 18636 + }, + { + "epoch": 0.18637, + "grad_norm": 1.0850780365833652, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 18637 + }, + { + "epoch": 0.18638, + "grad_norm": 1.386893023915246, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 18638 + }, + { + "epoch": 0.18639, + "grad_norm": 1.2265831060490398, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 18639 + }, + { + "epoch": 0.1864, + "grad_norm": 1.3292331057595888, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 18640 + }, + { + "epoch": 0.18641, + "grad_norm": 1.1445977131120186, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 18641 + }, + { + "epoch": 0.18642, + "grad_norm": 1.2961460433247391, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 18642 + }, + { + "epoch": 0.18643, + "grad_norm": 1.2967886889101183, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 18643 + }, + { + "epoch": 0.18644, + "grad_norm": 1.4426551743402376, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 18644 + }, + { + "epoch": 0.18645, + "grad_norm": 1.0748150099012093, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18645 + }, + { + "epoch": 0.18646, + "grad_norm": 1.2475923740165284, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 18646 + }, + { + "epoch": 0.18647, + "grad_norm": 1.3745529718226575, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 18647 + }, + { + "epoch": 0.18648, + "grad_norm": 1.092272608985131, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 18648 + }, + { + "epoch": 0.18649, + "grad_norm": 1.5267651586159152, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 18649 + }, + { + "epoch": 0.1865, + "grad_norm": 1.1513082707972615, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 18650 + }, + { + "epoch": 0.18651, + "grad_norm": 1.393832333231486, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 18651 + }, + { + "epoch": 0.18652, + "grad_norm": 1.2085993270070554, + "learning_rate": 0.003, + "loss": 4.083, + "step": 18652 + }, + { + "epoch": 0.18653, + "grad_norm": 1.1364757032034836, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 18653 + }, + { + "epoch": 0.18654, + "grad_norm": 1.4409042604239815, + "learning_rate": 0.003, + "loss": 4.052, + "step": 18654 + }, + { + "epoch": 0.18655, + "grad_norm": 1.3481757205484133, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 18655 + }, + { + "epoch": 0.18656, + "grad_norm": 1.7729058779742146, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 18656 + }, + { + "epoch": 0.18657, + "grad_norm": 1.2680266619449934, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 18657 + }, + { + "epoch": 0.18658, + "grad_norm": 1.5111394644392886, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18658 + }, + { + "epoch": 0.18659, + "grad_norm": 1.2973492131163422, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 18659 + }, + { + "epoch": 0.1866, + "grad_norm": 1.492565466964901, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18660 + }, + { + "epoch": 0.18661, + "grad_norm": 1.2836233918295634, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 18661 + }, + { + "epoch": 0.18662, + "grad_norm": 1.2841777141938255, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 18662 + }, + { + "epoch": 0.18663, + "grad_norm": 1.3870139131389745, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 18663 + }, + { + "epoch": 0.18664, + "grad_norm": 1.099793394818637, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 18664 + }, + { + "epoch": 0.18665, + "grad_norm": 1.644613665438987, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 18665 + }, + { + "epoch": 0.18666, + "grad_norm": 1.0836969998654264, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18666 + }, + { + "epoch": 0.18667, + "grad_norm": 1.47740999718926, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 18667 + }, + { + "epoch": 0.18668, + "grad_norm": 1.088842733935222, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18668 + }, + { + "epoch": 0.18669, + "grad_norm": 1.2804939696105566, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18669 + }, + { + "epoch": 0.1867, + "grad_norm": 1.2110539378980032, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18670 + }, + { + "epoch": 0.18671, + "grad_norm": 1.2941641843745262, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 18671 + }, + { + "epoch": 0.18672, + "grad_norm": 1.8119945201752838, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18672 + }, + { + "epoch": 0.18673, + "grad_norm": 1.1421809216299648, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 18673 + }, + { + "epoch": 0.18674, + "grad_norm": 1.3000509660383397, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 18674 + }, + { + "epoch": 0.18675, + "grad_norm": 1.2559014844496346, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18675 + }, + { + "epoch": 0.18676, + "grad_norm": 1.6155641495060489, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 18676 + }, + { + "epoch": 0.18677, + "grad_norm": 1.257363389798075, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 18677 + }, + { + "epoch": 0.18678, + "grad_norm": 1.3257530109944569, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 18678 + }, + { + "epoch": 0.18679, + "grad_norm": 1.314530369143493, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 18679 + }, + { + "epoch": 0.1868, + "grad_norm": 1.4568661866191923, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 18680 + }, + { + "epoch": 0.18681, + "grad_norm": 1.3720227968576448, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 18681 + }, + { + "epoch": 0.18682, + "grad_norm": 1.1044912654278982, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 18682 + }, + { + "epoch": 0.18683, + "grad_norm": 1.5181290307684798, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 18683 + }, + { + "epoch": 0.18684, + "grad_norm": 1.1483907499672095, + "learning_rate": 0.003, + "loss": 4.076, + "step": 18684 + }, + { + "epoch": 0.18685, + "grad_norm": 1.3808132590103672, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 18685 + }, + { + "epoch": 0.18686, + "grad_norm": 1.155800467326248, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 18686 + }, + { + "epoch": 0.18687, + "grad_norm": 1.3512800343171085, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 18687 + }, + { + "epoch": 0.18688, + "grad_norm": 1.1024999840102774, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 18688 + }, + { + "epoch": 0.18689, + "grad_norm": 1.4931714589099854, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 18689 + }, + { + "epoch": 0.1869, + "grad_norm": 1.2462176717919797, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 18690 + }, + { + "epoch": 0.18691, + "grad_norm": 1.4843451460985109, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 18691 + }, + { + "epoch": 0.18692, + "grad_norm": 1.1248574285383672, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18692 + }, + { + "epoch": 0.18693, + "grad_norm": 1.63210117841959, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 18693 + }, + { + "epoch": 0.18694, + "grad_norm": 1.3423671310621568, + "learning_rate": 0.003, + "loss": 4.091, + "step": 18694 + }, + { + "epoch": 0.18695, + "grad_norm": 1.2838447284722103, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 18695 + }, + { + "epoch": 0.18696, + "grad_norm": 1.502227661627234, + "learning_rate": 0.003, + "loss": 4.081, + "step": 18696 + }, + { + "epoch": 0.18697, + "grad_norm": 1.0560286188418697, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 18697 + }, + { + "epoch": 0.18698, + "grad_norm": 1.521082857669598, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18698 + }, + { + "epoch": 0.18699, + "grad_norm": 1.2431876001136009, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 18699 + }, + { + "epoch": 0.187, + "grad_norm": 1.372440298449826, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 18700 + }, + { + "epoch": 0.18701, + "grad_norm": 1.1941255846607965, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18701 + }, + { + "epoch": 0.18702, + "grad_norm": 1.3425501143162746, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 18702 + }, + { + "epoch": 0.18703, + "grad_norm": 1.2710114960563306, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 18703 + }, + { + "epoch": 0.18704, + "grad_norm": 1.1443549589368678, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 18704 + }, + { + "epoch": 0.18705, + "grad_norm": 1.2958357679471506, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 18705 + }, + { + "epoch": 0.18706, + "grad_norm": 1.082371359462474, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 18706 + }, + { + "epoch": 0.18707, + "grad_norm": 1.40762327599745, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 18707 + }, + { + "epoch": 0.18708, + "grad_norm": 1.0190726307980469, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18708 + }, + { + "epoch": 0.18709, + "grad_norm": 1.5810885741879948, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18709 + }, + { + "epoch": 0.1871, + "grad_norm": 1.1132235129114374, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 18710 + }, + { + "epoch": 0.18711, + "grad_norm": 1.4880544970134246, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 18711 + }, + { + "epoch": 0.18712, + "grad_norm": 1.269599237812824, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 18712 + }, + { + "epoch": 0.18713, + "grad_norm": 1.3346080176901836, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 18713 + }, + { + "epoch": 0.18714, + "grad_norm": 1.3880945710550838, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 18714 + }, + { + "epoch": 0.18715, + "grad_norm": 1.2721781497891678, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 18715 + }, + { + "epoch": 0.18716, + "grad_norm": 1.388109232318027, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18716 + }, + { + "epoch": 0.18717, + "grad_norm": 1.157295967954611, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18717 + }, + { + "epoch": 0.18718, + "grad_norm": 1.6700084966983302, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 18718 + }, + { + "epoch": 0.18719, + "grad_norm": 0.9754324085832572, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 18719 + }, + { + "epoch": 0.1872, + "grad_norm": 1.6455341310963458, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 18720 + }, + { + "epoch": 0.18721, + "grad_norm": 1.542051687037108, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 18721 + }, + { + "epoch": 0.18722, + "grad_norm": 1.2234910918244877, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 18722 + }, + { + "epoch": 0.18723, + "grad_norm": 1.3644131296122846, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 18723 + }, + { + "epoch": 0.18724, + "grad_norm": 1.189035087310562, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 18724 + }, + { + "epoch": 0.18725, + "grad_norm": 1.20247614424438, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 18725 + }, + { + "epoch": 0.18726, + "grad_norm": 1.304558388125256, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 18726 + }, + { + "epoch": 0.18727, + "grad_norm": 1.2216612061120402, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18727 + }, + { + "epoch": 0.18728, + "grad_norm": 1.5019065755274033, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 18728 + }, + { + "epoch": 0.18729, + "grad_norm": 1.2775538064993097, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 18729 + }, + { + "epoch": 0.1873, + "grad_norm": 1.217586267111922, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 18730 + }, + { + "epoch": 0.18731, + "grad_norm": 1.3627546761119038, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18731 + }, + { + "epoch": 0.18732, + "grad_norm": 1.2637175176188113, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 18732 + }, + { + "epoch": 0.18733, + "grad_norm": 1.2592185013033486, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 18733 + }, + { + "epoch": 0.18734, + "grad_norm": 1.3109836776164643, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18734 + }, + { + "epoch": 0.18735, + "grad_norm": 1.0865367938066122, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 18735 + }, + { + "epoch": 0.18736, + "grad_norm": 1.4847306463982128, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 18736 + }, + { + "epoch": 0.18737, + "grad_norm": 1.2358872148068392, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 18737 + }, + { + "epoch": 0.18738, + "grad_norm": 1.6299832826388085, + "learning_rate": 0.003, + "loss": 4.054, + "step": 18738 + }, + { + "epoch": 0.18739, + "grad_norm": 1.1126078032126734, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 18739 + }, + { + "epoch": 0.1874, + "grad_norm": 1.7625780690960753, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 18740 + }, + { + "epoch": 0.18741, + "grad_norm": 1.1070665620858755, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18741 + }, + { + "epoch": 0.18742, + "grad_norm": 1.3184582434762913, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 18742 + }, + { + "epoch": 0.18743, + "grad_norm": 1.4756540855700473, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 18743 + }, + { + "epoch": 0.18744, + "grad_norm": 1.111862833143605, + "learning_rate": 0.003, + "loss": 4.042, + "step": 18744 + }, + { + "epoch": 0.18745, + "grad_norm": 1.5106703817912401, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 18745 + }, + { + "epoch": 0.18746, + "grad_norm": 1.278505679171913, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 18746 + }, + { + "epoch": 0.18747, + "grad_norm": 1.178612535544609, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 18747 + }, + { + "epoch": 0.18748, + "grad_norm": 1.5664895067817362, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 18748 + }, + { + "epoch": 0.18749, + "grad_norm": 1.0697821444616817, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 18749 + }, + { + "epoch": 0.1875, + "grad_norm": 1.4680988615828654, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 18750 + }, + { + "epoch": 0.18751, + "grad_norm": 1.2208587576394223, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 18751 + }, + { + "epoch": 0.18752, + "grad_norm": 1.2564115783363097, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18752 + }, + { + "epoch": 0.18753, + "grad_norm": 1.2503742948372905, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 18753 + }, + { + "epoch": 0.18754, + "grad_norm": 1.3008110241640158, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 18754 + }, + { + "epoch": 0.18755, + "grad_norm": 1.4017182808098334, + "learning_rate": 0.003, + "loss": 4.048, + "step": 18755 + }, + { + "epoch": 0.18756, + "grad_norm": 1.1919745606184113, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 18756 + }, + { + "epoch": 0.18757, + "grad_norm": 1.4383827916124516, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 18757 + }, + { + "epoch": 0.18758, + "grad_norm": 1.1495324332570196, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18758 + }, + { + "epoch": 0.18759, + "grad_norm": 1.255199727608403, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 18759 + }, + { + "epoch": 0.1876, + "grad_norm": 1.4521720355970977, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18760 + }, + { + "epoch": 0.18761, + "grad_norm": 1.9297282228126031, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 18761 + }, + { + "epoch": 0.18762, + "grad_norm": 1.0610060005854558, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 18762 + }, + { + "epoch": 0.18763, + "grad_norm": 1.4842916207178807, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18763 + }, + { + "epoch": 0.18764, + "grad_norm": 1.2970477632330522, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 18764 + }, + { + "epoch": 0.18765, + "grad_norm": 1.3987190734282884, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 18765 + }, + { + "epoch": 0.18766, + "grad_norm": 1.0971797656792996, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18766 + }, + { + "epoch": 0.18767, + "grad_norm": 1.4950079914981427, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 18767 + }, + { + "epoch": 0.18768, + "grad_norm": 1.073136158356457, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 18768 + }, + { + "epoch": 0.18769, + "grad_norm": 1.4906139086402133, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 18769 + }, + { + "epoch": 0.1877, + "grad_norm": 1.0817221701374877, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 18770 + }, + { + "epoch": 0.18771, + "grad_norm": 1.5447318999519204, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 18771 + }, + { + "epoch": 0.18772, + "grad_norm": 1.399556958314134, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 18772 + }, + { + "epoch": 0.18773, + "grad_norm": 1.5172635406192099, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18773 + }, + { + "epoch": 0.18774, + "grad_norm": 1.308669597093372, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 18774 + }, + { + "epoch": 0.18775, + "grad_norm": 1.3594877497009628, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 18775 + }, + { + "epoch": 0.18776, + "grad_norm": 1.3904321936520263, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 18776 + }, + { + "epoch": 0.18777, + "grad_norm": 1.1499811214556928, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 18777 + }, + { + "epoch": 0.18778, + "grad_norm": 1.3851139884350505, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 18778 + }, + { + "epoch": 0.18779, + "grad_norm": 1.2638471704331713, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18779 + }, + { + "epoch": 0.1878, + "grad_norm": 1.094986868905266, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18780 + }, + { + "epoch": 0.18781, + "grad_norm": 1.5541715116868489, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18781 + }, + { + "epoch": 0.18782, + "grad_norm": 1.4316392568341905, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18782 + }, + { + "epoch": 0.18783, + "grad_norm": 1.4398117848926761, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 18783 + }, + { + "epoch": 0.18784, + "grad_norm": 1.2397901403540124, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18784 + }, + { + "epoch": 0.18785, + "grad_norm": 1.0714101734777415, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 18785 + }, + { + "epoch": 0.18786, + "grad_norm": 1.5671817059944262, + "learning_rate": 0.003, + "loss": 4.063, + "step": 18786 + }, + { + "epoch": 0.18787, + "grad_norm": 0.8931289534617091, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18787 + }, + { + "epoch": 0.18788, + "grad_norm": 1.1129403927682204, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 18788 + }, + { + "epoch": 0.18789, + "grad_norm": 1.163285805940526, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 18789 + }, + { + "epoch": 0.1879, + "grad_norm": 1.2622111365194049, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18790 + }, + { + "epoch": 0.18791, + "grad_norm": 1.372286907108706, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 18791 + }, + { + "epoch": 0.18792, + "grad_norm": 1.302717304034495, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 18792 + }, + { + "epoch": 0.18793, + "grad_norm": 1.5837673759866204, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 18793 + }, + { + "epoch": 0.18794, + "grad_norm": 1.217831445467684, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 18794 + }, + { + "epoch": 0.18795, + "grad_norm": 1.3698513239886787, + "learning_rate": 0.003, + "loss": 4.078, + "step": 18795 + }, + { + "epoch": 0.18796, + "grad_norm": 1.4402458918864507, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 18796 + }, + { + "epoch": 0.18797, + "grad_norm": 1.3757000551166048, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 18797 + }, + { + "epoch": 0.18798, + "grad_norm": 1.2050680386905128, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18798 + }, + { + "epoch": 0.18799, + "grad_norm": 1.4800546409412252, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 18799 + }, + { + "epoch": 0.188, + "grad_norm": 1.3658678067586925, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 18800 + }, + { + "epoch": 0.18801, + "grad_norm": 1.1888682259684493, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 18801 + }, + { + "epoch": 0.18802, + "grad_norm": 1.2904681428387519, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18802 + }, + { + "epoch": 0.18803, + "grad_norm": 1.2784827205733253, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 18803 + }, + { + "epoch": 0.18804, + "grad_norm": 1.2075887275062585, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 18804 + }, + { + "epoch": 0.18805, + "grad_norm": 1.1848057160336685, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 18805 + }, + { + "epoch": 0.18806, + "grad_norm": 1.406609644595669, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 18806 + }, + { + "epoch": 0.18807, + "grad_norm": 1.0692070001217224, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18807 + }, + { + "epoch": 0.18808, + "grad_norm": 1.7056009090567865, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18808 + }, + { + "epoch": 0.18809, + "grad_norm": 1.2059696983682684, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18809 + }, + { + "epoch": 0.1881, + "grad_norm": 1.4356000840083234, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 18810 + }, + { + "epoch": 0.18811, + "grad_norm": 1.192883530310836, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 18811 + }, + { + "epoch": 0.18812, + "grad_norm": 1.589611812778463, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 18812 + }, + { + "epoch": 0.18813, + "grad_norm": 1.1552372944192821, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 18813 + }, + { + "epoch": 0.18814, + "grad_norm": 1.266039013736394, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 18814 + }, + { + "epoch": 0.18815, + "grad_norm": 1.460033899528154, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 18815 + }, + { + "epoch": 0.18816, + "grad_norm": 1.2268797460089496, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 18816 + }, + { + "epoch": 0.18817, + "grad_norm": 1.3280622100516484, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 18817 + }, + { + "epoch": 0.18818, + "grad_norm": 1.3585101052879869, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 18818 + }, + { + "epoch": 0.18819, + "grad_norm": 1.1813963734584156, + "learning_rate": 0.003, + "loss": 4.068, + "step": 18819 + }, + { + "epoch": 0.1882, + "grad_norm": 1.5781017808410744, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18820 + }, + { + "epoch": 0.18821, + "grad_norm": 1.0583102367636024, + "learning_rate": 0.003, + "loss": 4.076, + "step": 18821 + }, + { + "epoch": 0.18822, + "grad_norm": 1.4281895673023899, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 18822 + }, + { + "epoch": 0.18823, + "grad_norm": 1.212896029624748, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 18823 + }, + { + "epoch": 0.18824, + "grad_norm": 1.173722315984396, + "learning_rate": 0.003, + "loss": 4.039, + "step": 18824 + }, + { + "epoch": 0.18825, + "grad_norm": 1.2548827205409894, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 18825 + }, + { + "epoch": 0.18826, + "grad_norm": 1.1790516501163488, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 18826 + }, + { + "epoch": 0.18827, + "grad_norm": 1.63894430103324, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 18827 + }, + { + "epoch": 0.18828, + "grad_norm": 1.3132756256823293, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 18828 + }, + { + "epoch": 0.18829, + "grad_norm": 1.711445563403175, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 18829 + }, + { + "epoch": 0.1883, + "grad_norm": 0.9668264349745779, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18830 + }, + { + "epoch": 0.18831, + "grad_norm": 1.1400888366794983, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 18831 + }, + { + "epoch": 0.18832, + "grad_norm": 1.5372213519021534, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 18832 + }, + { + "epoch": 0.18833, + "grad_norm": 1.310059535276661, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18833 + }, + { + "epoch": 0.18834, + "grad_norm": 1.363509962833014, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 18834 + }, + { + "epoch": 0.18835, + "grad_norm": 1.4236206910331353, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 18835 + }, + { + "epoch": 0.18836, + "grad_norm": 1.3900855704423931, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 18836 + }, + { + "epoch": 0.18837, + "grad_norm": 1.1567946586906692, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 18837 + }, + { + "epoch": 0.18838, + "grad_norm": 1.4650855471840627, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 18838 + }, + { + "epoch": 0.18839, + "grad_norm": 1.2082181065027349, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 18839 + }, + { + "epoch": 0.1884, + "grad_norm": 1.3100334618833482, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 18840 + }, + { + "epoch": 0.18841, + "grad_norm": 1.2898106179869984, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18841 + }, + { + "epoch": 0.18842, + "grad_norm": 1.3447100530946252, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 18842 + }, + { + "epoch": 0.18843, + "grad_norm": 1.4061202076551023, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18843 + }, + { + "epoch": 0.18844, + "grad_norm": 1.4189622114945963, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 18844 + }, + { + "epoch": 0.18845, + "grad_norm": 1.4081270069890108, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 18845 + }, + { + "epoch": 0.18846, + "grad_norm": 1.266243240453215, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18846 + }, + { + "epoch": 0.18847, + "grad_norm": 1.3470242607308647, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 18847 + }, + { + "epoch": 0.18848, + "grad_norm": 1.3204966376292842, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 18848 + }, + { + "epoch": 0.18849, + "grad_norm": 1.5356699983353543, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 18849 + }, + { + "epoch": 0.1885, + "grad_norm": 1.0159267477280214, + "learning_rate": 0.003, + "loss": 4.051, + "step": 18850 + }, + { + "epoch": 0.18851, + "grad_norm": 1.4983569056145334, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 18851 + }, + { + "epoch": 0.18852, + "grad_norm": 1.2704894434822271, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 18852 + }, + { + "epoch": 0.18853, + "grad_norm": 1.3369779797887722, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 18853 + }, + { + "epoch": 0.18854, + "grad_norm": 1.3035438538446587, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 18854 + }, + { + "epoch": 0.18855, + "grad_norm": 1.220516196000558, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18855 + }, + { + "epoch": 0.18856, + "grad_norm": 1.3808656416840817, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 18856 + }, + { + "epoch": 0.18857, + "grad_norm": 1.163717869471003, + "learning_rate": 0.003, + "loss": 4.057, + "step": 18857 + }, + { + "epoch": 0.18858, + "grad_norm": 1.5106764678560334, + "learning_rate": 0.003, + "loss": 4.056, + "step": 18858 + }, + { + "epoch": 0.18859, + "grad_norm": 1.153077200909646, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 18859 + }, + { + "epoch": 0.1886, + "grad_norm": 1.603833161245429, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 18860 + }, + { + "epoch": 0.18861, + "grad_norm": 1.2340982997403442, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 18861 + }, + { + "epoch": 0.18862, + "grad_norm": 1.5173359019104398, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 18862 + }, + { + "epoch": 0.18863, + "grad_norm": 1.1831778405636078, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18863 + }, + { + "epoch": 0.18864, + "grad_norm": 1.2135050745339184, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 18864 + }, + { + "epoch": 0.18865, + "grad_norm": 1.3365944063600408, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18865 + }, + { + "epoch": 0.18866, + "grad_norm": 1.1609925593180206, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18866 + }, + { + "epoch": 0.18867, + "grad_norm": 1.2892142345985795, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 18867 + }, + { + "epoch": 0.18868, + "grad_norm": 1.1249370401528647, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 18868 + }, + { + "epoch": 0.18869, + "grad_norm": 1.5062388002670444, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 18869 + }, + { + "epoch": 0.1887, + "grad_norm": 1.526121507824576, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 18870 + }, + { + "epoch": 0.18871, + "grad_norm": 1.420092814956437, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 18871 + }, + { + "epoch": 0.18872, + "grad_norm": 1.2955615793460946, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 18872 + }, + { + "epoch": 0.18873, + "grad_norm": 1.3497401636635546, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 18873 + }, + { + "epoch": 0.18874, + "grad_norm": 1.3588048814689848, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 18874 + }, + { + "epoch": 0.18875, + "grad_norm": 1.1401782635831865, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 18875 + }, + { + "epoch": 0.18876, + "grad_norm": 1.5040856671285598, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 18876 + }, + { + "epoch": 0.18877, + "grad_norm": 1.153114053739368, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18877 + }, + { + "epoch": 0.18878, + "grad_norm": 1.4876661397458029, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 18878 + }, + { + "epoch": 0.18879, + "grad_norm": 1.1516952756780263, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18879 + }, + { + "epoch": 0.1888, + "grad_norm": 1.5103888507625314, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 18880 + }, + { + "epoch": 0.18881, + "grad_norm": 1.102260589916007, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 18881 + }, + { + "epoch": 0.18882, + "grad_norm": 1.4428875482988193, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 18882 + }, + { + "epoch": 0.18883, + "grad_norm": 1.2718188780518804, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18883 + }, + { + "epoch": 0.18884, + "grad_norm": 1.3704925061754005, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 18884 + }, + { + "epoch": 0.18885, + "grad_norm": 1.4976345293649658, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 18885 + }, + { + "epoch": 0.18886, + "grad_norm": 1.2137579687025972, + "learning_rate": 0.003, + "loss": 4.056, + "step": 18886 + }, + { + "epoch": 0.18887, + "grad_norm": 1.4264884660689312, + "learning_rate": 0.003, + "loss": 4.071, + "step": 18887 + }, + { + "epoch": 0.18888, + "grad_norm": 1.0713536509148665, + "learning_rate": 0.003, + "loss": 4.076, + "step": 18888 + }, + { + "epoch": 0.18889, + "grad_norm": 1.5868036919963195, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 18889 + }, + { + "epoch": 0.1889, + "grad_norm": 1.180432126330319, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 18890 + }, + { + "epoch": 0.18891, + "grad_norm": 1.521492894557538, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 18891 + }, + { + "epoch": 0.18892, + "grad_norm": 1.1078770928910782, + "learning_rate": 0.003, + "loss": 4.069, + "step": 18892 + }, + { + "epoch": 0.18893, + "grad_norm": 1.7356931265094164, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 18893 + }, + { + "epoch": 0.18894, + "grad_norm": 1.317253082983154, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 18894 + }, + { + "epoch": 0.18895, + "grad_norm": 1.1648419331625475, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 18895 + }, + { + "epoch": 0.18896, + "grad_norm": 1.250255263445163, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 18896 + }, + { + "epoch": 0.18897, + "grad_norm": 1.4344943793092255, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 18897 + }, + { + "epoch": 0.18898, + "grad_norm": 1.3279246217317304, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 18898 + }, + { + "epoch": 0.18899, + "grad_norm": 1.1889584260873742, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 18899 + }, + { + "epoch": 0.189, + "grad_norm": 1.316494368314, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 18900 + }, + { + "epoch": 0.18901, + "grad_norm": 1.2230937412314116, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 18901 + }, + { + "epoch": 0.18902, + "grad_norm": 1.3945763924516088, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 18902 + }, + { + "epoch": 0.18903, + "grad_norm": 1.2739696174630883, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 18903 + }, + { + "epoch": 0.18904, + "grad_norm": 1.5827069235789044, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 18904 + }, + { + "epoch": 0.18905, + "grad_norm": 1.3636490487827087, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 18905 + }, + { + "epoch": 0.18906, + "grad_norm": 1.3408230077629284, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 18906 + }, + { + "epoch": 0.18907, + "grad_norm": 1.2225650484016408, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 18907 + }, + { + "epoch": 0.18908, + "grad_norm": 1.4123102155843825, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 18908 + }, + { + "epoch": 0.18909, + "grad_norm": 1.3793049252028673, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 18909 + }, + { + "epoch": 0.1891, + "grad_norm": 1.3107955629341879, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 18910 + }, + { + "epoch": 0.18911, + "grad_norm": 1.228759283452415, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 18911 + }, + { + "epoch": 0.18912, + "grad_norm": 1.4299315989229704, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 18912 + }, + { + "epoch": 0.18913, + "grad_norm": 1.1500243667154835, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 18913 + }, + { + "epoch": 0.18914, + "grad_norm": 1.2391305076804604, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 18914 + }, + { + "epoch": 0.18915, + "grad_norm": 1.383468763742884, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 18915 + }, + { + "epoch": 0.18916, + "grad_norm": 1.1880879138668168, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 18916 + }, + { + "epoch": 0.18917, + "grad_norm": 1.5951028679607262, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 18917 + }, + { + "epoch": 0.18918, + "grad_norm": 1.2260422838236318, + "learning_rate": 0.003, + "loss": 4.044, + "step": 18918 + }, + { + "epoch": 0.18919, + "grad_norm": 1.375698077238036, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 18919 + }, + { + "epoch": 0.1892, + "grad_norm": 1.262268901110128, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 18920 + }, + { + "epoch": 0.18921, + "grad_norm": 1.235518346439444, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 18921 + }, + { + "epoch": 0.18922, + "grad_norm": 1.447977057278561, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 18922 + }, + { + "epoch": 0.18923, + "grad_norm": 1.261849012928809, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 18923 + }, + { + "epoch": 0.18924, + "grad_norm": 1.5017822769392577, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 18924 + }, + { + "epoch": 0.18925, + "grad_norm": 1.3941361919123514, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 18925 + }, + { + "epoch": 0.18926, + "grad_norm": 1.5368017967326983, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 18926 + }, + { + "epoch": 0.18927, + "grad_norm": 1.1538779990492576, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 18927 + }, + { + "epoch": 0.18928, + "grad_norm": 1.5444160847420918, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 18928 + }, + { + "epoch": 0.18929, + "grad_norm": 0.9385130249989668, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 18929 + }, + { + "epoch": 0.1893, + "grad_norm": 1.3564797100958084, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 18930 + }, + { + "epoch": 0.18931, + "grad_norm": 1.0335793616368942, + "learning_rate": 0.003, + "loss": 4.064, + "step": 18931 + }, + { + "epoch": 0.18932, + "grad_norm": 1.3010390289418123, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 18932 + }, + { + "epoch": 0.18933, + "grad_norm": 1.437018943007172, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18933 + }, + { + "epoch": 0.18934, + "grad_norm": 1.4044142852135095, + "learning_rate": 0.003, + "loss": 4.071, + "step": 18934 + }, + { + "epoch": 0.18935, + "grad_norm": 1.4813851489409189, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 18935 + }, + { + "epoch": 0.18936, + "grad_norm": 0.9988193921991353, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 18936 + }, + { + "epoch": 0.18937, + "grad_norm": 1.3695145211207798, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 18937 + }, + { + "epoch": 0.18938, + "grad_norm": 1.2883443280092062, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 18938 + }, + { + "epoch": 0.18939, + "grad_norm": 1.1970784165724995, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 18939 + }, + { + "epoch": 0.1894, + "grad_norm": 1.2735363899786794, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 18940 + }, + { + "epoch": 0.18941, + "grad_norm": 1.3485978322242238, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 18941 + }, + { + "epoch": 0.18942, + "grad_norm": 0.9973824337445716, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 18942 + }, + { + "epoch": 0.18943, + "grad_norm": 1.7790042413316978, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 18943 + }, + { + "epoch": 0.18944, + "grad_norm": 1.4669139722319338, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 18944 + }, + { + "epoch": 0.18945, + "grad_norm": 1.4581928468055323, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 18945 + }, + { + "epoch": 0.18946, + "grad_norm": 1.1671636085692663, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 18946 + }, + { + "epoch": 0.18947, + "grad_norm": 1.3777204974748212, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 18947 + }, + { + "epoch": 0.18948, + "grad_norm": 1.1778477242902337, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 18948 + }, + { + "epoch": 0.18949, + "grad_norm": 1.3936343034445549, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 18949 + }, + { + "epoch": 0.1895, + "grad_norm": 1.222772542255314, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 18950 + }, + { + "epoch": 0.18951, + "grad_norm": 1.2931222145801997, + "learning_rate": 0.003, + "loss": 4.066, + "step": 18951 + }, + { + "epoch": 0.18952, + "grad_norm": 1.0958641638209, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 18952 + }, + { + "epoch": 0.18953, + "grad_norm": 1.497273939684197, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 18953 + }, + { + "epoch": 0.18954, + "grad_norm": 1.336175191600886, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18954 + }, + { + "epoch": 0.18955, + "grad_norm": 1.5511757924893705, + "learning_rate": 0.003, + "loss": 4.045, + "step": 18955 + }, + { + "epoch": 0.18956, + "grad_norm": 1.12151388342686, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 18956 + }, + { + "epoch": 0.18957, + "grad_norm": 1.2714774177311274, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18957 + }, + { + "epoch": 0.18958, + "grad_norm": 1.806335271820712, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 18958 + }, + { + "epoch": 0.18959, + "grad_norm": 1.1981557925984667, + "learning_rate": 0.003, + "loss": 4.031, + "step": 18959 + }, + { + "epoch": 0.1896, + "grad_norm": 1.2877466637724115, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 18960 + }, + { + "epoch": 0.18961, + "grad_norm": 1.2289050495891696, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 18961 + }, + { + "epoch": 0.18962, + "grad_norm": 1.3242121522869408, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 18962 + }, + { + "epoch": 0.18963, + "grad_norm": 1.2268292390487603, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 18963 + }, + { + "epoch": 0.18964, + "grad_norm": 1.4286199540648732, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 18964 + }, + { + "epoch": 0.18965, + "grad_norm": 1.1307354889118555, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 18965 + }, + { + "epoch": 0.18966, + "grad_norm": 1.3616380925033746, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 18966 + }, + { + "epoch": 0.18967, + "grad_norm": 1.3270887278372299, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 18967 + }, + { + "epoch": 0.18968, + "grad_norm": 1.180291146237884, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 18968 + }, + { + "epoch": 0.18969, + "grad_norm": 1.4248134973281465, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 18969 + }, + { + "epoch": 0.1897, + "grad_norm": 1.857430428677563, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 18970 + }, + { + "epoch": 0.18971, + "grad_norm": 1.513573264824216, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 18971 + }, + { + "epoch": 0.18972, + "grad_norm": 1.2313827860976658, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 18972 + }, + { + "epoch": 0.18973, + "grad_norm": 1.567017247104148, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18973 + }, + { + "epoch": 0.18974, + "grad_norm": 1.0781847216389804, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 18974 + }, + { + "epoch": 0.18975, + "grad_norm": 1.2342507716478786, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 18975 + }, + { + "epoch": 0.18976, + "grad_norm": 1.4390172817298563, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 18976 + }, + { + "epoch": 0.18977, + "grad_norm": 1.119764135955646, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 18977 + }, + { + "epoch": 0.18978, + "grad_norm": 1.4875556521970705, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 18978 + }, + { + "epoch": 0.18979, + "grad_norm": 1.3524860210818532, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 18979 + }, + { + "epoch": 0.1898, + "grad_norm": 1.0635357406263535, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 18980 + }, + { + "epoch": 0.18981, + "grad_norm": 1.3099436010284684, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 18981 + }, + { + "epoch": 0.18982, + "grad_norm": 1.1859148488860554, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 18982 + }, + { + "epoch": 0.18983, + "grad_norm": 1.6161576957549602, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 18983 + }, + { + "epoch": 0.18984, + "grad_norm": 1.8139644629208311, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 18984 + }, + { + "epoch": 0.18985, + "grad_norm": 1.0784425662683723, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 18985 + }, + { + "epoch": 0.18986, + "grad_norm": 1.363183551893584, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 18986 + }, + { + "epoch": 0.18987, + "grad_norm": 1.2599305508297793, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 18987 + }, + { + "epoch": 0.18988, + "grad_norm": 1.4277124939895625, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 18988 + }, + { + "epoch": 0.18989, + "grad_norm": 1.3098668231382307, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 18989 + }, + { + "epoch": 0.1899, + "grad_norm": 1.5675414246730872, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 18990 + }, + { + "epoch": 0.18991, + "grad_norm": 1.240194226915464, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 18991 + }, + { + "epoch": 0.18992, + "grad_norm": 1.1193047959233318, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 18992 + }, + { + "epoch": 0.18993, + "grad_norm": 1.420012937103303, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 18993 + }, + { + "epoch": 0.18994, + "grad_norm": 1.1123855253231913, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 18994 + }, + { + "epoch": 0.18995, + "grad_norm": 1.2576329907124797, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 18995 + }, + { + "epoch": 0.18996, + "grad_norm": 1.2610917729669, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 18996 + }, + { + "epoch": 0.18997, + "grad_norm": 1.3324235682068313, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 18997 + }, + { + "epoch": 0.18998, + "grad_norm": 1.2416667836230513, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 18998 + }, + { + "epoch": 0.18999, + "grad_norm": 1.4512750335089, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 18999 + }, + { + "epoch": 0.19, + "grad_norm": 1.2324685620970226, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 19000 + }, + { + "epoch": 0.19001, + "grad_norm": 1.5577462599745568, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 19001 + }, + { + "epoch": 0.19002, + "grad_norm": 1.2889729802796936, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 19002 + }, + { + "epoch": 0.19003, + "grad_norm": 1.233507256766918, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19003 + }, + { + "epoch": 0.19004, + "grad_norm": 1.1803096529839874, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 19004 + }, + { + "epoch": 0.19005, + "grad_norm": 1.1170026976717313, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 19005 + }, + { + "epoch": 0.19006, + "grad_norm": 1.325523253601187, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 19006 + }, + { + "epoch": 0.19007, + "grad_norm": 1.2934096707555685, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 19007 + }, + { + "epoch": 0.19008, + "grad_norm": 1.4935043714232368, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 19008 + }, + { + "epoch": 0.19009, + "grad_norm": 1.3216276747957934, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 19009 + }, + { + "epoch": 0.1901, + "grad_norm": 1.2591839637952178, + "learning_rate": 0.003, + "loss": 4.051, + "step": 19010 + }, + { + "epoch": 0.19011, + "grad_norm": 1.6761998989258382, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 19011 + }, + { + "epoch": 0.19012, + "grad_norm": 1.1244001332035305, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 19012 + }, + { + "epoch": 0.19013, + "grad_norm": 1.5815884330810812, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 19013 + }, + { + "epoch": 0.19014, + "grad_norm": 1.5778195952980243, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19014 + }, + { + "epoch": 0.19015, + "grad_norm": 1.153662316779612, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 19015 + }, + { + "epoch": 0.19016, + "grad_norm": 1.1217013138145722, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19016 + }, + { + "epoch": 0.19017, + "grad_norm": 1.3396995051527274, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 19017 + }, + { + "epoch": 0.19018, + "grad_norm": 1.3900040315464735, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 19018 + }, + { + "epoch": 0.19019, + "grad_norm": 1.392132345133188, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 19019 + }, + { + "epoch": 0.1902, + "grad_norm": 1.4664707424192767, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 19020 + }, + { + "epoch": 0.19021, + "grad_norm": 1.0816735042234542, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 19021 + }, + { + "epoch": 0.19022, + "grad_norm": 1.222139850489423, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 19022 + }, + { + "epoch": 0.19023, + "grad_norm": 1.3113447127507638, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 19023 + }, + { + "epoch": 0.19024, + "grad_norm": 1.41514397939476, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 19024 + }, + { + "epoch": 0.19025, + "grad_norm": 0.9802130337406161, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 19025 + }, + { + "epoch": 0.19026, + "grad_norm": 1.5998213443286031, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 19026 + }, + { + "epoch": 0.19027, + "grad_norm": 1.1665135813741723, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 19027 + }, + { + "epoch": 0.19028, + "grad_norm": 1.4020630133015164, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 19028 + }, + { + "epoch": 0.19029, + "grad_norm": 1.4448346034862, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 19029 + }, + { + "epoch": 0.1903, + "grad_norm": 1.0550181708031718, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19030 + }, + { + "epoch": 0.19031, + "grad_norm": 1.495096570441009, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19031 + }, + { + "epoch": 0.19032, + "grad_norm": 1.1932281477677271, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 19032 + }, + { + "epoch": 0.19033, + "grad_norm": 1.5091186667250431, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 19033 + }, + { + "epoch": 0.19034, + "grad_norm": 1.3843832633303428, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19034 + }, + { + "epoch": 0.19035, + "grad_norm": 1.314486318500889, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19035 + }, + { + "epoch": 0.19036, + "grad_norm": 1.2069363457552786, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 19036 + }, + { + "epoch": 0.19037, + "grad_norm": 1.3615790478931433, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 19037 + }, + { + "epoch": 0.19038, + "grad_norm": 1.193499137863031, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 19038 + }, + { + "epoch": 0.19039, + "grad_norm": 1.2939501145230716, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 19039 + }, + { + "epoch": 0.1904, + "grad_norm": 1.2528639930786927, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 19040 + }, + { + "epoch": 0.19041, + "grad_norm": 1.471674616815551, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19041 + }, + { + "epoch": 0.19042, + "grad_norm": 1.314049431909399, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19042 + }, + { + "epoch": 0.19043, + "grad_norm": 1.3624689771020175, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19043 + }, + { + "epoch": 0.19044, + "grad_norm": 1.4134190169893828, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 19044 + }, + { + "epoch": 0.19045, + "grad_norm": 1.0960238181783737, + "learning_rate": 0.003, + "loss": 4.07, + "step": 19045 + }, + { + "epoch": 0.19046, + "grad_norm": 1.5017184629957616, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19046 + }, + { + "epoch": 0.19047, + "grad_norm": 1.232097844273087, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 19047 + }, + { + "epoch": 0.19048, + "grad_norm": 1.3400956265663477, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19048 + }, + { + "epoch": 0.19049, + "grad_norm": 1.2470828259434086, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 19049 + }, + { + "epoch": 0.1905, + "grad_norm": 1.5793037547965862, + "learning_rate": 0.003, + "loss": 4.088, + "step": 19050 + }, + { + "epoch": 0.19051, + "grad_norm": 1.338329176891238, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 19051 + }, + { + "epoch": 0.19052, + "grad_norm": 1.1619816182998715, + "learning_rate": 0.003, + "loss": 4.068, + "step": 19052 + }, + { + "epoch": 0.19053, + "grad_norm": 1.386888683759865, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 19053 + }, + { + "epoch": 0.19054, + "grad_norm": 1.1400577701730739, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 19054 + }, + { + "epoch": 0.19055, + "grad_norm": 1.5485345322855846, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 19055 + }, + { + "epoch": 0.19056, + "grad_norm": 1.1029821679021268, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 19056 + }, + { + "epoch": 0.19057, + "grad_norm": 1.449446935040205, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19057 + }, + { + "epoch": 0.19058, + "grad_norm": 1.3637159846302798, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 19058 + }, + { + "epoch": 0.19059, + "grad_norm": 1.436568727485353, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19059 + }, + { + "epoch": 0.1906, + "grad_norm": 1.0690561626750321, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19060 + }, + { + "epoch": 0.19061, + "grad_norm": 1.602754281179194, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 19061 + }, + { + "epoch": 0.19062, + "grad_norm": 1.137154741975126, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 19062 + }, + { + "epoch": 0.19063, + "grad_norm": 1.4750047660200645, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 19063 + }, + { + "epoch": 0.19064, + "grad_norm": 1.0672597314943837, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 19064 + }, + { + "epoch": 0.19065, + "grad_norm": 1.6459411674301923, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19065 + }, + { + "epoch": 0.19066, + "grad_norm": 1.1575056516494613, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19066 + }, + { + "epoch": 0.19067, + "grad_norm": 1.232617210848777, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 19067 + }, + { + "epoch": 0.19068, + "grad_norm": 1.3765978890878052, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 19068 + }, + { + "epoch": 0.19069, + "grad_norm": 1.4613355184334849, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 19069 + }, + { + "epoch": 0.1907, + "grad_norm": 1.3304262962194269, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 19070 + }, + { + "epoch": 0.19071, + "grad_norm": 1.3375188461800747, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 19071 + }, + { + "epoch": 0.19072, + "grad_norm": 1.5392331002690696, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19072 + }, + { + "epoch": 0.19073, + "grad_norm": 1.4565796580099446, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 19073 + }, + { + "epoch": 0.19074, + "grad_norm": 1.2720043502537839, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 19074 + }, + { + "epoch": 0.19075, + "grad_norm": 1.3716543617079564, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 19075 + }, + { + "epoch": 0.19076, + "grad_norm": 1.3971309426935, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19076 + }, + { + "epoch": 0.19077, + "grad_norm": 1.30982799094468, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 19077 + }, + { + "epoch": 0.19078, + "grad_norm": 1.1911141610284892, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 19078 + }, + { + "epoch": 0.19079, + "grad_norm": 1.2083749238417725, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 19079 + }, + { + "epoch": 0.1908, + "grad_norm": 1.2449723095561924, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19080 + }, + { + "epoch": 0.19081, + "grad_norm": 1.3998045528501801, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 19081 + }, + { + "epoch": 0.19082, + "grad_norm": 1.4531284576949606, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19082 + }, + { + "epoch": 0.19083, + "grad_norm": 1.064278230314246, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19083 + }, + { + "epoch": 0.19084, + "grad_norm": 1.5083208229316238, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 19084 + }, + { + "epoch": 0.19085, + "grad_norm": 1.5000500657542444, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 19085 + }, + { + "epoch": 0.19086, + "grad_norm": 1.0974791363157106, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 19086 + }, + { + "epoch": 0.19087, + "grad_norm": 1.4513655729032795, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 19087 + }, + { + "epoch": 0.19088, + "grad_norm": 1.2977784897854003, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 19088 + }, + { + "epoch": 0.19089, + "grad_norm": 1.1013679680267832, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 19089 + }, + { + "epoch": 0.1909, + "grad_norm": 1.4129522572596733, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19090 + }, + { + "epoch": 0.19091, + "grad_norm": 1.207696078897892, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19091 + }, + { + "epoch": 0.19092, + "grad_norm": 1.4232598149015803, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 19092 + }, + { + "epoch": 0.19093, + "grad_norm": 1.1972431576898495, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19093 + }, + { + "epoch": 0.19094, + "grad_norm": 1.5183181645421888, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 19094 + }, + { + "epoch": 0.19095, + "grad_norm": 1.221872072035716, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19095 + }, + { + "epoch": 0.19096, + "grad_norm": 1.4396452999024958, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 19096 + }, + { + "epoch": 0.19097, + "grad_norm": 1.2246180936584108, + "learning_rate": 0.003, + "loss": 4.076, + "step": 19097 + }, + { + "epoch": 0.19098, + "grad_norm": 1.4125094881948392, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19098 + }, + { + "epoch": 0.19099, + "grad_norm": 1.190535711223137, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19099 + }, + { + "epoch": 0.191, + "grad_norm": 1.320284979939749, + "learning_rate": 0.003, + "loss": 4.094, + "step": 19100 + }, + { + "epoch": 0.19101, + "grad_norm": 1.2820570689852735, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 19101 + }, + { + "epoch": 0.19102, + "grad_norm": 1.640797367448142, + "learning_rate": 0.003, + "loss": 4.057, + "step": 19102 + }, + { + "epoch": 0.19103, + "grad_norm": 1.7711658876323253, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19103 + }, + { + "epoch": 0.19104, + "grad_norm": 1.1898220312387047, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 19104 + }, + { + "epoch": 0.19105, + "grad_norm": 1.242179886486909, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 19105 + }, + { + "epoch": 0.19106, + "grad_norm": 1.3358435037354868, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 19106 + }, + { + "epoch": 0.19107, + "grad_norm": 1.3206693791196868, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 19107 + }, + { + "epoch": 0.19108, + "grad_norm": 1.3219251111714507, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 19108 + }, + { + "epoch": 0.19109, + "grad_norm": 1.3692187282500543, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 19109 + }, + { + "epoch": 0.1911, + "grad_norm": 1.2516639706729331, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19110 + }, + { + "epoch": 0.19111, + "grad_norm": 1.3579748756501073, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 19111 + }, + { + "epoch": 0.19112, + "grad_norm": 1.2901685774005183, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 19112 + }, + { + "epoch": 0.19113, + "grad_norm": 1.3695313287734234, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 19113 + }, + { + "epoch": 0.19114, + "grad_norm": 1.1943068768193779, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 19114 + }, + { + "epoch": 0.19115, + "grad_norm": 1.3324190801436753, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 19115 + }, + { + "epoch": 0.19116, + "grad_norm": 1.0481218005906356, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19116 + }, + { + "epoch": 0.19117, + "grad_norm": 1.2175372768402644, + "learning_rate": 0.003, + "loss": 4.047, + "step": 19117 + }, + { + "epoch": 0.19118, + "grad_norm": 1.1851769264419512, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 19118 + }, + { + "epoch": 0.19119, + "grad_norm": 1.6243204774157487, + "learning_rate": 0.003, + "loss": 4.114, + "step": 19119 + }, + { + "epoch": 0.1912, + "grad_norm": 1.134277528181548, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 19120 + }, + { + "epoch": 0.19121, + "grad_norm": 1.3437022589300889, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 19121 + }, + { + "epoch": 0.19122, + "grad_norm": 1.3333603930491609, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 19122 + }, + { + "epoch": 0.19123, + "grad_norm": 1.2745079718137873, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19123 + }, + { + "epoch": 0.19124, + "grad_norm": 1.1960491080657072, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 19124 + }, + { + "epoch": 0.19125, + "grad_norm": 1.3172844462471818, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 19125 + }, + { + "epoch": 0.19126, + "grad_norm": 1.4834170903173642, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 19126 + }, + { + "epoch": 0.19127, + "grad_norm": 1.812018123397985, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 19127 + }, + { + "epoch": 0.19128, + "grad_norm": 1.1098226632687962, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19128 + }, + { + "epoch": 0.19129, + "grad_norm": 1.344914180637586, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19129 + }, + { + "epoch": 0.1913, + "grad_norm": 1.2581708893061798, + "learning_rate": 0.003, + "loss": 4.082, + "step": 19130 + }, + { + "epoch": 0.19131, + "grad_norm": 1.240574980726106, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 19131 + }, + { + "epoch": 0.19132, + "grad_norm": 1.6100387528656463, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19132 + }, + { + "epoch": 0.19133, + "grad_norm": 1.242291472166886, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19133 + }, + { + "epoch": 0.19134, + "grad_norm": 1.4652522360740017, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 19134 + }, + { + "epoch": 0.19135, + "grad_norm": 1.3220586428121326, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 19135 + }, + { + "epoch": 0.19136, + "grad_norm": 1.1979879985446065, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 19136 + }, + { + "epoch": 0.19137, + "grad_norm": 1.2245714794934948, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 19137 + }, + { + "epoch": 0.19138, + "grad_norm": 1.2533839335343562, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19138 + }, + { + "epoch": 0.19139, + "grad_norm": 1.2597911924182927, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 19139 + }, + { + "epoch": 0.1914, + "grad_norm": 1.0086352156745937, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 19140 + }, + { + "epoch": 0.19141, + "grad_norm": 1.5899559027790509, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 19141 + }, + { + "epoch": 0.19142, + "grad_norm": 1.5939739338876755, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19142 + }, + { + "epoch": 0.19143, + "grad_norm": 1.3287884363381772, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19143 + }, + { + "epoch": 0.19144, + "grad_norm": 1.3587858504048773, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19144 + }, + { + "epoch": 0.19145, + "grad_norm": 1.3404937183657344, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 19145 + }, + { + "epoch": 0.19146, + "grad_norm": 1.270922752836181, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19146 + }, + { + "epoch": 0.19147, + "grad_norm": 1.2147556481989903, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19147 + }, + { + "epoch": 0.19148, + "grad_norm": 1.3497960230796922, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19148 + }, + { + "epoch": 0.19149, + "grad_norm": 1.256723566268377, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19149 + }, + { + "epoch": 0.1915, + "grad_norm": 1.3643278964494576, + "learning_rate": 0.003, + "loss": 4.055, + "step": 19150 + }, + { + "epoch": 0.19151, + "grad_norm": 1.2185081050635174, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 19151 + }, + { + "epoch": 0.19152, + "grad_norm": 1.317530622128689, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 19152 + }, + { + "epoch": 0.19153, + "grad_norm": 1.2245685697754942, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19153 + }, + { + "epoch": 0.19154, + "grad_norm": 1.2319047829270333, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 19154 + }, + { + "epoch": 0.19155, + "grad_norm": 1.3030123854433888, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 19155 + }, + { + "epoch": 0.19156, + "grad_norm": 1.1758000665998163, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19156 + }, + { + "epoch": 0.19157, + "grad_norm": 1.3547531756373572, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19157 + }, + { + "epoch": 0.19158, + "grad_norm": 1.4880671874548606, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 19158 + }, + { + "epoch": 0.19159, + "grad_norm": 1.2509190691711998, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 19159 + }, + { + "epoch": 0.1916, + "grad_norm": 1.3289551859332054, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 19160 + }, + { + "epoch": 0.19161, + "grad_norm": 1.5236236955712956, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 19161 + }, + { + "epoch": 0.19162, + "grad_norm": 1.3370152625573213, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 19162 + }, + { + "epoch": 0.19163, + "grad_norm": 1.6548746772311307, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19163 + }, + { + "epoch": 0.19164, + "grad_norm": 1.5183876875667157, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 19164 + }, + { + "epoch": 0.19165, + "grad_norm": 1.3947780331195114, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 19165 + }, + { + "epoch": 0.19166, + "grad_norm": 1.2883469168643833, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19166 + }, + { + "epoch": 0.19167, + "grad_norm": 1.2956180033322826, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 19167 + }, + { + "epoch": 0.19168, + "grad_norm": 1.3990904680365333, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 19168 + }, + { + "epoch": 0.19169, + "grad_norm": 1.1364576534154611, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 19169 + }, + { + "epoch": 0.1917, + "grad_norm": 1.6642795045391565, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 19170 + }, + { + "epoch": 0.19171, + "grad_norm": 1.2925604382023206, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 19171 + }, + { + "epoch": 0.19172, + "grad_norm": 1.3464205038024264, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 19172 + }, + { + "epoch": 0.19173, + "grad_norm": 1.3202152463158157, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19173 + }, + { + "epoch": 0.19174, + "grad_norm": 1.382927722314029, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 19174 + }, + { + "epoch": 0.19175, + "grad_norm": 1.422565007939475, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 19175 + }, + { + "epoch": 0.19176, + "grad_norm": 1.677057405662981, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 19176 + }, + { + "epoch": 0.19177, + "grad_norm": 1.2214924954573159, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19177 + }, + { + "epoch": 0.19178, + "grad_norm": 1.3088438877680733, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19178 + }, + { + "epoch": 0.19179, + "grad_norm": 1.5170282665335622, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 19179 + }, + { + "epoch": 0.1918, + "grad_norm": 1.2040190205765118, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 19180 + }, + { + "epoch": 0.19181, + "grad_norm": 1.3822570608840876, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 19181 + }, + { + "epoch": 0.19182, + "grad_norm": 1.172484815976945, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 19182 + }, + { + "epoch": 0.19183, + "grad_norm": 1.274388763595249, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 19183 + }, + { + "epoch": 0.19184, + "grad_norm": 1.2924607758107507, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 19184 + }, + { + "epoch": 0.19185, + "grad_norm": 1.2128103145138238, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 19185 + }, + { + "epoch": 0.19186, + "grad_norm": 1.3739857604766152, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 19186 + }, + { + "epoch": 0.19187, + "grad_norm": 1.1002419192622068, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19187 + }, + { + "epoch": 0.19188, + "grad_norm": 1.297982852715103, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19188 + }, + { + "epoch": 0.19189, + "grad_norm": 1.3339832824292281, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 19189 + }, + { + "epoch": 0.1919, + "grad_norm": 1.0348721776220091, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 19190 + }, + { + "epoch": 0.19191, + "grad_norm": 1.6570408605160454, + "learning_rate": 0.003, + "loss": 4.071, + "step": 19191 + }, + { + "epoch": 0.19192, + "grad_norm": 1.3945724289140948, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 19192 + }, + { + "epoch": 0.19193, + "grad_norm": 1.5912266691413073, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19193 + }, + { + "epoch": 0.19194, + "grad_norm": 1.4058863699195427, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 19194 + }, + { + "epoch": 0.19195, + "grad_norm": 1.356988434286127, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 19195 + }, + { + "epoch": 0.19196, + "grad_norm": 1.1528145797759937, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19196 + }, + { + "epoch": 0.19197, + "grad_norm": 1.5981012584383123, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19197 + }, + { + "epoch": 0.19198, + "grad_norm": 0.9630358323900087, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19198 + }, + { + "epoch": 0.19199, + "grad_norm": 1.5678787468061148, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 19199 + }, + { + "epoch": 0.192, + "grad_norm": 1.162720578822619, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 19200 + }, + { + "epoch": 0.19201, + "grad_norm": 1.554253879167807, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 19201 + }, + { + "epoch": 0.19202, + "grad_norm": 1.3562060036479024, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 19202 + }, + { + "epoch": 0.19203, + "grad_norm": 1.572129714906069, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 19203 + }, + { + "epoch": 0.19204, + "grad_norm": 1.108132358920365, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 19204 + }, + { + "epoch": 0.19205, + "grad_norm": 1.4562272560833824, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 19205 + }, + { + "epoch": 0.19206, + "grad_norm": 1.1248392481966405, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 19206 + }, + { + "epoch": 0.19207, + "grad_norm": 1.347347178032185, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 19207 + }, + { + "epoch": 0.19208, + "grad_norm": 1.2569962037026983, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19208 + }, + { + "epoch": 0.19209, + "grad_norm": 1.3461303832239604, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 19209 + }, + { + "epoch": 0.1921, + "grad_norm": 0.9850114287517585, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19210 + }, + { + "epoch": 0.19211, + "grad_norm": 1.4005865113973734, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19211 + }, + { + "epoch": 0.19212, + "grad_norm": 1.3014172439729073, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 19212 + }, + { + "epoch": 0.19213, + "grad_norm": 1.3378102945952632, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 19213 + }, + { + "epoch": 0.19214, + "grad_norm": 1.3952909056018472, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19214 + }, + { + "epoch": 0.19215, + "grad_norm": 1.611663758916327, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 19215 + }, + { + "epoch": 0.19216, + "grad_norm": 0.8963660316350982, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 19216 + }, + { + "epoch": 0.19217, + "grad_norm": 1.1968197183576728, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 19217 + }, + { + "epoch": 0.19218, + "grad_norm": 1.288873508937206, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 19218 + }, + { + "epoch": 0.19219, + "grad_norm": 1.2116427947856974, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19219 + }, + { + "epoch": 0.1922, + "grad_norm": 1.40961153431515, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 19220 + }, + { + "epoch": 0.19221, + "grad_norm": 1.3174703939528276, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 19221 + }, + { + "epoch": 0.19222, + "grad_norm": 1.1947735338993128, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 19222 + }, + { + "epoch": 0.19223, + "grad_norm": 1.3706485111315745, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 19223 + }, + { + "epoch": 0.19224, + "grad_norm": 1.2645512304572817, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 19224 + }, + { + "epoch": 0.19225, + "grad_norm": 1.1543151542958139, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19225 + }, + { + "epoch": 0.19226, + "grad_norm": 1.291486822731222, + "learning_rate": 0.003, + "loss": 4.053, + "step": 19226 + }, + { + "epoch": 0.19227, + "grad_norm": 1.5269001763632155, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19227 + }, + { + "epoch": 0.19228, + "grad_norm": 1.3782331321295478, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19228 + }, + { + "epoch": 0.19229, + "grad_norm": 1.3237198432938115, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19229 + }, + { + "epoch": 0.1923, + "grad_norm": 1.323968945621076, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19230 + }, + { + "epoch": 0.19231, + "grad_norm": 1.3349763890268658, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 19231 + }, + { + "epoch": 0.19232, + "grad_norm": 1.3661924478328742, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 19232 + }, + { + "epoch": 0.19233, + "grad_norm": 1.194365790073095, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 19233 + }, + { + "epoch": 0.19234, + "grad_norm": 1.3965787017862838, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19234 + }, + { + "epoch": 0.19235, + "grad_norm": 1.196036206103378, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 19235 + }, + { + "epoch": 0.19236, + "grad_norm": 1.3711815439105595, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 19236 + }, + { + "epoch": 0.19237, + "grad_norm": 1.141165924450159, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 19237 + }, + { + "epoch": 0.19238, + "grad_norm": 1.308345863022952, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 19238 + }, + { + "epoch": 0.19239, + "grad_norm": 1.3249441038885172, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 19239 + }, + { + "epoch": 0.1924, + "grad_norm": 1.3966021263135535, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19240 + }, + { + "epoch": 0.19241, + "grad_norm": 1.3663512273311271, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 19241 + }, + { + "epoch": 0.19242, + "grad_norm": 1.2032887696702281, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19242 + }, + { + "epoch": 0.19243, + "grad_norm": 1.3339164408073845, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 19243 + }, + { + "epoch": 0.19244, + "grad_norm": 1.4303815180109472, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19244 + }, + { + "epoch": 0.19245, + "grad_norm": 1.423193016673088, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19245 + }, + { + "epoch": 0.19246, + "grad_norm": 1.4514350895399684, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 19246 + }, + { + "epoch": 0.19247, + "grad_norm": 1.495226265941209, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 19247 + }, + { + "epoch": 0.19248, + "grad_norm": 1.2397736348773147, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 19248 + }, + { + "epoch": 0.19249, + "grad_norm": 1.374416195938836, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 19249 + }, + { + "epoch": 0.1925, + "grad_norm": 1.3189947596094718, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19250 + }, + { + "epoch": 0.19251, + "grad_norm": 1.3593757512777345, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 19251 + }, + { + "epoch": 0.19252, + "grad_norm": 1.2451545340567312, + "learning_rate": 0.003, + "loss": 4.043, + "step": 19252 + }, + { + "epoch": 0.19253, + "grad_norm": 1.4531214541051583, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19253 + }, + { + "epoch": 0.19254, + "grad_norm": 1.1835021545730156, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 19254 + }, + { + "epoch": 0.19255, + "grad_norm": 1.576806190726499, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 19255 + }, + { + "epoch": 0.19256, + "grad_norm": 0.9618777802716235, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 19256 + }, + { + "epoch": 0.19257, + "grad_norm": 1.3588331222703336, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 19257 + }, + { + "epoch": 0.19258, + "grad_norm": 1.2878287437826246, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19258 + }, + { + "epoch": 0.19259, + "grad_norm": 1.4055216921584257, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 19259 + }, + { + "epoch": 0.1926, + "grad_norm": 1.3051453569308915, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 19260 + }, + { + "epoch": 0.19261, + "grad_norm": 1.326885316634315, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19261 + }, + { + "epoch": 0.19262, + "grad_norm": 1.1061915072445085, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 19262 + }, + { + "epoch": 0.19263, + "grad_norm": 1.5654195215504978, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 19263 + }, + { + "epoch": 0.19264, + "grad_norm": 1.0952973511418131, + "learning_rate": 0.003, + "loss": 4.064, + "step": 19264 + }, + { + "epoch": 0.19265, + "grad_norm": 1.6292694563212933, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 19265 + }, + { + "epoch": 0.19266, + "grad_norm": 1.0953996900593994, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19266 + }, + { + "epoch": 0.19267, + "grad_norm": 1.3387421529983907, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 19267 + }, + { + "epoch": 0.19268, + "grad_norm": 1.5660995372853055, + "learning_rate": 0.003, + "loss": 4.055, + "step": 19268 + }, + { + "epoch": 0.19269, + "grad_norm": 1.2087119276592098, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19269 + }, + { + "epoch": 0.1927, + "grad_norm": 1.3580668250597665, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19270 + }, + { + "epoch": 0.19271, + "grad_norm": 1.3348444888748765, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19271 + }, + { + "epoch": 0.19272, + "grad_norm": 1.2520883322083953, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 19272 + }, + { + "epoch": 0.19273, + "grad_norm": 1.1815566217193947, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 19273 + }, + { + "epoch": 0.19274, + "grad_norm": 1.3356315303300998, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 19274 + }, + { + "epoch": 0.19275, + "grad_norm": 1.2447575707031089, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 19275 + }, + { + "epoch": 0.19276, + "grad_norm": 1.4291321612692254, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19276 + }, + { + "epoch": 0.19277, + "grad_norm": 1.1091724810120003, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 19277 + }, + { + "epoch": 0.19278, + "grad_norm": 1.8999785592313487, + "learning_rate": 0.003, + "loss": 4.091, + "step": 19278 + }, + { + "epoch": 0.19279, + "grad_norm": 1.0056310117229874, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 19279 + }, + { + "epoch": 0.1928, + "grad_norm": 1.5151507602005392, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 19280 + }, + { + "epoch": 0.19281, + "grad_norm": 1.4595941770553433, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19281 + }, + { + "epoch": 0.19282, + "grad_norm": 1.537815087375208, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 19282 + }, + { + "epoch": 0.19283, + "grad_norm": 1.1238184726071319, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 19283 + }, + { + "epoch": 0.19284, + "grad_norm": 1.2056169917291322, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 19284 + }, + { + "epoch": 0.19285, + "grad_norm": 1.4513606586793888, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 19285 + }, + { + "epoch": 0.19286, + "grad_norm": 1.2977929469861313, + "learning_rate": 0.003, + "loss": 4.033, + "step": 19286 + }, + { + "epoch": 0.19287, + "grad_norm": 1.2631869724978644, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19287 + }, + { + "epoch": 0.19288, + "grad_norm": 1.4007130177214029, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 19288 + }, + { + "epoch": 0.19289, + "grad_norm": 1.1353526144337664, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 19289 + }, + { + "epoch": 0.1929, + "grad_norm": 1.212149995014371, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19290 + }, + { + "epoch": 0.19291, + "grad_norm": 1.3859392455790678, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 19291 + }, + { + "epoch": 0.19292, + "grad_norm": 1.2598101853711283, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19292 + }, + { + "epoch": 0.19293, + "grad_norm": 1.3911589489772473, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19293 + }, + { + "epoch": 0.19294, + "grad_norm": 1.210170004902706, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 19294 + }, + { + "epoch": 0.19295, + "grad_norm": 1.299972482005869, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 19295 + }, + { + "epoch": 0.19296, + "grad_norm": 1.2627534993763714, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19296 + }, + { + "epoch": 0.19297, + "grad_norm": 1.3443064312058892, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 19297 + }, + { + "epoch": 0.19298, + "grad_norm": 1.666208868220041, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 19298 + }, + { + "epoch": 0.19299, + "grad_norm": 1.113952270737019, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 19299 + }, + { + "epoch": 0.193, + "grad_norm": 1.7041944102215842, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 19300 + }, + { + "epoch": 0.19301, + "grad_norm": 1.6375690792017714, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 19301 + }, + { + "epoch": 0.19302, + "grad_norm": 1.1871622671795783, + "learning_rate": 0.003, + "loss": 4.059, + "step": 19302 + }, + { + "epoch": 0.19303, + "grad_norm": 1.2523017718203313, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 19303 + }, + { + "epoch": 0.19304, + "grad_norm": 1.3298620440401583, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19304 + }, + { + "epoch": 0.19305, + "grad_norm": 1.3646092971100463, + "learning_rate": 0.003, + "loss": 4.072, + "step": 19305 + }, + { + "epoch": 0.19306, + "grad_norm": 1.1620948348169835, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19306 + }, + { + "epoch": 0.19307, + "grad_norm": 1.4016613215927611, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 19307 + }, + { + "epoch": 0.19308, + "grad_norm": 1.1736152267860105, + "learning_rate": 0.003, + "loss": 4.057, + "step": 19308 + }, + { + "epoch": 0.19309, + "grad_norm": 1.3182358837677548, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 19309 + }, + { + "epoch": 0.1931, + "grad_norm": 1.4501204798766651, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 19310 + }, + { + "epoch": 0.19311, + "grad_norm": 1.1434160852236295, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 19311 + }, + { + "epoch": 0.19312, + "grad_norm": 1.431722256758653, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19312 + }, + { + "epoch": 0.19313, + "grad_norm": 1.29689321893418, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 19313 + }, + { + "epoch": 0.19314, + "grad_norm": 1.608081179380934, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19314 + }, + { + "epoch": 0.19315, + "grad_norm": 1.1555466880702878, + "learning_rate": 0.003, + "loss": 4.029, + "step": 19315 + }, + { + "epoch": 0.19316, + "grad_norm": 1.3106279749947531, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 19316 + }, + { + "epoch": 0.19317, + "grad_norm": 1.2154730846642248, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19317 + }, + { + "epoch": 0.19318, + "grad_norm": 1.3811202154259432, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19318 + }, + { + "epoch": 0.19319, + "grad_norm": 1.2952358672295878, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19319 + }, + { + "epoch": 0.1932, + "grad_norm": 1.349704612179986, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 19320 + }, + { + "epoch": 0.19321, + "grad_norm": 1.4391108882127606, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19321 + }, + { + "epoch": 0.19322, + "grad_norm": 1.0698517431090664, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19322 + }, + { + "epoch": 0.19323, + "grad_norm": 1.5377972878630841, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 19323 + }, + { + "epoch": 0.19324, + "grad_norm": 1.1620482234533458, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 19324 + }, + { + "epoch": 0.19325, + "grad_norm": 1.706295803169076, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19325 + }, + { + "epoch": 0.19326, + "grad_norm": 1.4627113037011565, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 19326 + }, + { + "epoch": 0.19327, + "grad_norm": 1.2952479122134402, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 19327 + }, + { + "epoch": 0.19328, + "grad_norm": 1.3647261235036796, + "learning_rate": 0.003, + "loss": 4.076, + "step": 19328 + }, + { + "epoch": 0.19329, + "grad_norm": 1.398775878899619, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19329 + }, + { + "epoch": 0.1933, + "grad_norm": 0.9840400767359754, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 19330 + }, + { + "epoch": 0.19331, + "grad_norm": 1.3387790757069982, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19331 + }, + { + "epoch": 0.19332, + "grad_norm": 1.1478228412156144, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 19332 + }, + { + "epoch": 0.19333, + "grad_norm": 1.363465300325813, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 19333 + }, + { + "epoch": 0.19334, + "grad_norm": 1.2228351145593437, + "learning_rate": 0.003, + "loss": 4.023, + "step": 19334 + }, + { + "epoch": 0.19335, + "grad_norm": 1.5490623373968835, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 19335 + }, + { + "epoch": 0.19336, + "grad_norm": 1.2547671310467585, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19336 + }, + { + "epoch": 0.19337, + "grad_norm": 1.378056199478073, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 19337 + }, + { + "epoch": 0.19338, + "grad_norm": 1.2951080135198598, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 19338 + }, + { + "epoch": 0.19339, + "grad_norm": 1.4443763625994586, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 19339 + }, + { + "epoch": 0.1934, + "grad_norm": 1.0753716014890167, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 19340 + }, + { + "epoch": 0.19341, + "grad_norm": 1.3421729672704792, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 19341 + }, + { + "epoch": 0.19342, + "grad_norm": 1.3002098104094193, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19342 + }, + { + "epoch": 0.19343, + "grad_norm": 1.1639841138101032, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19343 + }, + { + "epoch": 0.19344, + "grad_norm": 1.5256197722062088, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 19344 + }, + { + "epoch": 0.19345, + "grad_norm": 1.1923686050416904, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19345 + }, + { + "epoch": 0.19346, + "grad_norm": 1.4160350775090675, + "learning_rate": 0.003, + "loss": 4.037, + "step": 19346 + }, + { + "epoch": 0.19347, + "grad_norm": 1.1720319120661895, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19347 + }, + { + "epoch": 0.19348, + "grad_norm": 1.3593873642570558, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 19348 + }, + { + "epoch": 0.19349, + "grad_norm": 1.3505122712820288, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 19349 + }, + { + "epoch": 0.1935, + "grad_norm": 1.3306040730637425, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19350 + }, + { + "epoch": 0.19351, + "grad_norm": 1.655727960969945, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19351 + }, + { + "epoch": 0.19352, + "grad_norm": 1.2126789166642862, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 19352 + }, + { + "epoch": 0.19353, + "grad_norm": 1.5285402732283524, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19353 + }, + { + "epoch": 0.19354, + "grad_norm": 1.1052693417942343, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19354 + }, + { + "epoch": 0.19355, + "grad_norm": 1.5586897787894143, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19355 + }, + { + "epoch": 0.19356, + "grad_norm": 0.9362571491901476, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 19356 + }, + { + "epoch": 0.19357, + "grad_norm": 1.1845455160555813, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 19357 + }, + { + "epoch": 0.19358, + "grad_norm": 1.4794767646486355, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 19358 + }, + { + "epoch": 0.19359, + "grad_norm": 1.143734784255008, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 19359 + }, + { + "epoch": 0.1936, + "grad_norm": 1.4012925676613213, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 19360 + }, + { + "epoch": 0.19361, + "grad_norm": 1.4033702307588178, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19361 + }, + { + "epoch": 0.19362, + "grad_norm": 1.3481368007485075, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19362 + }, + { + "epoch": 0.19363, + "grad_norm": 1.4076040383533015, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 19363 + }, + { + "epoch": 0.19364, + "grad_norm": 1.1996152385665022, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 19364 + }, + { + "epoch": 0.19365, + "grad_norm": 1.4580495886259732, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19365 + }, + { + "epoch": 0.19366, + "grad_norm": 1.521856191699455, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19366 + }, + { + "epoch": 0.19367, + "grad_norm": 1.631340300360941, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 19367 + }, + { + "epoch": 0.19368, + "grad_norm": 1.3336661687622993, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19368 + }, + { + "epoch": 0.19369, + "grad_norm": 1.345784656866648, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 19369 + }, + { + "epoch": 0.1937, + "grad_norm": 1.2206180867788792, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19370 + }, + { + "epoch": 0.19371, + "grad_norm": 1.1654276116493354, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 19371 + }, + { + "epoch": 0.19372, + "grad_norm": 1.3810446201500377, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 19372 + }, + { + "epoch": 0.19373, + "grad_norm": 1.373336874783953, + "learning_rate": 0.003, + "loss": 4.068, + "step": 19373 + }, + { + "epoch": 0.19374, + "grad_norm": 1.629303583339713, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 19374 + }, + { + "epoch": 0.19375, + "grad_norm": 1.0735248207499326, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 19375 + }, + { + "epoch": 0.19376, + "grad_norm": 1.3276649318294715, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19376 + }, + { + "epoch": 0.19377, + "grad_norm": 1.3805067134683795, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19377 + }, + { + "epoch": 0.19378, + "grad_norm": 1.1123481119523868, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 19378 + }, + { + "epoch": 0.19379, + "grad_norm": 1.2822520851622643, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 19379 + }, + { + "epoch": 0.1938, + "grad_norm": 1.2518157359945186, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19380 + }, + { + "epoch": 0.19381, + "grad_norm": 1.3211605719658674, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 19381 + }, + { + "epoch": 0.19382, + "grad_norm": 1.2031288178131152, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 19382 + }, + { + "epoch": 0.19383, + "grad_norm": 1.534044515509392, + "learning_rate": 0.003, + "loss": 4.042, + "step": 19383 + }, + { + "epoch": 0.19384, + "grad_norm": 1.2372061620324233, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 19384 + }, + { + "epoch": 0.19385, + "grad_norm": 1.4711169660183832, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 19385 + }, + { + "epoch": 0.19386, + "grad_norm": 1.1991312259387814, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 19386 + }, + { + "epoch": 0.19387, + "grad_norm": 1.417263892946054, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19387 + }, + { + "epoch": 0.19388, + "grad_norm": 1.3587734414119863, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19388 + }, + { + "epoch": 0.19389, + "grad_norm": 1.2021437211625692, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 19389 + }, + { + "epoch": 0.1939, + "grad_norm": 1.3861074777535454, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 19390 + }, + { + "epoch": 0.19391, + "grad_norm": 1.1842067172210315, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19391 + }, + { + "epoch": 0.19392, + "grad_norm": 1.433987285114199, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 19392 + }, + { + "epoch": 0.19393, + "grad_norm": 1.4103147808509588, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19393 + }, + { + "epoch": 0.19394, + "grad_norm": 1.7161650832967164, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 19394 + }, + { + "epoch": 0.19395, + "grad_norm": 1.177453349617905, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 19395 + }, + { + "epoch": 0.19396, + "grad_norm": 1.4484004379686284, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 19396 + }, + { + "epoch": 0.19397, + "grad_norm": 1.2794772010946776, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19397 + }, + { + "epoch": 0.19398, + "grad_norm": 1.0829439718799576, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 19398 + }, + { + "epoch": 0.19399, + "grad_norm": 1.6847222832453004, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 19399 + }, + { + "epoch": 0.194, + "grad_norm": 1.1198867425614099, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19400 + }, + { + "epoch": 0.19401, + "grad_norm": 1.5339102604902262, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19401 + }, + { + "epoch": 0.19402, + "grad_norm": 1.4360483324581836, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 19402 + }, + { + "epoch": 0.19403, + "grad_norm": 1.1690678118895501, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 19403 + }, + { + "epoch": 0.19404, + "grad_norm": 1.4040136868442783, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 19404 + }, + { + "epoch": 0.19405, + "grad_norm": 1.1443562702751413, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 19405 + }, + { + "epoch": 0.19406, + "grad_norm": 1.3270118186775146, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19406 + }, + { + "epoch": 0.19407, + "grad_norm": 1.2485030701960447, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 19407 + }, + { + "epoch": 0.19408, + "grad_norm": 1.2257617847502387, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19408 + }, + { + "epoch": 0.19409, + "grad_norm": 1.339782896099073, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19409 + }, + { + "epoch": 0.1941, + "grad_norm": 1.1653312443080106, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 19410 + }, + { + "epoch": 0.19411, + "grad_norm": 1.3190368669455002, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 19411 + }, + { + "epoch": 0.19412, + "grad_norm": 1.2566581938102939, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19412 + }, + { + "epoch": 0.19413, + "grad_norm": 1.086513711853439, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 19413 + }, + { + "epoch": 0.19414, + "grad_norm": 1.3678932310645313, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 19414 + }, + { + "epoch": 0.19415, + "grad_norm": 1.8391709363358586, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 19415 + }, + { + "epoch": 0.19416, + "grad_norm": 1.319014463307184, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19416 + }, + { + "epoch": 0.19417, + "grad_norm": 1.8220178704496939, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 19417 + }, + { + "epoch": 0.19418, + "grad_norm": 1.124723875317562, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 19418 + }, + { + "epoch": 0.19419, + "grad_norm": 1.19884607098138, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19419 + }, + { + "epoch": 0.1942, + "grad_norm": 1.449264556093074, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 19420 + }, + { + "epoch": 0.19421, + "grad_norm": 1.3191374682439083, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 19421 + }, + { + "epoch": 0.19422, + "grad_norm": 1.3981067550540194, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 19422 + }, + { + "epoch": 0.19423, + "grad_norm": 1.3127266074302606, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 19423 + }, + { + "epoch": 0.19424, + "grad_norm": 1.2563177666759349, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 19424 + }, + { + "epoch": 0.19425, + "grad_norm": 1.2782899915280543, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 19425 + }, + { + "epoch": 0.19426, + "grad_norm": 1.182685010153037, + "learning_rate": 0.003, + "loss": 4.068, + "step": 19426 + }, + { + "epoch": 0.19427, + "grad_norm": 1.407694968111235, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19427 + }, + { + "epoch": 0.19428, + "grad_norm": 1.2142362395422055, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 19428 + }, + { + "epoch": 0.19429, + "grad_norm": 1.3321695542747696, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 19429 + }, + { + "epoch": 0.1943, + "grad_norm": 1.3455226385908636, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 19430 + }, + { + "epoch": 0.19431, + "grad_norm": 1.1975306942264474, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 19431 + }, + { + "epoch": 0.19432, + "grad_norm": 1.306164180647245, + "learning_rate": 0.003, + "loss": 4.069, + "step": 19432 + }, + { + "epoch": 0.19433, + "grad_norm": 1.2038121182877317, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19433 + }, + { + "epoch": 0.19434, + "grad_norm": 1.3100820335537733, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 19434 + }, + { + "epoch": 0.19435, + "grad_norm": 1.51000847595959, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 19435 + }, + { + "epoch": 0.19436, + "grad_norm": 1.5370351507869693, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 19436 + }, + { + "epoch": 0.19437, + "grad_norm": 1.1469400430869083, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 19437 + }, + { + "epoch": 0.19438, + "grad_norm": 1.5209830775292177, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 19438 + }, + { + "epoch": 0.19439, + "grad_norm": 1.1656555651542801, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 19439 + }, + { + "epoch": 0.1944, + "grad_norm": 1.3849667723604768, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 19440 + }, + { + "epoch": 0.19441, + "grad_norm": 1.026361630538749, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 19441 + }, + { + "epoch": 0.19442, + "grad_norm": 1.4677422327224452, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 19442 + }, + { + "epoch": 0.19443, + "grad_norm": 1.1064600363094863, + "learning_rate": 0.003, + "loss": 4.087, + "step": 19443 + }, + { + "epoch": 0.19444, + "grad_norm": 1.581247945244656, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 19444 + }, + { + "epoch": 0.19445, + "grad_norm": 1.3784402022713331, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 19445 + }, + { + "epoch": 0.19446, + "grad_norm": 1.4594290215569299, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19446 + }, + { + "epoch": 0.19447, + "grad_norm": 1.5644238177299843, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19447 + }, + { + "epoch": 0.19448, + "grad_norm": 1.4121667484118812, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19448 + }, + { + "epoch": 0.19449, + "grad_norm": 1.401840224088073, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 19449 + }, + { + "epoch": 0.1945, + "grad_norm": 1.3696362388852628, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 19450 + }, + { + "epoch": 0.19451, + "grad_norm": 1.5213077490416473, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 19451 + }, + { + "epoch": 0.19452, + "grad_norm": 1.0000764143350864, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 19452 + }, + { + "epoch": 0.19453, + "grad_norm": 1.4410855592453407, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 19453 + }, + { + "epoch": 0.19454, + "grad_norm": 1.3713228817734917, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 19454 + }, + { + "epoch": 0.19455, + "grad_norm": 1.3782633140229856, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 19455 + }, + { + "epoch": 0.19456, + "grad_norm": 1.137024405499252, + "learning_rate": 0.003, + "loss": 4.088, + "step": 19456 + }, + { + "epoch": 0.19457, + "grad_norm": 1.3749447670302388, + "learning_rate": 0.003, + "loss": 4.068, + "step": 19457 + }, + { + "epoch": 0.19458, + "grad_norm": 1.1077411007998608, + "learning_rate": 0.003, + "loss": 4.062, + "step": 19458 + }, + { + "epoch": 0.19459, + "grad_norm": 1.2389465617647617, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 19459 + }, + { + "epoch": 0.1946, + "grad_norm": 1.2397676138930445, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 19460 + }, + { + "epoch": 0.19461, + "grad_norm": 1.1978532223077072, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 19461 + }, + { + "epoch": 0.19462, + "grad_norm": 1.3514504239036356, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 19462 + }, + { + "epoch": 0.19463, + "grad_norm": 1.2523103199057817, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 19463 + }, + { + "epoch": 0.19464, + "grad_norm": 1.6969692608474285, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 19464 + }, + { + "epoch": 0.19465, + "grad_norm": 0.9414368684036143, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 19465 + }, + { + "epoch": 0.19466, + "grad_norm": 1.4974558776391218, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 19466 + }, + { + "epoch": 0.19467, + "grad_norm": 1.4638764195062006, + "learning_rate": 0.003, + "loss": 4.086, + "step": 19467 + }, + { + "epoch": 0.19468, + "grad_norm": 1.514719727744151, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 19468 + }, + { + "epoch": 0.19469, + "grad_norm": 1.439838514731369, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19469 + }, + { + "epoch": 0.1947, + "grad_norm": 1.203798999719418, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19470 + }, + { + "epoch": 0.19471, + "grad_norm": 1.2416918412369136, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 19471 + }, + { + "epoch": 0.19472, + "grad_norm": 1.537366145495566, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 19472 + }, + { + "epoch": 0.19473, + "grad_norm": 1.3688593584598943, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 19473 + }, + { + "epoch": 0.19474, + "grad_norm": 1.3891744221027942, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19474 + }, + { + "epoch": 0.19475, + "grad_norm": 1.263524168556553, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 19475 + }, + { + "epoch": 0.19476, + "grad_norm": 1.3439616213955559, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 19476 + }, + { + "epoch": 0.19477, + "grad_norm": 1.3250794091095022, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 19477 + }, + { + "epoch": 0.19478, + "grad_norm": 1.3995346362050427, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 19478 + }, + { + "epoch": 0.19479, + "grad_norm": 1.2639749929102022, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 19479 + }, + { + "epoch": 0.1948, + "grad_norm": 1.5526913027825195, + "learning_rate": 0.003, + "loss": 4.069, + "step": 19480 + }, + { + "epoch": 0.19481, + "grad_norm": 1.1375459439711735, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19481 + }, + { + "epoch": 0.19482, + "grad_norm": 1.3154176386784155, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19482 + }, + { + "epoch": 0.19483, + "grad_norm": 1.4716210996534527, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 19483 + }, + { + "epoch": 0.19484, + "grad_norm": 1.3905671527809653, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 19484 + }, + { + "epoch": 0.19485, + "grad_norm": 1.3117630558690683, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19485 + }, + { + "epoch": 0.19486, + "grad_norm": 1.5008569960991724, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 19486 + }, + { + "epoch": 0.19487, + "grad_norm": 1.12906638228282, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 19487 + }, + { + "epoch": 0.19488, + "grad_norm": 1.3657365714985676, + "learning_rate": 0.003, + "loss": 4.043, + "step": 19488 + }, + { + "epoch": 0.19489, + "grad_norm": 1.3018201857887008, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19489 + }, + { + "epoch": 0.1949, + "grad_norm": 1.3983987668216904, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 19490 + }, + { + "epoch": 0.19491, + "grad_norm": 1.481011022539193, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19491 + }, + { + "epoch": 0.19492, + "grad_norm": 1.1813655279877215, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 19492 + }, + { + "epoch": 0.19493, + "grad_norm": 1.4416587610837648, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 19493 + }, + { + "epoch": 0.19494, + "grad_norm": 1.1131622561587362, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 19494 + }, + { + "epoch": 0.19495, + "grad_norm": 1.4616149240882121, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 19495 + }, + { + "epoch": 0.19496, + "grad_norm": 1.0756483327498663, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 19496 + }, + { + "epoch": 0.19497, + "grad_norm": 1.1718359588897174, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 19497 + }, + { + "epoch": 0.19498, + "grad_norm": 1.4083804540585427, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 19498 + }, + { + "epoch": 0.19499, + "grad_norm": 1.2897914243706128, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 19499 + }, + { + "epoch": 0.195, + "grad_norm": 1.554466135239505, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 19500 + }, + { + "epoch": 0.19501, + "grad_norm": 1.167251633086409, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 19501 + }, + { + "epoch": 0.19502, + "grad_norm": 1.759459231435133, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 19502 + }, + { + "epoch": 0.19503, + "grad_norm": 1.0433124182408695, + "learning_rate": 0.003, + "loss": 4.039, + "step": 19503 + }, + { + "epoch": 0.19504, + "grad_norm": 1.4082628477633647, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19504 + }, + { + "epoch": 0.19505, + "grad_norm": 1.3823904406544583, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 19505 + }, + { + "epoch": 0.19506, + "grad_norm": 1.0638797553557129, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19506 + }, + { + "epoch": 0.19507, + "grad_norm": 1.6891751096651972, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 19507 + }, + { + "epoch": 0.19508, + "grad_norm": 0.9416716412099737, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19508 + }, + { + "epoch": 0.19509, + "grad_norm": 1.5765004581747462, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 19509 + }, + { + "epoch": 0.1951, + "grad_norm": 1.3089842021211013, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 19510 + }, + { + "epoch": 0.19511, + "grad_norm": 1.341192332501336, + "learning_rate": 0.003, + "loss": 4.051, + "step": 19511 + }, + { + "epoch": 0.19512, + "grad_norm": 1.7025869243643506, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 19512 + }, + { + "epoch": 0.19513, + "grad_norm": 1.3170221606778398, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 19513 + }, + { + "epoch": 0.19514, + "grad_norm": 1.2350694969711091, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 19514 + }, + { + "epoch": 0.19515, + "grad_norm": 1.290375283368854, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 19515 + }, + { + "epoch": 0.19516, + "grad_norm": 1.3103270446228672, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 19516 + }, + { + "epoch": 0.19517, + "grad_norm": 1.2217186690684785, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19517 + }, + { + "epoch": 0.19518, + "grad_norm": 1.3932532373420963, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 19518 + }, + { + "epoch": 0.19519, + "grad_norm": 1.6149310376647918, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19519 + }, + { + "epoch": 0.1952, + "grad_norm": 1.1468617142101982, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 19520 + }, + { + "epoch": 0.19521, + "grad_norm": 1.3430904754624255, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 19521 + }, + { + "epoch": 0.19522, + "grad_norm": 1.4246909909100243, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 19522 + }, + { + "epoch": 0.19523, + "grad_norm": 1.1588283886075517, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 19523 + }, + { + "epoch": 0.19524, + "grad_norm": 1.3071282399894546, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 19524 + }, + { + "epoch": 0.19525, + "grad_norm": 1.5768030414564296, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 19525 + }, + { + "epoch": 0.19526, + "grad_norm": 1.1685958781565342, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 19526 + }, + { + "epoch": 0.19527, + "grad_norm": 1.313093450897761, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 19527 + }, + { + "epoch": 0.19528, + "grad_norm": 1.2742119703672696, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 19528 + }, + { + "epoch": 0.19529, + "grad_norm": 1.7290547434394015, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 19529 + }, + { + "epoch": 0.1953, + "grad_norm": 1.0529733408830393, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 19530 + }, + { + "epoch": 0.19531, + "grad_norm": 1.5482101859966972, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 19531 + }, + { + "epoch": 0.19532, + "grad_norm": 1.2989756814178277, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 19532 + }, + { + "epoch": 0.19533, + "grad_norm": 1.1864556980596295, + "learning_rate": 0.003, + "loss": 4.078, + "step": 19533 + }, + { + "epoch": 0.19534, + "grad_norm": 1.1906409640998392, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 19534 + }, + { + "epoch": 0.19535, + "grad_norm": 1.3279386097222499, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 19535 + }, + { + "epoch": 0.19536, + "grad_norm": 1.3249211366390132, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 19536 + }, + { + "epoch": 0.19537, + "grad_norm": 1.2708672308772355, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 19537 + }, + { + "epoch": 0.19538, + "grad_norm": 1.3315379121969504, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19538 + }, + { + "epoch": 0.19539, + "grad_norm": 1.2429310832605476, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 19539 + }, + { + "epoch": 0.1954, + "grad_norm": 1.5054210239848758, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19540 + }, + { + "epoch": 0.19541, + "grad_norm": 1.0475191411451963, + "learning_rate": 0.003, + "loss": 4.064, + "step": 19541 + }, + { + "epoch": 0.19542, + "grad_norm": 1.4284508801628617, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 19542 + }, + { + "epoch": 0.19543, + "grad_norm": 1.0933611515009474, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 19543 + }, + { + "epoch": 0.19544, + "grad_norm": 1.661660218431094, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19544 + }, + { + "epoch": 0.19545, + "grad_norm": 1.2199410776915833, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19545 + }, + { + "epoch": 0.19546, + "grad_norm": 1.6905546886575713, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 19546 + }, + { + "epoch": 0.19547, + "grad_norm": 1.29296218202824, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19547 + }, + { + "epoch": 0.19548, + "grad_norm": 1.2837712876533047, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19548 + }, + { + "epoch": 0.19549, + "grad_norm": 1.4703643136338311, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 19549 + }, + { + "epoch": 0.1955, + "grad_norm": 1.2700676522488055, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 19550 + }, + { + "epoch": 0.19551, + "grad_norm": 1.1109703899269674, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19551 + }, + { + "epoch": 0.19552, + "grad_norm": 1.2673166776363936, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 19552 + }, + { + "epoch": 0.19553, + "grad_norm": 1.4214656677819935, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 19553 + }, + { + "epoch": 0.19554, + "grad_norm": 1.2423078438651163, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 19554 + }, + { + "epoch": 0.19555, + "grad_norm": 1.4398209252606844, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 19555 + }, + { + "epoch": 0.19556, + "grad_norm": 1.4883629305181807, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 19556 + }, + { + "epoch": 0.19557, + "grad_norm": 1.297674226675014, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19557 + }, + { + "epoch": 0.19558, + "grad_norm": 1.3053307952640294, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 19558 + }, + { + "epoch": 0.19559, + "grad_norm": 1.4310611165329026, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19559 + }, + { + "epoch": 0.1956, + "grad_norm": 1.5197492036910145, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 19560 + }, + { + "epoch": 0.19561, + "grad_norm": 1.1766594132642922, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 19561 + }, + { + "epoch": 0.19562, + "grad_norm": 1.2995395067215239, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 19562 + }, + { + "epoch": 0.19563, + "grad_norm": 1.2983980671464603, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 19563 + }, + { + "epoch": 0.19564, + "grad_norm": 1.3436105722348386, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19564 + }, + { + "epoch": 0.19565, + "grad_norm": 1.3194744920133412, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 19565 + }, + { + "epoch": 0.19566, + "grad_norm": 1.2171903493614817, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 19566 + }, + { + "epoch": 0.19567, + "grad_norm": 1.6401387912898335, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 19567 + }, + { + "epoch": 0.19568, + "grad_norm": 1.1441649428245726, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 19568 + }, + { + "epoch": 0.19569, + "grad_norm": 1.251626956814284, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19569 + }, + { + "epoch": 0.1957, + "grad_norm": 1.3502503512367485, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 19570 + }, + { + "epoch": 0.19571, + "grad_norm": 1.2426582395162038, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19571 + }, + { + "epoch": 0.19572, + "grad_norm": 1.2413271686787717, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 19572 + }, + { + "epoch": 0.19573, + "grad_norm": 1.1949950162191312, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 19573 + }, + { + "epoch": 0.19574, + "grad_norm": 1.288221508446464, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 19574 + }, + { + "epoch": 0.19575, + "grad_norm": 1.504647978943427, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 19575 + }, + { + "epoch": 0.19576, + "grad_norm": 1.2783626987122165, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 19576 + }, + { + "epoch": 0.19577, + "grad_norm": 1.5688010739625575, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 19577 + }, + { + "epoch": 0.19578, + "grad_norm": 1.1532961025165187, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19578 + }, + { + "epoch": 0.19579, + "grad_norm": 1.48442655140837, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 19579 + }, + { + "epoch": 0.1958, + "grad_norm": 1.1852336573135187, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 19580 + }, + { + "epoch": 0.19581, + "grad_norm": 1.4192082820545153, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 19581 + }, + { + "epoch": 0.19582, + "grad_norm": 1.2840117706091172, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 19582 + }, + { + "epoch": 0.19583, + "grad_norm": 1.2979628805575405, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 19583 + }, + { + "epoch": 0.19584, + "grad_norm": 1.0485564253954378, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 19584 + }, + { + "epoch": 0.19585, + "grad_norm": 1.698992243899307, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 19585 + }, + { + "epoch": 0.19586, + "grad_norm": 1.2007570479118221, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 19586 + }, + { + "epoch": 0.19587, + "grad_norm": 1.299143804009154, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 19587 + }, + { + "epoch": 0.19588, + "grad_norm": 1.4096556135900993, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19588 + }, + { + "epoch": 0.19589, + "grad_norm": 1.2516873856225414, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19589 + }, + { + "epoch": 0.1959, + "grad_norm": 1.3511913389802903, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 19590 + }, + { + "epoch": 0.19591, + "grad_norm": 1.5757957682280774, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 19591 + }, + { + "epoch": 0.19592, + "grad_norm": 1.3050377336538197, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 19592 + }, + { + "epoch": 0.19593, + "grad_norm": 1.3278524136963044, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 19593 + }, + { + "epoch": 0.19594, + "grad_norm": 1.2983540617735474, + "learning_rate": 0.003, + "loss": 4.078, + "step": 19594 + }, + { + "epoch": 0.19595, + "grad_norm": 1.461744079705968, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 19595 + }, + { + "epoch": 0.19596, + "grad_norm": 1.1097298106864095, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19596 + }, + { + "epoch": 0.19597, + "grad_norm": 1.5665459522554137, + "learning_rate": 0.003, + "loss": 4.088, + "step": 19597 + }, + { + "epoch": 0.19598, + "grad_norm": 1.167197822058204, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19598 + }, + { + "epoch": 0.19599, + "grad_norm": 1.23200999485626, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 19599 + }, + { + "epoch": 0.196, + "grad_norm": 1.5284174897054683, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 19600 + }, + { + "epoch": 0.19601, + "grad_norm": 1.5221817934907782, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 19601 + }, + { + "epoch": 0.19602, + "grad_norm": 1.243043726637542, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 19602 + }, + { + "epoch": 0.19603, + "grad_norm": 1.358168188861275, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 19603 + }, + { + "epoch": 0.19604, + "grad_norm": 1.3824418445757143, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 19604 + }, + { + "epoch": 0.19605, + "grad_norm": 1.4568231810713086, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 19605 + }, + { + "epoch": 0.19606, + "grad_norm": 1.4945389488733762, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 19606 + }, + { + "epoch": 0.19607, + "grad_norm": 1.1173944068333428, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 19607 + }, + { + "epoch": 0.19608, + "grad_norm": 1.4386483359475473, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19608 + }, + { + "epoch": 0.19609, + "grad_norm": 1.1587209555220592, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 19609 + }, + { + "epoch": 0.1961, + "grad_norm": 1.3855977084226376, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19610 + }, + { + "epoch": 0.19611, + "grad_norm": 1.191095261316692, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 19611 + }, + { + "epoch": 0.19612, + "grad_norm": 1.329141360808254, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 19612 + }, + { + "epoch": 0.19613, + "grad_norm": 1.2914757409323085, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 19613 + }, + { + "epoch": 0.19614, + "grad_norm": 1.1990703428141272, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 19614 + }, + { + "epoch": 0.19615, + "grad_norm": 1.8458551392426072, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 19615 + }, + { + "epoch": 0.19616, + "grad_norm": 1.2531170952353852, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19616 + }, + { + "epoch": 0.19617, + "grad_norm": 1.457888411021175, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 19617 + }, + { + "epoch": 0.19618, + "grad_norm": 1.2062617937354843, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 19618 + }, + { + "epoch": 0.19619, + "grad_norm": 1.2088629579297447, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 19619 + }, + { + "epoch": 0.1962, + "grad_norm": 1.3785785348043016, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 19620 + }, + { + "epoch": 0.19621, + "grad_norm": 1.1938751154673777, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 19621 + }, + { + "epoch": 0.19622, + "grad_norm": 1.405148392569504, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19622 + }, + { + "epoch": 0.19623, + "grad_norm": 1.4096251360863021, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 19623 + }, + { + "epoch": 0.19624, + "grad_norm": 1.2746598690678428, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19624 + }, + { + "epoch": 0.19625, + "grad_norm": 1.307968715382792, + "learning_rate": 0.003, + "loss": 4.057, + "step": 19625 + }, + { + "epoch": 0.19626, + "grad_norm": 1.3487640965516128, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19626 + }, + { + "epoch": 0.19627, + "grad_norm": 1.3014966503765235, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 19627 + }, + { + "epoch": 0.19628, + "grad_norm": 1.5196558968221845, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19628 + }, + { + "epoch": 0.19629, + "grad_norm": 1.351611078017904, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 19629 + }, + { + "epoch": 0.1963, + "grad_norm": 1.1982555578363527, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 19630 + }, + { + "epoch": 0.19631, + "grad_norm": 1.50858883084661, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 19631 + }, + { + "epoch": 0.19632, + "grad_norm": 1.209088040951843, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 19632 + }, + { + "epoch": 0.19633, + "grad_norm": 1.1706012946828515, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 19633 + }, + { + "epoch": 0.19634, + "grad_norm": 1.3884683601648247, + "learning_rate": 0.003, + "loss": 4.08, + "step": 19634 + }, + { + "epoch": 0.19635, + "grad_norm": 1.4084716436490483, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19635 + }, + { + "epoch": 0.19636, + "grad_norm": 1.2012068473180846, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19636 + }, + { + "epoch": 0.19637, + "grad_norm": 1.1904377831067214, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 19637 + }, + { + "epoch": 0.19638, + "grad_norm": 1.5549648250763126, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 19638 + }, + { + "epoch": 0.19639, + "grad_norm": 1.2924911454223567, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 19639 + }, + { + "epoch": 0.1964, + "grad_norm": 1.436073449635221, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 19640 + }, + { + "epoch": 0.19641, + "grad_norm": 1.2398135512071407, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19641 + }, + { + "epoch": 0.19642, + "grad_norm": 1.4762790518015563, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 19642 + }, + { + "epoch": 0.19643, + "grad_norm": 1.3289287324093857, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 19643 + }, + { + "epoch": 0.19644, + "grad_norm": 1.415053931855098, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 19644 + }, + { + "epoch": 0.19645, + "grad_norm": 1.3670122111536873, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 19645 + }, + { + "epoch": 0.19646, + "grad_norm": 1.2611243225559388, + "learning_rate": 0.003, + "loss": 4.078, + "step": 19646 + }, + { + "epoch": 0.19647, + "grad_norm": 1.290674837100807, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19647 + }, + { + "epoch": 0.19648, + "grad_norm": 1.3510109434285988, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 19648 + }, + { + "epoch": 0.19649, + "grad_norm": 1.2111258497810031, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 19649 + }, + { + "epoch": 0.1965, + "grad_norm": 1.626658891057458, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 19650 + }, + { + "epoch": 0.19651, + "grad_norm": 1.3400385588297339, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19651 + }, + { + "epoch": 0.19652, + "grad_norm": 1.3719636497684822, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 19652 + }, + { + "epoch": 0.19653, + "grad_norm": 1.2347975873177621, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 19653 + }, + { + "epoch": 0.19654, + "grad_norm": 1.447584926720447, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 19654 + }, + { + "epoch": 0.19655, + "grad_norm": 1.1502211212205573, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19655 + }, + { + "epoch": 0.19656, + "grad_norm": 1.5177566977359307, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 19656 + }, + { + "epoch": 0.19657, + "grad_norm": 1.6310376806213571, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 19657 + }, + { + "epoch": 0.19658, + "grad_norm": 1.1472209213061497, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 19658 + }, + { + "epoch": 0.19659, + "grad_norm": 1.4783673540103026, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 19659 + }, + { + "epoch": 0.1966, + "grad_norm": 1.442430981905889, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 19660 + }, + { + "epoch": 0.19661, + "grad_norm": 1.358558532685666, + "learning_rate": 0.003, + "loss": 4.049, + "step": 19661 + }, + { + "epoch": 0.19662, + "grad_norm": 1.2797478679943666, + "learning_rate": 0.003, + "loss": 4.077, + "step": 19662 + }, + { + "epoch": 0.19663, + "grad_norm": 1.1363377398858079, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19663 + }, + { + "epoch": 0.19664, + "grad_norm": 1.3989413774376376, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 19664 + }, + { + "epoch": 0.19665, + "grad_norm": 1.1939159607281624, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 19665 + }, + { + "epoch": 0.19666, + "grad_norm": 1.2299607667675085, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 19666 + }, + { + "epoch": 0.19667, + "grad_norm": 1.4127656745930717, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 19667 + }, + { + "epoch": 0.19668, + "grad_norm": 1.5367893456607715, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 19668 + }, + { + "epoch": 0.19669, + "grad_norm": 1.3147698787709958, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 19669 + }, + { + "epoch": 0.1967, + "grad_norm": 1.218187691026266, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 19670 + }, + { + "epoch": 0.19671, + "grad_norm": 1.6882610127488684, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 19671 + }, + { + "epoch": 0.19672, + "grad_norm": 0.9715104515027442, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19672 + }, + { + "epoch": 0.19673, + "grad_norm": 1.6056393856497801, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19673 + }, + { + "epoch": 0.19674, + "grad_norm": 1.1057824576585424, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 19674 + }, + { + "epoch": 0.19675, + "grad_norm": 1.5481958238517912, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 19675 + }, + { + "epoch": 0.19676, + "grad_norm": 1.1711742364191884, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 19676 + }, + { + "epoch": 0.19677, + "grad_norm": 1.7711689647842495, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 19677 + }, + { + "epoch": 0.19678, + "grad_norm": 1.219879003445908, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 19678 + }, + { + "epoch": 0.19679, + "grad_norm": 1.226992949333215, + "learning_rate": 0.003, + "loss": 4.052, + "step": 19679 + }, + { + "epoch": 0.1968, + "grad_norm": 1.5091792519248097, + "learning_rate": 0.003, + "loss": 4.054, + "step": 19680 + }, + { + "epoch": 0.19681, + "grad_norm": 1.1502942125089162, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 19681 + }, + { + "epoch": 0.19682, + "grad_norm": 1.369165848780662, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 19682 + }, + { + "epoch": 0.19683, + "grad_norm": 1.50610209047457, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 19683 + }, + { + "epoch": 0.19684, + "grad_norm": 1.3913243831111002, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19684 + }, + { + "epoch": 0.19685, + "grad_norm": 1.489411249466416, + "learning_rate": 0.003, + "loss": 4.026, + "step": 19685 + }, + { + "epoch": 0.19686, + "grad_norm": 1.2988468071078383, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 19686 + }, + { + "epoch": 0.19687, + "grad_norm": 1.2912204537705885, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19687 + }, + { + "epoch": 0.19688, + "grad_norm": 1.3233081118225472, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 19688 + }, + { + "epoch": 0.19689, + "grad_norm": 1.3256641452369216, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 19689 + }, + { + "epoch": 0.1969, + "grad_norm": 1.2660930257667473, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 19690 + }, + { + "epoch": 0.19691, + "grad_norm": 1.1883754721295503, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 19691 + }, + { + "epoch": 0.19692, + "grad_norm": 1.376985420827303, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 19692 + }, + { + "epoch": 0.19693, + "grad_norm": 1.3150305005470482, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 19693 + }, + { + "epoch": 0.19694, + "grad_norm": 1.47217745519008, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19694 + }, + { + "epoch": 0.19695, + "grad_norm": 1.1481972427979865, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19695 + }, + { + "epoch": 0.19696, + "grad_norm": 1.321599317780327, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19696 + }, + { + "epoch": 0.19697, + "grad_norm": 1.6388995502727124, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 19697 + }, + { + "epoch": 0.19698, + "grad_norm": 1.2484343120322956, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 19698 + }, + { + "epoch": 0.19699, + "grad_norm": 1.5216813620088356, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 19699 + }, + { + "epoch": 0.197, + "grad_norm": 1.0916232273133109, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 19700 + }, + { + "epoch": 0.19701, + "grad_norm": 1.1873914219973216, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 19701 + }, + { + "epoch": 0.19702, + "grad_norm": 1.4738247803097095, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 19702 + }, + { + "epoch": 0.19703, + "grad_norm": 1.1357491362604857, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 19703 + }, + { + "epoch": 0.19704, + "grad_norm": 1.4484137053164607, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 19704 + }, + { + "epoch": 0.19705, + "grad_norm": 1.3439057007523725, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 19705 + }, + { + "epoch": 0.19706, + "grad_norm": 1.4023196364405632, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 19706 + }, + { + "epoch": 0.19707, + "grad_norm": 1.2544428278487003, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 19707 + }, + { + "epoch": 0.19708, + "grad_norm": 1.224674119818485, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 19708 + }, + { + "epoch": 0.19709, + "grad_norm": 1.604438625568975, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 19709 + }, + { + "epoch": 0.1971, + "grad_norm": 1.2311551502434666, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19710 + }, + { + "epoch": 0.19711, + "grad_norm": 1.227087896947134, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19711 + }, + { + "epoch": 0.19712, + "grad_norm": 1.4217940509714437, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 19712 + }, + { + "epoch": 0.19713, + "grad_norm": 1.3413863733209068, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19713 + }, + { + "epoch": 0.19714, + "grad_norm": 1.4078100967643121, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 19714 + }, + { + "epoch": 0.19715, + "grad_norm": 1.2338926081565247, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19715 + }, + { + "epoch": 0.19716, + "grad_norm": 1.326378839883387, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 19716 + }, + { + "epoch": 0.19717, + "grad_norm": 1.3322748801981528, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 19717 + }, + { + "epoch": 0.19718, + "grad_norm": 1.374045360430272, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19718 + }, + { + "epoch": 0.19719, + "grad_norm": 1.2535400024902488, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 19719 + }, + { + "epoch": 0.1972, + "grad_norm": 1.7216545683033468, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 19720 + }, + { + "epoch": 0.19721, + "grad_norm": 1.143211416994933, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 19721 + }, + { + "epoch": 0.19722, + "grad_norm": 1.347843799576987, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 19722 + }, + { + "epoch": 0.19723, + "grad_norm": 1.2263043437640626, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 19723 + }, + { + "epoch": 0.19724, + "grad_norm": 1.4235896139380093, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19724 + }, + { + "epoch": 0.19725, + "grad_norm": 1.249363935710837, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 19725 + }, + { + "epoch": 0.19726, + "grad_norm": 1.3621808175319952, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19726 + }, + { + "epoch": 0.19727, + "grad_norm": 1.2027545345421449, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 19727 + }, + { + "epoch": 0.19728, + "grad_norm": 1.3819118273551108, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 19728 + }, + { + "epoch": 0.19729, + "grad_norm": 1.3139500700926068, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 19729 + }, + { + "epoch": 0.1973, + "grad_norm": 1.2412752229393393, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19730 + }, + { + "epoch": 0.19731, + "grad_norm": 1.2513695591662786, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 19731 + }, + { + "epoch": 0.19732, + "grad_norm": 1.6091074837307349, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 19732 + }, + { + "epoch": 0.19733, + "grad_norm": 1.1823045345431846, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19733 + }, + { + "epoch": 0.19734, + "grad_norm": 1.5925556387548963, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 19734 + }, + { + "epoch": 0.19735, + "grad_norm": 0.9643452763097676, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 19735 + }, + { + "epoch": 0.19736, + "grad_norm": 1.4569350256120528, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 19736 + }, + { + "epoch": 0.19737, + "grad_norm": 1.3691898829958777, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 19737 + }, + { + "epoch": 0.19738, + "grad_norm": 1.0934636594109788, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 19738 + }, + { + "epoch": 0.19739, + "grad_norm": 1.4965763361831188, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 19739 + }, + { + "epoch": 0.1974, + "grad_norm": 1.2917382252185112, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 19740 + }, + { + "epoch": 0.19741, + "grad_norm": 1.3573544639844923, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 19741 + }, + { + "epoch": 0.19742, + "grad_norm": 1.1576171567288007, + "learning_rate": 0.003, + "loss": 4.053, + "step": 19742 + }, + { + "epoch": 0.19743, + "grad_norm": 1.333131141716199, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 19743 + }, + { + "epoch": 0.19744, + "grad_norm": 1.3814560633215616, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 19744 + }, + { + "epoch": 0.19745, + "grad_norm": 1.1487887883235977, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 19745 + }, + { + "epoch": 0.19746, + "grad_norm": 1.3776024966932388, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 19746 + }, + { + "epoch": 0.19747, + "grad_norm": 1.3859940421713073, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 19747 + }, + { + "epoch": 0.19748, + "grad_norm": 1.5127485072971667, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19748 + }, + { + "epoch": 0.19749, + "grad_norm": 1.1785057150237257, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19749 + }, + { + "epoch": 0.1975, + "grad_norm": 1.589956479162368, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 19750 + }, + { + "epoch": 0.19751, + "grad_norm": 1.4266342589366159, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 19751 + }, + { + "epoch": 0.19752, + "grad_norm": 1.3121486648783165, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 19752 + }, + { + "epoch": 0.19753, + "grad_norm": 1.4121421014664186, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 19753 + }, + { + "epoch": 0.19754, + "grad_norm": 1.0924639918405121, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 19754 + }, + { + "epoch": 0.19755, + "grad_norm": 1.487611705583233, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 19755 + }, + { + "epoch": 0.19756, + "grad_norm": 1.2796519124302443, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 19756 + }, + { + "epoch": 0.19757, + "grad_norm": 1.6484103057642017, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 19757 + }, + { + "epoch": 0.19758, + "grad_norm": 1.311605176822461, + "learning_rate": 0.003, + "loss": 4.103, + "step": 19758 + }, + { + "epoch": 0.19759, + "grad_norm": 1.2889955881539013, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19759 + }, + { + "epoch": 0.1976, + "grad_norm": 1.4292082948589133, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 19760 + }, + { + "epoch": 0.19761, + "grad_norm": 1.3230632841660548, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 19761 + }, + { + "epoch": 0.19762, + "grad_norm": 1.3688875268505394, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 19762 + }, + { + "epoch": 0.19763, + "grad_norm": 1.2823318554850982, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19763 + }, + { + "epoch": 0.19764, + "grad_norm": 1.2166216140103328, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 19764 + }, + { + "epoch": 0.19765, + "grad_norm": 1.2852363984667483, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 19765 + }, + { + "epoch": 0.19766, + "grad_norm": 1.2898294841391993, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 19766 + }, + { + "epoch": 0.19767, + "grad_norm": 1.2367526470670498, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 19767 + }, + { + "epoch": 0.19768, + "grad_norm": 1.3154865687307928, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 19768 + }, + { + "epoch": 0.19769, + "grad_norm": 1.2603167654919585, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 19769 + }, + { + "epoch": 0.1977, + "grad_norm": 1.3007040435706034, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 19770 + }, + { + "epoch": 0.19771, + "grad_norm": 1.4114993924752812, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 19771 + }, + { + "epoch": 0.19772, + "grad_norm": 1.3096137540244503, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 19772 + }, + { + "epoch": 0.19773, + "grad_norm": 1.4355194693826703, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19773 + }, + { + "epoch": 0.19774, + "grad_norm": 1.4518512466537297, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 19774 + }, + { + "epoch": 0.19775, + "grad_norm": 1.6062944604381495, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 19775 + }, + { + "epoch": 0.19776, + "grad_norm": 1.4836222015556304, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 19776 + }, + { + "epoch": 0.19777, + "grad_norm": 1.3167608342443442, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 19777 + }, + { + "epoch": 0.19778, + "grad_norm": 1.3800670965071138, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 19778 + }, + { + "epoch": 0.19779, + "grad_norm": 1.2467694158107294, + "learning_rate": 0.003, + "loss": 4.076, + "step": 19779 + }, + { + "epoch": 0.1978, + "grad_norm": 1.1383855194669683, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 19780 + }, + { + "epoch": 0.19781, + "grad_norm": 1.4457824499750687, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 19781 + }, + { + "epoch": 0.19782, + "grad_norm": 1.1524489184514475, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 19782 + }, + { + "epoch": 0.19783, + "grad_norm": 1.4835000980643374, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 19783 + }, + { + "epoch": 0.19784, + "grad_norm": 1.1729540045494726, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 19784 + }, + { + "epoch": 0.19785, + "grad_norm": 1.5597271638773154, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 19785 + }, + { + "epoch": 0.19786, + "grad_norm": 1.3320044397616282, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19786 + }, + { + "epoch": 0.19787, + "grad_norm": 1.4422976842356665, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 19787 + }, + { + "epoch": 0.19788, + "grad_norm": 1.3253455241649499, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 19788 + }, + { + "epoch": 0.19789, + "grad_norm": 1.128730509489487, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 19789 + }, + { + "epoch": 0.1979, + "grad_norm": 1.5220058756348567, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 19790 + }, + { + "epoch": 0.19791, + "grad_norm": 1.0193149807952373, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 19791 + }, + { + "epoch": 0.19792, + "grad_norm": 1.7698679856468362, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 19792 + }, + { + "epoch": 0.19793, + "grad_norm": 1.1620624717348689, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19793 + }, + { + "epoch": 0.19794, + "grad_norm": 1.3691058623543266, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 19794 + }, + { + "epoch": 0.19795, + "grad_norm": 1.3043787042526258, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19795 + }, + { + "epoch": 0.19796, + "grad_norm": 1.4206097790958214, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 19796 + }, + { + "epoch": 0.19797, + "grad_norm": 1.236986530855312, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 19797 + }, + { + "epoch": 0.19798, + "grad_norm": 1.3913813954888499, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 19798 + }, + { + "epoch": 0.19799, + "grad_norm": 1.131110863878918, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19799 + }, + { + "epoch": 0.198, + "grad_norm": 1.4867324186283586, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 19800 + }, + { + "epoch": 0.19801, + "grad_norm": 1.419459874682956, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 19801 + }, + { + "epoch": 0.19802, + "grad_norm": 1.502929630810888, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 19802 + }, + { + "epoch": 0.19803, + "grad_norm": 1.35980660698995, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 19803 + }, + { + "epoch": 0.19804, + "grad_norm": 1.4439911170495572, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 19804 + }, + { + "epoch": 0.19805, + "grad_norm": 1.3834374075151001, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 19805 + }, + { + "epoch": 0.19806, + "grad_norm": 1.1914514918607964, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19806 + }, + { + "epoch": 0.19807, + "grad_norm": 1.290229652438182, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19807 + }, + { + "epoch": 0.19808, + "grad_norm": 1.2694731438203108, + "learning_rate": 0.003, + "loss": 4.111, + "step": 19808 + }, + { + "epoch": 0.19809, + "grad_norm": 1.3692766120784927, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 19809 + }, + { + "epoch": 0.1981, + "grad_norm": 1.1628895391517091, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 19810 + }, + { + "epoch": 0.19811, + "grad_norm": 1.424318988673435, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 19811 + }, + { + "epoch": 0.19812, + "grad_norm": 1.4851767350452065, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 19812 + }, + { + "epoch": 0.19813, + "grad_norm": 1.4167736869370577, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 19813 + }, + { + "epoch": 0.19814, + "grad_norm": 1.470995925875908, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 19814 + }, + { + "epoch": 0.19815, + "grad_norm": 1.2856835824921131, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19815 + }, + { + "epoch": 0.19816, + "grad_norm": 1.286169394231622, + "learning_rate": 0.003, + "loss": 4.05, + "step": 19816 + }, + { + "epoch": 0.19817, + "grad_norm": 1.1940010585749086, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 19817 + }, + { + "epoch": 0.19818, + "grad_norm": 1.695886874437397, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 19818 + }, + { + "epoch": 0.19819, + "grad_norm": 1.0606100857710243, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 19819 + }, + { + "epoch": 0.1982, + "grad_norm": 1.3438230427854372, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 19820 + }, + { + "epoch": 0.19821, + "grad_norm": 1.2955036264556594, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 19821 + }, + { + "epoch": 0.19822, + "grad_norm": 1.3098240891287187, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 19822 + }, + { + "epoch": 0.19823, + "grad_norm": 1.3909821017152115, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 19823 + }, + { + "epoch": 0.19824, + "grad_norm": 1.3030669319903594, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19824 + }, + { + "epoch": 0.19825, + "grad_norm": 1.2377995859660755, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 19825 + }, + { + "epoch": 0.19826, + "grad_norm": 1.138853085824324, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 19826 + }, + { + "epoch": 0.19827, + "grad_norm": 1.5446729952809068, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 19827 + }, + { + "epoch": 0.19828, + "grad_norm": 1.1720485330017334, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19828 + }, + { + "epoch": 0.19829, + "grad_norm": 1.4631446086855109, + "learning_rate": 0.003, + "loss": 4.038, + "step": 19829 + }, + { + "epoch": 0.1983, + "grad_norm": 1.27759947935634, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 19830 + }, + { + "epoch": 0.19831, + "grad_norm": 1.3392163108408761, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 19831 + }, + { + "epoch": 0.19832, + "grad_norm": 1.5133191924485834, + "learning_rate": 0.003, + "loss": 4.06, + "step": 19832 + }, + { + "epoch": 0.19833, + "grad_norm": 1.293294990560124, + "learning_rate": 0.003, + "loss": 4.048, + "step": 19833 + }, + { + "epoch": 0.19834, + "grad_norm": 1.6444223465147898, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 19834 + }, + { + "epoch": 0.19835, + "grad_norm": 1.1045129064447632, + "learning_rate": 0.003, + "loss": 4.055, + "step": 19835 + }, + { + "epoch": 0.19836, + "grad_norm": 1.3823524291233054, + "learning_rate": 0.003, + "loss": 4.046, + "step": 19836 + }, + { + "epoch": 0.19837, + "grad_norm": 1.3886344862669628, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 19837 + }, + { + "epoch": 0.19838, + "grad_norm": 1.310600756697459, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 19838 + }, + { + "epoch": 0.19839, + "grad_norm": 1.1322081362500147, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 19839 + }, + { + "epoch": 0.1984, + "grad_norm": 1.6079438090906444, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 19840 + }, + { + "epoch": 0.19841, + "grad_norm": 1.1272952523817377, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 19841 + }, + { + "epoch": 0.19842, + "grad_norm": 1.5750339813943295, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 19842 + }, + { + "epoch": 0.19843, + "grad_norm": 1.2076596936468764, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 19843 + }, + { + "epoch": 0.19844, + "grad_norm": 1.3084089414826487, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19844 + }, + { + "epoch": 0.19845, + "grad_norm": 1.2248463472094597, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 19845 + }, + { + "epoch": 0.19846, + "grad_norm": 1.3091388485192232, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 19846 + }, + { + "epoch": 0.19847, + "grad_norm": 1.2908395318229668, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 19847 + }, + { + "epoch": 0.19848, + "grad_norm": 1.477025855204809, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 19848 + }, + { + "epoch": 0.19849, + "grad_norm": 1.1826392681432139, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 19849 + }, + { + "epoch": 0.1985, + "grad_norm": 1.165434565753471, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 19850 + }, + { + "epoch": 0.19851, + "grad_norm": 1.498316728735124, + "learning_rate": 0.003, + "loss": 4.056, + "step": 19851 + }, + { + "epoch": 0.19852, + "grad_norm": 1.140564156512752, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 19852 + }, + { + "epoch": 0.19853, + "grad_norm": 1.3172906843177683, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 19853 + }, + { + "epoch": 0.19854, + "grad_norm": 1.2591174364083555, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 19854 + }, + { + "epoch": 0.19855, + "grad_norm": 1.209338093553082, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 19855 + }, + { + "epoch": 0.19856, + "grad_norm": 1.2883915670849042, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 19856 + }, + { + "epoch": 0.19857, + "grad_norm": 1.4796169194131135, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 19857 + }, + { + "epoch": 0.19858, + "grad_norm": 1.3598468046038348, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 19858 + }, + { + "epoch": 0.19859, + "grad_norm": 1.4518947651612657, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 19859 + }, + { + "epoch": 0.1986, + "grad_norm": 1.2877903450248325, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 19860 + }, + { + "epoch": 0.19861, + "grad_norm": 1.3644906622993722, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 19861 + }, + { + "epoch": 0.19862, + "grad_norm": 1.3852920489859368, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 19862 + }, + { + "epoch": 0.19863, + "grad_norm": 1.2485641830190142, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 19863 + }, + { + "epoch": 0.19864, + "grad_norm": 1.5908750559888494, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 19864 + }, + { + "epoch": 0.19865, + "grad_norm": 1.0458613828677508, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 19865 + }, + { + "epoch": 0.19866, + "grad_norm": 1.5942201250800003, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19866 + }, + { + "epoch": 0.19867, + "grad_norm": 1.4677573902918055, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 19867 + }, + { + "epoch": 0.19868, + "grad_norm": 1.6739373855135689, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 19868 + }, + { + "epoch": 0.19869, + "grad_norm": 1.2004253899879886, + "learning_rate": 0.003, + "loss": 4.069, + "step": 19869 + }, + { + "epoch": 0.1987, + "grad_norm": 1.1947915365252808, + "learning_rate": 0.003, + "loss": 4.058, + "step": 19870 + }, + { + "epoch": 0.19871, + "grad_norm": 1.3714915182793268, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 19871 + }, + { + "epoch": 0.19872, + "grad_norm": 1.303417716010588, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 19872 + }, + { + "epoch": 0.19873, + "grad_norm": 1.593425253178839, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 19873 + }, + { + "epoch": 0.19874, + "grad_norm": 1.4799633983016245, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 19874 + }, + { + "epoch": 0.19875, + "grad_norm": 1.1939553210421372, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 19875 + }, + { + "epoch": 0.19876, + "grad_norm": 1.4457963114860786, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 19876 + }, + { + "epoch": 0.19877, + "grad_norm": 1.0782302938928832, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 19877 + }, + { + "epoch": 0.19878, + "grad_norm": 1.5576792869516773, + "learning_rate": 0.003, + "loss": 4.075, + "step": 19878 + }, + { + "epoch": 0.19879, + "grad_norm": 0.9795673626634734, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19879 + }, + { + "epoch": 0.1988, + "grad_norm": 1.643192358745065, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 19880 + }, + { + "epoch": 0.19881, + "grad_norm": 1.2499734613264617, + "learning_rate": 0.003, + "loss": 4.069, + "step": 19881 + }, + { + "epoch": 0.19882, + "grad_norm": 1.3012922937892841, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 19882 + }, + { + "epoch": 0.19883, + "grad_norm": 1.681481905552871, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 19883 + }, + { + "epoch": 0.19884, + "grad_norm": 1.179966341422776, + "learning_rate": 0.003, + "loss": 4.086, + "step": 19884 + }, + { + "epoch": 0.19885, + "grad_norm": 1.6502341088662849, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 19885 + }, + { + "epoch": 0.19886, + "grad_norm": 1.4260864819610708, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 19886 + }, + { + "epoch": 0.19887, + "grad_norm": 1.5303099595712761, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 19887 + }, + { + "epoch": 0.19888, + "grad_norm": 1.3805490167116559, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 19888 + }, + { + "epoch": 0.19889, + "grad_norm": 1.1876169220641908, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 19889 + }, + { + "epoch": 0.1989, + "grad_norm": 1.3678792103686264, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 19890 + }, + { + "epoch": 0.19891, + "grad_norm": 1.2240000440553933, + "learning_rate": 0.003, + "loss": 4.067, + "step": 19891 + }, + { + "epoch": 0.19892, + "grad_norm": 1.5087633924656867, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 19892 + }, + { + "epoch": 0.19893, + "grad_norm": 1.0544259075986604, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 19893 + }, + { + "epoch": 0.19894, + "grad_norm": 1.678084831806876, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 19894 + }, + { + "epoch": 0.19895, + "grad_norm": 1.2068767730490293, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 19895 + }, + { + "epoch": 0.19896, + "grad_norm": 1.1556900429965884, + "learning_rate": 0.003, + "loss": 4.052, + "step": 19896 + }, + { + "epoch": 0.19897, + "grad_norm": 1.39828851238043, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 19897 + }, + { + "epoch": 0.19898, + "grad_norm": 1.1750349442371004, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 19898 + }, + { + "epoch": 0.19899, + "grad_norm": 1.7125131283056145, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 19899 + }, + { + "epoch": 0.199, + "grad_norm": 1.0903171062975907, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 19900 + }, + { + "epoch": 0.19901, + "grad_norm": 1.4278968872918296, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 19901 + }, + { + "epoch": 0.19902, + "grad_norm": 1.2769568163873293, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 19902 + }, + { + "epoch": 0.19903, + "grad_norm": 1.320969115171914, + "learning_rate": 0.003, + "loss": 4.084, + "step": 19903 + }, + { + "epoch": 0.19904, + "grad_norm": 1.4373934007265843, + "learning_rate": 0.003, + "loss": 4.071, + "step": 19904 + }, + { + "epoch": 0.19905, + "grad_norm": 1.3337403406343205, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19905 + }, + { + "epoch": 0.19906, + "grad_norm": 1.425015606638654, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 19906 + }, + { + "epoch": 0.19907, + "grad_norm": 1.364814039900154, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 19907 + }, + { + "epoch": 0.19908, + "grad_norm": 1.4000606683834802, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 19908 + }, + { + "epoch": 0.19909, + "grad_norm": 1.349820791493697, + "learning_rate": 0.003, + "loss": 4.045, + "step": 19909 + }, + { + "epoch": 0.1991, + "grad_norm": 1.2101690084272794, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 19910 + }, + { + "epoch": 0.19911, + "grad_norm": 1.328024974881233, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 19911 + }, + { + "epoch": 0.19912, + "grad_norm": 1.5214374495245637, + "learning_rate": 0.003, + "loss": 4.074, + "step": 19912 + }, + { + "epoch": 0.19913, + "grad_norm": 1.2319966053358085, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19913 + }, + { + "epoch": 0.19914, + "grad_norm": 1.624354784976297, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 19914 + }, + { + "epoch": 0.19915, + "grad_norm": 1.1680021177833477, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 19915 + }, + { + "epoch": 0.19916, + "grad_norm": 1.4351009845349718, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 19916 + }, + { + "epoch": 0.19917, + "grad_norm": 1.1904991323865397, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 19917 + }, + { + "epoch": 0.19918, + "grad_norm": 1.3005765935111615, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 19918 + }, + { + "epoch": 0.19919, + "grad_norm": 1.5136706664777948, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 19919 + }, + { + "epoch": 0.1992, + "grad_norm": 1.3631547180491872, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 19920 + }, + { + "epoch": 0.19921, + "grad_norm": 1.3177597213688967, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19921 + }, + { + "epoch": 0.19922, + "grad_norm": 1.3627078741513534, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 19922 + }, + { + "epoch": 0.19923, + "grad_norm": 1.7090650718346874, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 19923 + }, + { + "epoch": 0.19924, + "grad_norm": 1.3016585879344675, + "learning_rate": 0.003, + "loss": 4.063, + "step": 19924 + }, + { + "epoch": 0.19925, + "grad_norm": 1.321937135135949, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 19925 + }, + { + "epoch": 0.19926, + "grad_norm": 1.129513312386755, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 19926 + }, + { + "epoch": 0.19927, + "grad_norm": 1.3347963016520732, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 19927 + }, + { + "epoch": 0.19928, + "grad_norm": 1.1853099035095687, + "learning_rate": 0.003, + "loss": 4.061, + "step": 19928 + }, + { + "epoch": 0.19929, + "grad_norm": 1.335338652852557, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 19929 + }, + { + "epoch": 0.1993, + "grad_norm": 1.2792498490298683, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 19930 + }, + { + "epoch": 0.19931, + "grad_norm": 1.5528370483779794, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 19931 + }, + { + "epoch": 0.19932, + "grad_norm": 1.3599477542356222, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 19932 + }, + { + "epoch": 0.19933, + "grad_norm": 1.1688347790743336, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 19933 + }, + { + "epoch": 0.19934, + "grad_norm": 1.3690338688134256, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 19934 + }, + { + "epoch": 0.19935, + "grad_norm": 1.4862042578677468, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 19935 + }, + { + "epoch": 0.19936, + "grad_norm": 1.6043547474328443, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 19936 + }, + { + "epoch": 0.19937, + "grad_norm": 1.1050656290813126, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 19937 + }, + { + "epoch": 0.19938, + "grad_norm": 1.4952547269417187, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 19938 + }, + { + "epoch": 0.19939, + "grad_norm": 1.194603360033594, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 19939 + }, + { + "epoch": 0.1994, + "grad_norm": 1.665348114353188, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 19940 + }, + { + "epoch": 0.19941, + "grad_norm": 1.1753350624452628, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 19941 + }, + { + "epoch": 0.19942, + "grad_norm": 1.5595213126425294, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 19942 + }, + { + "epoch": 0.19943, + "grad_norm": 1.2802671819523601, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 19943 + }, + { + "epoch": 0.19944, + "grad_norm": 1.3149789591004202, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 19944 + }, + { + "epoch": 0.19945, + "grad_norm": 1.2048649365192403, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 19945 + }, + { + "epoch": 0.19946, + "grad_norm": 1.4394319122791803, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 19946 + }, + { + "epoch": 0.19947, + "grad_norm": 1.2301309484363132, + "learning_rate": 0.003, + "loss": 4.046, + "step": 19947 + }, + { + "epoch": 0.19948, + "grad_norm": 1.4508399657719824, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 19948 + }, + { + "epoch": 0.19949, + "grad_norm": 1.22998463320884, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 19949 + }, + { + "epoch": 0.1995, + "grad_norm": 1.600550233121151, + "learning_rate": 0.003, + "loss": 4.05, + "step": 19950 + }, + { + "epoch": 0.19951, + "grad_norm": 1.1852360744658244, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 19951 + }, + { + "epoch": 0.19952, + "grad_norm": 1.2761907028108868, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 19952 + }, + { + "epoch": 0.19953, + "grad_norm": 1.2186659520098557, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 19953 + }, + { + "epoch": 0.19954, + "grad_norm": 1.7912991145575592, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 19954 + }, + { + "epoch": 0.19955, + "grad_norm": 0.9624001380157303, + "learning_rate": 0.003, + "loss": 4.047, + "step": 19955 + }, + { + "epoch": 0.19956, + "grad_norm": 1.5260929370008876, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 19956 + }, + { + "epoch": 0.19957, + "grad_norm": 1.2051167624010655, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 19957 + }, + { + "epoch": 0.19958, + "grad_norm": 1.468663024710126, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 19958 + }, + { + "epoch": 0.19959, + "grad_norm": 1.2840580166998898, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 19959 + }, + { + "epoch": 0.1996, + "grad_norm": 1.3296993487934532, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 19960 + }, + { + "epoch": 0.19961, + "grad_norm": 1.2756426180152056, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 19961 + }, + { + "epoch": 0.19962, + "grad_norm": 1.1883889884680248, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 19962 + }, + { + "epoch": 0.19963, + "grad_norm": 1.2891980586589682, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 19963 + }, + { + "epoch": 0.19964, + "grad_norm": 1.2864650029059697, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 19964 + }, + { + "epoch": 0.19965, + "grad_norm": 1.364135770276665, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 19965 + }, + { + "epoch": 0.19966, + "grad_norm": 1.475210178191404, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 19966 + }, + { + "epoch": 0.19967, + "grad_norm": 1.180029806466963, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 19967 + }, + { + "epoch": 0.19968, + "grad_norm": 1.4155951959307755, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 19968 + }, + { + "epoch": 0.19969, + "grad_norm": 1.107023638411012, + "learning_rate": 0.003, + "loss": 4.034, + "step": 19969 + }, + { + "epoch": 0.1997, + "grad_norm": 1.4507236517739721, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 19970 + }, + { + "epoch": 0.19971, + "grad_norm": 1.2780514593119168, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 19971 + }, + { + "epoch": 0.19972, + "grad_norm": 1.1632612970969403, + "learning_rate": 0.003, + "loss": 4.052, + "step": 19972 + }, + { + "epoch": 0.19973, + "grad_norm": 1.3913863108877422, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 19973 + }, + { + "epoch": 0.19974, + "grad_norm": 1.143341976288878, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 19974 + }, + { + "epoch": 0.19975, + "grad_norm": 1.3601291769811557, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 19975 + }, + { + "epoch": 0.19976, + "grad_norm": 1.1314455011136493, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 19976 + }, + { + "epoch": 0.19977, + "grad_norm": 1.6772200213837052, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19977 + }, + { + "epoch": 0.19978, + "grad_norm": 1.165581605438778, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 19978 + }, + { + "epoch": 0.19979, + "grad_norm": 1.4402174169262436, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 19979 + }, + { + "epoch": 0.1998, + "grad_norm": 1.4146324527494631, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 19980 + }, + { + "epoch": 0.19981, + "grad_norm": 1.2198891774791902, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 19981 + }, + { + "epoch": 0.19982, + "grad_norm": 1.5295226286415056, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 19982 + }, + { + "epoch": 0.19983, + "grad_norm": 1.1364146683710168, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 19983 + }, + { + "epoch": 0.19984, + "grad_norm": 1.3098536703130816, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 19984 + }, + { + "epoch": 0.19985, + "grad_norm": 1.4299926040866193, + "learning_rate": 0.003, + "loss": 4.052, + "step": 19985 + }, + { + "epoch": 0.19986, + "grad_norm": 1.2957755651925076, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 19986 + }, + { + "epoch": 0.19987, + "grad_norm": 1.7681395903891073, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 19987 + }, + { + "epoch": 0.19988, + "grad_norm": 1.3033406689832867, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 19988 + }, + { + "epoch": 0.19989, + "grad_norm": 1.5434221339459064, + "learning_rate": 0.003, + "loss": 4.067, + "step": 19989 + }, + { + "epoch": 0.1999, + "grad_norm": 1.6724384889930606, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 19990 + }, + { + "epoch": 0.19991, + "grad_norm": 1.1969774103251682, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 19991 + }, + { + "epoch": 0.19992, + "grad_norm": 1.3784163899054613, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 19992 + }, + { + "epoch": 0.19993, + "grad_norm": 1.1385646467829325, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 19993 + }, + { + "epoch": 0.19994, + "grad_norm": 1.5850542355195134, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 19994 + }, + { + "epoch": 0.19995, + "grad_norm": 0.9839155399999102, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 19995 + }, + { + "epoch": 0.19996, + "grad_norm": 1.732207679268672, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 19996 + }, + { + "epoch": 0.19997, + "grad_norm": 1.587619606977346, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 19997 + }, + { + "epoch": 0.19998, + "grad_norm": 1.3720931438155322, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 19998 + }, + { + "epoch": 0.19999, + "grad_norm": 1.189825864203958, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 19999 + }, + { + "epoch": 0.2, + "grad_norm": 1.0731831679830781, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 20000 + }, + { + "epoch": 0.20001, + "grad_norm": 1.3452001360989823, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20001 + }, + { + "epoch": 0.20002, + "grad_norm": 1.2569999204028397, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20002 + }, + { + "epoch": 0.20003, + "grad_norm": 1.1248052065750045, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 20003 + }, + { + "epoch": 0.20004, + "grad_norm": 1.3231987893653487, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 20004 + }, + { + "epoch": 0.20005, + "grad_norm": 1.1075288655403044, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 20005 + }, + { + "epoch": 0.20006, + "grad_norm": 1.76241002894299, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 20006 + }, + { + "epoch": 0.20007, + "grad_norm": 1.1364575908920271, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 20007 + }, + { + "epoch": 0.20008, + "grad_norm": 1.34353856173137, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20008 + }, + { + "epoch": 0.20009, + "grad_norm": 1.713753754243954, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 20009 + }, + { + "epoch": 0.2001, + "grad_norm": 1.2003016467894756, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20010 + }, + { + "epoch": 0.20011, + "grad_norm": 1.3432840288651786, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 20011 + }, + { + "epoch": 0.20012, + "grad_norm": 1.2961175098904854, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 20012 + }, + { + "epoch": 0.20013, + "grad_norm": 1.2737245750977582, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 20013 + }, + { + "epoch": 0.20014, + "grad_norm": 1.3107480433607972, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 20014 + }, + { + "epoch": 0.20015, + "grad_norm": 1.2635527281555388, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20015 + }, + { + "epoch": 0.20016, + "grad_norm": 1.3073564491459384, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 20016 + }, + { + "epoch": 0.20017, + "grad_norm": 1.5120899435045736, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 20017 + }, + { + "epoch": 0.20018, + "grad_norm": 1.4893915911432425, + "learning_rate": 0.003, + "loss": 4.057, + "step": 20018 + }, + { + "epoch": 0.20019, + "grad_norm": 1.0854389602613164, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 20019 + }, + { + "epoch": 0.2002, + "grad_norm": 1.3465381909374021, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20020 + }, + { + "epoch": 0.20021, + "grad_norm": 1.192659101793909, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 20021 + }, + { + "epoch": 0.20022, + "grad_norm": 1.4715970544341472, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 20022 + }, + { + "epoch": 0.20023, + "grad_norm": 1.1248033760576712, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20023 + }, + { + "epoch": 0.20024, + "grad_norm": 1.2968157081760008, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20024 + }, + { + "epoch": 0.20025, + "grad_norm": 1.5507410282592036, + "learning_rate": 0.003, + "loss": 4.044, + "step": 20025 + }, + { + "epoch": 0.20026, + "grad_norm": 1.0551108520430137, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 20026 + }, + { + "epoch": 0.20027, + "grad_norm": 1.604496182642803, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20027 + }, + { + "epoch": 0.20028, + "grad_norm": 1.260339308218672, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 20028 + }, + { + "epoch": 0.20029, + "grad_norm": 1.4292429714224901, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 20029 + }, + { + "epoch": 0.2003, + "grad_norm": 1.332420901476742, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20030 + }, + { + "epoch": 0.20031, + "grad_norm": 1.2147558905404297, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 20031 + }, + { + "epoch": 0.20032, + "grad_norm": 1.261006618127135, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20032 + }, + { + "epoch": 0.20033, + "grad_norm": 1.3681351100810146, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20033 + }, + { + "epoch": 0.20034, + "grad_norm": 1.1859421534830465, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20034 + }, + { + "epoch": 0.20035, + "grad_norm": 1.37908980409705, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 20035 + }, + { + "epoch": 0.20036, + "grad_norm": 1.1969339621114272, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 20036 + }, + { + "epoch": 0.20037, + "grad_norm": 1.4180878572186497, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20037 + }, + { + "epoch": 0.20038, + "grad_norm": 1.2656692876512636, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 20038 + }, + { + "epoch": 0.20039, + "grad_norm": 1.8198151291611133, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20039 + }, + { + "epoch": 0.2004, + "grad_norm": 1.0161078553895786, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 20040 + }, + { + "epoch": 0.20041, + "grad_norm": 1.355432629523141, + "learning_rate": 0.003, + "loss": 4.064, + "step": 20041 + }, + { + "epoch": 0.20042, + "grad_norm": 1.383629325985083, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 20042 + }, + { + "epoch": 0.20043, + "grad_norm": 1.2228534181389905, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 20043 + }, + { + "epoch": 0.20044, + "grad_norm": 1.7296168020885307, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20044 + }, + { + "epoch": 0.20045, + "grad_norm": 1.3540347177465948, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20045 + }, + { + "epoch": 0.20046, + "grad_norm": 1.3239312908186236, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 20046 + }, + { + "epoch": 0.20047, + "grad_norm": 1.2775701830880886, + "learning_rate": 0.003, + "loss": 4.037, + "step": 20047 + }, + { + "epoch": 0.20048, + "grad_norm": 1.3203763932822852, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20048 + }, + { + "epoch": 0.20049, + "grad_norm": 1.1279264776919082, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20049 + }, + { + "epoch": 0.2005, + "grad_norm": 1.3189293017299164, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20050 + }, + { + "epoch": 0.20051, + "grad_norm": 0.9977932116097673, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 20051 + }, + { + "epoch": 0.20052, + "grad_norm": 1.571241154819268, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20052 + }, + { + "epoch": 0.20053, + "grad_norm": 1.1798043208059663, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20053 + }, + { + "epoch": 0.20054, + "grad_norm": 1.4574743878754446, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20054 + }, + { + "epoch": 0.20055, + "grad_norm": 1.4412622360713538, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 20055 + }, + { + "epoch": 0.20056, + "grad_norm": 1.4592961399760058, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 20056 + }, + { + "epoch": 0.20057, + "grad_norm": 1.3012474854060578, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20057 + }, + { + "epoch": 0.20058, + "grad_norm": 1.5786161375868784, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 20058 + }, + { + "epoch": 0.20059, + "grad_norm": 1.1360821546140618, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 20059 + }, + { + "epoch": 0.2006, + "grad_norm": 1.5675781861812292, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 20060 + }, + { + "epoch": 0.20061, + "grad_norm": 1.0718330140613528, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 20061 + }, + { + "epoch": 0.20062, + "grad_norm": 1.5152826138030726, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 20062 + }, + { + "epoch": 0.20063, + "grad_norm": 1.2354227782514513, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 20063 + }, + { + "epoch": 0.20064, + "grad_norm": 1.3823679920207135, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 20064 + }, + { + "epoch": 0.20065, + "grad_norm": 1.4637266564538518, + "learning_rate": 0.003, + "loss": 4.076, + "step": 20065 + }, + { + "epoch": 0.20066, + "grad_norm": 1.1454759865300959, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 20066 + }, + { + "epoch": 0.20067, + "grad_norm": 1.6057158909528744, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 20067 + }, + { + "epoch": 0.20068, + "grad_norm": 1.168781407246637, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 20068 + }, + { + "epoch": 0.20069, + "grad_norm": 1.591690861234611, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 20069 + }, + { + "epoch": 0.2007, + "grad_norm": 1.4144978587605654, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 20070 + }, + { + "epoch": 0.20071, + "grad_norm": 1.1947853938283506, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20071 + }, + { + "epoch": 0.20072, + "grad_norm": 1.5588563229128116, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 20072 + }, + { + "epoch": 0.20073, + "grad_norm": 1.1847033231436441, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 20073 + }, + { + "epoch": 0.20074, + "grad_norm": 1.513444662930399, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20074 + }, + { + "epoch": 0.20075, + "grad_norm": 1.10270547997938, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 20075 + }, + { + "epoch": 0.20076, + "grad_norm": 1.3780518119497263, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 20076 + }, + { + "epoch": 0.20077, + "grad_norm": 1.1144511408052817, + "learning_rate": 0.003, + "loss": 4.028, + "step": 20077 + }, + { + "epoch": 0.20078, + "grad_norm": 1.5146472285977763, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20078 + }, + { + "epoch": 0.20079, + "grad_norm": 1.4284349083888075, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20079 + }, + { + "epoch": 0.2008, + "grad_norm": 1.2989598073195823, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 20080 + }, + { + "epoch": 0.20081, + "grad_norm": 1.187298984538773, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20081 + }, + { + "epoch": 0.20082, + "grad_norm": 1.5113379046885078, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20082 + }, + { + "epoch": 0.20083, + "grad_norm": 1.1541632019018093, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20083 + }, + { + "epoch": 0.20084, + "grad_norm": 1.3668450870562816, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20084 + }, + { + "epoch": 0.20085, + "grad_norm": 1.4203238464339045, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 20085 + }, + { + "epoch": 0.20086, + "grad_norm": 1.5895090610935805, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20086 + }, + { + "epoch": 0.20087, + "grad_norm": 1.095686749121987, + "learning_rate": 0.003, + "loss": 4.063, + "step": 20087 + }, + { + "epoch": 0.20088, + "grad_norm": 1.5489389767527166, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 20088 + }, + { + "epoch": 0.20089, + "grad_norm": 1.1630214840475503, + "learning_rate": 0.003, + "loss": 4.059, + "step": 20089 + }, + { + "epoch": 0.2009, + "grad_norm": 1.3390333966051802, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 20090 + }, + { + "epoch": 0.20091, + "grad_norm": 1.2383088700061902, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 20091 + }, + { + "epoch": 0.20092, + "grad_norm": 1.4327085473639651, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 20092 + }, + { + "epoch": 0.20093, + "grad_norm": 1.1928407656732276, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20093 + }, + { + "epoch": 0.20094, + "grad_norm": 1.3324033273997509, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 20094 + }, + { + "epoch": 0.20095, + "grad_norm": 1.1639919920800843, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 20095 + }, + { + "epoch": 0.20096, + "grad_norm": 1.5326473703914898, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 20096 + }, + { + "epoch": 0.20097, + "grad_norm": 1.469273855798529, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 20097 + }, + { + "epoch": 0.20098, + "grad_norm": 1.4723035892612006, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20098 + }, + { + "epoch": 0.20099, + "grad_norm": 1.233952058675891, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 20099 + }, + { + "epoch": 0.201, + "grad_norm": 1.3370880960112455, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20100 + }, + { + "epoch": 0.20101, + "grad_norm": 1.3418940115999278, + "learning_rate": 0.003, + "loss": 4.038, + "step": 20101 + }, + { + "epoch": 0.20102, + "grad_norm": 1.3806154919944444, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 20102 + }, + { + "epoch": 0.20103, + "grad_norm": 1.280901208133101, + "learning_rate": 0.003, + "loss": 4.061, + "step": 20103 + }, + { + "epoch": 0.20104, + "grad_norm": 1.1118441129872672, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20104 + }, + { + "epoch": 0.20105, + "grad_norm": 1.3672865430967989, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 20105 + }, + { + "epoch": 0.20106, + "grad_norm": 1.262152710425421, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20106 + }, + { + "epoch": 0.20107, + "grad_norm": 1.358097389225668, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20107 + }, + { + "epoch": 0.20108, + "grad_norm": 1.15711306904865, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 20108 + }, + { + "epoch": 0.20109, + "grad_norm": 1.3936933683919779, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20109 + }, + { + "epoch": 0.2011, + "grad_norm": 1.2456943349460141, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 20110 + }, + { + "epoch": 0.20111, + "grad_norm": 1.4010785359454438, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20111 + }, + { + "epoch": 0.20112, + "grad_norm": 1.2792684280319735, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 20112 + }, + { + "epoch": 0.20113, + "grad_norm": 1.1347136211110727, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 20113 + }, + { + "epoch": 0.20114, + "grad_norm": 1.423287069113192, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20114 + }, + { + "epoch": 0.20115, + "grad_norm": 1.1749771510898996, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 20115 + }, + { + "epoch": 0.20116, + "grad_norm": 1.6274457970602212, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 20116 + }, + { + "epoch": 0.20117, + "grad_norm": 1.0200085761280606, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 20117 + }, + { + "epoch": 0.20118, + "grad_norm": 2.1369127590205155, + "learning_rate": 0.003, + "loss": 4.083, + "step": 20118 + }, + { + "epoch": 0.20119, + "grad_norm": 1.1187562929696346, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 20119 + }, + { + "epoch": 0.2012, + "grad_norm": 1.51805607519156, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 20120 + }, + { + "epoch": 0.20121, + "grad_norm": 1.6198867996201227, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 20121 + }, + { + "epoch": 0.20122, + "grad_norm": 1.4554496551998943, + "learning_rate": 0.003, + "loss": 4.107, + "step": 20122 + }, + { + "epoch": 0.20123, + "grad_norm": 1.2883640374402157, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 20123 + }, + { + "epoch": 0.20124, + "grad_norm": 1.620054827451497, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20124 + }, + { + "epoch": 0.20125, + "grad_norm": 1.2614383445274928, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20125 + }, + { + "epoch": 0.20126, + "grad_norm": 1.3521223171358985, + "learning_rate": 0.003, + "loss": 4.046, + "step": 20126 + }, + { + "epoch": 0.20127, + "grad_norm": 1.4850145105267474, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 20127 + }, + { + "epoch": 0.20128, + "grad_norm": 1.234214821383813, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 20128 + }, + { + "epoch": 0.20129, + "grad_norm": 1.29105799526028, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 20129 + }, + { + "epoch": 0.2013, + "grad_norm": 1.224626393915006, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 20130 + }, + { + "epoch": 0.20131, + "grad_norm": 1.3799965897460933, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20131 + }, + { + "epoch": 0.20132, + "grad_norm": 1.2221226479357916, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 20132 + }, + { + "epoch": 0.20133, + "grad_norm": 1.3700604460052845, + "learning_rate": 0.003, + "loss": 4.058, + "step": 20133 + }, + { + "epoch": 0.20134, + "grad_norm": 1.3710883696203422, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 20134 + }, + { + "epoch": 0.20135, + "grad_norm": 1.5385608161996616, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20135 + }, + { + "epoch": 0.20136, + "grad_norm": 1.3333863637238177, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20136 + }, + { + "epoch": 0.20137, + "grad_norm": 1.360397105862845, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 20137 + }, + { + "epoch": 0.20138, + "grad_norm": 1.3658203998710363, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 20138 + }, + { + "epoch": 0.20139, + "grad_norm": 1.5664289313768378, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 20139 + }, + { + "epoch": 0.2014, + "grad_norm": 1.0017139017476433, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 20140 + }, + { + "epoch": 0.20141, + "grad_norm": 1.4446306172886767, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20141 + }, + { + "epoch": 0.20142, + "grad_norm": 1.0921082499045707, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 20142 + }, + { + "epoch": 0.20143, + "grad_norm": 1.4709145539263713, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20143 + }, + { + "epoch": 0.20144, + "grad_norm": 0.9861328787696235, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 20144 + }, + { + "epoch": 0.20145, + "grad_norm": 1.7704188357606225, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20145 + }, + { + "epoch": 0.20146, + "grad_norm": 1.3164527385060978, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 20146 + }, + { + "epoch": 0.20147, + "grad_norm": 1.2712983978275954, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20147 + }, + { + "epoch": 0.20148, + "grad_norm": 1.337966940599898, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 20148 + }, + { + "epoch": 0.20149, + "grad_norm": 1.289917031299455, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 20149 + }, + { + "epoch": 0.2015, + "grad_norm": 1.3383005726772683, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 20150 + }, + { + "epoch": 0.20151, + "grad_norm": 1.419122846015543, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20151 + }, + { + "epoch": 0.20152, + "grad_norm": 1.1713851627259142, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20152 + }, + { + "epoch": 0.20153, + "grad_norm": 1.4930766676175529, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20153 + }, + { + "epoch": 0.20154, + "grad_norm": 1.151981171269532, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 20154 + }, + { + "epoch": 0.20155, + "grad_norm": 1.3762473501599417, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20155 + }, + { + "epoch": 0.20156, + "grad_norm": 1.3151396278511758, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20156 + }, + { + "epoch": 0.20157, + "grad_norm": 1.4038548922797431, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 20157 + }, + { + "epoch": 0.20158, + "grad_norm": 1.11717849011769, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20158 + }, + { + "epoch": 0.20159, + "grad_norm": 1.1990235231336113, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 20159 + }, + { + "epoch": 0.2016, + "grad_norm": 1.2462564898723882, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 20160 + }, + { + "epoch": 0.20161, + "grad_norm": 1.4524608490551825, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 20161 + }, + { + "epoch": 0.20162, + "grad_norm": 1.2838264373462425, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 20162 + }, + { + "epoch": 0.20163, + "grad_norm": 1.3131408545808823, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 20163 + }, + { + "epoch": 0.20164, + "grad_norm": 1.3843159691175067, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 20164 + }, + { + "epoch": 0.20165, + "grad_norm": 1.5100106083950458, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20165 + }, + { + "epoch": 0.20166, + "grad_norm": 1.419607775957611, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 20166 + }, + { + "epoch": 0.20167, + "grad_norm": 1.4835777811373496, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 20167 + }, + { + "epoch": 0.20168, + "grad_norm": 1.3865149314656044, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20168 + }, + { + "epoch": 0.20169, + "grad_norm": 1.1829050529432203, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 20169 + }, + { + "epoch": 0.2017, + "grad_norm": 1.528905891006897, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20170 + }, + { + "epoch": 0.20171, + "grad_norm": 1.06108764356655, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 20171 + }, + { + "epoch": 0.20172, + "grad_norm": 1.5104985676085916, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 20172 + }, + { + "epoch": 0.20173, + "grad_norm": 1.2729083111073791, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 20173 + }, + { + "epoch": 0.20174, + "grad_norm": 1.3503020296505024, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20174 + }, + { + "epoch": 0.20175, + "grad_norm": 1.4827398120112962, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 20175 + }, + { + "epoch": 0.20176, + "grad_norm": 1.2927001966130707, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20176 + }, + { + "epoch": 0.20177, + "grad_norm": 1.4896565621946782, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 20177 + }, + { + "epoch": 0.20178, + "grad_norm": 0.9985785220810762, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20178 + }, + { + "epoch": 0.20179, + "grad_norm": 1.4515332738871436, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 20179 + }, + { + "epoch": 0.2018, + "grad_norm": 1.1484456461245491, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20180 + }, + { + "epoch": 0.20181, + "grad_norm": 1.543886314020279, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20181 + }, + { + "epoch": 0.20182, + "grad_norm": 1.1660738695265007, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20182 + }, + { + "epoch": 0.20183, + "grad_norm": 1.4055135966478245, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 20183 + }, + { + "epoch": 0.20184, + "grad_norm": 1.5541115978355016, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 20184 + }, + { + "epoch": 0.20185, + "grad_norm": 1.2421326626388232, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 20185 + }, + { + "epoch": 0.20186, + "grad_norm": 1.5524946004265114, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 20186 + }, + { + "epoch": 0.20187, + "grad_norm": 1.1019332919822227, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20187 + }, + { + "epoch": 0.20188, + "grad_norm": 1.3527494610479813, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 20188 + }, + { + "epoch": 0.20189, + "grad_norm": 1.2439827387107196, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 20189 + }, + { + "epoch": 0.2019, + "grad_norm": 1.402906416601852, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 20190 + }, + { + "epoch": 0.20191, + "grad_norm": 1.336442867221924, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 20191 + }, + { + "epoch": 0.20192, + "grad_norm": 1.1787087646056902, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20192 + }, + { + "epoch": 0.20193, + "grad_norm": 1.5573011368805436, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 20193 + }, + { + "epoch": 0.20194, + "grad_norm": 1.3902073137712168, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20194 + }, + { + "epoch": 0.20195, + "grad_norm": 1.4901482909029486, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 20195 + }, + { + "epoch": 0.20196, + "grad_norm": 1.3014645418256428, + "learning_rate": 0.003, + "loss": 4.046, + "step": 20196 + }, + { + "epoch": 0.20197, + "grad_norm": 1.3414131733426882, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 20197 + }, + { + "epoch": 0.20198, + "grad_norm": 1.1407214861367891, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 20198 + }, + { + "epoch": 0.20199, + "grad_norm": 1.340005122470549, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 20199 + }, + { + "epoch": 0.202, + "grad_norm": 1.2230698647697285, + "learning_rate": 0.003, + "loss": 4.081, + "step": 20200 + }, + { + "epoch": 0.20201, + "grad_norm": 1.3501677002155696, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20201 + }, + { + "epoch": 0.20202, + "grad_norm": 1.1698990019155124, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 20202 + }, + { + "epoch": 0.20203, + "grad_norm": 1.5175614202195813, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20203 + }, + { + "epoch": 0.20204, + "grad_norm": 1.2589719409047873, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 20204 + }, + { + "epoch": 0.20205, + "grad_norm": 1.3787637881255508, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 20205 + }, + { + "epoch": 0.20206, + "grad_norm": 1.5965370500764264, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 20206 + }, + { + "epoch": 0.20207, + "grad_norm": 1.2845703200278211, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 20207 + }, + { + "epoch": 0.20208, + "grad_norm": 1.397008593604821, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 20208 + }, + { + "epoch": 0.20209, + "grad_norm": 1.2114512364999055, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 20209 + }, + { + "epoch": 0.2021, + "grad_norm": 1.5700928033091597, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20210 + }, + { + "epoch": 0.20211, + "grad_norm": 1.0960466247124028, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 20211 + }, + { + "epoch": 0.20212, + "grad_norm": 1.5013746356631381, + "learning_rate": 0.003, + "loss": 4.061, + "step": 20212 + }, + { + "epoch": 0.20213, + "grad_norm": 1.181832161284184, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20213 + }, + { + "epoch": 0.20214, + "grad_norm": 1.25126739260201, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20214 + }, + { + "epoch": 0.20215, + "grad_norm": 1.4633945227423286, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20215 + }, + { + "epoch": 0.20216, + "grad_norm": 1.4266661902588709, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 20216 + }, + { + "epoch": 0.20217, + "grad_norm": 1.2377326654245766, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 20217 + }, + { + "epoch": 0.20218, + "grad_norm": 1.2758062982531768, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 20218 + }, + { + "epoch": 0.20219, + "grad_norm": 1.2925038112892158, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 20219 + }, + { + "epoch": 0.2022, + "grad_norm": 1.353990795336194, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20220 + }, + { + "epoch": 0.20221, + "grad_norm": 1.5681784043895879, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 20221 + }, + { + "epoch": 0.20222, + "grad_norm": 1.4594748170477143, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20222 + }, + { + "epoch": 0.20223, + "grad_norm": 1.1328992001472526, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 20223 + }, + { + "epoch": 0.20224, + "grad_norm": 1.6070431944351944, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 20224 + }, + { + "epoch": 0.20225, + "grad_norm": 1.3345256871383968, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 20225 + }, + { + "epoch": 0.20226, + "grad_norm": 1.2039936279757504, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20226 + }, + { + "epoch": 0.20227, + "grad_norm": 1.4112783524690296, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 20227 + }, + { + "epoch": 0.20228, + "grad_norm": 1.0879874829746985, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 20228 + }, + { + "epoch": 0.20229, + "grad_norm": 1.406909773014083, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 20229 + }, + { + "epoch": 0.2023, + "grad_norm": 1.5073298460839546, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 20230 + }, + { + "epoch": 0.20231, + "grad_norm": 1.4725979263205242, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 20231 + }, + { + "epoch": 0.20232, + "grad_norm": 1.3097844389355746, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 20232 + }, + { + "epoch": 0.20233, + "grad_norm": 1.1974186108290308, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 20233 + }, + { + "epoch": 0.20234, + "grad_norm": 1.4660146901812041, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 20234 + }, + { + "epoch": 0.20235, + "grad_norm": 1.0145110696085622, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 20235 + }, + { + "epoch": 0.20236, + "grad_norm": 1.4059391654413957, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 20236 + }, + { + "epoch": 0.20237, + "grad_norm": 1.1952217406063508, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 20237 + }, + { + "epoch": 0.20238, + "grad_norm": 1.6725738519887603, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 20238 + }, + { + "epoch": 0.20239, + "grad_norm": 1.1162383117767734, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 20239 + }, + { + "epoch": 0.2024, + "grad_norm": 1.6663407668463703, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 20240 + }, + { + "epoch": 0.20241, + "grad_norm": 1.461105874407114, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 20241 + }, + { + "epoch": 0.20242, + "grad_norm": 1.7513584676973888, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 20242 + }, + { + "epoch": 0.20243, + "grad_norm": 1.0803860595038803, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 20243 + }, + { + "epoch": 0.20244, + "grad_norm": 1.3010710656545241, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 20244 + }, + { + "epoch": 0.20245, + "grad_norm": 1.3533895832189775, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 20245 + }, + { + "epoch": 0.20246, + "grad_norm": 1.4133369311925101, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 20246 + }, + { + "epoch": 0.20247, + "grad_norm": 1.2584835076219352, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 20247 + }, + { + "epoch": 0.20248, + "grad_norm": 1.4737037157476898, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 20248 + }, + { + "epoch": 0.20249, + "grad_norm": 1.1550526835609105, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 20249 + }, + { + "epoch": 0.2025, + "grad_norm": 1.3141195332757687, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 20250 + }, + { + "epoch": 0.20251, + "grad_norm": 1.2685653761651985, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 20251 + }, + { + "epoch": 0.20252, + "grad_norm": 1.3264547028903149, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20252 + }, + { + "epoch": 0.20253, + "grad_norm": 1.217633255514227, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20253 + }, + { + "epoch": 0.20254, + "grad_norm": 1.25743030026997, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 20254 + }, + { + "epoch": 0.20255, + "grad_norm": 1.4676744069862193, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 20255 + }, + { + "epoch": 0.20256, + "grad_norm": 1.2844555774592705, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20256 + }, + { + "epoch": 0.20257, + "grad_norm": 1.2091318047712682, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 20257 + }, + { + "epoch": 0.20258, + "grad_norm": 1.2943925744024822, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 20258 + }, + { + "epoch": 0.20259, + "grad_norm": 1.7294020948011957, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 20259 + }, + { + "epoch": 0.2026, + "grad_norm": 1.6668637325189721, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20260 + }, + { + "epoch": 0.20261, + "grad_norm": 1.185731719878189, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 20261 + }, + { + "epoch": 0.20262, + "grad_norm": 1.4799478765469893, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20262 + }, + { + "epoch": 0.20263, + "grad_norm": 1.5977095992613186, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20263 + }, + { + "epoch": 0.20264, + "grad_norm": 1.2166178118311102, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20264 + }, + { + "epoch": 0.20265, + "grad_norm": 1.1564296247113104, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 20265 + }, + { + "epoch": 0.20266, + "grad_norm": 1.7171201542819816, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20266 + }, + { + "epoch": 0.20267, + "grad_norm": 1.0281637559809484, + "learning_rate": 0.003, + "loss": 4.056, + "step": 20267 + }, + { + "epoch": 0.20268, + "grad_norm": 1.3917162972146064, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 20268 + }, + { + "epoch": 0.20269, + "grad_norm": 1.160640192991406, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20269 + }, + { + "epoch": 0.2027, + "grad_norm": 1.4939895821106073, + "learning_rate": 0.003, + "loss": 4.066, + "step": 20270 + }, + { + "epoch": 0.20271, + "grad_norm": 1.1561605112219357, + "learning_rate": 0.003, + "loss": 4.078, + "step": 20271 + }, + { + "epoch": 0.20272, + "grad_norm": 1.308674679046234, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 20272 + }, + { + "epoch": 0.20273, + "grad_norm": 1.1854461944313113, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 20273 + }, + { + "epoch": 0.20274, + "grad_norm": 1.5477263071366116, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 20274 + }, + { + "epoch": 0.20275, + "grad_norm": 1.2058426257177344, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 20275 + }, + { + "epoch": 0.20276, + "grad_norm": 1.3685886984542353, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20276 + }, + { + "epoch": 0.20277, + "grad_norm": 1.553754341867869, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 20277 + }, + { + "epoch": 0.20278, + "grad_norm": 1.2709523206304474, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 20278 + }, + { + "epoch": 0.20279, + "grad_norm": 1.3725987808318894, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 20279 + }, + { + "epoch": 0.2028, + "grad_norm": 1.2624941388485218, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 20280 + }, + { + "epoch": 0.20281, + "grad_norm": 1.233807790705302, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 20281 + }, + { + "epoch": 0.20282, + "grad_norm": 1.2521895323654215, + "learning_rate": 0.003, + "loss": 4.056, + "step": 20282 + }, + { + "epoch": 0.20283, + "grad_norm": 1.4753074404834448, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 20283 + }, + { + "epoch": 0.20284, + "grad_norm": 1.4409322853168969, + "learning_rate": 0.003, + "loss": 4.055, + "step": 20284 + }, + { + "epoch": 0.20285, + "grad_norm": 1.504716079962182, + "learning_rate": 0.003, + "loss": 4.053, + "step": 20285 + }, + { + "epoch": 0.20286, + "grad_norm": 1.2126978833429751, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 20286 + }, + { + "epoch": 0.20287, + "grad_norm": 1.3542455534211755, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 20287 + }, + { + "epoch": 0.20288, + "grad_norm": 1.076493558839879, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 20288 + }, + { + "epoch": 0.20289, + "grad_norm": 1.4439817719258055, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20289 + }, + { + "epoch": 0.2029, + "grad_norm": 1.303195151449773, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 20290 + }, + { + "epoch": 0.20291, + "grad_norm": 1.4547363156693787, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 20291 + }, + { + "epoch": 0.20292, + "grad_norm": 1.3927461978685791, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20292 + }, + { + "epoch": 0.20293, + "grad_norm": 1.2390192112833431, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20293 + }, + { + "epoch": 0.20294, + "grad_norm": 1.3637822144285798, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 20294 + }, + { + "epoch": 0.20295, + "grad_norm": 1.242130596796715, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20295 + }, + { + "epoch": 0.20296, + "grad_norm": 1.3901310772605913, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 20296 + }, + { + "epoch": 0.20297, + "grad_norm": 1.4037666856257323, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 20297 + }, + { + "epoch": 0.20298, + "grad_norm": 1.1758694063275068, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 20298 + }, + { + "epoch": 0.20299, + "grad_norm": 1.3623746577047566, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 20299 + }, + { + "epoch": 0.203, + "grad_norm": 1.4085414522927322, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 20300 + }, + { + "epoch": 0.20301, + "grad_norm": 1.5396625511929816, + "learning_rate": 0.003, + "loss": 4.056, + "step": 20301 + }, + { + "epoch": 0.20302, + "grad_norm": 1.266867217134575, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 20302 + }, + { + "epoch": 0.20303, + "grad_norm": 1.5175617191016293, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 20303 + }, + { + "epoch": 0.20304, + "grad_norm": 1.1821512523591957, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 20304 + }, + { + "epoch": 0.20305, + "grad_norm": 1.4226176539591662, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 20305 + }, + { + "epoch": 0.20306, + "grad_norm": 1.050703674193398, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 20306 + }, + { + "epoch": 0.20307, + "grad_norm": 1.521939154776105, + "learning_rate": 0.003, + "loss": 4.067, + "step": 20307 + }, + { + "epoch": 0.20308, + "grad_norm": 1.33044787968948, + "learning_rate": 0.003, + "loss": 4.054, + "step": 20308 + }, + { + "epoch": 0.20309, + "grad_norm": 1.4484830287900583, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 20309 + }, + { + "epoch": 0.2031, + "grad_norm": 1.5269447970098096, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 20310 + }, + { + "epoch": 0.20311, + "grad_norm": 1.4328106234116684, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20311 + }, + { + "epoch": 0.20312, + "grad_norm": 1.178726146404056, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20312 + }, + { + "epoch": 0.20313, + "grad_norm": 1.5048542215378433, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 20313 + }, + { + "epoch": 0.20314, + "grad_norm": 1.1369721257827141, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20314 + }, + { + "epoch": 0.20315, + "grad_norm": 1.5531515177048691, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 20315 + }, + { + "epoch": 0.20316, + "grad_norm": 1.2364073011949877, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 20316 + }, + { + "epoch": 0.20317, + "grad_norm": 1.5190111676246623, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 20317 + }, + { + "epoch": 0.20318, + "grad_norm": 1.281037111926793, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20318 + }, + { + "epoch": 0.20319, + "grad_norm": 1.2859070666158015, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20319 + }, + { + "epoch": 0.2032, + "grad_norm": 1.092940737832305, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 20320 + }, + { + "epoch": 0.20321, + "grad_norm": 1.3967347723396333, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 20321 + }, + { + "epoch": 0.20322, + "grad_norm": 1.2313847457891551, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20322 + }, + { + "epoch": 0.20323, + "grad_norm": 1.574469621303561, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 20323 + }, + { + "epoch": 0.20324, + "grad_norm": 1.2035069311206108, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20324 + }, + { + "epoch": 0.20325, + "grad_norm": 1.5439161847441938, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 20325 + }, + { + "epoch": 0.20326, + "grad_norm": 1.0239332953056914, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 20326 + }, + { + "epoch": 0.20327, + "grad_norm": 1.5932857943493817, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 20327 + }, + { + "epoch": 0.20328, + "grad_norm": 1.1140383020616065, + "learning_rate": 0.003, + "loss": 4.077, + "step": 20328 + }, + { + "epoch": 0.20329, + "grad_norm": 1.1682440481298149, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 20329 + }, + { + "epoch": 0.2033, + "grad_norm": 1.5287704639769066, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20330 + }, + { + "epoch": 0.20331, + "grad_norm": 1.2779345322703561, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20331 + }, + { + "epoch": 0.20332, + "grad_norm": 1.3815575764425685, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 20332 + }, + { + "epoch": 0.20333, + "grad_norm": 1.3061960024303134, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 20333 + }, + { + "epoch": 0.20334, + "grad_norm": 1.3891872739275988, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 20334 + }, + { + "epoch": 0.20335, + "grad_norm": 1.408257164633742, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20335 + }, + { + "epoch": 0.20336, + "grad_norm": 1.3995418314161383, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20336 + }, + { + "epoch": 0.20337, + "grad_norm": 1.3395419724950246, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 20337 + }, + { + "epoch": 0.20338, + "grad_norm": 1.2468307107462475, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20338 + }, + { + "epoch": 0.20339, + "grad_norm": 1.4289246140290377, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 20339 + }, + { + "epoch": 0.2034, + "grad_norm": 1.2476588970388962, + "learning_rate": 0.003, + "loss": 4.086, + "step": 20340 + }, + { + "epoch": 0.20341, + "grad_norm": 1.3422524370970599, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 20341 + }, + { + "epoch": 0.20342, + "grad_norm": 1.0577462129023343, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 20342 + }, + { + "epoch": 0.20343, + "grad_norm": 1.7264542407238392, + "learning_rate": 0.003, + "loss": 4.094, + "step": 20343 + }, + { + "epoch": 0.20344, + "grad_norm": 1.2643432223711861, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20344 + }, + { + "epoch": 0.20345, + "grad_norm": 1.529773062601106, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20345 + }, + { + "epoch": 0.20346, + "grad_norm": 1.4361889453096046, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 20346 + }, + { + "epoch": 0.20347, + "grad_norm": 1.499920858801668, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 20347 + }, + { + "epoch": 0.20348, + "grad_norm": 1.6738367304447712, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 20348 + }, + { + "epoch": 0.20349, + "grad_norm": 1.200930809138132, + "learning_rate": 0.003, + "loss": 4.079, + "step": 20349 + }, + { + "epoch": 0.2035, + "grad_norm": 1.467013383787434, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 20350 + }, + { + "epoch": 0.20351, + "grad_norm": 1.5679824072008346, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 20351 + }, + { + "epoch": 0.20352, + "grad_norm": 1.127251107861855, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20352 + }, + { + "epoch": 0.20353, + "grad_norm": 1.629146596022005, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 20353 + }, + { + "epoch": 0.20354, + "grad_norm": 1.183836849526088, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20354 + }, + { + "epoch": 0.20355, + "grad_norm": 1.228857414739491, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 20355 + }, + { + "epoch": 0.20356, + "grad_norm": 1.3416733588472254, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20356 + }, + { + "epoch": 0.20357, + "grad_norm": 1.3372111342595498, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 20357 + }, + { + "epoch": 0.20358, + "grad_norm": 1.2538288044204846, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 20358 + }, + { + "epoch": 0.20359, + "grad_norm": 1.3959090932702416, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 20359 + }, + { + "epoch": 0.2036, + "grad_norm": 1.3694363348608283, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 20360 + }, + { + "epoch": 0.20361, + "grad_norm": 1.5034844585728124, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 20361 + }, + { + "epoch": 0.20362, + "grad_norm": 1.4239047926456798, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20362 + }, + { + "epoch": 0.20363, + "grad_norm": 1.2520521246135627, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20363 + }, + { + "epoch": 0.20364, + "grad_norm": 1.2412612514439594, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 20364 + }, + { + "epoch": 0.20365, + "grad_norm": 1.3948056814887138, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20365 + }, + { + "epoch": 0.20366, + "grad_norm": 1.2792736518242112, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20366 + }, + { + "epoch": 0.20367, + "grad_norm": 1.1054230687849584, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20367 + }, + { + "epoch": 0.20368, + "grad_norm": 1.3011606662823947, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 20368 + }, + { + "epoch": 0.20369, + "grad_norm": 1.2803734189697848, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 20369 + }, + { + "epoch": 0.2037, + "grad_norm": 1.3831472596596728, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 20370 + }, + { + "epoch": 0.20371, + "grad_norm": 1.4680128368690626, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20371 + }, + { + "epoch": 0.20372, + "grad_norm": 1.364108438076807, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20372 + }, + { + "epoch": 0.20373, + "grad_norm": 1.4398845101211322, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 20373 + }, + { + "epoch": 0.20374, + "grad_norm": 1.3016675770978412, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20374 + }, + { + "epoch": 0.20375, + "grad_norm": 1.440576518128247, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 20375 + }, + { + "epoch": 0.20376, + "grad_norm": 1.222119920685402, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 20376 + }, + { + "epoch": 0.20377, + "grad_norm": 1.515461877623012, + "learning_rate": 0.003, + "loss": 4.081, + "step": 20377 + }, + { + "epoch": 0.20378, + "grad_norm": 1.2135329370512657, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 20378 + }, + { + "epoch": 0.20379, + "grad_norm": 1.5897915340651947, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 20379 + }, + { + "epoch": 0.2038, + "grad_norm": 1.0645381777011922, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20380 + }, + { + "epoch": 0.20381, + "grad_norm": 1.6386763563805693, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20381 + }, + { + "epoch": 0.20382, + "grad_norm": 1.1424167086168668, + "learning_rate": 0.003, + "loss": 4.042, + "step": 20382 + }, + { + "epoch": 0.20383, + "grad_norm": 1.3281632988062138, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 20383 + }, + { + "epoch": 0.20384, + "grad_norm": 1.5934718851990406, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 20384 + }, + { + "epoch": 0.20385, + "grad_norm": 1.0341667755535273, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20385 + }, + { + "epoch": 0.20386, + "grad_norm": 1.4459015184338975, + "learning_rate": 0.003, + "loss": 4.061, + "step": 20386 + }, + { + "epoch": 0.20387, + "grad_norm": 1.3366501262145274, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 20387 + }, + { + "epoch": 0.20388, + "grad_norm": 1.3915743056060734, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 20388 + }, + { + "epoch": 0.20389, + "grad_norm": 1.346962033954529, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 20389 + }, + { + "epoch": 0.2039, + "grad_norm": 1.1552000293677478, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 20390 + }, + { + "epoch": 0.20391, + "grad_norm": 1.5605764802292823, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 20391 + }, + { + "epoch": 0.20392, + "grad_norm": 1.1987219339139694, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 20392 + }, + { + "epoch": 0.20393, + "grad_norm": 1.2454690242319035, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 20393 + }, + { + "epoch": 0.20394, + "grad_norm": 1.2256636987965361, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20394 + }, + { + "epoch": 0.20395, + "grad_norm": 1.2867586718472315, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 20395 + }, + { + "epoch": 0.20396, + "grad_norm": 1.1965507261055377, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 20396 + }, + { + "epoch": 0.20397, + "grad_norm": 1.4592698624576914, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 20397 + }, + { + "epoch": 0.20398, + "grad_norm": 1.5161036939976629, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 20398 + }, + { + "epoch": 0.20399, + "grad_norm": 1.2383093577717579, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 20399 + }, + { + "epoch": 0.204, + "grad_norm": 1.569877039855125, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 20400 + }, + { + "epoch": 0.20401, + "grad_norm": 1.1692225184397718, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20401 + }, + { + "epoch": 0.20402, + "grad_norm": 1.5226657623961934, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20402 + }, + { + "epoch": 0.20403, + "grad_norm": 1.1749212004945278, + "learning_rate": 0.003, + "loss": 4.039, + "step": 20403 + }, + { + "epoch": 0.20404, + "grad_norm": 1.6261828343359845, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20404 + }, + { + "epoch": 0.20405, + "grad_norm": 1.2151944554580743, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 20405 + }, + { + "epoch": 0.20406, + "grad_norm": 1.2945642192787308, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 20406 + }, + { + "epoch": 0.20407, + "grad_norm": 1.4069555500329456, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 20407 + }, + { + "epoch": 0.20408, + "grad_norm": 1.3830004642020512, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 20408 + }, + { + "epoch": 0.20409, + "grad_norm": 1.4319706958306666, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20409 + }, + { + "epoch": 0.2041, + "grad_norm": 1.1870627800044244, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20410 + }, + { + "epoch": 0.20411, + "grad_norm": 1.3968933589457344, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 20411 + }, + { + "epoch": 0.20412, + "grad_norm": 1.2634991756889127, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20412 + }, + { + "epoch": 0.20413, + "grad_norm": 1.3074650979006743, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 20413 + }, + { + "epoch": 0.20414, + "grad_norm": 1.4663802224254836, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 20414 + }, + { + "epoch": 0.20415, + "grad_norm": 1.1745989685736145, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 20415 + }, + { + "epoch": 0.20416, + "grad_norm": 1.7716874707579042, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 20416 + }, + { + "epoch": 0.20417, + "grad_norm": 1.1465897204395348, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 20417 + }, + { + "epoch": 0.20418, + "grad_norm": 1.4541160183866086, + "learning_rate": 0.003, + "loss": 4.046, + "step": 20418 + }, + { + "epoch": 0.20419, + "grad_norm": 1.300364414594133, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20419 + }, + { + "epoch": 0.2042, + "grad_norm": 1.1566286092158102, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 20420 + }, + { + "epoch": 0.20421, + "grad_norm": 1.4019025899636917, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20421 + }, + { + "epoch": 0.20422, + "grad_norm": 1.209222857411128, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 20422 + }, + { + "epoch": 0.20423, + "grad_norm": 1.3496282306334784, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 20423 + }, + { + "epoch": 0.20424, + "grad_norm": 1.2433548633857987, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20424 + }, + { + "epoch": 0.20425, + "grad_norm": 1.3455150458622027, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 20425 + }, + { + "epoch": 0.20426, + "grad_norm": 1.1363945182804216, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 20426 + }, + { + "epoch": 0.20427, + "grad_norm": 1.447084854366867, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 20427 + }, + { + "epoch": 0.20428, + "grad_norm": 1.349916594398312, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20428 + }, + { + "epoch": 0.20429, + "grad_norm": 1.4837193509187336, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 20429 + }, + { + "epoch": 0.2043, + "grad_norm": 1.215639767974536, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 20430 + }, + { + "epoch": 0.20431, + "grad_norm": 1.8112911372440925, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 20431 + }, + { + "epoch": 0.20432, + "grad_norm": 1.3158951404821044, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20432 + }, + { + "epoch": 0.20433, + "grad_norm": 1.1216513036257185, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20433 + }, + { + "epoch": 0.20434, + "grad_norm": 1.5943047717634269, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 20434 + }, + { + "epoch": 0.20435, + "grad_norm": 1.343181365017431, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 20435 + }, + { + "epoch": 0.20436, + "grad_norm": 1.3464601497928512, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 20436 + }, + { + "epoch": 0.20437, + "grad_norm": 1.4358280375881376, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20437 + }, + { + "epoch": 0.20438, + "grad_norm": 1.4503941072369733, + "learning_rate": 0.003, + "loss": 4.039, + "step": 20438 + }, + { + "epoch": 0.20439, + "grad_norm": 1.145430636880698, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 20439 + }, + { + "epoch": 0.2044, + "grad_norm": 1.4974558111213008, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20440 + }, + { + "epoch": 0.20441, + "grad_norm": 1.3053873478983593, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20441 + }, + { + "epoch": 0.20442, + "grad_norm": 1.3672662373830282, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20442 + }, + { + "epoch": 0.20443, + "grad_norm": 1.3817825636432517, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 20443 + }, + { + "epoch": 0.20444, + "grad_norm": 1.1952694223765734, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 20444 + }, + { + "epoch": 0.20445, + "grad_norm": 1.4394061531641331, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 20445 + }, + { + "epoch": 0.20446, + "grad_norm": 1.3333207449315254, + "learning_rate": 0.003, + "loss": 4.057, + "step": 20446 + }, + { + "epoch": 0.20447, + "grad_norm": 1.4515805384860727, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 20447 + }, + { + "epoch": 0.20448, + "grad_norm": 1.2508524069591107, + "learning_rate": 0.003, + "loss": 4.058, + "step": 20448 + }, + { + "epoch": 0.20449, + "grad_norm": 1.29362758082749, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 20449 + }, + { + "epoch": 0.2045, + "grad_norm": 1.4251175560020428, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 20450 + }, + { + "epoch": 0.20451, + "grad_norm": 1.3396049861183335, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 20451 + }, + { + "epoch": 0.20452, + "grad_norm": 1.271441360928029, + "learning_rate": 0.003, + "loss": 4.032, + "step": 20452 + }, + { + "epoch": 0.20453, + "grad_norm": 1.208530742162209, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 20453 + }, + { + "epoch": 0.20454, + "grad_norm": 1.4584881237128287, + "learning_rate": 0.003, + "loss": 4.075, + "step": 20454 + }, + { + "epoch": 0.20455, + "grad_norm": 1.0092589910193805, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20455 + }, + { + "epoch": 0.20456, + "grad_norm": 1.4809515743739938, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20456 + }, + { + "epoch": 0.20457, + "grad_norm": 1.2503919073180787, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 20457 + }, + { + "epoch": 0.20458, + "grad_norm": 1.3325286512530745, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 20458 + }, + { + "epoch": 0.20459, + "grad_norm": 1.2101473095462296, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 20459 + }, + { + "epoch": 0.2046, + "grad_norm": 1.3251864349288356, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 20460 + }, + { + "epoch": 0.20461, + "grad_norm": 1.1389725466140332, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 20461 + }, + { + "epoch": 0.20462, + "grad_norm": 1.542600041658139, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 20462 + }, + { + "epoch": 0.20463, + "grad_norm": 1.3023259888264964, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 20463 + }, + { + "epoch": 0.20464, + "grad_norm": 1.4764643574700729, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 20464 + }, + { + "epoch": 0.20465, + "grad_norm": 1.1186142088971804, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20465 + }, + { + "epoch": 0.20466, + "grad_norm": 2.0460897971041487, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 20466 + }, + { + "epoch": 0.20467, + "grad_norm": 1.12972253606819, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 20467 + }, + { + "epoch": 0.20468, + "grad_norm": 1.5297144853412366, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 20468 + }, + { + "epoch": 0.20469, + "grad_norm": 1.0782410690750979, + "learning_rate": 0.003, + "loss": 4.081, + "step": 20469 + }, + { + "epoch": 0.2047, + "grad_norm": 1.5580879093447175, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 20470 + }, + { + "epoch": 0.20471, + "grad_norm": 1.0258224498191024, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20471 + }, + { + "epoch": 0.20472, + "grad_norm": 1.5514806475100573, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 20472 + }, + { + "epoch": 0.20473, + "grad_norm": 1.1637118520595569, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 20473 + }, + { + "epoch": 0.20474, + "grad_norm": 1.8429010158726027, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 20474 + }, + { + "epoch": 0.20475, + "grad_norm": 1.428374246002246, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 20475 + }, + { + "epoch": 0.20476, + "grad_norm": 1.2831870445809457, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 20476 + }, + { + "epoch": 0.20477, + "grad_norm": 1.2062427063284045, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20477 + }, + { + "epoch": 0.20478, + "grad_norm": 1.5416075304100199, + "learning_rate": 0.003, + "loss": 4.062, + "step": 20478 + }, + { + "epoch": 0.20479, + "grad_norm": 1.2029394842621126, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 20479 + }, + { + "epoch": 0.2048, + "grad_norm": 1.569242237429814, + "learning_rate": 0.003, + "loss": 4.082, + "step": 20480 + }, + { + "epoch": 0.20481, + "grad_norm": 1.2023060552514346, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20481 + }, + { + "epoch": 0.20482, + "grad_norm": 1.3671008704673404, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20482 + }, + { + "epoch": 0.20483, + "grad_norm": 1.2181550284200213, + "learning_rate": 0.003, + "loss": 4.072, + "step": 20483 + }, + { + "epoch": 0.20484, + "grad_norm": 1.2107098831651844, + "learning_rate": 0.003, + "loss": 4.08, + "step": 20484 + }, + { + "epoch": 0.20485, + "grad_norm": 1.2914939061169899, + "learning_rate": 0.003, + "loss": 4.094, + "step": 20485 + }, + { + "epoch": 0.20486, + "grad_norm": 1.6783530589798497, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 20486 + }, + { + "epoch": 0.20487, + "grad_norm": 1.485364830617809, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 20487 + }, + { + "epoch": 0.20488, + "grad_norm": 1.4099567115333826, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 20488 + }, + { + "epoch": 0.20489, + "grad_norm": 1.495916460389477, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 20489 + }, + { + "epoch": 0.2049, + "grad_norm": 1.178450710384942, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 20490 + }, + { + "epoch": 0.20491, + "grad_norm": 1.6912628484171894, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 20491 + }, + { + "epoch": 0.20492, + "grad_norm": 1.123540991095715, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20492 + }, + { + "epoch": 0.20493, + "grad_norm": 1.6529649747407045, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 20493 + }, + { + "epoch": 0.20494, + "grad_norm": 1.1509127886685253, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20494 + }, + { + "epoch": 0.20495, + "grad_norm": 1.4774965511998783, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20495 + }, + { + "epoch": 0.20496, + "grad_norm": 1.6530410144546603, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20496 + }, + { + "epoch": 0.20497, + "grad_norm": 0.9380240352308407, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 20497 + }, + { + "epoch": 0.20498, + "grad_norm": 1.3264095358263313, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 20498 + }, + { + "epoch": 0.20499, + "grad_norm": 1.4263106225259305, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 20499 + }, + { + "epoch": 0.205, + "grad_norm": 1.258423524127491, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 20500 + }, + { + "epoch": 0.20501, + "grad_norm": 1.0772749489987943, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 20501 + }, + { + "epoch": 0.20502, + "grad_norm": 1.6768436558193698, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 20502 + }, + { + "epoch": 0.20503, + "grad_norm": 1.3719749181356573, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20503 + }, + { + "epoch": 0.20504, + "grad_norm": 1.3076034976415154, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 20504 + }, + { + "epoch": 0.20505, + "grad_norm": 1.503886625040588, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20505 + }, + { + "epoch": 0.20506, + "grad_norm": 1.4609689972531943, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 20506 + }, + { + "epoch": 0.20507, + "grad_norm": 1.3549616497833068, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 20507 + }, + { + "epoch": 0.20508, + "grad_norm": 1.4939887798151368, + "learning_rate": 0.003, + "loss": 4.064, + "step": 20508 + }, + { + "epoch": 0.20509, + "grad_norm": 1.4217893187549624, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 20509 + }, + { + "epoch": 0.2051, + "grad_norm": 1.1819692988912474, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 20510 + }, + { + "epoch": 0.20511, + "grad_norm": 1.6665703317991702, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 20511 + }, + { + "epoch": 0.20512, + "grad_norm": 1.1581066444434676, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 20512 + }, + { + "epoch": 0.20513, + "grad_norm": 1.347662207439549, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 20513 + }, + { + "epoch": 0.20514, + "grad_norm": 1.3587040372607169, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 20514 + }, + { + "epoch": 0.20515, + "grad_norm": 1.3475403635569718, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 20515 + }, + { + "epoch": 0.20516, + "grad_norm": 1.1470955731703492, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20516 + }, + { + "epoch": 0.20517, + "grad_norm": 1.2038765316619247, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 20517 + }, + { + "epoch": 0.20518, + "grad_norm": 1.3109237023488451, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 20518 + }, + { + "epoch": 0.20519, + "grad_norm": 1.2361177572036106, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 20519 + }, + { + "epoch": 0.2052, + "grad_norm": 1.4796222856008112, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 20520 + }, + { + "epoch": 0.20521, + "grad_norm": 1.2531185153138025, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 20521 + }, + { + "epoch": 0.20522, + "grad_norm": 1.730612273552305, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 20522 + }, + { + "epoch": 0.20523, + "grad_norm": 1.2706772945225375, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20523 + }, + { + "epoch": 0.20524, + "grad_norm": 1.1398531348530765, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 20524 + }, + { + "epoch": 0.20525, + "grad_norm": 1.3519082464893908, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20525 + }, + { + "epoch": 0.20526, + "grad_norm": 1.4760340663790463, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 20526 + }, + { + "epoch": 0.20527, + "grad_norm": 1.2208782065493444, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 20527 + }, + { + "epoch": 0.20528, + "grad_norm": 1.5059266709517738, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 20528 + }, + { + "epoch": 0.20529, + "grad_norm": 1.195595525255474, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 20529 + }, + { + "epoch": 0.2053, + "grad_norm": 1.4505639511754715, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 20530 + }, + { + "epoch": 0.20531, + "grad_norm": 1.18422918279392, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20531 + }, + { + "epoch": 0.20532, + "grad_norm": 1.4236984485357382, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 20532 + }, + { + "epoch": 0.20533, + "grad_norm": 1.8410001515180068, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 20533 + }, + { + "epoch": 0.20534, + "grad_norm": 0.8771493260205179, + "learning_rate": 0.003, + "loss": 4.058, + "step": 20534 + }, + { + "epoch": 0.20535, + "grad_norm": 1.3573410739858818, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 20535 + }, + { + "epoch": 0.20536, + "grad_norm": 1.3291407846772125, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 20536 + }, + { + "epoch": 0.20537, + "grad_norm": 1.2363419084100955, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20537 + }, + { + "epoch": 0.20538, + "grad_norm": 1.2954414879710934, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 20538 + }, + { + "epoch": 0.20539, + "grad_norm": 1.244450601937851, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20539 + }, + { + "epoch": 0.2054, + "grad_norm": 1.3704835928868393, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 20540 + }, + { + "epoch": 0.20541, + "grad_norm": 1.2790535844257738, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20541 + }, + { + "epoch": 0.20542, + "grad_norm": 1.584307647377643, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 20542 + }, + { + "epoch": 0.20543, + "grad_norm": 1.5100470754899271, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 20543 + }, + { + "epoch": 0.20544, + "grad_norm": 1.2252937633668737, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 20544 + }, + { + "epoch": 0.20545, + "grad_norm": 1.2938591800291355, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 20545 + }, + { + "epoch": 0.20546, + "grad_norm": 1.4471883307764137, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 20546 + }, + { + "epoch": 0.20547, + "grad_norm": 1.5025869751133516, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 20547 + }, + { + "epoch": 0.20548, + "grad_norm": 1.0359233811834558, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 20548 + }, + { + "epoch": 0.20549, + "grad_norm": 1.566317033194877, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20549 + }, + { + "epoch": 0.2055, + "grad_norm": 1.2156013760342437, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 20550 + }, + { + "epoch": 0.20551, + "grad_norm": 1.5673897332688218, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20551 + }, + { + "epoch": 0.20552, + "grad_norm": 1.333953384074785, + "learning_rate": 0.003, + "loss": 4.052, + "step": 20552 + }, + { + "epoch": 0.20553, + "grad_norm": 1.5399730875589686, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20553 + }, + { + "epoch": 0.20554, + "grad_norm": 1.541981174557002, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 20554 + }, + { + "epoch": 0.20555, + "grad_norm": 1.1825785678189993, + "learning_rate": 0.003, + "loss": 4.075, + "step": 20555 + }, + { + "epoch": 0.20556, + "grad_norm": 1.3866878454557428, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 20556 + }, + { + "epoch": 0.20557, + "grad_norm": 1.2153274615526608, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 20557 + }, + { + "epoch": 0.20558, + "grad_norm": 1.495944994373923, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 20558 + }, + { + "epoch": 0.20559, + "grad_norm": 1.1103661783779841, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 20559 + }, + { + "epoch": 0.2056, + "grad_norm": 1.408891276822968, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 20560 + }, + { + "epoch": 0.20561, + "grad_norm": 1.0712000019832115, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20561 + }, + { + "epoch": 0.20562, + "grad_norm": 1.6283493784005738, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 20562 + }, + { + "epoch": 0.20563, + "grad_norm": 1.2292113338108412, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 20563 + }, + { + "epoch": 0.20564, + "grad_norm": 1.5663562204334438, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20564 + }, + { + "epoch": 0.20565, + "grad_norm": 1.3567179389875492, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 20565 + }, + { + "epoch": 0.20566, + "grad_norm": 1.434969226216492, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 20566 + }, + { + "epoch": 0.20567, + "grad_norm": 1.5394148286055918, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 20567 + }, + { + "epoch": 0.20568, + "grad_norm": 1.289344470377573, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 20568 + }, + { + "epoch": 0.20569, + "grad_norm": 1.3333326599327588, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20569 + }, + { + "epoch": 0.2057, + "grad_norm": 1.3239196660894166, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 20570 + }, + { + "epoch": 0.20571, + "grad_norm": 1.1103832018978599, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 20571 + }, + { + "epoch": 0.20572, + "grad_norm": 1.5524166451986205, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 20572 + }, + { + "epoch": 0.20573, + "grad_norm": 1.0716054149155247, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20573 + }, + { + "epoch": 0.20574, + "grad_norm": 1.4685398356964334, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 20574 + }, + { + "epoch": 0.20575, + "grad_norm": 1.1829256757367614, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 20575 + }, + { + "epoch": 0.20576, + "grad_norm": 1.5334565864330865, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 20576 + }, + { + "epoch": 0.20577, + "grad_norm": 1.5058034805054878, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 20577 + }, + { + "epoch": 0.20578, + "grad_norm": 1.3935883252454009, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 20578 + }, + { + "epoch": 0.20579, + "grad_norm": 1.4609226679617684, + "learning_rate": 0.003, + "loss": 4.047, + "step": 20579 + }, + { + "epoch": 0.2058, + "grad_norm": 1.359084702387566, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 20580 + }, + { + "epoch": 0.20581, + "grad_norm": 0.9425275373109557, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 20581 + }, + { + "epoch": 0.20582, + "grad_norm": 1.3324735626357085, + "learning_rate": 0.003, + "loss": 4.07, + "step": 20582 + }, + { + "epoch": 0.20583, + "grad_norm": 1.371511234401514, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 20583 + }, + { + "epoch": 0.20584, + "grad_norm": 1.3963853154347534, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 20584 + }, + { + "epoch": 0.20585, + "grad_norm": 1.4811417069310304, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 20585 + }, + { + "epoch": 0.20586, + "grad_norm": 1.376322539672415, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20586 + }, + { + "epoch": 0.20587, + "grad_norm": 1.038201050690109, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 20587 + }, + { + "epoch": 0.20588, + "grad_norm": 1.6625458179834598, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 20588 + }, + { + "epoch": 0.20589, + "grad_norm": 1.1475171467965672, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 20589 + }, + { + "epoch": 0.2059, + "grad_norm": 1.6370923525247905, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 20590 + }, + { + "epoch": 0.20591, + "grad_norm": 1.153261681585598, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 20591 + }, + { + "epoch": 0.20592, + "grad_norm": 1.2922002799285761, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 20592 + }, + { + "epoch": 0.20593, + "grad_norm": 1.3274509594898747, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 20593 + }, + { + "epoch": 0.20594, + "grad_norm": 1.155892313781376, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20594 + }, + { + "epoch": 0.20595, + "grad_norm": 1.3954880176684845, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 20595 + }, + { + "epoch": 0.20596, + "grad_norm": 1.2973600396488827, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 20596 + }, + { + "epoch": 0.20597, + "grad_norm": 1.5925782701536118, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 20597 + }, + { + "epoch": 0.20598, + "grad_norm": 1.4086120310400876, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 20598 + }, + { + "epoch": 0.20599, + "grad_norm": 1.4680185440620883, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 20599 + }, + { + "epoch": 0.206, + "grad_norm": 1.1989041267118707, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20600 + }, + { + "epoch": 0.20601, + "grad_norm": 1.3284651252197444, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 20601 + }, + { + "epoch": 0.20602, + "grad_norm": 1.363074375063444, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20602 + }, + { + "epoch": 0.20603, + "grad_norm": 1.4057815359007353, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 20603 + }, + { + "epoch": 0.20604, + "grad_norm": 1.2600489526130207, + "learning_rate": 0.003, + "loss": 4.054, + "step": 20604 + }, + { + "epoch": 0.20605, + "grad_norm": 1.4725082018922724, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 20605 + }, + { + "epoch": 0.20606, + "grad_norm": 1.0881565574525953, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 20606 + }, + { + "epoch": 0.20607, + "grad_norm": 1.5911806474969958, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 20607 + }, + { + "epoch": 0.20608, + "grad_norm": 1.2697992330638663, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 20608 + }, + { + "epoch": 0.20609, + "grad_norm": 1.4750340535495794, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 20609 + }, + { + "epoch": 0.2061, + "grad_norm": 1.4952831819412025, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 20610 + }, + { + "epoch": 0.20611, + "grad_norm": 1.402393234502162, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 20611 + }, + { + "epoch": 0.20612, + "grad_norm": 1.3257391433023868, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 20612 + }, + { + "epoch": 0.20613, + "grad_norm": 1.3717513598075812, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 20613 + }, + { + "epoch": 0.20614, + "grad_norm": 1.4537290474940516, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20614 + }, + { + "epoch": 0.20615, + "grad_norm": 1.5962756732291659, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20615 + }, + { + "epoch": 0.20616, + "grad_norm": 1.5051333484818226, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 20616 + }, + { + "epoch": 0.20617, + "grad_norm": 1.214466112433022, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 20617 + }, + { + "epoch": 0.20618, + "grad_norm": 1.4469079444714108, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20618 + }, + { + "epoch": 0.20619, + "grad_norm": 1.1397343693052815, + "learning_rate": 0.003, + "loss": 4.091, + "step": 20619 + }, + { + "epoch": 0.2062, + "grad_norm": 1.3456803832918454, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 20620 + }, + { + "epoch": 0.20621, + "grad_norm": 1.1544122825716558, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20621 + }, + { + "epoch": 0.20622, + "grad_norm": 1.4725408234925452, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20622 + }, + { + "epoch": 0.20623, + "grad_norm": 1.1801597668502901, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 20623 + }, + { + "epoch": 0.20624, + "grad_norm": 1.6559749019356147, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 20624 + }, + { + "epoch": 0.20625, + "grad_norm": 1.2627663038337578, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 20625 + }, + { + "epoch": 0.20626, + "grad_norm": 1.5817579773010755, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 20626 + }, + { + "epoch": 0.20627, + "grad_norm": 1.1749488660893088, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 20627 + }, + { + "epoch": 0.20628, + "grad_norm": 1.2607651323166291, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 20628 + }, + { + "epoch": 0.20629, + "grad_norm": 1.1521351642743436, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 20629 + }, + { + "epoch": 0.2063, + "grad_norm": 1.341773679744627, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20630 + }, + { + "epoch": 0.20631, + "grad_norm": 1.2009941594378983, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 20631 + }, + { + "epoch": 0.20632, + "grad_norm": 1.153179661594956, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 20632 + }, + { + "epoch": 0.20633, + "grad_norm": 1.3616113848885394, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 20633 + }, + { + "epoch": 0.20634, + "grad_norm": 1.2296475286632265, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 20634 + }, + { + "epoch": 0.20635, + "grad_norm": 1.2699110819449624, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 20635 + }, + { + "epoch": 0.20636, + "grad_norm": 1.6414333303584467, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 20636 + }, + { + "epoch": 0.20637, + "grad_norm": 1.0484630733656326, + "learning_rate": 0.003, + "loss": 4.034, + "step": 20637 + }, + { + "epoch": 0.20638, + "grad_norm": 1.5160647539076348, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20638 + }, + { + "epoch": 0.20639, + "grad_norm": 1.2813846351739706, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 20639 + }, + { + "epoch": 0.2064, + "grad_norm": 1.505955669199115, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 20640 + }, + { + "epoch": 0.20641, + "grad_norm": 1.3139142425717492, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20641 + }, + { + "epoch": 0.20642, + "grad_norm": 1.3777522538089098, + "learning_rate": 0.003, + "loss": 4.062, + "step": 20642 + }, + { + "epoch": 0.20643, + "grad_norm": 1.6087420201941252, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 20643 + }, + { + "epoch": 0.20644, + "grad_norm": 1.1610478132610833, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 20644 + }, + { + "epoch": 0.20645, + "grad_norm": 1.3894341858483343, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 20645 + }, + { + "epoch": 0.20646, + "grad_norm": 1.5936425142861472, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20646 + }, + { + "epoch": 0.20647, + "grad_norm": 1.3309219203024698, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 20647 + }, + { + "epoch": 0.20648, + "grad_norm": 1.531445145082477, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 20648 + }, + { + "epoch": 0.20649, + "grad_norm": 1.3212088065800662, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20649 + }, + { + "epoch": 0.2065, + "grad_norm": 1.5004849199510446, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 20650 + }, + { + "epoch": 0.20651, + "grad_norm": 1.3039083086214973, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20651 + }, + { + "epoch": 0.20652, + "grad_norm": 1.2837868305916982, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 20652 + }, + { + "epoch": 0.20653, + "grad_norm": 1.268203737086418, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 20653 + }, + { + "epoch": 0.20654, + "grad_norm": 1.3209765544553898, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 20654 + }, + { + "epoch": 0.20655, + "grad_norm": 1.3129637976579085, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20655 + }, + { + "epoch": 0.20656, + "grad_norm": 1.4278382500501035, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 20656 + }, + { + "epoch": 0.20657, + "grad_norm": 1.4090868893650896, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 20657 + }, + { + "epoch": 0.20658, + "grad_norm": 1.1532438735239938, + "learning_rate": 0.003, + "loss": 4.053, + "step": 20658 + }, + { + "epoch": 0.20659, + "grad_norm": 1.5435740786225585, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 20659 + }, + { + "epoch": 0.2066, + "grad_norm": 1.1721951629686629, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 20660 + }, + { + "epoch": 0.20661, + "grad_norm": 1.432834732135936, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 20661 + }, + { + "epoch": 0.20662, + "grad_norm": 1.3326968054391681, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 20662 + }, + { + "epoch": 0.20663, + "grad_norm": 1.2551462438480991, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 20663 + }, + { + "epoch": 0.20664, + "grad_norm": 1.6477065949606842, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 20664 + }, + { + "epoch": 0.20665, + "grad_norm": 1.2073682442471083, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 20665 + }, + { + "epoch": 0.20666, + "grad_norm": 1.5967910925825928, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20666 + }, + { + "epoch": 0.20667, + "grad_norm": 1.1702067473700242, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 20667 + }, + { + "epoch": 0.20668, + "grad_norm": 1.4490009490171216, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 20668 + }, + { + "epoch": 0.20669, + "grad_norm": 1.4982000874554064, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 20669 + }, + { + "epoch": 0.2067, + "grad_norm": 1.30080984292686, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20670 + }, + { + "epoch": 0.20671, + "grad_norm": 1.3677103611908143, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 20671 + }, + { + "epoch": 0.20672, + "grad_norm": 1.2772405357651315, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 20672 + }, + { + "epoch": 0.20673, + "grad_norm": 1.3588295599384557, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 20673 + }, + { + "epoch": 0.20674, + "grad_norm": 1.3946185325560583, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 20674 + }, + { + "epoch": 0.20675, + "grad_norm": 1.4712534004184314, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 20675 + }, + { + "epoch": 0.20676, + "grad_norm": 1.3305414169625183, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20676 + }, + { + "epoch": 0.20677, + "grad_norm": 1.5275630405565999, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 20677 + }, + { + "epoch": 0.20678, + "grad_norm": 1.1417928156030526, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 20678 + }, + { + "epoch": 0.20679, + "grad_norm": 1.3328961831124926, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 20679 + }, + { + "epoch": 0.2068, + "grad_norm": 1.3266191904489568, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 20680 + }, + { + "epoch": 0.20681, + "grad_norm": 1.1604107214621595, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 20681 + }, + { + "epoch": 0.20682, + "grad_norm": 1.6785511499365684, + "learning_rate": 0.003, + "loss": 4.023, + "step": 20682 + }, + { + "epoch": 0.20683, + "grad_norm": 1.0861493863370346, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20683 + }, + { + "epoch": 0.20684, + "grad_norm": 1.5560365428407226, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 20684 + }, + { + "epoch": 0.20685, + "grad_norm": 1.2498265822341013, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 20685 + }, + { + "epoch": 0.20686, + "grad_norm": 1.264505530010203, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 20686 + }, + { + "epoch": 0.20687, + "grad_norm": 1.2262354105769637, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 20687 + }, + { + "epoch": 0.20688, + "grad_norm": 1.3547014565798081, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 20688 + }, + { + "epoch": 0.20689, + "grad_norm": 1.3205301215705505, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 20689 + }, + { + "epoch": 0.2069, + "grad_norm": 1.4385638550422526, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20690 + }, + { + "epoch": 0.20691, + "grad_norm": 1.254260184322425, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20691 + }, + { + "epoch": 0.20692, + "grad_norm": 1.492638364859185, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 20692 + }, + { + "epoch": 0.20693, + "grad_norm": 1.6902384151982752, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20693 + }, + { + "epoch": 0.20694, + "grad_norm": 1.5486125304486253, + "learning_rate": 0.003, + "loss": 4.053, + "step": 20694 + }, + { + "epoch": 0.20695, + "grad_norm": 1.3571206194483856, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 20695 + }, + { + "epoch": 0.20696, + "grad_norm": 1.195684554846468, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 20696 + }, + { + "epoch": 0.20697, + "grad_norm": 1.2673737123930042, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 20697 + }, + { + "epoch": 0.20698, + "grad_norm": 1.4501427918214878, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 20698 + }, + { + "epoch": 0.20699, + "grad_norm": 1.4616677033308434, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 20699 + }, + { + "epoch": 0.207, + "grad_norm": 1.259657976909807, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 20700 + }, + { + "epoch": 0.20701, + "grad_norm": 1.467939418322868, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 20701 + }, + { + "epoch": 0.20702, + "grad_norm": 1.2539470183297858, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 20702 + }, + { + "epoch": 0.20703, + "grad_norm": 1.244938656450836, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 20703 + }, + { + "epoch": 0.20704, + "grad_norm": 1.4363533964769344, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 20704 + }, + { + "epoch": 0.20705, + "grad_norm": 1.1799204393485285, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 20705 + }, + { + "epoch": 0.20706, + "grad_norm": 1.8304053317183064, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 20706 + }, + { + "epoch": 0.20707, + "grad_norm": 0.9245806846188429, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 20707 + }, + { + "epoch": 0.20708, + "grad_norm": 1.3670050823744555, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 20708 + }, + { + "epoch": 0.20709, + "grad_norm": 1.148663547313698, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20709 + }, + { + "epoch": 0.2071, + "grad_norm": 1.4332837261687998, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 20710 + }, + { + "epoch": 0.20711, + "grad_norm": 1.2474709065086116, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 20711 + }, + { + "epoch": 0.20712, + "grad_norm": 1.518321959945295, + "learning_rate": 0.003, + "loss": 4.084, + "step": 20712 + }, + { + "epoch": 0.20713, + "grad_norm": 1.2589783945695787, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 20713 + }, + { + "epoch": 0.20714, + "grad_norm": 1.4680065605046984, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 20714 + }, + { + "epoch": 0.20715, + "grad_norm": 1.3689933685128204, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 20715 + }, + { + "epoch": 0.20716, + "grad_norm": 1.1690381414163973, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 20716 + }, + { + "epoch": 0.20717, + "grad_norm": 1.430373190789418, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20717 + }, + { + "epoch": 0.20718, + "grad_norm": 1.2445200041778108, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 20718 + }, + { + "epoch": 0.20719, + "grad_norm": 1.2360357908003283, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 20719 + }, + { + "epoch": 0.2072, + "grad_norm": 1.1518801939778631, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 20720 + }, + { + "epoch": 0.20721, + "grad_norm": 1.3540922092959804, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 20721 + }, + { + "epoch": 0.20722, + "grad_norm": 1.403886039911295, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 20722 + }, + { + "epoch": 0.20723, + "grad_norm": 1.408496319277612, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 20723 + }, + { + "epoch": 0.20724, + "grad_norm": 1.53624630820367, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 20724 + }, + { + "epoch": 0.20725, + "grad_norm": 1.1028599247352295, + "learning_rate": 0.003, + "loss": 4.083, + "step": 20725 + }, + { + "epoch": 0.20726, + "grad_norm": 1.4912493458663871, + "learning_rate": 0.003, + "loss": 4.049, + "step": 20726 + }, + { + "epoch": 0.20727, + "grad_norm": 1.400145332522727, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 20727 + }, + { + "epoch": 0.20728, + "grad_norm": 1.2550730824349292, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 20728 + }, + { + "epoch": 0.20729, + "grad_norm": 1.5020230627391862, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 20729 + }, + { + "epoch": 0.2073, + "grad_norm": 1.094855880312423, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20730 + }, + { + "epoch": 0.20731, + "grad_norm": 1.4454634777088968, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 20731 + }, + { + "epoch": 0.20732, + "grad_norm": 1.1960298178416016, + "learning_rate": 0.003, + "loss": 4.011, + "step": 20732 + }, + { + "epoch": 0.20733, + "grad_norm": 1.573553533514854, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 20733 + }, + { + "epoch": 0.20734, + "grad_norm": 1.3005374564992995, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 20734 + }, + { + "epoch": 0.20735, + "grad_norm": 1.4733758613998844, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 20735 + }, + { + "epoch": 0.20736, + "grad_norm": 1.192320375520012, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 20736 + }, + { + "epoch": 0.20737, + "grad_norm": 1.441776597098866, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20737 + }, + { + "epoch": 0.20738, + "grad_norm": 1.2019891485357035, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 20738 + }, + { + "epoch": 0.20739, + "grad_norm": 1.7143393481625193, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 20739 + }, + { + "epoch": 0.2074, + "grad_norm": 1.3011650865174773, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 20740 + }, + { + "epoch": 0.20741, + "grad_norm": 1.420040265271807, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 20741 + }, + { + "epoch": 0.20742, + "grad_norm": 1.271071548410552, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 20742 + }, + { + "epoch": 0.20743, + "grad_norm": 1.2725194915570306, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20743 + }, + { + "epoch": 0.20744, + "grad_norm": 1.407554084131202, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 20744 + }, + { + "epoch": 0.20745, + "grad_norm": 1.2723492493093125, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 20745 + }, + { + "epoch": 0.20746, + "grad_norm": 1.2789554944969392, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20746 + }, + { + "epoch": 0.20747, + "grad_norm": 1.2190173749910922, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 20747 + }, + { + "epoch": 0.20748, + "grad_norm": 1.3671246285439873, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 20748 + }, + { + "epoch": 0.20749, + "grad_norm": 1.2860610697540353, + "learning_rate": 0.003, + "loss": 4.047, + "step": 20749 + }, + { + "epoch": 0.2075, + "grad_norm": 1.5485327618678362, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 20750 + }, + { + "epoch": 0.20751, + "grad_norm": 1.0354250543920769, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 20751 + }, + { + "epoch": 0.20752, + "grad_norm": 1.5729608571657345, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 20752 + }, + { + "epoch": 0.20753, + "grad_norm": 1.293985990906778, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20753 + }, + { + "epoch": 0.20754, + "grad_norm": 1.383144680757959, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20754 + }, + { + "epoch": 0.20755, + "grad_norm": 1.5258289814465253, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 20755 + }, + { + "epoch": 0.20756, + "grad_norm": 1.2284186594380586, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20756 + }, + { + "epoch": 0.20757, + "grad_norm": 1.295884549935487, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 20757 + }, + { + "epoch": 0.20758, + "grad_norm": 1.2251701011240375, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 20758 + }, + { + "epoch": 0.20759, + "grad_norm": 1.4138209427266042, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20759 + }, + { + "epoch": 0.2076, + "grad_norm": 1.2928386514239687, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 20760 + }, + { + "epoch": 0.20761, + "grad_norm": 1.6584917470311489, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 20761 + }, + { + "epoch": 0.20762, + "grad_norm": 1.1000944720003125, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20762 + }, + { + "epoch": 0.20763, + "grad_norm": 1.5065086778699819, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 20763 + }, + { + "epoch": 0.20764, + "grad_norm": 1.2405226321846727, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 20764 + }, + { + "epoch": 0.20765, + "grad_norm": 1.5062568526642643, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 20765 + }, + { + "epoch": 0.20766, + "grad_norm": 1.2543886727598397, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 20766 + }, + { + "epoch": 0.20767, + "grad_norm": 1.4686589157048027, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 20767 + }, + { + "epoch": 0.20768, + "grad_norm": 1.0878892800689992, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 20768 + }, + { + "epoch": 0.20769, + "grad_norm": 1.2890076109933548, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 20769 + }, + { + "epoch": 0.2077, + "grad_norm": 1.2698804207351753, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 20770 + }, + { + "epoch": 0.20771, + "grad_norm": 1.2963430756336742, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20771 + }, + { + "epoch": 0.20772, + "grad_norm": 1.5503858672891013, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 20772 + }, + { + "epoch": 0.20773, + "grad_norm": 1.447824162200313, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 20773 + }, + { + "epoch": 0.20774, + "grad_norm": 1.0698241675703897, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 20774 + }, + { + "epoch": 0.20775, + "grad_norm": 1.5921843730582812, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20775 + }, + { + "epoch": 0.20776, + "grad_norm": 1.5676573158126632, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20776 + }, + { + "epoch": 0.20777, + "grad_norm": 1.2337925630873627, + "learning_rate": 0.003, + "loss": 4.085, + "step": 20777 + }, + { + "epoch": 0.20778, + "grad_norm": 1.558918139946969, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 20778 + }, + { + "epoch": 0.20779, + "grad_norm": 1.2576777157283208, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20779 + }, + { + "epoch": 0.2078, + "grad_norm": 1.4891119004045967, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 20780 + }, + { + "epoch": 0.20781, + "grad_norm": 1.4864117549677052, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 20781 + }, + { + "epoch": 0.20782, + "grad_norm": 1.3824042412785735, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 20782 + }, + { + "epoch": 0.20783, + "grad_norm": 1.2647212252167144, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20783 + }, + { + "epoch": 0.20784, + "grad_norm": 1.3682686819141703, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 20784 + }, + { + "epoch": 0.20785, + "grad_norm": 1.3134092379540208, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 20785 + }, + { + "epoch": 0.20786, + "grad_norm": 1.2000088386311523, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20786 + }, + { + "epoch": 0.20787, + "grad_norm": 1.358948748403982, + "learning_rate": 0.003, + "loss": 4.048, + "step": 20787 + }, + { + "epoch": 0.20788, + "grad_norm": 1.2266057165726971, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 20788 + }, + { + "epoch": 0.20789, + "grad_norm": 1.4599833485159968, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 20789 + }, + { + "epoch": 0.2079, + "grad_norm": 1.4274327253015713, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 20790 + }, + { + "epoch": 0.20791, + "grad_norm": 1.4618313253646975, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 20791 + }, + { + "epoch": 0.20792, + "grad_norm": 1.3711908786137523, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 20792 + }, + { + "epoch": 0.20793, + "grad_norm": 1.498250720442979, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 20793 + }, + { + "epoch": 0.20794, + "grad_norm": 1.2535367799442354, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 20794 + }, + { + "epoch": 0.20795, + "grad_norm": 1.3419060508276737, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 20795 + }, + { + "epoch": 0.20796, + "grad_norm": 1.4115902482667926, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20796 + }, + { + "epoch": 0.20797, + "grad_norm": 1.3850709340925633, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20797 + }, + { + "epoch": 0.20798, + "grad_norm": 1.3207471580223884, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 20798 + }, + { + "epoch": 0.20799, + "grad_norm": 1.3568939822112909, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 20799 + }, + { + "epoch": 0.208, + "grad_norm": 1.2019262130445871, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 20800 + }, + { + "epoch": 0.20801, + "grad_norm": 1.5522301951501158, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 20801 + }, + { + "epoch": 0.20802, + "grad_norm": 1.2738694402674298, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 20802 + }, + { + "epoch": 0.20803, + "grad_norm": 1.6066304757212384, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 20803 + }, + { + "epoch": 0.20804, + "grad_norm": 1.1511055437116582, + "learning_rate": 0.003, + "loss": 4.083, + "step": 20804 + }, + { + "epoch": 0.20805, + "grad_norm": 1.3737968339786775, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 20805 + }, + { + "epoch": 0.20806, + "grad_norm": 1.1696700332140544, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 20806 + }, + { + "epoch": 0.20807, + "grad_norm": 1.438958584346387, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 20807 + }, + { + "epoch": 0.20808, + "grad_norm": 1.3446405494157427, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20808 + }, + { + "epoch": 0.20809, + "grad_norm": 1.509586891470056, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 20809 + }, + { + "epoch": 0.2081, + "grad_norm": 1.2164155739319678, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 20810 + }, + { + "epoch": 0.20811, + "grad_norm": 1.4657159505696644, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 20811 + }, + { + "epoch": 0.20812, + "grad_norm": 1.2897275751829282, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 20812 + }, + { + "epoch": 0.20813, + "grad_norm": 1.2342837330568117, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 20813 + }, + { + "epoch": 0.20814, + "grad_norm": 1.2818396025311736, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 20814 + }, + { + "epoch": 0.20815, + "grad_norm": 1.138639834661771, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 20815 + }, + { + "epoch": 0.20816, + "grad_norm": 1.5138052170278105, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 20816 + }, + { + "epoch": 0.20817, + "grad_norm": 1.1634977361316547, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 20817 + }, + { + "epoch": 0.20818, + "grad_norm": 1.5215420540090645, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 20818 + }, + { + "epoch": 0.20819, + "grad_norm": 1.5668058020930633, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 20819 + }, + { + "epoch": 0.2082, + "grad_norm": 1.4531488544684097, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 20820 + }, + { + "epoch": 0.20821, + "grad_norm": 1.2865371605977516, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 20821 + }, + { + "epoch": 0.20822, + "grad_norm": 1.309351243002296, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 20822 + }, + { + "epoch": 0.20823, + "grad_norm": 1.4421691511982775, + "learning_rate": 0.003, + "loss": 4.055, + "step": 20823 + }, + { + "epoch": 0.20824, + "grad_norm": 1.2366488864526652, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 20824 + }, + { + "epoch": 0.20825, + "grad_norm": 1.5101299369902328, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 20825 + }, + { + "epoch": 0.20826, + "grad_norm": 1.0820147383838739, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20826 + }, + { + "epoch": 0.20827, + "grad_norm": 1.5934109137934334, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 20827 + }, + { + "epoch": 0.20828, + "grad_norm": 1.109182361190493, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20828 + }, + { + "epoch": 0.20829, + "grad_norm": 1.6282571708569145, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 20829 + }, + { + "epoch": 0.2083, + "grad_norm": 1.3254657626410726, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20830 + }, + { + "epoch": 0.20831, + "grad_norm": 1.4069406914196951, + "learning_rate": 0.003, + "loss": 4.069, + "step": 20831 + }, + { + "epoch": 0.20832, + "grad_norm": 1.1955776117205292, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 20832 + }, + { + "epoch": 0.20833, + "grad_norm": 1.6091706976751523, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 20833 + }, + { + "epoch": 0.20834, + "grad_norm": 1.3554567338878298, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 20834 + }, + { + "epoch": 0.20835, + "grad_norm": 1.2173982775230165, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 20835 + }, + { + "epoch": 0.20836, + "grad_norm": 1.5726560445228621, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 20836 + }, + { + "epoch": 0.20837, + "grad_norm": 1.3197429362875017, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 20837 + }, + { + "epoch": 0.20838, + "grad_norm": 1.3932221313372262, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 20838 + }, + { + "epoch": 0.20839, + "grad_norm": 1.3436690757123135, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 20839 + }, + { + "epoch": 0.2084, + "grad_norm": 1.2914233320150121, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 20840 + }, + { + "epoch": 0.20841, + "grad_norm": 1.4039904864137214, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20841 + }, + { + "epoch": 0.20842, + "grad_norm": 1.1224930480012258, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 20842 + }, + { + "epoch": 0.20843, + "grad_norm": 1.5006419635128865, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 20843 + }, + { + "epoch": 0.20844, + "grad_norm": 1.316535683710008, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 20844 + }, + { + "epoch": 0.20845, + "grad_norm": 1.6994979771996706, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 20845 + }, + { + "epoch": 0.20846, + "grad_norm": 1.4424776773399692, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 20846 + }, + { + "epoch": 0.20847, + "grad_norm": 1.292769059462763, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 20847 + }, + { + "epoch": 0.20848, + "grad_norm": 1.3931095132806635, + "learning_rate": 0.003, + "loss": 4.042, + "step": 20848 + }, + { + "epoch": 0.20849, + "grad_norm": 1.4169322101511408, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 20849 + }, + { + "epoch": 0.2085, + "grad_norm": 1.4162308539627977, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20850 + }, + { + "epoch": 0.20851, + "grad_norm": 1.3169184739063786, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 20851 + }, + { + "epoch": 0.20852, + "grad_norm": 1.2926076560780115, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 20852 + }, + { + "epoch": 0.20853, + "grad_norm": 1.4132541866511903, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 20853 + }, + { + "epoch": 0.20854, + "grad_norm": 1.2636015858679868, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 20854 + }, + { + "epoch": 0.20855, + "grad_norm": 1.1150503537596996, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20855 + }, + { + "epoch": 0.20856, + "grad_norm": 1.4027910960977807, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 20856 + }, + { + "epoch": 0.20857, + "grad_norm": 1.4297746470245551, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 20857 + }, + { + "epoch": 0.20858, + "grad_norm": 1.396782413637501, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 20858 + }, + { + "epoch": 0.20859, + "grad_norm": 1.164888722395929, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 20859 + }, + { + "epoch": 0.2086, + "grad_norm": 1.4318988786579787, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 20860 + }, + { + "epoch": 0.20861, + "grad_norm": 1.3869170185913235, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 20861 + }, + { + "epoch": 0.20862, + "grad_norm": 1.459312875995161, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 20862 + }, + { + "epoch": 0.20863, + "grad_norm": 1.353994075043403, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 20863 + }, + { + "epoch": 0.20864, + "grad_norm": 1.1460913522678278, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 20864 + }, + { + "epoch": 0.20865, + "grad_norm": 1.1715799872377994, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20865 + }, + { + "epoch": 0.20866, + "grad_norm": 1.5112689480891477, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 20866 + }, + { + "epoch": 0.20867, + "grad_norm": 1.2277638411317997, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 20867 + }, + { + "epoch": 0.20868, + "grad_norm": 1.3559174681191846, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 20868 + }, + { + "epoch": 0.20869, + "grad_norm": 1.1579036036919177, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20869 + }, + { + "epoch": 0.2087, + "grad_norm": 1.5168868574343963, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20870 + }, + { + "epoch": 0.20871, + "grad_norm": 1.192882139239422, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 20871 + }, + { + "epoch": 0.20872, + "grad_norm": 1.6397552745256645, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 20872 + }, + { + "epoch": 0.20873, + "grad_norm": 1.2998046635660576, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20873 + }, + { + "epoch": 0.20874, + "grad_norm": 1.170136001703282, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 20874 + }, + { + "epoch": 0.20875, + "grad_norm": 1.7712938964576301, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 20875 + }, + { + "epoch": 0.20876, + "grad_norm": 1.304514044762122, + "learning_rate": 0.003, + "loss": 4.079, + "step": 20876 + }, + { + "epoch": 0.20877, + "grad_norm": 1.5311149814383385, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 20877 + }, + { + "epoch": 0.20878, + "grad_norm": 1.377141579972561, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 20878 + }, + { + "epoch": 0.20879, + "grad_norm": 1.2885018804197217, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 20879 + }, + { + "epoch": 0.2088, + "grad_norm": 1.2805009293500598, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 20880 + }, + { + "epoch": 0.20881, + "grad_norm": 1.5271926042823611, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20881 + }, + { + "epoch": 0.20882, + "grad_norm": 1.313556059499855, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 20882 + }, + { + "epoch": 0.20883, + "grad_norm": 1.1464190081317402, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 20883 + }, + { + "epoch": 0.20884, + "grad_norm": 1.549942223937229, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 20884 + }, + { + "epoch": 0.20885, + "grad_norm": 1.1265602354973108, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 20885 + }, + { + "epoch": 0.20886, + "grad_norm": 1.707473848813536, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 20886 + }, + { + "epoch": 0.20887, + "grad_norm": 0.9961645801473272, + "learning_rate": 0.003, + "loss": 4.099, + "step": 20887 + }, + { + "epoch": 0.20888, + "grad_norm": 1.538759783989019, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 20888 + }, + { + "epoch": 0.20889, + "grad_norm": 1.3896655265307651, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 20889 + }, + { + "epoch": 0.2089, + "grad_norm": 1.1829799147541915, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 20890 + }, + { + "epoch": 0.20891, + "grad_norm": 1.4199392258570829, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 20891 + }, + { + "epoch": 0.20892, + "grad_norm": 1.3251457352384242, + "learning_rate": 0.003, + "loss": 4.074, + "step": 20892 + }, + { + "epoch": 0.20893, + "grad_norm": 1.3720741261766807, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 20893 + }, + { + "epoch": 0.20894, + "grad_norm": 1.4397315716619452, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 20894 + }, + { + "epoch": 0.20895, + "grad_norm": 1.2600167535017244, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 20895 + }, + { + "epoch": 0.20896, + "grad_norm": 1.4193530587315695, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20896 + }, + { + "epoch": 0.20897, + "grad_norm": 1.4613543310948776, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 20897 + }, + { + "epoch": 0.20898, + "grad_norm": 1.4334011717950574, + "learning_rate": 0.003, + "loss": 4.057, + "step": 20898 + }, + { + "epoch": 0.20899, + "grad_norm": 1.2649157110995044, + "learning_rate": 0.003, + "loss": 4.065, + "step": 20899 + }, + { + "epoch": 0.209, + "grad_norm": 1.4679698539406412, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 20900 + }, + { + "epoch": 0.20901, + "grad_norm": 1.1456737698413446, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 20901 + }, + { + "epoch": 0.20902, + "grad_norm": 1.260909399618353, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 20902 + }, + { + "epoch": 0.20903, + "grad_norm": 1.3774463340297802, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 20903 + }, + { + "epoch": 0.20904, + "grad_norm": 1.3792000896777399, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 20904 + }, + { + "epoch": 0.20905, + "grad_norm": 1.3145741389508931, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 20905 + }, + { + "epoch": 0.20906, + "grad_norm": 1.488555454898239, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 20906 + }, + { + "epoch": 0.20907, + "grad_norm": 0.9177311400044877, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 20907 + }, + { + "epoch": 0.20908, + "grad_norm": 1.4736044351887085, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 20908 + }, + { + "epoch": 0.20909, + "grad_norm": 1.236906959060174, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 20909 + }, + { + "epoch": 0.2091, + "grad_norm": 1.2164246192183843, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 20910 + }, + { + "epoch": 0.20911, + "grad_norm": 1.4078924617347541, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 20911 + }, + { + "epoch": 0.20912, + "grad_norm": 1.1626258082614278, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 20912 + }, + { + "epoch": 0.20913, + "grad_norm": 1.5000031481174514, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 20913 + }, + { + "epoch": 0.20914, + "grad_norm": 1.711991182599059, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 20914 + }, + { + "epoch": 0.20915, + "grad_norm": 1.1173024910508889, + "learning_rate": 0.003, + "loss": 4.066, + "step": 20915 + }, + { + "epoch": 0.20916, + "grad_norm": 1.6195944409721208, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 20916 + }, + { + "epoch": 0.20917, + "grad_norm": 1.1495495916677396, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 20917 + }, + { + "epoch": 0.20918, + "grad_norm": 1.3315584679143382, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 20918 + }, + { + "epoch": 0.20919, + "grad_norm": 1.3129474249266087, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 20919 + }, + { + "epoch": 0.2092, + "grad_norm": 1.45735198601752, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 20920 + }, + { + "epoch": 0.20921, + "grad_norm": 1.2008207342408617, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 20921 + }, + { + "epoch": 0.20922, + "grad_norm": 1.490154537659906, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 20922 + }, + { + "epoch": 0.20923, + "grad_norm": 1.3246868033101271, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 20923 + }, + { + "epoch": 0.20924, + "grad_norm": 1.4013961451103512, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 20924 + }, + { + "epoch": 0.20925, + "grad_norm": 1.3739218556715787, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 20925 + }, + { + "epoch": 0.20926, + "grad_norm": 1.272221983849928, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 20926 + }, + { + "epoch": 0.20927, + "grad_norm": 1.5647031482397493, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 20927 + }, + { + "epoch": 0.20928, + "grad_norm": 1.2508057694645471, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 20928 + }, + { + "epoch": 0.20929, + "grad_norm": 1.3826914101940824, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 20929 + }, + { + "epoch": 0.2093, + "grad_norm": 1.548268893130226, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 20930 + }, + { + "epoch": 0.20931, + "grad_norm": 1.2446087214126185, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 20931 + }, + { + "epoch": 0.20932, + "grad_norm": 1.6262942416468529, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 20932 + }, + { + "epoch": 0.20933, + "grad_norm": 1.1852497232306514, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20933 + }, + { + "epoch": 0.20934, + "grad_norm": 1.6696998384812047, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 20934 + }, + { + "epoch": 0.20935, + "grad_norm": 1.2181296766904035, + "learning_rate": 0.003, + "loss": 4.078, + "step": 20935 + }, + { + "epoch": 0.20936, + "grad_norm": 1.4900857645434027, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 20936 + }, + { + "epoch": 0.20937, + "grad_norm": 1.3021609810644803, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 20937 + }, + { + "epoch": 0.20938, + "grad_norm": 1.295584171371703, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 20938 + }, + { + "epoch": 0.20939, + "grad_norm": 1.7010666167860795, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 20939 + }, + { + "epoch": 0.2094, + "grad_norm": 0.9469238419501262, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 20940 + }, + { + "epoch": 0.20941, + "grad_norm": 1.3814181330827642, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 20941 + }, + { + "epoch": 0.20942, + "grad_norm": 1.4840720303537958, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 20942 + }, + { + "epoch": 0.20943, + "grad_norm": 1.3128491819583785, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20943 + }, + { + "epoch": 0.20944, + "grad_norm": 1.4681052115962732, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 20944 + }, + { + "epoch": 0.20945, + "grad_norm": 1.0928041953544825, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 20945 + }, + { + "epoch": 0.20946, + "grad_norm": 1.2776202521965, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 20946 + }, + { + "epoch": 0.20947, + "grad_norm": 1.2072092071595812, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 20947 + }, + { + "epoch": 0.20948, + "grad_norm": 1.4983914960524811, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 20948 + }, + { + "epoch": 0.20949, + "grad_norm": 1.3556837860308897, + "learning_rate": 0.003, + "loss": 4.049, + "step": 20949 + }, + { + "epoch": 0.2095, + "grad_norm": 1.1747432769790092, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 20950 + }, + { + "epoch": 0.20951, + "grad_norm": 1.4823779753416622, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 20951 + }, + { + "epoch": 0.20952, + "grad_norm": 0.9942073008067971, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 20952 + }, + { + "epoch": 0.20953, + "grad_norm": 1.6230027795029984, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 20953 + }, + { + "epoch": 0.20954, + "grad_norm": 1.1421677685740035, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20954 + }, + { + "epoch": 0.20955, + "grad_norm": 1.6820318386159103, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 20955 + }, + { + "epoch": 0.20956, + "grad_norm": 1.3487592414130567, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 20956 + }, + { + "epoch": 0.20957, + "grad_norm": 1.3066160841943804, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 20957 + }, + { + "epoch": 0.20958, + "grad_norm": 1.1619227542351784, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 20958 + }, + { + "epoch": 0.20959, + "grad_norm": 1.4546064403658636, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 20959 + }, + { + "epoch": 0.2096, + "grad_norm": 1.1838430608934516, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 20960 + }, + { + "epoch": 0.20961, + "grad_norm": 1.55115485747746, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20961 + }, + { + "epoch": 0.20962, + "grad_norm": 1.3047990384928365, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 20962 + }, + { + "epoch": 0.20963, + "grad_norm": 1.3312728907537128, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 20963 + }, + { + "epoch": 0.20964, + "grad_norm": 1.5045742901254493, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20964 + }, + { + "epoch": 0.20965, + "grad_norm": 1.216675557605364, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 20965 + }, + { + "epoch": 0.20966, + "grad_norm": 1.3876853879303457, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 20966 + }, + { + "epoch": 0.20967, + "grad_norm": 1.3182662675036918, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 20967 + }, + { + "epoch": 0.20968, + "grad_norm": 1.3461464413037225, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 20968 + }, + { + "epoch": 0.20969, + "grad_norm": 1.310372816887392, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 20969 + }, + { + "epoch": 0.2097, + "grad_norm": 1.4287852852752707, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 20970 + }, + { + "epoch": 0.20971, + "grad_norm": 1.2742591333006903, + "learning_rate": 0.003, + "loss": 4.056, + "step": 20971 + }, + { + "epoch": 0.20972, + "grad_norm": 1.410034093862182, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 20972 + }, + { + "epoch": 0.20973, + "grad_norm": 1.4844923925773001, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 20973 + }, + { + "epoch": 0.20974, + "grad_norm": 1.5172877935597546, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 20974 + }, + { + "epoch": 0.20975, + "grad_norm": 1.3044946275124711, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 20975 + }, + { + "epoch": 0.20976, + "grad_norm": 1.151846917353467, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20976 + }, + { + "epoch": 0.20977, + "grad_norm": 1.5449323600079314, + "learning_rate": 0.003, + "loss": 4.09, + "step": 20977 + }, + { + "epoch": 0.20978, + "grad_norm": 1.1036848639930728, + "learning_rate": 0.003, + "loss": 4.03, + "step": 20978 + }, + { + "epoch": 0.20979, + "grad_norm": 1.4178704958997441, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 20979 + }, + { + "epoch": 0.2098, + "grad_norm": 1.491666526743511, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 20980 + }, + { + "epoch": 0.20981, + "grad_norm": 1.259111988716843, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 20981 + }, + { + "epoch": 0.20982, + "grad_norm": 1.392424981454325, + "learning_rate": 0.003, + "loss": 4.063, + "step": 20982 + }, + { + "epoch": 0.20983, + "grad_norm": 1.3734294209193698, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 20983 + }, + { + "epoch": 0.20984, + "grad_norm": 1.2098709163289976, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 20984 + }, + { + "epoch": 0.20985, + "grad_norm": 1.5174873420676978, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 20985 + }, + { + "epoch": 0.20986, + "grad_norm": 1.109118359260239, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 20986 + }, + { + "epoch": 0.20987, + "grad_norm": 1.6151173655333941, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 20987 + }, + { + "epoch": 0.20988, + "grad_norm": 1.317078526794028, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20988 + }, + { + "epoch": 0.20989, + "grad_norm": 1.3960094316751979, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20989 + }, + { + "epoch": 0.2099, + "grad_norm": 1.502111690521901, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 20990 + }, + { + "epoch": 0.20991, + "grad_norm": 1.2405024142395, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 20991 + }, + { + "epoch": 0.20992, + "grad_norm": 1.5708082629079743, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 20992 + }, + { + "epoch": 0.20993, + "grad_norm": 1.2786993413574794, + "learning_rate": 0.003, + "loss": 4.115, + "step": 20993 + }, + { + "epoch": 0.20994, + "grad_norm": 1.3277837390722653, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 20994 + }, + { + "epoch": 0.20995, + "grad_norm": 1.43136863374496, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20995 + }, + { + "epoch": 0.20996, + "grad_norm": 1.1309822698757983, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 20996 + }, + { + "epoch": 0.20997, + "grad_norm": 1.567826172504886, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 20997 + }, + { + "epoch": 0.20998, + "grad_norm": 1.0645283417366327, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 20998 + }, + { + "epoch": 0.20999, + "grad_norm": 1.5062893700951852, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 20999 + }, + { + "epoch": 0.21, + "grad_norm": 1.2211254389217334, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 21000 + }, + { + "epoch": 0.21001, + "grad_norm": 1.461437401694092, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21001 + }, + { + "epoch": 0.21002, + "grad_norm": 1.422544853215876, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 21002 + }, + { + "epoch": 0.21003, + "grad_norm": 1.3366035931727727, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21003 + }, + { + "epoch": 0.21004, + "grad_norm": 1.5459735826737728, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 21004 + }, + { + "epoch": 0.21005, + "grad_norm": 1.2461805438471347, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21005 + }, + { + "epoch": 0.21006, + "grad_norm": 1.5147791701333184, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21006 + }, + { + "epoch": 0.21007, + "grad_norm": 1.2901714481509183, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21007 + }, + { + "epoch": 0.21008, + "grad_norm": 1.1429132481952657, + "learning_rate": 0.003, + "loss": 4.045, + "step": 21008 + }, + { + "epoch": 0.21009, + "grad_norm": 1.3931111686626179, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 21009 + }, + { + "epoch": 0.2101, + "grad_norm": 1.4620137134793034, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 21010 + }, + { + "epoch": 0.21011, + "grad_norm": 1.2342907771891274, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 21011 + }, + { + "epoch": 0.21012, + "grad_norm": 1.4617620465492411, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 21012 + }, + { + "epoch": 0.21013, + "grad_norm": 1.2649399673357389, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 21013 + }, + { + "epoch": 0.21014, + "grad_norm": 1.3435256308506187, + "learning_rate": 0.003, + "loss": 4.066, + "step": 21014 + }, + { + "epoch": 0.21015, + "grad_norm": 1.244101791785602, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 21015 + }, + { + "epoch": 0.21016, + "grad_norm": 1.1565299822725221, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21016 + }, + { + "epoch": 0.21017, + "grad_norm": 1.252277784984338, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 21017 + }, + { + "epoch": 0.21018, + "grad_norm": 1.4535768408396037, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 21018 + }, + { + "epoch": 0.21019, + "grad_norm": 1.316193671135584, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 21019 + }, + { + "epoch": 0.2102, + "grad_norm": 1.3432893741725889, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 21020 + }, + { + "epoch": 0.21021, + "grad_norm": 1.0648436783358863, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21021 + }, + { + "epoch": 0.21022, + "grad_norm": 1.6892575258933102, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21022 + }, + { + "epoch": 0.21023, + "grad_norm": 1.1008692586381736, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 21023 + }, + { + "epoch": 0.21024, + "grad_norm": 1.5431788160399122, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 21024 + }, + { + "epoch": 0.21025, + "grad_norm": 1.6087965706302754, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 21025 + }, + { + "epoch": 0.21026, + "grad_norm": 1.3207794878231496, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 21026 + }, + { + "epoch": 0.21027, + "grad_norm": 1.405276543970181, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 21027 + }, + { + "epoch": 0.21028, + "grad_norm": 1.376324143024024, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 21028 + }, + { + "epoch": 0.21029, + "grad_norm": 1.365003958090158, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 21029 + }, + { + "epoch": 0.2103, + "grad_norm": 1.4058150514705787, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 21030 + }, + { + "epoch": 0.21031, + "grad_norm": 1.153189792926779, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21031 + }, + { + "epoch": 0.21032, + "grad_norm": 1.3397695114600154, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21032 + }, + { + "epoch": 0.21033, + "grad_norm": 1.2726878278038598, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21033 + }, + { + "epoch": 0.21034, + "grad_norm": 1.1724902441253162, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 21034 + }, + { + "epoch": 0.21035, + "grad_norm": 1.4622195266657432, + "learning_rate": 0.003, + "loss": 4.049, + "step": 21035 + }, + { + "epoch": 0.21036, + "grad_norm": 1.2728210321227869, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 21036 + }, + { + "epoch": 0.21037, + "grad_norm": 1.337362962671212, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 21037 + }, + { + "epoch": 0.21038, + "grad_norm": 1.6430043034892616, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 21038 + }, + { + "epoch": 0.21039, + "grad_norm": 1.6129415272481018, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 21039 + }, + { + "epoch": 0.2104, + "grad_norm": 1.4475097526229443, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 21040 + }, + { + "epoch": 0.21041, + "grad_norm": 1.1784070418565082, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21041 + }, + { + "epoch": 0.21042, + "grad_norm": 1.622691663690252, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 21042 + }, + { + "epoch": 0.21043, + "grad_norm": 1.2557235638469777, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 21043 + }, + { + "epoch": 0.21044, + "grad_norm": 1.1002234268770594, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 21044 + }, + { + "epoch": 0.21045, + "grad_norm": 1.467101985799226, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 21045 + }, + { + "epoch": 0.21046, + "grad_norm": 1.4592275042934686, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21046 + }, + { + "epoch": 0.21047, + "grad_norm": 1.1961366095601582, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 21047 + }, + { + "epoch": 0.21048, + "grad_norm": 1.3259790306030865, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 21048 + }, + { + "epoch": 0.21049, + "grad_norm": 1.459604071228318, + "learning_rate": 0.003, + "loss": 4.046, + "step": 21049 + }, + { + "epoch": 0.2105, + "grad_norm": 1.3356775644760128, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 21050 + }, + { + "epoch": 0.21051, + "grad_norm": 1.2611272734027086, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 21051 + }, + { + "epoch": 0.21052, + "grad_norm": 1.4561435107404017, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 21052 + }, + { + "epoch": 0.21053, + "grad_norm": 1.3172832097982368, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 21053 + }, + { + "epoch": 0.21054, + "grad_norm": 1.5079848509312952, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 21054 + }, + { + "epoch": 0.21055, + "grad_norm": 1.0682100137209913, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 21055 + }, + { + "epoch": 0.21056, + "grad_norm": 1.3989052113249258, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 21056 + }, + { + "epoch": 0.21057, + "grad_norm": 1.2515853111972497, + "learning_rate": 0.003, + "loss": 4.047, + "step": 21057 + }, + { + "epoch": 0.21058, + "grad_norm": 1.8385061177094724, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 21058 + }, + { + "epoch": 0.21059, + "grad_norm": 1.2619283793884344, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 21059 + }, + { + "epoch": 0.2106, + "grad_norm": 1.1603808791470698, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 21060 + }, + { + "epoch": 0.21061, + "grad_norm": 1.5766504515523034, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 21061 + }, + { + "epoch": 0.21062, + "grad_norm": 1.0923398235748982, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 21062 + }, + { + "epoch": 0.21063, + "grad_norm": 1.3157689007715065, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 21063 + }, + { + "epoch": 0.21064, + "grad_norm": 1.2952466469961974, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 21064 + }, + { + "epoch": 0.21065, + "grad_norm": 1.361447178361404, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 21065 + }, + { + "epoch": 0.21066, + "grad_norm": 1.2877009617325028, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 21066 + }, + { + "epoch": 0.21067, + "grad_norm": 1.4508306716090476, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 21067 + }, + { + "epoch": 0.21068, + "grad_norm": 1.404906272655368, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 21068 + }, + { + "epoch": 0.21069, + "grad_norm": 1.2566162259120792, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21069 + }, + { + "epoch": 0.2107, + "grad_norm": 1.4371513027733627, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 21070 + }, + { + "epoch": 0.21071, + "grad_norm": 1.1741644539260434, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 21071 + }, + { + "epoch": 0.21072, + "grad_norm": 1.7495812264694768, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 21072 + }, + { + "epoch": 0.21073, + "grad_norm": 1.1914029554954333, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21073 + }, + { + "epoch": 0.21074, + "grad_norm": 1.358388855376881, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 21074 + }, + { + "epoch": 0.21075, + "grad_norm": 1.1295297359650347, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 21075 + }, + { + "epoch": 0.21076, + "grad_norm": 1.422826953405643, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21076 + }, + { + "epoch": 0.21077, + "grad_norm": 1.2238983295066468, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21077 + }, + { + "epoch": 0.21078, + "grad_norm": 1.4376730583978223, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 21078 + }, + { + "epoch": 0.21079, + "grad_norm": 1.312652953458203, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21079 + }, + { + "epoch": 0.2108, + "grad_norm": 1.2881762161594188, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 21080 + }, + { + "epoch": 0.21081, + "grad_norm": 1.3811842481459293, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 21081 + }, + { + "epoch": 0.21082, + "grad_norm": 1.4010350452207294, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21082 + }, + { + "epoch": 0.21083, + "grad_norm": 1.290332298644118, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 21083 + }, + { + "epoch": 0.21084, + "grad_norm": 1.489071854621563, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 21084 + }, + { + "epoch": 0.21085, + "grad_norm": 1.5024296642566897, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 21085 + }, + { + "epoch": 0.21086, + "grad_norm": 1.4062125961366714, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 21086 + }, + { + "epoch": 0.21087, + "grad_norm": 1.4251667457024468, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 21087 + }, + { + "epoch": 0.21088, + "grad_norm": 1.572228967046733, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21088 + }, + { + "epoch": 0.21089, + "grad_norm": 1.3883350111110444, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 21089 + }, + { + "epoch": 0.2109, + "grad_norm": 1.3950958358737215, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21090 + }, + { + "epoch": 0.21091, + "grad_norm": 1.2048175579643006, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 21091 + }, + { + "epoch": 0.21092, + "grad_norm": 1.5234295295873461, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 21092 + }, + { + "epoch": 0.21093, + "grad_norm": 1.294754161938382, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 21093 + }, + { + "epoch": 0.21094, + "grad_norm": 1.3512006946099722, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 21094 + }, + { + "epoch": 0.21095, + "grad_norm": 1.3137093691553654, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 21095 + }, + { + "epoch": 0.21096, + "grad_norm": 1.4209343670036005, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 21096 + }, + { + "epoch": 0.21097, + "grad_norm": 1.1895555560423718, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 21097 + }, + { + "epoch": 0.21098, + "grad_norm": 1.336801940566388, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 21098 + }, + { + "epoch": 0.21099, + "grad_norm": 1.2758607849242902, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 21099 + }, + { + "epoch": 0.211, + "grad_norm": 1.2117337581677836, + "learning_rate": 0.003, + "loss": 4.05, + "step": 21100 + }, + { + "epoch": 0.21101, + "grad_norm": 1.3525527482392288, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 21101 + }, + { + "epoch": 0.21102, + "grad_norm": 1.3040393589972092, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 21102 + }, + { + "epoch": 0.21103, + "grad_norm": 1.3424780630363968, + "learning_rate": 0.003, + "loss": 4.063, + "step": 21103 + }, + { + "epoch": 0.21104, + "grad_norm": 1.542667444735479, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 21104 + }, + { + "epoch": 0.21105, + "grad_norm": 1.6365698353753115, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 21105 + }, + { + "epoch": 0.21106, + "grad_norm": 1.2084286816553886, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 21106 + }, + { + "epoch": 0.21107, + "grad_norm": 1.5019987836084807, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 21107 + }, + { + "epoch": 0.21108, + "grad_norm": 1.318700513318784, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21108 + }, + { + "epoch": 0.21109, + "grad_norm": 1.3066584923179836, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21109 + }, + { + "epoch": 0.2111, + "grad_norm": 1.4399512148451805, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21110 + }, + { + "epoch": 0.21111, + "grad_norm": 1.1041215885148292, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 21111 + }, + { + "epoch": 0.21112, + "grad_norm": 1.501694680843476, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21112 + }, + { + "epoch": 0.21113, + "grad_norm": 1.2601037756571198, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 21113 + }, + { + "epoch": 0.21114, + "grad_norm": 1.2271546237973472, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21114 + }, + { + "epoch": 0.21115, + "grad_norm": 1.2376999598976492, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21115 + }, + { + "epoch": 0.21116, + "grad_norm": 1.5430782245530037, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 21116 + }, + { + "epoch": 0.21117, + "grad_norm": 1.3787545900202918, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 21117 + }, + { + "epoch": 0.21118, + "grad_norm": 1.4281319973902007, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 21118 + }, + { + "epoch": 0.21119, + "grad_norm": 1.6838838996168861, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21119 + }, + { + "epoch": 0.2112, + "grad_norm": 1.2215779967116982, + "learning_rate": 0.003, + "loss": 4.079, + "step": 21120 + }, + { + "epoch": 0.21121, + "grad_norm": 1.5199216079850038, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 21121 + }, + { + "epoch": 0.21122, + "grad_norm": 1.1448715515495345, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21122 + }, + { + "epoch": 0.21123, + "grad_norm": 1.4014754531269893, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 21123 + }, + { + "epoch": 0.21124, + "grad_norm": 1.143267669929633, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 21124 + }, + { + "epoch": 0.21125, + "grad_norm": 1.507889372373385, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 21125 + }, + { + "epoch": 0.21126, + "grad_norm": 1.1177387032574877, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 21126 + }, + { + "epoch": 0.21127, + "grad_norm": 1.4170981959098783, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 21127 + }, + { + "epoch": 0.21128, + "grad_norm": 1.193182993088027, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21128 + }, + { + "epoch": 0.21129, + "grad_norm": 1.4457888997085144, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 21129 + }, + { + "epoch": 0.2113, + "grad_norm": 1.3240888408523435, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 21130 + }, + { + "epoch": 0.21131, + "grad_norm": 1.4805545843347245, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 21131 + }, + { + "epoch": 0.21132, + "grad_norm": 1.4871800892653317, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 21132 + }, + { + "epoch": 0.21133, + "grad_norm": 1.2991302286864286, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 21133 + }, + { + "epoch": 0.21134, + "grad_norm": 1.6445450451190515, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 21134 + }, + { + "epoch": 0.21135, + "grad_norm": 1.4721111980365607, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 21135 + }, + { + "epoch": 0.21136, + "grad_norm": 1.2722739660811953, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21136 + }, + { + "epoch": 0.21137, + "grad_norm": 1.3982694574879595, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21137 + }, + { + "epoch": 0.21138, + "grad_norm": 1.226479278940514, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 21138 + }, + { + "epoch": 0.21139, + "grad_norm": 1.3917173367564466, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 21139 + }, + { + "epoch": 0.2114, + "grad_norm": 1.1647298403369704, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 21140 + }, + { + "epoch": 0.21141, + "grad_norm": 1.575793714531139, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 21141 + }, + { + "epoch": 0.21142, + "grad_norm": 1.199875711689336, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21142 + }, + { + "epoch": 0.21143, + "grad_norm": 1.6414373088741725, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 21143 + }, + { + "epoch": 0.21144, + "grad_norm": 1.091308331386974, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 21144 + }, + { + "epoch": 0.21145, + "grad_norm": 1.5590872193904934, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 21145 + }, + { + "epoch": 0.21146, + "grad_norm": 1.2853562163327288, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 21146 + }, + { + "epoch": 0.21147, + "grad_norm": 1.5550976773299796, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 21147 + }, + { + "epoch": 0.21148, + "grad_norm": 1.114369420327927, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21148 + }, + { + "epoch": 0.21149, + "grad_norm": 1.3614548969594964, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 21149 + }, + { + "epoch": 0.2115, + "grad_norm": 1.2148196019739446, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 21150 + }, + { + "epoch": 0.21151, + "grad_norm": 1.340682128882772, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 21151 + }, + { + "epoch": 0.21152, + "grad_norm": 1.4000709731369123, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 21152 + }, + { + "epoch": 0.21153, + "grad_norm": 1.1974593610511264, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21153 + }, + { + "epoch": 0.21154, + "grad_norm": 1.5536190722535004, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 21154 + }, + { + "epoch": 0.21155, + "grad_norm": 1.3833146429262424, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 21155 + }, + { + "epoch": 0.21156, + "grad_norm": 1.2868018237459313, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 21156 + }, + { + "epoch": 0.21157, + "grad_norm": 1.3479488477762227, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 21157 + }, + { + "epoch": 0.21158, + "grad_norm": 1.3625397818967828, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 21158 + }, + { + "epoch": 0.21159, + "grad_norm": 1.20914255417048, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 21159 + }, + { + "epoch": 0.2116, + "grad_norm": 1.3458094476032658, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21160 + }, + { + "epoch": 0.21161, + "grad_norm": 1.24337456561934, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 21161 + }, + { + "epoch": 0.21162, + "grad_norm": 1.2620528232172612, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21162 + }, + { + "epoch": 0.21163, + "grad_norm": 1.3810456235036637, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21163 + }, + { + "epoch": 0.21164, + "grad_norm": 1.511592471855777, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 21164 + }, + { + "epoch": 0.21165, + "grad_norm": 1.3431865078237681, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 21165 + }, + { + "epoch": 0.21166, + "grad_norm": 1.7412655610560623, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 21166 + }, + { + "epoch": 0.21167, + "grad_norm": 1.1109768866361736, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 21167 + }, + { + "epoch": 0.21168, + "grad_norm": 1.399082597999424, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 21168 + }, + { + "epoch": 0.21169, + "grad_norm": 1.1489789410732114, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21169 + }, + { + "epoch": 0.2117, + "grad_norm": 1.524057385496699, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21170 + }, + { + "epoch": 0.21171, + "grad_norm": 1.2092265332397514, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 21171 + }, + { + "epoch": 0.21172, + "grad_norm": 1.3782809880096014, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 21172 + }, + { + "epoch": 0.21173, + "grad_norm": 1.415031645517347, + "learning_rate": 0.003, + "loss": 4.073, + "step": 21173 + }, + { + "epoch": 0.21174, + "grad_norm": 1.4764342717559167, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 21174 + }, + { + "epoch": 0.21175, + "grad_norm": 1.5080392364169124, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 21175 + }, + { + "epoch": 0.21176, + "grad_norm": 1.4883776418571604, + "learning_rate": 0.003, + "loss": 4.079, + "step": 21176 + }, + { + "epoch": 0.21177, + "grad_norm": 1.5612744737054056, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21177 + }, + { + "epoch": 0.21178, + "grad_norm": 1.2024716611041812, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 21178 + }, + { + "epoch": 0.21179, + "grad_norm": 1.1880680580697465, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 21179 + }, + { + "epoch": 0.2118, + "grad_norm": 1.6787856254774396, + "learning_rate": 0.003, + "loss": 4.049, + "step": 21180 + }, + { + "epoch": 0.21181, + "grad_norm": 1.2348843210740865, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 21181 + }, + { + "epoch": 0.21182, + "grad_norm": 1.272340607329311, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21182 + }, + { + "epoch": 0.21183, + "grad_norm": 1.4679072254469465, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 21183 + }, + { + "epoch": 0.21184, + "grad_norm": 1.3585140016622648, + "learning_rate": 0.003, + "loss": 4.077, + "step": 21184 + }, + { + "epoch": 0.21185, + "grad_norm": 1.3043316251259256, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 21185 + }, + { + "epoch": 0.21186, + "grad_norm": 1.1942951148218348, + "learning_rate": 0.003, + "loss": 4.079, + "step": 21186 + }, + { + "epoch": 0.21187, + "grad_norm": 1.5555996789667779, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 21187 + }, + { + "epoch": 0.21188, + "grad_norm": 1.2566766113824286, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 21188 + }, + { + "epoch": 0.21189, + "grad_norm": 1.3926383314895503, + "learning_rate": 0.003, + "loss": 4.029, + "step": 21189 + }, + { + "epoch": 0.2119, + "grad_norm": 1.2036849778099206, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21190 + }, + { + "epoch": 0.21191, + "grad_norm": 1.3813535243352322, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 21191 + }, + { + "epoch": 0.21192, + "grad_norm": 1.2470340700518887, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 21192 + }, + { + "epoch": 0.21193, + "grad_norm": 1.2335173857517638, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 21193 + }, + { + "epoch": 0.21194, + "grad_norm": 1.4050847539015667, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 21194 + }, + { + "epoch": 0.21195, + "grad_norm": 1.405184145664891, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21195 + }, + { + "epoch": 0.21196, + "grad_norm": 1.1916257136074022, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 21196 + }, + { + "epoch": 0.21197, + "grad_norm": 1.2863503878206486, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21197 + }, + { + "epoch": 0.21198, + "grad_norm": 1.3920618485759595, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 21198 + }, + { + "epoch": 0.21199, + "grad_norm": 1.6393826888044547, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21199 + }, + { + "epoch": 0.212, + "grad_norm": 1.5873450028938483, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21200 + }, + { + "epoch": 0.21201, + "grad_norm": 1.276498523591141, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 21201 + }, + { + "epoch": 0.21202, + "grad_norm": 1.2141772863673175, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 21202 + }, + { + "epoch": 0.21203, + "grad_norm": 1.7306047496665609, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 21203 + }, + { + "epoch": 0.21204, + "grad_norm": 1.2487486082239754, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 21204 + }, + { + "epoch": 0.21205, + "grad_norm": 1.712695784117531, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 21205 + }, + { + "epoch": 0.21206, + "grad_norm": 1.1323199873225769, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21206 + }, + { + "epoch": 0.21207, + "grad_norm": 1.2652313130632524, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 21207 + }, + { + "epoch": 0.21208, + "grad_norm": 1.2512745025751575, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 21208 + }, + { + "epoch": 0.21209, + "grad_norm": 1.2466650859199238, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21209 + }, + { + "epoch": 0.2121, + "grad_norm": 1.2620167175109485, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21210 + }, + { + "epoch": 0.21211, + "grad_norm": 1.5212220521490063, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 21211 + }, + { + "epoch": 0.21212, + "grad_norm": 1.2635532006517542, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 21212 + }, + { + "epoch": 0.21213, + "grad_norm": 1.347810331585149, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21213 + }, + { + "epoch": 0.21214, + "grad_norm": 1.6564875740727845, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21214 + }, + { + "epoch": 0.21215, + "grad_norm": 1.5438927653902956, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 21215 + }, + { + "epoch": 0.21216, + "grad_norm": 1.208733121779013, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21216 + }, + { + "epoch": 0.21217, + "grad_norm": 1.2433388919071149, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 21217 + }, + { + "epoch": 0.21218, + "grad_norm": 1.4277012029638962, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21218 + }, + { + "epoch": 0.21219, + "grad_norm": 1.2923936162093763, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 21219 + }, + { + "epoch": 0.2122, + "grad_norm": 1.1617299223503543, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21220 + }, + { + "epoch": 0.21221, + "grad_norm": 1.7625335309163919, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 21221 + }, + { + "epoch": 0.21222, + "grad_norm": 1.0722691251519127, + "learning_rate": 0.003, + "loss": 4.045, + "step": 21222 + }, + { + "epoch": 0.21223, + "grad_norm": 1.523851439821383, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21223 + }, + { + "epoch": 0.21224, + "grad_norm": 1.5241769908828782, + "learning_rate": 0.003, + "loss": 4.049, + "step": 21224 + }, + { + "epoch": 0.21225, + "grad_norm": 1.5763055700460948, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 21225 + }, + { + "epoch": 0.21226, + "grad_norm": 1.288345375256502, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21226 + }, + { + "epoch": 0.21227, + "grad_norm": 1.3593170373457841, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 21227 + }, + { + "epoch": 0.21228, + "grad_norm": 1.3798720304842758, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21228 + }, + { + "epoch": 0.21229, + "grad_norm": 1.11347139995492, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21229 + }, + { + "epoch": 0.2123, + "grad_norm": 1.4881861926379105, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21230 + }, + { + "epoch": 0.21231, + "grad_norm": 1.222191148663261, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21231 + }, + { + "epoch": 0.21232, + "grad_norm": 1.4598446473774203, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 21232 + }, + { + "epoch": 0.21233, + "grad_norm": 1.3035913178757261, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 21233 + }, + { + "epoch": 0.21234, + "grad_norm": 1.4240272674012957, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 21234 + }, + { + "epoch": 0.21235, + "grad_norm": 1.7053319580459514, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21235 + }, + { + "epoch": 0.21236, + "grad_norm": 1.445897432070719, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 21236 + }, + { + "epoch": 0.21237, + "grad_norm": 1.3817227194504706, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 21237 + }, + { + "epoch": 0.21238, + "grad_norm": 1.2676941447795234, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21238 + }, + { + "epoch": 0.21239, + "grad_norm": 1.5641612851516646, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 21239 + }, + { + "epoch": 0.2124, + "grad_norm": 1.1241931151835491, + "learning_rate": 0.003, + "loss": 4.029, + "step": 21240 + }, + { + "epoch": 0.21241, + "grad_norm": 1.4868409328389938, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 21241 + }, + { + "epoch": 0.21242, + "grad_norm": 1.2405669154949257, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 21242 + }, + { + "epoch": 0.21243, + "grad_norm": 1.2224834326951193, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 21243 + }, + { + "epoch": 0.21244, + "grad_norm": 1.3189689925487944, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21244 + }, + { + "epoch": 0.21245, + "grad_norm": 1.347220053289232, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21245 + }, + { + "epoch": 0.21246, + "grad_norm": 1.2655707700006256, + "learning_rate": 0.003, + "loss": 4.058, + "step": 21246 + }, + { + "epoch": 0.21247, + "grad_norm": 1.362651105931732, + "learning_rate": 0.003, + "loss": 4.056, + "step": 21247 + }, + { + "epoch": 0.21248, + "grad_norm": 1.4785751933994802, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21248 + }, + { + "epoch": 0.21249, + "grad_norm": 1.2504349836530144, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 21249 + }, + { + "epoch": 0.2125, + "grad_norm": 1.2990606380438254, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 21250 + }, + { + "epoch": 0.21251, + "grad_norm": 1.42497627382582, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 21251 + }, + { + "epoch": 0.21252, + "grad_norm": 1.1735585685930139, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 21252 + }, + { + "epoch": 0.21253, + "grad_norm": 1.795106850549211, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21253 + }, + { + "epoch": 0.21254, + "grad_norm": 1.0061555841718746, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 21254 + }, + { + "epoch": 0.21255, + "grad_norm": 1.4122369561351449, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21255 + }, + { + "epoch": 0.21256, + "grad_norm": 1.4461580757000376, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21256 + }, + { + "epoch": 0.21257, + "grad_norm": 1.4284828541332442, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21257 + }, + { + "epoch": 0.21258, + "grad_norm": 1.2986806596599783, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21258 + }, + { + "epoch": 0.21259, + "grad_norm": 1.6524854377190041, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 21259 + }, + { + "epoch": 0.2126, + "grad_norm": 0.9870173335163619, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 21260 + }, + { + "epoch": 0.21261, + "grad_norm": 1.5495893972134074, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 21261 + }, + { + "epoch": 0.21262, + "grad_norm": 1.126423706366596, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 21262 + }, + { + "epoch": 0.21263, + "grad_norm": 1.3734463560813661, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21263 + }, + { + "epoch": 0.21264, + "grad_norm": 1.176126541735446, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21264 + }, + { + "epoch": 0.21265, + "grad_norm": 1.370153879548395, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21265 + }, + { + "epoch": 0.21266, + "grad_norm": 1.3532518679834689, + "learning_rate": 0.003, + "loss": 4.052, + "step": 21266 + }, + { + "epoch": 0.21267, + "grad_norm": 1.6946670175674308, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21267 + }, + { + "epoch": 0.21268, + "grad_norm": 1.0704729251671854, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21268 + }, + { + "epoch": 0.21269, + "grad_norm": 1.4596383231421428, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 21269 + }, + { + "epoch": 0.2127, + "grad_norm": 1.1808133393608502, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 21270 + }, + { + "epoch": 0.21271, + "grad_norm": 1.4127453609322973, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 21271 + }, + { + "epoch": 0.21272, + "grad_norm": 1.2825462550635216, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 21272 + }, + { + "epoch": 0.21273, + "grad_norm": 1.4332865788298836, + "learning_rate": 0.003, + "loss": 4.06, + "step": 21273 + }, + { + "epoch": 0.21274, + "grad_norm": 1.2555266101784204, + "learning_rate": 0.003, + "loss": 4.086, + "step": 21274 + }, + { + "epoch": 0.21275, + "grad_norm": 1.491116264139151, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 21275 + }, + { + "epoch": 0.21276, + "grad_norm": 1.357691091796868, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 21276 + }, + { + "epoch": 0.21277, + "grad_norm": 1.5137001660115919, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 21277 + }, + { + "epoch": 0.21278, + "grad_norm": 1.1125889653319778, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21278 + }, + { + "epoch": 0.21279, + "grad_norm": 1.3304734539024916, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21279 + }, + { + "epoch": 0.2128, + "grad_norm": 1.4679732364278968, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 21280 + }, + { + "epoch": 0.21281, + "grad_norm": 1.2033274494571464, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 21281 + }, + { + "epoch": 0.21282, + "grad_norm": 1.2190823850731825, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 21282 + }, + { + "epoch": 0.21283, + "grad_norm": 1.4288583356987408, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 21283 + }, + { + "epoch": 0.21284, + "grad_norm": 1.31602755126889, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 21284 + }, + { + "epoch": 0.21285, + "grad_norm": 1.2275327062705983, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 21285 + }, + { + "epoch": 0.21286, + "grad_norm": 1.5414860872345424, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21286 + }, + { + "epoch": 0.21287, + "grad_norm": 1.1120019871745512, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21287 + }, + { + "epoch": 0.21288, + "grad_norm": 1.4831050923685816, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21288 + }, + { + "epoch": 0.21289, + "grad_norm": 1.7494424733737561, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 21289 + }, + { + "epoch": 0.2129, + "grad_norm": 1.342926814624509, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21290 + }, + { + "epoch": 0.21291, + "grad_norm": 1.598270038306869, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21291 + }, + { + "epoch": 0.21292, + "grad_norm": 1.5961776628467343, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21292 + }, + { + "epoch": 0.21293, + "grad_norm": 0.9433010852092817, + "learning_rate": 0.003, + "loss": 4.05, + "step": 21293 + }, + { + "epoch": 0.21294, + "grad_norm": 1.3925449914517969, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 21294 + }, + { + "epoch": 0.21295, + "grad_norm": 1.3515145641464523, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 21295 + }, + { + "epoch": 0.21296, + "grad_norm": 1.4166695760130046, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21296 + }, + { + "epoch": 0.21297, + "grad_norm": 1.3591828142618334, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21297 + }, + { + "epoch": 0.21298, + "grad_norm": 1.5153716123119334, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 21298 + }, + { + "epoch": 0.21299, + "grad_norm": 1.6452501562883108, + "learning_rate": 0.003, + "loss": 4.075, + "step": 21299 + }, + { + "epoch": 0.213, + "grad_norm": 1.196597809663666, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21300 + }, + { + "epoch": 0.21301, + "grad_norm": 1.4346809471876099, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 21301 + }, + { + "epoch": 0.21302, + "grad_norm": 1.2582888762436075, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 21302 + }, + { + "epoch": 0.21303, + "grad_norm": 1.583365216904255, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 21303 + }, + { + "epoch": 0.21304, + "grad_norm": 1.2243987322696352, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 21304 + }, + { + "epoch": 0.21305, + "grad_norm": 1.3782757485019397, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 21305 + }, + { + "epoch": 0.21306, + "grad_norm": 1.3712449684226944, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 21306 + }, + { + "epoch": 0.21307, + "grad_norm": 1.452408027239848, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 21307 + }, + { + "epoch": 0.21308, + "grad_norm": 1.227024123708545, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 21308 + }, + { + "epoch": 0.21309, + "grad_norm": 1.2127778579692743, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21309 + }, + { + "epoch": 0.2131, + "grad_norm": 1.3640909424814787, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 21310 + }, + { + "epoch": 0.21311, + "grad_norm": 1.4625961453018894, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 21311 + }, + { + "epoch": 0.21312, + "grad_norm": 1.4381037832524861, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 21312 + }, + { + "epoch": 0.21313, + "grad_norm": 1.1144720293083312, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 21313 + }, + { + "epoch": 0.21314, + "grad_norm": 1.6391039384828037, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21314 + }, + { + "epoch": 0.21315, + "grad_norm": 1.249524180399814, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21315 + }, + { + "epoch": 0.21316, + "grad_norm": 1.306337595950138, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 21316 + }, + { + "epoch": 0.21317, + "grad_norm": 1.3883837054487016, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 21317 + }, + { + "epoch": 0.21318, + "grad_norm": 1.6026302139424504, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 21318 + }, + { + "epoch": 0.21319, + "grad_norm": 1.2604444747762127, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21319 + }, + { + "epoch": 0.2132, + "grad_norm": 1.3547506649531902, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 21320 + }, + { + "epoch": 0.21321, + "grad_norm": 1.5400351682445699, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 21321 + }, + { + "epoch": 0.21322, + "grad_norm": 1.4849351496306467, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21322 + }, + { + "epoch": 0.21323, + "grad_norm": 1.2063234257034268, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 21323 + }, + { + "epoch": 0.21324, + "grad_norm": 1.5292223315443574, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21324 + }, + { + "epoch": 0.21325, + "grad_norm": 1.151039941590078, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 21325 + }, + { + "epoch": 0.21326, + "grad_norm": 1.3318800995328697, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 21326 + }, + { + "epoch": 0.21327, + "grad_norm": 1.2652305263453445, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 21327 + }, + { + "epoch": 0.21328, + "grad_norm": 1.3515081434361438, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 21328 + }, + { + "epoch": 0.21329, + "grad_norm": 1.2875379433950092, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 21329 + }, + { + "epoch": 0.2133, + "grad_norm": 1.3972616385126326, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 21330 + }, + { + "epoch": 0.21331, + "grad_norm": 1.4816983405239224, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21331 + }, + { + "epoch": 0.21332, + "grad_norm": 1.503467852555335, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 21332 + }, + { + "epoch": 0.21333, + "grad_norm": 1.287556683534024, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 21333 + }, + { + "epoch": 0.21334, + "grad_norm": 1.293434698898435, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 21334 + }, + { + "epoch": 0.21335, + "grad_norm": 1.2973040929964068, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21335 + }, + { + "epoch": 0.21336, + "grad_norm": 1.4591818538831844, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 21336 + }, + { + "epoch": 0.21337, + "grad_norm": 1.31338047027029, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 21337 + }, + { + "epoch": 0.21338, + "grad_norm": 1.2076429007725722, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 21338 + }, + { + "epoch": 0.21339, + "grad_norm": 1.263751954787633, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 21339 + }, + { + "epoch": 0.2134, + "grad_norm": 1.2731326941221517, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 21340 + }, + { + "epoch": 0.21341, + "grad_norm": 1.4210042790778448, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21341 + }, + { + "epoch": 0.21342, + "grad_norm": 1.3568925479823957, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21342 + }, + { + "epoch": 0.21343, + "grad_norm": 1.5384252079790888, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 21343 + }, + { + "epoch": 0.21344, + "grad_norm": 1.4478375676361275, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21344 + }, + { + "epoch": 0.21345, + "grad_norm": 1.3680354906520085, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21345 + }, + { + "epoch": 0.21346, + "grad_norm": 1.3890887098322335, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 21346 + }, + { + "epoch": 0.21347, + "grad_norm": 1.1193721005748065, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 21347 + }, + { + "epoch": 0.21348, + "grad_norm": 1.5275977448430993, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21348 + }, + { + "epoch": 0.21349, + "grad_norm": 1.3959001019528146, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21349 + }, + { + "epoch": 0.2135, + "grad_norm": 1.490204399602071, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 21350 + }, + { + "epoch": 0.21351, + "grad_norm": 1.1499473326805703, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 21351 + }, + { + "epoch": 0.21352, + "grad_norm": 1.727939275779018, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21352 + }, + { + "epoch": 0.21353, + "grad_norm": 1.072972726456787, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 21353 + }, + { + "epoch": 0.21354, + "grad_norm": 1.483607518560974, + "learning_rate": 0.003, + "loss": 4.071, + "step": 21354 + }, + { + "epoch": 0.21355, + "grad_norm": 1.562193661576263, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 21355 + }, + { + "epoch": 0.21356, + "grad_norm": 1.5136521928411162, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21356 + }, + { + "epoch": 0.21357, + "grad_norm": 1.3266924162009406, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 21357 + }, + { + "epoch": 0.21358, + "grad_norm": 1.5459850487255908, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 21358 + }, + { + "epoch": 0.21359, + "grad_norm": 1.306340308286527, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 21359 + }, + { + "epoch": 0.2136, + "grad_norm": 1.4437227572721356, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 21360 + }, + { + "epoch": 0.21361, + "grad_norm": 1.1312331537840155, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 21361 + }, + { + "epoch": 0.21362, + "grad_norm": 1.4498115212610185, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 21362 + }, + { + "epoch": 0.21363, + "grad_norm": 1.2003047711896486, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 21363 + }, + { + "epoch": 0.21364, + "grad_norm": 1.330082599913098, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21364 + }, + { + "epoch": 0.21365, + "grad_norm": 1.5471781546909327, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 21365 + }, + { + "epoch": 0.21366, + "grad_norm": 1.4316421630210903, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21366 + }, + { + "epoch": 0.21367, + "grad_norm": 1.594803933302907, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21367 + }, + { + "epoch": 0.21368, + "grad_norm": 1.1617312218503295, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 21368 + }, + { + "epoch": 0.21369, + "grad_norm": 1.2293064917378977, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21369 + }, + { + "epoch": 0.2137, + "grad_norm": 1.418555917149502, + "learning_rate": 0.003, + "loss": 4.095, + "step": 21370 + }, + { + "epoch": 0.21371, + "grad_norm": 1.3902348667270346, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 21371 + }, + { + "epoch": 0.21372, + "grad_norm": 1.189623983536023, + "learning_rate": 0.003, + "loss": 4.084, + "step": 21372 + }, + { + "epoch": 0.21373, + "grad_norm": 1.5618531870219965, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 21373 + }, + { + "epoch": 0.21374, + "grad_norm": 1.065036404388025, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 21374 + }, + { + "epoch": 0.21375, + "grad_norm": 1.283965896979844, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21375 + }, + { + "epoch": 0.21376, + "grad_norm": 1.261048313827542, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 21376 + }, + { + "epoch": 0.21377, + "grad_norm": 1.4229495898524944, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 21377 + }, + { + "epoch": 0.21378, + "grad_norm": 1.3428569523264362, + "learning_rate": 0.003, + "loss": 4.052, + "step": 21378 + }, + { + "epoch": 0.21379, + "grad_norm": 1.5985336646598158, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 21379 + }, + { + "epoch": 0.2138, + "grad_norm": 1.5818368985579856, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21380 + }, + { + "epoch": 0.21381, + "grad_norm": 1.4472336461809965, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 21381 + }, + { + "epoch": 0.21382, + "grad_norm": 1.2563415448416158, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 21382 + }, + { + "epoch": 0.21383, + "grad_norm": 1.4436033902315786, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21383 + }, + { + "epoch": 0.21384, + "grad_norm": 1.2441850730232435, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 21384 + }, + { + "epoch": 0.21385, + "grad_norm": 1.5092420928285513, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 21385 + }, + { + "epoch": 0.21386, + "grad_norm": 1.1510500409342292, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 21386 + }, + { + "epoch": 0.21387, + "grad_norm": 1.3560229720526926, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21387 + }, + { + "epoch": 0.21388, + "grad_norm": 1.3342466552744916, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 21388 + }, + { + "epoch": 0.21389, + "grad_norm": 1.3020543951603012, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21389 + }, + { + "epoch": 0.2139, + "grad_norm": 1.608568950125023, + "learning_rate": 0.003, + "loss": 4.034, + "step": 21390 + }, + { + "epoch": 0.21391, + "grad_norm": 1.1353591119174902, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 21391 + }, + { + "epoch": 0.21392, + "grad_norm": 1.6950414767063793, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 21392 + }, + { + "epoch": 0.21393, + "grad_norm": 1.2387184056870462, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 21393 + }, + { + "epoch": 0.21394, + "grad_norm": 1.3613953563392753, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 21394 + }, + { + "epoch": 0.21395, + "grad_norm": 1.629304745379985, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 21395 + }, + { + "epoch": 0.21396, + "grad_norm": 1.0572390211011273, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 21396 + }, + { + "epoch": 0.21397, + "grad_norm": 1.3300832783882734, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 21397 + }, + { + "epoch": 0.21398, + "grad_norm": 1.3542246431342029, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 21398 + }, + { + "epoch": 0.21399, + "grad_norm": 1.39666540188732, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 21399 + }, + { + "epoch": 0.214, + "grad_norm": 1.4361003719188568, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 21400 + }, + { + "epoch": 0.21401, + "grad_norm": 1.300783711710657, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21401 + }, + { + "epoch": 0.21402, + "grad_norm": 1.290377078934423, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21402 + }, + { + "epoch": 0.21403, + "grad_norm": 1.6750658808471093, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 21403 + }, + { + "epoch": 0.21404, + "grad_norm": 0.9435626847165859, + "learning_rate": 0.003, + "loss": 4.055, + "step": 21404 + }, + { + "epoch": 0.21405, + "grad_norm": 1.4968566182794512, + "learning_rate": 0.003, + "loss": 4.058, + "step": 21405 + }, + { + "epoch": 0.21406, + "grad_norm": 1.2175482207573307, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 21406 + }, + { + "epoch": 0.21407, + "grad_norm": 1.577026013030714, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 21407 + }, + { + "epoch": 0.21408, + "grad_norm": 1.59911020245519, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 21408 + }, + { + "epoch": 0.21409, + "grad_norm": 1.4170936917729833, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 21409 + }, + { + "epoch": 0.2141, + "grad_norm": 1.3962209202302078, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 21410 + }, + { + "epoch": 0.21411, + "grad_norm": 1.3395683786301085, + "learning_rate": 0.003, + "loss": 4.059, + "step": 21411 + }, + { + "epoch": 0.21412, + "grad_norm": 1.4282438033940026, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21412 + }, + { + "epoch": 0.21413, + "grad_norm": 1.238506016974487, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21413 + }, + { + "epoch": 0.21414, + "grad_norm": 1.3051265514631454, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 21414 + }, + { + "epoch": 0.21415, + "grad_norm": 1.181389922210956, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21415 + }, + { + "epoch": 0.21416, + "grad_norm": 1.4710337042861845, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 21416 + }, + { + "epoch": 0.21417, + "grad_norm": 1.4680527309778317, + "learning_rate": 0.003, + "loss": 4.05, + "step": 21417 + }, + { + "epoch": 0.21418, + "grad_norm": 1.989767235792791, + "learning_rate": 0.003, + "loss": 4.077, + "step": 21418 + }, + { + "epoch": 0.21419, + "grad_norm": 0.9795602781230998, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21419 + }, + { + "epoch": 0.2142, + "grad_norm": 1.641796412983382, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21420 + }, + { + "epoch": 0.21421, + "grad_norm": 1.1101786702473517, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21421 + }, + { + "epoch": 0.21422, + "grad_norm": 1.2490099939055685, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 21422 + }, + { + "epoch": 0.21423, + "grad_norm": 1.4418455337047351, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21423 + }, + { + "epoch": 0.21424, + "grad_norm": 1.3036484430249282, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 21424 + }, + { + "epoch": 0.21425, + "grad_norm": 1.26627535365833, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 21425 + }, + { + "epoch": 0.21426, + "grad_norm": 1.595267433637214, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 21426 + }, + { + "epoch": 0.21427, + "grad_norm": 1.1415584876711913, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 21427 + }, + { + "epoch": 0.21428, + "grad_norm": 1.553396948370729, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 21428 + }, + { + "epoch": 0.21429, + "grad_norm": 1.3669329270555002, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21429 + }, + { + "epoch": 0.2143, + "grad_norm": 1.545399754488947, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 21430 + }, + { + "epoch": 0.21431, + "grad_norm": 1.3305307525785282, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 21431 + }, + { + "epoch": 0.21432, + "grad_norm": 1.340621650832546, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 21432 + }, + { + "epoch": 0.21433, + "grad_norm": 1.3656141783233655, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21433 + }, + { + "epoch": 0.21434, + "grad_norm": 1.350982127361377, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 21434 + }, + { + "epoch": 0.21435, + "grad_norm": 1.602265634975586, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21435 + }, + { + "epoch": 0.21436, + "grad_norm": 1.2430334534062013, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 21436 + }, + { + "epoch": 0.21437, + "grad_norm": 1.5552661004651134, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 21437 + }, + { + "epoch": 0.21438, + "grad_norm": 1.4210844535959881, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 21438 + }, + { + "epoch": 0.21439, + "grad_norm": 1.3154138728796594, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 21439 + }, + { + "epoch": 0.2144, + "grad_norm": 1.2269413756659808, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 21440 + }, + { + "epoch": 0.21441, + "grad_norm": 1.4585751482741935, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 21441 + }, + { + "epoch": 0.21442, + "grad_norm": 1.1810732345864168, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 21442 + }, + { + "epoch": 0.21443, + "grad_norm": 1.5843736873721943, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 21443 + }, + { + "epoch": 0.21444, + "grad_norm": 1.1867732253924455, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 21444 + }, + { + "epoch": 0.21445, + "grad_norm": 1.5217034118480697, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 21445 + }, + { + "epoch": 0.21446, + "grad_norm": 1.366518802802389, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 21446 + }, + { + "epoch": 0.21447, + "grad_norm": 1.2987407839800102, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 21447 + }, + { + "epoch": 0.21448, + "grad_norm": 1.4846561697401763, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 21448 + }, + { + "epoch": 0.21449, + "grad_norm": 1.4194658593227094, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 21449 + }, + { + "epoch": 0.2145, + "grad_norm": 1.2720148168821652, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 21450 + }, + { + "epoch": 0.21451, + "grad_norm": 1.4812306933833717, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 21451 + }, + { + "epoch": 0.21452, + "grad_norm": 1.2221215107688335, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 21452 + }, + { + "epoch": 0.21453, + "grad_norm": 1.4207095245756889, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21453 + }, + { + "epoch": 0.21454, + "grad_norm": 1.2603927311534966, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 21454 + }, + { + "epoch": 0.21455, + "grad_norm": 1.570954415430875, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 21455 + }, + { + "epoch": 0.21456, + "grad_norm": 1.3941117180325808, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21456 + }, + { + "epoch": 0.21457, + "grad_norm": 1.2417437923049768, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 21457 + }, + { + "epoch": 0.21458, + "grad_norm": 1.7488725029412966, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 21458 + }, + { + "epoch": 0.21459, + "grad_norm": 1.2214428170658385, + "learning_rate": 0.003, + "loss": 4.07, + "step": 21459 + }, + { + "epoch": 0.2146, + "grad_norm": 1.2970704147224645, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21460 + }, + { + "epoch": 0.21461, + "grad_norm": 1.2203965175276164, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 21461 + }, + { + "epoch": 0.21462, + "grad_norm": 1.50131381958523, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 21462 + }, + { + "epoch": 0.21463, + "grad_norm": 1.3875039130467537, + "learning_rate": 0.003, + "loss": 4.088, + "step": 21463 + }, + { + "epoch": 0.21464, + "grad_norm": 1.3995200427254, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 21464 + }, + { + "epoch": 0.21465, + "grad_norm": 1.1731856716385536, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21465 + }, + { + "epoch": 0.21466, + "grad_norm": 1.3101837349274383, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21466 + }, + { + "epoch": 0.21467, + "grad_norm": 1.3126630773851762, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 21467 + }, + { + "epoch": 0.21468, + "grad_norm": 1.4889895647025362, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 21468 + }, + { + "epoch": 0.21469, + "grad_norm": 1.3248932495708168, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 21469 + }, + { + "epoch": 0.2147, + "grad_norm": 1.74651213400279, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 21470 + }, + { + "epoch": 0.21471, + "grad_norm": 1.1062616432931969, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 21471 + }, + { + "epoch": 0.21472, + "grad_norm": 1.3867854769699446, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 21472 + }, + { + "epoch": 0.21473, + "grad_norm": 1.3543784557814584, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 21473 + }, + { + "epoch": 0.21474, + "grad_norm": 1.323656914816879, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21474 + }, + { + "epoch": 0.21475, + "grad_norm": 1.4335241743346856, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21475 + }, + { + "epoch": 0.21476, + "grad_norm": 1.4966983163562062, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 21476 + }, + { + "epoch": 0.21477, + "grad_norm": 1.341149323909709, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 21477 + }, + { + "epoch": 0.21478, + "grad_norm": 1.4698581882393453, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 21478 + }, + { + "epoch": 0.21479, + "grad_norm": 1.2797141996345442, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 21479 + }, + { + "epoch": 0.2148, + "grad_norm": 1.3373313242855844, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 21480 + }, + { + "epoch": 0.21481, + "grad_norm": 1.2637458978020324, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 21481 + }, + { + "epoch": 0.21482, + "grad_norm": 1.2091973333952728, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 21482 + }, + { + "epoch": 0.21483, + "grad_norm": 1.2198409255112599, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 21483 + }, + { + "epoch": 0.21484, + "grad_norm": 1.5400666985787543, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 21484 + }, + { + "epoch": 0.21485, + "grad_norm": 1.2893391802163316, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21485 + }, + { + "epoch": 0.21486, + "grad_norm": 1.607570309213407, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 21486 + }, + { + "epoch": 0.21487, + "grad_norm": 1.2989691916592647, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21487 + }, + { + "epoch": 0.21488, + "grad_norm": 1.5647644877402134, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 21488 + }, + { + "epoch": 0.21489, + "grad_norm": 1.1874910266671246, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21489 + }, + { + "epoch": 0.2149, + "grad_norm": 1.3350334919767148, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 21490 + }, + { + "epoch": 0.21491, + "grad_norm": 1.4462992266957824, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 21491 + }, + { + "epoch": 0.21492, + "grad_norm": 1.2328030980297424, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21492 + }, + { + "epoch": 0.21493, + "grad_norm": 1.4937031281531943, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 21493 + }, + { + "epoch": 0.21494, + "grad_norm": 1.3037103845090852, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 21494 + }, + { + "epoch": 0.21495, + "grad_norm": 1.5194341136682965, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21495 + }, + { + "epoch": 0.21496, + "grad_norm": 1.1070218887949796, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 21496 + }, + { + "epoch": 0.21497, + "grad_norm": 1.6035447800632763, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21497 + }, + { + "epoch": 0.21498, + "grad_norm": 1.659426705088667, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 21498 + }, + { + "epoch": 0.21499, + "grad_norm": 1.2973210627167238, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 21499 + }, + { + "epoch": 0.215, + "grad_norm": 1.1729428250012608, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 21500 + }, + { + "epoch": 0.21501, + "grad_norm": 1.4944754198342163, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 21501 + }, + { + "epoch": 0.21502, + "grad_norm": 1.183193279992519, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 21502 + }, + { + "epoch": 0.21503, + "grad_norm": 1.4494391545255134, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 21503 + }, + { + "epoch": 0.21504, + "grad_norm": 1.2729605463634632, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 21504 + }, + { + "epoch": 0.21505, + "grad_norm": 1.5520676393873234, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 21505 + }, + { + "epoch": 0.21506, + "grad_norm": 1.1765614188953981, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 21506 + }, + { + "epoch": 0.21507, + "grad_norm": 1.4259282001743225, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 21507 + }, + { + "epoch": 0.21508, + "grad_norm": 1.3792415358934287, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21508 + }, + { + "epoch": 0.21509, + "grad_norm": 1.3407684710741041, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21509 + }, + { + "epoch": 0.2151, + "grad_norm": 1.486477340588729, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 21510 + }, + { + "epoch": 0.21511, + "grad_norm": 1.206644971654592, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 21511 + }, + { + "epoch": 0.21512, + "grad_norm": 1.3279590824284087, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 21512 + }, + { + "epoch": 0.21513, + "grad_norm": 1.4300521784318503, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 21513 + }, + { + "epoch": 0.21514, + "grad_norm": 1.5089884939791547, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 21514 + }, + { + "epoch": 0.21515, + "grad_norm": 1.2972049813786273, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 21515 + }, + { + "epoch": 0.21516, + "grad_norm": 1.4342025494136799, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 21516 + }, + { + "epoch": 0.21517, + "grad_norm": 1.3277788729791116, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 21517 + }, + { + "epoch": 0.21518, + "grad_norm": 1.473741890578428, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 21518 + }, + { + "epoch": 0.21519, + "grad_norm": 1.3307439083792205, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 21519 + }, + { + "epoch": 0.2152, + "grad_norm": 1.345427902165864, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 21520 + }, + { + "epoch": 0.21521, + "grad_norm": 1.357117971587018, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21521 + }, + { + "epoch": 0.21522, + "grad_norm": 1.377549963602921, + "learning_rate": 0.003, + "loss": 4.049, + "step": 21522 + }, + { + "epoch": 0.21523, + "grad_norm": 1.5544818286222037, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 21523 + }, + { + "epoch": 0.21524, + "grad_norm": 1.1835800431228938, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21524 + }, + { + "epoch": 0.21525, + "grad_norm": 1.4275006507812762, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21525 + }, + { + "epoch": 0.21526, + "grad_norm": 1.2213527444171892, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 21526 + }, + { + "epoch": 0.21527, + "grad_norm": 1.372827038406382, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21527 + }, + { + "epoch": 0.21528, + "grad_norm": 1.3280158379319613, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 21528 + }, + { + "epoch": 0.21529, + "grad_norm": 1.4286487404610395, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21529 + }, + { + "epoch": 0.2153, + "grad_norm": 1.3629753919913346, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 21530 + }, + { + "epoch": 0.21531, + "grad_norm": 1.406135201149657, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 21531 + }, + { + "epoch": 0.21532, + "grad_norm": 1.1678754441716601, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21532 + }, + { + "epoch": 0.21533, + "grad_norm": 1.6796987843706, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21533 + }, + { + "epoch": 0.21534, + "grad_norm": 1.1519129898138445, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 21534 + }, + { + "epoch": 0.21535, + "grad_norm": 1.4082012232589747, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 21535 + }, + { + "epoch": 0.21536, + "grad_norm": 1.2124778273475414, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 21536 + }, + { + "epoch": 0.21537, + "grad_norm": 1.4000489102837348, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 21537 + }, + { + "epoch": 0.21538, + "grad_norm": 1.3097754581549077, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 21538 + }, + { + "epoch": 0.21539, + "grad_norm": 1.5649742201426702, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 21539 + }, + { + "epoch": 0.2154, + "grad_norm": 1.5444317447677847, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21540 + }, + { + "epoch": 0.21541, + "grad_norm": 1.40103127681567, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21541 + }, + { + "epoch": 0.21542, + "grad_norm": 1.3266392336491186, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 21542 + }, + { + "epoch": 0.21543, + "grad_norm": 1.5192252571422, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 21543 + }, + { + "epoch": 0.21544, + "grad_norm": 1.1035412439757208, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21544 + }, + { + "epoch": 0.21545, + "grad_norm": 1.3680108284675236, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 21545 + }, + { + "epoch": 0.21546, + "grad_norm": 1.2214251694454756, + "learning_rate": 0.003, + "loss": 4.043, + "step": 21546 + }, + { + "epoch": 0.21547, + "grad_norm": 1.1026129668183156, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 21547 + }, + { + "epoch": 0.21548, + "grad_norm": 1.5406280041130547, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 21548 + }, + { + "epoch": 0.21549, + "grad_norm": 1.1287473590446058, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 21549 + }, + { + "epoch": 0.2155, + "grad_norm": 1.4497523057720862, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21550 + }, + { + "epoch": 0.21551, + "grad_norm": 1.234154847974657, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 21551 + }, + { + "epoch": 0.21552, + "grad_norm": 1.3648088536447298, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 21552 + }, + { + "epoch": 0.21553, + "grad_norm": 1.452237570251351, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21553 + }, + { + "epoch": 0.21554, + "grad_norm": 1.3807680563664566, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 21554 + }, + { + "epoch": 0.21555, + "grad_norm": 1.3603024336725196, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 21555 + }, + { + "epoch": 0.21556, + "grad_norm": 1.2379192286946088, + "learning_rate": 0.003, + "loss": 4.057, + "step": 21556 + }, + { + "epoch": 0.21557, + "grad_norm": 1.3547139405076842, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21557 + }, + { + "epoch": 0.21558, + "grad_norm": 1.309257262851389, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 21558 + }, + { + "epoch": 0.21559, + "grad_norm": 1.3329537015540636, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 21559 + }, + { + "epoch": 0.2156, + "grad_norm": 1.6293792640850666, + "learning_rate": 0.003, + "loss": 4.059, + "step": 21560 + }, + { + "epoch": 0.21561, + "grad_norm": 1.2600053610320905, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 21561 + }, + { + "epoch": 0.21562, + "grad_norm": 1.419231945908033, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 21562 + }, + { + "epoch": 0.21563, + "grad_norm": 1.3226862447354382, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21563 + }, + { + "epoch": 0.21564, + "grad_norm": 1.339565374359852, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 21564 + }, + { + "epoch": 0.21565, + "grad_norm": 1.5140768921413992, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 21565 + }, + { + "epoch": 0.21566, + "grad_norm": 1.3598017958323683, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 21566 + }, + { + "epoch": 0.21567, + "grad_norm": 1.5207335031594, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 21567 + }, + { + "epoch": 0.21568, + "grad_norm": 1.3713189965898323, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21568 + }, + { + "epoch": 0.21569, + "grad_norm": 1.490099909551732, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 21569 + }, + { + "epoch": 0.2157, + "grad_norm": 1.2359100081394938, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 21570 + }, + { + "epoch": 0.21571, + "grad_norm": 1.593856573190763, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 21571 + }, + { + "epoch": 0.21572, + "grad_norm": 1.3689124722521349, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 21572 + }, + { + "epoch": 0.21573, + "grad_norm": 1.4359420290267144, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 21573 + }, + { + "epoch": 0.21574, + "grad_norm": 1.435224559342419, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 21574 + }, + { + "epoch": 0.21575, + "grad_norm": 1.3236478364948177, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 21575 + }, + { + "epoch": 0.21576, + "grad_norm": 1.3549316145333958, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 21576 + }, + { + "epoch": 0.21577, + "grad_norm": 1.1280127914712705, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21577 + }, + { + "epoch": 0.21578, + "grad_norm": 1.4183644314880084, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 21578 + }, + { + "epoch": 0.21579, + "grad_norm": 1.3812602074814249, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 21579 + }, + { + "epoch": 0.2158, + "grad_norm": 1.3967917065919164, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 21580 + }, + { + "epoch": 0.21581, + "grad_norm": 1.4160257381986296, + "learning_rate": 0.003, + "loss": 4.061, + "step": 21581 + }, + { + "epoch": 0.21582, + "grad_norm": 1.4757680075116717, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 21582 + }, + { + "epoch": 0.21583, + "grad_norm": 1.5358374207230423, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21583 + }, + { + "epoch": 0.21584, + "grad_norm": 1.5574011790728366, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 21584 + }, + { + "epoch": 0.21585, + "grad_norm": 1.1214992400217503, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 21585 + }, + { + "epoch": 0.21586, + "grad_norm": 1.6992726371130698, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 21586 + }, + { + "epoch": 0.21587, + "grad_norm": 1.1753772550111057, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 21587 + }, + { + "epoch": 0.21588, + "grad_norm": 1.3809315586928703, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 21588 + }, + { + "epoch": 0.21589, + "grad_norm": 1.2637905063856592, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21589 + }, + { + "epoch": 0.2159, + "grad_norm": 1.4473135353071278, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 21590 + }, + { + "epoch": 0.21591, + "grad_norm": 1.3323140250720638, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21591 + }, + { + "epoch": 0.21592, + "grad_norm": 1.422764327356364, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 21592 + }, + { + "epoch": 0.21593, + "grad_norm": 1.0475022562105123, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 21593 + }, + { + "epoch": 0.21594, + "grad_norm": 1.7374956068843377, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21594 + }, + { + "epoch": 0.21595, + "grad_norm": 1.3549828088244944, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 21595 + }, + { + "epoch": 0.21596, + "grad_norm": 1.3795446813878405, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 21596 + }, + { + "epoch": 0.21597, + "grad_norm": 1.3527697884769816, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 21597 + }, + { + "epoch": 0.21598, + "grad_norm": 1.443118614224528, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 21598 + }, + { + "epoch": 0.21599, + "grad_norm": 1.5789008710691697, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21599 + }, + { + "epoch": 0.216, + "grad_norm": 0.9971880523868474, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 21600 + }, + { + "epoch": 0.21601, + "grad_norm": 1.6959581327753726, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21601 + }, + { + "epoch": 0.21602, + "grad_norm": 1.1579673391638532, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 21602 + }, + { + "epoch": 0.21603, + "grad_norm": 1.430437164664146, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 21603 + }, + { + "epoch": 0.21604, + "grad_norm": 1.3046635963118043, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 21604 + }, + { + "epoch": 0.21605, + "grad_norm": 1.5287335485794018, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 21605 + }, + { + "epoch": 0.21606, + "grad_norm": 1.595175366047199, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 21606 + }, + { + "epoch": 0.21607, + "grad_norm": 1.2726329818538589, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 21607 + }, + { + "epoch": 0.21608, + "grad_norm": 1.4104958340286844, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21608 + }, + { + "epoch": 0.21609, + "grad_norm": 1.4477869337204952, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21609 + }, + { + "epoch": 0.2161, + "grad_norm": 1.1664696230932665, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21610 + }, + { + "epoch": 0.21611, + "grad_norm": 1.477087454913899, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 21611 + }, + { + "epoch": 0.21612, + "grad_norm": 1.3714088012117311, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 21612 + }, + { + "epoch": 0.21613, + "grad_norm": 1.4116579449268434, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 21613 + }, + { + "epoch": 0.21614, + "grad_norm": 1.3807817479659121, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 21614 + }, + { + "epoch": 0.21615, + "grad_norm": 1.3619293959221994, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 21615 + }, + { + "epoch": 0.21616, + "grad_norm": 1.3616990062115881, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21616 + }, + { + "epoch": 0.21617, + "grad_norm": 1.3587219297321471, + "learning_rate": 0.003, + "loss": 4.057, + "step": 21617 + }, + { + "epoch": 0.21618, + "grad_norm": 1.434026366402768, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 21618 + }, + { + "epoch": 0.21619, + "grad_norm": 1.3645624136467729, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 21619 + }, + { + "epoch": 0.2162, + "grad_norm": 1.3203288189146836, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21620 + }, + { + "epoch": 0.21621, + "grad_norm": 1.1929662451051506, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21621 + }, + { + "epoch": 0.21622, + "grad_norm": 1.3993197065686078, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 21622 + }, + { + "epoch": 0.21623, + "grad_norm": 1.237388037188455, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 21623 + }, + { + "epoch": 0.21624, + "grad_norm": 1.51925038792954, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 21624 + }, + { + "epoch": 0.21625, + "grad_norm": 1.297490591505803, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 21625 + }, + { + "epoch": 0.21626, + "grad_norm": 1.163973032156284, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 21626 + }, + { + "epoch": 0.21627, + "grad_norm": 1.2858400236240677, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 21627 + }, + { + "epoch": 0.21628, + "grad_norm": 1.144426443129442, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21628 + }, + { + "epoch": 0.21629, + "grad_norm": 1.8709954681847207, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21629 + }, + { + "epoch": 0.2163, + "grad_norm": 1.0437961753192782, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 21630 + }, + { + "epoch": 0.21631, + "grad_norm": 1.7225250200734556, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21631 + }, + { + "epoch": 0.21632, + "grad_norm": 1.2537767937813247, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 21632 + }, + { + "epoch": 0.21633, + "grad_norm": 1.3556325394402138, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 21633 + }, + { + "epoch": 0.21634, + "grad_norm": 1.2231110664289988, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21634 + }, + { + "epoch": 0.21635, + "grad_norm": 1.2845488504738347, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 21635 + }, + { + "epoch": 0.21636, + "grad_norm": 1.4193071665596668, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 21636 + }, + { + "epoch": 0.21637, + "grad_norm": 1.249477328209799, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 21637 + }, + { + "epoch": 0.21638, + "grad_norm": 1.5906557052491492, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 21638 + }, + { + "epoch": 0.21639, + "grad_norm": 1.3354898826295734, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 21639 + }, + { + "epoch": 0.2164, + "grad_norm": 1.2081118831734878, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 21640 + }, + { + "epoch": 0.21641, + "grad_norm": 1.4248452009874346, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 21641 + }, + { + "epoch": 0.21642, + "grad_norm": 1.1965572012589307, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 21642 + }, + { + "epoch": 0.21643, + "grad_norm": 1.564765311797686, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21643 + }, + { + "epoch": 0.21644, + "grad_norm": 1.21547877729175, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 21644 + }, + { + "epoch": 0.21645, + "grad_norm": 1.5468043681041297, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 21645 + }, + { + "epoch": 0.21646, + "grad_norm": 1.2890287381911754, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 21646 + }, + { + "epoch": 0.21647, + "grad_norm": 1.5472699496146884, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 21647 + }, + { + "epoch": 0.21648, + "grad_norm": 1.2753056416161248, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 21648 + }, + { + "epoch": 0.21649, + "grad_norm": 1.3225232253904111, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 21649 + }, + { + "epoch": 0.2165, + "grad_norm": 1.2158836680965754, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21650 + }, + { + "epoch": 0.21651, + "grad_norm": 1.2555133553073874, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 21651 + }, + { + "epoch": 0.21652, + "grad_norm": 1.4069853995644357, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21652 + }, + { + "epoch": 0.21653, + "grad_norm": 1.1577357999122015, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 21653 + }, + { + "epoch": 0.21654, + "grad_norm": 1.3273312106615434, + "learning_rate": 0.003, + "loss": 4.031, + "step": 21654 + }, + { + "epoch": 0.21655, + "grad_norm": 1.4568342365873415, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 21655 + }, + { + "epoch": 0.21656, + "grad_norm": 1.722474764085723, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21656 + }, + { + "epoch": 0.21657, + "grad_norm": 1.6570507191190962, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 21657 + }, + { + "epoch": 0.21658, + "grad_norm": 1.62144372785492, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21658 + }, + { + "epoch": 0.21659, + "grad_norm": 1.1890932469729452, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21659 + }, + { + "epoch": 0.2166, + "grad_norm": 1.4347718290022433, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 21660 + }, + { + "epoch": 0.21661, + "grad_norm": 1.2308231138075292, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 21661 + }, + { + "epoch": 0.21662, + "grad_norm": 1.5454350106985368, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 21662 + }, + { + "epoch": 0.21663, + "grad_norm": 1.046988003669505, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 21663 + }, + { + "epoch": 0.21664, + "grad_norm": 1.8212875291669854, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 21664 + }, + { + "epoch": 0.21665, + "grad_norm": 1.2163617415976309, + "learning_rate": 0.003, + "loss": 4.08, + "step": 21665 + }, + { + "epoch": 0.21666, + "grad_norm": 1.1954722007991765, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 21666 + }, + { + "epoch": 0.21667, + "grad_norm": 1.643670887358789, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21667 + }, + { + "epoch": 0.21668, + "grad_norm": 1.1819719137691378, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 21668 + }, + { + "epoch": 0.21669, + "grad_norm": 1.3883845641539347, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 21669 + }, + { + "epoch": 0.2167, + "grad_norm": 1.5432137729169304, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21670 + }, + { + "epoch": 0.21671, + "grad_norm": 1.0801449982316462, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 21671 + }, + { + "epoch": 0.21672, + "grad_norm": 1.4209952767389749, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 21672 + }, + { + "epoch": 0.21673, + "grad_norm": 1.5866298342325442, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 21673 + }, + { + "epoch": 0.21674, + "grad_norm": 1.2681859492203604, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 21674 + }, + { + "epoch": 0.21675, + "grad_norm": 1.5297042405494283, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 21675 + }, + { + "epoch": 0.21676, + "grad_norm": 1.3899495767717518, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 21676 + }, + { + "epoch": 0.21677, + "grad_norm": 1.25932941774649, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21677 + }, + { + "epoch": 0.21678, + "grad_norm": 1.221883870806337, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21678 + }, + { + "epoch": 0.21679, + "grad_norm": 1.3651250096100358, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 21679 + }, + { + "epoch": 0.2168, + "grad_norm": 1.1807414188959091, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 21680 + }, + { + "epoch": 0.21681, + "grad_norm": 1.5244914859217473, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 21681 + }, + { + "epoch": 0.21682, + "grad_norm": 1.1578109569084247, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 21682 + }, + { + "epoch": 0.21683, + "grad_norm": 1.6649032118906697, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 21683 + }, + { + "epoch": 0.21684, + "grad_norm": 1.2053851226183132, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21684 + }, + { + "epoch": 0.21685, + "grad_norm": 1.5142341801766461, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 21685 + }, + { + "epoch": 0.21686, + "grad_norm": 1.322200257992489, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 21686 + }, + { + "epoch": 0.21687, + "grad_norm": 1.3231026104256665, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 21687 + }, + { + "epoch": 0.21688, + "grad_norm": 1.606823587365431, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 21688 + }, + { + "epoch": 0.21689, + "grad_norm": 1.1984459130647076, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 21689 + }, + { + "epoch": 0.2169, + "grad_norm": 1.47229251005662, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 21690 + }, + { + "epoch": 0.21691, + "grad_norm": 1.4065269584718103, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21691 + }, + { + "epoch": 0.21692, + "grad_norm": 1.1773685317018276, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 21692 + }, + { + "epoch": 0.21693, + "grad_norm": 1.5216618999402516, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 21693 + }, + { + "epoch": 0.21694, + "grad_norm": 1.2676906916747757, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 21694 + }, + { + "epoch": 0.21695, + "grad_norm": 1.5011195436506621, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 21695 + }, + { + "epoch": 0.21696, + "grad_norm": 1.115549485955369, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 21696 + }, + { + "epoch": 0.21697, + "grad_norm": 1.8018502147911997, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21697 + }, + { + "epoch": 0.21698, + "grad_norm": 1.36356208110744, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 21698 + }, + { + "epoch": 0.21699, + "grad_norm": 1.5376830435927542, + "learning_rate": 0.003, + "loss": 4.068, + "step": 21699 + }, + { + "epoch": 0.217, + "grad_norm": 1.3219978376624038, + "learning_rate": 0.003, + "loss": 4.06, + "step": 21700 + }, + { + "epoch": 0.21701, + "grad_norm": 1.461548045741008, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 21701 + }, + { + "epoch": 0.21702, + "grad_norm": 1.2124060724442183, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 21702 + }, + { + "epoch": 0.21703, + "grad_norm": 1.4795227869985204, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 21703 + }, + { + "epoch": 0.21704, + "grad_norm": 1.3709728791071893, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21704 + }, + { + "epoch": 0.21705, + "grad_norm": 1.3277364546039956, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 21705 + }, + { + "epoch": 0.21706, + "grad_norm": 1.256696362963927, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21706 + }, + { + "epoch": 0.21707, + "grad_norm": 1.543516684423096, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 21707 + }, + { + "epoch": 0.21708, + "grad_norm": 1.2960958415082264, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 21708 + }, + { + "epoch": 0.21709, + "grad_norm": 1.3911681816473196, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 21709 + }, + { + "epoch": 0.2171, + "grad_norm": 1.2940636888664225, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 21710 + }, + { + "epoch": 0.21711, + "grad_norm": 1.4519873731834203, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 21711 + }, + { + "epoch": 0.21712, + "grad_norm": 1.3964960397836088, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 21712 + }, + { + "epoch": 0.21713, + "grad_norm": 1.1396648944563743, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21713 + }, + { + "epoch": 0.21714, + "grad_norm": 1.2894495531812444, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 21714 + }, + { + "epoch": 0.21715, + "grad_norm": 1.3753294352282759, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 21715 + }, + { + "epoch": 0.21716, + "grad_norm": 1.2255268338041903, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 21716 + }, + { + "epoch": 0.21717, + "grad_norm": 1.6193976605963485, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 21717 + }, + { + "epoch": 0.21718, + "grad_norm": 1.1317837812203877, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 21718 + }, + { + "epoch": 0.21719, + "grad_norm": 1.5081856943937748, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 21719 + }, + { + "epoch": 0.2172, + "grad_norm": 1.3034260129123711, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21720 + }, + { + "epoch": 0.21721, + "grad_norm": 1.5751825772886663, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21721 + }, + { + "epoch": 0.21722, + "grad_norm": 1.187587421016948, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 21722 + }, + { + "epoch": 0.21723, + "grad_norm": 1.4427287322645799, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21723 + }, + { + "epoch": 0.21724, + "grad_norm": 1.2827694304521151, + "learning_rate": 0.003, + "loss": 4.066, + "step": 21724 + }, + { + "epoch": 0.21725, + "grad_norm": 1.2441120564389594, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 21725 + }, + { + "epoch": 0.21726, + "grad_norm": 1.4617663722466212, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 21726 + }, + { + "epoch": 0.21727, + "grad_norm": 1.1876164644093088, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 21727 + }, + { + "epoch": 0.21728, + "grad_norm": 1.5713004814183846, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 21728 + }, + { + "epoch": 0.21729, + "grad_norm": 1.2712912632821525, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 21729 + }, + { + "epoch": 0.2173, + "grad_norm": 1.5295052971950158, + "learning_rate": 0.003, + "loss": 4.042, + "step": 21730 + }, + { + "epoch": 0.21731, + "grad_norm": 1.2662831509869186, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 21731 + }, + { + "epoch": 0.21732, + "grad_norm": 1.44190359333691, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 21732 + }, + { + "epoch": 0.21733, + "grad_norm": 1.352985815465497, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 21733 + }, + { + "epoch": 0.21734, + "grad_norm": 1.273980705478746, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 21734 + }, + { + "epoch": 0.21735, + "grad_norm": 1.3687623400104987, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21735 + }, + { + "epoch": 0.21736, + "grad_norm": 1.137617573178497, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21736 + }, + { + "epoch": 0.21737, + "grad_norm": 1.5544173363330809, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 21737 + }, + { + "epoch": 0.21738, + "grad_norm": 1.1422182207113591, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21738 + }, + { + "epoch": 0.21739, + "grad_norm": 1.8066904455344126, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 21739 + }, + { + "epoch": 0.2174, + "grad_norm": 1.1671789252959586, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 21740 + }, + { + "epoch": 0.21741, + "grad_norm": 1.4684617417972043, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 21741 + }, + { + "epoch": 0.21742, + "grad_norm": 1.399942284573787, + "learning_rate": 0.003, + "loss": 4.08, + "step": 21742 + }, + { + "epoch": 0.21743, + "grad_norm": 1.16995899577208, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 21743 + }, + { + "epoch": 0.21744, + "grad_norm": 1.3906569721727808, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 21744 + }, + { + "epoch": 0.21745, + "grad_norm": 1.477492393749993, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 21745 + }, + { + "epoch": 0.21746, + "grad_norm": 1.2952139975004873, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 21746 + }, + { + "epoch": 0.21747, + "grad_norm": 1.3901444632726925, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 21747 + }, + { + "epoch": 0.21748, + "grad_norm": 1.1549286366297145, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21748 + }, + { + "epoch": 0.21749, + "grad_norm": 1.7083361886105854, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 21749 + }, + { + "epoch": 0.2175, + "grad_norm": 1.2551477661927604, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 21750 + }, + { + "epoch": 0.21751, + "grad_norm": 1.3079787486564436, + "learning_rate": 0.003, + "loss": 4.024, + "step": 21751 + }, + { + "epoch": 0.21752, + "grad_norm": 1.5962397143865705, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 21752 + }, + { + "epoch": 0.21753, + "grad_norm": 1.596429040132191, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21753 + }, + { + "epoch": 0.21754, + "grad_norm": 1.2113879246711468, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 21754 + }, + { + "epoch": 0.21755, + "grad_norm": 1.6332028029615502, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 21755 + }, + { + "epoch": 0.21756, + "grad_norm": 1.6463052448187578, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 21756 + }, + { + "epoch": 0.21757, + "grad_norm": 1.137118479217971, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21757 + }, + { + "epoch": 0.21758, + "grad_norm": 1.3635544244652056, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21758 + }, + { + "epoch": 0.21759, + "grad_norm": 1.2322221825771982, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 21759 + }, + { + "epoch": 0.2176, + "grad_norm": 1.4530300513984373, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 21760 + }, + { + "epoch": 0.21761, + "grad_norm": 1.3612567174320283, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 21761 + }, + { + "epoch": 0.21762, + "grad_norm": 1.4689818427730725, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21762 + }, + { + "epoch": 0.21763, + "grad_norm": 1.3369737458324107, + "learning_rate": 0.003, + "loss": 4.064, + "step": 21763 + }, + { + "epoch": 0.21764, + "grad_norm": 1.3175865171698724, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 21764 + }, + { + "epoch": 0.21765, + "grad_norm": 1.246510114819442, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 21765 + }, + { + "epoch": 0.21766, + "grad_norm": 1.2665709182460367, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 21766 + }, + { + "epoch": 0.21767, + "grad_norm": 1.2461678747363165, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21767 + }, + { + "epoch": 0.21768, + "grad_norm": 1.277964271546957, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 21768 + }, + { + "epoch": 0.21769, + "grad_norm": 1.4382325065553294, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 21769 + }, + { + "epoch": 0.2177, + "grad_norm": 1.3483673137453749, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 21770 + }, + { + "epoch": 0.21771, + "grad_norm": 1.5007492645327565, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21771 + }, + { + "epoch": 0.21772, + "grad_norm": 1.4906521742284367, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21772 + }, + { + "epoch": 0.21773, + "grad_norm": 1.3829906249211739, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 21773 + }, + { + "epoch": 0.21774, + "grad_norm": 1.3756208190347674, + "learning_rate": 0.003, + "loss": 4.046, + "step": 21774 + }, + { + "epoch": 0.21775, + "grad_norm": 1.187419765112109, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 21775 + }, + { + "epoch": 0.21776, + "grad_norm": 1.2847796938167904, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 21776 + }, + { + "epoch": 0.21777, + "grad_norm": 1.1787148813225765, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21777 + }, + { + "epoch": 0.21778, + "grad_norm": 1.5384271887768861, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 21778 + }, + { + "epoch": 0.21779, + "grad_norm": 1.5211241536809952, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21779 + }, + { + "epoch": 0.2178, + "grad_norm": 1.3126100308265978, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 21780 + }, + { + "epoch": 0.21781, + "grad_norm": 1.3140837416202098, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21781 + }, + { + "epoch": 0.21782, + "grad_norm": 1.2506337097241051, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 21782 + }, + { + "epoch": 0.21783, + "grad_norm": 1.183549010980319, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21783 + }, + { + "epoch": 0.21784, + "grad_norm": 1.2120193027817996, + "learning_rate": 0.003, + "loss": 4.042, + "step": 21784 + }, + { + "epoch": 0.21785, + "grad_norm": 1.4705008753482647, + "learning_rate": 0.003, + "loss": 4.052, + "step": 21785 + }, + { + "epoch": 0.21786, + "grad_norm": 1.3037580962126534, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 21786 + }, + { + "epoch": 0.21787, + "grad_norm": 1.5873941885421359, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 21787 + }, + { + "epoch": 0.21788, + "grad_norm": 1.293752445907499, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 21788 + }, + { + "epoch": 0.21789, + "grad_norm": 1.6278839822338038, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 21789 + }, + { + "epoch": 0.2179, + "grad_norm": 1.2487031038570793, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 21790 + }, + { + "epoch": 0.21791, + "grad_norm": 1.497883497136813, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 21791 + }, + { + "epoch": 0.21792, + "grad_norm": 1.5202302105632721, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21792 + }, + { + "epoch": 0.21793, + "grad_norm": 1.2269791447971383, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 21793 + }, + { + "epoch": 0.21794, + "grad_norm": 1.449344551808254, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 21794 + }, + { + "epoch": 0.21795, + "grad_norm": 1.9388279199050673, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 21795 + }, + { + "epoch": 0.21796, + "grad_norm": 0.8841590120868, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21796 + }, + { + "epoch": 0.21797, + "grad_norm": 1.4292538185520196, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 21797 + }, + { + "epoch": 0.21798, + "grad_norm": 1.3789664901476326, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21798 + }, + { + "epoch": 0.21799, + "grad_norm": 1.3927764454629348, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 21799 + }, + { + "epoch": 0.218, + "grad_norm": 1.622972417186285, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21800 + }, + { + "epoch": 0.21801, + "grad_norm": 1.2911279890778193, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 21801 + }, + { + "epoch": 0.21802, + "grad_norm": 1.5380703604963368, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 21802 + }, + { + "epoch": 0.21803, + "grad_norm": 1.1990859205719844, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 21803 + }, + { + "epoch": 0.21804, + "grad_norm": 1.3908968327755922, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21804 + }, + { + "epoch": 0.21805, + "grad_norm": 1.6093157797807323, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 21805 + }, + { + "epoch": 0.21806, + "grad_norm": 1.082573623518282, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 21806 + }, + { + "epoch": 0.21807, + "grad_norm": 1.3874620059290594, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 21807 + }, + { + "epoch": 0.21808, + "grad_norm": 1.2340360635483378, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21808 + }, + { + "epoch": 0.21809, + "grad_norm": 1.48196030694557, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 21809 + }, + { + "epoch": 0.2181, + "grad_norm": 1.1829455082845057, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 21810 + }, + { + "epoch": 0.21811, + "grad_norm": 1.3656560031304386, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 21811 + }, + { + "epoch": 0.21812, + "grad_norm": 1.2827909957331247, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 21812 + }, + { + "epoch": 0.21813, + "grad_norm": 1.486494039463018, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 21813 + }, + { + "epoch": 0.21814, + "grad_norm": 1.3759143268638585, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21814 + }, + { + "epoch": 0.21815, + "grad_norm": 1.2142174582604783, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21815 + }, + { + "epoch": 0.21816, + "grad_norm": 1.4197772920590874, + "learning_rate": 0.003, + "loss": 4.068, + "step": 21816 + }, + { + "epoch": 0.21817, + "grad_norm": 1.323393658081733, + "learning_rate": 0.003, + "loss": 4.038, + "step": 21817 + }, + { + "epoch": 0.21818, + "grad_norm": 1.5159685668028264, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 21818 + }, + { + "epoch": 0.21819, + "grad_norm": 1.51023708455265, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 21819 + }, + { + "epoch": 0.2182, + "grad_norm": 1.2629682522797707, + "learning_rate": 0.003, + "loss": 4.056, + "step": 21820 + }, + { + "epoch": 0.21821, + "grad_norm": 1.236949904346332, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 21821 + }, + { + "epoch": 0.21822, + "grad_norm": 1.4900879106882725, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 21822 + }, + { + "epoch": 0.21823, + "grad_norm": 1.3545475172053465, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 21823 + }, + { + "epoch": 0.21824, + "grad_norm": 1.4291612969027583, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 21824 + }, + { + "epoch": 0.21825, + "grad_norm": 1.2378582878097866, + "learning_rate": 0.003, + "loss": 4.065, + "step": 21825 + }, + { + "epoch": 0.21826, + "grad_norm": 1.56258273065671, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21826 + }, + { + "epoch": 0.21827, + "grad_norm": 1.486960382792435, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 21827 + }, + { + "epoch": 0.21828, + "grad_norm": 1.2067952770408727, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 21828 + }, + { + "epoch": 0.21829, + "grad_norm": 1.3937543985084526, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 21829 + }, + { + "epoch": 0.2183, + "grad_norm": 1.3803056056357, + "learning_rate": 0.003, + "loss": 4.059, + "step": 21830 + }, + { + "epoch": 0.21831, + "grad_norm": 1.315611588775608, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 21831 + }, + { + "epoch": 0.21832, + "grad_norm": 1.4144492685364904, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 21832 + }, + { + "epoch": 0.21833, + "grad_norm": 1.2024873578795783, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 21833 + }, + { + "epoch": 0.21834, + "grad_norm": 1.2345346730467635, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 21834 + }, + { + "epoch": 0.21835, + "grad_norm": 1.590026520429568, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21835 + }, + { + "epoch": 0.21836, + "grad_norm": 1.4489599804000857, + "learning_rate": 0.003, + "loss": 4.054, + "step": 21836 + }, + { + "epoch": 0.21837, + "grad_norm": 1.3948991946463194, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21837 + }, + { + "epoch": 0.21838, + "grad_norm": 1.0839190492419373, + "learning_rate": 0.003, + "loss": 4.062, + "step": 21838 + }, + { + "epoch": 0.21839, + "grad_norm": 1.5310246196483528, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 21839 + }, + { + "epoch": 0.2184, + "grad_norm": 1.2972416568653322, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 21840 + }, + { + "epoch": 0.21841, + "grad_norm": 1.3518609380233026, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 21841 + }, + { + "epoch": 0.21842, + "grad_norm": 1.239134396927293, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 21842 + }, + { + "epoch": 0.21843, + "grad_norm": 1.4939191795927445, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 21843 + }, + { + "epoch": 0.21844, + "grad_norm": 1.2542783730907288, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 21844 + }, + { + "epoch": 0.21845, + "grad_norm": 1.3523336953828662, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 21845 + }, + { + "epoch": 0.21846, + "grad_norm": 1.3483181772009003, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 21846 + }, + { + "epoch": 0.21847, + "grad_norm": 1.4341188224770438, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21847 + }, + { + "epoch": 0.21848, + "grad_norm": 1.2701861347518604, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 21848 + }, + { + "epoch": 0.21849, + "grad_norm": 1.5735458276270475, + "learning_rate": 0.003, + "loss": 4.034, + "step": 21849 + }, + { + "epoch": 0.2185, + "grad_norm": 1.2977797392794104, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21850 + }, + { + "epoch": 0.21851, + "grad_norm": 1.3993707258427626, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 21851 + }, + { + "epoch": 0.21852, + "grad_norm": 1.3400140972805528, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 21852 + }, + { + "epoch": 0.21853, + "grad_norm": 1.3187261982616778, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 21853 + }, + { + "epoch": 0.21854, + "grad_norm": 1.373648276029539, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21854 + }, + { + "epoch": 0.21855, + "grad_norm": 1.2314813262996391, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 21855 + }, + { + "epoch": 0.21856, + "grad_norm": 1.4455662452529494, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 21856 + }, + { + "epoch": 0.21857, + "grad_norm": 1.1081284462682128, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 21857 + }, + { + "epoch": 0.21858, + "grad_norm": 1.6186299276515583, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 21858 + }, + { + "epoch": 0.21859, + "grad_norm": 1.342266584936963, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 21859 + }, + { + "epoch": 0.2186, + "grad_norm": 1.6222494766013495, + "learning_rate": 0.003, + "loss": 4.047, + "step": 21860 + }, + { + "epoch": 0.21861, + "grad_norm": 1.2756726672574457, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 21861 + }, + { + "epoch": 0.21862, + "grad_norm": 1.3800762617091014, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 21862 + }, + { + "epoch": 0.21863, + "grad_norm": 1.1219644793746457, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 21863 + }, + { + "epoch": 0.21864, + "grad_norm": 1.4807516433685504, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 21864 + }, + { + "epoch": 0.21865, + "grad_norm": 1.153569375518149, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 21865 + }, + { + "epoch": 0.21866, + "grad_norm": 1.3913084942040324, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 21866 + }, + { + "epoch": 0.21867, + "grad_norm": 1.3975902554296735, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 21867 + }, + { + "epoch": 0.21868, + "grad_norm": 1.6111752800832773, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 21868 + }, + { + "epoch": 0.21869, + "grad_norm": 1.3397324373724762, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 21869 + }, + { + "epoch": 0.2187, + "grad_norm": 1.5265125558745145, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 21870 + }, + { + "epoch": 0.21871, + "grad_norm": 1.5444100630413469, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21871 + }, + { + "epoch": 0.21872, + "grad_norm": 1.4136571138250893, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 21872 + }, + { + "epoch": 0.21873, + "grad_norm": 1.492531221152572, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 21873 + }, + { + "epoch": 0.21874, + "grad_norm": 1.3801935048510277, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 21874 + }, + { + "epoch": 0.21875, + "grad_norm": 1.3464300138449798, + "learning_rate": 0.003, + "loss": 4.053, + "step": 21875 + }, + { + "epoch": 0.21876, + "grad_norm": 1.3230002208921514, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 21876 + }, + { + "epoch": 0.21877, + "grad_norm": 1.3184776967515563, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 21877 + }, + { + "epoch": 0.21878, + "grad_norm": 1.3243604189039329, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 21878 + }, + { + "epoch": 0.21879, + "grad_norm": 1.350716120331371, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 21879 + }, + { + "epoch": 0.2188, + "grad_norm": 1.1887655380988411, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 21880 + }, + { + "epoch": 0.21881, + "grad_norm": 1.4732317039088851, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 21881 + }, + { + "epoch": 0.21882, + "grad_norm": 1.2481675039437126, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 21882 + }, + { + "epoch": 0.21883, + "grad_norm": 1.682173239924703, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 21883 + }, + { + "epoch": 0.21884, + "grad_norm": 1.408843425940714, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 21884 + }, + { + "epoch": 0.21885, + "grad_norm": 1.3991787530439368, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 21885 + }, + { + "epoch": 0.21886, + "grad_norm": 1.351255708264895, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 21886 + }, + { + "epoch": 0.21887, + "grad_norm": 1.3910581333259997, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 21887 + }, + { + "epoch": 0.21888, + "grad_norm": 1.3253623213565662, + "learning_rate": 0.003, + "loss": 4.035, + "step": 21888 + }, + { + "epoch": 0.21889, + "grad_norm": 1.3207359549564404, + "learning_rate": 0.003, + "loss": 4.041, + "step": 21889 + }, + { + "epoch": 0.2189, + "grad_norm": 1.3103691672670739, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 21890 + }, + { + "epoch": 0.21891, + "grad_norm": 1.4143579120442842, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 21891 + }, + { + "epoch": 0.21892, + "grad_norm": 1.3769560345949032, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 21892 + }, + { + "epoch": 0.21893, + "grad_norm": 1.6479910150324262, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21893 + }, + { + "epoch": 0.21894, + "grad_norm": 1.477498345410568, + "learning_rate": 0.003, + "loss": 4.02, + "step": 21894 + }, + { + "epoch": 0.21895, + "grad_norm": 1.6010316292528048, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 21895 + }, + { + "epoch": 0.21896, + "grad_norm": 1.0736524070408766, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 21896 + }, + { + "epoch": 0.21897, + "grad_norm": 1.5513404345546, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21897 + }, + { + "epoch": 0.21898, + "grad_norm": 1.4060130819583434, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21898 + }, + { + "epoch": 0.21899, + "grad_norm": 1.242919541121017, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21899 + }, + { + "epoch": 0.219, + "grad_norm": 1.419710336740735, + "learning_rate": 0.003, + "loss": 4.067, + "step": 21900 + }, + { + "epoch": 0.21901, + "grad_norm": 1.1142225146782796, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 21901 + }, + { + "epoch": 0.21902, + "grad_norm": 1.515604579657156, + "learning_rate": 0.003, + "loss": 4.049, + "step": 21902 + }, + { + "epoch": 0.21903, + "grad_norm": 1.383479495850097, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 21903 + }, + { + "epoch": 0.21904, + "grad_norm": 1.3652068263044148, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 21904 + }, + { + "epoch": 0.21905, + "grad_norm": 1.4900790425030694, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 21905 + }, + { + "epoch": 0.21906, + "grad_norm": 1.193768677638813, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 21906 + }, + { + "epoch": 0.21907, + "grad_norm": 1.3165246410732574, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 21907 + }, + { + "epoch": 0.21908, + "grad_norm": 1.2852506288954528, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 21908 + }, + { + "epoch": 0.21909, + "grad_norm": 1.4322754139851814, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 21909 + }, + { + "epoch": 0.2191, + "grad_norm": 1.1934054775387493, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 21910 + }, + { + "epoch": 0.21911, + "grad_norm": 1.427492098697218, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 21911 + }, + { + "epoch": 0.21912, + "grad_norm": 1.3050800021088695, + "learning_rate": 0.003, + "loss": 4.051, + "step": 21912 + }, + { + "epoch": 0.21913, + "grad_norm": 1.5633344870549826, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 21913 + }, + { + "epoch": 0.21914, + "grad_norm": 1.4370186550446122, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 21914 + }, + { + "epoch": 0.21915, + "grad_norm": 1.5627544102364597, + "learning_rate": 0.003, + "loss": 4.086, + "step": 21915 + }, + { + "epoch": 0.21916, + "grad_norm": 0.9544544959214196, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21916 + }, + { + "epoch": 0.21917, + "grad_norm": 1.3747323703717953, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 21917 + }, + { + "epoch": 0.21918, + "grad_norm": 1.356051125720396, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 21918 + }, + { + "epoch": 0.21919, + "grad_norm": 1.2647930960633598, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 21919 + }, + { + "epoch": 0.2192, + "grad_norm": 1.5382002387844138, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 21920 + }, + { + "epoch": 0.21921, + "grad_norm": 1.318015693485354, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 21921 + }, + { + "epoch": 0.21922, + "grad_norm": 1.5612089494631403, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 21922 + }, + { + "epoch": 0.21923, + "grad_norm": 1.2109777532096635, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 21923 + }, + { + "epoch": 0.21924, + "grad_norm": 1.5172583837326865, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 21924 + }, + { + "epoch": 0.21925, + "grad_norm": 1.2389141294335106, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 21925 + }, + { + "epoch": 0.21926, + "grad_norm": 1.6481160695458348, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 21926 + }, + { + "epoch": 0.21927, + "grad_norm": 1.1259072392279144, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 21927 + }, + { + "epoch": 0.21928, + "grad_norm": 1.523706799834442, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 21928 + }, + { + "epoch": 0.21929, + "grad_norm": 1.3838361375856523, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 21929 + }, + { + "epoch": 0.2193, + "grad_norm": 1.3183290648455395, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 21930 + }, + { + "epoch": 0.21931, + "grad_norm": 1.3157787075022724, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 21931 + }, + { + "epoch": 0.21932, + "grad_norm": 1.361708506842995, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 21932 + }, + { + "epoch": 0.21933, + "grad_norm": 1.5617791748578622, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 21933 + }, + { + "epoch": 0.21934, + "grad_norm": 1.2795490857302654, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 21934 + }, + { + "epoch": 0.21935, + "grad_norm": 1.4973845919237996, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 21935 + }, + { + "epoch": 0.21936, + "grad_norm": 1.2442587511940837, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 21936 + }, + { + "epoch": 0.21937, + "grad_norm": 1.62905009543493, + "learning_rate": 0.003, + "loss": 4.04, + "step": 21937 + }, + { + "epoch": 0.21938, + "grad_norm": 1.3083598097255102, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 21938 + }, + { + "epoch": 0.21939, + "grad_norm": 1.343908636553416, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 21939 + }, + { + "epoch": 0.2194, + "grad_norm": 1.2021113121993336, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 21940 + }, + { + "epoch": 0.21941, + "grad_norm": 1.491497563459595, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 21941 + }, + { + "epoch": 0.21942, + "grad_norm": 1.1311382475212675, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 21942 + }, + { + "epoch": 0.21943, + "grad_norm": 1.4006684967081473, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 21943 + }, + { + "epoch": 0.21944, + "grad_norm": 1.2935080502234235, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 21944 + }, + { + "epoch": 0.21945, + "grad_norm": 1.3453590992792903, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 21945 + }, + { + "epoch": 0.21946, + "grad_norm": 1.340367401947809, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 21946 + }, + { + "epoch": 0.21947, + "grad_norm": 1.3381634369619626, + "learning_rate": 0.003, + "loss": 4.074, + "step": 21947 + }, + { + "epoch": 0.21948, + "grad_norm": 1.6031288553416352, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21948 + }, + { + "epoch": 0.21949, + "grad_norm": 1.236788110522683, + "learning_rate": 0.003, + "loss": 4.057, + "step": 21949 + }, + { + "epoch": 0.2195, + "grad_norm": 1.3961868195729772, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 21950 + }, + { + "epoch": 0.21951, + "grad_norm": 1.3454923024665244, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 21951 + }, + { + "epoch": 0.21952, + "grad_norm": 1.1612319105690718, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21952 + }, + { + "epoch": 0.21953, + "grad_norm": 1.3979619770700362, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 21953 + }, + { + "epoch": 0.21954, + "grad_norm": 1.3488336642416268, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 21954 + }, + { + "epoch": 0.21955, + "grad_norm": 1.5890297765100796, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 21955 + }, + { + "epoch": 0.21956, + "grad_norm": 1.3013044515226226, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 21956 + }, + { + "epoch": 0.21957, + "grad_norm": 1.3205711846707797, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 21957 + }, + { + "epoch": 0.21958, + "grad_norm": 1.3960458913269216, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 21958 + }, + { + "epoch": 0.21959, + "grad_norm": 1.3410994689448847, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 21959 + }, + { + "epoch": 0.2196, + "grad_norm": 1.3622950495825286, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 21960 + }, + { + "epoch": 0.21961, + "grad_norm": 1.2439812562161565, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 21961 + }, + { + "epoch": 0.21962, + "grad_norm": 1.6495809689740977, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21962 + }, + { + "epoch": 0.21963, + "grad_norm": 1.1075823228719488, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 21963 + }, + { + "epoch": 0.21964, + "grad_norm": 1.40324106022888, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 21964 + }, + { + "epoch": 0.21965, + "grad_norm": 1.2884392999720742, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 21965 + }, + { + "epoch": 0.21966, + "grad_norm": 1.3539642794047264, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 21966 + }, + { + "epoch": 0.21967, + "grad_norm": 1.4955974362173503, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 21967 + }, + { + "epoch": 0.21968, + "grad_norm": 1.5116388453685434, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 21968 + }, + { + "epoch": 0.21969, + "grad_norm": 1.3549104667418375, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 21969 + }, + { + "epoch": 0.2197, + "grad_norm": 1.4079075297455776, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 21970 + }, + { + "epoch": 0.21971, + "grad_norm": 1.453401454507369, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 21971 + }, + { + "epoch": 0.21972, + "grad_norm": 1.4250613175756226, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 21972 + }, + { + "epoch": 0.21973, + "grad_norm": 1.3831072647546812, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 21973 + }, + { + "epoch": 0.21974, + "grad_norm": 1.4164304746590137, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 21974 + }, + { + "epoch": 0.21975, + "grad_norm": 1.3363938348699675, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 21975 + }, + { + "epoch": 0.21976, + "grad_norm": 1.2854685734366242, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21976 + }, + { + "epoch": 0.21977, + "grad_norm": 1.5134779219680778, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 21977 + }, + { + "epoch": 0.21978, + "grad_norm": 1.4153170955539442, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 21978 + }, + { + "epoch": 0.21979, + "grad_norm": 1.6458748964455412, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 21979 + }, + { + "epoch": 0.2198, + "grad_norm": 1.2724509650276048, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 21980 + }, + { + "epoch": 0.21981, + "grad_norm": 1.4023100840793203, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 21981 + }, + { + "epoch": 0.21982, + "grad_norm": 1.3912212765186005, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 21982 + }, + { + "epoch": 0.21983, + "grad_norm": 1.427322774860121, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 21983 + }, + { + "epoch": 0.21984, + "grad_norm": 1.3483426313043931, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 21984 + }, + { + "epoch": 0.21985, + "grad_norm": 1.3793512180367091, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 21985 + }, + { + "epoch": 0.21986, + "grad_norm": 1.1907907204772645, + "learning_rate": 0.003, + "loss": 4.066, + "step": 21986 + }, + { + "epoch": 0.21987, + "grad_norm": 1.1974306360117373, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 21987 + }, + { + "epoch": 0.21988, + "grad_norm": 1.470162481516743, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21988 + }, + { + "epoch": 0.21989, + "grad_norm": 1.1308975327673492, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 21989 + }, + { + "epoch": 0.2199, + "grad_norm": 1.5982872045018905, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 21990 + }, + { + "epoch": 0.21991, + "grad_norm": 1.3511960661468174, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 21991 + }, + { + "epoch": 0.21992, + "grad_norm": 1.2009597118512085, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21992 + }, + { + "epoch": 0.21993, + "grad_norm": 1.3066035698767409, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 21993 + }, + { + "epoch": 0.21994, + "grad_norm": 1.4236477958372746, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 21994 + }, + { + "epoch": 0.21995, + "grad_norm": 1.4189188999668838, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 21995 + }, + { + "epoch": 0.21996, + "grad_norm": 1.5128501434641455, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 21996 + }, + { + "epoch": 0.21997, + "grad_norm": 1.172286039870528, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 21997 + }, + { + "epoch": 0.21998, + "grad_norm": 1.6729050834879595, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 21998 + }, + { + "epoch": 0.21999, + "grad_norm": 0.9473969066854019, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 21999 + }, + { + "epoch": 0.22, + "grad_norm": 1.3857757584959134, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22000 + }, + { + "epoch": 0.22001, + "grad_norm": 1.3242205618236835, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 22001 + }, + { + "epoch": 0.22002, + "grad_norm": 1.4698493681572296, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 22002 + }, + { + "epoch": 0.22003, + "grad_norm": 1.2867339428696243, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 22003 + }, + { + "epoch": 0.22004, + "grad_norm": 1.3454886959099548, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 22004 + }, + { + "epoch": 0.22005, + "grad_norm": 1.253041919375533, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 22005 + }, + { + "epoch": 0.22006, + "grad_norm": 1.7307220189692687, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 22006 + }, + { + "epoch": 0.22007, + "grad_norm": 1.13504168347129, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22007 + }, + { + "epoch": 0.22008, + "grad_norm": 1.8296563772979586, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22008 + }, + { + "epoch": 0.22009, + "grad_norm": 1.1741589978484182, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 22009 + }, + { + "epoch": 0.2201, + "grad_norm": 1.5381026497279788, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22010 + }, + { + "epoch": 0.22011, + "grad_norm": 1.1094107114377345, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22011 + }, + { + "epoch": 0.22012, + "grad_norm": 1.442829519692605, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 22012 + }, + { + "epoch": 0.22013, + "grad_norm": 1.631016202833152, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22013 + }, + { + "epoch": 0.22014, + "grad_norm": 1.6106086231447079, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 22014 + }, + { + "epoch": 0.22015, + "grad_norm": 1.1814828776053918, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 22015 + }, + { + "epoch": 0.22016, + "grad_norm": 1.578015763537259, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 22016 + }, + { + "epoch": 0.22017, + "grad_norm": 1.3228701853660656, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 22017 + }, + { + "epoch": 0.22018, + "grad_norm": 1.3954939622633813, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 22018 + }, + { + "epoch": 0.22019, + "grad_norm": 1.4848139680187202, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 22019 + }, + { + "epoch": 0.2202, + "grad_norm": 1.1779000454188744, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 22020 + }, + { + "epoch": 0.22021, + "grad_norm": 1.359113863980639, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 22021 + }, + { + "epoch": 0.22022, + "grad_norm": 1.6020148930131255, + "learning_rate": 0.003, + "loss": 4.049, + "step": 22022 + }, + { + "epoch": 0.22023, + "grad_norm": 1.080821702668198, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 22023 + }, + { + "epoch": 0.22024, + "grad_norm": 1.3793835762293125, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 22024 + }, + { + "epoch": 0.22025, + "grad_norm": 1.3996018199293805, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 22025 + }, + { + "epoch": 0.22026, + "grad_norm": 1.4541457512077411, + "learning_rate": 0.003, + "loss": 4.064, + "step": 22026 + }, + { + "epoch": 0.22027, + "grad_norm": 1.5036697031980546, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 22027 + }, + { + "epoch": 0.22028, + "grad_norm": 1.3927468977215929, + "learning_rate": 0.003, + "loss": 4.083, + "step": 22028 + }, + { + "epoch": 0.22029, + "grad_norm": 1.1116074644616787, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22029 + }, + { + "epoch": 0.2203, + "grad_norm": 1.4137392822592407, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 22030 + }, + { + "epoch": 0.22031, + "grad_norm": 1.433020747516023, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 22031 + }, + { + "epoch": 0.22032, + "grad_norm": 1.191612135290802, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 22032 + }, + { + "epoch": 0.22033, + "grad_norm": 1.4952455930118238, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 22033 + }, + { + "epoch": 0.22034, + "grad_norm": 1.2605812991854766, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 22034 + }, + { + "epoch": 0.22035, + "grad_norm": 1.7377216985408868, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22035 + }, + { + "epoch": 0.22036, + "grad_norm": 1.9438553629326887, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 22036 + }, + { + "epoch": 0.22037, + "grad_norm": 1.0593080689450938, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 22037 + }, + { + "epoch": 0.22038, + "grad_norm": 1.6234239889397044, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 22038 + }, + { + "epoch": 0.22039, + "grad_norm": 1.4127420233740093, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 22039 + }, + { + "epoch": 0.2204, + "grad_norm": 1.4386928226103217, + "learning_rate": 0.003, + "loss": 4.083, + "step": 22040 + }, + { + "epoch": 0.22041, + "grad_norm": 1.3624686217299482, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 22041 + }, + { + "epoch": 0.22042, + "grad_norm": 1.3013724831468199, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22042 + }, + { + "epoch": 0.22043, + "grad_norm": 1.2419646753823672, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22043 + }, + { + "epoch": 0.22044, + "grad_norm": 1.4984428244005863, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 22044 + }, + { + "epoch": 0.22045, + "grad_norm": 1.5316081336786143, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 22045 + }, + { + "epoch": 0.22046, + "grad_norm": 1.4880937306232618, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 22046 + }, + { + "epoch": 0.22047, + "grad_norm": 1.1892740904767414, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 22047 + }, + { + "epoch": 0.22048, + "grad_norm": 1.483080169944513, + "learning_rate": 0.003, + "loss": 4.069, + "step": 22048 + }, + { + "epoch": 0.22049, + "grad_norm": 1.2408488390797463, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22049 + }, + { + "epoch": 0.2205, + "grad_norm": 1.3882050104985644, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 22050 + }, + { + "epoch": 0.22051, + "grad_norm": 1.3989381887034402, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 22051 + }, + { + "epoch": 0.22052, + "grad_norm": 1.225212641636643, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 22052 + }, + { + "epoch": 0.22053, + "grad_norm": 1.298685388696244, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 22053 + }, + { + "epoch": 0.22054, + "grad_norm": 1.2475364898339563, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 22054 + }, + { + "epoch": 0.22055, + "grad_norm": 1.2144002309450608, + "learning_rate": 0.003, + "loss": 4.017, + "step": 22055 + }, + { + "epoch": 0.22056, + "grad_norm": 1.618483149512431, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 22056 + }, + { + "epoch": 0.22057, + "grad_norm": 1.1606438126874004, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 22057 + }, + { + "epoch": 0.22058, + "grad_norm": 1.4745123089166658, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 22058 + }, + { + "epoch": 0.22059, + "grad_norm": 1.2769331651867426, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 22059 + }, + { + "epoch": 0.2206, + "grad_norm": 1.4767858939060083, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 22060 + }, + { + "epoch": 0.22061, + "grad_norm": 1.4935033078416218, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22061 + }, + { + "epoch": 0.22062, + "grad_norm": 1.1771267549329605, + "learning_rate": 0.003, + "loss": 4.025, + "step": 22062 + }, + { + "epoch": 0.22063, + "grad_norm": 1.3796150280199535, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 22063 + }, + { + "epoch": 0.22064, + "grad_norm": 1.3271035855948137, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 22064 + }, + { + "epoch": 0.22065, + "grad_norm": 1.1652481329836022, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22065 + }, + { + "epoch": 0.22066, + "grad_norm": 1.2905792226996244, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 22066 + }, + { + "epoch": 0.22067, + "grad_norm": 1.2851657372405187, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 22067 + }, + { + "epoch": 0.22068, + "grad_norm": 1.5742584223384706, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 22068 + }, + { + "epoch": 0.22069, + "grad_norm": 1.338394614444369, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 22069 + }, + { + "epoch": 0.2207, + "grad_norm": 1.3004002796280805, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22070 + }, + { + "epoch": 0.22071, + "grad_norm": 1.421639971712515, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 22071 + }, + { + "epoch": 0.22072, + "grad_norm": 1.428979973898089, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 22072 + }, + { + "epoch": 0.22073, + "grad_norm": 1.668612932878163, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 22073 + }, + { + "epoch": 0.22074, + "grad_norm": 1.4643693345898032, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 22074 + }, + { + "epoch": 0.22075, + "grad_norm": 1.3369643684227495, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 22075 + }, + { + "epoch": 0.22076, + "grad_norm": 1.3383140026879339, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 22076 + }, + { + "epoch": 0.22077, + "grad_norm": 1.2131089352572, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 22077 + }, + { + "epoch": 0.22078, + "grad_norm": 1.3283572949254057, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 22078 + }, + { + "epoch": 0.22079, + "grad_norm": 1.2187737844894437, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22079 + }, + { + "epoch": 0.2208, + "grad_norm": 1.450702948191589, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 22080 + }, + { + "epoch": 0.22081, + "grad_norm": 1.372454012378222, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22081 + }, + { + "epoch": 0.22082, + "grad_norm": 1.5001801474665366, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 22082 + }, + { + "epoch": 0.22083, + "grad_norm": 1.2899948523986073, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22083 + }, + { + "epoch": 0.22084, + "grad_norm": 1.5862468501136513, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 22084 + }, + { + "epoch": 0.22085, + "grad_norm": 1.7794074253172933, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22085 + }, + { + "epoch": 0.22086, + "grad_norm": 1.3740897458861343, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22086 + }, + { + "epoch": 0.22087, + "grad_norm": 1.4937486407431884, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22087 + }, + { + "epoch": 0.22088, + "grad_norm": 1.177792066169784, + "learning_rate": 0.003, + "loss": 4.054, + "step": 22088 + }, + { + "epoch": 0.22089, + "grad_norm": 1.5566010919211382, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22089 + }, + { + "epoch": 0.2209, + "grad_norm": 1.1499788653179301, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22090 + }, + { + "epoch": 0.22091, + "grad_norm": 1.4722548255020058, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22091 + }, + { + "epoch": 0.22092, + "grad_norm": 1.2267218318689077, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 22092 + }, + { + "epoch": 0.22093, + "grad_norm": 1.6261563629891562, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22093 + }, + { + "epoch": 0.22094, + "grad_norm": 1.1668655227439646, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 22094 + }, + { + "epoch": 0.22095, + "grad_norm": 1.3177353554533442, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 22095 + }, + { + "epoch": 0.22096, + "grad_norm": 1.37216423647877, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 22096 + }, + { + "epoch": 0.22097, + "grad_norm": 1.6901582244876612, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 22097 + }, + { + "epoch": 0.22098, + "grad_norm": 1.093044749690386, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 22098 + }, + { + "epoch": 0.22099, + "grad_norm": 1.358581512018892, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 22099 + }, + { + "epoch": 0.221, + "grad_norm": 1.4734676480778928, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 22100 + }, + { + "epoch": 0.22101, + "grad_norm": 1.1247671678322673, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22101 + }, + { + "epoch": 0.22102, + "grad_norm": 1.5760390934631532, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 22102 + }, + { + "epoch": 0.22103, + "grad_norm": 1.4089002766530339, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 22103 + }, + { + "epoch": 0.22104, + "grad_norm": 1.3022377509278193, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 22104 + }, + { + "epoch": 0.22105, + "grad_norm": 1.3879977222861182, + "learning_rate": 0.003, + "loss": 4.048, + "step": 22105 + }, + { + "epoch": 0.22106, + "grad_norm": 1.2008127244062086, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 22106 + }, + { + "epoch": 0.22107, + "grad_norm": 1.3873357243221884, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 22107 + }, + { + "epoch": 0.22108, + "grad_norm": 1.1851877743649342, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22108 + }, + { + "epoch": 0.22109, + "grad_norm": 1.7553480872377545, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22109 + }, + { + "epoch": 0.2211, + "grad_norm": 1.3783634274316432, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 22110 + }, + { + "epoch": 0.22111, + "grad_norm": 1.6975422750772762, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 22111 + }, + { + "epoch": 0.22112, + "grad_norm": 1.2598251750885294, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 22112 + }, + { + "epoch": 0.22113, + "grad_norm": 1.3488825181885948, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22113 + }, + { + "epoch": 0.22114, + "grad_norm": 1.1707139578745651, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 22114 + }, + { + "epoch": 0.22115, + "grad_norm": 1.4076970079350817, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22115 + }, + { + "epoch": 0.22116, + "grad_norm": 1.2035989898263957, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 22116 + }, + { + "epoch": 0.22117, + "grad_norm": 1.313990258923647, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 22117 + }, + { + "epoch": 0.22118, + "grad_norm": 1.4596496530806549, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 22118 + }, + { + "epoch": 0.22119, + "grad_norm": 1.3887979786596634, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 22119 + }, + { + "epoch": 0.2212, + "grad_norm": 1.2892662283743792, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 22120 + }, + { + "epoch": 0.22121, + "grad_norm": 1.3677195530909632, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 22121 + }, + { + "epoch": 0.22122, + "grad_norm": 1.0895280670903982, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 22122 + }, + { + "epoch": 0.22123, + "grad_norm": 1.7090804039595255, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 22123 + }, + { + "epoch": 0.22124, + "grad_norm": 1.2604584884079608, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 22124 + }, + { + "epoch": 0.22125, + "grad_norm": 1.5893034946385178, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22125 + }, + { + "epoch": 0.22126, + "grad_norm": 1.8171719489776013, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 22126 + }, + { + "epoch": 0.22127, + "grad_norm": 1.079785537190597, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 22127 + }, + { + "epoch": 0.22128, + "grad_norm": 1.4253921256548443, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 22128 + }, + { + "epoch": 0.22129, + "grad_norm": 1.2752312555966736, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 22129 + }, + { + "epoch": 0.2213, + "grad_norm": 1.5912898189212894, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22130 + }, + { + "epoch": 0.22131, + "grad_norm": 1.212781183286263, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22131 + }, + { + "epoch": 0.22132, + "grad_norm": 1.6288246243696782, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22132 + }, + { + "epoch": 0.22133, + "grad_norm": 1.364695014188986, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 22133 + }, + { + "epoch": 0.22134, + "grad_norm": 1.4164140732085475, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22134 + }, + { + "epoch": 0.22135, + "grad_norm": 1.2534625007837457, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 22135 + }, + { + "epoch": 0.22136, + "grad_norm": 1.196897126890033, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22136 + }, + { + "epoch": 0.22137, + "grad_norm": 1.502320147887795, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 22137 + }, + { + "epoch": 0.22138, + "grad_norm": 1.3266848797936062, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 22138 + }, + { + "epoch": 0.22139, + "grad_norm": 1.5001054181374793, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 22139 + }, + { + "epoch": 0.2214, + "grad_norm": 1.473091688110243, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 22140 + }, + { + "epoch": 0.22141, + "grad_norm": 1.3799393569486098, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 22141 + }, + { + "epoch": 0.22142, + "grad_norm": 1.4247663877479786, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 22142 + }, + { + "epoch": 0.22143, + "grad_norm": 1.0853672842292295, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 22143 + }, + { + "epoch": 0.22144, + "grad_norm": 1.6092501187145332, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22144 + }, + { + "epoch": 0.22145, + "grad_norm": 1.1489883941066004, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 22145 + }, + { + "epoch": 0.22146, + "grad_norm": 1.5132661551487747, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 22146 + }, + { + "epoch": 0.22147, + "grad_norm": 1.3046988580626422, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 22147 + }, + { + "epoch": 0.22148, + "grad_norm": 1.4132640202686817, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 22148 + }, + { + "epoch": 0.22149, + "grad_norm": 1.56435744408823, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 22149 + }, + { + "epoch": 0.2215, + "grad_norm": 1.686833161313362, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 22150 + }, + { + "epoch": 0.22151, + "grad_norm": 1.3313560145204009, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22151 + }, + { + "epoch": 0.22152, + "grad_norm": 1.3614519489465264, + "learning_rate": 0.003, + "loss": 4.093, + "step": 22152 + }, + { + "epoch": 0.22153, + "grad_norm": 1.4732927145628927, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 22153 + }, + { + "epoch": 0.22154, + "grad_norm": 1.3336301761906506, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22154 + }, + { + "epoch": 0.22155, + "grad_norm": 1.2845741404694027, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 22155 + }, + { + "epoch": 0.22156, + "grad_norm": 1.362731838707528, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22156 + }, + { + "epoch": 0.22157, + "grad_norm": 1.279897373506415, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 22157 + }, + { + "epoch": 0.22158, + "grad_norm": 1.3044229575795832, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 22158 + }, + { + "epoch": 0.22159, + "grad_norm": 1.3994874042785792, + "learning_rate": 0.003, + "loss": 4.077, + "step": 22159 + }, + { + "epoch": 0.2216, + "grad_norm": 1.274607452634766, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 22160 + }, + { + "epoch": 0.22161, + "grad_norm": 1.2943297130759688, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22161 + }, + { + "epoch": 0.22162, + "grad_norm": 1.362183302837162, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 22162 + }, + { + "epoch": 0.22163, + "grad_norm": 1.5660664006583445, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 22163 + }, + { + "epoch": 0.22164, + "grad_norm": 1.3757306851918003, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 22164 + }, + { + "epoch": 0.22165, + "grad_norm": 1.4412752627541876, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 22165 + }, + { + "epoch": 0.22166, + "grad_norm": 1.5658425271051613, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 22166 + }, + { + "epoch": 0.22167, + "grad_norm": 1.6523052232881559, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 22167 + }, + { + "epoch": 0.22168, + "grad_norm": 1.1567559762437325, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 22168 + }, + { + "epoch": 0.22169, + "grad_norm": 1.4328603257986592, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 22169 + }, + { + "epoch": 0.2217, + "grad_norm": 1.3470626463872521, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 22170 + }, + { + "epoch": 0.22171, + "grad_norm": 1.4284284243004235, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 22171 + }, + { + "epoch": 0.22172, + "grad_norm": 1.1893474315038903, + "learning_rate": 0.003, + "loss": 4.059, + "step": 22172 + }, + { + "epoch": 0.22173, + "grad_norm": 1.4289456979469684, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22173 + }, + { + "epoch": 0.22174, + "grad_norm": 1.057297957599464, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 22174 + }, + { + "epoch": 0.22175, + "grad_norm": 1.4257501085613518, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22175 + }, + { + "epoch": 0.22176, + "grad_norm": 1.3262331040893989, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 22176 + }, + { + "epoch": 0.22177, + "grad_norm": 1.4321319336927019, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 22177 + }, + { + "epoch": 0.22178, + "grad_norm": 1.5739569894444765, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 22178 + }, + { + "epoch": 0.22179, + "grad_norm": 1.3596252113335465, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 22179 + }, + { + "epoch": 0.2218, + "grad_norm": 1.4105314530330597, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22180 + }, + { + "epoch": 0.22181, + "grad_norm": 1.362897752269198, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 22181 + }, + { + "epoch": 0.22182, + "grad_norm": 1.4495305691893379, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 22182 + }, + { + "epoch": 0.22183, + "grad_norm": 1.3338275962033688, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 22183 + }, + { + "epoch": 0.22184, + "grad_norm": 1.3961143833545246, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 22184 + }, + { + "epoch": 0.22185, + "grad_norm": 1.2054362280955921, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 22185 + }, + { + "epoch": 0.22186, + "grad_norm": 1.4617105421301142, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 22186 + }, + { + "epoch": 0.22187, + "grad_norm": 1.3091955764656495, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 22187 + }, + { + "epoch": 0.22188, + "grad_norm": 1.3965949404899751, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 22188 + }, + { + "epoch": 0.22189, + "grad_norm": 1.7821877666649462, + "learning_rate": 0.003, + "loss": 4.043, + "step": 22189 + }, + { + "epoch": 0.2219, + "grad_norm": 1.295830581139901, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22190 + }, + { + "epoch": 0.22191, + "grad_norm": 1.445223537820084, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 22191 + }, + { + "epoch": 0.22192, + "grad_norm": 1.3071233557061357, + "learning_rate": 0.003, + "loss": 4.04, + "step": 22192 + }, + { + "epoch": 0.22193, + "grad_norm": 1.2753165325582236, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22193 + }, + { + "epoch": 0.22194, + "grad_norm": 1.5055299549441128, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 22194 + }, + { + "epoch": 0.22195, + "grad_norm": 1.1342693535556096, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 22195 + }, + { + "epoch": 0.22196, + "grad_norm": 1.3858983883526306, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 22196 + }, + { + "epoch": 0.22197, + "grad_norm": 1.6139655980808818, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22197 + }, + { + "epoch": 0.22198, + "grad_norm": 1.370292636226592, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 22198 + }, + { + "epoch": 0.22199, + "grad_norm": 1.4124649114097427, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 22199 + }, + { + "epoch": 0.222, + "grad_norm": 1.293709147338817, + "learning_rate": 0.003, + "loss": 4.049, + "step": 22200 + }, + { + "epoch": 0.22201, + "grad_norm": 1.4533704007836947, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 22201 + }, + { + "epoch": 0.22202, + "grad_norm": 1.567765777013424, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 22202 + }, + { + "epoch": 0.22203, + "grad_norm": 1.279871157610687, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 22203 + }, + { + "epoch": 0.22204, + "grad_norm": 1.385971844695112, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 22204 + }, + { + "epoch": 0.22205, + "grad_norm": 1.2180530423728437, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 22205 + }, + { + "epoch": 0.22206, + "grad_norm": 1.515881036126505, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 22206 + }, + { + "epoch": 0.22207, + "grad_norm": 1.305396320762747, + "learning_rate": 0.003, + "loss": 4.072, + "step": 22207 + }, + { + "epoch": 0.22208, + "grad_norm": 1.907422079706361, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 22208 + }, + { + "epoch": 0.22209, + "grad_norm": 1.21270690214005, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 22209 + }, + { + "epoch": 0.2221, + "grad_norm": 1.3404071885874145, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22210 + }, + { + "epoch": 0.22211, + "grad_norm": 1.2763298267522407, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22211 + }, + { + "epoch": 0.22212, + "grad_norm": 1.4845486426731773, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 22212 + }, + { + "epoch": 0.22213, + "grad_norm": 1.1597212035481745, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 22213 + }, + { + "epoch": 0.22214, + "grad_norm": 1.1880382563326437, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 22214 + }, + { + "epoch": 0.22215, + "grad_norm": 1.4824757540258113, + "learning_rate": 0.003, + "loss": 4.063, + "step": 22215 + }, + { + "epoch": 0.22216, + "grad_norm": 1.3396597534142682, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 22216 + }, + { + "epoch": 0.22217, + "grad_norm": 1.4647900070278586, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22217 + }, + { + "epoch": 0.22218, + "grad_norm": 1.1550937275289768, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 22218 + }, + { + "epoch": 0.22219, + "grad_norm": 1.4764592357132116, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 22219 + }, + { + "epoch": 0.2222, + "grad_norm": 1.7143856305412355, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 22220 + }, + { + "epoch": 0.22221, + "grad_norm": 1.5401353576578565, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22221 + }, + { + "epoch": 0.22222, + "grad_norm": 1.3416022332549657, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 22222 + }, + { + "epoch": 0.22223, + "grad_norm": 1.4242574036135836, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22223 + }, + { + "epoch": 0.22224, + "grad_norm": 1.1973358150028863, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 22224 + }, + { + "epoch": 0.22225, + "grad_norm": 1.2884641971513957, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 22225 + }, + { + "epoch": 0.22226, + "grad_norm": 1.424689662013279, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 22226 + }, + { + "epoch": 0.22227, + "grad_norm": 1.2250087938369854, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 22227 + }, + { + "epoch": 0.22228, + "grad_norm": 1.5751666610904282, + "learning_rate": 0.003, + "loss": 4.063, + "step": 22228 + }, + { + "epoch": 0.22229, + "grad_norm": 1.1494551555423635, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22229 + }, + { + "epoch": 0.2223, + "grad_norm": 1.4688453613121737, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22230 + }, + { + "epoch": 0.22231, + "grad_norm": 1.3699272268665363, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 22231 + }, + { + "epoch": 0.22232, + "grad_norm": 1.537492019731281, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 22232 + }, + { + "epoch": 0.22233, + "grad_norm": 1.0330617251989942, + "learning_rate": 0.003, + "loss": 4.038, + "step": 22233 + }, + { + "epoch": 0.22234, + "grad_norm": 1.4603664211895149, + "learning_rate": 0.003, + "loss": 4.068, + "step": 22234 + }, + { + "epoch": 0.22235, + "grad_norm": 1.2612609323108521, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22235 + }, + { + "epoch": 0.22236, + "grad_norm": 1.566572150437271, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22236 + }, + { + "epoch": 0.22237, + "grad_norm": 1.1865597068407436, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 22237 + }, + { + "epoch": 0.22238, + "grad_norm": 1.5926773568356256, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 22238 + }, + { + "epoch": 0.22239, + "grad_norm": 1.3216088389014924, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 22239 + }, + { + "epoch": 0.2224, + "grad_norm": 1.3976057640990152, + "learning_rate": 0.003, + "loss": 4.054, + "step": 22240 + }, + { + "epoch": 0.22241, + "grad_norm": 1.1496350226071208, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 22241 + }, + { + "epoch": 0.22242, + "grad_norm": 1.1580898237737296, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 22242 + }, + { + "epoch": 0.22243, + "grad_norm": 1.558939181904705, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 22243 + }, + { + "epoch": 0.22244, + "grad_norm": 1.3480081665211374, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 22244 + }, + { + "epoch": 0.22245, + "grad_norm": 1.4832107461145823, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22245 + }, + { + "epoch": 0.22246, + "grad_norm": 1.4375844492231902, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 22246 + }, + { + "epoch": 0.22247, + "grad_norm": 1.5835609335873375, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22247 + }, + { + "epoch": 0.22248, + "grad_norm": 1.4205729727127157, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 22248 + }, + { + "epoch": 0.22249, + "grad_norm": 1.6034593434402513, + "learning_rate": 0.003, + "loss": 4.04, + "step": 22249 + }, + { + "epoch": 0.2225, + "grad_norm": 1.3755340142441637, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 22250 + }, + { + "epoch": 0.22251, + "grad_norm": 1.4787785090773218, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22251 + }, + { + "epoch": 0.22252, + "grad_norm": 1.3389467168011282, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 22252 + }, + { + "epoch": 0.22253, + "grad_norm": 1.4218076740248158, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 22253 + }, + { + "epoch": 0.22254, + "grad_norm": 1.20919633978248, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22254 + }, + { + "epoch": 0.22255, + "grad_norm": 1.3390517491426153, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 22255 + }, + { + "epoch": 0.22256, + "grad_norm": 1.3712688774483852, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 22256 + }, + { + "epoch": 0.22257, + "grad_norm": 1.4771106606224886, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22257 + }, + { + "epoch": 0.22258, + "grad_norm": 1.2131879469373439, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 22258 + }, + { + "epoch": 0.22259, + "grad_norm": 1.430070864931875, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 22259 + }, + { + "epoch": 0.2226, + "grad_norm": 1.1127157666257408, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22260 + }, + { + "epoch": 0.22261, + "grad_norm": 1.4544316597453024, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22261 + }, + { + "epoch": 0.22262, + "grad_norm": 1.1531901204919115, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22262 + }, + { + "epoch": 0.22263, + "grad_norm": 1.5001268775560253, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 22263 + }, + { + "epoch": 0.22264, + "grad_norm": 1.388798488222104, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22264 + }, + { + "epoch": 0.22265, + "grad_norm": 1.7432748344689732, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 22265 + }, + { + "epoch": 0.22266, + "grad_norm": 1.036013823414377, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 22266 + }, + { + "epoch": 0.22267, + "grad_norm": 1.2522459940721222, + "learning_rate": 0.003, + "loss": 4.023, + "step": 22267 + }, + { + "epoch": 0.22268, + "grad_norm": 1.600614247711205, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 22268 + }, + { + "epoch": 0.22269, + "grad_norm": 1.3074802160740018, + "learning_rate": 0.003, + "loss": 4.044, + "step": 22269 + }, + { + "epoch": 0.2227, + "grad_norm": 1.4003433694119265, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 22270 + }, + { + "epoch": 0.22271, + "grad_norm": 1.3913342361356809, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 22271 + }, + { + "epoch": 0.22272, + "grad_norm": 1.2703376309876075, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 22272 + }, + { + "epoch": 0.22273, + "grad_norm": 1.500351184869441, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22273 + }, + { + "epoch": 0.22274, + "grad_norm": 1.3699969478225038, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22274 + }, + { + "epoch": 0.22275, + "grad_norm": 1.5223426025607947, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22275 + }, + { + "epoch": 0.22276, + "grad_norm": 1.4033689564381397, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22276 + }, + { + "epoch": 0.22277, + "grad_norm": 1.4601187426494533, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 22277 + }, + { + "epoch": 0.22278, + "grad_norm": 1.3888147570157965, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 22278 + }, + { + "epoch": 0.22279, + "grad_norm": 1.2404973573348714, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 22279 + }, + { + "epoch": 0.2228, + "grad_norm": 1.4555403624250252, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 22280 + }, + { + "epoch": 0.22281, + "grad_norm": 1.216160029817018, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22281 + }, + { + "epoch": 0.22282, + "grad_norm": 1.4058869811595323, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22282 + }, + { + "epoch": 0.22283, + "grad_norm": 1.2758387158134383, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 22283 + }, + { + "epoch": 0.22284, + "grad_norm": 1.2207169266410474, + "learning_rate": 0.003, + "loss": 4.081, + "step": 22284 + }, + { + "epoch": 0.22285, + "grad_norm": 1.412295249924611, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 22285 + }, + { + "epoch": 0.22286, + "grad_norm": 1.3732494203503594, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22286 + }, + { + "epoch": 0.22287, + "grad_norm": 1.3692184472505888, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 22287 + }, + { + "epoch": 0.22288, + "grad_norm": 1.3759831387998156, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 22288 + }, + { + "epoch": 0.22289, + "grad_norm": 1.2595618500573362, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 22289 + }, + { + "epoch": 0.2229, + "grad_norm": 1.6814775146119474, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 22290 + }, + { + "epoch": 0.22291, + "grad_norm": 1.3953387253409102, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 22291 + }, + { + "epoch": 0.22292, + "grad_norm": 1.7317713148633327, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22292 + }, + { + "epoch": 0.22293, + "grad_norm": 1.2844604150024208, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22293 + }, + { + "epoch": 0.22294, + "grad_norm": 1.4291699204264559, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 22294 + }, + { + "epoch": 0.22295, + "grad_norm": 1.5150130594599611, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 22295 + }, + { + "epoch": 0.22296, + "grad_norm": 1.3082256750627732, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 22296 + }, + { + "epoch": 0.22297, + "grad_norm": 1.3875180604880697, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 22297 + }, + { + "epoch": 0.22298, + "grad_norm": 1.3343820918708291, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 22298 + }, + { + "epoch": 0.22299, + "grad_norm": 1.3183796205639757, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 22299 + }, + { + "epoch": 0.223, + "grad_norm": 1.2510060244764296, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22300 + }, + { + "epoch": 0.22301, + "grad_norm": 1.3716672418427478, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 22301 + }, + { + "epoch": 0.22302, + "grad_norm": 1.2727298642982772, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 22302 + }, + { + "epoch": 0.22303, + "grad_norm": 1.488032564045237, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22303 + }, + { + "epoch": 0.22304, + "grad_norm": 1.4141357281721683, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 22304 + }, + { + "epoch": 0.22305, + "grad_norm": 1.3282741557292026, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 22305 + }, + { + "epoch": 0.22306, + "grad_norm": 1.6074682935491493, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 22306 + }, + { + "epoch": 0.22307, + "grad_norm": 1.393713371586898, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 22307 + }, + { + "epoch": 0.22308, + "grad_norm": 1.307301396543611, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 22308 + }, + { + "epoch": 0.22309, + "grad_norm": 1.3461150372960098, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 22309 + }, + { + "epoch": 0.2231, + "grad_norm": 1.3792478838566686, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 22310 + }, + { + "epoch": 0.22311, + "grad_norm": 1.4901498414341812, + "learning_rate": 0.003, + "loss": 4.075, + "step": 22311 + }, + { + "epoch": 0.22312, + "grad_norm": 1.485376306335906, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 22312 + }, + { + "epoch": 0.22313, + "grad_norm": 1.4336359621977617, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 22313 + }, + { + "epoch": 0.22314, + "grad_norm": 1.4223755734156847, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 22314 + }, + { + "epoch": 0.22315, + "grad_norm": 1.2564372412889622, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22315 + }, + { + "epoch": 0.22316, + "grad_norm": 1.2536470066684344, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22316 + }, + { + "epoch": 0.22317, + "grad_norm": 1.5964660266455553, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22317 + }, + { + "epoch": 0.22318, + "grad_norm": 1.1143586140085437, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 22318 + }, + { + "epoch": 0.22319, + "grad_norm": 1.8202754570487123, + "learning_rate": 0.003, + "loss": 4.082, + "step": 22319 + }, + { + "epoch": 0.2232, + "grad_norm": 1.1241456027960381, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 22320 + }, + { + "epoch": 0.22321, + "grad_norm": 1.4648023777671297, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 22321 + }, + { + "epoch": 0.22322, + "grad_norm": 1.4663830323431575, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 22322 + }, + { + "epoch": 0.22323, + "grad_norm": 1.1849994982508425, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 22323 + }, + { + "epoch": 0.22324, + "grad_norm": 1.371011478385604, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 22324 + }, + { + "epoch": 0.22325, + "grad_norm": 1.2628273562560974, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22325 + }, + { + "epoch": 0.22326, + "grad_norm": 1.3872267826245577, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 22326 + }, + { + "epoch": 0.22327, + "grad_norm": 1.3685132988827458, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 22327 + }, + { + "epoch": 0.22328, + "grad_norm": 1.1028904550243237, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 22328 + }, + { + "epoch": 0.22329, + "grad_norm": 1.3618561520281762, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 22329 + }, + { + "epoch": 0.2233, + "grad_norm": 1.1445670049550696, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22330 + }, + { + "epoch": 0.22331, + "grad_norm": 1.4022987676702379, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 22331 + }, + { + "epoch": 0.22332, + "grad_norm": 1.5599619432209575, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22332 + }, + { + "epoch": 0.22333, + "grad_norm": 1.878794688612716, + "learning_rate": 0.003, + "loss": 4.062, + "step": 22333 + }, + { + "epoch": 0.22334, + "grad_norm": 1.3050525918215132, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 22334 + }, + { + "epoch": 0.22335, + "grad_norm": 1.336371350588149, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 22335 + }, + { + "epoch": 0.22336, + "grad_norm": 1.208607895788011, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 22336 + }, + { + "epoch": 0.22337, + "grad_norm": 1.6457839457667778, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 22337 + }, + { + "epoch": 0.22338, + "grad_norm": 1.4127662001557084, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 22338 + }, + { + "epoch": 0.22339, + "grad_norm": 1.376876386888607, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22339 + }, + { + "epoch": 0.2234, + "grad_norm": 1.4570261167898284, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 22340 + }, + { + "epoch": 0.22341, + "grad_norm": 1.211546957361986, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 22341 + }, + { + "epoch": 0.22342, + "grad_norm": 1.3270741919713742, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 22342 + }, + { + "epoch": 0.22343, + "grad_norm": 1.397195166385156, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 22343 + }, + { + "epoch": 0.22344, + "grad_norm": 1.4763850759105857, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22344 + }, + { + "epoch": 0.22345, + "grad_norm": 1.3501067025222995, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22345 + }, + { + "epoch": 0.22346, + "grad_norm": 1.279130243319669, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 22346 + }, + { + "epoch": 0.22347, + "grad_norm": 1.352191370618253, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 22347 + }, + { + "epoch": 0.22348, + "grad_norm": 1.4361298838127596, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22348 + }, + { + "epoch": 0.22349, + "grad_norm": 1.2251495326785093, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22349 + }, + { + "epoch": 0.2235, + "grad_norm": 1.3910755839615114, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 22350 + }, + { + "epoch": 0.22351, + "grad_norm": 1.3125911961348145, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 22351 + }, + { + "epoch": 0.22352, + "grad_norm": 1.5656196258299442, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22352 + }, + { + "epoch": 0.22353, + "grad_norm": 1.1531059389308478, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 22353 + }, + { + "epoch": 0.22354, + "grad_norm": 1.4007020106647727, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 22354 + }, + { + "epoch": 0.22355, + "grad_norm": 1.2354692843363417, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 22355 + }, + { + "epoch": 0.22356, + "grad_norm": 1.590234105863267, + "learning_rate": 0.003, + "loss": 4.07, + "step": 22356 + }, + { + "epoch": 0.22357, + "grad_norm": 1.3509492624833452, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 22357 + }, + { + "epoch": 0.22358, + "grad_norm": 1.5275962199582724, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 22358 + }, + { + "epoch": 0.22359, + "grad_norm": 1.209076278040971, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 22359 + }, + { + "epoch": 0.2236, + "grad_norm": 1.5185901971561602, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22360 + }, + { + "epoch": 0.22361, + "grad_norm": 1.532891108295708, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22361 + }, + { + "epoch": 0.22362, + "grad_norm": 1.5865126216653196, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22362 + }, + { + "epoch": 0.22363, + "grad_norm": 1.270551236014406, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 22363 + }, + { + "epoch": 0.22364, + "grad_norm": 1.4329458288291133, + "learning_rate": 0.003, + "loss": 4.045, + "step": 22364 + }, + { + "epoch": 0.22365, + "grad_norm": 1.4096876484606953, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 22365 + }, + { + "epoch": 0.22366, + "grad_norm": 1.2142737042399505, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 22366 + }, + { + "epoch": 0.22367, + "grad_norm": 1.2614164115511466, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 22367 + }, + { + "epoch": 0.22368, + "grad_norm": 1.2051666649693737, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22368 + }, + { + "epoch": 0.22369, + "grad_norm": 1.590599099523706, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22369 + }, + { + "epoch": 0.2237, + "grad_norm": 1.1253766088301476, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 22370 + }, + { + "epoch": 0.22371, + "grad_norm": 1.5552956946763759, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 22371 + }, + { + "epoch": 0.22372, + "grad_norm": 1.3491888281502122, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 22372 + }, + { + "epoch": 0.22373, + "grad_norm": 1.3715023603072742, + "learning_rate": 0.003, + "loss": 4.075, + "step": 22373 + }, + { + "epoch": 0.22374, + "grad_norm": 1.3616540152037222, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 22374 + }, + { + "epoch": 0.22375, + "grad_norm": 1.177630374725669, + "learning_rate": 0.003, + "loss": 4.036, + "step": 22375 + }, + { + "epoch": 0.22376, + "grad_norm": 1.3473079774707535, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 22376 + }, + { + "epoch": 0.22377, + "grad_norm": 1.3537114757041389, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 22377 + }, + { + "epoch": 0.22378, + "grad_norm": 1.5142584430469408, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 22378 + }, + { + "epoch": 0.22379, + "grad_norm": 1.4421599606051467, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 22379 + }, + { + "epoch": 0.2238, + "grad_norm": 1.5732608949994045, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22380 + }, + { + "epoch": 0.22381, + "grad_norm": 1.2532908478937852, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 22381 + }, + { + "epoch": 0.22382, + "grad_norm": 1.4183163010238564, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 22382 + }, + { + "epoch": 0.22383, + "grad_norm": 1.3639666346782449, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22383 + }, + { + "epoch": 0.22384, + "grad_norm": 1.1796621129050568, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 22384 + }, + { + "epoch": 0.22385, + "grad_norm": 1.500752473050197, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 22385 + }, + { + "epoch": 0.22386, + "grad_norm": 1.299806986589891, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 22386 + }, + { + "epoch": 0.22387, + "grad_norm": 1.3628678395012956, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 22387 + }, + { + "epoch": 0.22388, + "grad_norm": 1.2145588780699879, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 22388 + }, + { + "epoch": 0.22389, + "grad_norm": 1.4354220668775866, + "learning_rate": 0.003, + "loss": 4.056, + "step": 22389 + }, + { + "epoch": 0.2239, + "grad_norm": 1.1753083460488114, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 22390 + }, + { + "epoch": 0.22391, + "grad_norm": 1.4450015905639986, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 22391 + }, + { + "epoch": 0.22392, + "grad_norm": 1.2241119805245964, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 22392 + }, + { + "epoch": 0.22393, + "grad_norm": 1.5411689184760144, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 22393 + }, + { + "epoch": 0.22394, + "grad_norm": 1.279268771909209, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22394 + }, + { + "epoch": 0.22395, + "grad_norm": 1.7097093223307864, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 22395 + }, + { + "epoch": 0.22396, + "grad_norm": 1.4902652883348568, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 22396 + }, + { + "epoch": 0.22397, + "grad_norm": 1.6807845293587749, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 22397 + }, + { + "epoch": 0.22398, + "grad_norm": 1.463050957154932, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 22398 + }, + { + "epoch": 0.22399, + "grad_norm": 1.4815674497276095, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 22399 + }, + { + "epoch": 0.224, + "grad_norm": 1.2927791947078966, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22400 + }, + { + "epoch": 0.22401, + "grad_norm": 1.3616976324807482, + "learning_rate": 0.003, + "loss": 4.095, + "step": 22401 + }, + { + "epoch": 0.22402, + "grad_norm": 1.3534396398652109, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 22402 + }, + { + "epoch": 0.22403, + "grad_norm": 1.286676318927532, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 22403 + }, + { + "epoch": 0.22404, + "grad_norm": 1.3768423147509237, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 22404 + }, + { + "epoch": 0.22405, + "grad_norm": 1.1635484876103743, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 22405 + }, + { + "epoch": 0.22406, + "grad_norm": 1.3140438382708493, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22406 + }, + { + "epoch": 0.22407, + "grad_norm": 1.3712284265314465, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 22407 + }, + { + "epoch": 0.22408, + "grad_norm": 1.1867705031365885, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 22408 + }, + { + "epoch": 0.22409, + "grad_norm": 1.1574130541241712, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 22409 + }, + { + "epoch": 0.2241, + "grad_norm": 1.2587040600977484, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 22410 + }, + { + "epoch": 0.22411, + "grad_norm": 1.3772805242419397, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 22411 + }, + { + "epoch": 0.22412, + "grad_norm": 1.4911914513850275, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22412 + }, + { + "epoch": 0.22413, + "grad_norm": 1.2854427393908894, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 22413 + }, + { + "epoch": 0.22414, + "grad_norm": 1.5290408026536262, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 22414 + }, + { + "epoch": 0.22415, + "grad_norm": 1.7649135826504136, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22415 + }, + { + "epoch": 0.22416, + "grad_norm": 1.1977781109776366, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 22416 + }, + { + "epoch": 0.22417, + "grad_norm": 1.293867976503959, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 22417 + }, + { + "epoch": 0.22418, + "grad_norm": 1.630642197670863, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 22418 + }, + { + "epoch": 0.22419, + "grad_norm": 1.4741383401959558, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 22419 + }, + { + "epoch": 0.2242, + "grad_norm": 1.2986315664115111, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 22420 + }, + { + "epoch": 0.22421, + "grad_norm": 1.3154813967475658, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 22421 + }, + { + "epoch": 0.22422, + "grad_norm": 1.5969126924782886, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22422 + }, + { + "epoch": 0.22423, + "grad_norm": 1.1954027795687365, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22423 + }, + { + "epoch": 0.22424, + "grad_norm": 1.3933097252550717, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 22424 + }, + { + "epoch": 0.22425, + "grad_norm": 1.04149405593488, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 22425 + }, + { + "epoch": 0.22426, + "grad_norm": 1.6382424532271298, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 22426 + }, + { + "epoch": 0.22427, + "grad_norm": 1.2027355428887891, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 22427 + }, + { + "epoch": 0.22428, + "grad_norm": 1.525117510391192, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 22428 + }, + { + "epoch": 0.22429, + "grad_norm": 1.426170080120505, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 22429 + }, + { + "epoch": 0.2243, + "grad_norm": 1.221673236416552, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 22430 + }, + { + "epoch": 0.22431, + "grad_norm": 1.5885387264266417, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 22431 + }, + { + "epoch": 0.22432, + "grad_norm": 1.2345654548738305, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 22432 + }, + { + "epoch": 0.22433, + "grad_norm": 1.5798732787241532, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22433 + }, + { + "epoch": 0.22434, + "grad_norm": 1.3471715133315665, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 22434 + }, + { + "epoch": 0.22435, + "grad_norm": 1.3854398618572008, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 22435 + }, + { + "epoch": 0.22436, + "grad_norm": 1.3779282433032272, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22436 + }, + { + "epoch": 0.22437, + "grad_norm": 1.5169256570561949, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22437 + }, + { + "epoch": 0.22438, + "grad_norm": 1.4647617156380661, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 22438 + }, + { + "epoch": 0.22439, + "grad_norm": 1.2177903916745678, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22439 + }, + { + "epoch": 0.2244, + "grad_norm": 1.4261753366390773, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 22440 + }, + { + "epoch": 0.22441, + "grad_norm": 1.1943996436170505, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22441 + }, + { + "epoch": 0.22442, + "grad_norm": 1.650187508331776, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 22442 + }, + { + "epoch": 0.22443, + "grad_norm": 1.2273277856836666, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 22443 + }, + { + "epoch": 0.22444, + "grad_norm": 1.563436201042193, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 22444 + }, + { + "epoch": 0.22445, + "grad_norm": 1.2477425551662131, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 22445 + }, + { + "epoch": 0.22446, + "grad_norm": 1.459547516657567, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22446 + }, + { + "epoch": 0.22447, + "grad_norm": 1.237427311213376, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 22447 + }, + { + "epoch": 0.22448, + "grad_norm": 1.582809499011086, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22448 + }, + { + "epoch": 0.22449, + "grad_norm": 1.189896586373214, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 22449 + }, + { + "epoch": 0.2245, + "grad_norm": 1.9901290265561125, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 22450 + }, + { + "epoch": 0.22451, + "grad_norm": 0.9412658196252865, + "learning_rate": 0.003, + "loss": 4.048, + "step": 22451 + }, + { + "epoch": 0.22452, + "grad_norm": 1.518676999335754, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 22452 + }, + { + "epoch": 0.22453, + "grad_norm": 1.457530808236886, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22453 + }, + { + "epoch": 0.22454, + "grad_norm": 1.2408992964524617, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 22454 + }, + { + "epoch": 0.22455, + "grad_norm": 1.5144145116345065, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 22455 + }, + { + "epoch": 0.22456, + "grad_norm": 1.2176355775449714, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 22456 + }, + { + "epoch": 0.22457, + "grad_norm": 1.578395937149088, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 22457 + }, + { + "epoch": 0.22458, + "grad_norm": 1.5822231270531464, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 22458 + }, + { + "epoch": 0.22459, + "grad_norm": 1.3717071898734479, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 22459 + }, + { + "epoch": 0.2246, + "grad_norm": 1.6047488710069298, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 22460 + }, + { + "epoch": 0.22461, + "grad_norm": 1.1601238559678562, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22461 + }, + { + "epoch": 0.22462, + "grad_norm": 1.3735324716539978, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22462 + }, + { + "epoch": 0.22463, + "grad_norm": 1.1844585024097143, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 22463 + }, + { + "epoch": 0.22464, + "grad_norm": 1.5323727852590565, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 22464 + }, + { + "epoch": 0.22465, + "grad_norm": 1.2472750170538465, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 22465 + }, + { + "epoch": 0.22466, + "grad_norm": 1.5245760113116062, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 22466 + }, + { + "epoch": 0.22467, + "grad_norm": 1.304704002247532, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 22467 + }, + { + "epoch": 0.22468, + "grad_norm": 1.2921716053001555, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22468 + }, + { + "epoch": 0.22469, + "grad_norm": 1.4888616674466386, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 22469 + }, + { + "epoch": 0.2247, + "grad_norm": 1.4069052197621945, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22470 + }, + { + "epoch": 0.22471, + "grad_norm": 1.2099995067052336, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22471 + }, + { + "epoch": 0.22472, + "grad_norm": 1.4435501977018366, + "learning_rate": 0.003, + "loss": 4.093, + "step": 22472 + }, + { + "epoch": 0.22473, + "grad_norm": 1.1184531466394783, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22473 + }, + { + "epoch": 0.22474, + "grad_norm": 1.42146086377367, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22474 + }, + { + "epoch": 0.22475, + "grad_norm": 1.2721389239149017, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 22475 + }, + { + "epoch": 0.22476, + "grad_norm": 1.5888927772852055, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 22476 + }, + { + "epoch": 0.22477, + "grad_norm": 1.2323738375159854, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22477 + }, + { + "epoch": 0.22478, + "grad_norm": 1.8480494916768275, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 22478 + }, + { + "epoch": 0.22479, + "grad_norm": 1.039961285327314, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22479 + }, + { + "epoch": 0.2248, + "grad_norm": 1.3240798637812554, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22480 + }, + { + "epoch": 0.22481, + "grad_norm": 1.138999054570129, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22481 + }, + { + "epoch": 0.22482, + "grad_norm": 1.6015792048656434, + "learning_rate": 0.003, + "loss": 4.052, + "step": 22482 + }, + { + "epoch": 0.22483, + "grad_norm": 1.2811650768298755, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22483 + }, + { + "epoch": 0.22484, + "grad_norm": 1.4050836723792486, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 22484 + }, + { + "epoch": 0.22485, + "grad_norm": 1.401263027541136, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 22485 + }, + { + "epoch": 0.22486, + "grad_norm": 1.3808037000615874, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 22486 + }, + { + "epoch": 0.22487, + "grad_norm": 1.0181960800189673, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 22487 + }, + { + "epoch": 0.22488, + "grad_norm": 1.4367139726826885, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 22488 + }, + { + "epoch": 0.22489, + "grad_norm": 1.2553541203514906, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22489 + }, + { + "epoch": 0.2249, + "grad_norm": 1.61788577800353, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22490 + }, + { + "epoch": 0.22491, + "grad_norm": 1.3568049795287969, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 22491 + }, + { + "epoch": 0.22492, + "grad_norm": 1.9470979963846207, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 22492 + }, + { + "epoch": 0.22493, + "grad_norm": 0.999144675700383, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22493 + }, + { + "epoch": 0.22494, + "grad_norm": 1.4578098514241924, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22494 + }, + { + "epoch": 0.22495, + "grad_norm": 1.5647267704157142, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 22495 + }, + { + "epoch": 0.22496, + "grad_norm": 1.2162677300726918, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 22496 + }, + { + "epoch": 0.22497, + "grad_norm": 1.3655534797327769, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 22497 + }, + { + "epoch": 0.22498, + "grad_norm": 1.3563680721289868, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22498 + }, + { + "epoch": 0.22499, + "grad_norm": 1.314289618928932, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 22499 + }, + { + "epoch": 0.225, + "grad_norm": 1.2412727130259196, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22500 + }, + { + "epoch": 0.22501, + "grad_norm": 1.407664446284243, + "learning_rate": 0.003, + "loss": 4.047, + "step": 22501 + }, + { + "epoch": 0.22502, + "grad_norm": 1.1452879237859683, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 22502 + }, + { + "epoch": 0.22503, + "grad_norm": 1.681630374166743, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 22503 + }, + { + "epoch": 0.22504, + "grad_norm": 1.5208064356506332, + "learning_rate": 0.003, + "loss": 4.07, + "step": 22504 + }, + { + "epoch": 0.22505, + "grad_norm": 1.9040342674938469, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22505 + }, + { + "epoch": 0.22506, + "grad_norm": 0.9039834512748233, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 22506 + }, + { + "epoch": 0.22507, + "grad_norm": 1.1289527494797158, + "learning_rate": 0.003, + "loss": 4.069, + "step": 22507 + }, + { + "epoch": 0.22508, + "grad_norm": 1.4814373394057403, + "learning_rate": 0.003, + "loss": 4.06, + "step": 22508 + }, + { + "epoch": 0.22509, + "grad_norm": 1.251319290489344, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 22509 + }, + { + "epoch": 0.2251, + "grad_norm": 1.5562981775401703, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 22510 + }, + { + "epoch": 0.22511, + "grad_norm": 1.3051826964869775, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 22511 + }, + { + "epoch": 0.22512, + "grad_norm": 1.6229240732536356, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 22512 + }, + { + "epoch": 0.22513, + "grad_norm": 1.3042315359192964, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 22513 + }, + { + "epoch": 0.22514, + "grad_norm": 1.4675307982498782, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 22514 + }, + { + "epoch": 0.22515, + "grad_norm": 1.6252411003222522, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 22515 + }, + { + "epoch": 0.22516, + "grad_norm": 1.0812707864662017, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 22516 + }, + { + "epoch": 0.22517, + "grad_norm": 1.3767570950385721, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 22517 + }, + { + "epoch": 0.22518, + "grad_norm": 1.4763795196532024, + "learning_rate": 0.003, + "loss": 4.055, + "step": 22518 + }, + { + "epoch": 0.22519, + "grad_norm": 1.4993782616115667, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 22519 + }, + { + "epoch": 0.2252, + "grad_norm": 1.3537945752575407, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 22520 + }, + { + "epoch": 0.22521, + "grad_norm": 1.3059347310012368, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 22521 + }, + { + "epoch": 0.22522, + "grad_norm": 1.4442979723184586, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22522 + }, + { + "epoch": 0.22523, + "grad_norm": 1.2281159833260513, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22523 + }, + { + "epoch": 0.22524, + "grad_norm": 1.4702059963582377, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 22524 + }, + { + "epoch": 0.22525, + "grad_norm": 1.2684940320670692, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 22525 + }, + { + "epoch": 0.22526, + "grad_norm": 1.5278258401588285, + "learning_rate": 0.003, + "loss": 4.077, + "step": 22526 + }, + { + "epoch": 0.22527, + "grad_norm": 1.2524578589278004, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 22527 + }, + { + "epoch": 0.22528, + "grad_norm": 1.520965957159605, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22528 + }, + { + "epoch": 0.22529, + "grad_norm": 1.2582154302824615, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22529 + }, + { + "epoch": 0.2253, + "grad_norm": 1.3265404712352202, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22530 + }, + { + "epoch": 0.22531, + "grad_norm": 1.4400352868471908, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 22531 + }, + { + "epoch": 0.22532, + "grad_norm": 1.1005549267295203, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 22532 + }, + { + "epoch": 0.22533, + "grad_norm": 1.3708248175839755, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 22533 + }, + { + "epoch": 0.22534, + "grad_norm": 1.3036441616788939, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 22534 + }, + { + "epoch": 0.22535, + "grad_norm": 1.3604752364885113, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 22535 + }, + { + "epoch": 0.22536, + "grad_norm": 1.516204836371943, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 22536 + }, + { + "epoch": 0.22537, + "grad_norm": 1.3429535542996642, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22537 + }, + { + "epoch": 0.22538, + "grad_norm": 1.5562585708934367, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 22538 + }, + { + "epoch": 0.22539, + "grad_norm": 1.36622916218615, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 22539 + }, + { + "epoch": 0.2254, + "grad_norm": 1.3929020539285881, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 22540 + }, + { + "epoch": 0.22541, + "grad_norm": 1.3343444959931712, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 22541 + }, + { + "epoch": 0.22542, + "grad_norm": 1.2985053494629357, + "learning_rate": 0.003, + "loss": 4.053, + "step": 22542 + }, + { + "epoch": 0.22543, + "grad_norm": 1.4792491364252658, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22543 + }, + { + "epoch": 0.22544, + "grad_norm": 1.1959731134436147, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 22544 + }, + { + "epoch": 0.22545, + "grad_norm": 1.4724669305686342, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22545 + }, + { + "epoch": 0.22546, + "grad_norm": 1.1761225949887955, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 22546 + }, + { + "epoch": 0.22547, + "grad_norm": 1.834610039270939, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22547 + }, + { + "epoch": 0.22548, + "grad_norm": 1.3844490196846349, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 22548 + }, + { + "epoch": 0.22549, + "grad_norm": 1.3630463640737251, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 22549 + }, + { + "epoch": 0.2255, + "grad_norm": 1.3436021286818365, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22550 + }, + { + "epoch": 0.22551, + "grad_norm": 1.126599897912119, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 22551 + }, + { + "epoch": 0.22552, + "grad_norm": 1.3418933257662087, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22552 + }, + { + "epoch": 0.22553, + "grad_norm": 1.5600682091030313, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 22553 + }, + { + "epoch": 0.22554, + "grad_norm": 1.27398638381522, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 22554 + }, + { + "epoch": 0.22555, + "grad_norm": 1.362818561676482, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 22555 + }, + { + "epoch": 0.22556, + "grad_norm": 1.4124436107553786, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 22556 + }, + { + "epoch": 0.22557, + "grad_norm": 1.266220853086558, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22557 + }, + { + "epoch": 0.22558, + "grad_norm": 1.426748845952024, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 22558 + }, + { + "epoch": 0.22559, + "grad_norm": 1.4317525615950089, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 22559 + }, + { + "epoch": 0.2256, + "grad_norm": 1.5982695632874984, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 22560 + }, + { + "epoch": 0.22561, + "grad_norm": 1.1566627687428226, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22561 + }, + { + "epoch": 0.22562, + "grad_norm": 1.521382546062163, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22562 + }, + { + "epoch": 0.22563, + "grad_norm": 1.3820456222187198, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 22563 + }, + { + "epoch": 0.22564, + "grad_norm": 1.5823811522579028, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 22564 + }, + { + "epoch": 0.22565, + "grad_norm": 1.500567266962183, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 22565 + }, + { + "epoch": 0.22566, + "grad_norm": 1.3315173635786572, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 22566 + }, + { + "epoch": 0.22567, + "grad_norm": 1.282929593176905, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22567 + }, + { + "epoch": 0.22568, + "grad_norm": 1.1853758748353218, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 22568 + }, + { + "epoch": 0.22569, + "grad_norm": 1.621094200564941, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 22569 + }, + { + "epoch": 0.2257, + "grad_norm": 1.046054275055912, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22570 + }, + { + "epoch": 0.22571, + "grad_norm": 1.6577346130398705, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22571 + }, + { + "epoch": 0.22572, + "grad_norm": 1.2758186688416238, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 22572 + }, + { + "epoch": 0.22573, + "grad_norm": 1.353975079659158, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 22573 + }, + { + "epoch": 0.22574, + "grad_norm": 1.5463450147980808, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22574 + }, + { + "epoch": 0.22575, + "grad_norm": 1.3217539352653154, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 22575 + }, + { + "epoch": 0.22576, + "grad_norm": 1.3559801332602135, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 22576 + }, + { + "epoch": 0.22577, + "grad_norm": 1.3706392752172982, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 22577 + }, + { + "epoch": 0.22578, + "grad_norm": 1.3520523513731522, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22578 + }, + { + "epoch": 0.22579, + "grad_norm": 1.4081367473851485, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 22579 + }, + { + "epoch": 0.2258, + "grad_norm": 1.39604049224214, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 22580 + }, + { + "epoch": 0.22581, + "grad_norm": 0.981022986380773, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 22581 + }, + { + "epoch": 0.22582, + "grad_norm": 1.47186423594832, + "learning_rate": 0.003, + "loss": 4.057, + "step": 22582 + }, + { + "epoch": 0.22583, + "grad_norm": 1.2225686547530379, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 22583 + }, + { + "epoch": 0.22584, + "grad_norm": 1.67045896861936, + "learning_rate": 0.003, + "loss": 4.08, + "step": 22584 + }, + { + "epoch": 0.22585, + "grad_norm": 1.4326038394314937, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 22585 + }, + { + "epoch": 0.22586, + "grad_norm": 1.172198451848986, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 22586 + }, + { + "epoch": 0.22587, + "grad_norm": 1.7019830866366101, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 22587 + }, + { + "epoch": 0.22588, + "grad_norm": 1.4090667334413138, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 22588 + }, + { + "epoch": 0.22589, + "grad_norm": 1.3827337695248374, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 22589 + }, + { + "epoch": 0.2259, + "grad_norm": 1.396615880672321, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 22590 + }, + { + "epoch": 0.22591, + "grad_norm": 1.4282167560265029, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 22591 + }, + { + "epoch": 0.22592, + "grad_norm": 1.3580130177571552, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 22592 + }, + { + "epoch": 0.22593, + "grad_norm": 1.3110588395587528, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 22593 + }, + { + "epoch": 0.22594, + "grad_norm": 1.249002226714932, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22594 + }, + { + "epoch": 0.22595, + "grad_norm": 1.3160504276882146, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 22595 + }, + { + "epoch": 0.22596, + "grad_norm": 1.5536352519305991, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 22596 + }, + { + "epoch": 0.22597, + "grad_norm": 1.3787725503319284, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 22597 + }, + { + "epoch": 0.22598, + "grad_norm": 1.3394153645608453, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 22598 + }, + { + "epoch": 0.22599, + "grad_norm": 1.931446622847827, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 22599 + }, + { + "epoch": 0.226, + "grad_norm": 1.06935267886322, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 22600 + }, + { + "epoch": 0.22601, + "grad_norm": 1.6100699849910967, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 22601 + }, + { + "epoch": 0.22602, + "grad_norm": 1.4042224864761799, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 22602 + }, + { + "epoch": 0.22603, + "grad_norm": 1.555809248028907, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 22603 + }, + { + "epoch": 0.22604, + "grad_norm": 1.3116509073024893, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 22604 + }, + { + "epoch": 0.22605, + "grad_norm": 1.2766533829004472, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22605 + }, + { + "epoch": 0.22606, + "grad_norm": 1.3952547284964383, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 22606 + }, + { + "epoch": 0.22607, + "grad_norm": 1.2181291883730188, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22607 + }, + { + "epoch": 0.22608, + "grad_norm": 1.2698844722199734, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 22608 + }, + { + "epoch": 0.22609, + "grad_norm": 1.2171669532200873, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 22609 + }, + { + "epoch": 0.2261, + "grad_norm": 1.2821859204849533, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 22610 + }, + { + "epoch": 0.22611, + "grad_norm": 1.697919090405346, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 22611 + }, + { + "epoch": 0.22612, + "grad_norm": 1.3176091918110264, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22612 + }, + { + "epoch": 0.22613, + "grad_norm": 1.717003828516172, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 22613 + }, + { + "epoch": 0.22614, + "grad_norm": 1.2890734962230233, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22614 + }, + { + "epoch": 0.22615, + "grad_norm": 1.4817984017078425, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 22615 + }, + { + "epoch": 0.22616, + "grad_norm": 1.3572368270716264, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 22616 + }, + { + "epoch": 0.22617, + "grad_norm": 1.2028491274514344, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22617 + }, + { + "epoch": 0.22618, + "grad_norm": 1.5242588280099894, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 22618 + }, + { + "epoch": 0.22619, + "grad_norm": 1.1436578845230336, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 22619 + }, + { + "epoch": 0.2262, + "grad_norm": 1.3811672576952223, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22620 + }, + { + "epoch": 0.22621, + "grad_norm": 1.2953398770733349, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22621 + }, + { + "epoch": 0.22622, + "grad_norm": 1.3431482846945377, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 22622 + }, + { + "epoch": 0.22623, + "grad_norm": 1.1372747823405238, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 22623 + }, + { + "epoch": 0.22624, + "grad_norm": 1.4052776077364333, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 22624 + }, + { + "epoch": 0.22625, + "grad_norm": 1.5489455531528675, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22625 + }, + { + "epoch": 0.22626, + "grad_norm": 1.4301500501178745, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22626 + }, + { + "epoch": 0.22627, + "grad_norm": 1.2121269881757493, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22627 + }, + { + "epoch": 0.22628, + "grad_norm": 1.3274725839129087, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 22628 + }, + { + "epoch": 0.22629, + "grad_norm": 1.2245562570378739, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 22629 + }, + { + "epoch": 0.2263, + "grad_norm": 1.5273451682665296, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 22630 + }, + { + "epoch": 0.22631, + "grad_norm": 1.279537679015296, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 22631 + }, + { + "epoch": 0.22632, + "grad_norm": 1.3074789891509484, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 22632 + }, + { + "epoch": 0.22633, + "grad_norm": 1.2957812732762986, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 22633 + }, + { + "epoch": 0.22634, + "grad_norm": 1.3134530453328914, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 22634 + }, + { + "epoch": 0.22635, + "grad_norm": 1.4303878963643286, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 22635 + }, + { + "epoch": 0.22636, + "grad_norm": 1.4462539730138868, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22636 + }, + { + "epoch": 0.22637, + "grad_norm": 1.348818328916082, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22637 + }, + { + "epoch": 0.22638, + "grad_norm": 1.5418063711053704, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 22638 + }, + { + "epoch": 0.22639, + "grad_norm": 1.341868096211138, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22639 + }, + { + "epoch": 0.2264, + "grad_norm": 1.5288327353418887, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 22640 + }, + { + "epoch": 0.22641, + "grad_norm": 1.1756331603935857, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 22641 + }, + { + "epoch": 0.22642, + "grad_norm": 1.7303641780587524, + "learning_rate": 0.003, + "loss": 4.057, + "step": 22642 + }, + { + "epoch": 0.22643, + "grad_norm": 1.675870297737419, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 22643 + }, + { + "epoch": 0.22644, + "grad_norm": 1.2903553844701023, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 22644 + }, + { + "epoch": 0.22645, + "grad_norm": 1.3563239081724314, + "learning_rate": 0.003, + "loss": 4.049, + "step": 22645 + }, + { + "epoch": 0.22646, + "grad_norm": 1.4468430507102832, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 22646 + }, + { + "epoch": 0.22647, + "grad_norm": 1.3259292173535169, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 22647 + }, + { + "epoch": 0.22648, + "grad_norm": 1.4509995946776215, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 22648 + }, + { + "epoch": 0.22649, + "grad_norm": 1.4294367517999493, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 22649 + }, + { + "epoch": 0.2265, + "grad_norm": 1.3177940217664728, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 22650 + }, + { + "epoch": 0.22651, + "grad_norm": 1.2960509928184432, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 22651 + }, + { + "epoch": 0.22652, + "grad_norm": 1.372940217189916, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 22652 + }, + { + "epoch": 0.22653, + "grad_norm": 1.2565000630216578, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 22653 + }, + { + "epoch": 0.22654, + "grad_norm": 1.4910404307810698, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 22654 + }, + { + "epoch": 0.22655, + "grad_norm": 1.4885483591490194, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 22655 + }, + { + "epoch": 0.22656, + "grad_norm": 1.2593129639972969, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22656 + }, + { + "epoch": 0.22657, + "grad_norm": 1.4690399184727785, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 22657 + }, + { + "epoch": 0.22658, + "grad_norm": 1.2480041075236903, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22658 + }, + { + "epoch": 0.22659, + "grad_norm": 1.5116333980188945, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 22659 + }, + { + "epoch": 0.2266, + "grad_norm": 1.398692993616762, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 22660 + }, + { + "epoch": 0.22661, + "grad_norm": 1.2407464218984163, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22661 + }, + { + "epoch": 0.22662, + "grad_norm": 1.2793070754179567, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 22662 + }, + { + "epoch": 0.22663, + "grad_norm": 1.4347225094402987, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 22663 + }, + { + "epoch": 0.22664, + "grad_norm": 1.4493947841028387, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 22664 + }, + { + "epoch": 0.22665, + "grad_norm": 1.267139810879918, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 22665 + }, + { + "epoch": 0.22666, + "grad_norm": 1.466853579717287, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 22666 + }, + { + "epoch": 0.22667, + "grad_norm": 1.2689565681150594, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 22667 + }, + { + "epoch": 0.22668, + "grad_norm": 1.7870831055293204, + "learning_rate": 0.003, + "loss": 4.033, + "step": 22668 + }, + { + "epoch": 0.22669, + "grad_norm": 1.320122758108175, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22669 + }, + { + "epoch": 0.2267, + "grad_norm": 1.5005533778962779, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 22670 + }, + { + "epoch": 0.22671, + "grad_norm": 1.2797600664813926, + "learning_rate": 0.003, + "loss": 4.055, + "step": 22671 + }, + { + "epoch": 0.22672, + "grad_norm": 1.4017309372932885, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22672 + }, + { + "epoch": 0.22673, + "grad_norm": 1.471016236023522, + "learning_rate": 0.003, + "loss": 4.015, + "step": 22673 + }, + { + "epoch": 0.22674, + "grad_norm": 1.2984861405736587, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22674 + }, + { + "epoch": 0.22675, + "grad_norm": 1.4666226372746194, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 22675 + }, + { + "epoch": 0.22676, + "grad_norm": 1.057823960025185, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 22676 + }, + { + "epoch": 0.22677, + "grad_norm": 1.6549346121176827, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22677 + }, + { + "epoch": 0.22678, + "grad_norm": 1.1585263518139224, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 22678 + }, + { + "epoch": 0.22679, + "grad_norm": 1.4487039629449114, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22679 + }, + { + "epoch": 0.2268, + "grad_norm": 1.325337987181312, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 22680 + }, + { + "epoch": 0.22681, + "grad_norm": 1.4040369566590036, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 22681 + }, + { + "epoch": 0.22682, + "grad_norm": 1.462985041434129, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 22682 + }, + { + "epoch": 0.22683, + "grad_norm": 1.4942923170342381, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22683 + }, + { + "epoch": 0.22684, + "grad_norm": 1.364818885404756, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 22684 + }, + { + "epoch": 0.22685, + "grad_norm": 1.3381450883990378, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 22685 + }, + { + "epoch": 0.22686, + "grad_norm": 1.4139855403410388, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 22686 + }, + { + "epoch": 0.22687, + "grad_norm": 1.283220290729647, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 22687 + }, + { + "epoch": 0.22688, + "grad_norm": 1.1335905218354496, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 22688 + }, + { + "epoch": 0.22689, + "grad_norm": 1.5542724056246844, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 22689 + }, + { + "epoch": 0.2269, + "grad_norm": 1.1854055008264828, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 22690 + }, + { + "epoch": 0.22691, + "grad_norm": 1.671151129810973, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 22691 + }, + { + "epoch": 0.22692, + "grad_norm": 1.4417314469168165, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 22692 + }, + { + "epoch": 0.22693, + "grad_norm": 1.4484208424018798, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 22693 + }, + { + "epoch": 0.22694, + "grad_norm": 1.226954351094281, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22694 + }, + { + "epoch": 0.22695, + "grad_norm": 1.4044018349908938, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 22695 + }, + { + "epoch": 0.22696, + "grad_norm": 1.2895303456317728, + "learning_rate": 0.003, + "loss": 4.05, + "step": 22696 + }, + { + "epoch": 0.22697, + "grad_norm": 1.3062241257820577, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 22697 + }, + { + "epoch": 0.22698, + "grad_norm": 1.2504406893191973, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 22698 + }, + { + "epoch": 0.22699, + "grad_norm": 1.3097849809865796, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 22699 + }, + { + "epoch": 0.227, + "grad_norm": 1.4801030945943425, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 22700 + }, + { + "epoch": 0.22701, + "grad_norm": 1.4569176256875882, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 22701 + }, + { + "epoch": 0.22702, + "grad_norm": 1.2775531479377897, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 22702 + }, + { + "epoch": 0.22703, + "grad_norm": 1.2828919103497352, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 22703 + }, + { + "epoch": 0.22704, + "grad_norm": 1.4993856414685836, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22704 + }, + { + "epoch": 0.22705, + "grad_norm": 1.4399009328303836, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 22705 + }, + { + "epoch": 0.22706, + "grad_norm": 1.3271074450231812, + "learning_rate": 0.003, + "loss": 4.042, + "step": 22706 + }, + { + "epoch": 0.22707, + "grad_norm": 1.6489946598338783, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 22707 + }, + { + "epoch": 0.22708, + "grad_norm": 1.6158436493310295, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22708 + }, + { + "epoch": 0.22709, + "grad_norm": 1.2037217417560346, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 22709 + }, + { + "epoch": 0.2271, + "grad_norm": 1.5531011340860523, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 22710 + }, + { + "epoch": 0.22711, + "grad_norm": 1.425885017699002, + "learning_rate": 0.003, + "loss": 4.055, + "step": 22711 + }, + { + "epoch": 0.22712, + "grad_norm": 1.2002858912145136, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 22712 + }, + { + "epoch": 0.22713, + "grad_norm": 1.430470037719689, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 22713 + }, + { + "epoch": 0.22714, + "grad_norm": 1.636990043396574, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22714 + }, + { + "epoch": 0.22715, + "grad_norm": 1.1099921951607443, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22715 + }, + { + "epoch": 0.22716, + "grad_norm": 1.5754495429131838, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 22716 + }, + { + "epoch": 0.22717, + "grad_norm": 1.3833784390319153, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 22717 + }, + { + "epoch": 0.22718, + "grad_norm": 1.324141264588012, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 22718 + }, + { + "epoch": 0.22719, + "grad_norm": 1.6033767035520747, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 22719 + }, + { + "epoch": 0.2272, + "grad_norm": 1.1777043738733637, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 22720 + }, + { + "epoch": 0.22721, + "grad_norm": 1.2804792369525233, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22721 + }, + { + "epoch": 0.22722, + "grad_norm": 1.369117070641414, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 22722 + }, + { + "epoch": 0.22723, + "grad_norm": 1.3113031825756496, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 22723 + }, + { + "epoch": 0.22724, + "grad_norm": 1.505547392760889, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 22724 + }, + { + "epoch": 0.22725, + "grad_norm": 1.3722136634024067, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 22725 + }, + { + "epoch": 0.22726, + "grad_norm": 1.3270259799031803, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22726 + }, + { + "epoch": 0.22727, + "grad_norm": 1.4597975785251383, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 22727 + }, + { + "epoch": 0.22728, + "grad_norm": 1.4347900265936195, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 22728 + }, + { + "epoch": 0.22729, + "grad_norm": 1.784177762127377, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 22729 + }, + { + "epoch": 0.2273, + "grad_norm": 1.119019016203389, + "learning_rate": 0.003, + "loss": 4.066, + "step": 22730 + }, + { + "epoch": 0.22731, + "grad_norm": 1.298704994703112, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 22731 + }, + { + "epoch": 0.22732, + "grad_norm": 1.6145253889696385, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 22732 + }, + { + "epoch": 0.22733, + "grad_norm": 1.009332942569693, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 22733 + }, + { + "epoch": 0.22734, + "grad_norm": 1.8197940796945007, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 22734 + }, + { + "epoch": 0.22735, + "grad_norm": 1.278516691136997, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22735 + }, + { + "epoch": 0.22736, + "grad_norm": 1.3859433272846655, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 22736 + }, + { + "epoch": 0.22737, + "grad_norm": 1.5925513615482265, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 22737 + }, + { + "epoch": 0.22738, + "grad_norm": 1.21997029965742, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 22738 + }, + { + "epoch": 0.22739, + "grad_norm": 1.193074086408873, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 22739 + }, + { + "epoch": 0.2274, + "grad_norm": 1.5345763007239461, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 22740 + }, + { + "epoch": 0.22741, + "grad_norm": 1.3989942151178685, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 22741 + }, + { + "epoch": 0.22742, + "grad_norm": 1.1211151425682409, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 22742 + }, + { + "epoch": 0.22743, + "grad_norm": 1.5827480809997463, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 22743 + }, + { + "epoch": 0.22744, + "grad_norm": 1.0515175238725416, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 22744 + }, + { + "epoch": 0.22745, + "grad_norm": 1.5357721268409588, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22745 + }, + { + "epoch": 0.22746, + "grad_norm": 1.3106788831436889, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 22746 + }, + { + "epoch": 0.22747, + "grad_norm": 1.5928894097233948, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 22747 + }, + { + "epoch": 0.22748, + "grad_norm": 1.7929971445680843, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 22748 + }, + { + "epoch": 0.22749, + "grad_norm": 1.3772959045324669, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 22749 + }, + { + "epoch": 0.2275, + "grad_norm": 1.3321748543633538, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 22750 + }, + { + "epoch": 0.22751, + "grad_norm": 1.556758363868485, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22751 + }, + { + "epoch": 0.22752, + "grad_norm": 1.3206813624861218, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 22752 + }, + { + "epoch": 0.22753, + "grad_norm": 1.2152563756581984, + "learning_rate": 0.003, + "loss": 4.015, + "step": 22753 + }, + { + "epoch": 0.22754, + "grad_norm": 1.544337384684816, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 22754 + }, + { + "epoch": 0.22755, + "grad_norm": 1.1660014149356033, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 22755 + }, + { + "epoch": 0.22756, + "grad_norm": 1.4894832389062866, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 22756 + }, + { + "epoch": 0.22757, + "grad_norm": 1.1558412605789987, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 22757 + }, + { + "epoch": 0.22758, + "grad_norm": 2.0043005159119542, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 22758 + }, + { + "epoch": 0.22759, + "grad_norm": 1.319424743303632, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 22759 + }, + { + "epoch": 0.2276, + "grad_norm": 1.414897525724111, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 22760 + }, + { + "epoch": 0.22761, + "grad_norm": 1.2681178406931337, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 22761 + }, + { + "epoch": 0.22762, + "grad_norm": 1.6161084330047177, + "learning_rate": 0.003, + "loss": 4.098, + "step": 22762 + }, + { + "epoch": 0.22763, + "grad_norm": 1.321014957858152, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 22763 + }, + { + "epoch": 0.22764, + "grad_norm": 1.4416448047819612, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 22764 + }, + { + "epoch": 0.22765, + "grad_norm": 1.3369984308013854, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 22765 + }, + { + "epoch": 0.22766, + "grad_norm": 1.2179835583670722, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 22766 + }, + { + "epoch": 0.22767, + "grad_norm": 1.8279900236105129, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 22767 + }, + { + "epoch": 0.22768, + "grad_norm": 1.213602320810376, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 22768 + }, + { + "epoch": 0.22769, + "grad_norm": 1.706127453137258, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 22769 + }, + { + "epoch": 0.2277, + "grad_norm": 1.1029359724223382, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22770 + }, + { + "epoch": 0.22771, + "grad_norm": 1.1785664252188008, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 22771 + }, + { + "epoch": 0.22772, + "grad_norm": 1.541480008398517, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 22772 + }, + { + "epoch": 0.22773, + "grad_norm": 1.2234950246108043, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 22773 + }, + { + "epoch": 0.22774, + "grad_norm": 1.4002365824636416, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22774 + }, + { + "epoch": 0.22775, + "grad_norm": 1.192025795982403, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22775 + }, + { + "epoch": 0.22776, + "grad_norm": 1.5961151778908518, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 22776 + }, + { + "epoch": 0.22777, + "grad_norm": 1.2336588502501258, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 22777 + }, + { + "epoch": 0.22778, + "grad_norm": 1.5534815136535918, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 22778 + }, + { + "epoch": 0.22779, + "grad_norm": 1.39564969745224, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 22779 + }, + { + "epoch": 0.2278, + "grad_norm": 1.1800592622090422, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 22780 + }, + { + "epoch": 0.22781, + "grad_norm": 1.2848512272701038, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22781 + }, + { + "epoch": 0.22782, + "grad_norm": 1.693306261388428, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 22782 + }, + { + "epoch": 0.22783, + "grad_norm": 1.1690051634797456, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 22783 + }, + { + "epoch": 0.22784, + "grad_norm": 1.6415685237943014, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 22784 + }, + { + "epoch": 0.22785, + "grad_norm": 1.0812010529935254, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 22785 + }, + { + "epoch": 0.22786, + "grad_norm": 1.4193971463042083, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 22786 + }, + { + "epoch": 0.22787, + "grad_norm": 1.2659239259709194, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 22787 + }, + { + "epoch": 0.22788, + "grad_norm": 1.291019294829421, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 22788 + }, + { + "epoch": 0.22789, + "grad_norm": 1.3837939932139751, + "learning_rate": 0.003, + "loss": 4.078, + "step": 22789 + }, + { + "epoch": 0.2279, + "grad_norm": 1.1401013654455603, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 22790 + }, + { + "epoch": 0.22791, + "grad_norm": 1.7155477896803115, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 22791 + }, + { + "epoch": 0.22792, + "grad_norm": 1.440121875217581, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 22792 + }, + { + "epoch": 0.22793, + "grad_norm": 1.1993406304256435, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 22793 + }, + { + "epoch": 0.22794, + "grad_norm": 1.499868426529385, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22794 + }, + { + "epoch": 0.22795, + "grad_norm": 1.5788791898780075, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 22795 + }, + { + "epoch": 0.22796, + "grad_norm": 1.3204890194348837, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 22796 + }, + { + "epoch": 0.22797, + "grad_norm": 1.3316817258024336, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 22797 + }, + { + "epoch": 0.22798, + "grad_norm": 1.5794818947540428, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 22798 + }, + { + "epoch": 0.22799, + "grad_norm": 1.3802337991872735, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 22799 + }, + { + "epoch": 0.228, + "grad_norm": 1.1524532883825167, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 22800 + }, + { + "epoch": 0.22801, + "grad_norm": 1.5056397569716424, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 22801 + }, + { + "epoch": 0.22802, + "grad_norm": 1.2668787670242587, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 22802 + }, + { + "epoch": 0.22803, + "grad_norm": 1.547512171508173, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 22803 + }, + { + "epoch": 0.22804, + "grad_norm": 1.4285910030488986, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 22804 + }, + { + "epoch": 0.22805, + "grad_norm": 1.2549136389487845, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 22805 + }, + { + "epoch": 0.22806, + "grad_norm": 1.3894638291845278, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 22806 + }, + { + "epoch": 0.22807, + "grad_norm": 1.1302487368272556, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 22807 + }, + { + "epoch": 0.22808, + "grad_norm": 1.4340720715074096, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 22808 + }, + { + "epoch": 0.22809, + "grad_norm": 1.0800651849205487, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 22809 + }, + { + "epoch": 0.2281, + "grad_norm": 1.5043339147042853, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 22810 + }, + { + "epoch": 0.22811, + "grad_norm": 1.4439286772604305, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22811 + }, + { + "epoch": 0.22812, + "grad_norm": 1.295586844963577, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22812 + }, + { + "epoch": 0.22813, + "grad_norm": 1.256345973930996, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 22813 + }, + { + "epoch": 0.22814, + "grad_norm": 1.3967598449398202, + "learning_rate": 0.003, + "loss": 4.035, + "step": 22814 + }, + { + "epoch": 0.22815, + "grad_norm": 1.5467402999818534, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 22815 + }, + { + "epoch": 0.22816, + "grad_norm": 1.206409645763743, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 22816 + }, + { + "epoch": 0.22817, + "grad_norm": 1.5334760029471426, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 22817 + }, + { + "epoch": 0.22818, + "grad_norm": 1.3003110524721266, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 22818 + }, + { + "epoch": 0.22819, + "grad_norm": 1.2978256734104014, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 22819 + }, + { + "epoch": 0.2282, + "grad_norm": 1.43622507146669, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 22820 + }, + { + "epoch": 0.22821, + "grad_norm": 1.3712213718832809, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 22821 + }, + { + "epoch": 0.22822, + "grad_norm": 1.417984294755835, + "learning_rate": 0.003, + "loss": 4.049, + "step": 22822 + }, + { + "epoch": 0.22823, + "grad_norm": 1.4077668676520234, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22823 + }, + { + "epoch": 0.22824, + "grad_norm": 1.5500761369323117, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 22824 + }, + { + "epoch": 0.22825, + "grad_norm": 1.4119672397602403, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 22825 + }, + { + "epoch": 0.22826, + "grad_norm": 1.4647195023239734, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 22826 + }, + { + "epoch": 0.22827, + "grad_norm": 1.5921915457413027, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 22827 + }, + { + "epoch": 0.22828, + "grad_norm": 1.175582461663063, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 22828 + }, + { + "epoch": 0.22829, + "grad_norm": 1.4089405036060674, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 22829 + }, + { + "epoch": 0.2283, + "grad_norm": 1.297291736433117, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 22830 + }, + { + "epoch": 0.22831, + "grad_norm": 1.4566281352328887, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 22831 + }, + { + "epoch": 0.22832, + "grad_norm": 1.2910121694672738, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 22832 + }, + { + "epoch": 0.22833, + "grad_norm": 1.4999606670033645, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 22833 + }, + { + "epoch": 0.22834, + "grad_norm": 1.2919907599807343, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 22834 + }, + { + "epoch": 0.22835, + "grad_norm": 1.3725973490557473, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 22835 + }, + { + "epoch": 0.22836, + "grad_norm": 1.4844399503007777, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 22836 + }, + { + "epoch": 0.22837, + "grad_norm": 1.2469283102796385, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 22837 + }, + { + "epoch": 0.22838, + "grad_norm": 1.5636974982482015, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 22838 + }, + { + "epoch": 0.22839, + "grad_norm": 1.1210394448162195, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 22839 + }, + { + "epoch": 0.2284, + "grad_norm": 1.3892339988548474, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 22840 + }, + { + "epoch": 0.22841, + "grad_norm": 1.4362686688740154, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 22841 + }, + { + "epoch": 0.22842, + "grad_norm": 1.4266672801548619, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 22842 + }, + { + "epoch": 0.22843, + "grad_norm": 1.5466784780867424, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 22843 + }, + { + "epoch": 0.22844, + "grad_norm": 1.2903751481327705, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 22844 + }, + { + "epoch": 0.22845, + "grad_norm": 1.4654023642465581, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 22845 + }, + { + "epoch": 0.22846, + "grad_norm": 1.346459458656631, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 22846 + }, + { + "epoch": 0.22847, + "grad_norm": 1.3280473063892437, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 22847 + }, + { + "epoch": 0.22848, + "grad_norm": 1.4291368851027328, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 22848 + }, + { + "epoch": 0.22849, + "grad_norm": 1.4043713279974217, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 22849 + }, + { + "epoch": 0.2285, + "grad_norm": 1.3779039823011487, + "learning_rate": 0.003, + "loss": 4.054, + "step": 22850 + }, + { + "epoch": 0.22851, + "grad_norm": 1.3587430350694432, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 22851 + }, + { + "epoch": 0.22852, + "grad_norm": 1.6487730120652673, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 22852 + }, + { + "epoch": 0.22853, + "grad_norm": 1.4694544507663712, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 22853 + }, + { + "epoch": 0.22854, + "grad_norm": 1.12884649591588, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 22854 + }, + { + "epoch": 0.22855, + "grad_norm": 1.38024726019724, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 22855 + }, + { + "epoch": 0.22856, + "grad_norm": 1.5847228529615476, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 22856 + }, + { + "epoch": 0.22857, + "grad_norm": 1.3817997015632875, + "learning_rate": 0.003, + "loss": 4.037, + "step": 22857 + }, + { + "epoch": 0.22858, + "grad_norm": 1.2984459309332053, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 22858 + }, + { + "epoch": 0.22859, + "grad_norm": 1.3225040141194015, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 22859 + }, + { + "epoch": 0.2286, + "grad_norm": 1.3417509700768964, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 22860 + }, + { + "epoch": 0.22861, + "grad_norm": 1.379279945212339, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22861 + }, + { + "epoch": 0.22862, + "grad_norm": 1.4159060377000825, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 22862 + }, + { + "epoch": 0.22863, + "grad_norm": 1.3739216498139826, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 22863 + }, + { + "epoch": 0.22864, + "grad_norm": 1.3711404096302824, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 22864 + }, + { + "epoch": 0.22865, + "grad_norm": 1.3548666387373958, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 22865 + }, + { + "epoch": 0.22866, + "grad_norm": 1.4234263619205612, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 22866 + }, + { + "epoch": 0.22867, + "grad_norm": 1.4163917455959771, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 22867 + }, + { + "epoch": 0.22868, + "grad_norm": 1.3975264010223116, + "learning_rate": 0.003, + "loss": 4.086, + "step": 22868 + }, + { + "epoch": 0.22869, + "grad_norm": 1.3400257381170582, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22869 + }, + { + "epoch": 0.2287, + "grad_norm": 1.176772266744342, + "learning_rate": 0.003, + "loss": 4.046, + "step": 22870 + }, + { + "epoch": 0.22871, + "grad_norm": 1.2383970809001525, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 22871 + }, + { + "epoch": 0.22872, + "grad_norm": 1.333779816791978, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 22872 + }, + { + "epoch": 0.22873, + "grad_norm": 1.3122308334037585, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 22873 + }, + { + "epoch": 0.22874, + "grad_norm": 1.364839848854087, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 22874 + }, + { + "epoch": 0.22875, + "grad_norm": 1.60895472345263, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 22875 + }, + { + "epoch": 0.22876, + "grad_norm": 1.4546016585888952, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 22876 + }, + { + "epoch": 0.22877, + "grad_norm": 1.3148719840863317, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 22877 + }, + { + "epoch": 0.22878, + "grad_norm": 1.392800427880589, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 22878 + }, + { + "epoch": 0.22879, + "grad_norm": 1.2159669449941268, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 22879 + }, + { + "epoch": 0.2288, + "grad_norm": 1.3155450912631526, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 22880 + }, + { + "epoch": 0.22881, + "grad_norm": 1.3312969299174433, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 22881 + }, + { + "epoch": 0.22882, + "grad_norm": 1.3002076627978985, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 22882 + }, + { + "epoch": 0.22883, + "grad_norm": 1.2661962155110686, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 22883 + }, + { + "epoch": 0.22884, + "grad_norm": 1.3450381125574369, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 22884 + }, + { + "epoch": 0.22885, + "grad_norm": 1.339805467192508, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 22885 + }, + { + "epoch": 0.22886, + "grad_norm": 1.4809585998745207, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 22886 + }, + { + "epoch": 0.22887, + "grad_norm": 1.162420575641933, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 22887 + }, + { + "epoch": 0.22888, + "grad_norm": 1.7168073083852065, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 22888 + }, + { + "epoch": 0.22889, + "grad_norm": 1.3367044738475202, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 22889 + }, + { + "epoch": 0.2289, + "grad_norm": 1.5371849697118831, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 22890 + }, + { + "epoch": 0.22891, + "grad_norm": 1.5009762139201366, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 22891 + }, + { + "epoch": 0.22892, + "grad_norm": 1.3995811493563683, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 22892 + }, + { + "epoch": 0.22893, + "grad_norm": 1.283622543765794, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22893 + }, + { + "epoch": 0.22894, + "grad_norm": 1.3468349419979935, + "learning_rate": 0.003, + "loss": 4.085, + "step": 22894 + }, + { + "epoch": 0.22895, + "grad_norm": 1.408573251621135, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 22895 + }, + { + "epoch": 0.22896, + "grad_norm": 1.5397575419788314, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 22896 + }, + { + "epoch": 0.22897, + "grad_norm": 1.3886972589073308, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 22897 + }, + { + "epoch": 0.22898, + "grad_norm": 1.3145346513690945, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22898 + }, + { + "epoch": 0.22899, + "grad_norm": 1.518352083951177, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 22899 + }, + { + "epoch": 0.229, + "grad_norm": 1.1782434367507242, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 22900 + }, + { + "epoch": 0.22901, + "grad_norm": 1.5610356730680504, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 22901 + }, + { + "epoch": 0.22902, + "grad_norm": 1.614204415348391, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 22902 + }, + { + "epoch": 0.22903, + "grad_norm": 1.5406045784687927, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 22903 + }, + { + "epoch": 0.22904, + "grad_norm": 1.170225557948686, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 22904 + }, + { + "epoch": 0.22905, + "grad_norm": 1.4559568853599187, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22905 + }, + { + "epoch": 0.22906, + "grad_norm": 1.3330053767174777, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 22906 + }, + { + "epoch": 0.22907, + "grad_norm": 1.447820799841297, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 22907 + }, + { + "epoch": 0.22908, + "grad_norm": 1.5794080995014115, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 22908 + }, + { + "epoch": 0.22909, + "grad_norm": 1.3907820618217002, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 22909 + }, + { + "epoch": 0.2291, + "grad_norm": 1.425359578922196, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 22910 + }, + { + "epoch": 0.22911, + "grad_norm": 1.58183015644073, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 22911 + }, + { + "epoch": 0.22912, + "grad_norm": 1.2066358438817404, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 22912 + }, + { + "epoch": 0.22913, + "grad_norm": 1.419448658573045, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 22913 + }, + { + "epoch": 0.22914, + "grad_norm": 1.175171432981624, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 22914 + }, + { + "epoch": 0.22915, + "grad_norm": 1.235510844625249, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 22915 + }, + { + "epoch": 0.22916, + "grad_norm": 1.3219877756870189, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 22916 + }, + { + "epoch": 0.22917, + "grad_norm": 1.4428557821369838, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 22917 + }, + { + "epoch": 0.22918, + "grad_norm": 1.1858581525800913, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 22918 + }, + { + "epoch": 0.22919, + "grad_norm": 1.6051683298512007, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 22919 + }, + { + "epoch": 0.2292, + "grad_norm": 1.3032162782771097, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 22920 + }, + { + "epoch": 0.22921, + "grad_norm": 1.2802020113054942, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 22921 + }, + { + "epoch": 0.22922, + "grad_norm": 1.6529037593282858, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 22922 + }, + { + "epoch": 0.22923, + "grad_norm": 1.2700712794416136, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22923 + }, + { + "epoch": 0.22924, + "grad_norm": 1.4259915181989866, + "learning_rate": 0.003, + "loss": 4.043, + "step": 22924 + }, + { + "epoch": 0.22925, + "grad_norm": 1.4124012855645303, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 22925 + }, + { + "epoch": 0.22926, + "grad_norm": 1.364736100490055, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 22926 + }, + { + "epoch": 0.22927, + "grad_norm": 1.6581967917988738, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 22927 + }, + { + "epoch": 0.22928, + "grad_norm": 1.0919513674529981, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 22928 + }, + { + "epoch": 0.22929, + "grad_norm": 1.606412865866531, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 22929 + }, + { + "epoch": 0.2293, + "grad_norm": 1.2609937857636244, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 22930 + }, + { + "epoch": 0.22931, + "grad_norm": 1.7529582876795367, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 22931 + }, + { + "epoch": 0.22932, + "grad_norm": 1.4139012140387022, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 22932 + }, + { + "epoch": 0.22933, + "grad_norm": 1.566417900134426, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 22933 + }, + { + "epoch": 0.22934, + "grad_norm": 1.1962950032689532, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 22934 + }, + { + "epoch": 0.22935, + "grad_norm": 1.4383831293450509, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 22935 + }, + { + "epoch": 0.22936, + "grad_norm": 1.565994399753327, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 22936 + }, + { + "epoch": 0.22937, + "grad_norm": 1.0786662983194601, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 22937 + }, + { + "epoch": 0.22938, + "grad_norm": 1.4133924513402512, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 22938 + }, + { + "epoch": 0.22939, + "grad_norm": 1.084571229371841, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 22939 + }, + { + "epoch": 0.2294, + "grad_norm": 1.3965331813781878, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 22940 + }, + { + "epoch": 0.22941, + "grad_norm": 1.0325201682266654, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 22941 + }, + { + "epoch": 0.22942, + "grad_norm": 1.542890426626316, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 22942 + }, + { + "epoch": 0.22943, + "grad_norm": 1.1180506726697423, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 22943 + }, + { + "epoch": 0.22944, + "grad_norm": 1.491061165134877, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 22944 + }, + { + "epoch": 0.22945, + "grad_norm": 1.1811620917207506, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 22945 + }, + { + "epoch": 0.22946, + "grad_norm": 1.4797081927263973, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 22946 + }, + { + "epoch": 0.22947, + "grad_norm": 1.6598278116824179, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 22947 + }, + { + "epoch": 0.22948, + "grad_norm": 1.4760840712503893, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 22948 + }, + { + "epoch": 0.22949, + "grad_norm": 1.7641302893273025, + "learning_rate": 0.003, + "loss": 4.073, + "step": 22949 + }, + { + "epoch": 0.2295, + "grad_norm": 1.5219615885729352, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 22950 + }, + { + "epoch": 0.22951, + "grad_norm": 1.0801487313124043, + "learning_rate": 0.003, + "loss": 4.067, + "step": 22951 + }, + { + "epoch": 0.22952, + "grad_norm": 1.7030498238077243, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 22952 + }, + { + "epoch": 0.22953, + "grad_norm": 1.228314594691771, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 22953 + }, + { + "epoch": 0.22954, + "grad_norm": 1.3193889796150344, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 22954 + }, + { + "epoch": 0.22955, + "grad_norm": 1.3953681681989965, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 22955 + }, + { + "epoch": 0.22956, + "grad_norm": 1.3948156590524627, + "learning_rate": 0.003, + "loss": 4.061, + "step": 22956 + }, + { + "epoch": 0.22957, + "grad_norm": 1.212636409383338, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 22957 + }, + { + "epoch": 0.22958, + "grad_norm": 1.3611995845323064, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 22958 + }, + { + "epoch": 0.22959, + "grad_norm": 1.3668827353309014, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 22959 + }, + { + "epoch": 0.2296, + "grad_norm": 1.4707145337750223, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 22960 + }, + { + "epoch": 0.22961, + "grad_norm": 1.4328658070926283, + "learning_rate": 0.003, + "loss": 4.048, + "step": 22961 + }, + { + "epoch": 0.22962, + "grad_norm": 1.3462902085348085, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 22962 + }, + { + "epoch": 0.22963, + "grad_norm": 1.2726388097540406, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 22963 + }, + { + "epoch": 0.22964, + "grad_norm": 1.390529523454449, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 22964 + }, + { + "epoch": 0.22965, + "grad_norm": 1.341336906374233, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 22965 + }, + { + "epoch": 0.22966, + "grad_norm": 1.403688994902358, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 22966 + }, + { + "epoch": 0.22967, + "grad_norm": 1.2501517024262214, + "learning_rate": 0.003, + "loss": 4.074, + "step": 22967 + }, + { + "epoch": 0.22968, + "grad_norm": 1.2657217447996258, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 22968 + }, + { + "epoch": 0.22969, + "grad_norm": 1.4040568075913151, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 22969 + }, + { + "epoch": 0.2297, + "grad_norm": 1.3447201613229363, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 22970 + }, + { + "epoch": 0.22971, + "grad_norm": 1.512430617630756, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 22971 + }, + { + "epoch": 0.22972, + "grad_norm": 1.1426672822249624, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 22972 + }, + { + "epoch": 0.22973, + "grad_norm": 1.6517270229492123, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 22973 + }, + { + "epoch": 0.22974, + "grad_norm": 1.4281794579999305, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 22974 + }, + { + "epoch": 0.22975, + "grad_norm": 1.3955411968895046, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 22975 + }, + { + "epoch": 0.22976, + "grad_norm": 1.615919374855493, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 22976 + }, + { + "epoch": 0.22977, + "grad_norm": 1.2222241116553751, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 22977 + }, + { + "epoch": 0.22978, + "grad_norm": 1.6735990138092902, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 22978 + }, + { + "epoch": 0.22979, + "grad_norm": 1.2673683420332382, + "learning_rate": 0.003, + "loss": 4.054, + "step": 22979 + }, + { + "epoch": 0.2298, + "grad_norm": 1.4814718469317947, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 22980 + }, + { + "epoch": 0.22981, + "grad_norm": 1.350122169637812, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 22981 + }, + { + "epoch": 0.22982, + "grad_norm": 1.533561415066937, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 22982 + }, + { + "epoch": 0.22983, + "grad_norm": 1.2700255494324815, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 22983 + }, + { + "epoch": 0.22984, + "grad_norm": 1.4737021321360557, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 22984 + }, + { + "epoch": 0.22985, + "grad_norm": 1.174831944097366, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 22985 + }, + { + "epoch": 0.22986, + "grad_norm": 1.3997706918456425, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 22986 + }, + { + "epoch": 0.22987, + "grad_norm": 1.1127635429152034, + "learning_rate": 0.003, + "loss": 4.056, + "step": 22987 + }, + { + "epoch": 0.22988, + "grad_norm": 1.538052023411139, + "learning_rate": 0.003, + "loss": 4.058, + "step": 22988 + }, + { + "epoch": 0.22989, + "grad_norm": 1.1620311378531678, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 22989 + }, + { + "epoch": 0.2299, + "grad_norm": 1.2928872617805722, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 22990 + }, + { + "epoch": 0.22991, + "grad_norm": 1.3814882984970216, + "learning_rate": 0.003, + "loss": 4.082, + "step": 22991 + }, + { + "epoch": 0.22992, + "grad_norm": 1.6545724967470796, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 22992 + }, + { + "epoch": 0.22993, + "grad_norm": 1.6784809434047772, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 22993 + }, + { + "epoch": 0.22994, + "grad_norm": 1.4447138372507262, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 22994 + }, + { + "epoch": 0.22995, + "grad_norm": 1.3250502982354588, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 22995 + }, + { + "epoch": 0.22996, + "grad_norm": 1.801497709744894, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 22996 + }, + { + "epoch": 0.22997, + "grad_norm": 1.254220967590485, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 22997 + }, + { + "epoch": 0.22998, + "grad_norm": 1.7448802253970186, + "learning_rate": 0.003, + "loss": 4.079, + "step": 22998 + }, + { + "epoch": 0.22999, + "grad_norm": 1.1891084182034939, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 22999 + }, + { + "epoch": 0.23, + "grad_norm": 1.4394272234556105, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 23000 + }, + { + "epoch": 0.23001, + "grad_norm": 1.3168780938686961, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 23001 + }, + { + "epoch": 0.23002, + "grad_norm": 1.6278056183432175, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23002 + }, + { + "epoch": 0.23003, + "grad_norm": 1.2519319632700479, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23003 + }, + { + "epoch": 0.23004, + "grad_norm": 1.262104534876486, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23004 + }, + { + "epoch": 0.23005, + "grad_norm": 1.243282803795832, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 23005 + }, + { + "epoch": 0.23006, + "grad_norm": 1.4441124180883653, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23006 + }, + { + "epoch": 0.23007, + "grad_norm": 1.434945131236988, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 23007 + }, + { + "epoch": 0.23008, + "grad_norm": 1.4197288535722148, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 23008 + }, + { + "epoch": 0.23009, + "grad_norm": 1.3982519755985698, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23009 + }, + { + "epoch": 0.2301, + "grad_norm": 1.469397525657271, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 23010 + }, + { + "epoch": 0.23011, + "grad_norm": 1.3102803657496553, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23011 + }, + { + "epoch": 0.23012, + "grad_norm": 1.278005363063069, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 23012 + }, + { + "epoch": 0.23013, + "grad_norm": 1.3300809198050145, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23013 + }, + { + "epoch": 0.23014, + "grad_norm": 1.5934259801017507, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23014 + }, + { + "epoch": 0.23015, + "grad_norm": 1.196117130422267, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 23015 + }, + { + "epoch": 0.23016, + "grad_norm": 1.6251479607161354, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 23016 + }, + { + "epoch": 0.23017, + "grad_norm": 1.4486473889099827, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 23017 + }, + { + "epoch": 0.23018, + "grad_norm": 1.3693501972873945, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 23018 + }, + { + "epoch": 0.23019, + "grad_norm": 1.27548053234981, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 23019 + }, + { + "epoch": 0.2302, + "grad_norm": 1.3200342429436345, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 23020 + }, + { + "epoch": 0.23021, + "grad_norm": 1.5248744753660666, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 23021 + }, + { + "epoch": 0.23022, + "grad_norm": 1.158348168545915, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23022 + }, + { + "epoch": 0.23023, + "grad_norm": 1.9071166989661332, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 23023 + }, + { + "epoch": 0.23024, + "grad_norm": 1.1970788055193202, + "learning_rate": 0.003, + "loss": 4.024, + "step": 23024 + }, + { + "epoch": 0.23025, + "grad_norm": 1.2160555414860683, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23025 + }, + { + "epoch": 0.23026, + "grad_norm": 1.344484298259978, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 23026 + }, + { + "epoch": 0.23027, + "grad_norm": 1.5255176934262091, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23027 + }, + { + "epoch": 0.23028, + "grad_norm": 1.1681419830245958, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23028 + }, + { + "epoch": 0.23029, + "grad_norm": 1.394941371974529, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 23029 + }, + { + "epoch": 0.2303, + "grad_norm": 1.1122218157565824, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 23030 + }, + { + "epoch": 0.23031, + "grad_norm": 1.424333834084624, + "learning_rate": 0.003, + "loss": 4.052, + "step": 23031 + }, + { + "epoch": 0.23032, + "grad_norm": 1.5497192043982402, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 23032 + }, + { + "epoch": 0.23033, + "grad_norm": 1.5981620336165951, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23033 + }, + { + "epoch": 0.23034, + "grad_norm": 1.252140877826222, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 23034 + }, + { + "epoch": 0.23035, + "grad_norm": 1.401712325898123, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 23035 + }, + { + "epoch": 0.23036, + "grad_norm": 1.6003688749238476, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 23036 + }, + { + "epoch": 0.23037, + "grad_norm": 1.3118799647869948, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 23037 + }, + { + "epoch": 0.23038, + "grad_norm": 1.3595215125288516, + "learning_rate": 0.003, + "loss": 4.052, + "step": 23038 + }, + { + "epoch": 0.23039, + "grad_norm": 1.3915882753467457, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 23039 + }, + { + "epoch": 0.2304, + "grad_norm": 1.2549520097406257, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 23040 + }, + { + "epoch": 0.23041, + "grad_norm": 1.3439948582731243, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 23041 + }, + { + "epoch": 0.23042, + "grad_norm": 1.2951894916424498, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 23042 + }, + { + "epoch": 0.23043, + "grad_norm": 1.4622390794486304, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 23043 + }, + { + "epoch": 0.23044, + "grad_norm": 1.4544637414603367, + "learning_rate": 0.003, + "loss": 4.035, + "step": 23044 + }, + { + "epoch": 0.23045, + "grad_norm": 1.6228262982357775, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 23045 + }, + { + "epoch": 0.23046, + "grad_norm": 1.3595395138767727, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 23046 + }, + { + "epoch": 0.23047, + "grad_norm": 1.3348288664332195, + "learning_rate": 0.003, + "loss": 4.052, + "step": 23047 + }, + { + "epoch": 0.23048, + "grad_norm": 1.3961785392373596, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 23048 + }, + { + "epoch": 0.23049, + "grad_norm": 1.3894074047128306, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 23049 + }, + { + "epoch": 0.2305, + "grad_norm": 1.4851582462848276, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 23050 + }, + { + "epoch": 0.23051, + "grad_norm": 1.019450939862282, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 23051 + }, + { + "epoch": 0.23052, + "grad_norm": 1.4883474037119848, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 23052 + }, + { + "epoch": 0.23053, + "grad_norm": 1.1632807575688127, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 23053 + }, + { + "epoch": 0.23054, + "grad_norm": 1.7056925565619752, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 23054 + }, + { + "epoch": 0.23055, + "grad_norm": 1.3118817855524907, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 23055 + }, + { + "epoch": 0.23056, + "grad_norm": 1.523639956822756, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 23056 + }, + { + "epoch": 0.23057, + "grad_norm": 1.5461451176446483, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 23057 + }, + { + "epoch": 0.23058, + "grad_norm": 1.314297129018635, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 23058 + }, + { + "epoch": 0.23059, + "grad_norm": 1.533021870348019, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 23059 + }, + { + "epoch": 0.2306, + "grad_norm": 1.3037588670864102, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 23060 + }, + { + "epoch": 0.23061, + "grad_norm": 1.2486908498374083, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 23061 + }, + { + "epoch": 0.23062, + "grad_norm": 1.3545675816325136, + "learning_rate": 0.003, + "loss": 4.024, + "step": 23062 + }, + { + "epoch": 0.23063, + "grad_norm": 1.4238703700593431, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23063 + }, + { + "epoch": 0.23064, + "grad_norm": 1.260590410799713, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23064 + }, + { + "epoch": 0.23065, + "grad_norm": 1.5345428707631013, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23065 + }, + { + "epoch": 0.23066, + "grad_norm": 1.2083117870553177, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23066 + }, + { + "epoch": 0.23067, + "grad_norm": 1.6470988351741727, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23067 + }, + { + "epoch": 0.23068, + "grad_norm": 1.1630818070944566, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 23068 + }, + { + "epoch": 0.23069, + "grad_norm": 1.6895136508341708, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23069 + }, + { + "epoch": 0.2307, + "grad_norm": 1.25827195763974, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23070 + }, + { + "epoch": 0.23071, + "grad_norm": 1.410811113011283, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 23071 + }, + { + "epoch": 0.23072, + "grad_norm": 1.3395227633208335, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 23072 + }, + { + "epoch": 0.23073, + "grad_norm": 1.6835649017578809, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 23073 + }, + { + "epoch": 0.23074, + "grad_norm": 1.1112159433910216, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 23074 + }, + { + "epoch": 0.23075, + "grad_norm": 1.7472991571570606, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 23075 + }, + { + "epoch": 0.23076, + "grad_norm": 1.2145662256286143, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 23076 + }, + { + "epoch": 0.23077, + "grad_norm": 1.3231062274195922, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 23077 + }, + { + "epoch": 0.23078, + "grad_norm": 1.4200853882677293, + "learning_rate": 0.003, + "loss": 4.061, + "step": 23078 + }, + { + "epoch": 0.23079, + "grad_norm": 1.3677150569072831, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 23079 + }, + { + "epoch": 0.2308, + "grad_norm": 1.3676110212542516, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 23080 + }, + { + "epoch": 0.23081, + "grad_norm": 1.4137468889436755, + "learning_rate": 0.003, + "loss": 4.075, + "step": 23081 + }, + { + "epoch": 0.23082, + "grad_norm": 1.177594064115716, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 23082 + }, + { + "epoch": 0.23083, + "grad_norm": 1.5074512855725697, + "learning_rate": 0.003, + "loss": 4.052, + "step": 23083 + }, + { + "epoch": 0.23084, + "grad_norm": 1.3238048485838252, + "learning_rate": 0.003, + "loss": 4.059, + "step": 23084 + }, + { + "epoch": 0.23085, + "grad_norm": 1.4820599142183168, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 23085 + }, + { + "epoch": 0.23086, + "grad_norm": 1.3445418917237837, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23086 + }, + { + "epoch": 0.23087, + "grad_norm": 1.2570486310307096, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23087 + }, + { + "epoch": 0.23088, + "grad_norm": 1.436336772648397, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23088 + }, + { + "epoch": 0.23089, + "grad_norm": 1.1637129945506097, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 23089 + }, + { + "epoch": 0.2309, + "grad_norm": 1.6038071535423457, + "learning_rate": 0.003, + "loss": 4.039, + "step": 23090 + }, + { + "epoch": 0.23091, + "grad_norm": 1.3679500974508012, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 23091 + }, + { + "epoch": 0.23092, + "grad_norm": 1.6065914153722922, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 23092 + }, + { + "epoch": 0.23093, + "grad_norm": 1.1361562542425891, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23093 + }, + { + "epoch": 0.23094, + "grad_norm": 1.2749205805039103, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 23094 + }, + { + "epoch": 0.23095, + "grad_norm": 1.2988769294460014, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 23095 + }, + { + "epoch": 0.23096, + "grad_norm": 1.3538875084098738, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 23096 + }, + { + "epoch": 0.23097, + "grad_norm": 1.315085808107658, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 23097 + }, + { + "epoch": 0.23098, + "grad_norm": 1.4182471702035666, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23098 + }, + { + "epoch": 0.23099, + "grad_norm": 1.3566694693414556, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 23099 + }, + { + "epoch": 0.231, + "grad_norm": 1.2740335308945778, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23100 + }, + { + "epoch": 0.23101, + "grad_norm": 1.3224127108449382, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 23101 + }, + { + "epoch": 0.23102, + "grad_norm": 1.2037559306869496, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23102 + }, + { + "epoch": 0.23103, + "grad_norm": 1.9817962812717091, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 23103 + }, + { + "epoch": 0.23104, + "grad_norm": 1.2225055196978887, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 23104 + }, + { + "epoch": 0.23105, + "grad_norm": 1.5000579479300924, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 23105 + }, + { + "epoch": 0.23106, + "grad_norm": 1.5623842054630377, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 23106 + }, + { + "epoch": 0.23107, + "grad_norm": 1.1491403271789082, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 23107 + }, + { + "epoch": 0.23108, + "grad_norm": 1.2366893852890424, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23108 + }, + { + "epoch": 0.23109, + "grad_norm": 1.3032847184870682, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23109 + }, + { + "epoch": 0.2311, + "grad_norm": 1.2902243588838969, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23110 + }, + { + "epoch": 0.23111, + "grad_norm": 1.589132549648097, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23111 + }, + { + "epoch": 0.23112, + "grad_norm": 1.4206133407013364, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 23112 + }, + { + "epoch": 0.23113, + "grad_norm": 1.5505661618510957, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 23113 + }, + { + "epoch": 0.23114, + "grad_norm": 1.4278368033703972, + "learning_rate": 0.003, + "loss": 4.032, + "step": 23114 + }, + { + "epoch": 0.23115, + "grad_norm": 1.49234650598739, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 23115 + }, + { + "epoch": 0.23116, + "grad_norm": 1.1263272541904785, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 23116 + }, + { + "epoch": 0.23117, + "grad_norm": 1.5748913296976446, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 23117 + }, + { + "epoch": 0.23118, + "grad_norm": 1.1986532243015542, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 23118 + }, + { + "epoch": 0.23119, + "grad_norm": 1.4866078782519772, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 23119 + }, + { + "epoch": 0.2312, + "grad_norm": 1.122292274862112, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 23120 + }, + { + "epoch": 0.23121, + "grad_norm": 1.554413091342367, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23121 + }, + { + "epoch": 0.23122, + "grad_norm": 1.5887694598592788, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 23122 + }, + { + "epoch": 0.23123, + "grad_norm": 1.5626775116592848, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23123 + }, + { + "epoch": 0.23124, + "grad_norm": 1.395390976413349, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23124 + }, + { + "epoch": 0.23125, + "grad_norm": 1.360377731365266, + "learning_rate": 0.003, + "loss": 4.071, + "step": 23125 + }, + { + "epoch": 0.23126, + "grad_norm": 1.437002287152256, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 23126 + }, + { + "epoch": 0.23127, + "grad_norm": 1.2982733064086367, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 23127 + }, + { + "epoch": 0.23128, + "grad_norm": 1.2165050994472264, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 23128 + }, + { + "epoch": 0.23129, + "grad_norm": 1.484541274078872, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 23129 + }, + { + "epoch": 0.2313, + "grad_norm": 1.3126693478781408, + "learning_rate": 0.003, + "loss": 4.041, + "step": 23130 + }, + { + "epoch": 0.23131, + "grad_norm": 1.6188912436746827, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23131 + }, + { + "epoch": 0.23132, + "grad_norm": 0.9608496433254575, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 23132 + }, + { + "epoch": 0.23133, + "grad_norm": 1.4943965654805949, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 23133 + }, + { + "epoch": 0.23134, + "grad_norm": 1.6103602703785342, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 23134 + }, + { + "epoch": 0.23135, + "grad_norm": 1.4865506551792285, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 23135 + }, + { + "epoch": 0.23136, + "grad_norm": 1.4997999500696977, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 23136 + }, + { + "epoch": 0.23137, + "grad_norm": 1.2820657856007582, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 23137 + }, + { + "epoch": 0.23138, + "grad_norm": 1.1256573261519152, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 23138 + }, + { + "epoch": 0.23139, + "grad_norm": 1.4841915233502543, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 23139 + }, + { + "epoch": 0.2314, + "grad_norm": 1.4319827874652815, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 23140 + }, + { + "epoch": 0.23141, + "grad_norm": 1.649943338937173, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 23141 + }, + { + "epoch": 0.23142, + "grad_norm": 1.2014528257151893, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 23142 + }, + { + "epoch": 0.23143, + "grad_norm": 1.581294329507776, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 23143 + }, + { + "epoch": 0.23144, + "grad_norm": 1.3873995621357795, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23144 + }, + { + "epoch": 0.23145, + "grad_norm": 1.741841217549336, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 23145 + }, + { + "epoch": 0.23146, + "grad_norm": 1.0427242420105038, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 23146 + }, + { + "epoch": 0.23147, + "grad_norm": 1.2988860658950285, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 23147 + }, + { + "epoch": 0.23148, + "grad_norm": 1.224340615320788, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 23148 + }, + { + "epoch": 0.23149, + "grad_norm": 1.602496600517891, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 23149 + }, + { + "epoch": 0.2315, + "grad_norm": 1.1586195794603469, + "learning_rate": 0.003, + "loss": 4.057, + "step": 23150 + }, + { + "epoch": 0.23151, + "grad_norm": 1.6108298467371582, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23151 + }, + { + "epoch": 0.23152, + "grad_norm": 1.2344649499198077, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 23152 + }, + { + "epoch": 0.23153, + "grad_norm": 1.3800617425480675, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 23153 + }, + { + "epoch": 0.23154, + "grad_norm": 1.3758833483827941, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23154 + }, + { + "epoch": 0.23155, + "grad_norm": 1.5336184022458934, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 23155 + }, + { + "epoch": 0.23156, + "grad_norm": 1.370913566210212, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 23156 + }, + { + "epoch": 0.23157, + "grad_norm": 1.2596585620803824, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 23157 + }, + { + "epoch": 0.23158, + "grad_norm": 1.6195621423029691, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 23158 + }, + { + "epoch": 0.23159, + "grad_norm": 1.3963011760127697, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23159 + }, + { + "epoch": 0.2316, + "grad_norm": 1.427369476982424, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23160 + }, + { + "epoch": 0.23161, + "grad_norm": 1.6023339723130183, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 23161 + }, + { + "epoch": 0.23162, + "grad_norm": 1.3977799439014873, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 23162 + }, + { + "epoch": 0.23163, + "grad_norm": 1.1471109711506093, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 23163 + }, + { + "epoch": 0.23164, + "grad_norm": 1.4522518793825152, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 23164 + }, + { + "epoch": 0.23165, + "grad_norm": 1.2770313310326598, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23165 + }, + { + "epoch": 0.23166, + "grad_norm": 1.5415520221080945, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 23166 + }, + { + "epoch": 0.23167, + "grad_norm": 1.192403516683385, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23167 + }, + { + "epoch": 0.23168, + "grad_norm": 1.4369307951722878, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 23168 + }, + { + "epoch": 0.23169, + "grad_norm": 1.3428022781460058, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23169 + }, + { + "epoch": 0.2317, + "grad_norm": 1.4573974960234255, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 23170 + }, + { + "epoch": 0.23171, + "grad_norm": 1.3559560376745463, + "learning_rate": 0.003, + "loss": 4.06, + "step": 23171 + }, + { + "epoch": 0.23172, + "grad_norm": 1.290824733352766, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23172 + }, + { + "epoch": 0.23173, + "grad_norm": 1.4016706474171015, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23173 + }, + { + "epoch": 0.23174, + "grad_norm": 1.2672525599673197, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 23174 + }, + { + "epoch": 0.23175, + "grad_norm": 1.505934042929275, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 23175 + }, + { + "epoch": 0.23176, + "grad_norm": 1.4555875781210392, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 23176 + }, + { + "epoch": 0.23177, + "grad_norm": 1.4200226637953215, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23177 + }, + { + "epoch": 0.23178, + "grad_norm": 1.3511209540027005, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 23178 + }, + { + "epoch": 0.23179, + "grad_norm": 1.535066650779469, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23179 + }, + { + "epoch": 0.2318, + "grad_norm": 1.040862751941025, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23180 + }, + { + "epoch": 0.23181, + "grad_norm": 1.5481846575702471, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 23181 + }, + { + "epoch": 0.23182, + "grad_norm": 1.0461838488185702, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 23182 + }, + { + "epoch": 0.23183, + "grad_norm": 1.4058362774518611, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 23183 + }, + { + "epoch": 0.23184, + "grad_norm": 1.2187833980533604, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 23184 + }, + { + "epoch": 0.23185, + "grad_norm": 1.399106467132836, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 23185 + }, + { + "epoch": 0.23186, + "grad_norm": 1.4194814086397585, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 23186 + }, + { + "epoch": 0.23187, + "grad_norm": 1.7670927996006467, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 23187 + }, + { + "epoch": 0.23188, + "grad_norm": 1.3519469406355538, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 23188 + }, + { + "epoch": 0.23189, + "grad_norm": 1.5863237688284646, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 23189 + }, + { + "epoch": 0.2319, + "grad_norm": 1.419398324412055, + "learning_rate": 0.003, + "loss": 4.039, + "step": 23190 + }, + { + "epoch": 0.23191, + "grad_norm": 1.5093101000095643, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23191 + }, + { + "epoch": 0.23192, + "grad_norm": 1.3126495336212163, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 23192 + }, + { + "epoch": 0.23193, + "grad_norm": 1.5041173424379344, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 23193 + }, + { + "epoch": 0.23194, + "grad_norm": 1.2081253656400714, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 23194 + }, + { + "epoch": 0.23195, + "grad_norm": 1.6018004083847395, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 23195 + }, + { + "epoch": 0.23196, + "grad_norm": 1.1038385697305264, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 23196 + }, + { + "epoch": 0.23197, + "grad_norm": 1.4140298665465247, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 23197 + }, + { + "epoch": 0.23198, + "grad_norm": 1.2455816829302624, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 23198 + }, + { + "epoch": 0.23199, + "grad_norm": 1.293493351059705, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 23199 + }, + { + "epoch": 0.232, + "grad_norm": 1.4399840994518005, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23200 + }, + { + "epoch": 0.23201, + "grad_norm": 1.8503743395333405, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 23201 + }, + { + "epoch": 0.23202, + "grad_norm": 1.1866492005161333, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23202 + }, + { + "epoch": 0.23203, + "grad_norm": 1.5410260579949557, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 23203 + }, + { + "epoch": 0.23204, + "grad_norm": 1.2874196651584062, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23204 + }, + { + "epoch": 0.23205, + "grad_norm": 1.558736019283087, + "learning_rate": 0.003, + "loss": 4.067, + "step": 23205 + }, + { + "epoch": 0.23206, + "grad_norm": 1.1155157635475197, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 23206 + }, + { + "epoch": 0.23207, + "grad_norm": 1.6284966721910028, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 23207 + }, + { + "epoch": 0.23208, + "grad_norm": 1.4187063451456476, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 23208 + }, + { + "epoch": 0.23209, + "grad_norm": 1.428265865797565, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 23209 + }, + { + "epoch": 0.2321, + "grad_norm": 1.3748224800991098, + "learning_rate": 0.003, + "loss": 4.067, + "step": 23210 + }, + { + "epoch": 0.23211, + "grad_norm": 1.6895160144380876, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 23211 + }, + { + "epoch": 0.23212, + "grad_norm": 1.0415828454277831, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 23212 + }, + { + "epoch": 0.23213, + "grad_norm": 1.517687929459321, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 23213 + }, + { + "epoch": 0.23214, + "grad_norm": 1.2141206614580975, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 23214 + }, + { + "epoch": 0.23215, + "grad_norm": 1.3926596116826988, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23215 + }, + { + "epoch": 0.23216, + "grad_norm": 1.5309294204579795, + "learning_rate": 0.003, + "loss": 4.039, + "step": 23216 + }, + { + "epoch": 0.23217, + "grad_norm": 1.4013125958772263, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 23217 + }, + { + "epoch": 0.23218, + "grad_norm": 1.5455137081320858, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23218 + }, + { + "epoch": 0.23219, + "grad_norm": 1.3460428580661958, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23219 + }, + { + "epoch": 0.2322, + "grad_norm": 1.5581252782994168, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 23220 + }, + { + "epoch": 0.23221, + "grad_norm": 1.2972087312487386, + "learning_rate": 0.003, + "loss": 4.046, + "step": 23221 + }, + { + "epoch": 0.23222, + "grad_norm": 1.3393994009577652, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 23222 + }, + { + "epoch": 0.23223, + "grad_norm": 1.4743440652844046, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23223 + }, + { + "epoch": 0.23224, + "grad_norm": 1.2399382638443128, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 23224 + }, + { + "epoch": 0.23225, + "grad_norm": 1.4274691304657934, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23225 + }, + { + "epoch": 0.23226, + "grad_norm": 1.2153235809358585, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 23226 + }, + { + "epoch": 0.23227, + "grad_norm": 1.4742385055043994, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 23227 + }, + { + "epoch": 0.23228, + "grad_norm": 1.201506525492165, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23228 + }, + { + "epoch": 0.23229, + "grad_norm": 1.55977523256643, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23229 + }, + { + "epoch": 0.2323, + "grad_norm": 1.3237454623368943, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23230 + }, + { + "epoch": 0.23231, + "grad_norm": 1.4198197981554923, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 23231 + }, + { + "epoch": 0.23232, + "grad_norm": 1.200003748296258, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23232 + }, + { + "epoch": 0.23233, + "grad_norm": 1.364562004030165, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23233 + }, + { + "epoch": 0.23234, + "grad_norm": 1.2354970282473423, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 23234 + }, + { + "epoch": 0.23235, + "grad_norm": 1.6048801376951722, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 23235 + }, + { + "epoch": 0.23236, + "grad_norm": 1.144768589358892, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 23236 + }, + { + "epoch": 0.23237, + "grad_norm": 1.6146400549623448, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23237 + }, + { + "epoch": 0.23238, + "grad_norm": 1.2845332033949788, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 23238 + }, + { + "epoch": 0.23239, + "grad_norm": 1.5198619928502966, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 23239 + }, + { + "epoch": 0.2324, + "grad_norm": 1.4796043315387597, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 23240 + }, + { + "epoch": 0.23241, + "grad_norm": 1.3232047527976964, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23241 + }, + { + "epoch": 0.23242, + "grad_norm": 1.5909351606870596, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23242 + }, + { + "epoch": 0.23243, + "grad_norm": 1.295351887818341, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 23243 + }, + { + "epoch": 0.23244, + "grad_norm": 1.7195299870472625, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23244 + }, + { + "epoch": 0.23245, + "grad_norm": 1.1073968338402673, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23245 + }, + { + "epoch": 0.23246, + "grad_norm": 1.3603625036297036, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 23246 + }, + { + "epoch": 0.23247, + "grad_norm": 1.648006095868824, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 23247 + }, + { + "epoch": 0.23248, + "grad_norm": 1.0838339786857158, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23248 + }, + { + "epoch": 0.23249, + "grad_norm": 1.8525512974473193, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 23249 + }, + { + "epoch": 0.2325, + "grad_norm": 1.0995067287459384, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 23250 + }, + { + "epoch": 0.23251, + "grad_norm": 1.501016569416679, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 23251 + }, + { + "epoch": 0.23252, + "grad_norm": 1.1455643535743112, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23252 + }, + { + "epoch": 0.23253, + "grad_norm": 1.3518525655773297, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23253 + }, + { + "epoch": 0.23254, + "grad_norm": 1.4105817535103777, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 23254 + }, + { + "epoch": 0.23255, + "grad_norm": 1.169917404981231, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 23255 + }, + { + "epoch": 0.23256, + "grad_norm": 1.501594932765973, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 23256 + }, + { + "epoch": 0.23257, + "grad_norm": 1.295753874583137, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23257 + }, + { + "epoch": 0.23258, + "grad_norm": 2.039609126896833, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 23258 + }, + { + "epoch": 0.23259, + "grad_norm": 1.1959699585543309, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 23259 + }, + { + "epoch": 0.2326, + "grad_norm": 1.4181179411384424, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 23260 + }, + { + "epoch": 0.23261, + "grad_norm": 1.143132251049764, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 23261 + }, + { + "epoch": 0.23262, + "grad_norm": 1.2022803882982567, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 23262 + }, + { + "epoch": 0.23263, + "grad_norm": 1.4850903264816182, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 23263 + }, + { + "epoch": 0.23264, + "grad_norm": 1.1552796642446281, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 23264 + }, + { + "epoch": 0.23265, + "grad_norm": 1.4577981987212594, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 23265 + }, + { + "epoch": 0.23266, + "grad_norm": 1.5465790001414537, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23266 + }, + { + "epoch": 0.23267, + "grad_norm": 1.2356899901345653, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23267 + }, + { + "epoch": 0.23268, + "grad_norm": 1.4966204405392933, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 23268 + }, + { + "epoch": 0.23269, + "grad_norm": 1.1490991958453594, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23269 + }, + { + "epoch": 0.2327, + "grad_norm": 1.632207849955671, + "learning_rate": 0.003, + "loss": 4.082, + "step": 23270 + }, + { + "epoch": 0.23271, + "grad_norm": 1.1523017348795916, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23271 + }, + { + "epoch": 0.23272, + "grad_norm": 1.4942461998769034, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 23272 + }, + { + "epoch": 0.23273, + "grad_norm": 1.524916558472294, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 23273 + }, + { + "epoch": 0.23274, + "grad_norm": 1.3375224337440415, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 23274 + }, + { + "epoch": 0.23275, + "grad_norm": 1.640461322521533, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23275 + }, + { + "epoch": 0.23276, + "grad_norm": 1.376257675894278, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 23276 + }, + { + "epoch": 0.23277, + "grad_norm": 1.3720448538424719, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 23277 + }, + { + "epoch": 0.23278, + "grad_norm": 1.330648971532328, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 23278 + }, + { + "epoch": 0.23279, + "grad_norm": 1.4403782791511024, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 23279 + }, + { + "epoch": 0.2328, + "grad_norm": 1.3486738279518276, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 23280 + }, + { + "epoch": 0.23281, + "grad_norm": 1.5742117054538272, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 23281 + }, + { + "epoch": 0.23282, + "grad_norm": 1.360354340322476, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23282 + }, + { + "epoch": 0.23283, + "grad_norm": 1.573803399401895, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23283 + }, + { + "epoch": 0.23284, + "grad_norm": 1.380470787810996, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 23284 + }, + { + "epoch": 0.23285, + "grad_norm": 1.1481081913957012, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 23285 + }, + { + "epoch": 0.23286, + "grad_norm": 1.378426484420779, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 23286 + }, + { + "epoch": 0.23287, + "grad_norm": 1.4891976572396248, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23287 + }, + { + "epoch": 0.23288, + "grad_norm": 1.274367116949293, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 23288 + }, + { + "epoch": 0.23289, + "grad_norm": 1.2995741354141095, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 23289 + }, + { + "epoch": 0.2329, + "grad_norm": 1.3751707069043315, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 23290 + }, + { + "epoch": 0.23291, + "grad_norm": 1.1318703341623686, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 23291 + }, + { + "epoch": 0.23292, + "grad_norm": 1.3848642530821986, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 23292 + }, + { + "epoch": 0.23293, + "grad_norm": 1.3580898429055006, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23293 + }, + { + "epoch": 0.23294, + "grad_norm": 1.3435476391696566, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 23294 + }, + { + "epoch": 0.23295, + "grad_norm": 1.2607504054385046, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 23295 + }, + { + "epoch": 0.23296, + "grad_norm": 1.456243137673101, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 23296 + }, + { + "epoch": 0.23297, + "grad_norm": 1.3895092586474425, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23297 + }, + { + "epoch": 0.23298, + "grad_norm": 1.3805235974651768, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23298 + }, + { + "epoch": 0.23299, + "grad_norm": 1.4239808245437895, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23299 + }, + { + "epoch": 0.233, + "grad_norm": 1.530693199442528, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 23300 + }, + { + "epoch": 0.23301, + "grad_norm": 1.2764472475073199, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 23301 + }, + { + "epoch": 0.23302, + "grad_norm": 1.4247669730558707, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23302 + }, + { + "epoch": 0.23303, + "grad_norm": 1.2172673406242815, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 23303 + }, + { + "epoch": 0.23304, + "grad_norm": 1.4074013609359952, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23304 + }, + { + "epoch": 0.23305, + "grad_norm": 1.789281302975788, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 23305 + }, + { + "epoch": 0.23306, + "grad_norm": 1.1198034725501445, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 23306 + }, + { + "epoch": 0.23307, + "grad_norm": 1.5216483930759517, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 23307 + }, + { + "epoch": 0.23308, + "grad_norm": 1.3337214740061536, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 23308 + }, + { + "epoch": 0.23309, + "grad_norm": 1.379853589673503, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 23309 + }, + { + "epoch": 0.2331, + "grad_norm": 1.3763096443011047, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 23310 + }, + { + "epoch": 0.23311, + "grad_norm": 1.324091611104481, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23311 + }, + { + "epoch": 0.23312, + "grad_norm": 1.469904166008016, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23312 + }, + { + "epoch": 0.23313, + "grad_norm": 1.3533478110113275, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 23313 + }, + { + "epoch": 0.23314, + "grad_norm": 1.3417430157453967, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 23314 + }, + { + "epoch": 0.23315, + "grad_norm": 1.4765814465557936, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23315 + }, + { + "epoch": 0.23316, + "grad_norm": 1.5731600437764057, + "learning_rate": 0.003, + "loss": 4.079, + "step": 23316 + }, + { + "epoch": 0.23317, + "grad_norm": 1.4499194957120853, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 23317 + }, + { + "epoch": 0.23318, + "grad_norm": 1.461271136541248, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 23318 + }, + { + "epoch": 0.23319, + "grad_norm": 1.4191884278936935, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 23319 + }, + { + "epoch": 0.2332, + "grad_norm": 1.3395844738642793, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 23320 + }, + { + "epoch": 0.23321, + "grad_norm": 1.2927872666501299, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 23321 + }, + { + "epoch": 0.23322, + "grad_norm": 1.3284300302078813, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 23322 + }, + { + "epoch": 0.23323, + "grad_norm": 1.388065516587412, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 23323 + }, + { + "epoch": 0.23324, + "grad_norm": 1.2730160726961077, + "learning_rate": 0.003, + "loss": 4.065, + "step": 23324 + }, + { + "epoch": 0.23325, + "grad_norm": 1.6958348103450889, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 23325 + }, + { + "epoch": 0.23326, + "grad_norm": 1.218806575339292, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23326 + }, + { + "epoch": 0.23327, + "grad_norm": 1.4049694173549079, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 23327 + }, + { + "epoch": 0.23328, + "grad_norm": 1.3725084742900742, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 23328 + }, + { + "epoch": 0.23329, + "grad_norm": 1.4354762724536505, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 23329 + }, + { + "epoch": 0.2333, + "grad_norm": 1.2729372688233613, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 23330 + }, + { + "epoch": 0.23331, + "grad_norm": 1.3739779736425317, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 23331 + }, + { + "epoch": 0.23332, + "grad_norm": 1.2797412439650528, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 23332 + }, + { + "epoch": 0.23333, + "grad_norm": 1.1359805215254921, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 23333 + }, + { + "epoch": 0.23334, + "grad_norm": 1.4859752571815634, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 23334 + }, + { + "epoch": 0.23335, + "grad_norm": 1.2644116192581694, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 23335 + }, + { + "epoch": 0.23336, + "grad_norm": 1.7288192148564303, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23336 + }, + { + "epoch": 0.23337, + "grad_norm": 1.4522526863132863, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 23337 + }, + { + "epoch": 0.23338, + "grad_norm": 1.3746559438134074, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23338 + }, + { + "epoch": 0.23339, + "grad_norm": 1.4896974277707722, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 23339 + }, + { + "epoch": 0.2334, + "grad_norm": 1.115931811769656, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23340 + }, + { + "epoch": 0.23341, + "grad_norm": 1.4310574265050977, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 23341 + }, + { + "epoch": 0.23342, + "grad_norm": 1.242085521227107, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 23342 + }, + { + "epoch": 0.23343, + "grad_norm": 1.487393175276797, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 23343 + }, + { + "epoch": 0.23344, + "grad_norm": 1.1575642661862828, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23344 + }, + { + "epoch": 0.23345, + "grad_norm": 1.399116039326976, + "learning_rate": 0.003, + "loss": 4.053, + "step": 23345 + }, + { + "epoch": 0.23346, + "grad_norm": 1.2296764806064457, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 23346 + }, + { + "epoch": 0.23347, + "grad_norm": 1.4347408253791656, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 23347 + }, + { + "epoch": 0.23348, + "grad_norm": 1.7674929643590882, + "learning_rate": 0.003, + "loss": 4.079, + "step": 23348 + }, + { + "epoch": 0.23349, + "grad_norm": 1.3096582728277626, + "learning_rate": 0.003, + "loss": 4.018, + "step": 23349 + }, + { + "epoch": 0.2335, + "grad_norm": 1.5564933789167448, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 23350 + }, + { + "epoch": 0.23351, + "grad_norm": 1.3566410676533962, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 23351 + }, + { + "epoch": 0.23352, + "grad_norm": 1.4891266643637302, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 23352 + }, + { + "epoch": 0.23353, + "grad_norm": 1.2880685163878165, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 23353 + }, + { + "epoch": 0.23354, + "grad_norm": 1.4053611881721693, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 23354 + }, + { + "epoch": 0.23355, + "grad_norm": 1.4909643929490488, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 23355 + }, + { + "epoch": 0.23356, + "grad_norm": 1.305293227109484, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 23356 + }, + { + "epoch": 0.23357, + "grad_norm": 1.8078193423309594, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 23357 + }, + { + "epoch": 0.23358, + "grad_norm": 1.0891309342760909, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 23358 + }, + { + "epoch": 0.23359, + "grad_norm": 1.6236321420258064, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 23359 + }, + { + "epoch": 0.2336, + "grad_norm": 1.3116676798911029, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 23360 + }, + { + "epoch": 0.23361, + "grad_norm": 1.3478344689475437, + "learning_rate": 0.003, + "loss": 4.083, + "step": 23361 + }, + { + "epoch": 0.23362, + "grad_norm": 1.5128040823850337, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 23362 + }, + { + "epoch": 0.23363, + "grad_norm": 1.3541693429615973, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 23363 + }, + { + "epoch": 0.23364, + "grad_norm": 1.205625666698049, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 23364 + }, + { + "epoch": 0.23365, + "grad_norm": 1.3524023350765948, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 23365 + }, + { + "epoch": 0.23366, + "grad_norm": 1.3072262571821724, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 23366 + }, + { + "epoch": 0.23367, + "grad_norm": 1.222400294982097, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 23367 + }, + { + "epoch": 0.23368, + "grad_norm": 1.3857796257043968, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 23368 + }, + { + "epoch": 0.23369, + "grad_norm": 1.5039811217841277, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 23369 + }, + { + "epoch": 0.2337, + "grad_norm": 1.592357064418626, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23370 + }, + { + "epoch": 0.23371, + "grad_norm": 1.3307194504706745, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 23371 + }, + { + "epoch": 0.23372, + "grad_norm": 1.7265129605754046, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 23372 + }, + { + "epoch": 0.23373, + "grad_norm": 1.2791146988045174, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 23373 + }, + { + "epoch": 0.23374, + "grad_norm": 1.2658211887182347, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 23374 + }, + { + "epoch": 0.23375, + "grad_norm": 1.615940418471002, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 23375 + }, + { + "epoch": 0.23376, + "grad_norm": 1.2488341062337922, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23376 + }, + { + "epoch": 0.23377, + "grad_norm": 1.64317541007616, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23377 + }, + { + "epoch": 0.23378, + "grad_norm": 1.1794176936796756, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 23378 + }, + { + "epoch": 0.23379, + "grad_norm": 1.5511462295838057, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 23379 + }, + { + "epoch": 0.2338, + "grad_norm": 1.2101383495455327, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 23380 + }, + { + "epoch": 0.23381, + "grad_norm": 1.4288709848721513, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23381 + }, + { + "epoch": 0.23382, + "grad_norm": 1.145074381396933, + "learning_rate": 0.003, + "loss": 4.059, + "step": 23382 + }, + { + "epoch": 0.23383, + "grad_norm": 1.3238362213026706, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23383 + }, + { + "epoch": 0.23384, + "grad_norm": 1.5058887305002904, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23384 + }, + { + "epoch": 0.23385, + "grad_norm": 1.2497972129327182, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 23385 + }, + { + "epoch": 0.23386, + "grad_norm": 1.5232675410756014, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 23386 + }, + { + "epoch": 0.23387, + "grad_norm": 1.3182771221034846, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23387 + }, + { + "epoch": 0.23388, + "grad_norm": 1.6502823950348506, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 23388 + }, + { + "epoch": 0.23389, + "grad_norm": 1.513712969471974, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23389 + }, + { + "epoch": 0.2339, + "grad_norm": 1.123901985534363, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 23390 + }, + { + "epoch": 0.23391, + "grad_norm": 1.7476835044221801, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 23391 + }, + { + "epoch": 0.23392, + "grad_norm": 1.1481292324836674, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 23392 + }, + { + "epoch": 0.23393, + "grad_norm": 1.572492849609787, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 23393 + }, + { + "epoch": 0.23394, + "grad_norm": 1.3732563583360473, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23394 + }, + { + "epoch": 0.23395, + "grad_norm": 1.275735982851037, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 23395 + }, + { + "epoch": 0.23396, + "grad_norm": 1.380916624018907, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 23396 + }, + { + "epoch": 0.23397, + "grad_norm": 1.3210506383537612, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23397 + }, + { + "epoch": 0.23398, + "grad_norm": 1.3783554506449633, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23398 + }, + { + "epoch": 0.23399, + "grad_norm": 1.209440185167752, + "learning_rate": 0.003, + "loss": 4.08, + "step": 23399 + }, + { + "epoch": 0.234, + "grad_norm": 1.4810129691853908, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23400 + }, + { + "epoch": 0.23401, + "grad_norm": 1.3207923745865922, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 23401 + }, + { + "epoch": 0.23402, + "grad_norm": 1.4871117893603158, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 23402 + }, + { + "epoch": 0.23403, + "grad_norm": 1.2725559733965675, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 23403 + }, + { + "epoch": 0.23404, + "grad_norm": 1.6976241417923523, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 23404 + }, + { + "epoch": 0.23405, + "grad_norm": 1.2316079912852995, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 23405 + }, + { + "epoch": 0.23406, + "grad_norm": 1.379303060405435, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23406 + }, + { + "epoch": 0.23407, + "grad_norm": 1.5705799116220056, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23407 + }, + { + "epoch": 0.23408, + "grad_norm": 1.3894820157771752, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 23408 + }, + { + "epoch": 0.23409, + "grad_norm": 1.3482000288335323, + "learning_rate": 0.003, + "loss": 4.03, + "step": 23409 + }, + { + "epoch": 0.2341, + "grad_norm": 1.290661024844589, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23410 + }, + { + "epoch": 0.23411, + "grad_norm": 1.3901306049252855, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 23411 + }, + { + "epoch": 0.23412, + "grad_norm": 1.2377484876085936, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 23412 + }, + { + "epoch": 0.23413, + "grad_norm": 1.3366616591804248, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 23413 + }, + { + "epoch": 0.23414, + "grad_norm": 1.4331132534925277, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23414 + }, + { + "epoch": 0.23415, + "grad_norm": 1.3126353689905936, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23415 + }, + { + "epoch": 0.23416, + "grad_norm": 1.4781456314539463, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 23416 + }, + { + "epoch": 0.23417, + "grad_norm": 1.4561657212061085, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 23417 + }, + { + "epoch": 0.23418, + "grad_norm": 1.4285506376814623, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23418 + }, + { + "epoch": 0.23419, + "grad_norm": 1.370247952989015, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 23419 + }, + { + "epoch": 0.2342, + "grad_norm": 1.2749684833331565, + "learning_rate": 0.003, + "loss": 4.072, + "step": 23420 + }, + { + "epoch": 0.23421, + "grad_norm": 1.5250548904588213, + "learning_rate": 0.003, + "loss": 4.05, + "step": 23421 + }, + { + "epoch": 0.23422, + "grad_norm": 1.187391022985111, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23422 + }, + { + "epoch": 0.23423, + "grad_norm": 1.6147469186980699, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23423 + }, + { + "epoch": 0.23424, + "grad_norm": 1.0474255437095539, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 23424 + }, + { + "epoch": 0.23425, + "grad_norm": 1.7465891540813754, + "learning_rate": 0.003, + "loss": 4.084, + "step": 23425 + }, + { + "epoch": 0.23426, + "grad_norm": 1.2244151716829588, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 23426 + }, + { + "epoch": 0.23427, + "grad_norm": 1.3679252555123618, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23427 + }, + { + "epoch": 0.23428, + "grad_norm": 1.4915060486582232, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 23428 + }, + { + "epoch": 0.23429, + "grad_norm": 1.4721209458612163, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 23429 + }, + { + "epoch": 0.2343, + "grad_norm": 1.771403773607354, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 23430 + }, + { + "epoch": 0.23431, + "grad_norm": 1.2437319053095972, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23431 + }, + { + "epoch": 0.23432, + "grad_norm": 1.4602353551740295, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 23432 + }, + { + "epoch": 0.23433, + "grad_norm": 1.212846807146865, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23433 + }, + { + "epoch": 0.23434, + "grad_norm": 1.4052700980998136, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 23434 + }, + { + "epoch": 0.23435, + "grad_norm": 1.4041306961096618, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 23435 + }, + { + "epoch": 0.23436, + "grad_norm": 1.374945230547344, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 23436 + }, + { + "epoch": 0.23437, + "grad_norm": 1.6102002724792641, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 23437 + }, + { + "epoch": 0.23438, + "grad_norm": 1.3578097499182225, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 23438 + }, + { + "epoch": 0.23439, + "grad_norm": 1.2544931946268074, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 23439 + }, + { + "epoch": 0.2344, + "grad_norm": 1.435947825713841, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 23440 + }, + { + "epoch": 0.23441, + "grad_norm": 1.1505188288965753, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 23441 + }, + { + "epoch": 0.23442, + "grad_norm": 1.5148787047929175, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 23442 + }, + { + "epoch": 0.23443, + "grad_norm": 1.3109857199586687, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 23443 + }, + { + "epoch": 0.23444, + "grad_norm": 1.6302184377208593, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 23444 + }, + { + "epoch": 0.23445, + "grad_norm": 1.0958710629638364, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23445 + }, + { + "epoch": 0.23446, + "grad_norm": 1.6449759672702116, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 23446 + }, + { + "epoch": 0.23447, + "grad_norm": 1.1804608987453538, + "learning_rate": 0.003, + "loss": 4.064, + "step": 23447 + }, + { + "epoch": 0.23448, + "grad_norm": 1.5954880830901266, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 23448 + }, + { + "epoch": 0.23449, + "grad_norm": 1.4107087451702547, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 23449 + }, + { + "epoch": 0.2345, + "grad_norm": 1.6456755065393895, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 23450 + }, + { + "epoch": 0.23451, + "grad_norm": 1.4246644443506047, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 23451 + }, + { + "epoch": 0.23452, + "grad_norm": 1.2505246247554676, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23452 + }, + { + "epoch": 0.23453, + "grad_norm": 1.3988344508462989, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 23453 + }, + { + "epoch": 0.23454, + "grad_norm": 1.4150773598301845, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 23454 + }, + { + "epoch": 0.23455, + "grad_norm": 1.0643776084089818, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 23455 + }, + { + "epoch": 0.23456, + "grad_norm": 1.4902059310863867, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23456 + }, + { + "epoch": 0.23457, + "grad_norm": 1.2956298542669884, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23457 + }, + { + "epoch": 0.23458, + "grad_norm": 1.5402630597607896, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 23458 + }, + { + "epoch": 0.23459, + "grad_norm": 1.379775772744756, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23459 + }, + { + "epoch": 0.2346, + "grad_norm": 1.4297407268821414, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23460 + }, + { + "epoch": 0.23461, + "grad_norm": 1.167866307132083, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23461 + }, + { + "epoch": 0.23462, + "grad_norm": 1.5211096038407923, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 23462 + }, + { + "epoch": 0.23463, + "grad_norm": 1.0167766892965715, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23463 + }, + { + "epoch": 0.23464, + "grad_norm": 1.7284721847452755, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23464 + }, + { + "epoch": 0.23465, + "grad_norm": 1.2114230935605876, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 23465 + }, + { + "epoch": 0.23466, + "grad_norm": 1.5320697409612032, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23466 + }, + { + "epoch": 0.23467, + "grad_norm": 1.4745310006514158, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 23467 + }, + { + "epoch": 0.23468, + "grad_norm": 1.5868292231327064, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23468 + }, + { + "epoch": 0.23469, + "grad_norm": 1.4583457832566906, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 23469 + }, + { + "epoch": 0.2347, + "grad_norm": 1.2634409313940942, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 23470 + }, + { + "epoch": 0.23471, + "grad_norm": 1.7875095908216263, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 23471 + }, + { + "epoch": 0.23472, + "grad_norm": 1.331252310075803, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 23472 + }, + { + "epoch": 0.23473, + "grad_norm": 1.4729126453494954, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23473 + }, + { + "epoch": 0.23474, + "grad_norm": 1.6809016283735423, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 23474 + }, + { + "epoch": 0.23475, + "grad_norm": 1.1312357787690661, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 23475 + }, + { + "epoch": 0.23476, + "grad_norm": 1.216519874183779, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 23476 + }, + { + "epoch": 0.23477, + "grad_norm": 1.4284770336089858, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23477 + }, + { + "epoch": 0.23478, + "grad_norm": 1.5610670537451417, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 23478 + }, + { + "epoch": 0.23479, + "grad_norm": 1.3721097626999792, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 23479 + }, + { + "epoch": 0.2348, + "grad_norm": 1.2099749031404357, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 23480 + }, + { + "epoch": 0.23481, + "grad_norm": 1.4651803028437265, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 23481 + }, + { + "epoch": 0.23482, + "grad_norm": 1.0910292321437485, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23482 + }, + { + "epoch": 0.23483, + "grad_norm": 1.4626230500530066, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23483 + }, + { + "epoch": 0.23484, + "grad_norm": 1.2148026937237315, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 23484 + }, + { + "epoch": 0.23485, + "grad_norm": 1.4405634035580306, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 23485 + }, + { + "epoch": 0.23486, + "grad_norm": 1.5469730217530646, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23486 + }, + { + "epoch": 0.23487, + "grad_norm": 1.3719537071087946, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 23487 + }, + { + "epoch": 0.23488, + "grad_norm": 1.8411077853873343, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23488 + }, + { + "epoch": 0.23489, + "grad_norm": 1.1661254921174309, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 23489 + }, + { + "epoch": 0.2349, + "grad_norm": 1.317792002689811, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 23490 + }, + { + "epoch": 0.23491, + "grad_norm": 1.4869573300670418, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23491 + }, + { + "epoch": 0.23492, + "grad_norm": 1.290033490913107, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 23492 + }, + { + "epoch": 0.23493, + "grad_norm": 1.7556965055797964, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 23493 + }, + { + "epoch": 0.23494, + "grad_norm": 1.2639642209153696, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23494 + }, + { + "epoch": 0.23495, + "grad_norm": 1.5050470251308685, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 23495 + }, + { + "epoch": 0.23496, + "grad_norm": 1.269189098340193, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 23496 + }, + { + "epoch": 0.23497, + "grad_norm": 1.3678242319170064, + "learning_rate": 0.003, + "loss": 4.032, + "step": 23497 + }, + { + "epoch": 0.23498, + "grad_norm": 1.2846911244072163, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 23498 + }, + { + "epoch": 0.23499, + "grad_norm": 1.4047687043640837, + "learning_rate": 0.003, + "loss": 4.06, + "step": 23499 + }, + { + "epoch": 0.235, + "grad_norm": 1.4032908436053146, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 23500 + }, + { + "epoch": 0.23501, + "grad_norm": 1.4658861420478264, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23501 + }, + { + "epoch": 0.23502, + "grad_norm": 1.4790123184827366, + "learning_rate": 0.003, + "loss": 4.073, + "step": 23502 + }, + { + "epoch": 0.23503, + "grad_norm": 1.3415461361994145, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 23503 + }, + { + "epoch": 0.23504, + "grad_norm": 1.6424081294110777, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23504 + }, + { + "epoch": 0.23505, + "grad_norm": 1.2811723199997707, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 23505 + }, + { + "epoch": 0.23506, + "grad_norm": 1.304736790261805, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 23506 + }, + { + "epoch": 0.23507, + "grad_norm": 1.2371494547114326, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23507 + }, + { + "epoch": 0.23508, + "grad_norm": 1.536752203983268, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 23508 + }, + { + "epoch": 0.23509, + "grad_norm": 1.1185890320027325, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 23509 + }, + { + "epoch": 0.2351, + "grad_norm": 1.577830003306467, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23510 + }, + { + "epoch": 0.23511, + "grad_norm": 1.2613423722497226, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 23511 + }, + { + "epoch": 0.23512, + "grad_norm": 1.399843933996001, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 23512 + }, + { + "epoch": 0.23513, + "grad_norm": 1.6404305390784972, + "learning_rate": 0.003, + "loss": 4.057, + "step": 23513 + }, + { + "epoch": 0.23514, + "grad_norm": 1.069714512392399, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 23514 + }, + { + "epoch": 0.23515, + "grad_norm": 1.5734915129907068, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 23515 + }, + { + "epoch": 0.23516, + "grad_norm": 1.465511969395423, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 23516 + }, + { + "epoch": 0.23517, + "grad_norm": 1.3599987691119655, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23517 + }, + { + "epoch": 0.23518, + "grad_norm": 1.7162863697047077, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 23518 + }, + { + "epoch": 0.23519, + "grad_norm": 1.0657951760927105, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 23519 + }, + { + "epoch": 0.2352, + "grad_norm": 1.4633888179031667, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 23520 + }, + { + "epoch": 0.23521, + "grad_norm": 1.1581660133998444, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23521 + }, + { + "epoch": 0.23522, + "grad_norm": 1.245044795349148, + "learning_rate": 0.003, + "loss": 4.033, + "step": 23522 + }, + { + "epoch": 0.23523, + "grad_norm": 1.4553285563825764, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 23523 + }, + { + "epoch": 0.23524, + "grad_norm": 1.2864694674159676, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 23524 + }, + { + "epoch": 0.23525, + "grad_norm": 1.367469797784762, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23525 + }, + { + "epoch": 0.23526, + "grad_norm": 1.5847550660382819, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23526 + }, + { + "epoch": 0.23527, + "grad_norm": 1.1267107355595851, + "learning_rate": 0.003, + "loss": 4.051, + "step": 23527 + }, + { + "epoch": 0.23528, + "grad_norm": 1.1254508854791694, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 23528 + }, + { + "epoch": 0.23529, + "grad_norm": 1.499019604704434, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 23529 + }, + { + "epoch": 0.2353, + "grad_norm": 1.3221042060715784, + "learning_rate": 0.003, + "loss": 4.05, + "step": 23530 + }, + { + "epoch": 0.23531, + "grad_norm": 1.500259870860935, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23531 + }, + { + "epoch": 0.23532, + "grad_norm": 1.3774606507063807, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 23532 + }, + { + "epoch": 0.23533, + "grad_norm": 1.488951266188294, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 23533 + }, + { + "epoch": 0.23534, + "grad_norm": 1.2468561355909225, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 23534 + }, + { + "epoch": 0.23535, + "grad_norm": 1.4126856395387974, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23535 + }, + { + "epoch": 0.23536, + "grad_norm": 1.3393107042206556, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 23536 + }, + { + "epoch": 0.23537, + "grad_norm": 1.4675866749726905, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 23537 + }, + { + "epoch": 0.23538, + "grad_norm": 1.6632827300737214, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23538 + }, + { + "epoch": 0.23539, + "grad_norm": 1.2609099121791207, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23539 + }, + { + "epoch": 0.2354, + "grad_norm": 1.2448403795611518, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23540 + }, + { + "epoch": 0.23541, + "grad_norm": 1.644576973005076, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23541 + }, + { + "epoch": 0.23542, + "grad_norm": 1.1833374314746115, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23542 + }, + { + "epoch": 0.23543, + "grad_norm": 1.5165897407239832, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 23543 + }, + { + "epoch": 0.23544, + "grad_norm": 1.3440478216017815, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 23544 + }, + { + "epoch": 0.23545, + "grad_norm": 1.1487889574704027, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23545 + }, + { + "epoch": 0.23546, + "grad_norm": 1.551788326118772, + "learning_rate": 0.003, + "loss": 4.031, + "step": 23546 + }, + { + "epoch": 0.23547, + "grad_norm": 1.1445624709477686, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23547 + }, + { + "epoch": 0.23548, + "grad_norm": 1.4151480868609965, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23548 + }, + { + "epoch": 0.23549, + "grad_norm": 1.3238955882498833, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 23549 + }, + { + "epoch": 0.2355, + "grad_norm": 1.3394488361646577, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 23550 + }, + { + "epoch": 0.23551, + "grad_norm": 1.1350649434168805, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 23551 + }, + { + "epoch": 0.23552, + "grad_norm": 1.6686796802772081, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 23552 + }, + { + "epoch": 0.23553, + "grad_norm": 1.2501725999871147, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 23553 + }, + { + "epoch": 0.23554, + "grad_norm": 1.4502452262016516, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 23554 + }, + { + "epoch": 0.23555, + "grad_norm": 1.2627182999775564, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 23555 + }, + { + "epoch": 0.23556, + "grad_norm": 1.5865820054319144, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 23556 + }, + { + "epoch": 0.23557, + "grad_norm": 1.212166402140409, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23557 + }, + { + "epoch": 0.23558, + "grad_norm": 1.3941679528516313, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 23558 + }, + { + "epoch": 0.23559, + "grad_norm": 1.1675756947254337, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 23559 + }, + { + "epoch": 0.2356, + "grad_norm": 1.769189889288519, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 23560 + }, + { + "epoch": 0.23561, + "grad_norm": 1.2395302226234761, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23561 + }, + { + "epoch": 0.23562, + "grad_norm": 1.7037482148406256, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23562 + }, + { + "epoch": 0.23563, + "grad_norm": 1.3006614121391928, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 23563 + }, + { + "epoch": 0.23564, + "grad_norm": 1.7178535780951354, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 23564 + }, + { + "epoch": 0.23565, + "grad_norm": 1.2511621782698048, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 23565 + }, + { + "epoch": 0.23566, + "grad_norm": 1.5401991286903127, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 23566 + }, + { + "epoch": 0.23567, + "grad_norm": 1.127857362854384, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23567 + }, + { + "epoch": 0.23568, + "grad_norm": 1.4358199888136134, + "learning_rate": 0.003, + "loss": 4.059, + "step": 23568 + }, + { + "epoch": 0.23569, + "grad_norm": 1.4887109954764102, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 23569 + }, + { + "epoch": 0.2357, + "grad_norm": 1.0503727518447026, + "learning_rate": 0.003, + "loss": 4.047, + "step": 23570 + }, + { + "epoch": 0.23571, + "grad_norm": 1.646107473638904, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 23571 + }, + { + "epoch": 0.23572, + "grad_norm": 1.1858287766157665, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 23572 + }, + { + "epoch": 0.23573, + "grad_norm": 1.6227116841989064, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 23573 + }, + { + "epoch": 0.23574, + "grad_norm": 1.2894256998778117, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 23574 + }, + { + "epoch": 0.23575, + "grad_norm": 1.6329288467353305, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 23575 + }, + { + "epoch": 0.23576, + "grad_norm": 1.0797845743008918, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 23576 + }, + { + "epoch": 0.23577, + "grad_norm": 1.4520211993906205, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 23577 + }, + { + "epoch": 0.23578, + "grad_norm": 1.3284805516276323, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 23578 + }, + { + "epoch": 0.23579, + "grad_norm": 1.3618990962596662, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 23579 + }, + { + "epoch": 0.2358, + "grad_norm": 1.145652833990343, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 23580 + }, + { + "epoch": 0.23581, + "grad_norm": 1.514144047022743, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 23581 + }, + { + "epoch": 0.23582, + "grad_norm": 1.3285574107066616, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 23582 + }, + { + "epoch": 0.23583, + "grad_norm": 1.664632172552976, + "learning_rate": 0.003, + "loss": 4.039, + "step": 23583 + }, + { + "epoch": 0.23584, + "grad_norm": 1.5167512332361512, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23584 + }, + { + "epoch": 0.23585, + "grad_norm": 1.7125282195167086, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 23585 + }, + { + "epoch": 0.23586, + "grad_norm": 1.130931265910693, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23586 + }, + { + "epoch": 0.23587, + "grad_norm": 1.4755149683551616, + "learning_rate": 0.003, + "loss": 4.064, + "step": 23587 + }, + { + "epoch": 0.23588, + "grad_norm": 1.5338596236600828, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 23588 + }, + { + "epoch": 0.23589, + "grad_norm": 1.6684215415562105, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 23589 + }, + { + "epoch": 0.2359, + "grad_norm": 1.2147513914037273, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 23590 + }, + { + "epoch": 0.23591, + "grad_norm": 1.7053040574485951, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 23591 + }, + { + "epoch": 0.23592, + "grad_norm": 1.2224905964458432, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 23592 + }, + { + "epoch": 0.23593, + "grad_norm": 1.3395516006038173, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23593 + }, + { + "epoch": 0.23594, + "grad_norm": 1.0564481744672038, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 23594 + }, + { + "epoch": 0.23595, + "grad_norm": 1.5427023138415519, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 23595 + }, + { + "epoch": 0.23596, + "grad_norm": 1.1457494099338927, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 23596 + }, + { + "epoch": 0.23597, + "grad_norm": 1.4923339434817735, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 23597 + }, + { + "epoch": 0.23598, + "grad_norm": 1.3421229492117712, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 23598 + }, + { + "epoch": 0.23599, + "grad_norm": 1.3173561641282392, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 23599 + }, + { + "epoch": 0.236, + "grad_norm": 1.5246088725331406, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 23600 + }, + { + "epoch": 0.23601, + "grad_norm": 1.3983035419160177, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 23601 + }, + { + "epoch": 0.23602, + "grad_norm": 1.7450572159300748, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 23602 + }, + { + "epoch": 0.23603, + "grad_norm": 1.3108541616789215, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 23603 + }, + { + "epoch": 0.23604, + "grad_norm": 1.496498747979966, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23604 + }, + { + "epoch": 0.23605, + "grad_norm": 1.335927751276949, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 23605 + }, + { + "epoch": 0.23606, + "grad_norm": 1.376043104839073, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 23606 + }, + { + "epoch": 0.23607, + "grad_norm": 1.1983800794163468, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 23607 + }, + { + "epoch": 0.23608, + "grad_norm": 1.4116727332495218, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 23608 + }, + { + "epoch": 0.23609, + "grad_norm": 1.327947045088272, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 23609 + }, + { + "epoch": 0.2361, + "grad_norm": 1.3483372317783033, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 23610 + }, + { + "epoch": 0.23611, + "grad_norm": 1.1669563261202986, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 23611 + }, + { + "epoch": 0.23612, + "grad_norm": 1.6665118008124027, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 23612 + }, + { + "epoch": 0.23613, + "grad_norm": 1.3739468002730715, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23613 + }, + { + "epoch": 0.23614, + "grad_norm": 1.233809584077891, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 23614 + }, + { + "epoch": 0.23615, + "grad_norm": 1.572931478480147, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23615 + }, + { + "epoch": 0.23616, + "grad_norm": 1.1738648294276521, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23616 + }, + { + "epoch": 0.23617, + "grad_norm": 1.4122694266819866, + "learning_rate": 0.003, + "loss": 4.062, + "step": 23617 + }, + { + "epoch": 0.23618, + "grad_norm": 1.5351955831772994, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 23618 + }, + { + "epoch": 0.23619, + "grad_norm": 1.1480549871649792, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 23619 + }, + { + "epoch": 0.2362, + "grad_norm": 1.5886702088393145, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 23620 + }, + { + "epoch": 0.23621, + "grad_norm": 1.192709069786014, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 23621 + }, + { + "epoch": 0.23622, + "grad_norm": 1.3516921026147195, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 23622 + }, + { + "epoch": 0.23623, + "grad_norm": 1.658909331920533, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 23623 + }, + { + "epoch": 0.23624, + "grad_norm": 1.2287851205103206, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 23624 + }, + { + "epoch": 0.23625, + "grad_norm": 1.556326240733628, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23625 + }, + { + "epoch": 0.23626, + "grad_norm": 1.1704765854460446, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 23626 + }, + { + "epoch": 0.23627, + "grad_norm": 1.5200472734273278, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 23627 + }, + { + "epoch": 0.23628, + "grad_norm": 1.319051527043223, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 23628 + }, + { + "epoch": 0.23629, + "grad_norm": 1.416119703817279, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 23629 + }, + { + "epoch": 0.2363, + "grad_norm": 1.3133879921714262, + "learning_rate": 0.003, + "loss": 4.064, + "step": 23630 + }, + { + "epoch": 0.23631, + "grad_norm": 1.437023945086885, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 23631 + }, + { + "epoch": 0.23632, + "grad_norm": 1.1513591766520368, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23632 + }, + { + "epoch": 0.23633, + "grad_norm": 1.5193434001319606, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23633 + }, + { + "epoch": 0.23634, + "grad_norm": 1.1623495610484178, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 23634 + }, + { + "epoch": 0.23635, + "grad_norm": 1.5844147563868947, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 23635 + }, + { + "epoch": 0.23636, + "grad_norm": 1.2318230489096145, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 23636 + }, + { + "epoch": 0.23637, + "grad_norm": 1.432070004872038, + "learning_rate": 0.003, + "loss": 4.061, + "step": 23637 + }, + { + "epoch": 0.23638, + "grad_norm": 1.4564620691728094, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 23638 + }, + { + "epoch": 0.23639, + "grad_norm": 1.3721671151108517, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 23639 + }, + { + "epoch": 0.2364, + "grad_norm": 1.5991308480593305, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 23640 + }, + { + "epoch": 0.23641, + "grad_norm": 1.4340117617888781, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 23641 + }, + { + "epoch": 0.23642, + "grad_norm": 1.4102524067431634, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 23642 + }, + { + "epoch": 0.23643, + "grad_norm": 1.4785654806317532, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 23643 + }, + { + "epoch": 0.23644, + "grad_norm": 1.3197359125305035, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 23644 + }, + { + "epoch": 0.23645, + "grad_norm": 1.225739894839928, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 23645 + }, + { + "epoch": 0.23646, + "grad_norm": 1.549236165081848, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 23646 + }, + { + "epoch": 0.23647, + "grad_norm": 1.3554814343574306, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 23647 + }, + { + "epoch": 0.23648, + "grad_norm": 1.4858102492073069, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 23648 + }, + { + "epoch": 0.23649, + "grad_norm": 1.1460061042209861, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23649 + }, + { + "epoch": 0.2365, + "grad_norm": 1.4003691538584704, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23650 + }, + { + "epoch": 0.23651, + "grad_norm": 1.3420310754803368, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 23651 + }, + { + "epoch": 0.23652, + "grad_norm": 1.4427500744279627, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 23652 + }, + { + "epoch": 0.23653, + "grad_norm": 0.9981497956929709, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 23653 + }, + { + "epoch": 0.23654, + "grad_norm": 1.5434548242262909, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 23654 + }, + { + "epoch": 0.23655, + "grad_norm": 1.282829108353015, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23655 + }, + { + "epoch": 0.23656, + "grad_norm": 1.4051043368170837, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 23656 + }, + { + "epoch": 0.23657, + "grad_norm": 1.696644316180369, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23657 + }, + { + "epoch": 0.23658, + "grad_norm": 1.3853789495479925, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 23658 + }, + { + "epoch": 0.23659, + "grad_norm": 1.465777658921139, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23659 + }, + { + "epoch": 0.2366, + "grad_norm": 1.535637065863811, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 23660 + }, + { + "epoch": 0.23661, + "grad_norm": 1.2407916991708687, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 23661 + }, + { + "epoch": 0.23662, + "grad_norm": 1.3224541634369102, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23662 + }, + { + "epoch": 0.23663, + "grad_norm": 1.2735361590109953, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23663 + }, + { + "epoch": 0.23664, + "grad_norm": 1.1599813143934437, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 23664 + }, + { + "epoch": 0.23665, + "grad_norm": 1.5231624381414275, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 23665 + }, + { + "epoch": 0.23666, + "grad_norm": 1.3112522071404034, + "learning_rate": 0.003, + "loss": 4.115, + "step": 23666 + }, + { + "epoch": 0.23667, + "grad_norm": 1.6521600127322806, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 23667 + }, + { + "epoch": 0.23668, + "grad_norm": 1.424585037913398, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 23668 + }, + { + "epoch": 0.23669, + "grad_norm": 1.3715206623300467, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 23669 + }, + { + "epoch": 0.2367, + "grad_norm": 1.4647596590040015, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 23670 + }, + { + "epoch": 0.23671, + "grad_norm": 1.2028551548756445, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 23671 + }, + { + "epoch": 0.23672, + "grad_norm": 1.4944334500334975, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 23672 + }, + { + "epoch": 0.23673, + "grad_norm": 1.5016964241815314, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23673 + }, + { + "epoch": 0.23674, + "grad_norm": 1.3039651288047591, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23674 + }, + { + "epoch": 0.23675, + "grad_norm": 1.3626413074604664, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23675 + }, + { + "epoch": 0.23676, + "grad_norm": 1.4496866940216593, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 23676 + }, + { + "epoch": 0.23677, + "grad_norm": 1.183866697465119, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 23677 + }, + { + "epoch": 0.23678, + "grad_norm": 1.2890338921266171, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23678 + }, + { + "epoch": 0.23679, + "grad_norm": 1.1721886344381682, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 23679 + }, + { + "epoch": 0.2368, + "grad_norm": 1.6278739187063578, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 23680 + }, + { + "epoch": 0.23681, + "grad_norm": 1.2920736306502882, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 23681 + }, + { + "epoch": 0.23682, + "grad_norm": 1.5685630119267018, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 23682 + }, + { + "epoch": 0.23683, + "grad_norm": 1.3732226614011787, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 23683 + }, + { + "epoch": 0.23684, + "grad_norm": 1.5887688049102116, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23684 + }, + { + "epoch": 0.23685, + "grad_norm": 1.3730346857562286, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 23685 + }, + { + "epoch": 0.23686, + "grad_norm": 1.4689113599286796, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 23686 + }, + { + "epoch": 0.23687, + "grad_norm": 1.4095915160540724, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23687 + }, + { + "epoch": 0.23688, + "grad_norm": 1.257061519749002, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23688 + }, + { + "epoch": 0.23689, + "grad_norm": 1.5853550954893474, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 23689 + }, + { + "epoch": 0.2369, + "grad_norm": 1.1632440511212632, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 23690 + }, + { + "epoch": 0.23691, + "grad_norm": 1.4970128095142774, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 23691 + }, + { + "epoch": 0.23692, + "grad_norm": 1.3638489336068609, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23692 + }, + { + "epoch": 0.23693, + "grad_norm": 1.6696265294868635, + "learning_rate": 0.003, + "loss": 4.05, + "step": 23693 + }, + { + "epoch": 0.23694, + "grad_norm": 1.2733102986013913, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 23694 + }, + { + "epoch": 0.23695, + "grad_norm": 1.2757334531279756, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23695 + }, + { + "epoch": 0.23696, + "grad_norm": 1.592273124766524, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23696 + }, + { + "epoch": 0.23697, + "grad_norm": 1.2613170095067048, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 23697 + }, + { + "epoch": 0.23698, + "grad_norm": 1.4239489887722652, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 23698 + }, + { + "epoch": 0.23699, + "grad_norm": 1.5151470075544053, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 23699 + }, + { + "epoch": 0.237, + "grad_norm": 1.3190693526624828, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 23700 + }, + { + "epoch": 0.23701, + "grad_norm": 1.2625702194076969, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 23701 + }, + { + "epoch": 0.23702, + "grad_norm": 1.3163639093166508, + "learning_rate": 0.003, + "loss": 4.043, + "step": 23702 + }, + { + "epoch": 0.23703, + "grad_norm": 1.4585642788834488, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 23703 + }, + { + "epoch": 0.23704, + "grad_norm": 1.2463201430820265, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23704 + }, + { + "epoch": 0.23705, + "grad_norm": 1.399710281982181, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23705 + }, + { + "epoch": 0.23706, + "grad_norm": 1.2254111778445447, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 23706 + }, + { + "epoch": 0.23707, + "grad_norm": 1.4938337992820137, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 23707 + }, + { + "epoch": 0.23708, + "grad_norm": 1.471848255121138, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23708 + }, + { + "epoch": 0.23709, + "grad_norm": 1.6508759526546322, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 23709 + }, + { + "epoch": 0.2371, + "grad_norm": 1.3579753007102955, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 23710 + }, + { + "epoch": 0.23711, + "grad_norm": 1.2823517492476701, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23711 + }, + { + "epoch": 0.23712, + "grad_norm": 1.4418318137878419, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23712 + }, + { + "epoch": 0.23713, + "grad_norm": 1.3365225311086284, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 23713 + }, + { + "epoch": 0.23714, + "grad_norm": 1.4961142610676437, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 23714 + }, + { + "epoch": 0.23715, + "grad_norm": 1.2835543909532312, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 23715 + }, + { + "epoch": 0.23716, + "grad_norm": 1.4022699343194178, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 23716 + }, + { + "epoch": 0.23717, + "grad_norm": 1.2939316952306115, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 23717 + }, + { + "epoch": 0.23718, + "grad_norm": 1.4151173691120322, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 23718 + }, + { + "epoch": 0.23719, + "grad_norm": 1.460839796525741, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23719 + }, + { + "epoch": 0.2372, + "grad_norm": 1.5745424089000837, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 23720 + }, + { + "epoch": 0.23721, + "grad_norm": 1.3016430484884807, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 23721 + }, + { + "epoch": 0.23722, + "grad_norm": 1.4548967565753528, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 23722 + }, + { + "epoch": 0.23723, + "grad_norm": 1.308716035128893, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23723 + }, + { + "epoch": 0.23724, + "grad_norm": 1.5032293477234924, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 23724 + }, + { + "epoch": 0.23725, + "grad_norm": 1.0581030438208996, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23725 + }, + { + "epoch": 0.23726, + "grad_norm": 1.4633059635079007, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23726 + }, + { + "epoch": 0.23727, + "grad_norm": 1.2753225451283543, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 23727 + }, + { + "epoch": 0.23728, + "grad_norm": 1.3293575075976654, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23728 + }, + { + "epoch": 0.23729, + "grad_norm": 1.4792940776092705, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 23729 + }, + { + "epoch": 0.2373, + "grad_norm": 1.2234705801377, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23730 + }, + { + "epoch": 0.23731, + "grad_norm": 1.3607529940793255, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 23731 + }, + { + "epoch": 0.23732, + "grad_norm": 1.4660074124153655, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23732 + }, + { + "epoch": 0.23733, + "grad_norm": 1.3950089055417283, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 23733 + }, + { + "epoch": 0.23734, + "grad_norm": 1.4855922809082651, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 23734 + }, + { + "epoch": 0.23735, + "grad_norm": 1.1619249957598732, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 23735 + }, + { + "epoch": 0.23736, + "grad_norm": 1.515395474163773, + "learning_rate": 0.003, + "loss": 4.026, + "step": 23736 + }, + { + "epoch": 0.23737, + "grad_norm": 1.2710030388482008, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 23737 + }, + { + "epoch": 0.23738, + "grad_norm": 1.5125834984056037, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 23738 + }, + { + "epoch": 0.23739, + "grad_norm": 1.4517012803959244, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 23739 + }, + { + "epoch": 0.2374, + "grad_norm": 1.67274048647102, + "learning_rate": 0.003, + "loss": 4.061, + "step": 23740 + }, + { + "epoch": 0.23741, + "grad_norm": 1.292461838734173, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 23741 + }, + { + "epoch": 0.23742, + "grad_norm": 1.2820159233178188, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 23742 + }, + { + "epoch": 0.23743, + "grad_norm": 1.7082296540956725, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 23743 + }, + { + "epoch": 0.23744, + "grad_norm": 1.4325084137542587, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23744 + }, + { + "epoch": 0.23745, + "grad_norm": 1.4595301237648413, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 23745 + }, + { + "epoch": 0.23746, + "grad_norm": 1.4245005043264383, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 23746 + }, + { + "epoch": 0.23747, + "grad_norm": 1.304128097745544, + "learning_rate": 0.003, + "loss": 4.049, + "step": 23747 + }, + { + "epoch": 0.23748, + "grad_norm": 1.2330748303873393, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 23748 + }, + { + "epoch": 0.23749, + "grad_norm": 1.2902122167464298, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 23749 + }, + { + "epoch": 0.2375, + "grad_norm": 1.357336213341285, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23750 + }, + { + "epoch": 0.23751, + "grad_norm": 1.4101977731690332, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 23751 + }, + { + "epoch": 0.23752, + "grad_norm": 1.5671052386243283, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 23752 + }, + { + "epoch": 0.23753, + "grad_norm": 1.4561874458523836, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 23753 + }, + { + "epoch": 0.23754, + "grad_norm": 1.6496280310214197, + "learning_rate": 0.003, + "loss": 4.058, + "step": 23754 + }, + { + "epoch": 0.23755, + "grad_norm": 1.6085701026443695, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 23755 + }, + { + "epoch": 0.23756, + "grad_norm": 1.1582347693150215, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 23756 + }, + { + "epoch": 0.23757, + "grad_norm": 1.562853208179962, + "learning_rate": 0.003, + "loss": 4.057, + "step": 23757 + }, + { + "epoch": 0.23758, + "grad_norm": 1.2483496470144355, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 23758 + }, + { + "epoch": 0.23759, + "grad_norm": 1.3270835572321285, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 23759 + }, + { + "epoch": 0.2376, + "grad_norm": 1.3381027466485975, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 23760 + }, + { + "epoch": 0.23761, + "grad_norm": 1.3795053871405227, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23761 + }, + { + "epoch": 0.23762, + "grad_norm": 1.4713292606496422, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 23762 + }, + { + "epoch": 0.23763, + "grad_norm": 1.13310181388237, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 23763 + }, + { + "epoch": 0.23764, + "grad_norm": 1.4365549016004904, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 23764 + }, + { + "epoch": 0.23765, + "grad_norm": 1.1297188680876304, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 23765 + }, + { + "epoch": 0.23766, + "grad_norm": 1.4116402948496631, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 23766 + }, + { + "epoch": 0.23767, + "grad_norm": 1.3164830502265807, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 23767 + }, + { + "epoch": 0.23768, + "grad_norm": 1.2918661745473947, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 23768 + }, + { + "epoch": 0.23769, + "grad_norm": 1.5637865132804893, + "learning_rate": 0.003, + "loss": 4.056, + "step": 23769 + }, + { + "epoch": 0.2377, + "grad_norm": 1.3124835634283885, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23770 + }, + { + "epoch": 0.23771, + "grad_norm": 1.8116991522849801, + "learning_rate": 0.003, + "loss": 4.043, + "step": 23771 + }, + { + "epoch": 0.23772, + "grad_norm": 1.350272520718708, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 23772 + }, + { + "epoch": 0.23773, + "grad_norm": 1.4044574815881063, + "learning_rate": 0.003, + "loss": 4.036, + "step": 23773 + }, + { + "epoch": 0.23774, + "grad_norm": 1.4913650240195102, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 23774 + }, + { + "epoch": 0.23775, + "grad_norm": 1.3945788119693396, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 23775 + }, + { + "epoch": 0.23776, + "grad_norm": 1.4266427732762825, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 23776 + }, + { + "epoch": 0.23777, + "grad_norm": 1.3372979525506867, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 23777 + }, + { + "epoch": 0.23778, + "grad_norm": 1.3489631019606934, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23778 + }, + { + "epoch": 0.23779, + "grad_norm": 1.3421415503862049, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 23779 + }, + { + "epoch": 0.2378, + "grad_norm": 1.675774883772357, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23780 + }, + { + "epoch": 0.23781, + "grad_norm": 1.1704258782292785, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23781 + }, + { + "epoch": 0.23782, + "grad_norm": 1.4356232365793093, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 23782 + }, + { + "epoch": 0.23783, + "grad_norm": 1.364742331801125, + "learning_rate": 0.003, + "loss": 4.054, + "step": 23783 + }, + { + "epoch": 0.23784, + "grad_norm": 1.275911914822051, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 23784 + }, + { + "epoch": 0.23785, + "grad_norm": 1.4887763898630668, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 23785 + }, + { + "epoch": 0.23786, + "grad_norm": 1.331803732350777, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23786 + }, + { + "epoch": 0.23787, + "grad_norm": 1.3698619919514319, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23787 + }, + { + "epoch": 0.23788, + "grad_norm": 1.5433500274899132, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 23788 + }, + { + "epoch": 0.23789, + "grad_norm": 1.20440428110954, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23789 + }, + { + "epoch": 0.2379, + "grad_norm": 1.3825456419984896, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23790 + }, + { + "epoch": 0.23791, + "grad_norm": 1.5089479161473718, + "learning_rate": 0.003, + "loss": 4.027, + "step": 23791 + }, + { + "epoch": 0.23792, + "grad_norm": 1.4206593693592973, + "learning_rate": 0.003, + "loss": 4.036, + "step": 23792 + }, + { + "epoch": 0.23793, + "grad_norm": 1.5144565745928913, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 23793 + }, + { + "epoch": 0.23794, + "grad_norm": 1.3032704970872415, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 23794 + }, + { + "epoch": 0.23795, + "grad_norm": 1.3063518106157503, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 23795 + }, + { + "epoch": 0.23796, + "grad_norm": 1.4414575007722408, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 23796 + }, + { + "epoch": 0.23797, + "grad_norm": 1.446387987669235, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 23797 + }, + { + "epoch": 0.23798, + "grad_norm": 1.54667755487038, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 23798 + }, + { + "epoch": 0.23799, + "grad_norm": 1.266682287593711, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 23799 + }, + { + "epoch": 0.238, + "grad_norm": 1.312003475435821, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 23800 + }, + { + "epoch": 0.23801, + "grad_norm": 1.4088559393432345, + "learning_rate": 0.003, + "loss": 4.092, + "step": 23801 + }, + { + "epoch": 0.23802, + "grad_norm": 1.6838138418644581, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 23802 + }, + { + "epoch": 0.23803, + "grad_norm": 1.215442220818897, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23803 + }, + { + "epoch": 0.23804, + "grad_norm": 1.6589090290364392, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 23804 + }, + { + "epoch": 0.23805, + "grad_norm": 0.9436150220405386, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 23805 + }, + { + "epoch": 0.23806, + "grad_norm": 1.7691316974642126, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23806 + }, + { + "epoch": 0.23807, + "grad_norm": 1.477498995331464, + "learning_rate": 0.003, + "loss": 4.041, + "step": 23807 + }, + { + "epoch": 0.23808, + "grad_norm": 1.304295383588438, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 23808 + }, + { + "epoch": 0.23809, + "grad_norm": 1.614707971805161, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 23809 + }, + { + "epoch": 0.2381, + "grad_norm": 1.0162396579929045, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 23810 + }, + { + "epoch": 0.23811, + "grad_norm": 1.5559334686374409, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 23811 + }, + { + "epoch": 0.23812, + "grad_norm": 1.5982991422660973, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 23812 + }, + { + "epoch": 0.23813, + "grad_norm": 1.276807995696549, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23813 + }, + { + "epoch": 0.23814, + "grad_norm": 1.2772523958571014, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 23814 + }, + { + "epoch": 0.23815, + "grad_norm": 1.3132961813792767, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 23815 + }, + { + "epoch": 0.23816, + "grad_norm": 1.309818868243345, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 23816 + }, + { + "epoch": 0.23817, + "grad_norm": 1.4249962150156523, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 23817 + }, + { + "epoch": 0.23818, + "grad_norm": 1.2297845460360541, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 23818 + }, + { + "epoch": 0.23819, + "grad_norm": 1.4499472388015047, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 23819 + }, + { + "epoch": 0.2382, + "grad_norm": 1.6833199053099153, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23820 + }, + { + "epoch": 0.23821, + "grad_norm": 1.432908495373751, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 23821 + }, + { + "epoch": 0.23822, + "grad_norm": 1.0842052591519995, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 23822 + }, + { + "epoch": 0.23823, + "grad_norm": 1.4007313293813313, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 23823 + }, + { + "epoch": 0.23824, + "grad_norm": 1.6372371721978412, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 23824 + }, + { + "epoch": 0.23825, + "grad_norm": 1.1039519693304536, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 23825 + }, + { + "epoch": 0.23826, + "grad_norm": 1.7034974107340861, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23826 + }, + { + "epoch": 0.23827, + "grad_norm": 1.1050668562005355, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23827 + }, + { + "epoch": 0.23828, + "grad_norm": 1.5089155098277798, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23828 + }, + { + "epoch": 0.23829, + "grad_norm": 1.2680754184912342, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 23829 + }, + { + "epoch": 0.2383, + "grad_norm": 1.3294701242887348, + "learning_rate": 0.003, + "loss": 4.074, + "step": 23830 + }, + { + "epoch": 0.23831, + "grad_norm": 1.4721807512161498, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 23831 + }, + { + "epoch": 0.23832, + "grad_norm": 1.2474083008461583, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 23832 + }, + { + "epoch": 0.23833, + "grad_norm": 1.5024746972000815, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 23833 + }, + { + "epoch": 0.23834, + "grad_norm": 1.511506487520023, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 23834 + }, + { + "epoch": 0.23835, + "grad_norm": 1.6908269131540616, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 23835 + }, + { + "epoch": 0.23836, + "grad_norm": 1.3790388819882633, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 23836 + }, + { + "epoch": 0.23837, + "grad_norm": 1.3285309395252207, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 23837 + }, + { + "epoch": 0.23838, + "grad_norm": 1.2743391450563046, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 23838 + }, + { + "epoch": 0.23839, + "grad_norm": 1.6217943225615725, + "learning_rate": 0.003, + "loss": 4.063, + "step": 23839 + }, + { + "epoch": 0.2384, + "grad_norm": 1.1152541200610773, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 23840 + }, + { + "epoch": 0.23841, + "grad_norm": 1.6173808582178677, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 23841 + }, + { + "epoch": 0.23842, + "grad_norm": 1.1833309294393404, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 23842 + }, + { + "epoch": 0.23843, + "grad_norm": 1.7304148497630552, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 23843 + }, + { + "epoch": 0.23844, + "grad_norm": 1.2326838908494742, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 23844 + }, + { + "epoch": 0.23845, + "grad_norm": 1.47916800059893, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 23845 + }, + { + "epoch": 0.23846, + "grad_norm": 1.1302090760844912, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 23846 + }, + { + "epoch": 0.23847, + "grad_norm": 1.5664279076266978, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 23847 + }, + { + "epoch": 0.23848, + "grad_norm": 1.3737071177931959, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 23848 + }, + { + "epoch": 0.23849, + "grad_norm": 1.3654776330550957, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23849 + }, + { + "epoch": 0.2385, + "grad_norm": 1.5070315706528226, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 23850 + }, + { + "epoch": 0.23851, + "grad_norm": 1.3873273151489298, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 23851 + }, + { + "epoch": 0.23852, + "grad_norm": 1.33552502928097, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 23852 + }, + { + "epoch": 0.23853, + "grad_norm": 1.4264683902740416, + "learning_rate": 0.003, + "loss": 4.069, + "step": 23853 + }, + { + "epoch": 0.23854, + "grad_norm": 1.1861298323305576, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 23854 + }, + { + "epoch": 0.23855, + "grad_norm": 1.4632123550935487, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23855 + }, + { + "epoch": 0.23856, + "grad_norm": 1.064430085926257, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 23856 + }, + { + "epoch": 0.23857, + "grad_norm": 1.9269232944162906, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 23857 + }, + { + "epoch": 0.23858, + "grad_norm": 1.204356903464043, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 23858 + }, + { + "epoch": 0.23859, + "grad_norm": 1.9279509446903418, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 23859 + }, + { + "epoch": 0.2386, + "grad_norm": 1.3213780425828707, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 23860 + }, + { + "epoch": 0.23861, + "grad_norm": 1.2273116828630017, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 23861 + }, + { + "epoch": 0.23862, + "grad_norm": 1.4482013010905208, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23862 + }, + { + "epoch": 0.23863, + "grad_norm": 1.5255913592993389, + "learning_rate": 0.003, + "loss": 4.065, + "step": 23863 + }, + { + "epoch": 0.23864, + "grad_norm": 1.2667835745332046, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 23864 + }, + { + "epoch": 0.23865, + "grad_norm": 1.4638919124615997, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23865 + }, + { + "epoch": 0.23866, + "grad_norm": 1.6537109980718634, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 23866 + }, + { + "epoch": 0.23867, + "grad_norm": 1.144264157861541, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 23867 + }, + { + "epoch": 0.23868, + "grad_norm": 1.45679501662993, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23868 + }, + { + "epoch": 0.23869, + "grad_norm": 1.2851293505655565, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 23869 + }, + { + "epoch": 0.2387, + "grad_norm": 1.53453233126814, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 23870 + }, + { + "epoch": 0.23871, + "grad_norm": 1.2324819540586525, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 23871 + }, + { + "epoch": 0.23872, + "grad_norm": 1.4576342550602204, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 23872 + }, + { + "epoch": 0.23873, + "grad_norm": 1.3095921949579354, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 23873 + }, + { + "epoch": 0.23874, + "grad_norm": 1.6303465975553029, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23874 + }, + { + "epoch": 0.23875, + "grad_norm": 1.5271946277937312, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 23875 + }, + { + "epoch": 0.23876, + "grad_norm": 1.3545128670544826, + "learning_rate": 0.003, + "loss": 4.033, + "step": 23876 + }, + { + "epoch": 0.23877, + "grad_norm": 1.4882755565741435, + "learning_rate": 0.003, + "loss": 4.057, + "step": 23877 + }, + { + "epoch": 0.23878, + "grad_norm": 1.4555933551625828, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 23878 + }, + { + "epoch": 0.23879, + "grad_norm": 1.4442452418839198, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 23879 + }, + { + "epoch": 0.2388, + "grad_norm": 1.3838006327606363, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 23880 + }, + { + "epoch": 0.23881, + "grad_norm": 1.353045979434275, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 23881 + }, + { + "epoch": 0.23882, + "grad_norm": 1.2035868230847475, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 23882 + }, + { + "epoch": 0.23883, + "grad_norm": 1.3151637884857132, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 23883 + }, + { + "epoch": 0.23884, + "grad_norm": 1.3469918439785171, + "learning_rate": 0.003, + "loss": 4.049, + "step": 23884 + }, + { + "epoch": 0.23885, + "grad_norm": 1.6276977312150183, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 23885 + }, + { + "epoch": 0.23886, + "grad_norm": 0.9870363651716452, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 23886 + }, + { + "epoch": 0.23887, + "grad_norm": 1.687312232772964, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 23887 + }, + { + "epoch": 0.23888, + "grad_norm": 1.2719392440031376, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 23888 + }, + { + "epoch": 0.23889, + "grad_norm": 1.7781724738599882, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 23889 + }, + { + "epoch": 0.2389, + "grad_norm": 1.429271397424128, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 23890 + }, + { + "epoch": 0.23891, + "grad_norm": 1.2064605716494519, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 23891 + }, + { + "epoch": 0.23892, + "grad_norm": 1.574314808234865, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23892 + }, + { + "epoch": 0.23893, + "grad_norm": 1.2306207363246748, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 23893 + }, + { + "epoch": 0.23894, + "grad_norm": 1.4800015072971937, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 23894 + }, + { + "epoch": 0.23895, + "grad_norm": 1.1766164342089749, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 23895 + }, + { + "epoch": 0.23896, + "grad_norm": 2.038374218099904, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23896 + }, + { + "epoch": 0.23897, + "grad_norm": 1.209997615477808, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 23897 + }, + { + "epoch": 0.23898, + "grad_norm": 1.527994793198077, + "learning_rate": 0.003, + "loss": 4.073, + "step": 23898 + }, + { + "epoch": 0.23899, + "grad_norm": 1.5598152877419624, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 23899 + }, + { + "epoch": 0.239, + "grad_norm": 1.0747574357175296, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 23900 + }, + { + "epoch": 0.23901, + "grad_norm": 1.5540337504735566, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 23901 + }, + { + "epoch": 0.23902, + "grad_norm": 1.2864827706828081, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 23902 + }, + { + "epoch": 0.23903, + "grad_norm": 1.5679024715751169, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 23903 + }, + { + "epoch": 0.23904, + "grad_norm": 1.2940647261642273, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 23904 + }, + { + "epoch": 0.23905, + "grad_norm": 1.4599982353565195, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 23905 + }, + { + "epoch": 0.23906, + "grad_norm": 1.1975209818326515, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 23906 + }, + { + "epoch": 0.23907, + "grad_norm": 1.4749190018515188, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23907 + }, + { + "epoch": 0.23908, + "grad_norm": 1.3128009532242484, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 23908 + }, + { + "epoch": 0.23909, + "grad_norm": 1.3401151975136683, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 23909 + }, + { + "epoch": 0.2391, + "grad_norm": 1.759963670211569, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 23910 + }, + { + "epoch": 0.23911, + "grad_norm": 1.1500155017694818, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23911 + }, + { + "epoch": 0.23912, + "grad_norm": 1.3889662282538109, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 23912 + }, + { + "epoch": 0.23913, + "grad_norm": 1.5685165044324842, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23913 + }, + { + "epoch": 0.23914, + "grad_norm": 1.35687960144389, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 23914 + }, + { + "epoch": 0.23915, + "grad_norm": 1.5036410374791425, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23915 + }, + { + "epoch": 0.23916, + "grad_norm": 1.3774942534493801, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 23916 + }, + { + "epoch": 0.23917, + "grad_norm": 1.3622200146525703, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 23917 + }, + { + "epoch": 0.23918, + "grad_norm": 1.4785813270006374, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 23918 + }, + { + "epoch": 0.23919, + "grad_norm": 1.4166758330222942, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 23919 + }, + { + "epoch": 0.2392, + "grad_norm": 1.137203057813097, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 23920 + }, + { + "epoch": 0.23921, + "grad_norm": 1.4881849967255394, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 23921 + }, + { + "epoch": 0.23922, + "grad_norm": 1.2638340477372327, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 23922 + }, + { + "epoch": 0.23923, + "grad_norm": 1.232610847895491, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 23923 + }, + { + "epoch": 0.23924, + "grad_norm": 1.428442090677148, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23924 + }, + { + "epoch": 0.23925, + "grad_norm": 1.229763622838009, + "learning_rate": 0.003, + "loss": 4.041, + "step": 23925 + }, + { + "epoch": 0.23926, + "grad_norm": 1.5704113702178863, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 23926 + }, + { + "epoch": 0.23927, + "grad_norm": 1.3759363736172487, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 23927 + }, + { + "epoch": 0.23928, + "grad_norm": 1.479622646250979, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 23928 + }, + { + "epoch": 0.23929, + "grad_norm": 1.5193335088531157, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 23929 + }, + { + "epoch": 0.2393, + "grad_norm": 1.2385853878345658, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 23930 + }, + { + "epoch": 0.23931, + "grad_norm": 1.3301591585359105, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 23931 + }, + { + "epoch": 0.23932, + "grad_norm": 1.4967629933181466, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 23932 + }, + { + "epoch": 0.23933, + "grad_norm": 1.239061230910692, + "learning_rate": 0.003, + "loss": 4.033, + "step": 23933 + }, + { + "epoch": 0.23934, + "grad_norm": 1.5911761601168937, + "learning_rate": 0.003, + "loss": 4.028, + "step": 23934 + }, + { + "epoch": 0.23935, + "grad_norm": 1.426816165529343, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 23935 + }, + { + "epoch": 0.23936, + "grad_norm": 1.2027913915117059, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 23936 + }, + { + "epoch": 0.23937, + "grad_norm": 1.4237114034771414, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 23937 + }, + { + "epoch": 0.23938, + "grad_norm": 1.3361100789464557, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 23938 + }, + { + "epoch": 0.23939, + "grad_norm": 1.7015890090381045, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 23939 + }, + { + "epoch": 0.2394, + "grad_norm": 1.1599771038186024, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 23940 + }, + { + "epoch": 0.23941, + "grad_norm": 1.389903527458184, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 23941 + }, + { + "epoch": 0.23942, + "grad_norm": 1.2259445035966947, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 23942 + }, + { + "epoch": 0.23943, + "grad_norm": 1.3392741013180798, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 23943 + }, + { + "epoch": 0.23944, + "grad_norm": 1.2351941856468855, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 23944 + }, + { + "epoch": 0.23945, + "grad_norm": 1.5485213693209419, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 23945 + }, + { + "epoch": 0.23946, + "grad_norm": 1.2991648162126856, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 23946 + }, + { + "epoch": 0.23947, + "grad_norm": 1.4474743768636082, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 23947 + }, + { + "epoch": 0.23948, + "grad_norm": 1.3183307411077938, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 23948 + }, + { + "epoch": 0.23949, + "grad_norm": 1.3685403069930697, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 23949 + }, + { + "epoch": 0.2395, + "grad_norm": 1.3755717188009384, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 23950 + }, + { + "epoch": 0.23951, + "grad_norm": 1.6721210950286103, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 23951 + }, + { + "epoch": 0.23952, + "grad_norm": 1.3083439455606527, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 23952 + }, + { + "epoch": 0.23953, + "grad_norm": 1.4422213495504526, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 23953 + }, + { + "epoch": 0.23954, + "grad_norm": 1.387026715996028, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 23954 + }, + { + "epoch": 0.23955, + "grad_norm": 1.440346263084508, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 23955 + }, + { + "epoch": 0.23956, + "grad_norm": 1.590818189868967, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 23956 + }, + { + "epoch": 0.23957, + "grad_norm": 1.3889756185660593, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 23957 + }, + { + "epoch": 0.23958, + "grad_norm": 1.0926375524677692, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 23958 + }, + { + "epoch": 0.23959, + "grad_norm": 1.558191918429594, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23959 + }, + { + "epoch": 0.2396, + "grad_norm": 1.5299441214630756, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 23960 + }, + { + "epoch": 0.23961, + "grad_norm": 1.239989674867137, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 23961 + }, + { + "epoch": 0.23962, + "grad_norm": 1.6305351077706236, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 23962 + }, + { + "epoch": 0.23963, + "grad_norm": 1.1062324901786194, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23963 + }, + { + "epoch": 0.23964, + "grad_norm": 1.4694025967846787, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 23964 + }, + { + "epoch": 0.23965, + "grad_norm": 1.333277997412035, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23965 + }, + { + "epoch": 0.23966, + "grad_norm": 1.5332268470613066, + "learning_rate": 0.003, + "loss": 4.08, + "step": 23966 + }, + { + "epoch": 0.23967, + "grad_norm": 1.2938365569303085, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 23967 + }, + { + "epoch": 0.23968, + "grad_norm": 1.1061559486835393, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 23968 + }, + { + "epoch": 0.23969, + "grad_norm": 1.4417066760279533, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23969 + }, + { + "epoch": 0.2397, + "grad_norm": 1.0927795175523582, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 23970 + }, + { + "epoch": 0.23971, + "grad_norm": 1.5746775711428718, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 23971 + }, + { + "epoch": 0.23972, + "grad_norm": 1.4299606579525728, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 23972 + }, + { + "epoch": 0.23973, + "grad_norm": 1.6227309934797456, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 23973 + }, + { + "epoch": 0.23974, + "grad_norm": 1.517868044457317, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 23974 + }, + { + "epoch": 0.23975, + "grad_norm": 1.5062625468166382, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 23975 + }, + { + "epoch": 0.23976, + "grad_norm": 1.6528882048371647, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 23976 + }, + { + "epoch": 0.23977, + "grad_norm": 1.2776247453207008, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 23977 + }, + { + "epoch": 0.23978, + "grad_norm": 1.4202344502240514, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 23978 + }, + { + "epoch": 0.23979, + "grad_norm": 1.168376450636879, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 23979 + }, + { + "epoch": 0.2398, + "grad_norm": 1.3600901694620695, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 23980 + }, + { + "epoch": 0.23981, + "grad_norm": 1.3406922849924334, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 23981 + }, + { + "epoch": 0.23982, + "grad_norm": 1.4085494383855803, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 23982 + }, + { + "epoch": 0.23983, + "grad_norm": 1.1953433239102236, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 23983 + }, + { + "epoch": 0.23984, + "grad_norm": 1.7073756513123728, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 23984 + }, + { + "epoch": 0.23985, + "grad_norm": 1.1673486075400417, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 23985 + }, + { + "epoch": 0.23986, + "grad_norm": 1.6603190878239664, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 23986 + }, + { + "epoch": 0.23987, + "grad_norm": 1.2459795657565698, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 23987 + }, + { + "epoch": 0.23988, + "grad_norm": 1.2623862909418935, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 23988 + }, + { + "epoch": 0.23989, + "grad_norm": 1.4939180244219257, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 23989 + }, + { + "epoch": 0.2399, + "grad_norm": 1.4221846092754766, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 23990 + }, + { + "epoch": 0.23991, + "grad_norm": 1.3174146944497975, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 23991 + }, + { + "epoch": 0.23992, + "grad_norm": 1.231965197781564, + "learning_rate": 0.003, + "loss": 4.048, + "step": 23992 + }, + { + "epoch": 0.23993, + "grad_norm": 1.4611481267350466, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 23993 + }, + { + "epoch": 0.23994, + "grad_norm": 1.4764328218480265, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 23994 + }, + { + "epoch": 0.23995, + "grad_norm": 1.2652128922069787, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 23995 + }, + { + "epoch": 0.23996, + "grad_norm": 1.6664184416192886, + "learning_rate": 0.003, + "loss": 4.065, + "step": 23996 + }, + { + "epoch": 0.23997, + "grad_norm": 1.2647859807629176, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 23997 + }, + { + "epoch": 0.23998, + "grad_norm": 1.6670294215699575, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 23998 + }, + { + "epoch": 0.23999, + "grad_norm": 1.3493151760352913, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 23999 + }, + { + "epoch": 0.24, + "grad_norm": 1.5097572940765318, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 24000 + }, + { + "epoch": 0.24001, + "grad_norm": 1.3208655695161489, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24001 + }, + { + "epoch": 0.24002, + "grad_norm": 1.369927948056816, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 24002 + }, + { + "epoch": 0.24003, + "grad_norm": 1.5871583498951725, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 24003 + }, + { + "epoch": 0.24004, + "grad_norm": 1.6102392093801234, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 24004 + }, + { + "epoch": 0.24005, + "grad_norm": 1.34507653505968, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 24005 + }, + { + "epoch": 0.24006, + "grad_norm": 1.5630318864677974, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 24006 + }, + { + "epoch": 0.24007, + "grad_norm": 1.0674417926438253, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24007 + }, + { + "epoch": 0.24008, + "grad_norm": 1.6015581748821677, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 24008 + }, + { + "epoch": 0.24009, + "grad_norm": 1.2303992524679244, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 24009 + }, + { + "epoch": 0.2401, + "grad_norm": 1.4950705278510021, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 24010 + }, + { + "epoch": 0.24011, + "grad_norm": 1.133061847566615, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 24011 + }, + { + "epoch": 0.24012, + "grad_norm": 1.6713085072539149, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 24012 + }, + { + "epoch": 0.24013, + "grad_norm": 1.6075171742907977, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 24013 + }, + { + "epoch": 0.24014, + "grad_norm": 1.2231569309114756, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 24014 + }, + { + "epoch": 0.24015, + "grad_norm": 1.4613631005558372, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 24015 + }, + { + "epoch": 0.24016, + "grad_norm": 1.574652594837066, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24016 + }, + { + "epoch": 0.24017, + "grad_norm": 1.4814976148306211, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24017 + }, + { + "epoch": 0.24018, + "grad_norm": 1.2243178435485185, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24018 + }, + { + "epoch": 0.24019, + "grad_norm": 1.5514268268011124, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 24019 + }, + { + "epoch": 0.2402, + "grad_norm": 1.2261703220940252, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24020 + }, + { + "epoch": 0.24021, + "grad_norm": 1.253688213915527, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 24021 + }, + { + "epoch": 0.24022, + "grad_norm": 1.3775097145449553, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 24022 + }, + { + "epoch": 0.24023, + "grad_norm": 1.321556310405547, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24023 + }, + { + "epoch": 0.24024, + "grad_norm": 1.607243932832859, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 24024 + }, + { + "epoch": 0.24025, + "grad_norm": 1.341042583470396, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 24025 + }, + { + "epoch": 0.24026, + "grad_norm": 1.4578221252401997, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 24026 + }, + { + "epoch": 0.24027, + "grad_norm": 1.3098491479749361, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 24027 + }, + { + "epoch": 0.24028, + "grad_norm": 1.4545823917202343, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24028 + }, + { + "epoch": 0.24029, + "grad_norm": 1.2893726150280804, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 24029 + }, + { + "epoch": 0.2403, + "grad_norm": 1.3363865057682953, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24030 + }, + { + "epoch": 0.24031, + "grad_norm": 1.6045281258319553, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24031 + }, + { + "epoch": 0.24032, + "grad_norm": 1.3380129367690536, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 24032 + }, + { + "epoch": 0.24033, + "grad_norm": 1.5289132545255149, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24033 + }, + { + "epoch": 0.24034, + "grad_norm": 1.2731089616102533, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24034 + }, + { + "epoch": 0.24035, + "grad_norm": 1.3720113652692365, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24035 + }, + { + "epoch": 0.24036, + "grad_norm": 1.1645157447080348, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24036 + }, + { + "epoch": 0.24037, + "grad_norm": 1.4602016963125448, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 24037 + }, + { + "epoch": 0.24038, + "grad_norm": 1.17121783826053, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24038 + }, + { + "epoch": 0.24039, + "grad_norm": 1.565316292029105, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24039 + }, + { + "epoch": 0.2404, + "grad_norm": 1.4431977852581739, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 24040 + }, + { + "epoch": 0.24041, + "grad_norm": 1.2971511719510866, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 24041 + }, + { + "epoch": 0.24042, + "grad_norm": 1.5378137044571127, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24042 + }, + { + "epoch": 0.24043, + "grad_norm": 1.538340658409097, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 24043 + }, + { + "epoch": 0.24044, + "grad_norm": 1.7107100056471156, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 24044 + }, + { + "epoch": 0.24045, + "grad_norm": 1.1631157122618356, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 24045 + }, + { + "epoch": 0.24046, + "grad_norm": 1.3447005065554585, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24046 + }, + { + "epoch": 0.24047, + "grad_norm": 1.6684331329788673, + "learning_rate": 0.003, + "loss": 4.075, + "step": 24047 + }, + { + "epoch": 0.24048, + "grad_norm": 1.199564465199791, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24048 + }, + { + "epoch": 0.24049, + "grad_norm": 1.4851422530365161, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 24049 + }, + { + "epoch": 0.2405, + "grad_norm": 1.562186527883096, + "learning_rate": 0.003, + "loss": 4.058, + "step": 24050 + }, + { + "epoch": 0.24051, + "grad_norm": 1.2951359561812292, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 24051 + }, + { + "epoch": 0.24052, + "grad_norm": 1.4070257444670804, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 24052 + }, + { + "epoch": 0.24053, + "grad_norm": 1.2996402174222985, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24053 + }, + { + "epoch": 0.24054, + "grad_norm": 1.4405012769108854, + "learning_rate": 0.003, + "loss": 4.058, + "step": 24054 + }, + { + "epoch": 0.24055, + "grad_norm": 1.3124257817813214, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 24055 + }, + { + "epoch": 0.24056, + "grad_norm": 1.5220651917606063, + "learning_rate": 0.003, + "loss": 4.04, + "step": 24056 + }, + { + "epoch": 0.24057, + "grad_norm": 1.8254056278058537, + "learning_rate": 0.003, + "loss": 4.03, + "step": 24057 + }, + { + "epoch": 0.24058, + "grad_norm": 1.4577066872880862, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24058 + }, + { + "epoch": 0.24059, + "grad_norm": 1.3197777921445313, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 24059 + }, + { + "epoch": 0.2406, + "grad_norm": 1.3289676010536744, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 24060 + }, + { + "epoch": 0.24061, + "grad_norm": 1.363120612686246, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 24061 + }, + { + "epoch": 0.24062, + "grad_norm": 1.251584936235425, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24062 + }, + { + "epoch": 0.24063, + "grad_norm": 1.629686642487354, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24063 + }, + { + "epoch": 0.24064, + "grad_norm": 1.1974476306693564, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24064 + }, + { + "epoch": 0.24065, + "grad_norm": 1.5052029223205199, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 24065 + }, + { + "epoch": 0.24066, + "grad_norm": 1.3714932148527221, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 24066 + }, + { + "epoch": 0.24067, + "grad_norm": 1.5123863749372173, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 24067 + }, + { + "epoch": 0.24068, + "grad_norm": 1.427258807086581, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 24068 + }, + { + "epoch": 0.24069, + "grad_norm": 1.3667381992617014, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 24069 + }, + { + "epoch": 0.2407, + "grad_norm": 1.336350845274785, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24070 + }, + { + "epoch": 0.24071, + "grad_norm": 1.3208790369639019, + "learning_rate": 0.003, + "loss": 4.047, + "step": 24071 + }, + { + "epoch": 0.24072, + "grad_norm": 1.1108388793306836, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 24072 + }, + { + "epoch": 0.24073, + "grad_norm": 1.4814179727466374, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24073 + }, + { + "epoch": 0.24074, + "grad_norm": 1.305090190150312, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 24074 + }, + { + "epoch": 0.24075, + "grad_norm": 1.553580445107228, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 24075 + }, + { + "epoch": 0.24076, + "grad_norm": 1.3514565595428543, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 24076 + }, + { + "epoch": 0.24077, + "grad_norm": 1.8428150937722338, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24077 + }, + { + "epoch": 0.24078, + "grad_norm": 1.15136528152379, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 24078 + }, + { + "epoch": 0.24079, + "grad_norm": 1.2394692787565729, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 24079 + }, + { + "epoch": 0.2408, + "grad_norm": 1.5167698228484925, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 24080 + }, + { + "epoch": 0.24081, + "grad_norm": 1.4595376744030748, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24081 + }, + { + "epoch": 0.24082, + "grad_norm": 1.2838236761478816, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 24082 + }, + { + "epoch": 0.24083, + "grad_norm": 1.4465599909295825, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 24083 + }, + { + "epoch": 0.24084, + "grad_norm": 1.4072058842213704, + "learning_rate": 0.003, + "loss": 4.095, + "step": 24084 + }, + { + "epoch": 0.24085, + "grad_norm": 1.2014843558169472, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 24085 + }, + { + "epoch": 0.24086, + "grad_norm": 1.7230270251193622, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 24086 + }, + { + "epoch": 0.24087, + "grad_norm": 1.1460395686823723, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 24087 + }, + { + "epoch": 0.24088, + "grad_norm": 1.9497605446726536, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 24088 + }, + { + "epoch": 0.24089, + "grad_norm": 1.0301781552385405, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 24089 + }, + { + "epoch": 0.2409, + "grad_norm": 1.6469408464810509, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 24090 + }, + { + "epoch": 0.24091, + "grad_norm": 1.273926825135361, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 24091 + }, + { + "epoch": 0.24092, + "grad_norm": 1.4330414027792062, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 24092 + }, + { + "epoch": 0.24093, + "grad_norm": 1.4755580105387949, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 24093 + }, + { + "epoch": 0.24094, + "grad_norm": 1.3552857859565144, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 24094 + }, + { + "epoch": 0.24095, + "grad_norm": 1.2449461527733876, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24095 + }, + { + "epoch": 0.24096, + "grad_norm": 1.3337860286244896, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24096 + }, + { + "epoch": 0.24097, + "grad_norm": 1.3142564338112297, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 24097 + }, + { + "epoch": 0.24098, + "grad_norm": 1.649099599663682, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24098 + }, + { + "epoch": 0.24099, + "grad_norm": 1.3280762681798943, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 24099 + }, + { + "epoch": 0.241, + "grad_norm": 1.654781855622568, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 24100 + }, + { + "epoch": 0.24101, + "grad_norm": 1.2119325482687349, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 24101 + }, + { + "epoch": 0.24102, + "grad_norm": 1.5160643133768692, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 24102 + }, + { + "epoch": 0.24103, + "grad_norm": 1.3073909264012171, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 24103 + }, + { + "epoch": 0.24104, + "grad_norm": 1.2471836493913577, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 24104 + }, + { + "epoch": 0.24105, + "grad_norm": 1.2986683022158774, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24105 + }, + { + "epoch": 0.24106, + "grad_norm": 1.4753876400658665, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 24106 + }, + { + "epoch": 0.24107, + "grad_norm": 1.2841130899235613, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 24107 + }, + { + "epoch": 0.24108, + "grad_norm": 1.298390054165471, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24108 + }, + { + "epoch": 0.24109, + "grad_norm": 1.458470898975051, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 24109 + }, + { + "epoch": 0.2411, + "grad_norm": 1.0671846941033294, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24110 + }, + { + "epoch": 0.24111, + "grad_norm": 1.5408516818685556, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 24111 + }, + { + "epoch": 0.24112, + "grad_norm": 0.9886222221704883, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24112 + }, + { + "epoch": 0.24113, + "grad_norm": 1.6609339760141486, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 24113 + }, + { + "epoch": 0.24114, + "grad_norm": 1.1177725002355285, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 24114 + }, + { + "epoch": 0.24115, + "grad_norm": 1.6618753087966223, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 24115 + }, + { + "epoch": 0.24116, + "grad_norm": 1.593161191206839, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 24116 + }, + { + "epoch": 0.24117, + "grad_norm": 1.2873678293225177, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 24117 + }, + { + "epoch": 0.24118, + "grad_norm": 1.7329763851569069, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24118 + }, + { + "epoch": 0.24119, + "grad_norm": 1.2714595493373622, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 24119 + }, + { + "epoch": 0.2412, + "grad_norm": 1.5502666064407715, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 24120 + }, + { + "epoch": 0.24121, + "grad_norm": 1.2214682016660192, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24121 + }, + { + "epoch": 0.24122, + "grad_norm": 1.4618454169690436, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 24122 + }, + { + "epoch": 0.24123, + "grad_norm": 1.2804479803066813, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 24123 + }, + { + "epoch": 0.24124, + "grad_norm": 1.5313227284482278, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24124 + }, + { + "epoch": 0.24125, + "grad_norm": 1.5613304791889409, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 24125 + }, + { + "epoch": 0.24126, + "grad_norm": 1.0656993644403323, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24126 + }, + { + "epoch": 0.24127, + "grad_norm": 1.539046537649022, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 24127 + }, + { + "epoch": 0.24128, + "grad_norm": 1.1492346054391356, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 24128 + }, + { + "epoch": 0.24129, + "grad_norm": 1.8960752256071003, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 24129 + }, + { + "epoch": 0.2413, + "grad_norm": 1.2089011267547258, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24130 + }, + { + "epoch": 0.24131, + "grad_norm": 1.3396534965253424, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24131 + }, + { + "epoch": 0.24132, + "grad_norm": 1.4842610594009207, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24132 + }, + { + "epoch": 0.24133, + "grad_norm": 1.3976309450236981, + "learning_rate": 0.003, + "loss": 4.042, + "step": 24133 + }, + { + "epoch": 0.24134, + "grad_norm": 1.5902169018715517, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 24134 + }, + { + "epoch": 0.24135, + "grad_norm": 1.4054456342875623, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 24135 + }, + { + "epoch": 0.24136, + "grad_norm": 1.1633319820102586, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24136 + }, + { + "epoch": 0.24137, + "grad_norm": 1.6028449933793758, + "learning_rate": 0.003, + "loss": 4.041, + "step": 24137 + }, + { + "epoch": 0.24138, + "grad_norm": 1.7416615377617044, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 24138 + }, + { + "epoch": 0.24139, + "grad_norm": 1.1040978857034935, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 24139 + }, + { + "epoch": 0.2414, + "grad_norm": 1.4326342747758016, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24140 + }, + { + "epoch": 0.24141, + "grad_norm": 1.3173080638906012, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 24141 + }, + { + "epoch": 0.24142, + "grad_norm": 1.199641285163167, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 24142 + }, + { + "epoch": 0.24143, + "grad_norm": 1.3731662312621917, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 24143 + }, + { + "epoch": 0.24144, + "grad_norm": 1.4187046283538451, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 24144 + }, + { + "epoch": 0.24145, + "grad_norm": 1.2963124746011216, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24145 + }, + { + "epoch": 0.24146, + "grad_norm": 1.2583703473186578, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 24146 + }, + { + "epoch": 0.24147, + "grad_norm": 1.378754668740218, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 24147 + }, + { + "epoch": 0.24148, + "grad_norm": 1.3044351423436642, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 24148 + }, + { + "epoch": 0.24149, + "grad_norm": 1.3123845621937755, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24149 + }, + { + "epoch": 0.2415, + "grad_norm": 1.3925292135149083, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 24150 + }, + { + "epoch": 0.24151, + "grad_norm": 1.229044455895039, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24151 + }, + { + "epoch": 0.24152, + "grad_norm": 1.5548229529941022, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 24152 + }, + { + "epoch": 0.24153, + "grad_norm": 1.226909830678718, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 24153 + }, + { + "epoch": 0.24154, + "grad_norm": 1.7686656047696907, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 24154 + }, + { + "epoch": 0.24155, + "grad_norm": 1.2456571851748708, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 24155 + }, + { + "epoch": 0.24156, + "grad_norm": 1.3966131277600515, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 24156 + }, + { + "epoch": 0.24157, + "grad_norm": 1.2597456621338394, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 24157 + }, + { + "epoch": 0.24158, + "grad_norm": 1.5881093924879088, + "learning_rate": 0.003, + "loss": 4.046, + "step": 24158 + }, + { + "epoch": 0.24159, + "grad_norm": 1.3635991818037387, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24159 + }, + { + "epoch": 0.2416, + "grad_norm": 1.9670775413511101, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 24160 + }, + { + "epoch": 0.24161, + "grad_norm": 1.3054958691614253, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 24161 + }, + { + "epoch": 0.24162, + "grad_norm": 1.0894464236970764, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 24162 + }, + { + "epoch": 0.24163, + "grad_norm": 1.6673970714260016, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 24163 + }, + { + "epoch": 0.24164, + "grad_norm": 1.3907948894269935, + "learning_rate": 0.003, + "loss": 4.022, + "step": 24164 + }, + { + "epoch": 0.24165, + "grad_norm": 1.303472901631941, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24165 + }, + { + "epoch": 0.24166, + "grad_norm": 1.5688314266097538, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24166 + }, + { + "epoch": 0.24167, + "grad_norm": 1.4828007101737333, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 24167 + }, + { + "epoch": 0.24168, + "grad_norm": 1.153020032594496, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 24168 + }, + { + "epoch": 0.24169, + "grad_norm": 1.6807951751398211, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 24169 + }, + { + "epoch": 0.2417, + "grad_norm": 1.1395687023333403, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24170 + }, + { + "epoch": 0.24171, + "grad_norm": 1.710114409140257, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24171 + }, + { + "epoch": 0.24172, + "grad_norm": 1.2453411558166612, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24172 + }, + { + "epoch": 0.24173, + "grad_norm": 1.4640730242339421, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24173 + }, + { + "epoch": 0.24174, + "grad_norm": 1.2053949287522918, + "learning_rate": 0.003, + "loss": 4.021, + "step": 24174 + }, + { + "epoch": 0.24175, + "grad_norm": 1.674565837253066, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 24175 + }, + { + "epoch": 0.24176, + "grad_norm": 1.3836774832370602, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 24176 + }, + { + "epoch": 0.24177, + "grad_norm": 1.3490456470272745, + "learning_rate": 0.003, + "loss": 4.054, + "step": 24177 + }, + { + "epoch": 0.24178, + "grad_norm": 1.369329690781937, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 24178 + }, + { + "epoch": 0.24179, + "grad_norm": 1.1655394574473443, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 24179 + }, + { + "epoch": 0.2418, + "grad_norm": 1.637493211341957, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 24180 + }, + { + "epoch": 0.24181, + "grad_norm": 1.1968948824780248, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 24181 + }, + { + "epoch": 0.24182, + "grad_norm": 1.5602879931997047, + "learning_rate": 0.003, + "loss": 4.061, + "step": 24182 + }, + { + "epoch": 0.24183, + "grad_norm": 1.0345435895912405, + "learning_rate": 0.003, + "loss": 4.03, + "step": 24183 + }, + { + "epoch": 0.24184, + "grad_norm": 1.6200732593921277, + "learning_rate": 0.003, + "loss": 4.07, + "step": 24184 + }, + { + "epoch": 0.24185, + "grad_norm": 1.2023060083441939, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24185 + }, + { + "epoch": 0.24186, + "grad_norm": 1.4663230827324905, + "learning_rate": 0.003, + "loss": 4.08, + "step": 24186 + }, + { + "epoch": 0.24187, + "grad_norm": 1.5532612101258794, + "learning_rate": 0.003, + "loss": 4.077, + "step": 24187 + }, + { + "epoch": 0.24188, + "grad_norm": 1.4739841593131926, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 24188 + }, + { + "epoch": 0.24189, + "grad_norm": 1.3376844071359788, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24189 + }, + { + "epoch": 0.2419, + "grad_norm": 1.3448150107076327, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 24190 + }, + { + "epoch": 0.24191, + "grad_norm": 1.3102973802513778, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 24191 + }, + { + "epoch": 0.24192, + "grad_norm": 1.2912464168049576, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 24192 + }, + { + "epoch": 0.24193, + "grad_norm": 1.3284029491857814, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 24193 + }, + { + "epoch": 0.24194, + "grad_norm": 1.419056951370366, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24194 + }, + { + "epoch": 0.24195, + "grad_norm": 1.4665238512294811, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24195 + }, + { + "epoch": 0.24196, + "grad_norm": 1.4440262793436895, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 24196 + }, + { + "epoch": 0.24197, + "grad_norm": 1.3368358354618906, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 24197 + }, + { + "epoch": 0.24198, + "grad_norm": 1.294532460452846, + "learning_rate": 0.003, + "loss": 4.041, + "step": 24198 + }, + { + "epoch": 0.24199, + "grad_norm": 1.155168883050232, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24199 + }, + { + "epoch": 0.242, + "grad_norm": 1.5366001992696707, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 24200 + }, + { + "epoch": 0.24201, + "grad_norm": 1.186991280202697, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 24201 + }, + { + "epoch": 0.24202, + "grad_norm": 1.5046129820830885, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24202 + }, + { + "epoch": 0.24203, + "grad_norm": 1.5490891294740057, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24203 + }, + { + "epoch": 0.24204, + "grad_norm": 1.1992394383361793, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24204 + }, + { + "epoch": 0.24205, + "grad_norm": 1.540735703971023, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 24205 + }, + { + "epoch": 0.24206, + "grad_norm": 1.1360236479870494, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24206 + }, + { + "epoch": 0.24207, + "grad_norm": 1.5499068140833971, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24207 + }, + { + "epoch": 0.24208, + "grad_norm": 1.3230316006938978, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 24208 + }, + { + "epoch": 0.24209, + "grad_norm": 1.2977193847469652, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24209 + }, + { + "epoch": 0.2421, + "grad_norm": 1.8675181796009408, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 24210 + }, + { + "epoch": 0.24211, + "grad_norm": 1.2728049259065852, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 24211 + }, + { + "epoch": 0.24212, + "grad_norm": 1.3735599001217964, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 24212 + }, + { + "epoch": 0.24213, + "grad_norm": 1.361356516951775, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 24213 + }, + { + "epoch": 0.24214, + "grad_norm": 1.4157889372005399, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 24214 + }, + { + "epoch": 0.24215, + "grad_norm": 1.2223427196954062, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 24215 + }, + { + "epoch": 0.24216, + "grad_norm": 1.441977318289308, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 24216 + }, + { + "epoch": 0.24217, + "grad_norm": 1.2370739232957682, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 24217 + }, + { + "epoch": 0.24218, + "grad_norm": 1.2690633201751553, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 24218 + }, + { + "epoch": 0.24219, + "grad_norm": 1.6734778404795343, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 24219 + }, + { + "epoch": 0.2422, + "grad_norm": 1.2219899883140208, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 24220 + }, + { + "epoch": 0.24221, + "grad_norm": 1.4097330247284796, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24221 + }, + { + "epoch": 0.24222, + "grad_norm": 1.4938325003799242, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 24222 + }, + { + "epoch": 0.24223, + "grad_norm": 1.3487625356047497, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 24223 + }, + { + "epoch": 0.24224, + "grad_norm": 1.4262319837233057, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 24224 + }, + { + "epoch": 0.24225, + "grad_norm": 1.5446225618111487, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 24225 + }, + { + "epoch": 0.24226, + "grad_norm": 1.3232341728510078, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 24226 + }, + { + "epoch": 0.24227, + "grad_norm": 1.683012094414617, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 24227 + }, + { + "epoch": 0.24228, + "grad_norm": 1.295498521942109, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24228 + }, + { + "epoch": 0.24229, + "grad_norm": 1.3345779702149414, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 24229 + }, + { + "epoch": 0.2423, + "grad_norm": 1.220376470555496, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24230 + }, + { + "epoch": 0.24231, + "grad_norm": 1.4614746663436433, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 24231 + }, + { + "epoch": 0.24232, + "grad_norm": 1.105429062896927, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24232 + }, + { + "epoch": 0.24233, + "grad_norm": 1.40791526406954, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 24233 + }, + { + "epoch": 0.24234, + "grad_norm": 1.5286186035685638, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 24234 + }, + { + "epoch": 0.24235, + "grad_norm": 1.2760029744853594, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 24235 + }, + { + "epoch": 0.24236, + "grad_norm": 1.6618425248039033, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 24236 + }, + { + "epoch": 0.24237, + "grad_norm": 1.124526517736685, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24237 + }, + { + "epoch": 0.24238, + "grad_norm": 1.416640233410491, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 24238 + }, + { + "epoch": 0.24239, + "grad_norm": 1.347448856826729, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 24239 + }, + { + "epoch": 0.2424, + "grad_norm": 1.4961608163368172, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24240 + }, + { + "epoch": 0.24241, + "grad_norm": 1.3471709137260668, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24241 + }, + { + "epoch": 0.24242, + "grad_norm": 1.3148583875188227, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24242 + }, + { + "epoch": 0.24243, + "grad_norm": 1.1937711995588174, + "learning_rate": 0.003, + "loss": 4.042, + "step": 24243 + }, + { + "epoch": 0.24244, + "grad_norm": 1.4749043608917174, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24244 + }, + { + "epoch": 0.24245, + "grad_norm": 1.498681909173045, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 24245 + }, + { + "epoch": 0.24246, + "grad_norm": 1.578179945223917, + "learning_rate": 0.003, + "loss": 4.071, + "step": 24246 + }, + { + "epoch": 0.24247, + "grad_norm": 1.44493706097333, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24247 + }, + { + "epoch": 0.24248, + "grad_norm": 1.565795094573324, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 24248 + }, + { + "epoch": 0.24249, + "grad_norm": 1.3342083156197733, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 24249 + }, + { + "epoch": 0.2425, + "grad_norm": 1.2513184429047661, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 24250 + }, + { + "epoch": 0.24251, + "grad_norm": 1.2281229389318682, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 24251 + }, + { + "epoch": 0.24252, + "grad_norm": 1.0860596610360937, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 24252 + }, + { + "epoch": 0.24253, + "grad_norm": 1.5039491917515584, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 24253 + }, + { + "epoch": 0.24254, + "grad_norm": 1.4064376154439924, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24254 + }, + { + "epoch": 0.24255, + "grad_norm": 1.6241444988948979, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24255 + }, + { + "epoch": 0.24256, + "grad_norm": 1.2205957992114362, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 24256 + }, + { + "epoch": 0.24257, + "grad_norm": 1.5260255640938487, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 24257 + }, + { + "epoch": 0.24258, + "grad_norm": 1.5989556429015173, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24258 + }, + { + "epoch": 0.24259, + "grad_norm": 1.0751015815729699, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24259 + }, + { + "epoch": 0.2426, + "grad_norm": 1.702472906850061, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 24260 + }, + { + "epoch": 0.24261, + "grad_norm": 1.5207047352202259, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 24261 + }, + { + "epoch": 0.24262, + "grad_norm": 1.2409427212254576, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24262 + }, + { + "epoch": 0.24263, + "grad_norm": 1.4261002465171797, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24263 + }, + { + "epoch": 0.24264, + "grad_norm": 1.3528463451310437, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 24264 + }, + { + "epoch": 0.24265, + "grad_norm": 1.4199721692575715, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24265 + }, + { + "epoch": 0.24266, + "grad_norm": 1.1801013207495208, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24266 + }, + { + "epoch": 0.24267, + "grad_norm": 1.4976461476618232, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24267 + }, + { + "epoch": 0.24268, + "grad_norm": 1.450681518453227, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 24268 + }, + { + "epoch": 0.24269, + "grad_norm": 1.5218552775333807, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24269 + }, + { + "epoch": 0.2427, + "grad_norm": 1.3646692833450003, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 24270 + }, + { + "epoch": 0.24271, + "grad_norm": 1.372212462846825, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 24271 + }, + { + "epoch": 0.24272, + "grad_norm": 1.476927481122054, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24272 + }, + { + "epoch": 0.24273, + "grad_norm": 1.6055376498452612, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 24273 + }, + { + "epoch": 0.24274, + "grad_norm": 1.1388426209900409, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 24274 + }, + { + "epoch": 0.24275, + "grad_norm": 1.6332333310164349, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 24275 + }, + { + "epoch": 0.24276, + "grad_norm": 1.1006605511883762, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 24276 + }, + { + "epoch": 0.24277, + "grad_norm": 1.5254223487723864, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 24277 + }, + { + "epoch": 0.24278, + "grad_norm": 1.0211464488351767, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24278 + }, + { + "epoch": 0.24279, + "grad_norm": 1.6620715130789216, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24279 + }, + { + "epoch": 0.2428, + "grad_norm": 1.2899780131352225, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24280 + }, + { + "epoch": 0.24281, + "grad_norm": 1.6287916717460589, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 24281 + }, + { + "epoch": 0.24282, + "grad_norm": 1.3662707506483824, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24282 + }, + { + "epoch": 0.24283, + "grad_norm": 1.26474848782294, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24283 + }, + { + "epoch": 0.24284, + "grad_norm": 1.3738238600431016, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 24284 + }, + { + "epoch": 0.24285, + "grad_norm": 1.4668232371604881, + "learning_rate": 0.003, + "loss": 4.037, + "step": 24285 + }, + { + "epoch": 0.24286, + "grad_norm": 1.3705409294532989, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 24286 + }, + { + "epoch": 0.24287, + "grad_norm": 1.4100367607106197, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 24287 + }, + { + "epoch": 0.24288, + "grad_norm": 1.3674569502200074, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 24288 + }, + { + "epoch": 0.24289, + "grad_norm": 1.5153245137137263, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 24289 + }, + { + "epoch": 0.2429, + "grad_norm": 1.2030234602807022, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24290 + }, + { + "epoch": 0.24291, + "grad_norm": 1.453601677945708, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 24291 + }, + { + "epoch": 0.24292, + "grad_norm": 1.4259857001108331, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 24292 + }, + { + "epoch": 0.24293, + "grad_norm": 1.3952410742227412, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 24293 + }, + { + "epoch": 0.24294, + "grad_norm": 1.294020014575058, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 24294 + }, + { + "epoch": 0.24295, + "grad_norm": 1.4638737452182335, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24295 + }, + { + "epoch": 0.24296, + "grad_norm": 1.363088574657565, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 24296 + }, + { + "epoch": 0.24297, + "grad_norm": 1.5362387615785396, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24297 + }, + { + "epoch": 0.24298, + "grad_norm": 1.4637221422097164, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 24298 + }, + { + "epoch": 0.24299, + "grad_norm": 1.4573293428199658, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 24299 + }, + { + "epoch": 0.243, + "grad_norm": 1.3336338228792428, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 24300 + }, + { + "epoch": 0.24301, + "grad_norm": 1.4459234739305402, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 24301 + }, + { + "epoch": 0.24302, + "grad_norm": 1.128787365746679, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24302 + }, + { + "epoch": 0.24303, + "grad_norm": 1.530955486995232, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 24303 + }, + { + "epoch": 0.24304, + "grad_norm": 1.1271872822650284, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24304 + }, + { + "epoch": 0.24305, + "grad_norm": 1.2707499654126093, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 24305 + }, + { + "epoch": 0.24306, + "grad_norm": 1.4113113896743796, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 24306 + }, + { + "epoch": 0.24307, + "grad_norm": 1.522721069518732, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 24307 + }, + { + "epoch": 0.24308, + "grad_norm": 1.3787396396927958, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 24308 + }, + { + "epoch": 0.24309, + "grad_norm": 1.456770842668709, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 24309 + }, + { + "epoch": 0.2431, + "grad_norm": 1.4023116403124465, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 24310 + }, + { + "epoch": 0.24311, + "grad_norm": 1.4478873108427872, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 24311 + }, + { + "epoch": 0.24312, + "grad_norm": 1.2028607088997458, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24312 + }, + { + "epoch": 0.24313, + "grad_norm": 1.3631454952837638, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 24313 + }, + { + "epoch": 0.24314, + "grad_norm": 1.1105654090559485, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 24314 + }, + { + "epoch": 0.24315, + "grad_norm": 1.5136857873274912, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 24315 + }, + { + "epoch": 0.24316, + "grad_norm": 1.3313712629597232, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24316 + }, + { + "epoch": 0.24317, + "grad_norm": 1.5155104946271194, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 24317 + }, + { + "epoch": 0.24318, + "grad_norm": 1.522851616703085, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24318 + }, + { + "epoch": 0.24319, + "grad_norm": 1.3383762316729688, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 24319 + }, + { + "epoch": 0.2432, + "grad_norm": 1.3100848345872873, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 24320 + }, + { + "epoch": 0.24321, + "grad_norm": 1.3435673846156737, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 24321 + }, + { + "epoch": 0.24322, + "grad_norm": 1.2516632175398126, + "learning_rate": 0.003, + "loss": 4.071, + "step": 24322 + }, + { + "epoch": 0.24323, + "grad_norm": 1.5232425443425288, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 24323 + }, + { + "epoch": 0.24324, + "grad_norm": 1.4367358145086015, + "learning_rate": 0.003, + "loss": 4.055, + "step": 24324 + }, + { + "epoch": 0.24325, + "grad_norm": 1.4684628974815295, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24325 + }, + { + "epoch": 0.24326, + "grad_norm": 1.3624352636286248, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24326 + }, + { + "epoch": 0.24327, + "grad_norm": 1.5706965201661032, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 24327 + }, + { + "epoch": 0.24328, + "grad_norm": 1.5638520719360105, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 24328 + }, + { + "epoch": 0.24329, + "grad_norm": 1.2353108973359153, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 24329 + }, + { + "epoch": 0.2433, + "grad_norm": 1.3871393259640707, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24330 + }, + { + "epoch": 0.24331, + "grad_norm": 1.2881701036380848, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 24331 + }, + { + "epoch": 0.24332, + "grad_norm": 1.3615269539507815, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 24332 + }, + { + "epoch": 0.24333, + "grad_norm": 1.5615727208485255, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 24333 + }, + { + "epoch": 0.24334, + "grad_norm": 1.2817409757654774, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24334 + }, + { + "epoch": 0.24335, + "grad_norm": 1.4257980343726475, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 24335 + }, + { + "epoch": 0.24336, + "grad_norm": 1.2940428962366473, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 24336 + }, + { + "epoch": 0.24337, + "grad_norm": 1.2305778335972597, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 24337 + }, + { + "epoch": 0.24338, + "grad_norm": 1.4471247474330844, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 24338 + }, + { + "epoch": 0.24339, + "grad_norm": 1.385055081545386, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 24339 + }, + { + "epoch": 0.2434, + "grad_norm": 1.3640820062649563, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 24340 + }, + { + "epoch": 0.24341, + "grad_norm": 1.483300219861865, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 24341 + }, + { + "epoch": 0.24342, + "grad_norm": 1.270825362616498, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24342 + }, + { + "epoch": 0.24343, + "grad_norm": 1.3257500639714206, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24343 + }, + { + "epoch": 0.24344, + "grad_norm": 1.3964866735472208, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24344 + }, + { + "epoch": 0.24345, + "grad_norm": 1.5052621243505104, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24345 + }, + { + "epoch": 0.24346, + "grad_norm": 1.4123858896856982, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24346 + }, + { + "epoch": 0.24347, + "grad_norm": 1.4832504684088665, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 24347 + }, + { + "epoch": 0.24348, + "grad_norm": 1.4298142844095485, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 24348 + }, + { + "epoch": 0.24349, + "grad_norm": 1.5737336294592716, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 24349 + }, + { + "epoch": 0.2435, + "grad_norm": 1.4858935974430687, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24350 + }, + { + "epoch": 0.24351, + "grad_norm": 1.4123014696818081, + "learning_rate": 0.003, + "loss": 4.037, + "step": 24351 + }, + { + "epoch": 0.24352, + "grad_norm": 1.521035894910265, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24352 + }, + { + "epoch": 0.24353, + "grad_norm": 1.2365922099599098, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24353 + }, + { + "epoch": 0.24354, + "grad_norm": 1.429933876519162, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24354 + }, + { + "epoch": 0.24355, + "grad_norm": 1.19182519647841, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24355 + }, + { + "epoch": 0.24356, + "grad_norm": 1.3457778123474984, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 24356 + }, + { + "epoch": 0.24357, + "grad_norm": 1.3159769497093734, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 24357 + }, + { + "epoch": 0.24358, + "grad_norm": 1.592912543197476, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24358 + }, + { + "epoch": 0.24359, + "grad_norm": 1.0522976764991934, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24359 + }, + { + "epoch": 0.2436, + "grad_norm": 1.8177060944721388, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24360 + }, + { + "epoch": 0.24361, + "grad_norm": 1.3556473724145826, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24361 + }, + { + "epoch": 0.24362, + "grad_norm": 1.3194578770130425, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 24362 + }, + { + "epoch": 0.24363, + "grad_norm": 1.352932434724565, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 24363 + }, + { + "epoch": 0.24364, + "grad_norm": 1.3340297145132445, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24364 + }, + { + "epoch": 0.24365, + "grad_norm": 1.4323645307855688, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 24365 + }, + { + "epoch": 0.24366, + "grad_norm": 1.0952847900510025, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 24366 + }, + { + "epoch": 0.24367, + "grad_norm": 1.9483655270056768, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24367 + }, + { + "epoch": 0.24368, + "grad_norm": 1.184671412937592, + "learning_rate": 0.003, + "loss": 4.074, + "step": 24368 + }, + { + "epoch": 0.24369, + "grad_norm": 1.323735956311098, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24369 + }, + { + "epoch": 0.2437, + "grad_norm": 1.2287185524637567, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 24370 + }, + { + "epoch": 0.24371, + "grad_norm": 1.5477768258982456, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 24371 + }, + { + "epoch": 0.24372, + "grad_norm": 1.5782593451961935, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24372 + }, + { + "epoch": 0.24373, + "grad_norm": 1.4108284316744677, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24373 + }, + { + "epoch": 0.24374, + "grad_norm": 1.2284705573126384, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 24374 + }, + { + "epoch": 0.24375, + "grad_norm": 1.37467768781896, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24375 + }, + { + "epoch": 0.24376, + "grad_norm": 1.471349874890904, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 24376 + }, + { + "epoch": 0.24377, + "grad_norm": 1.3749388319854385, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 24377 + }, + { + "epoch": 0.24378, + "grad_norm": 1.274363233056151, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24378 + }, + { + "epoch": 0.24379, + "grad_norm": 1.6912229841315816, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 24379 + }, + { + "epoch": 0.2438, + "grad_norm": 1.400093247837791, + "learning_rate": 0.003, + "loss": 4.048, + "step": 24380 + }, + { + "epoch": 0.24381, + "grad_norm": 1.2895066337843406, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24381 + }, + { + "epoch": 0.24382, + "grad_norm": 1.4194758712654583, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 24382 + }, + { + "epoch": 0.24383, + "grad_norm": 1.2433499914638473, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 24383 + }, + { + "epoch": 0.24384, + "grad_norm": 1.4821397422065814, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 24384 + }, + { + "epoch": 0.24385, + "grad_norm": 1.295262408381157, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 24385 + }, + { + "epoch": 0.24386, + "grad_norm": 1.2954067866727659, + "learning_rate": 0.003, + "loss": 4.065, + "step": 24386 + }, + { + "epoch": 0.24387, + "grad_norm": 1.5892041704504634, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 24387 + }, + { + "epoch": 0.24388, + "grad_norm": 1.430285687375732, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24388 + }, + { + "epoch": 0.24389, + "grad_norm": 1.712218677082825, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 24389 + }, + { + "epoch": 0.2439, + "grad_norm": 1.4151262536645157, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 24390 + }, + { + "epoch": 0.24391, + "grad_norm": 1.3112013316467788, + "learning_rate": 0.003, + "loss": 4.032, + "step": 24391 + }, + { + "epoch": 0.24392, + "grad_norm": 1.7240463615981416, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 24392 + }, + { + "epoch": 0.24393, + "grad_norm": 0.8858424824078097, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 24393 + }, + { + "epoch": 0.24394, + "grad_norm": 1.4116201647707651, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 24394 + }, + { + "epoch": 0.24395, + "grad_norm": 1.3334782424316771, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 24395 + }, + { + "epoch": 0.24396, + "grad_norm": 1.4632406948876087, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 24396 + }, + { + "epoch": 0.24397, + "grad_norm": 1.2817814009624928, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 24397 + }, + { + "epoch": 0.24398, + "grad_norm": 1.487348617246001, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 24398 + }, + { + "epoch": 0.24399, + "grad_norm": 1.5460169285214382, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24399 + }, + { + "epoch": 0.244, + "grad_norm": 1.4556184245199653, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24400 + }, + { + "epoch": 0.24401, + "grad_norm": 1.3171795632224563, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24401 + }, + { + "epoch": 0.24402, + "grad_norm": 1.185485189642518, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24402 + }, + { + "epoch": 0.24403, + "grad_norm": 1.3650193870209284, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 24403 + }, + { + "epoch": 0.24404, + "grad_norm": 1.3693011354662386, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 24404 + }, + { + "epoch": 0.24405, + "grad_norm": 1.3979817367995035, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 24405 + }, + { + "epoch": 0.24406, + "grad_norm": 1.4139725650330204, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 24406 + }, + { + "epoch": 0.24407, + "grad_norm": 1.5673551822882998, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24407 + }, + { + "epoch": 0.24408, + "grad_norm": 1.4348223488692182, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24408 + }, + { + "epoch": 0.24409, + "grad_norm": 1.2870111952470822, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24409 + }, + { + "epoch": 0.2441, + "grad_norm": 1.5607075385910172, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 24410 + }, + { + "epoch": 0.24411, + "grad_norm": 1.465916095422506, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 24411 + }, + { + "epoch": 0.24412, + "grad_norm": 1.3937935936403099, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 24412 + }, + { + "epoch": 0.24413, + "grad_norm": 1.2495468899332896, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 24413 + }, + { + "epoch": 0.24414, + "grad_norm": 1.4251258114352572, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 24414 + }, + { + "epoch": 0.24415, + "grad_norm": 1.516055056087047, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 24415 + }, + { + "epoch": 0.24416, + "grad_norm": 1.1966696834797699, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 24416 + }, + { + "epoch": 0.24417, + "grad_norm": 1.5138749495250303, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 24417 + }, + { + "epoch": 0.24418, + "grad_norm": 1.5030877361485186, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 24418 + }, + { + "epoch": 0.24419, + "grad_norm": 1.4994207623423623, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 24419 + }, + { + "epoch": 0.2442, + "grad_norm": 1.5317987422931272, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24420 + }, + { + "epoch": 0.24421, + "grad_norm": 1.358310738936398, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 24421 + }, + { + "epoch": 0.24422, + "grad_norm": 1.4750672972351306, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 24422 + }, + { + "epoch": 0.24423, + "grad_norm": 1.353864942876393, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 24423 + }, + { + "epoch": 0.24424, + "grad_norm": 1.3545312285939408, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24424 + }, + { + "epoch": 0.24425, + "grad_norm": 1.3341578590560708, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 24425 + }, + { + "epoch": 0.24426, + "grad_norm": 1.3757113987183092, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24426 + }, + { + "epoch": 0.24427, + "grad_norm": 1.5416574636024183, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24427 + }, + { + "epoch": 0.24428, + "grad_norm": 1.570906714197422, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 24428 + }, + { + "epoch": 0.24429, + "grad_norm": 1.4222210092104266, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24429 + }, + { + "epoch": 0.2443, + "grad_norm": 1.0247992703851363, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 24430 + }, + { + "epoch": 0.24431, + "grad_norm": 1.810736259743895, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 24431 + }, + { + "epoch": 0.24432, + "grad_norm": 1.2365585636560195, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 24432 + }, + { + "epoch": 0.24433, + "grad_norm": 1.674257260914585, + "learning_rate": 0.003, + "loss": 4.072, + "step": 24433 + }, + { + "epoch": 0.24434, + "grad_norm": 1.3286671090444853, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24434 + }, + { + "epoch": 0.24435, + "grad_norm": 1.0904345767145893, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24435 + }, + { + "epoch": 0.24436, + "grad_norm": 1.5762453636013694, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 24436 + }, + { + "epoch": 0.24437, + "grad_norm": 1.158325997917941, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 24437 + }, + { + "epoch": 0.24438, + "grad_norm": 1.680406328144425, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24438 + }, + { + "epoch": 0.24439, + "grad_norm": 1.311732222890648, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 24439 + }, + { + "epoch": 0.2444, + "grad_norm": 1.2699507423947751, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 24440 + }, + { + "epoch": 0.24441, + "grad_norm": 1.3243951161664762, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 24441 + }, + { + "epoch": 0.24442, + "grad_norm": 1.5229048088252097, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 24442 + }, + { + "epoch": 0.24443, + "grad_norm": 1.4887025761334038, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 24443 + }, + { + "epoch": 0.24444, + "grad_norm": 1.586094301592182, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 24444 + }, + { + "epoch": 0.24445, + "grad_norm": 1.302344126469741, + "learning_rate": 0.003, + "loss": 4.085, + "step": 24445 + }, + { + "epoch": 0.24446, + "grad_norm": 1.186296232078591, + "learning_rate": 0.003, + "loss": 4.061, + "step": 24446 + }, + { + "epoch": 0.24447, + "grad_norm": 1.6377893578773677, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 24447 + }, + { + "epoch": 0.24448, + "grad_norm": 1.3549837039624482, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24448 + }, + { + "epoch": 0.24449, + "grad_norm": 1.4502058738900307, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24449 + }, + { + "epoch": 0.2445, + "grad_norm": 1.3386576209257963, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 24450 + }, + { + "epoch": 0.24451, + "grad_norm": 1.1957847782425934, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24451 + }, + { + "epoch": 0.24452, + "grad_norm": 1.412730486244832, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 24452 + }, + { + "epoch": 0.24453, + "grad_norm": 1.2428989416283058, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 24453 + }, + { + "epoch": 0.24454, + "grad_norm": 1.517494911075393, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24454 + }, + { + "epoch": 0.24455, + "grad_norm": 1.271593906889155, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 24455 + }, + { + "epoch": 0.24456, + "grad_norm": 1.7310280058005147, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24456 + }, + { + "epoch": 0.24457, + "grad_norm": 1.387073538895397, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 24457 + }, + { + "epoch": 0.24458, + "grad_norm": 1.782320903754428, + "learning_rate": 0.003, + "loss": 4.079, + "step": 24458 + }, + { + "epoch": 0.24459, + "grad_norm": 1.1960676787134221, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 24459 + }, + { + "epoch": 0.2446, + "grad_norm": 1.5858316416307017, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 24460 + }, + { + "epoch": 0.24461, + "grad_norm": 1.709659418447422, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 24461 + }, + { + "epoch": 0.24462, + "grad_norm": 1.2395379160011137, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 24462 + }, + { + "epoch": 0.24463, + "grad_norm": 1.4529047209427233, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24463 + }, + { + "epoch": 0.24464, + "grad_norm": 1.3186117408301783, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24464 + }, + { + "epoch": 0.24465, + "grad_norm": 1.2927826551197905, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 24465 + }, + { + "epoch": 0.24466, + "grad_norm": 1.6199789647127618, + "learning_rate": 0.003, + "loss": 4.074, + "step": 24466 + }, + { + "epoch": 0.24467, + "grad_norm": 1.0021142676261792, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 24467 + }, + { + "epoch": 0.24468, + "grad_norm": 1.4961842732304016, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 24468 + }, + { + "epoch": 0.24469, + "grad_norm": 1.30394600125921, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 24469 + }, + { + "epoch": 0.2447, + "grad_norm": 1.5561301183312808, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 24470 + }, + { + "epoch": 0.24471, + "grad_norm": 1.2900322797382657, + "learning_rate": 0.003, + "loss": 4.042, + "step": 24471 + }, + { + "epoch": 0.24472, + "grad_norm": 1.3229581521811908, + "learning_rate": 0.003, + "loss": 4.046, + "step": 24472 + }, + { + "epoch": 0.24473, + "grad_norm": 1.3858105877881572, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 24473 + }, + { + "epoch": 0.24474, + "grad_norm": 1.496441007578832, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 24474 + }, + { + "epoch": 0.24475, + "grad_norm": 1.4317586032144547, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 24475 + }, + { + "epoch": 0.24476, + "grad_norm": 1.6698655719421154, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 24476 + }, + { + "epoch": 0.24477, + "grad_norm": 1.6093669618707407, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 24477 + }, + { + "epoch": 0.24478, + "grad_norm": 1.2589770699717462, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 24478 + }, + { + "epoch": 0.24479, + "grad_norm": 1.4768781549212349, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 24479 + }, + { + "epoch": 0.2448, + "grad_norm": 1.6449614413144684, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 24480 + }, + { + "epoch": 0.24481, + "grad_norm": 1.2132655621083248, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 24481 + }, + { + "epoch": 0.24482, + "grad_norm": 1.3357216242531542, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 24482 + }, + { + "epoch": 0.24483, + "grad_norm": 1.582000784823852, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24483 + }, + { + "epoch": 0.24484, + "grad_norm": 1.2914068837901693, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24484 + }, + { + "epoch": 0.24485, + "grad_norm": 1.2532583593279065, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 24485 + }, + { + "epoch": 0.24486, + "grad_norm": 1.4148623858568774, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 24486 + }, + { + "epoch": 0.24487, + "grad_norm": 1.3746186948771522, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 24487 + }, + { + "epoch": 0.24488, + "grad_norm": 1.335901034283889, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 24488 + }, + { + "epoch": 0.24489, + "grad_norm": 1.3127438995233012, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 24489 + }, + { + "epoch": 0.2449, + "grad_norm": 1.2952384595072532, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 24490 + }, + { + "epoch": 0.24491, + "grad_norm": 1.3252811911960327, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 24491 + }, + { + "epoch": 0.24492, + "grad_norm": 1.416293001547004, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24492 + }, + { + "epoch": 0.24493, + "grad_norm": 1.6077015621828852, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 24493 + }, + { + "epoch": 0.24494, + "grad_norm": 1.653231955767635, + "learning_rate": 0.003, + "loss": 4.046, + "step": 24494 + }, + { + "epoch": 0.24495, + "grad_norm": 1.3480367016954977, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24495 + }, + { + "epoch": 0.24496, + "grad_norm": 1.3594085884946376, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 24496 + }, + { + "epoch": 0.24497, + "grad_norm": 1.1981812674244252, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 24497 + }, + { + "epoch": 0.24498, + "grad_norm": 1.498331640084471, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 24498 + }, + { + "epoch": 0.24499, + "grad_norm": 1.3094852746117118, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24499 + }, + { + "epoch": 0.245, + "grad_norm": 1.472095960571987, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 24500 + }, + { + "epoch": 0.24501, + "grad_norm": 1.3808112648803745, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24501 + }, + { + "epoch": 0.24502, + "grad_norm": 1.2129734417411704, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 24502 + }, + { + "epoch": 0.24503, + "grad_norm": 1.5611790014750202, + "learning_rate": 0.003, + "loss": 4.037, + "step": 24503 + }, + { + "epoch": 0.24504, + "grad_norm": 1.1533309815977, + "learning_rate": 0.003, + "loss": 4.036, + "step": 24504 + }, + { + "epoch": 0.24505, + "grad_norm": 1.6488448674654328, + "learning_rate": 0.003, + "loss": 4.031, + "step": 24505 + }, + { + "epoch": 0.24506, + "grad_norm": 1.294658197306135, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 24506 + }, + { + "epoch": 0.24507, + "grad_norm": 1.4197784989873545, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24507 + }, + { + "epoch": 0.24508, + "grad_norm": 1.2750191872365566, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 24508 + }, + { + "epoch": 0.24509, + "grad_norm": 1.4359010091783646, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24509 + }, + { + "epoch": 0.2451, + "grad_norm": 1.3417831220305074, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 24510 + }, + { + "epoch": 0.24511, + "grad_norm": 1.4232112183549983, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24511 + }, + { + "epoch": 0.24512, + "grad_norm": 1.2173747227926879, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24512 + }, + { + "epoch": 0.24513, + "grad_norm": 1.2648743666521953, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24513 + }, + { + "epoch": 0.24514, + "grad_norm": 1.3260981005183057, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 24514 + }, + { + "epoch": 0.24515, + "grad_norm": 1.862719067652409, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24515 + }, + { + "epoch": 0.24516, + "grad_norm": 1.202946518556536, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 24516 + }, + { + "epoch": 0.24517, + "grad_norm": 2.04262011083375, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24517 + }, + { + "epoch": 0.24518, + "grad_norm": 1.0726008388250743, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 24518 + }, + { + "epoch": 0.24519, + "grad_norm": 1.6750839788348069, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 24519 + }, + { + "epoch": 0.2452, + "grad_norm": 1.1668555665196068, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 24520 + }, + { + "epoch": 0.24521, + "grad_norm": 1.5162086637467458, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 24521 + }, + { + "epoch": 0.24522, + "grad_norm": 1.5733418179620733, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 24522 + }, + { + "epoch": 0.24523, + "grad_norm": 1.4948107770526557, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 24523 + }, + { + "epoch": 0.24524, + "grad_norm": 1.6909722365205078, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 24524 + }, + { + "epoch": 0.24525, + "grad_norm": 1.6588137622291679, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 24525 + }, + { + "epoch": 0.24526, + "grad_norm": 1.40047849015704, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 24526 + }, + { + "epoch": 0.24527, + "grad_norm": 1.6987966478577776, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 24527 + }, + { + "epoch": 0.24528, + "grad_norm": 1.1681302398291613, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 24528 + }, + { + "epoch": 0.24529, + "grad_norm": 1.4112970283358313, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 24529 + }, + { + "epoch": 0.2453, + "grad_norm": 1.2115141575097237, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 24530 + }, + { + "epoch": 0.24531, + "grad_norm": 1.3956583977814705, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24531 + }, + { + "epoch": 0.24532, + "grad_norm": 1.2277481321412032, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 24532 + }, + { + "epoch": 0.24533, + "grad_norm": 1.1711346197082155, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 24533 + }, + { + "epoch": 0.24534, + "grad_norm": 1.361342262928631, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 24534 + }, + { + "epoch": 0.24535, + "grad_norm": 1.465196438698652, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 24535 + }, + { + "epoch": 0.24536, + "grad_norm": 1.3227436156788686, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24536 + }, + { + "epoch": 0.24537, + "grad_norm": 1.7352730851433897, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 24537 + }, + { + "epoch": 0.24538, + "grad_norm": 1.4420467752851898, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 24538 + }, + { + "epoch": 0.24539, + "grad_norm": 1.8879252281652532, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 24539 + }, + { + "epoch": 0.2454, + "grad_norm": 1.0139930085724325, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 24540 + }, + { + "epoch": 0.24541, + "grad_norm": 1.4078807796121886, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 24541 + }, + { + "epoch": 0.24542, + "grad_norm": 1.5171695189987984, + "learning_rate": 0.003, + "loss": 4.072, + "step": 24542 + }, + { + "epoch": 0.24543, + "grad_norm": 1.2287151872288782, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 24543 + }, + { + "epoch": 0.24544, + "grad_norm": 1.827233745827479, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 24544 + }, + { + "epoch": 0.24545, + "grad_norm": 1.5425358275520964, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 24545 + }, + { + "epoch": 0.24546, + "grad_norm": 1.1074978733238126, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 24546 + }, + { + "epoch": 0.24547, + "grad_norm": 1.41572862405677, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 24547 + }, + { + "epoch": 0.24548, + "grad_norm": 1.4618538064043483, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 24548 + }, + { + "epoch": 0.24549, + "grad_norm": 1.5358867638614377, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24549 + }, + { + "epoch": 0.2455, + "grad_norm": 1.4764067874278128, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 24550 + }, + { + "epoch": 0.24551, + "grad_norm": 1.2656017796107097, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 24551 + }, + { + "epoch": 0.24552, + "grad_norm": 1.422792356176076, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 24552 + }, + { + "epoch": 0.24553, + "grad_norm": 1.1951801034404221, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24553 + }, + { + "epoch": 0.24554, + "grad_norm": 1.3188185892123274, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 24554 + }, + { + "epoch": 0.24555, + "grad_norm": 1.3115026082608847, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 24555 + }, + { + "epoch": 0.24556, + "grad_norm": 1.3329702336810645, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 24556 + }, + { + "epoch": 0.24557, + "grad_norm": 1.3978328960150062, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24557 + }, + { + "epoch": 0.24558, + "grad_norm": 1.1175116108372913, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 24558 + }, + { + "epoch": 0.24559, + "grad_norm": 1.773148591028893, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 24559 + }, + { + "epoch": 0.2456, + "grad_norm": 1.2411366591922672, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 24560 + }, + { + "epoch": 0.24561, + "grad_norm": 1.3765841855025884, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 24561 + }, + { + "epoch": 0.24562, + "grad_norm": 1.2549496069463526, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24562 + }, + { + "epoch": 0.24563, + "grad_norm": 1.346449356926344, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 24563 + }, + { + "epoch": 0.24564, + "grad_norm": 1.413674698156181, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 24564 + }, + { + "epoch": 0.24565, + "grad_norm": 1.2789473351579235, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 24565 + }, + { + "epoch": 0.24566, + "grad_norm": 1.339566785711176, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24566 + }, + { + "epoch": 0.24567, + "grad_norm": 1.3650356101008385, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24567 + }, + { + "epoch": 0.24568, + "grad_norm": 1.3876420647866352, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 24568 + }, + { + "epoch": 0.24569, + "grad_norm": 1.4338418726194089, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 24569 + }, + { + "epoch": 0.2457, + "grad_norm": 1.753807450464722, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 24570 + }, + { + "epoch": 0.24571, + "grad_norm": 1.5146053908835122, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 24571 + }, + { + "epoch": 0.24572, + "grad_norm": 1.4814182439570196, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 24572 + }, + { + "epoch": 0.24573, + "grad_norm": 1.3029561351089674, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 24573 + }, + { + "epoch": 0.24574, + "grad_norm": 1.5578404170827305, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 24574 + }, + { + "epoch": 0.24575, + "grad_norm": 1.1894949968217112, + "learning_rate": 0.003, + "loss": 4.036, + "step": 24575 + }, + { + "epoch": 0.24576, + "grad_norm": 1.5470838354769207, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24576 + }, + { + "epoch": 0.24577, + "grad_norm": 1.3500795966059036, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 24577 + }, + { + "epoch": 0.24578, + "grad_norm": 1.4092926111890176, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24578 + }, + { + "epoch": 0.24579, + "grad_norm": 1.52709199684596, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24579 + }, + { + "epoch": 0.2458, + "grad_norm": 1.1814001688213163, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 24580 + }, + { + "epoch": 0.24581, + "grad_norm": 1.4261722649888424, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 24581 + }, + { + "epoch": 0.24582, + "grad_norm": 1.1442515720367656, + "learning_rate": 0.003, + "loss": 4.094, + "step": 24582 + }, + { + "epoch": 0.24583, + "grad_norm": 1.487270336446602, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 24583 + }, + { + "epoch": 0.24584, + "grad_norm": 1.0401735167002224, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 24584 + }, + { + "epoch": 0.24585, + "grad_norm": 1.4650482110188905, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 24585 + }, + { + "epoch": 0.24586, + "grad_norm": 1.242945969827525, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 24586 + }, + { + "epoch": 0.24587, + "grad_norm": 1.6521344087701795, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 24587 + }, + { + "epoch": 0.24588, + "grad_norm": 1.5645848951245152, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 24588 + }, + { + "epoch": 0.24589, + "grad_norm": 1.4555479561093512, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 24589 + }, + { + "epoch": 0.2459, + "grad_norm": 1.694459714484806, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24590 + }, + { + "epoch": 0.24591, + "grad_norm": 1.4258825484612274, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 24591 + }, + { + "epoch": 0.24592, + "grad_norm": 1.3288767041200285, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 24592 + }, + { + "epoch": 0.24593, + "grad_norm": 1.393988006884798, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 24593 + }, + { + "epoch": 0.24594, + "grad_norm": 1.2835427756796285, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24594 + }, + { + "epoch": 0.24595, + "grad_norm": 1.4058980884817271, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 24595 + }, + { + "epoch": 0.24596, + "grad_norm": 1.3398051499422285, + "learning_rate": 0.003, + "loss": 4.052, + "step": 24596 + }, + { + "epoch": 0.24597, + "grad_norm": 1.6835506575380206, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 24597 + }, + { + "epoch": 0.24598, + "grad_norm": 1.3022608969836875, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 24598 + }, + { + "epoch": 0.24599, + "grad_norm": 1.866922719544958, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 24599 + }, + { + "epoch": 0.246, + "grad_norm": 1.1664846654114946, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 24600 + }, + { + "epoch": 0.24601, + "grad_norm": 1.5861335434926875, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24601 + }, + { + "epoch": 0.24602, + "grad_norm": 1.2087072445416438, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 24602 + }, + { + "epoch": 0.24603, + "grad_norm": 1.6703543958167697, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 24603 + }, + { + "epoch": 0.24604, + "grad_norm": 1.2758211439689626, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24604 + }, + { + "epoch": 0.24605, + "grad_norm": 1.4340962875835834, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 24605 + }, + { + "epoch": 0.24606, + "grad_norm": 1.1954652855240864, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 24606 + }, + { + "epoch": 0.24607, + "grad_norm": 1.4462666793842442, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 24607 + }, + { + "epoch": 0.24608, + "grad_norm": 1.4156766202597262, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 24608 + }, + { + "epoch": 0.24609, + "grad_norm": 1.323020574912954, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 24609 + }, + { + "epoch": 0.2461, + "grad_norm": 1.387088035102824, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 24610 + }, + { + "epoch": 0.24611, + "grad_norm": 1.3002607368516486, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 24611 + }, + { + "epoch": 0.24612, + "grad_norm": 1.5218306101556822, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 24612 + }, + { + "epoch": 0.24613, + "grad_norm": 1.1749153785406177, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24613 + }, + { + "epoch": 0.24614, + "grad_norm": 1.5835294723077846, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 24614 + }, + { + "epoch": 0.24615, + "grad_norm": 1.3003192097520786, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 24615 + }, + { + "epoch": 0.24616, + "grad_norm": 1.373273185063886, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 24616 + }, + { + "epoch": 0.24617, + "grad_norm": 1.3320281377586736, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 24617 + }, + { + "epoch": 0.24618, + "grad_norm": 1.611281347154727, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 24618 + }, + { + "epoch": 0.24619, + "grad_norm": 1.3408690334115998, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24619 + }, + { + "epoch": 0.2462, + "grad_norm": 1.6464263087993103, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24620 + }, + { + "epoch": 0.24621, + "grad_norm": 1.3693516194600759, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 24621 + }, + { + "epoch": 0.24622, + "grad_norm": 1.458470456610755, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 24622 + }, + { + "epoch": 0.24623, + "grad_norm": 1.399964577553382, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24623 + }, + { + "epoch": 0.24624, + "grad_norm": 1.3420808827799742, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24624 + }, + { + "epoch": 0.24625, + "grad_norm": 1.3963823107514195, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24625 + }, + { + "epoch": 0.24626, + "grad_norm": 1.2297394037438159, + "learning_rate": 0.003, + "loss": 4.018, + "step": 24626 + }, + { + "epoch": 0.24627, + "grad_norm": 1.1268844287216007, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 24627 + }, + { + "epoch": 0.24628, + "grad_norm": 1.3406638305777105, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24628 + }, + { + "epoch": 0.24629, + "grad_norm": 1.3450601486899039, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24629 + }, + { + "epoch": 0.2463, + "grad_norm": 1.5864310279963865, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 24630 + }, + { + "epoch": 0.24631, + "grad_norm": 1.133502531197477, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 24631 + }, + { + "epoch": 0.24632, + "grad_norm": 1.6448342809364231, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 24632 + }, + { + "epoch": 0.24633, + "grad_norm": 1.4034757528785844, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 24633 + }, + { + "epoch": 0.24634, + "grad_norm": 1.556867051293386, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 24634 + }, + { + "epoch": 0.24635, + "grad_norm": 1.5121569647113713, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 24635 + }, + { + "epoch": 0.24636, + "grad_norm": 1.1961739620650575, + "learning_rate": 0.003, + "loss": 4.055, + "step": 24636 + }, + { + "epoch": 0.24637, + "grad_norm": 1.4239970723771453, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 24637 + }, + { + "epoch": 0.24638, + "grad_norm": 1.102409428379748, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24638 + }, + { + "epoch": 0.24639, + "grad_norm": 1.7719140279355743, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24639 + }, + { + "epoch": 0.2464, + "grad_norm": 1.3712803644169622, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 24640 + }, + { + "epoch": 0.24641, + "grad_norm": 1.5045699212811086, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 24641 + }, + { + "epoch": 0.24642, + "grad_norm": 1.3761270797693337, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 24642 + }, + { + "epoch": 0.24643, + "grad_norm": 1.4943490813640457, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 24643 + }, + { + "epoch": 0.24644, + "grad_norm": 1.5344596961688393, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 24644 + }, + { + "epoch": 0.24645, + "grad_norm": 1.22811239361762, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24645 + }, + { + "epoch": 0.24646, + "grad_norm": 1.6091391728534283, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 24646 + }, + { + "epoch": 0.24647, + "grad_norm": 1.0788366757862968, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 24647 + }, + { + "epoch": 0.24648, + "grad_norm": 1.7664175405023235, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24648 + }, + { + "epoch": 0.24649, + "grad_norm": 1.3672920496192789, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24649 + }, + { + "epoch": 0.2465, + "grad_norm": 1.4143340175968577, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 24650 + }, + { + "epoch": 0.24651, + "grad_norm": 1.5061316213091294, + "learning_rate": 0.003, + "loss": 4.031, + "step": 24651 + }, + { + "epoch": 0.24652, + "grad_norm": 1.3424203535012875, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 24652 + }, + { + "epoch": 0.24653, + "grad_norm": 1.5086800307564445, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 24653 + }, + { + "epoch": 0.24654, + "grad_norm": 1.415792946661826, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 24654 + }, + { + "epoch": 0.24655, + "grad_norm": 1.33784359813273, + "learning_rate": 0.003, + "loss": 4.094, + "step": 24655 + }, + { + "epoch": 0.24656, + "grad_norm": 1.2431502701811876, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 24656 + }, + { + "epoch": 0.24657, + "grad_norm": 1.3791456290761512, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 24657 + }, + { + "epoch": 0.24658, + "grad_norm": 1.234715877137626, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 24658 + }, + { + "epoch": 0.24659, + "grad_norm": 1.5439065876747495, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 24659 + }, + { + "epoch": 0.2466, + "grad_norm": 1.0873679317458538, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 24660 + }, + { + "epoch": 0.24661, + "grad_norm": 1.6673409970089326, + "learning_rate": 0.003, + "loss": 4.044, + "step": 24661 + }, + { + "epoch": 0.24662, + "grad_norm": 1.0444852034726346, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24662 + }, + { + "epoch": 0.24663, + "grad_norm": 1.4755195416697107, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 24663 + }, + { + "epoch": 0.24664, + "grad_norm": 1.5490630187821017, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 24664 + }, + { + "epoch": 0.24665, + "grad_norm": 1.4389212193285406, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 24665 + }, + { + "epoch": 0.24666, + "grad_norm": 1.5898485789007988, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 24666 + }, + { + "epoch": 0.24667, + "grad_norm": 1.3414327554826224, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 24667 + }, + { + "epoch": 0.24668, + "grad_norm": 1.5083138034996968, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 24668 + }, + { + "epoch": 0.24669, + "grad_norm": 1.4375052700800133, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24669 + }, + { + "epoch": 0.2467, + "grad_norm": 1.5605833834304783, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 24670 + }, + { + "epoch": 0.24671, + "grad_norm": 1.3886377108529842, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 24671 + }, + { + "epoch": 0.24672, + "grad_norm": 1.32058004497019, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 24672 + }, + { + "epoch": 0.24673, + "grad_norm": 1.4354588066798009, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24673 + }, + { + "epoch": 0.24674, + "grad_norm": 1.1846883296170725, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 24674 + }, + { + "epoch": 0.24675, + "grad_norm": 1.6357411637273922, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 24675 + }, + { + "epoch": 0.24676, + "grad_norm": 1.6304879496370692, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 24676 + }, + { + "epoch": 0.24677, + "grad_norm": 1.1608645715542714, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 24677 + }, + { + "epoch": 0.24678, + "grad_norm": 1.426252735297271, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 24678 + }, + { + "epoch": 0.24679, + "grad_norm": 1.3048441360426093, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 24679 + }, + { + "epoch": 0.2468, + "grad_norm": 1.9295635246293055, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 24680 + }, + { + "epoch": 0.24681, + "grad_norm": 1.1741759928736533, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 24681 + }, + { + "epoch": 0.24682, + "grad_norm": 1.5398171618549163, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24682 + }, + { + "epoch": 0.24683, + "grad_norm": 1.516394595940718, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 24683 + }, + { + "epoch": 0.24684, + "grad_norm": 1.1237933799901774, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 24684 + }, + { + "epoch": 0.24685, + "grad_norm": 1.6189441512891052, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 24685 + }, + { + "epoch": 0.24686, + "grad_norm": 1.3418548017049141, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24686 + }, + { + "epoch": 0.24687, + "grad_norm": 1.2101976318774845, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 24687 + }, + { + "epoch": 0.24688, + "grad_norm": 1.5088613613462525, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 24688 + }, + { + "epoch": 0.24689, + "grad_norm": 1.396871327334466, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24689 + }, + { + "epoch": 0.2469, + "grad_norm": 1.432637212109617, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 24690 + }, + { + "epoch": 0.24691, + "grad_norm": 1.4407538030132299, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 24691 + }, + { + "epoch": 0.24692, + "grad_norm": 1.3150026238012944, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 24692 + }, + { + "epoch": 0.24693, + "grad_norm": 1.454035732610563, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 24693 + }, + { + "epoch": 0.24694, + "grad_norm": 1.4155672294803454, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 24694 + }, + { + "epoch": 0.24695, + "grad_norm": 1.4857061136107959, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24695 + }, + { + "epoch": 0.24696, + "grad_norm": 1.2884400314255022, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 24696 + }, + { + "epoch": 0.24697, + "grad_norm": 1.2763712061446344, + "learning_rate": 0.003, + "loss": 4.034, + "step": 24697 + }, + { + "epoch": 0.24698, + "grad_norm": 1.404145056826194, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 24698 + }, + { + "epoch": 0.24699, + "grad_norm": 1.4530050644988275, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24699 + }, + { + "epoch": 0.247, + "grad_norm": 1.2924308540326528, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24700 + }, + { + "epoch": 0.24701, + "grad_norm": 1.437510979020967, + "learning_rate": 0.003, + "loss": 4.03, + "step": 24701 + }, + { + "epoch": 0.24702, + "grad_norm": 1.4454755117993687, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 24702 + }, + { + "epoch": 0.24703, + "grad_norm": 1.3855527548886926, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 24703 + }, + { + "epoch": 0.24704, + "grad_norm": 1.2122761754217306, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 24704 + }, + { + "epoch": 0.24705, + "grad_norm": 1.5411351433529628, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 24705 + }, + { + "epoch": 0.24706, + "grad_norm": 1.36596179369511, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24706 + }, + { + "epoch": 0.24707, + "grad_norm": 1.5830397978622857, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 24707 + }, + { + "epoch": 0.24708, + "grad_norm": 1.5662877541642863, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 24708 + }, + { + "epoch": 0.24709, + "grad_norm": 1.2411658668010157, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 24709 + }, + { + "epoch": 0.2471, + "grad_norm": 1.372268665719959, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24710 + }, + { + "epoch": 0.24711, + "grad_norm": 1.2982392326994294, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 24711 + }, + { + "epoch": 0.24712, + "grad_norm": 1.317207009060153, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 24712 + }, + { + "epoch": 0.24713, + "grad_norm": 1.295956405394906, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 24713 + }, + { + "epoch": 0.24714, + "grad_norm": 1.47415354394144, + "learning_rate": 0.003, + "loss": 4.031, + "step": 24714 + }, + { + "epoch": 0.24715, + "grad_norm": 1.1830980536372186, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24715 + }, + { + "epoch": 0.24716, + "grad_norm": 1.4562620130790478, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 24716 + }, + { + "epoch": 0.24717, + "grad_norm": 1.4242449287789543, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 24717 + }, + { + "epoch": 0.24718, + "grad_norm": 1.3875396746958109, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 24718 + }, + { + "epoch": 0.24719, + "grad_norm": 1.2596669274496077, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 24719 + }, + { + "epoch": 0.2472, + "grad_norm": 1.3599444745526024, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 24720 + }, + { + "epoch": 0.24721, + "grad_norm": 1.6367321665569092, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 24721 + }, + { + "epoch": 0.24722, + "grad_norm": 1.3975867186835018, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24722 + }, + { + "epoch": 0.24723, + "grad_norm": 1.492204042599613, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 24723 + }, + { + "epoch": 0.24724, + "grad_norm": 1.276593404764828, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 24724 + }, + { + "epoch": 0.24725, + "grad_norm": 1.4397073350127796, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 24725 + }, + { + "epoch": 0.24726, + "grad_norm": 1.2502827618745087, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 24726 + }, + { + "epoch": 0.24727, + "grad_norm": 1.5028846024404838, + "learning_rate": 0.003, + "loss": 4.027, + "step": 24727 + }, + { + "epoch": 0.24728, + "grad_norm": 1.367116680061631, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 24728 + }, + { + "epoch": 0.24729, + "grad_norm": 1.3241220634655897, + "learning_rate": 0.003, + "loss": 4.071, + "step": 24729 + }, + { + "epoch": 0.2473, + "grad_norm": 1.494587052328523, + "learning_rate": 0.003, + "loss": 4.026, + "step": 24730 + }, + { + "epoch": 0.24731, + "grad_norm": 1.463582765087985, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 24731 + }, + { + "epoch": 0.24732, + "grad_norm": 1.5693980068020192, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24732 + }, + { + "epoch": 0.24733, + "grad_norm": 1.1560774062930037, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 24733 + }, + { + "epoch": 0.24734, + "grad_norm": 1.2579879998812238, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 24734 + }, + { + "epoch": 0.24735, + "grad_norm": 1.6737570852668278, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 24735 + }, + { + "epoch": 0.24736, + "grad_norm": 1.422753624810444, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 24736 + }, + { + "epoch": 0.24737, + "grad_norm": 1.3081567642689587, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 24737 + }, + { + "epoch": 0.24738, + "grad_norm": 1.4681337506181662, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 24738 + }, + { + "epoch": 0.24739, + "grad_norm": 1.353013517803222, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 24739 + }, + { + "epoch": 0.2474, + "grad_norm": 1.5044503178582012, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 24740 + }, + { + "epoch": 0.24741, + "grad_norm": 1.2147016615036763, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 24741 + }, + { + "epoch": 0.24742, + "grad_norm": 1.3548029930576675, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24742 + }, + { + "epoch": 0.24743, + "grad_norm": 1.1736661393469732, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 24743 + }, + { + "epoch": 0.24744, + "grad_norm": 1.523797189567375, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 24744 + }, + { + "epoch": 0.24745, + "grad_norm": 1.4938887946099033, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 24745 + }, + { + "epoch": 0.24746, + "grad_norm": 1.6129495765344535, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 24746 + }, + { + "epoch": 0.24747, + "grad_norm": 1.5465153082862477, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 24747 + }, + { + "epoch": 0.24748, + "grad_norm": 1.3470657073245074, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24748 + }, + { + "epoch": 0.24749, + "grad_norm": 1.4144434361944425, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 24749 + }, + { + "epoch": 0.2475, + "grad_norm": 1.6091144458303441, + "learning_rate": 0.003, + "loss": 4.026, + "step": 24750 + }, + { + "epoch": 0.24751, + "grad_norm": 1.151243467795758, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 24751 + }, + { + "epoch": 0.24752, + "grad_norm": 1.487955722009135, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 24752 + }, + { + "epoch": 0.24753, + "grad_norm": 1.3489445381704375, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24753 + }, + { + "epoch": 0.24754, + "grad_norm": 1.4814548584561162, + "learning_rate": 0.003, + "loss": 4.075, + "step": 24754 + }, + { + "epoch": 0.24755, + "grad_norm": 1.3338594890886353, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24755 + }, + { + "epoch": 0.24756, + "grad_norm": 1.3463998506375616, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24756 + }, + { + "epoch": 0.24757, + "grad_norm": 1.4057127181307358, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 24757 + }, + { + "epoch": 0.24758, + "grad_norm": 1.3194816805840566, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 24758 + }, + { + "epoch": 0.24759, + "grad_norm": 1.3604347150425697, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24759 + }, + { + "epoch": 0.2476, + "grad_norm": 1.4327687092121575, + "learning_rate": 0.003, + "loss": 4.057, + "step": 24760 + }, + { + "epoch": 0.24761, + "grad_norm": 1.2454851239598583, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 24761 + }, + { + "epoch": 0.24762, + "grad_norm": 1.304836654064837, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 24762 + }, + { + "epoch": 0.24763, + "grad_norm": 1.4698849853851503, + "learning_rate": 0.003, + "loss": 4.076, + "step": 24763 + }, + { + "epoch": 0.24764, + "grad_norm": 1.331294383622434, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 24764 + }, + { + "epoch": 0.24765, + "grad_norm": 1.541256627652232, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 24765 + }, + { + "epoch": 0.24766, + "grad_norm": 1.577679557555772, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 24766 + }, + { + "epoch": 0.24767, + "grad_norm": 1.1012168018809079, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 24767 + }, + { + "epoch": 0.24768, + "grad_norm": 1.5592017317726659, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 24768 + }, + { + "epoch": 0.24769, + "grad_norm": 1.2909868546716334, + "learning_rate": 0.003, + "loss": 4.028, + "step": 24769 + }, + { + "epoch": 0.2477, + "grad_norm": 1.4914432464569773, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 24770 + }, + { + "epoch": 0.24771, + "grad_norm": 1.3316743496743026, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 24771 + }, + { + "epoch": 0.24772, + "grad_norm": 1.4832628542594846, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 24772 + }, + { + "epoch": 0.24773, + "grad_norm": 1.3558204825540894, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24773 + }, + { + "epoch": 0.24774, + "grad_norm": 1.3845284802193545, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 24774 + }, + { + "epoch": 0.24775, + "grad_norm": 1.6918078209206067, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 24775 + }, + { + "epoch": 0.24776, + "grad_norm": 1.218607779578139, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 24776 + }, + { + "epoch": 0.24777, + "grad_norm": 1.5667096008125976, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 24777 + }, + { + "epoch": 0.24778, + "grad_norm": 1.2915817404671388, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 24778 + }, + { + "epoch": 0.24779, + "grad_norm": 1.35905369335716, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 24779 + }, + { + "epoch": 0.2478, + "grad_norm": 1.403991505646697, + "learning_rate": 0.003, + "loss": 4.065, + "step": 24780 + }, + { + "epoch": 0.24781, + "grad_norm": 1.3525827146462097, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 24781 + }, + { + "epoch": 0.24782, + "grad_norm": 1.3993274299502594, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 24782 + }, + { + "epoch": 0.24783, + "grad_norm": 1.3927380904293887, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 24783 + }, + { + "epoch": 0.24784, + "grad_norm": 1.4019995753607755, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 24784 + }, + { + "epoch": 0.24785, + "grad_norm": 1.456500031497397, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 24785 + }, + { + "epoch": 0.24786, + "grad_norm": 1.3937884265844593, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24786 + }, + { + "epoch": 0.24787, + "grad_norm": 1.499569326172104, + "learning_rate": 0.003, + "loss": 4.068, + "step": 24787 + }, + { + "epoch": 0.24788, + "grad_norm": 1.2575954064825454, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 24788 + }, + { + "epoch": 0.24789, + "grad_norm": 1.49778299411736, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 24789 + }, + { + "epoch": 0.2479, + "grad_norm": 1.3586828791251255, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 24790 + }, + { + "epoch": 0.24791, + "grad_norm": 1.3173905183874424, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 24791 + }, + { + "epoch": 0.24792, + "grad_norm": 1.5420085687784748, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24792 + }, + { + "epoch": 0.24793, + "grad_norm": 1.959099866201587, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 24793 + }, + { + "epoch": 0.24794, + "grad_norm": 1.2520886598351242, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 24794 + }, + { + "epoch": 0.24795, + "grad_norm": 1.370389659314976, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 24795 + }, + { + "epoch": 0.24796, + "grad_norm": 1.3983090798660438, + "learning_rate": 0.003, + "loss": 4.066, + "step": 24796 + }, + { + "epoch": 0.24797, + "grad_norm": 1.646825361741049, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 24797 + }, + { + "epoch": 0.24798, + "grad_norm": 1.2272973826331968, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24798 + }, + { + "epoch": 0.24799, + "grad_norm": 1.466392956421392, + "learning_rate": 0.003, + "loss": 4.04, + "step": 24799 + }, + { + "epoch": 0.248, + "grad_norm": 1.2716286009691005, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 24800 + }, + { + "epoch": 0.24801, + "grad_norm": 1.3834781240407417, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 24801 + }, + { + "epoch": 0.24802, + "grad_norm": 1.3723594513286763, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 24802 + }, + { + "epoch": 0.24803, + "grad_norm": 1.2643603060613744, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 24803 + }, + { + "epoch": 0.24804, + "grad_norm": 1.319402698735918, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 24804 + }, + { + "epoch": 0.24805, + "grad_norm": 1.5612727928705017, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 24805 + }, + { + "epoch": 0.24806, + "grad_norm": 1.2740989483540295, + "learning_rate": 0.003, + "loss": 4.015, + "step": 24806 + }, + { + "epoch": 0.24807, + "grad_norm": 1.40594330473953, + "learning_rate": 0.003, + "loss": 4.043, + "step": 24807 + }, + { + "epoch": 0.24808, + "grad_norm": 1.3658506105532042, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 24808 + }, + { + "epoch": 0.24809, + "grad_norm": 1.5299280710476353, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24809 + }, + { + "epoch": 0.2481, + "grad_norm": 1.43084151160238, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 24810 + }, + { + "epoch": 0.24811, + "grad_norm": 1.41360182111721, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 24811 + }, + { + "epoch": 0.24812, + "grad_norm": 1.5230535965569425, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 24812 + }, + { + "epoch": 0.24813, + "grad_norm": 1.1047908473832206, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 24813 + }, + { + "epoch": 0.24814, + "grad_norm": 1.2984868528247686, + "learning_rate": 0.003, + "loss": 4.053, + "step": 24814 + }, + { + "epoch": 0.24815, + "grad_norm": 1.3578828905489941, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 24815 + }, + { + "epoch": 0.24816, + "grad_norm": 1.5434596633809294, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 24816 + }, + { + "epoch": 0.24817, + "grad_norm": 1.123125374566577, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 24817 + }, + { + "epoch": 0.24818, + "grad_norm": 1.7423886550183643, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 24818 + }, + { + "epoch": 0.24819, + "grad_norm": 1.2024239741592904, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 24819 + }, + { + "epoch": 0.2482, + "grad_norm": 1.4649236260931853, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 24820 + }, + { + "epoch": 0.24821, + "grad_norm": 1.2166068174318763, + "learning_rate": 0.003, + "loss": 4.081, + "step": 24821 + }, + { + "epoch": 0.24822, + "grad_norm": 1.4652583334148943, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24822 + }, + { + "epoch": 0.24823, + "grad_norm": 1.585850347497253, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 24823 + }, + { + "epoch": 0.24824, + "grad_norm": 1.3916526111881091, + "learning_rate": 0.003, + "loss": 4.021, + "step": 24824 + }, + { + "epoch": 0.24825, + "grad_norm": 1.6536254524236857, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 24825 + }, + { + "epoch": 0.24826, + "grad_norm": 1.4018207123308375, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 24826 + }, + { + "epoch": 0.24827, + "grad_norm": 1.7208651229403833, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 24827 + }, + { + "epoch": 0.24828, + "grad_norm": 1.2741941037851814, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 24828 + }, + { + "epoch": 0.24829, + "grad_norm": 1.486304567488449, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 24829 + }, + { + "epoch": 0.2483, + "grad_norm": 1.4911574450730498, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 24830 + }, + { + "epoch": 0.24831, + "grad_norm": 1.2176633831592192, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 24831 + }, + { + "epoch": 0.24832, + "grad_norm": 1.3517968058011713, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 24832 + }, + { + "epoch": 0.24833, + "grad_norm": 1.301456914131885, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24833 + }, + { + "epoch": 0.24834, + "grad_norm": 1.2259073306915889, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 24834 + }, + { + "epoch": 0.24835, + "grad_norm": 1.4886572650448724, + "learning_rate": 0.003, + "loss": 4.096, + "step": 24835 + }, + { + "epoch": 0.24836, + "grad_norm": 1.5991193231226448, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 24836 + }, + { + "epoch": 0.24837, + "grad_norm": 1.172590946791768, + "learning_rate": 0.003, + "loss": 4.062, + "step": 24837 + }, + { + "epoch": 0.24838, + "grad_norm": 1.6931173646800393, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 24838 + }, + { + "epoch": 0.24839, + "grad_norm": 1.251729650022794, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24839 + }, + { + "epoch": 0.2484, + "grad_norm": 1.8671057562896713, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 24840 + }, + { + "epoch": 0.24841, + "grad_norm": 1.1413728903592155, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 24841 + }, + { + "epoch": 0.24842, + "grad_norm": 1.4479764225510083, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24842 + }, + { + "epoch": 0.24843, + "grad_norm": 1.473154459160328, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 24843 + }, + { + "epoch": 0.24844, + "grad_norm": 1.3335312756841913, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 24844 + }, + { + "epoch": 0.24845, + "grad_norm": 1.3417832457179788, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 24845 + }, + { + "epoch": 0.24846, + "grad_norm": 1.5881214403623038, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 24846 + }, + { + "epoch": 0.24847, + "grad_norm": 0.9745628327870453, + "learning_rate": 0.003, + "loss": 4.039, + "step": 24847 + }, + { + "epoch": 0.24848, + "grad_norm": 1.5852238210899334, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 24848 + }, + { + "epoch": 0.24849, + "grad_norm": 1.3723568450317898, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 24849 + }, + { + "epoch": 0.2485, + "grad_norm": 1.2900442154879015, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24850 + }, + { + "epoch": 0.24851, + "grad_norm": 1.3980421100565765, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24851 + }, + { + "epoch": 0.24852, + "grad_norm": 1.3663843712744006, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 24852 + }, + { + "epoch": 0.24853, + "grad_norm": 1.282759220882982, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24853 + }, + { + "epoch": 0.24854, + "grad_norm": 1.4978189660133514, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 24854 + }, + { + "epoch": 0.24855, + "grad_norm": 1.2896105125080048, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 24855 + }, + { + "epoch": 0.24856, + "grad_norm": 1.402952601496086, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24856 + }, + { + "epoch": 0.24857, + "grad_norm": 1.2786785021417044, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 24857 + }, + { + "epoch": 0.24858, + "grad_norm": 1.5481757204827589, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 24858 + }, + { + "epoch": 0.24859, + "grad_norm": 1.351570389855419, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24859 + }, + { + "epoch": 0.2486, + "grad_norm": 1.397430048986852, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 24860 + }, + { + "epoch": 0.24861, + "grad_norm": 1.3551863557561121, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 24861 + }, + { + "epoch": 0.24862, + "grad_norm": 1.4185473871724321, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24862 + }, + { + "epoch": 0.24863, + "grad_norm": 1.0862842495971277, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 24863 + }, + { + "epoch": 0.24864, + "grad_norm": 1.6818731455903855, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 24864 + }, + { + "epoch": 0.24865, + "grad_norm": 1.2511071361237098, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 24865 + }, + { + "epoch": 0.24866, + "grad_norm": 1.4338302577905506, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 24866 + }, + { + "epoch": 0.24867, + "grad_norm": 1.4837660041330187, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 24867 + }, + { + "epoch": 0.24868, + "grad_norm": 1.3639552229035876, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 24868 + }, + { + "epoch": 0.24869, + "grad_norm": 1.6062538649364084, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 24869 + }, + { + "epoch": 0.2487, + "grad_norm": 1.306422248173135, + "learning_rate": 0.003, + "loss": 4.075, + "step": 24870 + }, + { + "epoch": 0.24871, + "grad_norm": 1.2408179046414147, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 24871 + }, + { + "epoch": 0.24872, + "grad_norm": 1.3788783268952154, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 24872 + }, + { + "epoch": 0.24873, + "grad_norm": 1.1628844349626586, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 24873 + }, + { + "epoch": 0.24874, + "grad_norm": 1.4655332628857765, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 24874 + }, + { + "epoch": 0.24875, + "grad_norm": 1.384914918385266, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 24875 + }, + { + "epoch": 0.24876, + "grad_norm": 1.4130782634140597, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 24876 + }, + { + "epoch": 0.24877, + "grad_norm": 1.355909140788544, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 24877 + }, + { + "epoch": 0.24878, + "grad_norm": 1.1676441454980664, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 24878 + }, + { + "epoch": 0.24879, + "grad_norm": 1.5883908599015018, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 24879 + }, + { + "epoch": 0.2488, + "grad_norm": 1.3787898491179174, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24880 + }, + { + "epoch": 0.24881, + "grad_norm": 1.5499210880341256, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 24881 + }, + { + "epoch": 0.24882, + "grad_norm": 1.3268620001276943, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 24882 + }, + { + "epoch": 0.24883, + "grad_norm": 1.3964274047136032, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24883 + }, + { + "epoch": 0.24884, + "grad_norm": 1.6682214680093097, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 24884 + }, + { + "epoch": 0.24885, + "grad_norm": 1.1521562131146341, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 24885 + }, + { + "epoch": 0.24886, + "grad_norm": 1.6896948860704655, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24886 + }, + { + "epoch": 0.24887, + "grad_norm": 1.1755749713396866, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 24887 + }, + { + "epoch": 0.24888, + "grad_norm": 1.3413675390733668, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 24888 + }, + { + "epoch": 0.24889, + "grad_norm": 1.416136348657915, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 24889 + }, + { + "epoch": 0.2489, + "grad_norm": 1.225528924865507, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 24890 + }, + { + "epoch": 0.24891, + "grad_norm": 2.094430239331432, + "learning_rate": 0.003, + "loss": 4.078, + "step": 24891 + }, + { + "epoch": 0.24892, + "grad_norm": 1.0868170182591497, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 24892 + }, + { + "epoch": 0.24893, + "grad_norm": 1.5105464776711766, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 24893 + }, + { + "epoch": 0.24894, + "grad_norm": 1.2482318290860162, + "learning_rate": 0.003, + "loss": 4.045, + "step": 24894 + }, + { + "epoch": 0.24895, + "grad_norm": 1.5186923811710138, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 24895 + }, + { + "epoch": 0.24896, + "grad_norm": 1.309405107142853, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 24896 + }, + { + "epoch": 0.24897, + "grad_norm": 1.5169604372824392, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 24897 + }, + { + "epoch": 0.24898, + "grad_norm": 1.5921853381062598, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 24898 + }, + { + "epoch": 0.24899, + "grad_norm": 1.4648493377128722, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 24899 + }, + { + "epoch": 0.249, + "grad_norm": 1.2439150950204947, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 24900 + }, + { + "epoch": 0.24901, + "grad_norm": 1.74117714099156, + "learning_rate": 0.003, + "loss": 4.034, + "step": 24901 + }, + { + "epoch": 0.24902, + "grad_norm": 1.3645948388309788, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24902 + }, + { + "epoch": 0.24903, + "grad_norm": 1.227528824834206, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 24903 + }, + { + "epoch": 0.24904, + "grad_norm": 1.4694408437430944, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 24904 + }, + { + "epoch": 0.24905, + "grad_norm": 1.6293815404863183, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 24905 + }, + { + "epoch": 0.24906, + "grad_norm": 1.229020343389747, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 24906 + }, + { + "epoch": 0.24907, + "grad_norm": 1.3833909364053913, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 24907 + }, + { + "epoch": 0.24908, + "grad_norm": 1.5717682911050141, + "learning_rate": 0.003, + "loss": 4.078, + "step": 24908 + }, + { + "epoch": 0.24909, + "grad_norm": 1.3164742775962954, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24909 + }, + { + "epoch": 0.2491, + "grad_norm": 1.5423291639592938, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 24910 + }, + { + "epoch": 0.24911, + "grad_norm": 1.286871212691636, + "learning_rate": 0.003, + "loss": 4.071, + "step": 24911 + }, + { + "epoch": 0.24912, + "grad_norm": 1.5492135276404744, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 24912 + }, + { + "epoch": 0.24913, + "grad_norm": 1.3480728891768714, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 24913 + }, + { + "epoch": 0.24914, + "grad_norm": 1.4481688106902018, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 24914 + }, + { + "epoch": 0.24915, + "grad_norm": 1.3349488120589184, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 24915 + }, + { + "epoch": 0.24916, + "grad_norm": 1.3652534876243616, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 24916 + }, + { + "epoch": 0.24917, + "grad_norm": 1.5501035773819616, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24917 + }, + { + "epoch": 0.24918, + "grad_norm": 1.1708953688393553, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 24918 + }, + { + "epoch": 0.24919, + "grad_norm": 1.4982117647734141, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 24919 + }, + { + "epoch": 0.2492, + "grad_norm": 0.9986467472437912, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 24920 + }, + { + "epoch": 0.24921, + "grad_norm": 1.6723162577163804, + "learning_rate": 0.003, + "loss": 4.076, + "step": 24921 + }, + { + "epoch": 0.24922, + "grad_norm": 1.1854858068943663, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 24922 + }, + { + "epoch": 0.24923, + "grad_norm": 1.5399502472417759, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 24923 + }, + { + "epoch": 0.24924, + "grad_norm": 1.3444110408533056, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 24924 + }, + { + "epoch": 0.24925, + "grad_norm": 1.370379849328035, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24925 + }, + { + "epoch": 0.24926, + "grad_norm": 1.1605349139968573, + "learning_rate": 0.003, + "loss": 4.051, + "step": 24926 + }, + { + "epoch": 0.24927, + "grad_norm": 1.3521096311721168, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 24927 + }, + { + "epoch": 0.24928, + "grad_norm": 1.2241155366659324, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 24928 + }, + { + "epoch": 0.24929, + "grad_norm": 1.674015221783975, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 24929 + }, + { + "epoch": 0.2493, + "grad_norm": 1.3273645716043994, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 24930 + }, + { + "epoch": 0.24931, + "grad_norm": 1.42044499953458, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 24931 + }, + { + "epoch": 0.24932, + "grad_norm": 1.4143382443837755, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 24932 + }, + { + "epoch": 0.24933, + "grad_norm": 1.7790182925022033, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 24933 + }, + { + "epoch": 0.24934, + "grad_norm": 1.2015171015364932, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 24934 + }, + { + "epoch": 0.24935, + "grad_norm": 1.6905300146912643, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 24935 + }, + { + "epoch": 0.24936, + "grad_norm": 1.5317808660507717, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 24936 + }, + { + "epoch": 0.24937, + "grad_norm": 1.510811048729096, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24937 + }, + { + "epoch": 0.24938, + "grad_norm": 1.5739336242025765, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 24938 + }, + { + "epoch": 0.24939, + "grad_norm": 1.3278074605778212, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24939 + }, + { + "epoch": 0.2494, + "grad_norm": 1.461548147720003, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 24940 + }, + { + "epoch": 0.24941, + "grad_norm": 1.0999535839830032, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24941 + }, + { + "epoch": 0.24942, + "grad_norm": 1.4332565291133144, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 24942 + }, + { + "epoch": 0.24943, + "grad_norm": 1.515141909449798, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 24943 + }, + { + "epoch": 0.24944, + "grad_norm": 1.5380607304798846, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 24944 + }, + { + "epoch": 0.24945, + "grad_norm": 1.3992941597860196, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 24945 + }, + { + "epoch": 0.24946, + "grad_norm": 1.4545334976515218, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 24946 + }, + { + "epoch": 0.24947, + "grad_norm": 1.2777979900092966, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 24947 + }, + { + "epoch": 0.24948, + "grad_norm": 1.6710438025633039, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 24948 + }, + { + "epoch": 0.24949, + "grad_norm": 1.3390451864209725, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 24949 + }, + { + "epoch": 0.2495, + "grad_norm": 1.4898233288977503, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 24950 + }, + { + "epoch": 0.24951, + "grad_norm": 1.2959355284493708, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 24951 + }, + { + "epoch": 0.24952, + "grad_norm": 1.413154750131716, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 24952 + }, + { + "epoch": 0.24953, + "grad_norm": 1.3960260011173238, + "learning_rate": 0.003, + "loss": 4.036, + "step": 24953 + }, + { + "epoch": 0.24954, + "grad_norm": 1.1679144432834494, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 24954 + }, + { + "epoch": 0.24955, + "grad_norm": 1.557044717174055, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 24955 + }, + { + "epoch": 0.24956, + "grad_norm": 1.3195799815444853, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 24956 + }, + { + "epoch": 0.24957, + "grad_norm": 1.2553260762254992, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 24957 + }, + { + "epoch": 0.24958, + "grad_norm": 1.3083190563309135, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 24958 + }, + { + "epoch": 0.24959, + "grad_norm": 1.7430249716467296, + "learning_rate": 0.003, + "loss": 4.039, + "step": 24959 + }, + { + "epoch": 0.2496, + "grad_norm": 1.2486684097059337, + "learning_rate": 0.003, + "loss": 4.059, + "step": 24960 + }, + { + "epoch": 0.24961, + "grad_norm": 1.462288326704534, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 24961 + }, + { + "epoch": 0.24962, + "grad_norm": 1.6236630141571975, + "learning_rate": 0.003, + "loss": 4.064, + "step": 24962 + }, + { + "epoch": 0.24963, + "grad_norm": 1.1187082060403732, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 24963 + }, + { + "epoch": 0.24964, + "grad_norm": 1.5318702321218611, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 24964 + }, + { + "epoch": 0.24965, + "grad_norm": 1.3332413918668171, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 24965 + }, + { + "epoch": 0.24966, + "grad_norm": 1.1564175191270192, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 24966 + }, + { + "epoch": 0.24967, + "grad_norm": 1.428895740734984, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 24967 + }, + { + "epoch": 0.24968, + "grad_norm": 1.2742530340491622, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 24968 + }, + { + "epoch": 0.24969, + "grad_norm": 1.541860631267734, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 24969 + }, + { + "epoch": 0.2497, + "grad_norm": 1.2837649030426541, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 24970 + }, + { + "epoch": 0.24971, + "grad_norm": 1.4671433508690153, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 24971 + }, + { + "epoch": 0.24972, + "grad_norm": 1.4063292914102563, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 24972 + }, + { + "epoch": 0.24973, + "grad_norm": 1.5678528858891951, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 24973 + }, + { + "epoch": 0.24974, + "grad_norm": 1.6838405654390767, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 24974 + }, + { + "epoch": 0.24975, + "grad_norm": 1.4335037590487514, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 24975 + }, + { + "epoch": 0.24976, + "grad_norm": 1.294616511862058, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 24976 + }, + { + "epoch": 0.24977, + "grad_norm": 1.477863145139648, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 24977 + }, + { + "epoch": 0.24978, + "grad_norm": 1.1322766215325784, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24978 + }, + { + "epoch": 0.24979, + "grad_norm": 1.397481628428317, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 24979 + }, + { + "epoch": 0.2498, + "grad_norm": 1.2790383466875175, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 24980 + }, + { + "epoch": 0.24981, + "grad_norm": 1.4565413286850377, + "learning_rate": 0.003, + "loss": 4.047, + "step": 24981 + }, + { + "epoch": 0.24982, + "grad_norm": 1.3285689190026917, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 24982 + }, + { + "epoch": 0.24983, + "grad_norm": 1.5237320103316905, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 24983 + }, + { + "epoch": 0.24984, + "grad_norm": 1.4263999868936332, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24984 + }, + { + "epoch": 0.24985, + "grad_norm": 1.5723829723682632, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 24985 + }, + { + "epoch": 0.24986, + "grad_norm": 1.3118673213997734, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 24986 + }, + { + "epoch": 0.24987, + "grad_norm": 1.291787280514421, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 24987 + }, + { + "epoch": 0.24988, + "grad_norm": 1.7411938909897147, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 24988 + }, + { + "epoch": 0.24989, + "grad_norm": 1.1754841604041872, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 24989 + }, + { + "epoch": 0.2499, + "grad_norm": 1.5195210592200663, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 24990 + }, + { + "epoch": 0.24991, + "grad_norm": 1.3669007011543888, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 24991 + }, + { + "epoch": 0.24992, + "grad_norm": 1.221543626014081, + "learning_rate": 0.003, + "loss": 4.072, + "step": 24992 + }, + { + "epoch": 0.24993, + "grad_norm": 1.228415036087146, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 24993 + }, + { + "epoch": 0.24994, + "grad_norm": 1.4075007937712496, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 24994 + }, + { + "epoch": 0.24995, + "grad_norm": 1.1173673761004927, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 24995 + }, + { + "epoch": 0.24996, + "grad_norm": 1.5485276760896818, + "learning_rate": 0.003, + "loss": 4.06, + "step": 24996 + }, + { + "epoch": 0.24997, + "grad_norm": 1.1573506280325772, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 24997 + }, + { + "epoch": 0.24998, + "grad_norm": 1.448005405494121, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 24998 + }, + { + "epoch": 0.24999, + "grad_norm": 1.4040346319688222, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 24999 + }, + { + "epoch": 0.25, + "grad_norm": 1.8838509897983278, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 25000 + }, + { + "epoch": 0.25001, + "grad_norm": 1.2748017475324664, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 25001 + }, + { + "epoch": 0.25002, + "grad_norm": 1.4801517152524883, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 25002 + }, + { + "epoch": 0.25003, + "grad_norm": 1.3330920461551459, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25003 + }, + { + "epoch": 0.25004, + "grad_norm": 1.135724062116385, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25004 + }, + { + "epoch": 0.25005, + "grad_norm": 1.7225988356124133, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 25005 + }, + { + "epoch": 0.25006, + "grad_norm": 1.4821539595897881, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 25006 + }, + { + "epoch": 0.25007, + "grad_norm": 1.344104848645791, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 25007 + }, + { + "epoch": 0.25008, + "grad_norm": 1.2711149140336329, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 25008 + }, + { + "epoch": 0.25009, + "grad_norm": 1.8915259742696044, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25009 + }, + { + "epoch": 0.2501, + "grad_norm": 1.1617670539810043, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 25010 + }, + { + "epoch": 0.25011, + "grad_norm": 1.4106537485260828, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 25011 + }, + { + "epoch": 0.25012, + "grad_norm": 1.6012213409493314, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 25012 + }, + { + "epoch": 0.25013, + "grad_norm": 1.1638164507160436, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25013 + }, + { + "epoch": 0.25014, + "grad_norm": 1.4016524508438235, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 25014 + }, + { + "epoch": 0.25015, + "grad_norm": 1.2380151210316184, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25015 + }, + { + "epoch": 0.25016, + "grad_norm": 1.5357464557476528, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 25016 + }, + { + "epoch": 0.25017, + "grad_norm": 1.538934875163108, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 25017 + }, + { + "epoch": 0.25018, + "grad_norm": 1.3479982683605563, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 25018 + }, + { + "epoch": 0.25019, + "grad_norm": 1.46247496258233, + "learning_rate": 0.003, + "loss": 4.086, + "step": 25019 + }, + { + "epoch": 0.2502, + "grad_norm": 1.1007571816329893, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 25020 + }, + { + "epoch": 0.25021, + "grad_norm": 1.4440317923634345, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 25021 + }, + { + "epoch": 0.25022, + "grad_norm": 1.6107645955759808, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 25022 + }, + { + "epoch": 0.25023, + "grad_norm": 1.2881296338059511, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25023 + }, + { + "epoch": 0.25024, + "grad_norm": 1.4144028382262763, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 25024 + }, + { + "epoch": 0.25025, + "grad_norm": 1.5649497075553052, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 25025 + }, + { + "epoch": 0.25026, + "grad_norm": 1.1499792536365803, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25026 + }, + { + "epoch": 0.25027, + "grad_norm": 1.436633311360459, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25027 + }, + { + "epoch": 0.25028, + "grad_norm": 1.2893043720731796, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 25028 + }, + { + "epoch": 0.25029, + "grad_norm": 1.3135017011424721, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 25029 + }, + { + "epoch": 0.2503, + "grad_norm": 1.3907877043148116, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 25030 + }, + { + "epoch": 0.25031, + "grad_norm": 1.529935768172165, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 25031 + }, + { + "epoch": 0.25032, + "grad_norm": 1.5475090834382632, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 25032 + }, + { + "epoch": 0.25033, + "grad_norm": 1.3376815131287971, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 25033 + }, + { + "epoch": 0.25034, + "grad_norm": 1.3254297699267132, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25034 + }, + { + "epoch": 0.25035, + "grad_norm": 1.359602359603589, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 25035 + }, + { + "epoch": 0.25036, + "grad_norm": 1.1715099649607268, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 25036 + }, + { + "epoch": 0.25037, + "grad_norm": 1.5000779841111187, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25037 + }, + { + "epoch": 0.25038, + "grad_norm": 1.2164945876141462, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25038 + }, + { + "epoch": 0.25039, + "grad_norm": 1.3229696305792358, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 25039 + }, + { + "epoch": 0.2504, + "grad_norm": 1.4793524610030588, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25040 + }, + { + "epoch": 0.25041, + "grad_norm": 1.550371363016915, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 25041 + }, + { + "epoch": 0.25042, + "grad_norm": 1.5546323928883719, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 25042 + }, + { + "epoch": 0.25043, + "grad_norm": 1.27409114915906, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 25043 + }, + { + "epoch": 0.25044, + "grad_norm": 1.6421988949703035, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 25044 + }, + { + "epoch": 0.25045, + "grad_norm": 1.207263186027553, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25045 + }, + { + "epoch": 0.25046, + "grad_norm": 1.628015587652915, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25046 + }, + { + "epoch": 0.25047, + "grad_norm": 1.2303938396009948, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 25047 + }, + { + "epoch": 0.25048, + "grad_norm": 1.2675196435956482, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 25048 + }, + { + "epoch": 0.25049, + "grad_norm": 1.449009113313399, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 25049 + }, + { + "epoch": 0.2505, + "grad_norm": 1.3391571994921696, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25050 + }, + { + "epoch": 0.25051, + "grad_norm": 1.5504349144109788, + "learning_rate": 0.003, + "loss": 4.043, + "step": 25051 + }, + { + "epoch": 0.25052, + "grad_norm": 1.273381778899984, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 25052 + }, + { + "epoch": 0.25053, + "grad_norm": 1.2570499703351385, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25053 + }, + { + "epoch": 0.25054, + "grad_norm": 1.3450530257740156, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 25054 + }, + { + "epoch": 0.25055, + "grad_norm": 1.3723494824881228, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25055 + }, + { + "epoch": 0.25056, + "grad_norm": 1.5026395755157202, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 25056 + }, + { + "epoch": 0.25057, + "grad_norm": 1.3738937049339626, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25057 + }, + { + "epoch": 0.25058, + "grad_norm": 1.6493501919757914, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25058 + }, + { + "epoch": 0.25059, + "grad_norm": 2.067010137092943, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 25059 + }, + { + "epoch": 0.2506, + "grad_norm": 1.332216595669157, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 25060 + }, + { + "epoch": 0.25061, + "grad_norm": 1.2092073858482, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25061 + }, + { + "epoch": 0.25062, + "grad_norm": 1.3714865804427254, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25062 + }, + { + "epoch": 0.25063, + "grad_norm": 1.3298674353430395, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25063 + }, + { + "epoch": 0.25064, + "grad_norm": 1.4359353365071301, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 25064 + }, + { + "epoch": 0.25065, + "grad_norm": 1.471839901228785, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 25065 + }, + { + "epoch": 0.25066, + "grad_norm": 1.406811254447624, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 25066 + }, + { + "epoch": 0.25067, + "grad_norm": 1.384814010059198, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 25067 + }, + { + "epoch": 0.25068, + "grad_norm": 1.2837779920048389, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 25068 + }, + { + "epoch": 0.25069, + "grad_norm": 1.3865271012867195, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 25069 + }, + { + "epoch": 0.2507, + "grad_norm": 1.5067677462514886, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25070 + }, + { + "epoch": 0.25071, + "grad_norm": 1.2625014372525856, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 25071 + }, + { + "epoch": 0.25072, + "grad_norm": 1.6130587445811615, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 25072 + }, + { + "epoch": 0.25073, + "grad_norm": 1.2906004924780379, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25073 + }, + { + "epoch": 0.25074, + "grad_norm": 1.3010535286702674, + "learning_rate": 0.003, + "loss": 4.077, + "step": 25074 + }, + { + "epoch": 0.25075, + "grad_norm": 1.3461636181166996, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 25075 + }, + { + "epoch": 0.25076, + "grad_norm": 1.5663004864494208, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 25076 + }, + { + "epoch": 0.25077, + "grad_norm": 1.2892589301736825, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 25077 + }, + { + "epoch": 0.25078, + "grad_norm": 1.2352698688091392, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25078 + }, + { + "epoch": 0.25079, + "grad_norm": 1.3442533519893591, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 25079 + }, + { + "epoch": 0.2508, + "grad_norm": 1.4542872670051366, + "learning_rate": 0.003, + "loss": 4.043, + "step": 25080 + }, + { + "epoch": 0.25081, + "grad_norm": 1.2801046803551044, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 25081 + }, + { + "epoch": 0.25082, + "grad_norm": 1.420217542975841, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 25082 + }, + { + "epoch": 0.25083, + "grad_norm": 1.7141721616532908, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 25083 + }, + { + "epoch": 0.25084, + "grad_norm": 1.331859307732692, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 25084 + }, + { + "epoch": 0.25085, + "grad_norm": 1.2380851877688794, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 25085 + }, + { + "epoch": 0.25086, + "grad_norm": 1.9382257193389576, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 25086 + }, + { + "epoch": 0.25087, + "grad_norm": 0.9514925501712712, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25087 + }, + { + "epoch": 0.25088, + "grad_norm": 1.5746573300504183, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 25088 + }, + { + "epoch": 0.25089, + "grad_norm": 1.1497653308679063, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25089 + }, + { + "epoch": 0.2509, + "grad_norm": 1.4877931758857417, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25090 + }, + { + "epoch": 0.25091, + "grad_norm": 1.5172605757609134, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 25091 + }, + { + "epoch": 0.25092, + "grad_norm": 1.2571439915881732, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 25092 + }, + { + "epoch": 0.25093, + "grad_norm": 1.6019225827112673, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 25093 + }, + { + "epoch": 0.25094, + "grad_norm": 1.3508186049678066, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 25094 + }, + { + "epoch": 0.25095, + "grad_norm": 1.2758583196980002, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 25095 + }, + { + "epoch": 0.25096, + "grad_norm": 1.4366718068242452, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 25096 + }, + { + "epoch": 0.25097, + "grad_norm": 1.6128743785652717, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25097 + }, + { + "epoch": 0.25098, + "grad_norm": 1.421883497722055, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25098 + }, + { + "epoch": 0.25099, + "grad_norm": 1.5359042444124875, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 25099 + }, + { + "epoch": 0.251, + "grad_norm": 1.5560587269101052, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25100 + }, + { + "epoch": 0.25101, + "grad_norm": 1.1870444886318026, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25101 + }, + { + "epoch": 0.25102, + "grad_norm": 1.5022136824037668, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25102 + }, + { + "epoch": 0.25103, + "grad_norm": 1.2523542449419982, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 25103 + }, + { + "epoch": 0.25104, + "grad_norm": 1.6053678501454782, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25104 + }, + { + "epoch": 0.25105, + "grad_norm": 1.1830984434372966, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 25105 + }, + { + "epoch": 0.25106, + "grad_norm": 1.7028503508424542, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 25106 + }, + { + "epoch": 0.25107, + "grad_norm": 1.276465655350452, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 25107 + }, + { + "epoch": 0.25108, + "grad_norm": 1.642978517199037, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 25108 + }, + { + "epoch": 0.25109, + "grad_norm": 1.3285579061365733, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 25109 + }, + { + "epoch": 0.2511, + "grad_norm": 1.5164706875427652, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 25110 + }, + { + "epoch": 0.25111, + "grad_norm": 1.2564825677002043, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 25111 + }, + { + "epoch": 0.25112, + "grad_norm": 1.3576272139447902, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 25112 + }, + { + "epoch": 0.25113, + "grad_norm": 1.4842960772650207, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 25113 + }, + { + "epoch": 0.25114, + "grad_norm": 1.5729813113545792, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 25114 + }, + { + "epoch": 0.25115, + "grad_norm": 1.201038438127107, + "learning_rate": 0.003, + "loss": 4.025, + "step": 25115 + }, + { + "epoch": 0.25116, + "grad_norm": 1.3821215330755574, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 25116 + }, + { + "epoch": 0.25117, + "grad_norm": 1.361043474531906, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 25117 + }, + { + "epoch": 0.25118, + "grad_norm": 1.429792382444159, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25118 + }, + { + "epoch": 0.25119, + "grad_norm": 1.2984450847415157, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 25119 + }, + { + "epoch": 0.2512, + "grad_norm": 1.251387180293601, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25120 + }, + { + "epoch": 0.25121, + "grad_norm": 1.6212657568444315, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 25121 + }, + { + "epoch": 0.25122, + "grad_norm": 1.101721795296369, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 25122 + }, + { + "epoch": 0.25123, + "grad_norm": 1.7175279157835124, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 25123 + }, + { + "epoch": 0.25124, + "grad_norm": 1.0755713648383642, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 25124 + }, + { + "epoch": 0.25125, + "grad_norm": 1.7882614117403746, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 25125 + }, + { + "epoch": 0.25126, + "grad_norm": 1.436601426219007, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25126 + }, + { + "epoch": 0.25127, + "grad_norm": 1.4731193208094815, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25127 + }, + { + "epoch": 0.25128, + "grad_norm": 1.8291166228073243, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 25128 + }, + { + "epoch": 0.25129, + "grad_norm": 1.2940235491169902, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25129 + }, + { + "epoch": 0.2513, + "grad_norm": 1.401892583160338, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 25130 + }, + { + "epoch": 0.25131, + "grad_norm": 1.5127252425077868, + "learning_rate": 0.003, + "loss": 4.078, + "step": 25131 + }, + { + "epoch": 0.25132, + "grad_norm": 1.3940964706079149, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25132 + }, + { + "epoch": 0.25133, + "grad_norm": 1.357102229005622, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 25133 + }, + { + "epoch": 0.25134, + "grad_norm": 1.4684367662800373, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25134 + }, + { + "epoch": 0.25135, + "grad_norm": 1.4236044972482442, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25135 + }, + { + "epoch": 0.25136, + "grad_norm": 1.2394641907887927, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25136 + }, + { + "epoch": 0.25137, + "grad_norm": 1.2520251277594996, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25137 + }, + { + "epoch": 0.25138, + "grad_norm": 1.3601293011990994, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25138 + }, + { + "epoch": 0.25139, + "grad_norm": 1.3660228536740826, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25139 + }, + { + "epoch": 0.2514, + "grad_norm": 1.4537555643947053, + "learning_rate": 0.003, + "loss": 4.048, + "step": 25140 + }, + { + "epoch": 0.25141, + "grad_norm": 1.5232967625090827, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 25141 + }, + { + "epoch": 0.25142, + "grad_norm": 1.4963666099124064, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25142 + }, + { + "epoch": 0.25143, + "grad_norm": 1.4013324726574117, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25143 + }, + { + "epoch": 0.25144, + "grad_norm": 1.3046829409451521, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25144 + }, + { + "epoch": 0.25145, + "grad_norm": 1.537355149465918, + "learning_rate": 0.003, + "loss": 4.056, + "step": 25145 + }, + { + "epoch": 0.25146, + "grad_norm": 1.0734506048543841, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 25146 + }, + { + "epoch": 0.25147, + "grad_norm": 1.6936472664094977, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25147 + }, + { + "epoch": 0.25148, + "grad_norm": 1.184668133259422, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 25148 + }, + { + "epoch": 0.25149, + "grad_norm": 1.6557792673181408, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25149 + }, + { + "epoch": 0.2515, + "grad_norm": 1.2444052042044118, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 25150 + }, + { + "epoch": 0.25151, + "grad_norm": 1.7276337200454241, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25151 + }, + { + "epoch": 0.25152, + "grad_norm": 1.134267436836086, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 25152 + }, + { + "epoch": 0.25153, + "grad_norm": 1.4210589908143996, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 25153 + }, + { + "epoch": 0.25154, + "grad_norm": 1.3573451675136259, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 25154 + }, + { + "epoch": 0.25155, + "grad_norm": 1.5440684094347121, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25155 + }, + { + "epoch": 0.25156, + "grad_norm": 1.4149666272875177, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 25156 + }, + { + "epoch": 0.25157, + "grad_norm": 1.516900080469773, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 25157 + }, + { + "epoch": 0.25158, + "grad_norm": 1.439199693827732, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25158 + }, + { + "epoch": 0.25159, + "grad_norm": 1.1999588454352765, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 25159 + }, + { + "epoch": 0.2516, + "grad_norm": 1.637985824999422, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 25160 + }, + { + "epoch": 0.25161, + "grad_norm": 1.2078896254049998, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25161 + }, + { + "epoch": 0.25162, + "grad_norm": 1.5540013680699354, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25162 + }, + { + "epoch": 0.25163, + "grad_norm": 1.1133042461942622, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25163 + }, + { + "epoch": 0.25164, + "grad_norm": 1.8615331735865226, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 25164 + }, + { + "epoch": 0.25165, + "grad_norm": 1.1747796455899546, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 25165 + }, + { + "epoch": 0.25166, + "grad_norm": 1.4022974315076093, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 25166 + }, + { + "epoch": 0.25167, + "grad_norm": 1.6004397743087728, + "learning_rate": 0.003, + "loss": 4.077, + "step": 25167 + }, + { + "epoch": 0.25168, + "grad_norm": 1.2728967979811856, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 25168 + }, + { + "epoch": 0.25169, + "grad_norm": 1.560009772990896, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 25169 + }, + { + "epoch": 0.2517, + "grad_norm": 1.202206915278775, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 25170 + }, + { + "epoch": 0.25171, + "grad_norm": 1.4360783404367445, + "learning_rate": 0.003, + "loss": 4.029, + "step": 25171 + }, + { + "epoch": 0.25172, + "grad_norm": 1.3178475774984317, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25172 + }, + { + "epoch": 0.25173, + "grad_norm": 1.3286607421612373, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25173 + }, + { + "epoch": 0.25174, + "grad_norm": 1.7317664582148886, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 25174 + }, + { + "epoch": 0.25175, + "grad_norm": 1.1587860139328252, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25175 + }, + { + "epoch": 0.25176, + "grad_norm": 1.6043646579839161, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25176 + }, + { + "epoch": 0.25177, + "grad_norm": 1.5277343890698323, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 25177 + }, + { + "epoch": 0.25178, + "grad_norm": 1.272206851210595, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 25178 + }, + { + "epoch": 0.25179, + "grad_norm": 1.643293505702538, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25179 + }, + { + "epoch": 0.2518, + "grad_norm": 0.9575810638322739, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 25180 + }, + { + "epoch": 0.25181, + "grad_norm": 1.5991334240527573, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 25181 + }, + { + "epoch": 0.25182, + "grad_norm": 1.192800793906798, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 25182 + }, + { + "epoch": 0.25183, + "grad_norm": 1.8500253568772047, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 25183 + }, + { + "epoch": 0.25184, + "grad_norm": 1.3156410732671433, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 25184 + }, + { + "epoch": 0.25185, + "grad_norm": 1.3647590890373258, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 25185 + }, + { + "epoch": 0.25186, + "grad_norm": 1.3464265359380858, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25186 + }, + { + "epoch": 0.25187, + "grad_norm": 1.4572536800134575, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 25187 + }, + { + "epoch": 0.25188, + "grad_norm": 1.3175917856223025, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 25188 + }, + { + "epoch": 0.25189, + "grad_norm": 1.288567220097874, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 25189 + }, + { + "epoch": 0.2519, + "grad_norm": 1.484782114060467, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25190 + }, + { + "epoch": 0.25191, + "grad_norm": 1.2869256707428494, + "learning_rate": 0.003, + "loss": 4.043, + "step": 25191 + }, + { + "epoch": 0.25192, + "grad_norm": 1.3245246568168316, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 25192 + }, + { + "epoch": 0.25193, + "grad_norm": 1.5690624476461152, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 25193 + }, + { + "epoch": 0.25194, + "grad_norm": 1.2659452680962804, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 25194 + }, + { + "epoch": 0.25195, + "grad_norm": 1.4583903814389454, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25195 + }, + { + "epoch": 0.25196, + "grad_norm": 1.2397058449658471, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 25196 + }, + { + "epoch": 0.25197, + "grad_norm": 1.339395723429851, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 25197 + }, + { + "epoch": 0.25198, + "grad_norm": 1.328676936692069, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 25198 + }, + { + "epoch": 0.25199, + "grad_norm": 1.6238119822845876, + "learning_rate": 0.003, + "loss": 4.024, + "step": 25199 + }, + { + "epoch": 0.252, + "grad_norm": 1.269888360143885, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 25200 + }, + { + "epoch": 0.25201, + "grad_norm": 1.4267044277087728, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25201 + }, + { + "epoch": 0.25202, + "grad_norm": 1.4218801507011976, + "learning_rate": 0.003, + "loss": 4.033, + "step": 25202 + }, + { + "epoch": 0.25203, + "grad_norm": 1.3688132860757418, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 25203 + }, + { + "epoch": 0.25204, + "grad_norm": 1.609201944142336, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 25204 + }, + { + "epoch": 0.25205, + "grad_norm": 1.3787918809355866, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 25205 + }, + { + "epoch": 0.25206, + "grad_norm": 1.6419170881232383, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25206 + }, + { + "epoch": 0.25207, + "grad_norm": 1.3773995047148953, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25207 + }, + { + "epoch": 0.25208, + "grad_norm": 1.467466751074815, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 25208 + }, + { + "epoch": 0.25209, + "grad_norm": 1.2283268077267135, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25209 + }, + { + "epoch": 0.2521, + "grad_norm": 1.4422807603201062, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25210 + }, + { + "epoch": 0.25211, + "grad_norm": 1.2376673380059546, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 25211 + }, + { + "epoch": 0.25212, + "grad_norm": 1.6167020652642925, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25212 + }, + { + "epoch": 0.25213, + "grad_norm": 1.1562475019273848, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 25213 + }, + { + "epoch": 0.25214, + "grad_norm": 1.6284489472696146, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 25214 + }, + { + "epoch": 0.25215, + "grad_norm": 1.4914284717135833, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25215 + }, + { + "epoch": 0.25216, + "grad_norm": 1.5671720529987918, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 25216 + }, + { + "epoch": 0.25217, + "grad_norm": 1.279759773672114, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 25217 + }, + { + "epoch": 0.25218, + "grad_norm": 1.3908419171119077, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25218 + }, + { + "epoch": 0.25219, + "grad_norm": 1.4238686154367661, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 25219 + }, + { + "epoch": 0.2522, + "grad_norm": 1.1857520037593197, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 25220 + }, + { + "epoch": 0.25221, + "grad_norm": 1.3730914307492301, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 25221 + }, + { + "epoch": 0.25222, + "grad_norm": 1.2651121726250283, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 25222 + }, + { + "epoch": 0.25223, + "grad_norm": 1.4139279888094562, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25223 + }, + { + "epoch": 0.25224, + "grad_norm": 1.1753096524842168, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 25224 + }, + { + "epoch": 0.25225, + "grad_norm": 1.4300409723200138, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 25225 + }, + { + "epoch": 0.25226, + "grad_norm": 1.1391466203413618, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 25226 + }, + { + "epoch": 0.25227, + "grad_norm": 1.6894102544410714, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 25227 + }, + { + "epoch": 0.25228, + "grad_norm": 1.1246710077672635, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25228 + }, + { + "epoch": 0.25229, + "grad_norm": 1.5565339525433832, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 25229 + }, + { + "epoch": 0.2523, + "grad_norm": 1.2863038928226818, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 25230 + }, + { + "epoch": 0.25231, + "grad_norm": 1.676789127526627, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25231 + }, + { + "epoch": 0.25232, + "grad_norm": 1.2318633956853078, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25232 + }, + { + "epoch": 0.25233, + "grad_norm": 1.7270268075070208, + "learning_rate": 0.003, + "loss": 4.084, + "step": 25233 + }, + { + "epoch": 0.25234, + "grad_norm": 1.8843912582486868, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 25234 + }, + { + "epoch": 0.25235, + "grad_norm": 1.2539211454200057, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 25235 + }, + { + "epoch": 0.25236, + "grad_norm": 1.0934542740270976, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 25236 + }, + { + "epoch": 0.25237, + "grad_norm": 1.664552691951597, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 25237 + }, + { + "epoch": 0.25238, + "grad_norm": 1.3523323860632475, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 25238 + }, + { + "epoch": 0.25239, + "grad_norm": 1.7030880227832823, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 25239 + }, + { + "epoch": 0.2524, + "grad_norm": 1.4292686011186586, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 25240 + }, + { + "epoch": 0.25241, + "grad_norm": 1.1257082469293838, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25241 + }, + { + "epoch": 0.25242, + "grad_norm": 1.3709438714996827, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 25242 + }, + { + "epoch": 0.25243, + "grad_norm": 1.363709261568603, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25243 + }, + { + "epoch": 0.25244, + "grad_norm": 1.5986290113544008, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 25244 + }, + { + "epoch": 0.25245, + "grad_norm": 1.5283901507907833, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 25245 + }, + { + "epoch": 0.25246, + "grad_norm": 1.3552740125178642, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 25246 + }, + { + "epoch": 0.25247, + "grad_norm": 1.4763989174990026, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25247 + }, + { + "epoch": 0.25248, + "grad_norm": 1.3853022269493331, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 25248 + }, + { + "epoch": 0.25249, + "grad_norm": 1.380827051311956, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25249 + }, + { + "epoch": 0.2525, + "grad_norm": 1.4990658995688257, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 25250 + }, + { + "epoch": 0.25251, + "grad_norm": 1.3669744332486737, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 25251 + }, + { + "epoch": 0.25252, + "grad_norm": 1.3500189881507845, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 25252 + }, + { + "epoch": 0.25253, + "grad_norm": 1.245215618245871, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 25253 + }, + { + "epoch": 0.25254, + "grad_norm": 1.3994491652014052, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 25254 + }, + { + "epoch": 0.25255, + "grad_norm": 1.12317261140991, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25255 + }, + { + "epoch": 0.25256, + "grad_norm": 1.636785252583222, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25256 + }, + { + "epoch": 0.25257, + "grad_norm": 1.1551172065144573, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 25257 + }, + { + "epoch": 0.25258, + "grad_norm": 1.4391707178789181, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 25258 + }, + { + "epoch": 0.25259, + "grad_norm": 1.518728799871394, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 25259 + }, + { + "epoch": 0.2526, + "grad_norm": 1.291050329160415, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25260 + }, + { + "epoch": 0.25261, + "grad_norm": 1.3886281933205307, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 25261 + }, + { + "epoch": 0.25262, + "grad_norm": 1.2639215308699423, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 25262 + }, + { + "epoch": 0.25263, + "grad_norm": 1.2449024687391204, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 25263 + }, + { + "epoch": 0.25264, + "grad_norm": 1.6348950672560394, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 25264 + }, + { + "epoch": 0.25265, + "grad_norm": 1.7705397155507583, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 25265 + }, + { + "epoch": 0.25266, + "grad_norm": 1.5916546516380479, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25266 + }, + { + "epoch": 0.25267, + "grad_norm": 1.4357868472063537, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25267 + }, + { + "epoch": 0.25268, + "grad_norm": 1.2137546219048938, + "learning_rate": 0.003, + "loss": 4.061, + "step": 25268 + }, + { + "epoch": 0.25269, + "grad_norm": 1.5021698413914775, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 25269 + }, + { + "epoch": 0.2527, + "grad_norm": 1.3569008962632036, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 25270 + }, + { + "epoch": 0.25271, + "grad_norm": 1.6240817817906592, + "learning_rate": 0.003, + "loss": 4.047, + "step": 25271 + }, + { + "epoch": 0.25272, + "grad_norm": 1.1948694351999662, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25272 + }, + { + "epoch": 0.25273, + "grad_norm": 1.493489422202077, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 25273 + }, + { + "epoch": 0.25274, + "grad_norm": 1.3185568053747854, + "learning_rate": 0.003, + "loss": 4.029, + "step": 25274 + }, + { + "epoch": 0.25275, + "grad_norm": 1.447258551626343, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25275 + }, + { + "epoch": 0.25276, + "grad_norm": 1.3908947661523872, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 25276 + }, + { + "epoch": 0.25277, + "grad_norm": 1.4495112107734367, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25277 + }, + { + "epoch": 0.25278, + "grad_norm": 1.5572027865147067, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 25278 + }, + { + "epoch": 0.25279, + "grad_norm": 1.2244474680635473, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 25279 + }, + { + "epoch": 0.2528, + "grad_norm": 1.4015021802087044, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 25280 + }, + { + "epoch": 0.25281, + "grad_norm": 1.5014720273624802, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 25281 + }, + { + "epoch": 0.25282, + "grad_norm": 1.4039071490433639, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 25282 + }, + { + "epoch": 0.25283, + "grad_norm": 1.5923833113653523, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 25283 + }, + { + "epoch": 0.25284, + "grad_norm": 1.4991002940986895, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 25284 + }, + { + "epoch": 0.25285, + "grad_norm": 1.846203411630082, + "learning_rate": 0.003, + "loss": 4.071, + "step": 25285 + }, + { + "epoch": 0.25286, + "grad_norm": 1.1134305180409663, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25286 + }, + { + "epoch": 0.25287, + "grad_norm": 1.6651144200303902, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 25287 + }, + { + "epoch": 0.25288, + "grad_norm": 1.3111997370810007, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 25288 + }, + { + "epoch": 0.25289, + "grad_norm": 1.5443862247101616, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 25289 + }, + { + "epoch": 0.2529, + "grad_norm": 1.5555542843265413, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25290 + }, + { + "epoch": 0.25291, + "grad_norm": 1.222826730149652, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 25291 + }, + { + "epoch": 0.25292, + "grad_norm": 1.2559938765362613, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 25292 + }, + { + "epoch": 0.25293, + "grad_norm": 1.5355502562598609, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 25293 + }, + { + "epoch": 0.25294, + "grad_norm": 1.3170346789557994, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 25294 + }, + { + "epoch": 0.25295, + "grad_norm": 1.237767382911788, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 25295 + }, + { + "epoch": 0.25296, + "grad_norm": 1.413900295177227, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25296 + }, + { + "epoch": 0.25297, + "grad_norm": 1.257466286911203, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 25297 + }, + { + "epoch": 0.25298, + "grad_norm": 1.4962081789133097, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25298 + }, + { + "epoch": 0.25299, + "grad_norm": 1.2628159445568492, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25299 + }, + { + "epoch": 0.253, + "grad_norm": 1.8174857011302987, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 25300 + }, + { + "epoch": 0.25301, + "grad_norm": 1.2035343529271079, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 25301 + }, + { + "epoch": 0.25302, + "grad_norm": 1.6560577664185787, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25302 + }, + { + "epoch": 0.25303, + "grad_norm": 1.3482229308444569, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 25303 + }, + { + "epoch": 0.25304, + "grad_norm": 1.688565516809209, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25304 + }, + { + "epoch": 0.25305, + "grad_norm": 1.1769296822558077, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25305 + }, + { + "epoch": 0.25306, + "grad_norm": 1.6848923174320518, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 25306 + }, + { + "epoch": 0.25307, + "grad_norm": 1.376158276833462, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 25307 + }, + { + "epoch": 0.25308, + "grad_norm": 1.3371662072263286, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 25308 + }, + { + "epoch": 0.25309, + "grad_norm": 1.2956318033021297, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 25309 + }, + { + "epoch": 0.2531, + "grad_norm": 1.6719974740738792, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 25310 + }, + { + "epoch": 0.25311, + "grad_norm": 1.2701102834006786, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 25311 + }, + { + "epoch": 0.25312, + "grad_norm": 1.155714883626628, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25312 + }, + { + "epoch": 0.25313, + "grad_norm": 1.6508745898760928, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 25313 + }, + { + "epoch": 0.25314, + "grad_norm": 1.1854889941185416, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 25314 + }, + { + "epoch": 0.25315, + "grad_norm": 1.751531085829464, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 25315 + }, + { + "epoch": 0.25316, + "grad_norm": 1.15988291149224, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 25316 + }, + { + "epoch": 0.25317, + "grad_norm": 1.545473273781796, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 25317 + }, + { + "epoch": 0.25318, + "grad_norm": 1.1814362412392383, + "learning_rate": 0.003, + "loss": 4.031, + "step": 25318 + }, + { + "epoch": 0.25319, + "grad_norm": 1.417648847282533, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25319 + }, + { + "epoch": 0.2532, + "grad_norm": 1.3642856434639747, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 25320 + }, + { + "epoch": 0.25321, + "grad_norm": 1.3247494612287611, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 25321 + }, + { + "epoch": 0.25322, + "grad_norm": 1.72306792822049, + "learning_rate": 0.003, + "loss": 4.074, + "step": 25322 + }, + { + "epoch": 0.25323, + "grad_norm": 1.470529794566376, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 25323 + }, + { + "epoch": 0.25324, + "grad_norm": 1.2170346363395002, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 25324 + }, + { + "epoch": 0.25325, + "grad_norm": 1.568263916775638, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 25325 + }, + { + "epoch": 0.25326, + "grad_norm": 1.5491882537504937, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 25326 + }, + { + "epoch": 0.25327, + "grad_norm": 1.0427265548153881, + "learning_rate": 0.003, + "loss": 4.035, + "step": 25327 + }, + { + "epoch": 0.25328, + "grad_norm": 1.6310818781840875, + "learning_rate": 0.003, + "loss": 4.062, + "step": 25328 + }, + { + "epoch": 0.25329, + "grad_norm": 1.0730968676960242, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25329 + }, + { + "epoch": 0.2533, + "grad_norm": 1.5533450401337339, + "learning_rate": 0.003, + "loss": 4.06, + "step": 25330 + }, + { + "epoch": 0.25331, + "grad_norm": 1.091160367129258, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 25331 + }, + { + "epoch": 0.25332, + "grad_norm": 1.4582832354874897, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 25332 + }, + { + "epoch": 0.25333, + "grad_norm": 1.1904747000840705, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 25333 + }, + { + "epoch": 0.25334, + "grad_norm": 1.6746989650572792, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 25334 + }, + { + "epoch": 0.25335, + "grad_norm": 1.3943335118458742, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 25335 + }, + { + "epoch": 0.25336, + "grad_norm": 1.3007989107596187, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 25336 + }, + { + "epoch": 0.25337, + "grad_norm": 1.2758247238218996, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 25337 + }, + { + "epoch": 0.25338, + "grad_norm": 1.9076240102285584, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 25338 + }, + { + "epoch": 0.25339, + "grad_norm": 1.1404905957926628, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 25339 + }, + { + "epoch": 0.2534, + "grad_norm": 1.8523073011824807, + "learning_rate": 0.003, + "loss": 4.069, + "step": 25340 + }, + { + "epoch": 0.25341, + "grad_norm": 1.1194040882344323, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25341 + }, + { + "epoch": 0.25342, + "grad_norm": 1.549142310175357, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25342 + }, + { + "epoch": 0.25343, + "grad_norm": 1.2006496823268282, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 25343 + }, + { + "epoch": 0.25344, + "grad_norm": 1.6631956676763877, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 25344 + }, + { + "epoch": 0.25345, + "grad_norm": 1.2454918394718975, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 25345 + }, + { + "epoch": 0.25346, + "grad_norm": 1.7132471770695137, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 25346 + }, + { + "epoch": 0.25347, + "grad_norm": 1.306744519301823, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 25347 + }, + { + "epoch": 0.25348, + "grad_norm": 1.8369378981985613, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 25348 + }, + { + "epoch": 0.25349, + "grad_norm": 1.2057702365716583, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 25349 + }, + { + "epoch": 0.2535, + "grad_norm": 1.3978518034395102, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 25350 + }, + { + "epoch": 0.25351, + "grad_norm": 1.5167059645036451, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 25351 + }, + { + "epoch": 0.25352, + "grad_norm": 1.1736513542821312, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 25352 + }, + { + "epoch": 0.25353, + "grad_norm": 1.366786288754727, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 25353 + }, + { + "epoch": 0.25354, + "grad_norm": 1.5529538620022543, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25354 + }, + { + "epoch": 0.25355, + "grad_norm": 1.4534921537422647, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25355 + }, + { + "epoch": 0.25356, + "grad_norm": 1.1231021103701653, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 25356 + }, + { + "epoch": 0.25357, + "grad_norm": 1.398818329809975, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 25357 + }, + { + "epoch": 0.25358, + "grad_norm": 1.3066952279349477, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 25358 + }, + { + "epoch": 0.25359, + "grad_norm": 1.5085211200614557, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25359 + }, + { + "epoch": 0.2536, + "grad_norm": 1.3792822156418256, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 25360 + }, + { + "epoch": 0.25361, + "grad_norm": 1.437914146825342, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 25361 + }, + { + "epoch": 0.25362, + "grad_norm": 1.3442472799101133, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25362 + }, + { + "epoch": 0.25363, + "grad_norm": 1.2775616597390271, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 25363 + }, + { + "epoch": 0.25364, + "grad_norm": 1.4903277099078978, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 25364 + }, + { + "epoch": 0.25365, + "grad_norm": 1.0867282015038917, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 25365 + }, + { + "epoch": 0.25366, + "grad_norm": 1.5969745409458787, + "learning_rate": 0.003, + "loss": 4.083, + "step": 25366 + }, + { + "epoch": 0.25367, + "grad_norm": 1.354055787199915, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 25367 + }, + { + "epoch": 0.25368, + "grad_norm": 1.4261475991246289, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 25368 + }, + { + "epoch": 0.25369, + "grad_norm": 1.4494309302630086, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25369 + }, + { + "epoch": 0.2537, + "grad_norm": 1.6908373777218801, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 25370 + }, + { + "epoch": 0.25371, + "grad_norm": 1.2693320946214979, + "learning_rate": 0.003, + "loss": 4.036, + "step": 25371 + }, + { + "epoch": 0.25372, + "grad_norm": 1.5912014458095827, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25372 + }, + { + "epoch": 0.25373, + "grad_norm": 1.429442354639141, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25373 + }, + { + "epoch": 0.25374, + "grad_norm": 1.2106835598717447, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 25374 + }, + { + "epoch": 0.25375, + "grad_norm": 1.2291352644114955, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 25375 + }, + { + "epoch": 0.25376, + "grad_norm": 1.5491030380805002, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 25376 + }, + { + "epoch": 0.25377, + "grad_norm": 1.349354695469733, + "learning_rate": 0.003, + "loss": 4.032, + "step": 25377 + }, + { + "epoch": 0.25378, + "grad_norm": 1.2550556685135807, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25378 + }, + { + "epoch": 0.25379, + "grad_norm": 1.70959528564685, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 25379 + }, + { + "epoch": 0.2538, + "grad_norm": 1.071350400387467, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 25380 + }, + { + "epoch": 0.25381, + "grad_norm": 1.4934836952308728, + "learning_rate": 0.003, + "loss": 4.058, + "step": 25381 + }, + { + "epoch": 0.25382, + "grad_norm": 1.253833647497224, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 25382 + }, + { + "epoch": 0.25383, + "grad_norm": 1.4168675618424733, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 25383 + }, + { + "epoch": 0.25384, + "grad_norm": 1.4055443780686308, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25384 + }, + { + "epoch": 0.25385, + "grad_norm": 1.3802201908647092, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 25385 + }, + { + "epoch": 0.25386, + "grad_norm": 1.61344484365093, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 25386 + }, + { + "epoch": 0.25387, + "grad_norm": 1.4473526834834542, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 25387 + }, + { + "epoch": 0.25388, + "grad_norm": 1.3572345545708926, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25388 + }, + { + "epoch": 0.25389, + "grad_norm": 1.4380307115348705, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25389 + }, + { + "epoch": 0.2539, + "grad_norm": 1.5985142399858883, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 25390 + }, + { + "epoch": 0.25391, + "grad_norm": 1.1681266889310944, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 25391 + }, + { + "epoch": 0.25392, + "grad_norm": 1.6881797846404536, + "learning_rate": 0.003, + "loss": 4.056, + "step": 25392 + }, + { + "epoch": 0.25393, + "grad_norm": 0.9404537895376099, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 25393 + }, + { + "epoch": 0.25394, + "grad_norm": 1.5971862911816268, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 25394 + }, + { + "epoch": 0.25395, + "grad_norm": 1.3150112109260046, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 25395 + }, + { + "epoch": 0.25396, + "grad_norm": 1.5841399325369836, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25396 + }, + { + "epoch": 0.25397, + "grad_norm": 1.2896557300443863, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 25397 + }, + { + "epoch": 0.25398, + "grad_norm": 1.4597374381102795, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 25398 + }, + { + "epoch": 0.25399, + "grad_norm": 1.2407152853191694, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 25399 + }, + { + "epoch": 0.254, + "grad_norm": 1.7234638239605042, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25400 + }, + { + "epoch": 0.25401, + "grad_norm": 1.1062642598266486, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 25401 + }, + { + "epoch": 0.25402, + "grad_norm": 1.5092134498788248, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25402 + }, + { + "epoch": 0.25403, + "grad_norm": 1.3190106082444915, + "learning_rate": 0.003, + "loss": 4.052, + "step": 25403 + }, + { + "epoch": 0.25404, + "grad_norm": 1.3631759003200474, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25404 + }, + { + "epoch": 0.25405, + "grad_norm": 1.2454031135005021, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 25405 + }, + { + "epoch": 0.25406, + "grad_norm": 1.3316368945211745, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 25406 + }, + { + "epoch": 0.25407, + "grad_norm": 1.216654701130413, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 25407 + }, + { + "epoch": 0.25408, + "grad_norm": 1.3947282269274737, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 25408 + }, + { + "epoch": 0.25409, + "grad_norm": 1.045474083004841, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 25409 + }, + { + "epoch": 0.2541, + "grad_norm": 1.3733577526266227, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 25410 + }, + { + "epoch": 0.25411, + "grad_norm": 1.356634390863771, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 25411 + }, + { + "epoch": 0.25412, + "grad_norm": 1.4267529153713783, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25412 + }, + { + "epoch": 0.25413, + "grad_norm": 1.3964855542087644, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 25413 + }, + { + "epoch": 0.25414, + "grad_norm": 1.7791704413924154, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25414 + }, + { + "epoch": 0.25415, + "grad_norm": 1.7759494880818907, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 25415 + }, + { + "epoch": 0.25416, + "grad_norm": 1.0945053015374115, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 25416 + }, + { + "epoch": 0.25417, + "grad_norm": 1.5732961289276437, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25417 + }, + { + "epoch": 0.25418, + "grad_norm": 1.7129251975752757, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 25418 + }, + { + "epoch": 0.25419, + "grad_norm": 1.213056209907701, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 25419 + }, + { + "epoch": 0.2542, + "grad_norm": 1.3752080791935162, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25420 + }, + { + "epoch": 0.25421, + "grad_norm": 1.6566408869926255, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 25421 + }, + { + "epoch": 0.25422, + "grad_norm": 1.1857458251141733, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 25422 + }, + { + "epoch": 0.25423, + "grad_norm": 1.5654506277682456, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 25423 + }, + { + "epoch": 0.25424, + "grad_norm": 1.3534971447060011, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 25424 + }, + { + "epoch": 0.25425, + "grad_norm": 1.3945769427685644, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 25425 + }, + { + "epoch": 0.25426, + "grad_norm": 1.417318171662554, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25426 + }, + { + "epoch": 0.25427, + "grad_norm": 1.1261987609957194, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 25427 + }, + { + "epoch": 0.25428, + "grad_norm": 1.3967926642107085, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 25428 + }, + { + "epoch": 0.25429, + "grad_norm": 1.3087608074691588, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 25429 + }, + { + "epoch": 0.2543, + "grad_norm": 1.515608708485061, + "learning_rate": 0.003, + "loss": 4.048, + "step": 25430 + }, + { + "epoch": 0.25431, + "grad_norm": 1.6166523837265636, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 25431 + }, + { + "epoch": 0.25432, + "grad_norm": 1.290464585157251, + "learning_rate": 0.003, + "loss": 4.042, + "step": 25432 + }, + { + "epoch": 0.25433, + "grad_norm": 1.4610107350983912, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 25433 + }, + { + "epoch": 0.25434, + "grad_norm": 1.3292558713759466, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 25434 + }, + { + "epoch": 0.25435, + "grad_norm": 1.4190979527478365, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25435 + }, + { + "epoch": 0.25436, + "grad_norm": 1.4469213125140579, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 25436 + }, + { + "epoch": 0.25437, + "grad_norm": 1.5297520558522897, + "learning_rate": 0.003, + "loss": 4.025, + "step": 25437 + }, + { + "epoch": 0.25438, + "grad_norm": 1.631054984413253, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 25438 + }, + { + "epoch": 0.25439, + "grad_norm": 1.4467444768516275, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25439 + }, + { + "epoch": 0.2544, + "grad_norm": 1.4579086730313882, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 25440 + }, + { + "epoch": 0.25441, + "grad_norm": 1.5767669818177323, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 25441 + }, + { + "epoch": 0.25442, + "grad_norm": 1.0968711455614821, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 25442 + }, + { + "epoch": 0.25443, + "grad_norm": 1.419885756274376, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 25443 + }, + { + "epoch": 0.25444, + "grad_norm": 1.3317378902711083, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 25444 + }, + { + "epoch": 0.25445, + "grad_norm": 1.540388880840489, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 25445 + }, + { + "epoch": 0.25446, + "grad_norm": 1.2031396547544115, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 25446 + }, + { + "epoch": 0.25447, + "grad_norm": 1.2472168293952166, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 25447 + }, + { + "epoch": 0.25448, + "grad_norm": 1.4286423213250237, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 25448 + }, + { + "epoch": 0.25449, + "grad_norm": 1.3767987734319982, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 25449 + }, + { + "epoch": 0.2545, + "grad_norm": 1.2072036752512636, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 25450 + }, + { + "epoch": 0.25451, + "grad_norm": 1.3885889860302425, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 25451 + }, + { + "epoch": 0.25452, + "grad_norm": 1.2807127650982557, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 25452 + }, + { + "epoch": 0.25453, + "grad_norm": 1.4622716565462128, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25453 + }, + { + "epoch": 0.25454, + "grad_norm": 1.3984585133403915, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 25454 + }, + { + "epoch": 0.25455, + "grad_norm": 1.722589543330776, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25455 + }, + { + "epoch": 0.25456, + "grad_norm": 1.049234477652921, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 25456 + }, + { + "epoch": 0.25457, + "grad_norm": 1.6585121093231572, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 25457 + }, + { + "epoch": 0.25458, + "grad_norm": 1.6568689101169995, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25458 + }, + { + "epoch": 0.25459, + "grad_norm": 1.3065026531125563, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 25459 + }, + { + "epoch": 0.2546, + "grad_norm": 1.6461219953234143, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 25460 + }, + { + "epoch": 0.25461, + "grad_norm": 1.4595907393033671, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 25461 + }, + { + "epoch": 0.25462, + "grad_norm": 1.4673569175305712, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 25462 + }, + { + "epoch": 0.25463, + "grad_norm": 1.343289518619863, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 25463 + }, + { + "epoch": 0.25464, + "grad_norm": 1.5496183489886368, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25464 + }, + { + "epoch": 0.25465, + "grad_norm": 1.245319571820456, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25465 + }, + { + "epoch": 0.25466, + "grad_norm": 1.5463524494688146, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25466 + }, + { + "epoch": 0.25467, + "grad_norm": 1.4283336557576143, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 25467 + }, + { + "epoch": 0.25468, + "grad_norm": 1.3001518310217621, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25468 + }, + { + "epoch": 0.25469, + "grad_norm": 1.2773717276405157, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 25469 + }, + { + "epoch": 0.2547, + "grad_norm": 1.38340031269305, + "learning_rate": 0.003, + "loss": 4.039, + "step": 25470 + }, + { + "epoch": 0.25471, + "grad_norm": 1.4828503726044868, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25471 + }, + { + "epoch": 0.25472, + "grad_norm": 1.1984016203073116, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25472 + }, + { + "epoch": 0.25473, + "grad_norm": 1.5418996007598897, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 25473 + }, + { + "epoch": 0.25474, + "grad_norm": 1.2062283675408971, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 25474 + }, + { + "epoch": 0.25475, + "grad_norm": 1.5981634597954275, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 25475 + }, + { + "epoch": 0.25476, + "grad_norm": 1.24141003092722, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 25476 + }, + { + "epoch": 0.25477, + "grad_norm": 1.4308010302219056, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 25477 + }, + { + "epoch": 0.25478, + "grad_norm": 1.3492380519691354, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25478 + }, + { + "epoch": 0.25479, + "grad_norm": 1.6521006888376721, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 25479 + }, + { + "epoch": 0.2548, + "grad_norm": 1.2105233944318274, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 25480 + }, + { + "epoch": 0.25481, + "grad_norm": 1.8273802969537531, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25481 + }, + { + "epoch": 0.25482, + "grad_norm": 1.4317374315117672, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 25482 + }, + { + "epoch": 0.25483, + "grad_norm": 1.6709654773685005, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 25483 + }, + { + "epoch": 0.25484, + "grad_norm": 1.122015636225175, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25484 + }, + { + "epoch": 0.25485, + "grad_norm": 1.8720606423002597, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25485 + }, + { + "epoch": 0.25486, + "grad_norm": 1.2414872521489775, + "learning_rate": 0.003, + "loss": 4.057, + "step": 25486 + }, + { + "epoch": 0.25487, + "grad_norm": 1.3501391275763572, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 25487 + }, + { + "epoch": 0.25488, + "grad_norm": 1.2657488936108148, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 25488 + }, + { + "epoch": 0.25489, + "grad_norm": 1.2866693287512825, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25489 + }, + { + "epoch": 0.2549, + "grad_norm": 1.1507462943989544, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 25490 + }, + { + "epoch": 0.25491, + "grad_norm": 1.6505210383597093, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 25491 + }, + { + "epoch": 0.25492, + "grad_norm": 1.655999320337114, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 25492 + }, + { + "epoch": 0.25493, + "grad_norm": 1.489003942159722, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 25493 + }, + { + "epoch": 0.25494, + "grad_norm": 1.4469608474854367, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25494 + }, + { + "epoch": 0.25495, + "grad_norm": 1.5283449341925222, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 25495 + }, + { + "epoch": 0.25496, + "grad_norm": 1.3635115107183724, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 25496 + }, + { + "epoch": 0.25497, + "grad_norm": 1.1995007646145033, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25497 + }, + { + "epoch": 0.25498, + "grad_norm": 1.5854153604104129, + "learning_rate": 0.003, + "loss": 4.055, + "step": 25498 + }, + { + "epoch": 0.25499, + "grad_norm": 1.3648984887424485, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 25499 + }, + { + "epoch": 0.255, + "grad_norm": 1.250932743135752, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25500 + }, + { + "epoch": 0.25501, + "grad_norm": 1.287839439068214, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25501 + }, + { + "epoch": 0.25502, + "grad_norm": 1.3817175731409814, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25502 + }, + { + "epoch": 0.25503, + "grad_norm": 1.3792964016289855, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 25503 + }, + { + "epoch": 0.25504, + "grad_norm": 1.204147996878464, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25504 + }, + { + "epoch": 0.25505, + "grad_norm": 1.5397284941873677, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 25505 + }, + { + "epoch": 0.25506, + "grad_norm": 1.0146692603730831, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 25506 + }, + { + "epoch": 0.25507, + "grad_norm": 1.8067172621579008, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 25507 + }, + { + "epoch": 0.25508, + "grad_norm": 1.3891274380035143, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 25508 + }, + { + "epoch": 0.25509, + "grad_norm": 1.3754320305357979, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 25509 + }, + { + "epoch": 0.2551, + "grad_norm": 1.8460865750632554, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 25510 + }, + { + "epoch": 0.25511, + "grad_norm": 1.4117599576402156, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 25511 + }, + { + "epoch": 0.25512, + "grad_norm": 1.2375998029830446, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 25512 + }, + { + "epoch": 0.25513, + "grad_norm": 1.7249260792517664, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25513 + }, + { + "epoch": 0.25514, + "grad_norm": 1.3667986530658038, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 25514 + }, + { + "epoch": 0.25515, + "grad_norm": 1.2065183453708739, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 25515 + }, + { + "epoch": 0.25516, + "grad_norm": 1.499885123957226, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25516 + }, + { + "epoch": 0.25517, + "grad_norm": 1.6053158692127931, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 25517 + }, + { + "epoch": 0.25518, + "grad_norm": 1.1784143439425059, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 25518 + }, + { + "epoch": 0.25519, + "grad_norm": 1.663841417551379, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 25519 + }, + { + "epoch": 0.2552, + "grad_norm": 1.2454679350478837, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 25520 + }, + { + "epoch": 0.25521, + "grad_norm": 1.4526655998402291, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 25521 + }, + { + "epoch": 0.25522, + "grad_norm": 1.3503624720970302, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 25522 + }, + { + "epoch": 0.25523, + "grad_norm": 1.5321714143195468, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25523 + }, + { + "epoch": 0.25524, + "grad_norm": 1.2074238680178804, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25524 + }, + { + "epoch": 0.25525, + "grad_norm": 1.3414394790024693, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 25525 + }, + { + "epoch": 0.25526, + "grad_norm": 1.5064976773081298, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 25526 + }, + { + "epoch": 0.25527, + "grad_norm": 1.574143445729999, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 25527 + }, + { + "epoch": 0.25528, + "grad_norm": 1.3103603861903474, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 25528 + }, + { + "epoch": 0.25529, + "grad_norm": 1.4242531305185333, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25529 + }, + { + "epoch": 0.2553, + "grad_norm": 1.7002054477903017, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 25530 + }, + { + "epoch": 0.25531, + "grad_norm": 1.305686598220321, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 25531 + }, + { + "epoch": 0.25532, + "grad_norm": 1.6031756885553046, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 25532 + }, + { + "epoch": 0.25533, + "grad_norm": 1.6535578698077482, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 25533 + }, + { + "epoch": 0.25534, + "grad_norm": 1.4011224076301971, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25534 + }, + { + "epoch": 0.25535, + "grad_norm": 1.3194927226083388, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 25535 + }, + { + "epoch": 0.25536, + "grad_norm": 1.4719984387637652, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 25536 + }, + { + "epoch": 0.25537, + "grad_norm": 1.413405276629304, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 25537 + }, + { + "epoch": 0.25538, + "grad_norm": 1.2029370892731581, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 25538 + }, + { + "epoch": 0.25539, + "grad_norm": 1.4745098513515444, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 25539 + }, + { + "epoch": 0.2554, + "grad_norm": 1.4683139081476744, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 25540 + }, + { + "epoch": 0.25541, + "grad_norm": 1.4207176040469283, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 25541 + }, + { + "epoch": 0.25542, + "grad_norm": 1.345149981613017, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 25542 + }, + { + "epoch": 0.25543, + "grad_norm": 1.2976518051036532, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 25543 + }, + { + "epoch": 0.25544, + "grad_norm": 1.302915675300831, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25544 + }, + { + "epoch": 0.25545, + "grad_norm": 1.3851046737962214, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 25545 + }, + { + "epoch": 0.25546, + "grad_norm": 1.1542478170592156, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 25546 + }, + { + "epoch": 0.25547, + "grad_norm": 1.2668581097510914, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 25547 + }, + { + "epoch": 0.25548, + "grad_norm": 1.408565277981055, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 25548 + }, + { + "epoch": 0.25549, + "grad_norm": 1.379448097975971, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 25549 + }, + { + "epoch": 0.2555, + "grad_norm": 1.47746390899516, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 25550 + }, + { + "epoch": 0.25551, + "grad_norm": 1.4060110220394637, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25551 + }, + { + "epoch": 0.25552, + "grad_norm": 1.5521657059476335, + "learning_rate": 0.003, + "loss": 4.032, + "step": 25552 + }, + { + "epoch": 0.25553, + "grad_norm": 1.3460182335681414, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 25553 + }, + { + "epoch": 0.25554, + "grad_norm": 1.3951274717807889, + "learning_rate": 0.003, + "loss": 4.024, + "step": 25554 + }, + { + "epoch": 0.25555, + "grad_norm": 1.2605631411112124, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 25555 + }, + { + "epoch": 0.25556, + "grad_norm": 1.387365403509073, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25556 + }, + { + "epoch": 0.25557, + "grad_norm": 1.1365117688393942, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 25557 + }, + { + "epoch": 0.25558, + "grad_norm": 1.5523552618741925, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 25558 + }, + { + "epoch": 0.25559, + "grad_norm": 1.2852773056090736, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25559 + }, + { + "epoch": 0.2556, + "grad_norm": 1.4867682850994497, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 25560 + }, + { + "epoch": 0.25561, + "grad_norm": 1.4433288723808106, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 25561 + }, + { + "epoch": 0.25562, + "grad_norm": 1.484708219144454, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 25562 + }, + { + "epoch": 0.25563, + "grad_norm": 1.2217009187906631, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25563 + }, + { + "epoch": 0.25564, + "grad_norm": 1.6273487398124253, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25564 + }, + { + "epoch": 0.25565, + "grad_norm": 1.1418039625357443, + "learning_rate": 0.003, + "loss": 4.041, + "step": 25565 + }, + { + "epoch": 0.25566, + "grad_norm": 1.9081119274794787, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 25566 + }, + { + "epoch": 0.25567, + "grad_norm": 1.1309008990510911, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 25567 + }, + { + "epoch": 0.25568, + "grad_norm": 1.4415724374870398, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 25568 + }, + { + "epoch": 0.25569, + "grad_norm": 1.5751933103145532, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 25569 + }, + { + "epoch": 0.2557, + "grad_norm": 1.4707785212510334, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 25570 + }, + { + "epoch": 0.25571, + "grad_norm": 1.4149629158922243, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 25571 + }, + { + "epoch": 0.25572, + "grad_norm": 1.374790990623414, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25572 + }, + { + "epoch": 0.25573, + "grad_norm": 1.210199376353704, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 25573 + }, + { + "epoch": 0.25574, + "grad_norm": 1.5109900352152457, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 25574 + }, + { + "epoch": 0.25575, + "grad_norm": 1.3628665564438698, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 25575 + }, + { + "epoch": 0.25576, + "grad_norm": 1.3713555492479177, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25576 + }, + { + "epoch": 0.25577, + "grad_norm": 1.456544716391132, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25577 + }, + { + "epoch": 0.25578, + "grad_norm": 1.424732368650631, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 25578 + }, + { + "epoch": 0.25579, + "grad_norm": 1.4844816880175222, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 25579 + }, + { + "epoch": 0.2558, + "grad_norm": 1.5947452454680795, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 25580 + }, + { + "epoch": 0.25581, + "grad_norm": 1.277538620835759, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 25581 + }, + { + "epoch": 0.25582, + "grad_norm": 1.441032111193491, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25582 + }, + { + "epoch": 0.25583, + "grad_norm": 1.5593337561585785, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25583 + }, + { + "epoch": 0.25584, + "grad_norm": 1.2656265666580455, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 25584 + }, + { + "epoch": 0.25585, + "grad_norm": 1.572161089478733, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25585 + }, + { + "epoch": 0.25586, + "grad_norm": 1.3725402641669586, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 25586 + }, + { + "epoch": 0.25587, + "grad_norm": 1.2928108511865175, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 25587 + }, + { + "epoch": 0.25588, + "grad_norm": 1.7000502042071954, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 25588 + }, + { + "epoch": 0.25589, + "grad_norm": 1.0391278695442063, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25589 + }, + { + "epoch": 0.2559, + "grad_norm": 1.670549943871537, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 25590 + }, + { + "epoch": 0.25591, + "grad_norm": 1.3502080895288906, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 25591 + }, + { + "epoch": 0.25592, + "grad_norm": 1.5998127949138552, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 25592 + }, + { + "epoch": 0.25593, + "grad_norm": 1.4927779283903124, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 25593 + }, + { + "epoch": 0.25594, + "grad_norm": 1.3512325946577588, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25594 + }, + { + "epoch": 0.25595, + "grad_norm": 1.545728011810376, + "learning_rate": 0.003, + "loss": 4.054, + "step": 25595 + }, + { + "epoch": 0.25596, + "grad_norm": 1.143924173770256, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 25596 + }, + { + "epoch": 0.25597, + "grad_norm": 1.4719395316672117, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 25597 + }, + { + "epoch": 0.25598, + "grad_norm": 1.2823896317577903, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25598 + }, + { + "epoch": 0.25599, + "grad_norm": 1.4977329146274463, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 25599 + }, + { + "epoch": 0.256, + "grad_norm": 1.2146057423019825, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25600 + }, + { + "epoch": 0.25601, + "grad_norm": 1.5364502447600348, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 25601 + }, + { + "epoch": 0.25602, + "grad_norm": 1.2953270765922653, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 25602 + }, + { + "epoch": 0.25603, + "grad_norm": 1.5422478467612095, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 25603 + }, + { + "epoch": 0.25604, + "grad_norm": 1.504897084769102, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25604 + }, + { + "epoch": 0.25605, + "grad_norm": 1.227914010864788, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 25605 + }, + { + "epoch": 0.25606, + "grad_norm": 1.3701605516147022, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 25606 + }, + { + "epoch": 0.25607, + "grad_norm": 1.4373366965215018, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 25607 + }, + { + "epoch": 0.25608, + "grad_norm": 1.3863671750769027, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25608 + }, + { + "epoch": 0.25609, + "grad_norm": 1.3586256980547762, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 25609 + }, + { + "epoch": 0.2561, + "grad_norm": 1.4221874924687905, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 25610 + }, + { + "epoch": 0.25611, + "grad_norm": 1.2933945943071143, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 25611 + }, + { + "epoch": 0.25612, + "grad_norm": 1.8888523979511467, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25612 + }, + { + "epoch": 0.25613, + "grad_norm": 1.1553774649680226, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 25613 + }, + { + "epoch": 0.25614, + "grad_norm": 1.382682284795602, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 25614 + }, + { + "epoch": 0.25615, + "grad_norm": 1.5200154382208988, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 25615 + }, + { + "epoch": 0.25616, + "grad_norm": 1.3677227848629085, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 25616 + }, + { + "epoch": 0.25617, + "grad_norm": 1.3587337029225195, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 25617 + }, + { + "epoch": 0.25618, + "grad_norm": 1.3218479755322596, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25618 + }, + { + "epoch": 0.25619, + "grad_norm": 1.6378284934670104, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25619 + }, + { + "epoch": 0.2562, + "grad_norm": 1.1891262651888395, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 25620 + }, + { + "epoch": 0.25621, + "grad_norm": 1.8150342195290958, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 25621 + }, + { + "epoch": 0.25622, + "grad_norm": 1.2266477034103502, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25622 + }, + { + "epoch": 0.25623, + "grad_norm": 1.4550482710301471, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 25623 + }, + { + "epoch": 0.25624, + "grad_norm": 1.6517687336703877, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25624 + }, + { + "epoch": 0.25625, + "grad_norm": 1.2874298181925097, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 25625 + }, + { + "epoch": 0.25626, + "grad_norm": 1.5608521150812815, + "learning_rate": 0.003, + "loss": 4.026, + "step": 25626 + }, + { + "epoch": 0.25627, + "grad_norm": 1.2844718717936994, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 25627 + }, + { + "epoch": 0.25628, + "grad_norm": 1.5219868861506365, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 25628 + }, + { + "epoch": 0.25629, + "grad_norm": 1.1072690521254682, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25629 + }, + { + "epoch": 0.2563, + "grad_norm": 1.6770526752316803, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 25630 + }, + { + "epoch": 0.25631, + "grad_norm": 1.1924272460918408, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 25631 + }, + { + "epoch": 0.25632, + "grad_norm": 1.6524297052037216, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 25632 + }, + { + "epoch": 0.25633, + "grad_norm": 1.3786302641111716, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 25633 + }, + { + "epoch": 0.25634, + "grad_norm": 1.4738184669781345, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25634 + }, + { + "epoch": 0.25635, + "grad_norm": 1.4271402770465842, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25635 + }, + { + "epoch": 0.25636, + "grad_norm": 1.5840391634191457, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 25636 + }, + { + "epoch": 0.25637, + "grad_norm": 1.3924577605942232, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 25637 + }, + { + "epoch": 0.25638, + "grad_norm": 1.5127636962257702, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 25638 + }, + { + "epoch": 0.25639, + "grad_norm": 1.125998535681846, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 25639 + }, + { + "epoch": 0.2564, + "grad_norm": 1.4315293157848545, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 25640 + }, + { + "epoch": 0.25641, + "grad_norm": 1.166717423192097, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 25641 + }, + { + "epoch": 0.25642, + "grad_norm": 1.4611707778140268, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 25642 + }, + { + "epoch": 0.25643, + "grad_norm": 1.3541098870906236, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25643 + }, + { + "epoch": 0.25644, + "grad_norm": 1.439131328468955, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 25644 + }, + { + "epoch": 0.25645, + "grad_norm": 1.5520526523202323, + "learning_rate": 0.003, + "loss": 4.072, + "step": 25645 + }, + { + "epoch": 0.25646, + "grad_norm": 1.3276712651446125, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 25646 + }, + { + "epoch": 0.25647, + "grad_norm": 1.3719425216373475, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 25647 + }, + { + "epoch": 0.25648, + "grad_norm": 1.3012783655688513, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25648 + }, + { + "epoch": 0.25649, + "grad_norm": 1.3732277507595656, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 25649 + }, + { + "epoch": 0.2565, + "grad_norm": 1.2922094344270323, + "learning_rate": 0.003, + "loss": 4.047, + "step": 25650 + }, + { + "epoch": 0.25651, + "grad_norm": 1.4127963947497395, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 25651 + }, + { + "epoch": 0.25652, + "grad_norm": 1.6479457181059827, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 25652 + }, + { + "epoch": 0.25653, + "grad_norm": 1.5008279493265089, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25653 + }, + { + "epoch": 0.25654, + "grad_norm": 1.3590110056169176, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 25654 + }, + { + "epoch": 0.25655, + "grad_norm": 1.6594833834902327, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 25655 + }, + { + "epoch": 0.25656, + "grad_norm": 1.0363940422854796, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 25656 + }, + { + "epoch": 0.25657, + "grad_norm": 1.9050085010301476, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25657 + }, + { + "epoch": 0.25658, + "grad_norm": 1.0781457885298993, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 25658 + }, + { + "epoch": 0.25659, + "grad_norm": 1.5519917422087668, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 25659 + }, + { + "epoch": 0.2566, + "grad_norm": 1.432387047987245, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 25660 + }, + { + "epoch": 0.25661, + "grad_norm": 1.4014516490127664, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 25661 + }, + { + "epoch": 0.25662, + "grad_norm": 1.5124493225855538, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 25662 + }, + { + "epoch": 0.25663, + "grad_norm": 1.3104131918819921, + "learning_rate": 0.003, + "loss": 4.076, + "step": 25663 + }, + { + "epoch": 0.25664, + "grad_norm": 1.6319781966993143, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25664 + }, + { + "epoch": 0.25665, + "grad_norm": 1.40101755375911, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 25665 + }, + { + "epoch": 0.25666, + "grad_norm": 1.2432996400599061, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25666 + }, + { + "epoch": 0.25667, + "grad_norm": 1.3739814362869518, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25667 + }, + { + "epoch": 0.25668, + "grad_norm": 1.3465405971500355, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25668 + }, + { + "epoch": 0.25669, + "grad_norm": 1.0850199712691657, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 25669 + }, + { + "epoch": 0.2567, + "grad_norm": 1.5972457607952169, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 25670 + }, + { + "epoch": 0.25671, + "grad_norm": 1.3516584668589968, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25671 + }, + { + "epoch": 0.25672, + "grad_norm": 1.1213905471187569, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 25672 + }, + { + "epoch": 0.25673, + "grad_norm": 1.6695261828303127, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 25673 + }, + { + "epoch": 0.25674, + "grad_norm": 1.6851991452851967, + "learning_rate": 0.003, + "loss": 4.057, + "step": 25674 + }, + { + "epoch": 0.25675, + "grad_norm": 1.4957603568627118, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 25675 + }, + { + "epoch": 0.25676, + "grad_norm": 1.4393328321796806, + "learning_rate": 0.003, + "loss": 4.064, + "step": 25676 + }, + { + "epoch": 0.25677, + "grad_norm": 1.318697217303443, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 25677 + }, + { + "epoch": 0.25678, + "grad_norm": 1.6135854417095907, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 25678 + }, + { + "epoch": 0.25679, + "grad_norm": 1.500621109228186, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 25679 + }, + { + "epoch": 0.2568, + "grad_norm": 1.2092819112978743, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25680 + }, + { + "epoch": 0.25681, + "grad_norm": 1.5647121698206703, + "learning_rate": 0.003, + "loss": 4.045, + "step": 25681 + }, + { + "epoch": 0.25682, + "grad_norm": 1.2488500341440882, + "learning_rate": 0.003, + "loss": 4.078, + "step": 25682 + }, + { + "epoch": 0.25683, + "grad_norm": 1.3307927660995165, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 25683 + }, + { + "epoch": 0.25684, + "grad_norm": 1.2020345417632035, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25684 + }, + { + "epoch": 0.25685, + "grad_norm": 1.3690097645369093, + "learning_rate": 0.003, + "loss": 4.034, + "step": 25685 + }, + { + "epoch": 0.25686, + "grad_norm": 1.3572202738675507, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 25686 + }, + { + "epoch": 0.25687, + "grad_norm": 1.674455321942972, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 25687 + }, + { + "epoch": 0.25688, + "grad_norm": 1.5836338673174564, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 25688 + }, + { + "epoch": 0.25689, + "grad_norm": 1.3288305723084326, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 25689 + }, + { + "epoch": 0.2569, + "grad_norm": 1.2606732885831478, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 25690 + }, + { + "epoch": 0.25691, + "grad_norm": 1.6622957764526318, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 25691 + }, + { + "epoch": 0.25692, + "grad_norm": 1.4238687590002466, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 25692 + }, + { + "epoch": 0.25693, + "grad_norm": 1.2600048246169704, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25693 + }, + { + "epoch": 0.25694, + "grad_norm": 1.6758560804396716, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25694 + }, + { + "epoch": 0.25695, + "grad_norm": 1.3004927686085215, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 25695 + }, + { + "epoch": 0.25696, + "grad_norm": 1.3234999045842284, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25696 + }, + { + "epoch": 0.25697, + "grad_norm": 1.3130852781360365, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 25697 + }, + { + "epoch": 0.25698, + "grad_norm": 1.409466515261072, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 25698 + }, + { + "epoch": 0.25699, + "grad_norm": 1.953456834564288, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 25699 + }, + { + "epoch": 0.257, + "grad_norm": 1.1811700948959156, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 25700 + }, + { + "epoch": 0.25701, + "grad_norm": 1.3793797791911686, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25701 + }, + { + "epoch": 0.25702, + "grad_norm": 1.4917660759265097, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 25702 + }, + { + "epoch": 0.25703, + "grad_norm": 1.1470374359095978, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 25703 + }, + { + "epoch": 0.25704, + "grad_norm": 1.5744649941338507, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 25704 + }, + { + "epoch": 0.25705, + "grad_norm": 1.239173001518195, + "learning_rate": 0.003, + "loss": 4.042, + "step": 25705 + }, + { + "epoch": 0.25706, + "grad_norm": 1.459386466978485, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 25706 + }, + { + "epoch": 0.25707, + "grad_norm": 1.2680497930950785, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 25707 + }, + { + "epoch": 0.25708, + "grad_norm": 1.4292873426686685, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25708 + }, + { + "epoch": 0.25709, + "grad_norm": 1.4518833240958424, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 25709 + }, + { + "epoch": 0.2571, + "grad_norm": 1.4613610877265584, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25710 + }, + { + "epoch": 0.25711, + "grad_norm": 1.3905270100159586, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 25711 + }, + { + "epoch": 0.25712, + "grad_norm": 1.5639175758899766, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25712 + }, + { + "epoch": 0.25713, + "grad_norm": 1.4409150133627313, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 25713 + }, + { + "epoch": 0.25714, + "grad_norm": 1.252541035045259, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 25714 + }, + { + "epoch": 0.25715, + "grad_norm": 1.6163380686679378, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 25715 + }, + { + "epoch": 0.25716, + "grad_norm": 1.1820498209170367, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 25716 + }, + { + "epoch": 0.25717, + "grad_norm": 1.301598833424317, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 25717 + }, + { + "epoch": 0.25718, + "grad_norm": 1.2792018926899467, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25718 + }, + { + "epoch": 0.25719, + "grad_norm": 1.3501303322912859, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 25719 + }, + { + "epoch": 0.2572, + "grad_norm": 1.3177568924249792, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 25720 + }, + { + "epoch": 0.25721, + "grad_norm": 1.4874371828159054, + "learning_rate": 0.003, + "loss": 4.07, + "step": 25721 + }, + { + "epoch": 0.25722, + "grad_norm": 1.4517723696697118, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 25722 + }, + { + "epoch": 0.25723, + "grad_norm": 1.7467544869978353, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 25723 + }, + { + "epoch": 0.25724, + "grad_norm": 1.1058695475593083, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25724 + }, + { + "epoch": 0.25725, + "grad_norm": 1.6426173155731811, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25725 + }, + { + "epoch": 0.25726, + "grad_norm": 1.6369307065972698, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 25726 + }, + { + "epoch": 0.25727, + "grad_norm": 1.2751965527246372, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 25727 + }, + { + "epoch": 0.25728, + "grad_norm": 1.4786255229428193, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25728 + }, + { + "epoch": 0.25729, + "grad_norm": 1.456060724550526, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 25729 + }, + { + "epoch": 0.2573, + "grad_norm": 1.33075720510049, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 25730 + }, + { + "epoch": 0.25731, + "grad_norm": 1.366940157978389, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 25731 + }, + { + "epoch": 0.25732, + "grad_norm": 1.518974674670828, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25732 + }, + { + "epoch": 0.25733, + "grad_norm": 1.4038927932142937, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 25733 + }, + { + "epoch": 0.25734, + "grad_norm": 1.3190757994773379, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25734 + }, + { + "epoch": 0.25735, + "grad_norm": 1.4512224467330683, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 25735 + }, + { + "epoch": 0.25736, + "grad_norm": 1.3863419204178877, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 25736 + }, + { + "epoch": 0.25737, + "grad_norm": 1.6052178635467413, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 25737 + }, + { + "epoch": 0.25738, + "grad_norm": 1.2400021749999985, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 25738 + }, + { + "epoch": 0.25739, + "grad_norm": 1.4807450909830673, + "learning_rate": 0.003, + "loss": 4.028, + "step": 25739 + }, + { + "epoch": 0.2574, + "grad_norm": 1.1886029489537362, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25740 + }, + { + "epoch": 0.25741, + "grad_norm": 1.4392580724147757, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 25741 + }, + { + "epoch": 0.25742, + "grad_norm": 1.2713361839644788, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25742 + }, + { + "epoch": 0.25743, + "grad_norm": 1.4427922784725438, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 25743 + }, + { + "epoch": 0.25744, + "grad_norm": 1.5399738532177305, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 25744 + }, + { + "epoch": 0.25745, + "grad_norm": 1.1017972537118668, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 25745 + }, + { + "epoch": 0.25746, + "grad_norm": 1.4519954814509688, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25746 + }, + { + "epoch": 0.25747, + "grad_norm": 1.3861194903314558, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25747 + }, + { + "epoch": 0.25748, + "grad_norm": 1.4039774307830237, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 25748 + }, + { + "epoch": 0.25749, + "grad_norm": 1.4899174733046028, + "learning_rate": 0.003, + "loss": 4.052, + "step": 25749 + }, + { + "epoch": 0.2575, + "grad_norm": 1.2485636621023675, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 25750 + }, + { + "epoch": 0.25751, + "grad_norm": 1.3900415948773714, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25751 + }, + { + "epoch": 0.25752, + "grad_norm": 1.412189065251773, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 25752 + }, + { + "epoch": 0.25753, + "grad_norm": 1.8201969398357412, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25753 + }, + { + "epoch": 0.25754, + "grad_norm": 1.2402487925046348, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 25754 + }, + { + "epoch": 0.25755, + "grad_norm": 1.512449877552699, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25755 + }, + { + "epoch": 0.25756, + "grad_norm": 1.2832081561586905, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 25756 + }, + { + "epoch": 0.25757, + "grad_norm": 1.4919667998007922, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 25757 + }, + { + "epoch": 0.25758, + "grad_norm": 1.6293179742207706, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25758 + }, + { + "epoch": 0.25759, + "grad_norm": 1.4613255786136594, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 25759 + }, + { + "epoch": 0.2576, + "grad_norm": 1.5924353041415318, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 25760 + }, + { + "epoch": 0.25761, + "grad_norm": 1.4793877081838278, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 25761 + }, + { + "epoch": 0.25762, + "grad_norm": 1.2943966666867441, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 25762 + }, + { + "epoch": 0.25763, + "grad_norm": 1.5622737201337684, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 25763 + }, + { + "epoch": 0.25764, + "grad_norm": 1.1941422050212882, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 25764 + }, + { + "epoch": 0.25765, + "grad_norm": 1.495325959856617, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 25765 + }, + { + "epoch": 0.25766, + "grad_norm": 1.130020433845397, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 25766 + }, + { + "epoch": 0.25767, + "grad_norm": 1.708171967378693, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 25767 + }, + { + "epoch": 0.25768, + "grad_norm": 0.9975118265844913, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 25768 + }, + { + "epoch": 0.25769, + "grad_norm": 1.732516829154019, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 25769 + }, + { + "epoch": 0.2577, + "grad_norm": 1.2597843860679456, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 25770 + }, + { + "epoch": 0.25771, + "grad_norm": 1.5335957054617897, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 25771 + }, + { + "epoch": 0.25772, + "grad_norm": 1.31186682266505, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 25772 + }, + { + "epoch": 0.25773, + "grad_norm": 1.6527018481229452, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 25773 + }, + { + "epoch": 0.25774, + "grad_norm": 1.2172463731152112, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 25774 + }, + { + "epoch": 0.25775, + "grad_norm": 1.6520069739135184, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 25775 + }, + { + "epoch": 0.25776, + "grad_norm": 1.4380218738376331, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 25776 + }, + { + "epoch": 0.25777, + "grad_norm": 1.3792681041955943, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25777 + }, + { + "epoch": 0.25778, + "grad_norm": 1.3921307262827065, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 25778 + }, + { + "epoch": 0.25779, + "grad_norm": 1.3682029684365404, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 25779 + }, + { + "epoch": 0.2578, + "grad_norm": 1.233905160603916, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 25780 + }, + { + "epoch": 0.25781, + "grad_norm": 1.4994398708565877, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 25781 + }, + { + "epoch": 0.25782, + "grad_norm": 1.3428404440572903, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 25782 + }, + { + "epoch": 0.25783, + "grad_norm": 1.4965462579487645, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 25783 + }, + { + "epoch": 0.25784, + "grad_norm": 1.4961692116748513, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 25784 + }, + { + "epoch": 0.25785, + "grad_norm": 1.2021955570297993, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 25785 + }, + { + "epoch": 0.25786, + "grad_norm": 1.4632467341558268, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 25786 + }, + { + "epoch": 0.25787, + "grad_norm": 1.3477823937379005, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25787 + }, + { + "epoch": 0.25788, + "grad_norm": 1.2806958470443732, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 25788 + }, + { + "epoch": 0.25789, + "grad_norm": 1.2430195123888563, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 25789 + }, + { + "epoch": 0.2579, + "grad_norm": 1.9115407089694612, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 25790 + }, + { + "epoch": 0.25791, + "grad_norm": 1.213164328000405, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 25791 + }, + { + "epoch": 0.25792, + "grad_norm": 1.5413253583696145, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 25792 + }, + { + "epoch": 0.25793, + "grad_norm": 1.5369040950220545, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 25793 + }, + { + "epoch": 0.25794, + "grad_norm": 1.287590219552472, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 25794 + }, + { + "epoch": 0.25795, + "grad_norm": 1.3687037765605559, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25795 + }, + { + "epoch": 0.25796, + "grad_norm": 1.282031588292454, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 25796 + }, + { + "epoch": 0.25797, + "grad_norm": 1.5006802561169799, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 25797 + }, + { + "epoch": 0.25798, + "grad_norm": 1.303402137518414, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25798 + }, + { + "epoch": 0.25799, + "grad_norm": 1.326103615117249, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 25799 + }, + { + "epoch": 0.258, + "grad_norm": 1.3655312105769033, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 25800 + }, + { + "epoch": 0.25801, + "grad_norm": 1.5553004008696156, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 25801 + }, + { + "epoch": 0.25802, + "grad_norm": 1.3217126095448872, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 25802 + }, + { + "epoch": 0.25803, + "grad_norm": 1.5199853871559588, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 25803 + }, + { + "epoch": 0.25804, + "grad_norm": 1.3853183223047778, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 25804 + }, + { + "epoch": 0.25805, + "grad_norm": 1.5399550439739562, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 25805 + }, + { + "epoch": 0.25806, + "grad_norm": 1.3326718898696486, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 25806 + }, + { + "epoch": 0.25807, + "grad_norm": 1.2571037602440756, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 25807 + }, + { + "epoch": 0.25808, + "grad_norm": 1.3764438141405884, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25808 + }, + { + "epoch": 0.25809, + "grad_norm": 1.2389826627729208, + "learning_rate": 0.003, + "loss": 4.04, + "step": 25809 + }, + { + "epoch": 0.2581, + "grad_norm": 1.7141548719383164, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 25810 + }, + { + "epoch": 0.25811, + "grad_norm": 1.2884270080294644, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 25811 + }, + { + "epoch": 0.25812, + "grad_norm": 1.3525926731880507, + "learning_rate": 0.003, + "loss": 4.049, + "step": 25812 + }, + { + "epoch": 0.25813, + "grad_norm": 1.2412140720868727, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 25813 + }, + { + "epoch": 0.25814, + "grad_norm": 1.434993971607768, + "learning_rate": 0.003, + "loss": 4.081, + "step": 25814 + }, + { + "epoch": 0.25815, + "grad_norm": 1.5584543317064383, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 25815 + }, + { + "epoch": 0.25816, + "grad_norm": 1.088982180340461, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 25816 + }, + { + "epoch": 0.25817, + "grad_norm": 1.458333364605955, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 25817 + }, + { + "epoch": 0.25818, + "grad_norm": 1.2596061817867001, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 25818 + }, + { + "epoch": 0.25819, + "grad_norm": 1.3691866607641443, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 25819 + }, + { + "epoch": 0.2582, + "grad_norm": 1.4594653395400055, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 25820 + }, + { + "epoch": 0.25821, + "grad_norm": 1.633895425150386, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 25821 + }, + { + "epoch": 0.25822, + "grad_norm": 1.3673656056026295, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 25822 + }, + { + "epoch": 0.25823, + "grad_norm": 1.2212225322647978, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 25823 + }, + { + "epoch": 0.25824, + "grad_norm": 1.7094744529305244, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 25824 + }, + { + "epoch": 0.25825, + "grad_norm": 0.9633848464898301, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 25825 + }, + { + "epoch": 0.25826, + "grad_norm": 1.6949635919488364, + "learning_rate": 0.003, + "loss": 4.064, + "step": 25826 + }, + { + "epoch": 0.25827, + "grad_norm": 1.5900793526759451, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 25827 + }, + { + "epoch": 0.25828, + "grad_norm": 1.6316238400329914, + "learning_rate": 0.003, + "loss": 4.037, + "step": 25828 + }, + { + "epoch": 0.25829, + "grad_norm": 1.4877724100021106, + "learning_rate": 0.003, + "loss": 4.083, + "step": 25829 + }, + { + "epoch": 0.2583, + "grad_norm": 1.063959535273306, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 25830 + }, + { + "epoch": 0.25831, + "grad_norm": 1.693513770148217, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25831 + }, + { + "epoch": 0.25832, + "grad_norm": 1.1227150917589217, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 25832 + }, + { + "epoch": 0.25833, + "grad_norm": 1.555061526083666, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 25833 + }, + { + "epoch": 0.25834, + "grad_norm": 1.1771953174490324, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 25834 + }, + { + "epoch": 0.25835, + "grad_norm": 1.4463921611936497, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 25835 + }, + { + "epoch": 0.25836, + "grad_norm": 1.209374213224575, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 25836 + }, + { + "epoch": 0.25837, + "grad_norm": 1.4100623060802875, + "learning_rate": 0.003, + "loss": 4.034, + "step": 25837 + }, + { + "epoch": 0.25838, + "grad_norm": 1.4893210326625075, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 25838 + }, + { + "epoch": 0.25839, + "grad_norm": 1.3983396255015352, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 25839 + }, + { + "epoch": 0.2584, + "grad_norm": 1.4738249081823351, + "learning_rate": 0.003, + "loss": 4.049, + "step": 25840 + }, + { + "epoch": 0.25841, + "grad_norm": 1.3411580400428451, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 25841 + }, + { + "epoch": 0.25842, + "grad_norm": 1.590027162773852, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 25842 + }, + { + "epoch": 0.25843, + "grad_norm": 1.4321730226213878, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 25843 + }, + { + "epoch": 0.25844, + "grad_norm": 1.5059054946020332, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 25844 + }, + { + "epoch": 0.25845, + "grad_norm": 1.3981638150751274, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 25845 + }, + { + "epoch": 0.25846, + "grad_norm": 1.3661536368106877, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 25846 + }, + { + "epoch": 0.25847, + "grad_norm": 1.4269252423372771, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 25847 + }, + { + "epoch": 0.25848, + "grad_norm": 1.2746697136898264, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 25848 + }, + { + "epoch": 0.25849, + "grad_norm": 1.4284738821220324, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 25849 + }, + { + "epoch": 0.2585, + "grad_norm": 1.0158981420254636, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 25850 + }, + { + "epoch": 0.25851, + "grad_norm": 1.5779896004785696, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 25851 + }, + { + "epoch": 0.25852, + "grad_norm": 1.2983258197763992, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 25852 + }, + { + "epoch": 0.25853, + "grad_norm": 1.6845078781875717, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 25853 + }, + { + "epoch": 0.25854, + "grad_norm": 1.1986544876414655, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 25854 + }, + { + "epoch": 0.25855, + "grad_norm": 1.5729150886691499, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 25855 + }, + { + "epoch": 0.25856, + "grad_norm": 1.3392915306796007, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25856 + }, + { + "epoch": 0.25857, + "grad_norm": 1.3631361177319512, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 25857 + }, + { + "epoch": 0.25858, + "grad_norm": 1.7809939238505188, + "learning_rate": 0.003, + "loss": 4.067, + "step": 25858 + }, + { + "epoch": 0.25859, + "grad_norm": 1.543515764385466, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 25859 + }, + { + "epoch": 0.2586, + "grad_norm": 1.3737959323869435, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 25860 + }, + { + "epoch": 0.25861, + "grad_norm": 1.3540438262391012, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 25861 + }, + { + "epoch": 0.25862, + "grad_norm": 1.5414773649869888, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 25862 + }, + { + "epoch": 0.25863, + "grad_norm": 1.243365671387855, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 25863 + }, + { + "epoch": 0.25864, + "grad_norm": 1.3895325389113213, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25864 + }, + { + "epoch": 0.25865, + "grad_norm": 1.151712296482096, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 25865 + }, + { + "epoch": 0.25866, + "grad_norm": 1.4360038086544018, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25866 + }, + { + "epoch": 0.25867, + "grad_norm": 1.2929408018680493, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25867 + }, + { + "epoch": 0.25868, + "grad_norm": 1.3149036231710793, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25868 + }, + { + "epoch": 0.25869, + "grad_norm": 1.4299760667538082, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 25869 + }, + { + "epoch": 0.2587, + "grad_norm": 1.3049319738197966, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25870 + }, + { + "epoch": 0.25871, + "grad_norm": 1.8497211546857757, + "learning_rate": 0.003, + "loss": 4.039, + "step": 25871 + }, + { + "epoch": 0.25872, + "grad_norm": 1.3471229107379368, + "learning_rate": 0.003, + "loss": 4.058, + "step": 25872 + }, + { + "epoch": 0.25873, + "grad_norm": 1.5283683520592855, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 25873 + }, + { + "epoch": 0.25874, + "grad_norm": 1.3668622164834656, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25874 + }, + { + "epoch": 0.25875, + "grad_norm": 1.4353794924885628, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 25875 + }, + { + "epoch": 0.25876, + "grad_norm": 1.3529544417632908, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 25876 + }, + { + "epoch": 0.25877, + "grad_norm": 1.4498900800868253, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 25877 + }, + { + "epoch": 0.25878, + "grad_norm": 1.4364274602052436, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 25878 + }, + { + "epoch": 0.25879, + "grad_norm": 1.29028391936865, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 25879 + }, + { + "epoch": 0.2588, + "grad_norm": 1.4779018540409474, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 25880 + }, + { + "epoch": 0.25881, + "grad_norm": 1.2336714565414102, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 25881 + }, + { + "epoch": 0.25882, + "grad_norm": 1.7462766422344733, + "learning_rate": 0.003, + "loss": 4.061, + "step": 25882 + }, + { + "epoch": 0.25883, + "grad_norm": 1.3169954918792477, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 25883 + }, + { + "epoch": 0.25884, + "grad_norm": 1.584913381609622, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 25884 + }, + { + "epoch": 0.25885, + "grad_norm": 1.3720890571185853, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 25885 + }, + { + "epoch": 0.25886, + "grad_norm": 1.6966985311526994, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 25886 + }, + { + "epoch": 0.25887, + "grad_norm": 1.2111718973440997, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 25887 + }, + { + "epoch": 0.25888, + "grad_norm": 1.5375861801955801, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 25888 + }, + { + "epoch": 0.25889, + "grad_norm": 1.5573593268619124, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 25889 + }, + { + "epoch": 0.2589, + "grad_norm": 1.3598938148443485, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25890 + }, + { + "epoch": 0.25891, + "grad_norm": 1.536737696286913, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 25891 + }, + { + "epoch": 0.25892, + "grad_norm": 1.2306886748574855, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25892 + }, + { + "epoch": 0.25893, + "grad_norm": 1.5911207196467794, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 25893 + }, + { + "epoch": 0.25894, + "grad_norm": 1.5960649998147831, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25894 + }, + { + "epoch": 0.25895, + "grad_norm": 1.8014799959070236, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 25895 + }, + { + "epoch": 0.25896, + "grad_norm": 1.1943238857273202, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 25896 + }, + { + "epoch": 0.25897, + "grad_norm": 1.4254415322533194, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 25897 + }, + { + "epoch": 0.25898, + "grad_norm": 1.7471029558675286, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 25898 + }, + { + "epoch": 0.25899, + "grad_norm": 1.0702999858831728, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 25899 + }, + { + "epoch": 0.259, + "grad_norm": 1.6960684513089446, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 25900 + }, + { + "epoch": 0.25901, + "grad_norm": 1.4095364530171197, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 25901 + }, + { + "epoch": 0.25902, + "grad_norm": 1.5962430791332929, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 25902 + }, + { + "epoch": 0.25903, + "grad_norm": 1.2283704094120844, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 25903 + }, + { + "epoch": 0.25904, + "grad_norm": 1.4244908969617915, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 25904 + }, + { + "epoch": 0.25905, + "grad_norm": 1.2567023770419468, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 25905 + }, + { + "epoch": 0.25906, + "grad_norm": 1.4056920028431878, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 25906 + }, + { + "epoch": 0.25907, + "grad_norm": 1.1940805887856745, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 25907 + }, + { + "epoch": 0.25908, + "grad_norm": 1.443914232137034, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 25908 + }, + { + "epoch": 0.25909, + "grad_norm": 1.2586347841342749, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 25909 + }, + { + "epoch": 0.2591, + "grad_norm": 1.4355098144136065, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 25910 + }, + { + "epoch": 0.25911, + "grad_norm": 1.9966617665864, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 25911 + }, + { + "epoch": 0.25912, + "grad_norm": 1.2705773935075908, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 25912 + }, + { + "epoch": 0.25913, + "grad_norm": 1.6154203845070347, + "learning_rate": 0.003, + "loss": 4.045, + "step": 25913 + }, + { + "epoch": 0.25914, + "grad_norm": 1.2523853860935699, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 25914 + }, + { + "epoch": 0.25915, + "grad_norm": 1.3920245565841947, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 25915 + }, + { + "epoch": 0.25916, + "grad_norm": 1.3577179122459062, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 25916 + }, + { + "epoch": 0.25917, + "grad_norm": 1.4608612031271264, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 25917 + }, + { + "epoch": 0.25918, + "grad_norm": 1.3977284702206003, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 25918 + }, + { + "epoch": 0.25919, + "grad_norm": 1.2728415589647335, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 25919 + }, + { + "epoch": 0.2592, + "grad_norm": 1.4042861080005589, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25920 + }, + { + "epoch": 0.25921, + "grad_norm": 1.367977769991152, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25921 + }, + { + "epoch": 0.25922, + "grad_norm": 1.4001055246272784, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 25922 + }, + { + "epoch": 0.25923, + "grad_norm": 1.5724131637743832, + "learning_rate": 0.003, + "loss": 4.056, + "step": 25923 + }, + { + "epoch": 0.25924, + "grad_norm": 1.7570110282193514, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25924 + }, + { + "epoch": 0.25925, + "grad_norm": 1.2471720108652042, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 25925 + }, + { + "epoch": 0.25926, + "grad_norm": 1.7590488070068033, + "learning_rate": 0.003, + "loss": 4.061, + "step": 25926 + }, + { + "epoch": 0.25927, + "grad_norm": 1.1566800277862708, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25927 + }, + { + "epoch": 0.25928, + "grad_norm": 1.6654547912856716, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 25928 + }, + { + "epoch": 0.25929, + "grad_norm": 1.4548410622588641, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 25929 + }, + { + "epoch": 0.2593, + "grad_norm": 1.3240441326667984, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 25930 + }, + { + "epoch": 0.25931, + "grad_norm": 1.3468820633157403, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 25931 + }, + { + "epoch": 0.25932, + "grad_norm": 1.234752623714711, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 25932 + }, + { + "epoch": 0.25933, + "grad_norm": 1.489783748503926, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 25933 + }, + { + "epoch": 0.25934, + "grad_norm": 1.5703388537052494, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 25934 + }, + { + "epoch": 0.25935, + "grad_norm": 1.359146624253078, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 25935 + }, + { + "epoch": 0.25936, + "grad_norm": 1.4311469619932031, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 25936 + }, + { + "epoch": 0.25937, + "grad_norm": 1.548093724859345, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 25937 + }, + { + "epoch": 0.25938, + "grad_norm": 1.1705270753155963, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25938 + }, + { + "epoch": 0.25939, + "grad_norm": 1.4512579070685991, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 25939 + }, + { + "epoch": 0.2594, + "grad_norm": 1.383681518901621, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 25940 + }, + { + "epoch": 0.25941, + "grad_norm": 1.4528343122391336, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 25941 + }, + { + "epoch": 0.25942, + "grad_norm": 1.2724837522093335, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 25942 + }, + { + "epoch": 0.25943, + "grad_norm": 1.335361592598783, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 25943 + }, + { + "epoch": 0.25944, + "grad_norm": 1.8521552296619705, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 25944 + }, + { + "epoch": 0.25945, + "grad_norm": 1.2096373900506994, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 25945 + }, + { + "epoch": 0.25946, + "grad_norm": 1.811153102155675, + "learning_rate": 0.003, + "loss": 4.056, + "step": 25946 + }, + { + "epoch": 0.25947, + "grad_norm": 1.318550659821529, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 25947 + }, + { + "epoch": 0.25948, + "grad_norm": 1.4040401329610572, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 25948 + }, + { + "epoch": 0.25949, + "grad_norm": 1.2728069476774184, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25949 + }, + { + "epoch": 0.2595, + "grad_norm": 1.572538744908175, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 25950 + }, + { + "epoch": 0.25951, + "grad_norm": 1.3586307321276074, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 25951 + }, + { + "epoch": 0.25952, + "grad_norm": 1.6110430967188807, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 25952 + }, + { + "epoch": 0.25953, + "grad_norm": 1.2698854533034047, + "learning_rate": 0.003, + "loss": 4.07, + "step": 25953 + }, + { + "epoch": 0.25954, + "grad_norm": 1.375021416491968, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 25954 + }, + { + "epoch": 0.25955, + "grad_norm": 1.1818258227142686, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 25955 + }, + { + "epoch": 0.25956, + "grad_norm": 1.4977947701076593, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 25956 + }, + { + "epoch": 0.25957, + "grad_norm": 1.0816061410918312, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 25957 + }, + { + "epoch": 0.25958, + "grad_norm": 1.6780020100966895, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 25958 + }, + { + "epoch": 0.25959, + "grad_norm": 1.0487390752286232, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 25959 + }, + { + "epoch": 0.2596, + "grad_norm": 1.4932867315397373, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 25960 + }, + { + "epoch": 0.25961, + "grad_norm": 1.2606805109056465, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25961 + }, + { + "epoch": 0.25962, + "grad_norm": 1.4666911533275648, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 25962 + }, + { + "epoch": 0.25963, + "grad_norm": 1.541871918588217, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 25963 + }, + { + "epoch": 0.25964, + "grad_norm": 1.6697087055813573, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 25964 + }, + { + "epoch": 0.25965, + "grad_norm": 1.2259853941984973, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 25965 + }, + { + "epoch": 0.25966, + "grad_norm": 1.5250675842116186, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 25966 + }, + { + "epoch": 0.25967, + "grad_norm": 1.3753817342983226, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 25967 + }, + { + "epoch": 0.25968, + "grad_norm": 1.376575143593774, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 25968 + }, + { + "epoch": 0.25969, + "grad_norm": 1.2994905161975587, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 25969 + }, + { + "epoch": 0.2597, + "grad_norm": 1.5169319110368726, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 25970 + }, + { + "epoch": 0.25971, + "grad_norm": 1.2386824205147875, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 25971 + }, + { + "epoch": 0.25972, + "grad_norm": 1.7383666529143638, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 25972 + }, + { + "epoch": 0.25973, + "grad_norm": 1.2663148660338621, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 25973 + }, + { + "epoch": 0.25974, + "grad_norm": 1.410954093833394, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 25974 + }, + { + "epoch": 0.25975, + "grad_norm": 1.3959748346323602, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 25975 + }, + { + "epoch": 0.25976, + "grad_norm": 1.6033639859116886, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 25976 + }, + { + "epoch": 0.25977, + "grad_norm": 1.015590349717408, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 25977 + }, + { + "epoch": 0.25978, + "grad_norm": 1.5386007579331171, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 25978 + }, + { + "epoch": 0.25979, + "grad_norm": 1.2730667385894219, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 25979 + }, + { + "epoch": 0.2598, + "grad_norm": 1.6525263209786583, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 25980 + }, + { + "epoch": 0.25981, + "grad_norm": 1.320399834548752, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 25981 + }, + { + "epoch": 0.25982, + "grad_norm": 1.4240212741695537, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 25982 + }, + { + "epoch": 0.25983, + "grad_norm": 1.2681146865078845, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 25983 + }, + { + "epoch": 0.25984, + "grad_norm": 1.350591087072189, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 25984 + }, + { + "epoch": 0.25985, + "grad_norm": 1.3509651928041615, + "learning_rate": 0.003, + "loss": 4.016, + "step": 25985 + }, + { + "epoch": 0.25986, + "grad_norm": 1.7312014585479307, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 25986 + }, + { + "epoch": 0.25987, + "grad_norm": 1.3613277487845337, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 25987 + }, + { + "epoch": 0.25988, + "grad_norm": 1.5766365237863522, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 25988 + }, + { + "epoch": 0.25989, + "grad_norm": 1.3156267855734007, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 25989 + }, + { + "epoch": 0.2599, + "grad_norm": 1.3163374138023691, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 25990 + }, + { + "epoch": 0.25991, + "grad_norm": 1.2630937017814599, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 25991 + }, + { + "epoch": 0.25992, + "grad_norm": 1.5870309025748812, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 25992 + }, + { + "epoch": 0.25993, + "grad_norm": 1.1307578692165547, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 25993 + }, + { + "epoch": 0.25994, + "grad_norm": 1.656157416411466, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 25994 + }, + { + "epoch": 0.25995, + "grad_norm": 1.0316959825670473, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 25995 + }, + { + "epoch": 0.25996, + "grad_norm": 1.6570808197962392, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 25996 + }, + { + "epoch": 0.25997, + "grad_norm": 1.2620136405913034, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 25997 + }, + { + "epoch": 0.25998, + "grad_norm": 1.550969458116143, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 25998 + }, + { + "epoch": 0.25999, + "grad_norm": 1.634554826716528, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 25999 + }, + { + "epoch": 0.26, + "grad_norm": 1.6771422883617966, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 26000 + }, + { + "epoch": 0.26001, + "grad_norm": 1.3000847438120544, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26001 + }, + { + "epoch": 0.26002, + "grad_norm": 1.3364011285435617, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 26002 + }, + { + "epoch": 0.26003, + "grad_norm": 1.4479150126116704, + "learning_rate": 0.003, + "loss": 4.034, + "step": 26003 + }, + { + "epoch": 0.26004, + "grad_norm": 1.394729084370741, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 26004 + }, + { + "epoch": 0.26005, + "grad_norm": 1.5138617264075938, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 26005 + }, + { + "epoch": 0.26006, + "grad_norm": 1.2558673059973169, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 26006 + }, + { + "epoch": 0.26007, + "grad_norm": 1.4482098902920977, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 26007 + }, + { + "epoch": 0.26008, + "grad_norm": 1.1075884570881904, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 26008 + }, + { + "epoch": 0.26009, + "grad_norm": 1.8200571984759495, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26009 + }, + { + "epoch": 0.2601, + "grad_norm": 1.123013228323243, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 26010 + }, + { + "epoch": 0.26011, + "grad_norm": 1.6880099673690845, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 26011 + }, + { + "epoch": 0.26012, + "grad_norm": 1.245545733649582, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 26012 + }, + { + "epoch": 0.26013, + "grad_norm": 1.4617701896086133, + "learning_rate": 0.003, + "loss": 4.045, + "step": 26013 + }, + { + "epoch": 0.26014, + "grad_norm": 1.4665438188745272, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26014 + }, + { + "epoch": 0.26015, + "grad_norm": 1.1128924571298922, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 26015 + }, + { + "epoch": 0.26016, + "grad_norm": 1.5196637771774784, + "learning_rate": 0.003, + "loss": 4.046, + "step": 26016 + }, + { + "epoch": 0.26017, + "grad_norm": 1.2243750284669412, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26017 + }, + { + "epoch": 0.26018, + "grad_norm": 1.3147878411080076, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 26018 + }, + { + "epoch": 0.26019, + "grad_norm": 1.4777129697682345, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 26019 + }, + { + "epoch": 0.2602, + "grad_norm": 1.5746945322454176, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26020 + }, + { + "epoch": 0.26021, + "grad_norm": 0.9744066819755666, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26021 + }, + { + "epoch": 0.26022, + "grad_norm": 1.7475202563292984, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 26022 + }, + { + "epoch": 0.26023, + "grad_norm": 1.0612220389125266, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 26023 + }, + { + "epoch": 0.26024, + "grad_norm": 1.7908487427214232, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 26024 + }, + { + "epoch": 0.26025, + "grad_norm": 1.2681021000773447, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26025 + }, + { + "epoch": 0.26026, + "grad_norm": 1.670406405104956, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 26026 + }, + { + "epoch": 0.26027, + "grad_norm": 1.3180229925142966, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26027 + }, + { + "epoch": 0.26028, + "grad_norm": 1.333434148640585, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 26028 + }, + { + "epoch": 0.26029, + "grad_norm": 1.3816312904431067, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26029 + }, + { + "epoch": 0.2603, + "grad_norm": 1.347053548585626, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 26030 + }, + { + "epoch": 0.26031, + "grad_norm": 1.20570786935628, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 26031 + }, + { + "epoch": 0.26032, + "grad_norm": 1.2960635634395246, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 26032 + }, + { + "epoch": 0.26033, + "grad_norm": 1.6796038115054543, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26033 + }, + { + "epoch": 0.26034, + "grad_norm": 1.5689534390815558, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26034 + }, + { + "epoch": 0.26035, + "grad_norm": 1.7653666673430233, + "learning_rate": 0.003, + "loss": 4.05, + "step": 26035 + }, + { + "epoch": 0.26036, + "grad_norm": 1.1682564293007847, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 26036 + }, + { + "epoch": 0.26037, + "grad_norm": 1.2662943422251716, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 26037 + }, + { + "epoch": 0.26038, + "grad_norm": 1.4628023392545204, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26038 + }, + { + "epoch": 0.26039, + "grad_norm": 1.6367725642503304, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 26039 + }, + { + "epoch": 0.2604, + "grad_norm": 1.5564871568348417, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 26040 + }, + { + "epoch": 0.26041, + "grad_norm": 1.3672150892008819, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 26041 + }, + { + "epoch": 0.26042, + "grad_norm": 1.3562490234381208, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 26042 + }, + { + "epoch": 0.26043, + "grad_norm": 1.3009097512533194, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 26043 + }, + { + "epoch": 0.26044, + "grad_norm": 1.427006353192353, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 26044 + }, + { + "epoch": 0.26045, + "grad_norm": 1.4650344780835123, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26045 + }, + { + "epoch": 0.26046, + "grad_norm": 1.1682246232436817, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26046 + }, + { + "epoch": 0.26047, + "grad_norm": 1.5647224861528404, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 26047 + }, + { + "epoch": 0.26048, + "grad_norm": 1.344323043651862, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 26048 + }, + { + "epoch": 0.26049, + "grad_norm": 1.1804076945784996, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 26049 + }, + { + "epoch": 0.2605, + "grad_norm": 1.3565233886072303, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26050 + }, + { + "epoch": 0.26051, + "grad_norm": 1.334867485121708, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 26051 + }, + { + "epoch": 0.26052, + "grad_norm": 1.5604126320035954, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 26052 + }, + { + "epoch": 0.26053, + "grad_norm": 1.4028125607794886, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26053 + }, + { + "epoch": 0.26054, + "grad_norm": 1.212674808976009, + "learning_rate": 0.003, + "loss": 4.071, + "step": 26054 + }, + { + "epoch": 0.26055, + "grad_norm": 1.4854827836792286, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 26055 + }, + { + "epoch": 0.26056, + "grad_norm": 1.1907581807444654, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26056 + }, + { + "epoch": 0.26057, + "grad_norm": 1.395862081954804, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 26057 + }, + { + "epoch": 0.26058, + "grad_norm": 1.2386895651185899, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 26058 + }, + { + "epoch": 0.26059, + "grad_norm": 1.7491747734174596, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 26059 + }, + { + "epoch": 0.2606, + "grad_norm": 1.0133824314529083, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26060 + }, + { + "epoch": 0.26061, + "grad_norm": 1.5318839883866462, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26061 + }, + { + "epoch": 0.26062, + "grad_norm": 1.2517014631237353, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 26062 + }, + { + "epoch": 0.26063, + "grad_norm": 1.4869871466358777, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 26063 + }, + { + "epoch": 0.26064, + "grad_norm": 1.2907941090981765, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26064 + }, + { + "epoch": 0.26065, + "grad_norm": 1.7331314533588165, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 26065 + }, + { + "epoch": 0.26066, + "grad_norm": 1.3532005116567623, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 26066 + }, + { + "epoch": 0.26067, + "grad_norm": 1.5586726948735754, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 26067 + }, + { + "epoch": 0.26068, + "grad_norm": 1.4994645976368646, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 26068 + }, + { + "epoch": 0.26069, + "grad_norm": 1.4540661450748456, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 26069 + }, + { + "epoch": 0.2607, + "grad_norm": 1.2753167114488906, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 26070 + }, + { + "epoch": 0.26071, + "grad_norm": 1.3989215677074158, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 26071 + }, + { + "epoch": 0.26072, + "grad_norm": 1.2254396520873279, + "learning_rate": 0.003, + "loss": 4.011, + "step": 26072 + }, + { + "epoch": 0.26073, + "grad_norm": 1.3416193311526765, + "learning_rate": 0.003, + "loss": 4.041, + "step": 26073 + }, + { + "epoch": 0.26074, + "grad_norm": 1.5703609339738938, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 26074 + }, + { + "epoch": 0.26075, + "grad_norm": 1.198930576691343, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 26075 + }, + { + "epoch": 0.26076, + "grad_norm": 1.3862403195184618, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 26076 + }, + { + "epoch": 0.26077, + "grad_norm": 1.3063365646694558, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26077 + }, + { + "epoch": 0.26078, + "grad_norm": 1.3596624551541052, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 26078 + }, + { + "epoch": 0.26079, + "grad_norm": 1.303304808060919, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26079 + }, + { + "epoch": 0.2608, + "grad_norm": 1.6157004753339963, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 26080 + }, + { + "epoch": 0.26081, + "grad_norm": 1.6829511918828395, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 26081 + }, + { + "epoch": 0.26082, + "grad_norm": 1.5093893903882467, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 26082 + }, + { + "epoch": 0.26083, + "grad_norm": 1.2861196379696564, + "learning_rate": 0.003, + "loss": 4.066, + "step": 26083 + }, + { + "epoch": 0.26084, + "grad_norm": 1.5708613832964653, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 26084 + }, + { + "epoch": 0.26085, + "grad_norm": 1.2759365548080701, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 26085 + }, + { + "epoch": 0.26086, + "grad_norm": 1.5720956478548702, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26086 + }, + { + "epoch": 0.26087, + "grad_norm": 1.3410929907507505, + "learning_rate": 0.003, + "loss": 4.04, + "step": 26087 + }, + { + "epoch": 0.26088, + "grad_norm": 1.3123356345568025, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 26088 + }, + { + "epoch": 0.26089, + "grad_norm": 1.468790186261554, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 26089 + }, + { + "epoch": 0.2609, + "grad_norm": 1.2836672722071651, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26090 + }, + { + "epoch": 0.26091, + "grad_norm": 1.4875782079669955, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 26091 + }, + { + "epoch": 0.26092, + "grad_norm": 1.1871928856044425, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 26092 + }, + { + "epoch": 0.26093, + "grad_norm": 1.529144439853646, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26093 + }, + { + "epoch": 0.26094, + "grad_norm": 1.2230356183303084, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 26094 + }, + { + "epoch": 0.26095, + "grad_norm": 1.2569375486731982, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 26095 + }, + { + "epoch": 0.26096, + "grad_norm": 1.6203264525752543, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 26096 + }, + { + "epoch": 0.26097, + "grad_norm": 1.1712889735812024, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26097 + }, + { + "epoch": 0.26098, + "grad_norm": 1.4452967529793423, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26098 + }, + { + "epoch": 0.26099, + "grad_norm": 1.3439918476548127, + "learning_rate": 0.003, + "loss": 4.057, + "step": 26099 + }, + { + "epoch": 0.261, + "grad_norm": 1.4196043060398031, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 26100 + }, + { + "epoch": 0.26101, + "grad_norm": 1.132301369225921, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 26101 + }, + { + "epoch": 0.26102, + "grad_norm": 1.420314610063409, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26102 + }, + { + "epoch": 0.26103, + "grad_norm": 1.4900084238655948, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 26103 + }, + { + "epoch": 0.26104, + "grad_norm": 1.5644797925687497, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 26104 + }, + { + "epoch": 0.26105, + "grad_norm": 1.21011200393285, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 26105 + }, + { + "epoch": 0.26106, + "grad_norm": 1.5729983104308554, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26106 + }, + { + "epoch": 0.26107, + "grad_norm": 1.4182777580338495, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 26107 + }, + { + "epoch": 0.26108, + "grad_norm": 1.6761744297836594, + "learning_rate": 0.003, + "loss": 4.032, + "step": 26108 + }, + { + "epoch": 0.26109, + "grad_norm": 1.2999454462146738, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26109 + }, + { + "epoch": 0.2611, + "grad_norm": 1.718531044055941, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 26110 + }, + { + "epoch": 0.26111, + "grad_norm": 1.0557692032890076, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 26111 + }, + { + "epoch": 0.26112, + "grad_norm": 1.4509710765078558, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 26112 + }, + { + "epoch": 0.26113, + "grad_norm": 1.2929556973991132, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 26113 + }, + { + "epoch": 0.26114, + "grad_norm": 1.613834605259719, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26114 + }, + { + "epoch": 0.26115, + "grad_norm": 1.605778920199912, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 26115 + }, + { + "epoch": 0.26116, + "grad_norm": 1.6291330716330643, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26116 + }, + { + "epoch": 0.26117, + "grad_norm": 1.3018931004143781, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 26117 + }, + { + "epoch": 0.26118, + "grad_norm": 1.378576274603408, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26118 + }, + { + "epoch": 0.26119, + "grad_norm": 1.2897866159616231, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 26119 + }, + { + "epoch": 0.2612, + "grad_norm": 1.5726795541953154, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26120 + }, + { + "epoch": 0.26121, + "grad_norm": 1.1841061233920367, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26121 + }, + { + "epoch": 0.26122, + "grad_norm": 1.6248303048286197, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 26122 + }, + { + "epoch": 0.26123, + "grad_norm": 1.2611826962440553, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 26123 + }, + { + "epoch": 0.26124, + "grad_norm": 1.3165503712721986, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 26124 + }, + { + "epoch": 0.26125, + "grad_norm": 1.3352395005418352, + "learning_rate": 0.003, + "loss": 4.06, + "step": 26125 + }, + { + "epoch": 0.26126, + "grad_norm": 1.3588871224573988, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 26126 + }, + { + "epoch": 0.26127, + "grad_norm": 1.2687247275123623, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26127 + }, + { + "epoch": 0.26128, + "grad_norm": 1.3342028538639636, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26128 + }, + { + "epoch": 0.26129, + "grad_norm": 1.6600813844693358, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 26129 + }, + { + "epoch": 0.2613, + "grad_norm": 1.562162027433235, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 26130 + }, + { + "epoch": 0.26131, + "grad_norm": 1.7165530678900138, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 26131 + }, + { + "epoch": 0.26132, + "grad_norm": 1.2377635845969857, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26132 + }, + { + "epoch": 0.26133, + "grad_norm": 1.4506144001930945, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26133 + }, + { + "epoch": 0.26134, + "grad_norm": 1.2414406753248592, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 26134 + }, + { + "epoch": 0.26135, + "grad_norm": 1.6487287546806624, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26135 + }, + { + "epoch": 0.26136, + "grad_norm": 1.40980531444157, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26136 + }, + { + "epoch": 0.26137, + "grad_norm": 1.6771782258636159, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26137 + }, + { + "epoch": 0.26138, + "grad_norm": 1.0209183995295215, + "learning_rate": 0.003, + "loss": 4.068, + "step": 26138 + }, + { + "epoch": 0.26139, + "grad_norm": 1.5714213742001164, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 26139 + }, + { + "epoch": 0.2614, + "grad_norm": 1.3640034611247536, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 26140 + }, + { + "epoch": 0.26141, + "grad_norm": 1.7439780190307175, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 26141 + }, + { + "epoch": 0.26142, + "grad_norm": 1.1002244994552863, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 26142 + }, + { + "epoch": 0.26143, + "grad_norm": 1.5445100061547528, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 26143 + }, + { + "epoch": 0.26144, + "grad_norm": 1.340549551361928, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26144 + }, + { + "epoch": 0.26145, + "grad_norm": 1.7559105408691897, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 26145 + }, + { + "epoch": 0.26146, + "grad_norm": 1.5417134728901325, + "learning_rate": 0.003, + "loss": 4.066, + "step": 26146 + }, + { + "epoch": 0.26147, + "grad_norm": 1.1164700429186558, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26147 + }, + { + "epoch": 0.26148, + "grad_norm": 1.8238427013694711, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 26148 + }, + { + "epoch": 0.26149, + "grad_norm": 1.2023986489271605, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 26149 + }, + { + "epoch": 0.2615, + "grad_norm": 1.6193026800592802, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26150 + }, + { + "epoch": 0.26151, + "grad_norm": 1.215661480012003, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 26151 + }, + { + "epoch": 0.26152, + "grad_norm": 1.236927177532983, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26152 + }, + { + "epoch": 0.26153, + "grad_norm": 1.2642404562366423, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 26153 + }, + { + "epoch": 0.26154, + "grad_norm": 1.3474149875884414, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 26154 + }, + { + "epoch": 0.26155, + "grad_norm": 1.4057804630329396, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 26155 + }, + { + "epoch": 0.26156, + "grad_norm": 1.7384617155217947, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 26156 + }, + { + "epoch": 0.26157, + "grad_norm": 1.1690457614701564, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26157 + }, + { + "epoch": 0.26158, + "grad_norm": 1.4545145134844872, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 26158 + }, + { + "epoch": 0.26159, + "grad_norm": 1.3912544867201813, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 26159 + }, + { + "epoch": 0.2616, + "grad_norm": 1.4182157160391877, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 26160 + }, + { + "epoch": 0.26161, + "grad_norm": 1.2535501834765355, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26161 + }, + { + "epoch": 0.26162, + "grad_norm": 1.3663070848440444, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 26162 + }, + { + "epoch": 0.26163, + "grad_norm": 1.411853630704683, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26163 + }, + { + "epoch": 0.26164, + "grad_norm": 1.5003938456176518, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26164 + }, + { + "epoch": 0.26165, + "grad_norm": 1.7363738718673913, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26165 + }, + { + "epoch": 0.26166, + "grad_norm": 1.158663450338267, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 26166 + }, + { + "epoch": 0.26167, + "grad_norm": 1.5926689818851438, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26167 + }, + { + "epoch": 0.26168, + "grad_norm": 1.474893731592287, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 26168 + }, + { + "epoch": 0.26169, + "grad_norm": 1.1851905784448797, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 26169 + }, + { + "epoch": 0.2617, + "grad_norm": 1.4296578520326382, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 26170 + }, + { + "epoch": 0.26171, + "grad_norm": 1.3784317760204978, + "learning_rate": 0.003, + "loss": 4.038, + "step": 26171 + }, + { + "epoch": 0.26172, + "grad_norm": 1.3585245642319488, + "learning_rate": 0.003, + "loss": 4.079, + "step": 26172 + }, + { + "epoch": 0.26173, + "grad_norm": 1.3469048046676608, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26173 + }, + { + "epoch": 0.26174, + "grad_norm": 1.3757326570783457, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 26174 + }, + { + "epoch": 0.26175, + "grad_norm": 1.2809953459872485, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26175 + }, + { + "epoch": 0.26176, + "grad_norm": 1.4820119328543484, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26176 + }, + { + "epoch": 0.26177, + "grad_norm": 1.2176131165152875, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 26177 + }, + { + "epoch": 0.26178, + "grad_norm": 1.5740547046925897, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26178 + }, + { + "epoch": 0.26179, + "grad_norm": 1.687256505072092, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 26179 + }, + { + "epoch": 0.2618, + "grad_norm": 1.0744011624875078, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26180 + }, + { + "epoch": 0.26181, + "grad_norm": 1.5819509337109658, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 26181 + }, + { + "epoch": 0.26182, + "grad_norm": 1.441596399458077, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 26182 + }, + { + "epoch": 0.26183, + "grad_norm": 1.49076580291498, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26183 + }, + { + "epoch": 0.26184, + "grad_norm": 1.2531679192189777, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26184 + }, + { + "epoch": 0.26185, + "grad_norm": 1.668708512916372, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 26185 + }, + { + "epoch": 0.26186, + "grad_norm": 1.3429557786750703, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26186 + }, + { + "epoch": 0.26187, + "grad_norm": 1.3410163224181295, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 26187 + }, + { + "epoch": 0.26188, + "grad_norm": 1.5167209664874686, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 26188 + }, + { + "epoch": 0.26189, + "grad_norm": 1.3196959086096505, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26189 + }, + { + "epoch": 0.2619, + "grad_norm": 1.377865847049796, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26190 + }, + { + "epoch": 0.26191, + "grad_norm": 1.3606484364838556, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 26191 + }, + { + "epoch": 0.26192, + "grad_norm": 1.405228152947912, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26192 + }, + { + "epoch": 0.26193, + "grad_norm": 1.4958682252343638, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26193 + }, + { + "epoch": 0.26194, + "grad_norm": 1.5464522694505232, + "learning_rate": 0.003, + "loss": 4.052, + "step": 26194 + }, + { + "epoch": 0.26195, + "grad_norm": 1.2999542839068643, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 26195 + }, + { + "epoch": 0.26196, + "grad_norm": 1.6439013389799124, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 26196 + }, + { + "epoch": 0.26197, + "grad_norm": 1.3845269995013796, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26197 + }, + { + "epoch": 0.26198, + "grad_norm": 1.7605915041581846, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 26198 + }, + { + "epoch": 0.26199, + "grad_norm": 1.1733014698747701, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 26199 + }, + { + "epoch": 0.262, + "grad_norm": 1.346721988523116, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 26200 + }, + { + "epoch": 0.26201, + "grad_norm": 1.4203141148771923, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 26201 + }, + { + "epoch": 0.26202, + "grad_norm": 1.3944407996055148, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 26202 + }, + { + "epoch": 0.26203, + "grad_norm": 1.3333047484787508, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 26203 + }, + { + "epoch": 0.26204, + "grad_norm": 1.1671716223923356, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 26204 + }, + { + "epoch": 0.26205, + "grad_norm": 1.7269720964430693, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 26205 + }, + { + "epoch": 0.26206, + "grad_norm": 1.111604609749631, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 26206 + }, + { + "epoch": 0.26207, + "grad_norm": 1.861499187817077, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 26207 + }, + { + "epoch": 0.26208, + "grad_norm": 1.5604505810121505, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 26208 + }, + { + "epoch": 0.26209, + "grad_norm": 1.2957786835868728, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 26209 + }, + { + "epoch": 0.2621, + "grad_norm": 1.5926041459461269, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 26210 + }, + { + "epoch": 0.26211, + "grad_norm": 1.2057306756570318, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 26211 + }, + { + "epoch": 0.26212, + "grad_norm": 1.6854075225083975, + "learning_rate": 0.003, + "loss": 4.092, + "step": 26212 + }, + { + "epoch": 0.26213, + "grad_norm": 1.1756623323461275, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 26213 + }, + { + "epoch": 0.26214, + "grad_norm": 1.2793257040300614, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 26214 + }, + { + "epoch": 0.26215, + "grad_norm": 1.5137041084884015, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 26215 + }, + { + "epoch": 0.26216, + "grad_norm": 1.3735024908467404, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26216 + }, + { + "epoch": 0.26217, + "grad_norm": 1.3518635043812337, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 26217 + }, + { + "epoch": 0.26218, + "grad_norm": 1.2460345054016173, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 26218 + }, + { + "epoch": 0.26219, + "grad_norm": 1.7674396601832787, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 26219 + }, + { + "epoch": 0.2622, + "grad_norm": 0.9790477903863751, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26220 + }, + { + "epoch": 0.26221, + "grad_norm": 1.5248865202414532, + "learning_rate": 0.003, + "loss": 4.072, + "step": 26221 + }, + { + "epoch": 0.26222, + "grad_norm": 1.4469930100574064, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 26222 + }, + { + "epoch": 0.26223, + "grad_norm": 1.3838414761253588, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26223 + }, + { + "epoch": 0.26224, + "grad_norm": 1.3251044917956059, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26224 + }, + { + "epoch": 0.26225, + "grad_norm": 1.2023077488595437, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 26225 + }, + { + "epoch": 0.26226, + "grad_norm": 1.3199495580280958, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 26226 + }, + { + "epoch": 0.26227, + "grad_norm": 1.3656447357660833, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 26227 + }, + { + "epoch": 0.26228, + "grad_norm": 1.4365917311393643, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 26228 + }, + { + "epoch": 0.26229, + "grad_norm": 1.3183694841280178, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 26229 + }, + { + "epoch": 0.2623, + "grad_norm": 1.6021764597492052, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 26230 + }, + { + "epoch": 0.26231, + "grad_norm": 1.303049980529435, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 26231 + }, + { + "epoch": 0.26232, + "grad_norm": 1.5870852828044721, + "learning_rate": 0.003, + "loss": 4.014, + "step": 26232 + }, + { + "epoch": 0.26233, + "grad_norm": 1.6571121715598225, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 26233 + }, + { + "epoch": 0.26234, + "grad_norm": 1.6019282407384785, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 26234 + }, + { + "epoch": 0.26235, + "grad_norm": 1.3826571284802207, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 26235 + }, + { + "epoch": 0.26236, + "grad_norm": 1.4213060779168911, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 26236 + }, + { + "epoch": 0.26237, + "grad_norm": 1.3390722294705972, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 26237 + }, + { + "epoch": 0.26238, + "grad_norm": 1.2496220029832934, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 26238 + }, + { + "epoch": 0.26239, + "grad_norm": 1.243823615657916, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 26239 + }, + { + "epoch": 0.2624, + "grad_norm": 1.41881863647092, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 26240 + }, + { + "epoch": 0.26241, + "grad_norm": 1.161735037995058, + "learning_rate": 0.003, + "loss": 4.042, + "step": 26241 + }, + { + "epoch": 0.26242, + "grad_norm": 1.9727697722000974, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 26242 + }, + { + "epoch": 0.26243, + "grad_norm": 4.98863487457687, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 26243 + }, + { + "epoch": 0.26244, + "grad_norm": 3.251471943297102, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 26244 + }, + { + "epoch": 0.26245, + "grad_norm": 1.7071262449606392, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 26245 + }, + { + "epoch": 0.26246, + "grad_norm": 1.9077218399369058, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 26246 + }, + { + "epoch": 0.26247, + "grad_norm": 1.5430601254647671, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 26247 + }, + { + "epoch": 0.26248, + "grad_norm": 2.0475358081580626, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 26248 + }, + { + "epoch": 0.26249, + "grad_norm": 1.8647316427059468, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 26249 + }, + { + "epoch": 0.2625, + "grad_norm": 1.5606108245247248, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 26250 + }, + { + "epoch": 0.26251, + "grad_norm": 2.7683194641326607, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 26251 + }, + { + "epoch": 0.26252, + "grad_norm": 1.9640777024009546, + "learning_rate": 0.003, + "loss": 4.1391, + "step": 26252 + }, + { + "epoch": 0.26253, + "grad_norm": 1.9297617487010124, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 26253 + }, + { + "epoch": 0.26254, + "grad_norm": 1.3772224349477864, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 26254 + }, + { + "epoch": 0.26255, + "grad_norm": 1.5040160414181971, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 26255 + }, + { + "epoch": 0.26256, + "grad_norm": 2.4282021395209235, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 26256 + }, + { + "epoch": 0.26257, + "grad_norm": 3.8147420217243426, + "learning_rate": 0.003, + "loss": 4.1677, + "step": 26257 + }, + { + "epoch": 0.26258, + "grad_norm": 2.1150950127805617, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 26258 + }, + { + "epoch": 0.26259, + "grad_norm": 1.576712999581666, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 26259 + }, + { + "epoch": 0.2626, + "grad_norm": 1.9658788091234465, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 26260 + }, + { + "epoch": 0.26261, + "grad_norm": 2.6000816822962216, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 26261 + }, + { + "epoch": 0.26262, + "grad_norm": 3.5717161979199754, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 26262 + }, + { + "epoch": 0.26263, + "grad_norm": 2.154212411936737, + "learning_rate": 0.003, + "loss": 4.124, + "step": 26263 + }, + { + "epoch": 0.26264, + "grad_norm": 2.057221801449474, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 26264 + }, + { + "epoch": 0.26265, + "grad_norm": 1.9590177762184788, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 26265 + }, + { + "epoch": 0.26266, + "grad_norm": 2.529267936504621, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 26266 + }, + { + "epoch": 0.26267, + "grad_norm": 1.5830102764097558, + "learning_rate": 0.003, + "loss": 4.096, + "step": 26267 + }, + { + "epoch": 0.26268, + "grad_norm": 1.6063673986410207, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 26268 + }, + { + "epoch": 0.26269, + "grad_norm": 1.5947727640830085, + "learning_rate": 0.003, + "loss": 4.092, + "step": 26269 + }, + { + "epoch": 0.2627, + "grad_norm": 1.5596608644181722, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 26270 + }, + { + "epoch": 0.26271, + "grad_norm": 1.563423437907866, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 26271 + }, + { + "epoch": 0.26272, + "grad_norm": 1.6304174960222102, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 26272 + }, + { + "epoch": 0.26273, + "grad_norm": 1.3313782409787416, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 26273 + }, + { + "epoch": 0.26274, + "grad_norm": 1.6272100830673748, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26274 + }, + { + "epoch": 0.26275, + "grad_norm": 1.46571347703826, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26275 + }, + { + "epoch": 0.26276, + "grad_norm": 1.8643886043659925, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 26276 + }, + { + "epoch": 0.26277, + "grad_norm": 1.3516215011222086, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 26277 + }, + { + "epoch": 0.26278, + "grad_norm": 1.9922299546329671, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 26278 + }, + { + "epoch": 0.26279, + "grad_norm": 1.4856090978306589, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 26279 + }, + { + "epoch": 0.2628, + "grad_norm": 2.075256061819583, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 26280 + }, + { + "epoch": 0.26281, + "grad_norm": 1.4539297692246258, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 26281 + }, + { + "epoch": 0.26282, + "grad_norm": 1.236689712675275, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 26282 + }, + { + "epoch": 0.26283, + "grad_norm": 1.413164368093047, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 26283 + }, + { + "epoch": 0.26284, + "grad_norm": 1.5300251307022683, + "learning_rate": 0.003, + "loss": 4.081, + "step": 26284 + }, + { + "epoch": 0.26285, + "grad_norm": 1.0770764987890273, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 26285 + }, + { + "epoch": 0.26286, + "grad_norm": 1.2877089638518247, + "learning_rate": 0.003, + "loss": 4.076, + "step": 26286 + }, + { + "epoch": 0.26287, + "grad_norm": 1.4989123798834958, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 26287 + }, + { + "epoch": 0.26288, + "grad_norm": 1.474425776527829, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 26288 + }, + { + "epoch": 0.26289, + "grad_norm": 1.6779276049297838, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 26289 + }, + { + "epoch": 0.2629, + "grad_norm": 1.4239586529377959, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 26290 + }, + { + "epoch": 0.26291, + "grad_norm": 1.6547216862380802, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 26291 + }, + { + "epoch": 0.26292, + "grad_norm": 1.4712033582256399, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26292 + }, + { + "epoch": 0.26293, + "grad_norm": 1.6583959360904992, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 26293 + }, + { + "epoch": 0.26294, + "grad_norm": 1.1894077095129167, + "learning_rate": 0.003, + "loss": 4.064, + "step": 26294 + }, + { + "epoch": 0.26295, + "grad_norm": 1.53987053821478, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26295 + }, + { + "epoch": 0.26296, + "grad_norm": 1.2229334096956905, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26296 + }, + { + "epoch": 0.26297, + "grad_norm": 1.7858235228041421, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 26297 + }, + { + "epoch": 0.26298, + "grad_norm": 1.1967295089219727, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26298 + }, + { + "epoch": 0.26299, + "grad_norm": 1.2983337113726472, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26299 + }, + { + "epoch": 0.263, + "grad_norm": 1.1784023706904119, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 26300 + }, + { + "epoch": 0.26301, + "grad_norm": 1.8676431207594637, + "learning_rate": 0.003, + "loss": 4.075, + "step": 26301 + }, + { + "epoch": 0.26302, + "grad_norm": 1.4506490595672212, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26302 + }, + { + "epoch": 0.26303, + "grad_norm": 1.5334210978102238, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 26303 + }, + { + "epoch": 0.26304, + "grad_norm": 1.4304494860277484, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 26304 + }, + { + "epoch": 0.26305, + "grad_norm": 1.1803113671860817, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 26305 + }, + { + "epoch": 0.26306, + "grad_norm": 1.5303164308285027, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26306 + }, + { + "epoch": 0.26307, + "grad_norm": 1.4296904003743498, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 26307 + }, + { + "epoch": 0.26308, + "grad_norm": 1.2107264555341897, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 26308 + }, + { + "epoch": 0.26309, + "grad_norm": 1.365111318685289, + "learning_rate": 0.003, + "loss": 4.045, + "step": 26309 + }, + { + "epoch": 0.2631, + "grad_norm": 1.2506000752637285, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 26310 + }, + { + "epoch": 0.26311, + "grad_norm": 1.5818385434558915, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 26311 + }, + { + "epoch": 0.26312, + "grad_norm": 1.5131035970509348, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 26312 + }, + { + "epoch": 0.26313, + "grad_norm": 1.8527869556863024, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 26313 + }, + { + "epoch": 0.26314, + "grad_norm": 1.1463593049955008, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 26314 + }, + { + "epoch": 0.26315, + "grad_norm": 1.590462079727231, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26315 + }, + { + "epoch": 0.26316, + "grad_norm": 1.3566859842665382, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 26316 + }, + { + "epoch": 0.26317, + "grad_norm": 1.486157132281985, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26317 + }, + { + "epoch": 0.26318, + "grad_norm": 1.304147564432363, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26318 + }, + { + "epoch": 0.26319, + "grad_norm": 1.4300415496651901, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 26319 + }, + { + "epoch": 0.2632, + "grad_norm": 1.140569885479183, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 26320 + }, + { + "epoch": 0.26321, + "grad_norm": 1.3121704201302227, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 26321 + }, + { + "epoch": 0.26322, + "grad_norm": 1.2965961033998787, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 26322 + }, + { + "epoch": 0.26323, + "grad_norm": 1.2394802900252793, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26323 + }, + { + "epoch": 0.26324, + "grad_norm": 1.2830896831295615, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 26324 + }, + { + "epoch": 0.26325, + "grad_norm": 1.3270305122734911, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26325 + }, + { + "epoch": 0.26326, + "grad_norm": 1.2914152322632022, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 26326 + }, + { + "epoch": 0.26327, + "grad_norm": 1.5442194716462705, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 26327 + }, + { + "epoch": 0.26328, + "grad_norm": 1.956039927351964, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 26328 + }, + { + "epoch": 0.26329, + "grad_norm": 1.465317600199115, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 26329 + }, + { + "epoch": 0.2633, + "grad_norm": 1.3988899584302488, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 26330 + }, + { + "epoch": 0.26331, + "grad_norm": 1.4126495003092239, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 26331 + }, + { + "epoch": 0.26332, + "grad_norm": 1.2772306349590197, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 26332 + }, + { + "epoch": 0.26333, + "grad_norm": 1.5023353436137612, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 26333 + }, + { + "epoch": 0.26334, + "grad_norm": 1.4017453876932504, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 26334 + }, + { + "epoch": 0.26335, + "grad_norm": 1.426317232801471, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 26335 + }, + { + "epoch": 0.26336, + "grad_norm": 1.2473925554276797, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26336 + }, + { + "epoch": 0.26337, + "grad_norm": 1.467251218149138, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 26337 + }, + { + "epoch": 0.26338, + "grad_norm": 1.328384975214918, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 26338 + }, + { + "epoch": 0.26339, + "grad_norm": 1.5565179976456558, + "learning_rate": 0.003, + "loss": 4.06, + "step": 26339 + }, + { + "epoch": 0.2634, + "grad_norm": 1.4853992604233175, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 26340 + }, + { + "epoch": 0.26341, + "grad_norm": 1.3223745247547076, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 26341 + }, + { + "epoch": 0.26342, + "grad_norm": 1.412250186481602, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26342 + }, + { + "epoch": 0.26343, + "grad_norm": 1.3060397517047007, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 26343 + }, + { + "epoch": 0.26344, + "grad_norm": 1.4784935420710108, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 26344 + }, + { + "epoch": 0.26345, + "grad_norm": 1.3642391328937271, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 26345 + }, + { + "epoch": 0.26346, + "grad_norm": 1.4984377860805318, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26346 + }, + { + "epoch": 0.26347, + "grad_norm": 1.2630133987081738, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 26347 + }, + { + "epoch": 0.26348, + "grad_norm": 1.3109751876813567, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 26348 + }, + { + "epoch": 0.26349, + "grad_norm": 1.400908828247379, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 26349 + }, + { + "epoch": 0.2635, + "grad_norm": 1.399647655859998, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 26350 + }, + { + "epoch": 0.26351, + "grad_norm": 1.550865086666066, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 26351 + }, + { + "epoch": 0.26352, + "grad_norm": 1.1559958358273263, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 26352 + }, + { + "epoch": 0.26353, + "grad_norm": 1.5093607039876318, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 26353 + }, + { + "epoch": 0.26354, + "grad_norm": 1.1685137529916978, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 26354 + }, + { + "epoch": 0.26355, + "grad_norm": 1.4577107063289916, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26355 + }, + { + "epoch": 0.26356, + "grad_norm": 1.1431745543202865, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26356 + }, + { + "epoch": 0.26357, + "grad_norm": 1.367827798153738, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 26357 + }, + { + "epoch": 0.26358, + "grad_norm": 1.3852091026806508, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 26358 + }, + { + "epoch": 0.26359, + "grad_norm": 1.3762897844325424, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26359 + }, + { + "epoch": 0.2636, + "grad_norm": 1.567807433226251, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 26360 + }, + { + "epoch": 0.26361, + "grad_norm": 1.5248486565494674, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 26361 + }, + { + "epoch": 0.26362, + "grad_norm": 1.6830583612412935, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 26362 + }, + { + "epoch": 0.26363, + "grad_norm": 1.2883107764849413, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 26363 + }, + { + "epoch": 0.26364, + "grad_norm": 1.3805501409822747, + "learning_rate": 0.003, + "loss": 4.028, + "step": 26364 + }, + { + "epoch": 0.26365, + "grad_norm": 1.2167454432571607, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 26365 + }, + { + "epoch": 0.26366, + "grad_norm": 1.3028802014727991, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 26366 + }, + { + "epoch": 0.26367, + "grad_norm": 1.5556408507593034, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 26367 + }, + { + "epoch": 0.26368, + "grad_norm": 1.4148182108775957, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26368 + }, + { + "epoch": 0.26369, + "grad_norm": 1.2896025085089171, + "learning_rate": 0.003, + "loss": 4.05, + "step": 26369 + }, + { + "epoch": 0.2637, + "grad_norm": 1.3061914874614393, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 26370 + }, + { + "epoch": 0.26371, + "grad_norm": 1.495167312666796, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 26371 + }, + { + "epoch": 0.26372, + "grad_norm": 1.207990081391306, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 26372 + }, + { + "epoch": 0.26373, + "grad_norm": 1.3785725955595374, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 26373 + }, + { + "epoch": 0.26374, + "grad_norm": 1.806726257693243, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 26374 + }, + { + "epoch": 0.26375, + "grad_norm": 1.0446899521878341, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 26375 + }, + { + "epoch": 0.26376, + "grad_norm": 1.8644536967634364, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 26376 + }, + { + "epoch": 0.26377, + "grad_norm": 1.0931435758522796, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26377 + }, + { + "epoch": 0.26378, + "grad_norm": 1.6829561761030838, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 26378 + }, + { + "epoch": 0.26379, + "grad_norm": 1.3699488619175868, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26379 + }, + { + "epoch": 0.2638, + "grad_norm": 1.2179350306689694, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 26380 + }, + { + "epoch": 0.26381, + "grad_norm": 1.6566521311099491, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 26381 + }, + { + "epoch": 0.26382, + "grad_norm": 1.1567618333370784, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 26382 + }, + { + "epoch": 0.26383, + "grad_norm": 1.5981558448837376, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 26383 + }, + { + "epoch": 0.26384, + "grad_norm": 1.604476184530225, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 26384 + }, + { + "epoch": 0.26385, + "grad_norm": 1.334581572166795, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 26385 + }, + { + "epoch": 0.26386, + "grad_norm": 1.3334517758491247, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 26386 + }, + { + "epoch": 0.26387, + "grad_norm": 1.3324830741844196, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 26387 + }, + { + "epoch": 0.26388, + "grad_norm": 1.0880497546414944, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 26388 + }, + { + "epoch": 0.26389, + "grad_norm": 1.3734744715691038, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26389 + }, + { + "epoch": 0.2639, + "grad_norm": 1.2666390133448293, + "learning_rate": 0.003, + "loss": 4.033, + "step": 26390 + }, + { + "epoch": 0.26391, + "grad_norm": 1.3727676204388488, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 26391 + }, + { + "epoch": 0.26392, + "grad_norm": 1.276203194953432, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 26392 + }, + { + "epoch": 0.26393, + "grad_norm": 1.2791272182907754, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 26393 + }, + { + "epoch": 0.26394, + "grad_norm": 1.2850800333764225, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 26394 + }, + { + "epoch": 0.26395, + "grad_norm": 1.8105952698254157, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 26395 + }, + { + "epoch": 0.26396, + "grad_norm": 1.480316767480441, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 26396 + }, + { + "epoch": 0.26397, + "grad_norm": 1.3671369416635804, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 26397 + }, + { + "epoch": 0.26398, + "grad_norm": 1.2840192928075207, + "learning_rate": 0.003, + "loss": 4.073, + "step": 26398 + }, + { + "epoch": 0.26399, + "grad_norm": 1.5829432650648492, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 26399 + }, + { + "epoch": 0.264, + "grad_norm": 1.0750432882611756, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 26400 + }, + { + "epoch": 0.26401, + "grad_norm": 1.8531124545281596, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 26401 + }, + { + "epoch": 0.26402, + "grad_norm": 1.3027982087025305, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 26402 + }, + { + "epoch": 0.26403, + "grad_norm": 1.5019723128560831, + "learning_rate": 0.003, + "loss": 4.074, + "step": 26403 + }, + { + "epoch": 0.26404, + "grad_norm": 1.4187341683887447, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 26404 + }, + { + "epoch": 0.26405, + "grad_norm": 1.300151528760964, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 26405 + }, + { + "epoch": 0.26406, + "grad_norm": 1.2134019759794605, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 26406 + }, + { + "epoch": 0.26407, + "grad_norm": 1.2811129148554012, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26407 + }, + { + "epoch": 0.26408, + "grad_norm": 1.472048504376434, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 26408 + }, + { + "epoch": 0.26409, + "grad_norm": 1.410996990111969, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26409 + }, + { + "epoch": 0.2641, + "grad_norm": 1.3836702134775527, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26410 + }, + { + "epoch": 0.26411, + "grad_norm": 1.420598617391154, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 26411 + }, + { + "epoch": 0.26412, + "grad_norm": 1.417083080615042, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26412 + }, + { + "epoch": 0.26413, + "grad_norm": 1.3684454790980072, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 26413 + }, + { + "epoch": 0.26414, + "grad_norm": 1.130327684798827, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 26414 + }, + { + "epoch": 0.26415, + "grad_norm": 1.2804934484373638, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26415 + }, + { + "epoch": 0.26416, + "grad_norm": 1.3770833405862966, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 26416 + }, + { + "epoch": 0.26417, + "grad_norm": 1.2816660437187601, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 26417 + }, + { + "epoch": 0.26418, + "grad_norm": 1.2056075922969773, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26418 + }, + { + "epoch": 0.26419, + "grad_norm": 1.5624193286866106, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 26419 + }, + { + "epoch": 0.2642, + "grad_norm": 1.5233755497261448, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 26420 + }, + { + "epoch": 0.26421, + "grad_norm": 1.7351626636330122, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 26421 + }, + { + "epoch": 0.26422, + "grad_norm": 1.2114583890891013, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 26422 + }, + { + "epoch": 0.26423, + "grad_norm": 1.3177857269011544, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 26423 + }, + { + "epoch": 0.26424, + "grad_norm": 1.5853493929426459, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 26424 + }, + { + "epoch": 0.26425, + "grad_norm": 1.2346126734201186, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 26425 + }, + { + "epoch": 0.26426, + "grad_norm": 1.548257650459047, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26426 + }, + { + "epoch": 0.26427, + "grad_norm": 1.259965821401596, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 26427 + }, + { + "epoch": 0.26428, + "grad_norm": 1.6811854476240204, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 26428 + }, + { + "epoch": 0.26429, + "grad_norm": 1.3200001140836572, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 26429 + }, + { + "epoch": 0.2643, + "grad_norm": 1.5700332405895359, + "learning_rate": 0.003, + "loss": 4.037, + "step": 26430 + }, + { + "epoch": 0.26431, + "grad_norm": 1.3092737344451022, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 26431 + }, + { + "epoch": 0.26432, + "grad_norm": 1.3372602935597275, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 26432 + }, + { + "epoch": 0.26433, + "grad_norm": 1.3908836840833942, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 26433 + }, + { + "epoch": 0.26434, + "grad_norm": 1.2952435869142087, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 26434 + }, + { + "epoch": 0.26435, + "grad_norm": 1.7677379196137475, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 26435 + }, + { + "epoch": 0.26436, + "grad_norm": 1.2100101753204981, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 26436 + }, + { + "epoch": 0.26437, + "grad_norm": 1.6242274999250743, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 26437 + }, + { + "epoch": 0.26438, + "grad_norm": 1.279728552536087, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26438 + }, + { + "epoch": 0.26439, + "grad_norm": 1.3981034679957967, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 26439 + }, + { + "epoch": 0.2644, + "grad_norm": 1.310247523513397, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 26440 + }, + { + "epoch": 0.26441, + "grad_norm": 1.5749812269295789, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 26441 + }, + { + "epoch": 0.26442, + "grad_norm": 1.3498359320661142, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 26442 + }, + { + "epoch": 0.26443, + "grad_norm": 1.3058893640900553, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 26443 + }, + { + "epoch": 0.26444, + "grad_norm": 1.3007918582885691, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 26444 + }, + { + "epoch": 0.26445, + "grad_norm": 1.2985127133093222, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 26445 + }, + { + "epoch": 0.26446, + "grad_norm": 1.3016043634086165, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 26446 + }, + { + "epoch": 0.26447, + "grad_norm": 1.4798207658348388, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 26447 + }, + { + "epoch": 0.26448, + "grad_norm": 1.4437105131433268, + "learning_rate": 0.003, + "loss": 4.037, + "step": 26448 + }, + { + "epoch": 0.26449, + "grad_norm": 1.2713103464406874, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 26449 + }, + { + "epoch": 0.2645, + "grad_norm": 1.3243671576205467, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 26450 + }, + { + "epoch": 0.26451, + "grad_norm": 1.3993392025418867, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26451 + }, + { + "epoch": 0.26452, + "grad_norm": 1.539713533424695, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 26452 + }, + { + "epoch": 0.26453, + "grad_norm": 1.2372215737727703, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 26453 + }, + { + "epoch": 0.26454, + "grad_norm": 1.5848885430305055, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26454 + }, + { + "epoch": 0.26455, + "grad_norm": 1.165973456298834, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 26455 + }, + { + "epoch": 0.26456, + "grad_norm": 1.6135637505366633, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 26456 + }, + { + "epoch": 0.26457, + "grad_norm": 1.4658855618176627, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 26457 + }, + { + "epoch": 0.26458, + "grad_norm": 1.2620077469093867, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26458 + }, + { + "epoch": 0.26459, + "grad_norm": 1.523819221957462, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 26459 + }, + { + "epoch": 0.2646, + "grad_norm": 1.208158236418585, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 26460 + }, + { + "epoch": 0.26461, + "grad_norm": 1.4900442235831142, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 26461 + }, + { + "epoch": 0.26462, + "grad_norm": 1.3993997712925976, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 26462 + }, + { + "epoch": 0.26463, + "grad_norm": 1.6609438992341004, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 26463 + }, + { + "epoch": 0.26464, + "grad_norm": 1.3224511623138693, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26464 + }, + { + "epoch": 0.26465, + "grad_norm": 1.2927206362533432, + "learning_rate": 0.003, + "loss": 4.03, + "step": 26465 + }, + { + "epoch": 0.26466, + "grad_norm": 1.3700770711643673, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 26466 + }, + { + "epoch": 0.26467, + "grad_norm": 1.2880916457764788, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 26467 + }, + { + "epoch": 0.26468, + "grad_norm": 1.4545534860376705, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26468 + }, + { + "epoch": 0.26469, + "grad_norm": 1.5120376217138751, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 26469 + }, + { + "epoch": 0.2647, + "grad_norm": 1.1038769147128789, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 26470 + }, + { + "epoch": 0.26471, + "grad_norm": 1.4796941380167772, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 26471 + }, + { + "epoch": 0.26472, + "grad_norm": 1.183997559550462, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 26472 + }, + { + "epoch": 0.26473, + "grad_norm": 1.358993415092744, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 26473 + }, + { + "epoch": 0.26474, + "grad_norm": 1.3016087278651214, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 26474 + }, + { + "epoch": 0.26475, + "grad_norm": 1.6259919265246756, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 26475 + }, + { + "epoch": 0.26476, + "grad_norm": 1.5413674671496929, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26476 + }, + { + "epoch": 0.26477, + "grad_norm": 1.845984113941337, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 26477 + }, + { + "epoch": 0.26478, + "grad_norm": 1.184571792706768, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 26478 + }, + { + "epoch": 0.26479, + "grad_norm": 1.31686858705688, + "learning_rate": 0.003, + "loss": 4.085, + "step": 26479 + }, + { + "epoch": 0.2648, + "grad_norm": 1.3770774225533302, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 26480 + }, + { + "epoch": 0.26481, + "grad_norm": 1.502360149218931, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26481 + }, + { + "epoch": 0.26482, + "grad_norm": 1.2808825701861903, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 26482 + }, + { + "epoch": 0.26483, + "grad_norm": 1.6882235665187937, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 26483 + }, + { + "epoch": 0.26484, + "grad_norm": 1.2528135863600487, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 26484 + }, + { + "epoch": 0.26485, + "grad_norm": 1.332883548509537, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26485 + }, + { + "epoch": 0.26486, + "grad_norm": 1.538959742028728, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 26486 + }, + { + "epoch": 0.26487, + "grad_norm": 1.2458311168430272, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 26487 + }, + { + "epoch": 0.26488, + "grad_norm": 1.4805679246259875, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 26488 + }, + { + "epoch": 0.26489, + "grad_norm": 1.5474048289640714, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 26489 + }, + { + "epoch": 0.2649, + "grad_norm": 1.2904174529454082, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 26490 + }, + { + "epoch": 0.26491, + "grad_norm": 1.4151633648518986, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 26491 + }, + { + "epoch": 0.26492, + "grad_norm": 1.444964938835607, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 26492 + }, + { + "epoch": 0.26493, + "grad_norm": 1.1900036581065663, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 26493 + }, + { + "epoch": 0.26494, + "grad_norm": 1.389474921292151, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 26494 + }, + { + "epoch": 0.26495, + "grad_norm": 1.4103611013604245, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 26495 + }, + { + "epoch": 0.26496, + "grad_norm": 1.3704181692472104, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 26496 + }, + { + "epoch": 0.26497, + "grad_norm": 1.255833925595971, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26497 + }, + { + "epoch": 0.26498, + "grad_norm": 1.643130125882077, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 26498 + }, + { + "epoch": 0.26499, + "grad_norm": 1.280130536039726, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 26499 + }, + { + "epoch": 0.265, + "grad_norm": 1.4028816572835092, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 26500 + }, + { + "epoch": 0.26501, + "grad_norm": 1.3627672647248674, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26501 + }, + { + "epoch": 0.26502, + "grad_norm": 1.4118577421712515, + "learning_rate": 0.003, + "loss": 4.045, + "step": 26502 + }, + { + "epoch": 0.26503, + "grad_norm": 1.2587746905585997, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 26503 + }, + { + "epoch": 0.26504, + "grad_norm": 1.5435460951680058, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 26504 + }, + { + "epoch": 0.26505, + "grad_norm": 1.458485975087471, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 26505 + }, + { + "epoch": 0.26506, + "grad_norm": 1.297497752864237, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 26506 + }, + { + "epoch": 0.26507, + "grad_norm": 1.2898133059665031, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26507 + }, + { + "epoch": 0.26508, + "grad_norm": 1.468979312987715, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 26508 + }, + { + "epoch": 0.26509, + "grad_norm": 1.317900008982059, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 26509 + }, + { + "epoch": 0.2651, + "grad_norm": 1.5130820063607329, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 26510 + }, + { + "epoch": 0.26511, + "grad_norm": 1.4658540694905673, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 26511 + }, + { + "epoch": 0.26512, + "grad_norm": 1.646551958336409, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26512 + }, + { + "epoch": 0.26513, + "grad_norm": 1.359703460671303, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26513 + }, + { + "epoch": 0.26514, + "grad_norm": 1.428539101317104, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 26514 + }, + { + "epoch": 0.26515, + "grad_norm": 1.2102381252124543, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 26515 + }, + { + "epoch": 0.26516, + "grad_norm": 1.4922656070900626, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26516 + }, + { + "epoch": 0.26517, + "grad_norm": 1.3928942143528114, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 26517 + }, + { + "epoch": 0.26518, + "grad_norm": 1.6028078991367598, + "learning_rate": 0.003, + "loss": 4.026, + "step": 26518 + }, + { + "epoch": 0.26519, + "grad_norm": 1.4472363310644616, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 26519 + }, + { + "epoch": 0.2652, + "grad_norm": 1.3243354349778993, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 26520 + }, + { + "epoch": 0.26521, + "grad_norm": 1.371298986836318, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 26521 + }, + { + "epoch": 0.26522, + "grad_norm": 1.2777362975315294, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 26522 + }, + { + "epoch": 0.26523, + "grad_norm": 1.2938954194099437, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 26523 + }, + { + "epoch": 0.26524, + "grad_norm": 1.255556164836365, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 26524 + }, + { + "epoch": 0.26525, + "grad_norm": 1.5766030160442248, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 26525 + }, + { + "epoch": 0.26526, + "grad_norm": 1.232221906749862, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 26526 + }, + { + "epoch": 0.26527, + "grad_norm": 1.5427576262210476, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 26527 + }, + { + "epoch": 0.26528, + "grad_norm": 1.5743998732587614, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 26528 + }, + { + "epoch": 0.26529, + "grad_norm": 1.6271056791328082, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26529 + }, + { + "epoch": 0.2653, + "grad_norm": 1.6692200606509024, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 26530 + }, + { + "epoch": 0.26531, + "grad_norm": 1.155749670619981, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 26531 + }, + { + "epoch": 0.26532, + "grad_norm": 1.3910719625953152, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 26532 + }, + { + "epoch": 0.26533, + "grad_norm": 1.5099004199612194, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 26533 + }, + { + "epoch": 0.26534, + "grad_norm": 1.3444640613133454, + "learning_rate": 0.003, + "loss": 4.071, + "step": 26534 + }, + { + "epoch": 0.26535, + "grad_norm": 1.3768425823927868, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 26535 + }, + { + "epoch": 0.26536, + "grad_norm": 1.3267072477809978, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 26536 + }, + { + "epoch": 0.26537, + "grad_norm": 1.521282739062617, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 26537 + }, + { + "epoch": 0.26538, + "grad_norm": 1.1897853655415969, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 26538 + }, + { + "epoch": 0.26539, + "grad_norm": 1.557838720894475, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 26539 + }, + { + "epoch": 0.2654, + "grad_norm": 1.146889514112428, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 26540 + }, + { + "epoch": 0.26541, + "grad_norm": 1.4410343698873225, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 26541 + }, + { + "epoch": 0.26542, + "grad_norm": 1.39654914986918, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 26542 + }, + { + "epoch": 0.26543, + "grad_norm": 1.3932150826911989, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 26543 + }, + { + "epoch": 0.26544, + "grad_norm": 1.6912061795003013, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 26544 + }, + { + "epoch": 0.26545, + "grad_norm": 1.4081992268223429, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 26545 + }, + { + "epoch": 0.26546, + "grad_norm": 1.175510899869732, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 26546 + }, + { + "epoch": 0.26547, + "grad_norm": 1.454068205406492, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 26547 + }, + { + "epoch": 0.26548, + "grad_norm": 1.307485097503662, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26548 + }, + { + "epoch": 0.26549, + "grad_norm": 1.1847491907758152, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 26549 + }, + { + "epoch": 0.2655, + "grad_norm": 1.5088034124087886, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26550 + }, + { + "epoch": 0.26551, + "grad_norm": 1.3151469182442708, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 26551 + }, + { + "epoch": 0.26552, + "grad_norm": 1.5739805009042185, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 26552 + }, + { + "epoch": 0.26553, + "grad_norm": 1.3229863723007396, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 26553 + }, + { + "epoch": 0.26554, + "grad_norm": 1.4959674735880553, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 26554 + }, + { + "epoch": 0.26555, + "grad_norm": 1.390439400848731, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26555 + }, + { + "epoch": 0.26556, + "grad_norm": 1.9292370226554332, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 26556 + }, + { + "epoch": 0.26557, + "grad_norm": 1.2143543163773436, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 26557 + }, + { + "epoch": 0.26558, + "grad_norm": 1.885415651757958, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 26558 + }, + { + "epoch": 0.26559, + "grad_norm": 1.1552311732100167, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26559 + }, + { + "epoch": 0.2656, + "grad_norm": 1.2641199019201002, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 26560 + }, + { + "epoch": 0.26561, + "grad_norm": 1.5138472366064266, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26561 + }, + { + "epoch": 0.26562, + "grad_norm": 1.4944035782048763, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 26562 + }, + { + "epoch": 0.26563, + "grad_norm": 1.421606738010415, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 26563 + }, + { + "epoch": 0.26564, + "grad_norm": 1.4680794729488469, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26564 + }, + { + "epoch": 0.26565, + "grad_norm": 1.397061686212988, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 26565 + }, + { + "epoch": 0.26566, + "grad_norm": 1.2195805128749087, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 26566 + }, + { + "epoch": 0.26567, + "grad_norm": 1.657781686892417, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26567 + }, + { + "epoch": 0.26568, + "grad_norm": 1.1384979867307303, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 26568 + }, + { + "epoch": 0.26569, + "grad_norm": 1.7625402120268798, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26569 + }, + { + "epoch": 0.2657, + "grad_norm": 1.2930644876031063, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 26570 + }, + { + "epoch": 0.26571, + "grad_norm": 1.680706487550516, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26571 + }, + { + "epoch": 0.26572, + "grad_norm": 1.2635218272181141, + "learning_rate": 0.003, + "loss": 4.035, + "step": 26572 + }, + { + "epoch": 0.26573, + "grad_norm": 1.752646202960821, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 26573 + }, + { + "epoch": 0.26574, + "grad_norm": 1.3990518587265557, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26574 + }, + { + "epoch": 0.26575, + "grad_norm": 1.2245343803504636, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 26575 + }, + { + "epoch": 0.26576, + "grad_norm": 1.2726789254680226, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 26576 + }, + { + "epoch": 0.26577, + "grad_norm": 1.4214910878006832, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 26577 + }, + { + "epoch": 0.26578, + "grad_norm": 1.4525866843428985, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 26578 + }, + { + "epoch": 0.26579, + "grad_norm": 1.137525275177525, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 26579 + }, + { + "epoch": 0.2658, + "grad_norm": 1.6212406745024965, + "learning_rate": 0.003, + "loss": 4.062, + "step": 26580 + }, + { + "epoch": 0.26581, + "grad_norm": 1.3748697820728475, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 26581 + }, + { + "epoch": 0.26582, + "grad_norm": 1.4351955538430876, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 26582 + }, + { + "epoch": 0.26583, + "grad_norm": 1.4699460741415482, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 26583 + }, + { + "epoch": 0.26584, + "grad_norm": 1.400717890226722, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 26584 + }, + { + "epoch": 0.26585, + "grad_norm": 1.4762936473687691, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 26585 + }, + { + "epoch": 0.26586, + "grad_norm": 1.3017320434736661, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 26586 + }, + { + "epoch": 0.26587, + "grad_norm": 1.3879232336736764, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 26587 + }, + { + "epoch": 0.26588, + "grad_norm": 1.2111718182760498, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 26588 + }, + { + "epoch": 0.26589, + "grad_norm": 1.3870104767770424, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26589 + }, + { + "epoch": 0.2659, + "grad_norm": 1.3775679905045883, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 26590 + }, + { + "epoch": 0.26591, + "grad_norm": 1.4071678635557108, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 26591 + }, + { + "epoch": 0.26592, + "grad_norm": 1.5778747313913009, + "learning_rate": 0.003, + "loss": 4.034, + "step": 26592 + }, + { + "epoch": 0.26593, + "grad_norm": 1.4031872097823, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 26593 + }, + { + "epoch": 0.26594, + "grad_norm": 1.6254873005052273, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26594 + }, + { + "epoch": 0.26595, + "grad_norm": 1.1774700131266542, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26595 + }, + { + "epoch": 0.26596, + "grad_norm": 1.4751688705461288, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26596 + }, + { + "epoch": 0.26597, + "grad_norm": 1.194953974194251, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26597 + }, + { + "epoch": 0.26598, + "grad_norm": 1.6979839837686603, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26598 + }, + { + "epoch": 0.26599, + "grad_norm": 1.1721667446439101, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 26599 + }, + { + "epoch": 0.266, + "grad_norm": 1.5437811053645742, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26600 + }, + { + "epoch": 0.26601, + "grad_norm": 1.5010223397625628, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 26601 + }, + { + "epoch": 0.26602, + "grad_norm": 1.4229416220491784, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 26602 + }, + { + "epoch": 0.26603, + "grad_norm": 1.3012040010707442, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 26603 + }, + { + "epoch": 0.26604, + "grad_norm": 1.4129557943721927, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 26604 + }, + { + "epoch": 0.26605, + "grad_norm": 1.4297395185994772, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26605 + }, + { + "epoch": 0.26606, + "grad_norm": 1.2421400998371457, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 26606 + }, + { + "epoch": 0.26607, + "grad_norm": 1.5039280800709824, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 26607 + }, + { + "epoch": 0.26608, + "grad_norm": 1.227485710547346, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 26608 + }, + { + "epoch": 0.26609, + "grad_norm": 1.3834437109031155, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 26609 + }, + { + "epoch": 0.2661, + "grad_norm": 1.4456933676321557, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 26610 + }, + { + "epoch": 0.26611, + "grad_norm": 1.415428262807568, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 26611 + }, + { + "epoch": 0.26612, + "grad_norm": 1.499721419718272, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 26612 + }, + { + "epoch": 0.26613, + "grad_norm": 1.7044551126133753, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 26613 + }, + { + "epoch": 0.26614, + "grad_norm": 1.3837747060605985, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 26614 + }, + { + "epoch": 0.26615, + "grad_norm": 1.562712125854259, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26615 + }, + { + "epoch": 0.26616, + "grad_norm": 1.2686473961612914, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 26616 + }, + { + "epoch": 0.26617, + "grad_norm": 1.4212883176134603, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 26617 + }, + { + "epoch": 0.26618, + "grad_norm": 1.251755176008923, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 26618 + }, + { + "epoch": 0.26619, + "grad_norm": 1.4047637364472372, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 26619 + }, + { + "epoch": 0.2662, + "grad_norm": 1.3836330018471155, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 26620 + }, + { + "epoch": 0.26621, + "grad_norm": 1.2257589106630797, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26621 + }, + { + "epoch": 0.26622, + "grad_norm": 1.5415113669599678, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26622 + }, + { + "epoch": 0.26623, + "grad_norm": 1.3079425039077317, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 26623 + }, + { + "epoch": 0.26624, + "grad_norm": 1.6076558026796917, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 26624 + }, + { + "epoch": 0.26625, + "grad_norm": 1.1264441081507357, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26625 + }, + { + "epoch": 0.26626, + "grad_norm": 1.7042468579570513, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 26626 + }, + { + "epoch": 0.26627, + "grad_norm": 1.239493905511783, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 26627 + }, + { + "epoch": 0.26628, + "grad_norm": 1.4316636743948694, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 26628 + }, + { + "epoch": 0.26629, + "grad_norm": 1.4640639382253755, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26629 + }, + { + "epoch": 0.2663, + "grad_norm": 1.2076833121594834, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 26630 + }, + { + "epoch": 0.26631, + "grad_norm": 1.5718276378035247, + "learning_rate": 0.003, + "loss": 4.055, + "step": 26631 + }, + { + "epoch": 0.26632, + "grad_norm": 1.3282627907104636, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 26632 + }, + { + "epoch": 0.26633, + "grad_norm": 1.5364187347359675, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26633 + }, + { + "epoch": 0.26634, + "grad_norm": 1.1748763015714745, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26634 + }, + { + "epoch": 0.26635, + "grad_norm": 1.7751094452303018, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 26635 + }, + { + "epoch": 0.26636, + "grad_norm": 1.3939325696511429, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 26636 + }, + { + "epoch": 0.26637, + "grad_norm": 1.598249046158806, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 26637 + }, + { + "epoch": 0.26638, + "grad_norm": 1.4795342727262144, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 26638 + }, + { + "epoch": 0.26639, + "grad_norm": 1.473774832750527, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 26639 + }, + { + "epoch": 0.2664, + "grad_norm": 1.6397333447949245, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 26640 + }, + { + "epoch": 0.26641, + "grad_norm": 1.1440561667997002, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 26641 + }, + { + "epoch": 0.26642, + "grad_norm": 1.4866523533621452, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 26642 + }, + { + "epoch": 0.26643, + "grad_norm": 1.486703616877837, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26643 + }, + { + "epoch": 0.26644, + "grad_norm": 1.3294002658902675, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 26644 + }, + { + "epoch": 0.26645, + "grad_norm": 1.4305549534316375, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 26645 + }, + { + "epoch": 0.26646, + "grad_norm": 1.3189172219349308, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 26646 + }, + { + "epoch": 0.26647, + "grad_norm": 1.6170132201732395, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 26647 + }, + { + "epoch": 0.26648, + "grad_norm": 1.3024130880237221, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 26648 + }, + { + "epoch": 0.26649, + "grad_norm": 1.441643756864115, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 26649 + }, + { + "epoch": 0.2665, + "grad_norm": 1.1050787040068037, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 26650 + }, + { + "epoch": 0.26651, + "grad_norm": 1.5081658606749688, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 26651 + }, + { + "epoch": 0.26652, + "grad_norm": 1.6096245679341064, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26652 + }, + { + "epoch": 0.26653, + "grad_norm": 1.444188286739025, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 26653 + }, + { + "epoch": 0.26654, + "grad_norm": 1.575607022024991, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 26654 + }, + { + "epoch": 0.26655, + "grad_norm": 1.3341777581550218, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 26655 + }, + { + "epoch": 0.26656, + "grad_norm": 1.277379156142287, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 26656 + }, + { + "epoch": 0.26657, + "grad_norm": 1.4080840193654476, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26657 + }, + { + "epoch": 0.26658, + "grad_norm": 1.4308670749025845, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 26658 + }, + { + "epoch": 0.26659, + "grad_norm": 1.1995662202783495, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 26659 + }, + { + "epoch": 0.2666, + "grad_norm": 1.4762507303284917, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 26660 + }, + { + "epoch": 0.26661, + "grad_norm": 1.3361662292899537, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 26661 + }, + { + "epoch": 0.26662, + "grad_norm": 1.3916920419713978, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 26662 + }, + { + "epoch": 0.26663, + "grad_norm": 1.3487127606959335, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 26663 + }, + { + "epoch": 0.26664, + "grad_norm": 1.621328277461404, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26664 + }, + { + "epoch": 0.26665, + "grad_norm": 1.318003777904212, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 26665 + }, + { + "epoch": 0.26666, + "grad_norm": 1.6600251242115296, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 26666 + }, + { + "epoch": 0.26667, + "grad_norm": 1.7185013936962605, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 26667 + }, + { + "epoch": 0.26668, + "grad_norm": 1.0075962021411984, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26668 + }, + { + "epoch": 0.26669, + "grad_norm": 1.5145603293236707, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 26669 + }, + { + "epoch": 0.2667, + "grad_norm": 1.4329681267653327, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 26670 + }, + { + "epoch": 0.26671, + "grad_norm": 1.7190586186968038, + "learning_rate": 0.003, + "loss": 4.025, + "step": 26671 + }, + { + "epoch": 0.26672, + "grad_norm": 1.4283975240556575, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 26672 + }, + { + "epoch": 0.26673, + "grad_norm": 1.3721731383108082, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 26673 + }, + { + "epoch": 0.26674, + "grad_norm": 1.4263423460027052, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26674 + }, + { + "epoch": 0.26675, + "grad_norm": 1.0879194397162553, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 26675 + }, + { + "epoch": 0.26676, + "grad_norm": 1.4861957009421436, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 26676 + }, + { + "epoch": 0.26677, + "grad_norm": 1.077344043647858, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 26677 + }, + { + "epoch": 0.26678, + "grad_norm": 1.795627945637991, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 26678 + }, + { + "epoch": 0.26679, + "grad_norm": 0.9887926421608312, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 26679 + }, + { + "epoch": 0.2668, + "grad_norm": 1.6831560884539751, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26680 + }, + { + "epoch": 0.26681, + "grad_norm": 1.2462354003360974, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 26681 + }, + { + "epoch": 0.26682, + "grad_norm": 1.887098857152304, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 26682 + }, + { + "epoch": 0.26683, + "grad_norm": 1.169003897298711, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 26683 + }, + { + "epoch": 0.26684, + "grad_norm": 1.563119193307645, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 26684 + }, + { + "epoch": 0.26685, + "grad_norm": 1.525373259564046, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 26685 + }, + { + "epoch": 0.26686, + "grad_norm": 1.3910737591131808, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 26686 + }, + { + "epoch": 0.26687, + "grad_norm": 1.443960619755905, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 26687 + }, + { + "epoch": 0.26688, + "grad_norm": 1.248297180438855, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 26688 + }, + { + "epoch": 0.26689, + "grad_norm": 1.5930055630912743, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 26689 + }, + { + "epoch": 0.2669, + "grad_norm": 1.3078357967589715, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 26690 + }, + { + "epoch": 0.26691, + "grad_norm": 1.8150040476882754, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 26691 + }, + { + "epoch": 0.26692, + "grad_norm": 1.3004955427267604, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 26692 + }, + { + "epoch": 0.26693, + "grad_norm": 1.3914185398434433, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 26693 + }, + { + "epoch": 0.26694, + "grad_norm": 1.4733650853651634, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 26694 + }, + { + "epoch": 0.26695, + "grad_norm": 1.2429041325997519, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26695 + }, + { + "epoch": 0.26696, + "grad_norm": 1.2420234859428614, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 26696 + }, + { + "epoch": 0.26697, + "grad_norm": 1.609116667466942, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 26697 + }, + { + "epoch": 0.26698, + "grad_norm": 1.2775683510352815, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 26698 + }, + { + "epoch": 0.26699, + "grad_norm": 1.6015169808656107, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 26699 + }, + { + "epoch": 0.267, + "grad_norm": 1.4267003009982748, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 26700 + }, + { + "epoch": 0.26701, + "grad_norm": 1.2398052573158167, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26701 + }, + { + "epoch": 0.26702, + "grad_norm": 1.357852223550653, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 26702 + }, + { + "epoch": 0.26703, + "grad_norm": 1.2993273646763723, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 26703 + }, + { + "epoch": 0.26704, + "grad_norm": 1.448174955309298, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26704 + }, + { + "epoch": 0.26705, + "grad_norm": 1.3612762439017818, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26705 + }, + { + "epoch": 0.26706, + "grad_norm": 1.4145825031613775, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 26706 + }, + { + "epoch": 0.26707, + "grad_norm": 1.3893187595035736, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 26707 + }, + { + "epoch": 0.26708, + "grad_norm": 1.2494549363880378, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 26708 + }, + { + "epoch": 0.26709, + "grad_norm": 1.5701896748284385, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 26709 + }, + { + "epoch": 0.2671, + "grad_norm": 1.2813445610476117, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 26710 + }, + { + "epoch": 0.26711, + "grad_norm": 1.4751156692351428, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 26711 + }, + { + "epoch": 0.26712, + "grad_norm": 1.0964039027052632, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 26712 + }, + { + "epoch": 0.26713, + "grad_norm": 1.6601356880774643, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 26713 + }, + { + "epoch": 0.26714, + "grad_norm": 1.2808241026424707, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 26714 + }, + { + "epoch": 0.26715, + "grad_norm": 1.5874311842456554, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26715 + }, + { + "epoch": 0.26716, + "grad_norm": 1.3055511743138744, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 26716 + }, + { + "epoch": 0.26717, + "grad_norm": 1.391847701134864, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 26717 + }, + { + "epoch": 0.26718, + "grad_norm": 1.1537920721436423, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 26718 + }, + { + "epoch": 0.26719, + "grad_norm": 1.6713678810454224, + "learning_rate": 0.003, + "loss": 4.079, + "step": 26719 + }, + { + "epoch": 0.2672, + "grad_norm": 1.2601348748671402, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 26720 + }, + { + "epoch": 0.26721, + "grad_norm": 1.65288077923644, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 26721 + }, + { + "epoch": 0.26722, + "grad_norm": 1.514849752335802, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 26722 + }, + { + "epoch": 0.26723, + "grad_norm": 1.1924930314327284, + "learning_rate": 0.003, + "loss": 4.062, + "step": 26723 + }, + { + "epoch": 0.26724, + "grad_norm": 1.2464048115638495, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 26724 + }, + { + "epoch": 0.26725, + "grad_norm": 1.678095610090629, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 26725 + }, + { + "epoch": 0.26726, + "grad_norm": 1.2941162185542716, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 26726 + }, + { + "epoch": 0.26727, + "grad_norm": 1.3689493705530362, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26727 + }, + { + "epoch": 0.26728, + "grad_norm": 1.6655014318606527, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 26728 + }, + { + "epoch": 0.26729, + "grad_norm": 1.3791551257053105, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 26729 + }, + { + "epoch": 0.2673, + "grad_norm": 1.160057919295123, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 26730 + }, + { + "epoch": 0.26731, + "grad_norm": 1.3486112294084487, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 26731 + }, + { + "epoch": 0.26732, + "grad_norm": 1.2593127558019679, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 26732 + }, + { + "epoch": 0.26733, + "grad_norm": 1.5789746172460193, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 26733 + }, + { + "epoch": 0.26734, + "grad_norm": 1.139960374263176, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 26734 + }, + { + "epoch": 0.26735, + "grad_norm": 1.5384121819697114, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 26735 + }, + { + "epoch": 0.26736, + "grad_norm": 1.2455264198470122, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 26736 + }, + { + "epoch": 0.26737, + "grad_norm": 1.7040055407930985, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 26737 + }, + { + "epoch": 0.26738, + "grad_norm": 1.2507371865813217, + "learning_rate": 0.003, + "loss": 4.064, + "step": 26738 + }, + { + "epoch": 0.26739, + "grad_norm": 1.3136200056835032, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 26739 + }, + { + "epoch": 0.2674, + "grad_norm": 1.6433004642545033, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 26740 + }, + { + "epoch": 0.26741, + "grad_norm": 1.811349377102091, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 26741 + }, + { + "epoch": 0.26742, + "grad_norm": 1.2645005804641716, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 26742 + }, + { + "epoch": 0.26743, + "grad_norm": 1.4989175306253504, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 26743 + }, + { + "epoch": 0.26744, + "grad_norm": 1.3419378100318529, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 26744 + }, + { + "epoch": 0.26745, + "grad_norm": 1.475979758659051, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 26745 + }, + { + "epoch": 0.26746, + "grad_norm": 1.3526237013293743, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 26746 + }, + { + "epoch": 0.26747, + "grad_norm": 1.5886307893543838, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 26747 + }, + { + "epoch": 0.26748, + "grad_norm": 1.0333426424195729, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 26748 + }, + { + "epoch": 0.26749, + "grad_norm": 1.860476346950679, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 26749 + }, + { + "epoch": 0.2675, + "grad_norm": 1.2739342996942238, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 26750 + }, + { + "epoch": 0.26751, + "grad_norm": 1.412470499344418, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 26751 + }, + { + "epoch": 0.26752, + "grad_norm": 1.5718093036210654, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 26752 + }, + { + "epoch": 0.26753, + "grad_norm": 1.2095481859562627, + "learning_rate": 0.003, + "loss": 4.071, + "step": 26753 + }, + { + "epoch": 0.26754, + "grad_norm": 1.2201404700020122, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 26754 + }, + { + "epoch": 0.26755, + "grad_norm": 1.5417125378859038, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 26755 + }, + { + "epoch": 0.26756, + "grad_norm": 1.1628657342438544, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26756 + }, + { + "epoch": 0.26757, + "grad_norm": 1.5537215907099826, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 26757 + }, + { + "epoch": 0.26758, + "grad_norm": 1.1981122574820653, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 26758 + }, + { + "epoch": 0.26759, + "grad_norm": 1.5531356317476965, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 26759 + }, + { + "epoch": 0.2676, + "grad_norm": 1.4026060156131315, + "learning_rate": 0.003, + "loss": 4.054, + "step": 26760 + }, + { + "epoch": 0.26761, + "grad_norm": 1.2878920312365374, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 26761 + }, + { + "epoch": 0.26762, + "grad_norm": 1.692193226555058, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 26762 + }, + { + "epoch": 0.26763, + "grad_norm": 0.9995441559853426, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26763 + }, + { + "epoch": 0.26764, + "grad_norm": 1.4497466485428148, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 26764 + }, + { + "epoch": 0.26765, + "grad_norm": 1.5227159735919722, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 26765 + }, + { + "epoch": 0.26766, + "grad_norm": 1.226228854088631, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 26766 + }, + { + "epoch": 0.26767, + "grad_norm": 1.4566401073824464, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 26767 + }, + { + "epoch": 0.26768, + "grad_norm": 1.2829119074065778, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 26768 + }, + { + "epoch": 0.26769, + "grad_norm": 1.5052578388657378, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26769 + }, + { + "epoch": 0.2677, + "grad_norm": 1.4795089978020641, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 26770 + }, + { + "epoch": 0.26771, + "grad_norm": 1.3759837551979457, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 26771 + }, + { + "epoch": 0.26772, + "grad_norm": 1.490728750107944, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26772 + }, + { + "epoch": 0.26773, + "grad_norm": 1.1848120366326307, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 26773 + }, + { + "epoch": 0.26774, + "grad_norm": 1.4686262208296472, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 26774 + }, + { + "epoch": 0.26775, + "grad_norm": 1.0422335019591413, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26775 + }, + { + "epoch": 0.26776, + "grad_norm": 1.3640469616901836, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 26776 + }, + { + "epoch": 0.26777, + "grad_norm": 1.332380212772556, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 26777 + }, + { + "epoch": 0.26778, + "grad_norm": 1.6446103778058, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 26778 + }, + { + "epoch": 0.26779, + "grad_norm": 1.5160755591581956, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26779 + }, + { + "epoch": 0.2678, + "grad_norm": 1.2379652768837655, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 26780 + }, + { + "epoch": 0.26781, + "grad_norm": 1.508171446063977, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 26781 + }, + { + "epoch": 0.26782, + "grad_norm": 1.2827793244204475, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 26782 + }, + { + "epoch": 0.26783, + "grad_norm": 1.645009664688264, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 26783 + }, + { + "epoch": 0.26784, + "grad_norm": 1.3926350175372146, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 26784 + }, + { + "epoch": 0.26785, + "grad_norm": 1.7444573124472338, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 26785 + }, + { + "epoch": 0.26786, + "grad_norm": 1.4946259010035567, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 26786 + }, + { + "epoch": 0.26787, + "grad_norm": 1.219175604721552, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 26787 + }, + { + "epoch": 0.26788, + "grad_norm": 1.3365341143108815, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 26788 + }, + { + "epoch": 0.26789, + "grad_norm": 1.590008139276348, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26789 + }, + { + "epoch": 0.2679, + "grad_norm": 1.321991766112513, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 26790 + }, + { + "epoch": 0.26791, + "grad_norm": 1.3350698044963298, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 26791 + }, + { + "epoch": 0.26792, + "grad_norm": 1.4268843002119211, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 26792 + }, + { + "epoch": 0.26793, + "grad_norm": 1.229048703764711, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26793 + }, + { + "epoch": 0.26794, + "grad_norm": 1.4579019533311723, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 26794 + }, + { + "epoch": 0.26795, + "grad_norm": 1.4323657940724823, + "learning_rate": 0.003, + "loss": 4.054, + "step": 26795 + }, + { + "epoch": 0.26796, + "grad_norm": 1.2531894761756757, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 26796 + }, + { + "epoch": 0.26797, + "grad_norm": 1.8532576334881454, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 26797 + }, + { + "epoch": 0.26798, + "grad_norm": 1.120507559098618, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26798 + }, + { + "epoch": 0.26799, + "grad_norm": 1.5110845993294952, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 26799 + }, + { + "epoch": 0.268, + "grad_norm": 1.216534501771361, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 26800 + }, + { + "epoch": 0.26801, + "grad_norm": 1.3905991917005631, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26801 + }, + { + "epoch": 0.26802, + "grad_norm": 1.3081984973977188, + "learning_rate": 0.003, + "loss": 4.028, + "step": 26802 + }, + { + "epoch": 0.26803, + "grad_norm": 1.3312103225008123, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26803 + }, + { + "epoch": 0.26804, + "grad_norm": 1.6216818096029721, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26804 + }, + { + "epoch": 0.26805, + "grad_norm": 1.4510930483944957, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 26805 + }, + { + "epoch": 0.26806, + "grad_norm": 1.5619254699459706, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 26806 + }, + { + "epoch": 0.26807, + "grad_norm": 1.227826346378614, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 26807 + }, + { + "epoch": 0.26808, + "grad_norm": 1.456228209293136, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 26808 + }, + { + "epoch": 0.26809, + "grad_norm": 1.3282454658012883, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 26809 + }, + { + "epoch": 0.2681, + "grad_norm": 1.6245077245392012, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26810 + }, + { + "epoch": 0.26811, + "grad_norm": 1.1209073264458933, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26811 + }, + { + "epoch": 0.26812, + "grad_norm": 1.5624253959581609, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 26812 + }, + { + "epoch": 0.26813, + "grad_norm": 1.3100162231949806, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 26813 + }, + { + "epoch": 0.26814, + "grad_norm": 1.5694313411575602, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 26814 + }, + { + "epoch": 0.26815, + "grad_norm": 1.3936581836546258, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 26815 + }, + { + "epoch": 0.26816, + "grad_norm": 1.4070365029742116, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 26816 + }, + { + "epoch": 0.26817, + "grad_norm": 1.3597589680632352, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26817 + }, + { + "epoch": 0.26818, + "grad_norm": 1.476241686996813, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 26818 + }, + { + "epoch": 0.26819, + "grad_norm": 1.2608756215142225, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 26819 + }, + { + "epoch": 0.2682, + "grad_norm": 1.6273872161367138, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26820 + }, + { + "epoch": 0.26821, + "grad_norm": 1.2760113350100195, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 26821 + }, + { + "epoch": 0.26822, + "grad_norm": 1.3526881523503895, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 26822 + }, + { + "epoch": 0.26823, + "grad_norm": 1.3915836719219856, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 26823 + }, + { + "epoch": 0.26824, + "grad_norm": 1.2969708680601533, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 26824 + }, + { + "epoch": 0.26825, + "grad_norm": 1.3404907235478547, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 26825 + }, + { + "epoch": 0.26826, + "grad_norm": 1.3464278639740996, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 26826 + }, + { + "epoch": 0.26827, + "grad_norm": 1.4236523563075614, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 26827 + }, + { + "epoch": 0.26828, + "grad_norm": 1.3079575348497892, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 26828 + }, + { + "epoch": 0.26829, + "grad_norm": 1.6092111165787808, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 26829 + }, + { + "epoch": 0.2683, + "grad_norm": 1.3690363192484618, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 26830 + }, + { + "epoch": 0.26831, + "grad_norm": 1.3039671413927016, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 26831 + }, + { + "epoch": 0.26832, + "grad_norm": 1.3294528766018092, + "learning_rate": 0.003, + "loss": 4.057, + "step": 26832 + }, + { + "epoch": 0.26833, + "grad_norm": 1.4038453848439114, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 26833 + }, + { + "epoch": 0.26834, + "grad_norm": 1.2969357990632502, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 26834 + }, + { + "epoch": 0.26835, + "grad_norm": 1.7090827273228895, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 26835 + }, + { + "epoch": 0.26836, + "grad_norm": 1.6565515436435998, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 26836 + }, + { + "epoch": 0.26837, + "grad_norm": 1.3749899986650358, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 26837 + }, + { + "epoch": 0.26838, + "grad_norm": 1.598280785236757, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 26838 + }, + { + "epoch": 0.26839, + "grad_norm": 1.2878588326575873, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 26839 + }, + { + "epoch": 0.2684, + "grad_norm": 1.2666573756946866, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 26840 + }, + { + "epoch": 0.26841, + "grad_norm": 1.5480404689547376, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26841 + }, + { + "epoch": 0.26842, + "grad_norm": 0.9553755254723726, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 26842 + }, + { + "epoch": 0.26843, + "grad_norm": 1.6748648464523408, + "learning_rate": 0.003, + "loss": 4.065, + "step": 26843 + }, + { + "epoch": 0.26844, + "grad_norm": 1.2505498849467764, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 26844 + }, + { + "epoch": 0.26845, + "grad_norm": 1.2884070869028958, + "learning_rate": 0.003, + "loss": 4.048, + "step": 26845 + }, + { + "epoch": 0.26846, + "grad_norm": 1.4255973179306118, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 26846 + }, + { + "epoch": 0.26847, + "grad_norm": 1.52902141763805, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 26847 + }, + { + "epoch": 0.26848, + "grad_norm": 1.7192362040162268, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 26848 + }, + { + "epoch": 0.26849, + "grad_norm": 1.4473154902280878, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26849 + }, + { + "epoch": 0.2685, + "grad_norm": 1.730780000932733, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 26850 + }, + { + "epoch": 0.26851, + "grad_norm": 1.300328260910973, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 26851 + }, + { + "epoch": 0.26852, + "grad_norm": 1.5751850766458364, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 26852 + }, + { + "epoch": 0.26853, + "grad_norm": 1.3357406251892854, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26853 + }, + { + "epoch": 0.26854, + "grad_norm": 1.3107020230555733, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 26854 + }, + { + "epoch": 0.26855, + "grad_norm": 1.5706405859974324, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 26855 + }, + { + "epoch": 0.26856, + "grad_norm": 1.4040494706017694, + "learning_rate": 0.003, + "loss": 4.022, + "step": 26856 + }, + { + "epoch": 0.26857, + "grad_norm": 1.4637263263996925, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 26857 + }, + { + "epoch": 0.26858, + "grad_norm": 1.1681114936323322, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 26858 + }, + { + "epoch": 0.26859, + "grad_norm": 1.4681635205450614, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 26859 + }, + { + "epoch": 0.2686, + "grad_norm": 1.3149799504751696, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 26860 + }, + { + "epoch": 0.26861, + "grad_norm": 1.4288761649881827, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 26861 + }, + { + "epoch": 0.26862, + "grad_norm": 1.4088333331237297, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 26862 + }, + { + "epoch": 0.26863, + "grad_norm": 1.2109938465844476, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 26863 + }, + { + "epoch": 0.26864, + "grad_norm": 1.4959131459809287, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 26864 + }, + { + "epoch": 0.26865, + "grad_norm": 1.4617272211262822, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 26865 + }, + { + "epoch": 0.26866, + "grad_norm": 1.659121827063586, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 26866 + }, + { + "epoch": 0.26867, + "grad_norm": 1.6753068957604234, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 26867 + }, + { + "epoch": 0.26868, + "grad_norm": 1.0083243264810444, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 26868 + }, + { + "epoch": 0.26869, + "grad_norm": 1.4145587081688176, + "learning_rate": 0.003, + "loss": 4.049, + "step": 26869 + }, + { + "epoch": 0.2687, + "grad_norm": 1.5789982511431508, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 26870 + }, + { + "epoch": 0.26871, + "grad_norm": 1.5193259710184623, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 26871 + }, + { + "epoch": 0.26872, + "grad_norm": 1.28966632269514, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 26872 + }, + { + "epoch": 0.26873, + "grad_norm": 1.4826463573225643, + "learning_rate": 0.003, + "loss": 4.03, + "step": 26873 + }, + { + "epoch": 0.26874, + "grad_norm": 1.3986788889232473, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 26874 + }, + { + "epoch": 0.26875, + "grad_norm": 1.1691091003801144, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 26875 + }, + { + "epoch": 0.26876, + "grad_norm": 1.7018725220215554, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 26876 + }, + { + "epoch": 0.26877, + "grad_norm": 1.2396349065110162, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 26877 + }, + { + "epoch": 0.26878, + "grad_norm": 1.7977148030057208, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 26878 + }, + { + "epoch": 0.26879, + "grad_norm": 1.1668135220490337, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26879 + }, + { + "epoch": 0.2688, + "grad_norm": 1.9302813227954556, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 26880 + }, + { + "epoch": 0.26881, + "grad_norm": 1.339633915201141, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26881 + }, + { + "epoch": 0.26882, + "grad_norm": 1.110366401452878, + "learning_rate": 0.003, + "loss": 4.056, + "step": 26882 + }, + { + "epoch": 0.26883, + "grad_norm": 1.5170962202439364, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 26883 + }, + { + "epoch": 0.26884, + "grad_norm": 1.477071470440326, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 26884 + }, + { + "epoch": 0.26885, + "grad_norm": 1.4033744025986659, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 26885 + }, + { + "epoch": 0.26886, + "grad_norm": 1.3992716750646348, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 26886 + }, + { + "epoch": 0.26887, + "grad_norm": 1.6994356007170142, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 26887 + }, + { + "epoch": 0.26888, + "grad_norm": 1.145865107689341, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 26888 + }, + { + "epoch": 0.26889, + "grad_norm": 1.5502804004331356, + "learning_rate": 0.003, + "loss": 4.052, + "step": 26889 + }, + { + "epoch": 0.2689, + "grad_norm": 1.175113972051805, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 26890 + }, + { + "epoch": 0.26891, + "grad_norm": 1.4721661172913143, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 26891 + }, + { + "epoch": 0.26892, + "grad_norm": 1.64214551645058, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 26892 + }, + { + "epoch": 0.26893, + "grad_norm": 1.1744381423141377, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 26893 + }, + { + "epoch": 0.26894, + "grad_norm": 1.7230640132263173, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 26894 + }, + { + "epoch": 0.26895, + "grad_norm": 1.3040505722554383, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 26895 + }, + { + "epoch": 0.26896, + "grad_norm": 1.4814148529431346, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 26896 + }, + { + "epoch": 0.26897, + "grad_norm": 1.5206733067864147, + "learning_rate": 0.003, + "loss": 4.071, + "step": 26897 + }, + { + "epoch": 0.26898, + "grad_norm": 1.4493183315850378, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 26898 + }, + { + "epoch": 0.26899, + "grad_norm": 1.5075361863585997, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 26899 + }, + { + "epoch": 0.269, + "grad_norm": 1.322562068302338, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 26900 + }, + { + "epoch": 0.26901, + "grad_norm": 1.5695661187682435, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 26901 + }, + { + "epoch": 0.26902, + "grad_norm": 1.400105379418201, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 26902 + }, + { + "epoch": 0.26903, + "grad_norm": 1.4648430004159272, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 26903 + }, + { + "epoch": 0.26904, + "grad_norm": 1.428402273004599, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 26904 + }, + { + "epoch": 0.26905, + "grad_norm": 1.5812415682341812, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26905 + }, + { + "epoch": 0.26906, + "grad_norm": 1.1289033207075927, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26906 + }, + { + "epoch": 0.26907, + "grad_norm": 1.6643535306324175, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 26907 + }, + { + "epoch": 0.26908, + "grad_norm": 1.4107770970214417, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 26908 + }, + { + "epoch": 0.26909, + "grad_norm": 1.623864648019138, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 26909 + }, + { + "epoch": 0.2691, + "grad_norm": 1.4981731658546473, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 26910 + }, + { + "epoch": 0.26911, + "grad_norm": 1.1966724422211688, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 26911 + }, + { + "epoch": 0.26912, + "grad_norm": 1.3507615067690277, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 26912 + }, + { + "epoch": 0.26913, + "grad_norm": 1.3429202284179753, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 26913 + }, + { + "epoch": 0.26914, + "grad_norm": 1.298984931088869, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 26914 + }, + { + "epoch": 0.26915, + "grad_norm": 1.3904684848623365, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 26915 + }, + { + "epoch": 0.26916, + "grad_norm": 1.1466261959578405, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 26916 + }, + { + "epoch": 0.26917, + "grad_norm": 1.5245580342371137, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 26917 + }, + { + "epoch": 0.26918, + "grad_norm": 1.3181214907181629, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26918 + }, + { + "epoch": 0.26919, + "grad_norm": 1.4111655571987154, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26919 + }, + { + "epoch": 0.2692, + "grad_norm": 1.3726850267671402, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 26920 + }, + { + "epoch": 0.26921, + "grad_norm": 1.6139401163669058, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 26921 + }, + { + "epoch": 0.26922, + "grad_norm": 1.7913209101078997, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 26922 + }, + { + "epoch": 0.26923, + "grad_norm": 1.2116951032973255, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 26923 + }, + { + "epoch": 0.26924, + "grad_norm": 1.9439445492141105, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 26924 + }, + { + "epoch": 0.26925, + "grad_norm": 1.189897810768441, + "learning_rate": 0.003, + "loss": 4.045, + "step": 26925 + }, + { + "epoch": 0.26926, + "grad_norm": 1.4109478638317463, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 26926 + }, + { + "epoch": 0.26927, + "grad_norm": 1.491732381661467, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 26927 + }, + { + "epoch": 0.26928, + "grad_norm": 1.3872302417343345, + "learning_rate": 0.003, + "loss": 4.061, + "step": 26928 + }, + { + "epoch": 0.26929, + "grad_norm": 1.5121569350855746, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 26929 + }, + { + "epoch": 0.2693, + "grad_norm": 1.2024525529919292, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 26930 + }, + { + "epoch": 0.26931, + "grad_norm": 1.4868103352810171, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 26931 + }, + { + "epoch": 0.26932, + "grad_norm": 1.3063005646308754, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 26932 + }, + { + "epoch": 0.26933, + "grad_norm": 1.4743952802284086, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 26933 + }, + { + "epoch": 0.26934, + "grad_norm": 1.2711776598731461, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 26934 + }, + { + "epoch": 0.26935, + "grad_norm": 1.6020251560254544, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 26935 + }, + { + "epoch": 0.26936, + "grad_norm": 1.0807709953442832, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 26936 + }, + { + "epoch": 0.26937, + "grad_norm": 1.5088290880464597, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 26937 + }, + { + "epoch": 0.26938, + "grad_norm": 1.4803442548973353, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 26938 + }, + { + "epoch": 0.26939, + "grad_norm": 1.7146387814168376, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 26939 + }, + { + "epoch": 0.2694, + "grad_norm": 1.281327267113094, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 26940 + }, + { + "epoch": 0.26941, + "grad_norm": 1.8014755733073073, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 26941 + }, + { + "epoch": 0.26942, + "grad_norm": 1.1350028880986807, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 26942 + }, + { + "epoch": 0.26943, + "grad_norm": 1.617819061822884, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 26943 + }, + { + "epoch": 0.26944, + "grad_norm": 1.4265451784875753, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 26944 + }, + { + "epoch": 0.26945, + "grad_norm": 1.5216398980778816, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 26945 + }, + { + "epoch": 0.26946, + "grad_norm": 1.2949741073630905, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 26946 + }, + { + "epoch": 0.26947, + "grad_norm": 1.1360898834117885, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 26947 + }, + { + "epoch": 0.26948, + "grad_norm": 1.3066068903693955, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 26948 + }, + { + "epoch": 0.26949, + "grad_norm": 1.3294727445823642, + "learning_rate": 0.003, + "loss": 4.047, + "step": 26949 + }, + { + "epoch": 0.2695, + "grad_norm": 1.3909053131963922, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 26950 + }, + { + "epoch": 0.26951, + "grad_norm": 1.244618150519542, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 26951 + }, + { + "epoch": 0.26952, + "grad_norm": 1.5010808846232542, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 26952 + }, + { + "epoch": 0.26953, + "grad_norm": 1.2718852285392248, + "learning_rate": 0.003, + "loss": 4.063, + "step": 26953 + }, + { + "epoch": 0.26954, + "grad_norm": 1.667665541434983, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 26954 + }, + { + "epoch": 0.26955, + "grad_norm": 1.2498361279004684, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 26955 + }, + { + "epoch": 0.26956, + "grad_norm": 1.3252299637521912, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 26956 + }, + { + "epoch": 0.26957, + "grad_norm": 1.273359342237186, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 26957 + }, + { + "epoch": 0.26958, + "grad_norm": 1.5262325387779256, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 26958 + }, + { + "epoch": 0.26959, + "grad_norm": 1.558715078624298, + "learning_rate": 0.003, + "loss": 4.054, + "step": 26959 + }, + { + "epoch": 0.2696, + "grad_norm": 1.5625175798482533, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 26960 + }, + { + "epoch": 0.26961, + "grad_norm": 1.6530087391489634, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26961 + }, + { + "epoch": 0.26962, + "grad_norm": 1.2663337960358343, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 26962 + }, + { + "epoch": 0.26963, + "grad_norm": 1.4978731797305536, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 26963 + }, + { + "epoch": 0.26964, + "grad_norm": 1.2139635136411364, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 26964 + }, + { + "epoch": 0.26965, + "grad_norm": 1.372079275381656, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 26965 + }, + { + "epoch": 0.26966, + "grad_norm": 1.4511833233468558, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 26966 + }, + { + "epoch": 0.26967, + "grad_norm": 1.287265821201264, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 26967 + }, + { + "epoch": 0.26968, + "grad_norm": 1.5434970745550556, + "learning_rate": 0.003, + "loss": 4.07, + "step": 26968 + }, + { + "epoch": 0.26969, + "grad_norm": 1.26080975091427, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 26969 + }, + { + "epoch": 0.2697, + "grad_norm": 1.6265506701284393, + "learning_rate": 0.003, + "loss": 4.034, + "step": 26970 + }, + { + "epoch": 0.26971, + "grad_norm": 1.4329653844027668, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 26971 + }, + { + "epoch": 0.26972, + "grad_norm": 1.3926932219641774, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 26972 + }, + { + "epoch": 0.26973, + "grad_norm": 1.474740943395521, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26973 + }, + { + "epoch": 0.26974, + "grad_norm": 1.3463684405418366, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 26974 + }, + { + "epoch": 0.26975, + "grad_norm": 1.2840107605856372, + "learning_rate": 0.003, + "loss": 4.027, + "step": 26975 + }, + { + "epoch": 0.26976, + "grad_norm": 1.59987433311809, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 26976 + }, + { + "epoch": 0.26977, + "grad_norm": 1.2760454059637383, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 26977 + }, + { + "epoch": 0.26978, + "grad_norm": 1.5689777664552804, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 26978 + }, + { + "epoch": 0.26979, + "grad_norm": 1.4346515384365428, + "learning_rate": 0.003, + "loss": 4.058, + "step": 26979 + }, + { + "epoch": 0.2698, + "grad_norm": 1.5532981218688446, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 26980 + }, + { + "epoch": 0.26981, + "grad_norm": 1.1496306080548377, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 26981 + }, + { + "epoch": 0.26982, + "grad_norm": 1.363673233715384, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 26982 + }, + { + "epoch": 0.26983, + "grad_norm": 1.3945920714195634, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 26983 + }, + { + "epoch": 0.26984, + "grad_norm": 1.4431893933123496, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 26984 + }, + { + "epoch": 0.26985, + "grad_norm": 1.5077127933518377, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 26985 + }, + { + "epoch": 0.26986, + "grad_norm": 1.4428321538113, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 26986 + }, + { + "epoch": 0.26987, + "grad_norm": 1.6482647504833332, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 26987 + }, + { + "epoch": 0.26988, + "grad_norm": 1.2302944098223152, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 26988 + }, + { + "epoch": 0.26989, + "grad_norm": 1.517876615750449, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 26989 + }, + { + "epoch": 0.2699, + "grad_norm": 1.307990613071044, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 26990 + }, + { + "epoch": 0.26991, + "grad_norm": 1.4690057657332023, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 26991 + }, + { + "epoch": 0.26992, + "grad_norm": 1.502124111661996, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 26992 + }, + { + "epoch": 0.26993, + "grad_norm": 1.4971276448081872, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 26993 + }, + { + "epoch": 0.26994, + "grad_norm": 1.3710139904651368, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 26994 + }, + { + "epoch": 0.26995, + "grad_norm": 1.809733948526321, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 26995 + }, + { + "epoch": 0.26996, + "grad_norm": 1.403513974935053, + "learning_rate": 0.003, + "loss": 4.047, + "step": 26996 + }, + { + "epoch": 0.26997, + "grad_norm": 1.2927365127167703, + "learning_rate": 0.003, + "loss": 4.044, + "step": 26997 + }, + { + "epoch": 0.26998, + "grad_norm": 1.4501912761118056, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 26998 + }, + { + "epoch": 0.26999, + "grad_norm": 1.4335991027055612, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 26999 + }, + { + "epoch": 0.27, + "grad_norm": 1.2106866346660732, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27000 + }, + { + "epoch": 0.27001, + "grad_norm": 1.2808612932819157, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 27001 + }, + { + "epoch": 0.27002, + "grad_norm": 1.8627591580209275, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 27002 + }, + { + "epoch": 0.27003, + "grad_norm": 1.203153649960348, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 27003 + }, + { + "epoch": 0.27004, + "grad_norm": 1.472087660169873, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 27004 + }, + { + "epoch": 0.27005, + "grad_norm": 1.590562605273029, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27005 + }, + { + "epoch": 0.27006, + "grad_norm": 1.4417208157135912, + "learning_rate": 0.003, + "loss": 4.043, + "step": 27006 + }, + { + "epoch": 0.27007, + "grad_norm": 1.308395403120845, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 27007 + }, + { + "epoch": 0.27008, + "grad_norm": 1.2873421252158301, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 27008 + }, + { + "epoch": 0.27009, + "grad_norm": 1.3730061086095349, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27009 + }, + { + "epoch": 0.2701, + "grad_norm": 1.2676168207196974, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 27010 + }, + { + "epoch": 0.27011, + "grad_norm": 1.5866094989545132, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27011 + }, + { + "epoch": 0.27012, + "grad_norm": 1.2387802322133599, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27012 + }, + { + "epoch": 0.27013, + "grad_norm": 1.5049658763267166, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 27013 + }, + { + "epoch": 0.27014, + "grad_norm": 1.4182899670466922, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27014 + }, + { + "epoch": 0.27015, + "grad_norm": 1.303088264042947, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 27015 + }, + { + "epoch": 0.27016, + "grad_norm": 1.266621224861267, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 27016 + }, + { + "epoch": 0.27017, + "grad_norm": 1.452083707468586, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 27017 + }, + { + "epoch": 0.27018, + "grad_norm": 1.5143702977322762, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 27018 + }, + { + "epoch": 0.27019, + "grad_norm": 1.1962565850727893, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 27019 + }, + { + "epoch": 0.2702, + "grad_norm": 1.5329823823978, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 27020 + }, + { + "epoch": 0.27021, + "grad_norm": 1.2621218074794365, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 27021 + }, + { + "epoch": 0.27022, + "grad_norm": 1.8223193132892326, + "learning_rate": 0.003, + "loss": 4.031, + "step": 27022 + }, + { + "epoch": 0.27023, + "grad_norm": 1.1676222831624252, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 27023 + }, + { + "epoch": 0.27024, + "grad_norm": 1.4203179089645541, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 27024 + }, + { + "epoch": 0.27025, + "grad_norm": 1.7091521647437877, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 27025 + }, + { + "epoch": 0.27026, + "grad_norm": 1.3311679636837177, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27026 + }, + { + "epoch": 0.27027, + "grad_norm": 1.6126818900464512, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 27027 + }, + { + "epoch": 0.27028, + "grad_norm": 1.4386645392703201, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27028 + }, + { + "epoch": 0.27029, + "grad_norm": 1.340605252642936, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27029 + }, + { + "epoch": 0.2703, + "grad_norm": 1.185219013669222, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 27030 + }, + { + "epoch": 0.27031, + "grad_norm": 1.6717160228808219, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 27031 + }, + { + "epoch": 0.27032, + "grad_norm": 1.24850162148913, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27032 + }, + { + "epoch": 0.27033, + "grad_norm": 1.351341854616695, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27033 + }, + { + "epoch": 0.27034, + "grad_norm": 1.5871957035196431, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 27034 + }, + { + "epoch": 0.27035, + "grad_norm": 1.1507730024564422, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27035 + }, + { + "epoch": 0.27036, + "grad_norm": 1.6557972745564298, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 27036 + }, + { + "epoch": 0.27037, + "grad_norm": 1.6824100716015578, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 27037 + }, + { + "epoch": 0.27038, + "grad_norm": 1.634757917857535, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27038 + }, + { + "epoch": 0.27039, + "grad_norm": 1.3582900749756506, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27039 + }, + { + "epoch": 0.2704, + "grad_norm": 1.559975525929497, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 27040 + }, + { + "epoch": 0.27041, + "grad_norm": 1.2497874345961653, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27041 + }, + { + "epoch": 0.27042, + "grad_norm": 1.4481457271225144, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 27042 + }, + { + "epoch": 0.27043, + "grad_norm": 1.1026101225706322, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27043 + }, + { + "epoch": 0.27044, + "grad_norm": 1.6022764246127537, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27044 + }, + { + "epoch": 0.27045, + "grad_norm": 1.2721582962638713, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 27045 + }, + { + "epoch": 0.27046, + "grad_norm": 1.4016674233349977, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 27046 + }, + { + "epoch": 0.27047, + "grad_norm": 1.302535804696179, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 27047 + }, + { + "epoch": 0.27048, + "grad_norm": 1.5260954840703185, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27048 + }, + { + "epoch": 0.27049, + "grad_norm": 1.3540254090863386, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 27049 + }, + { + "epoch": 0.2705, + "grad_norm": 1.6020259597077122, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 27050 + }, + { + "epoch": 0.27051, + "grad_norm": 1.489817413698479, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 27051 + }, + { + "epoch": 0.27052, + "grad_norm": 1.5748622516507622, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 27052 + }, + { + "epoch": 0.27053, + "grad_norm": 1.324709186027098, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 27053 + }, + { + "epoch": 0.27054, + "grad_norm": 1.8023721595248579, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 27054 + }, + { + "epoch": 0.27055, + "grad_norm": 1.2868825147743732, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27055 + }, + { + "epoch": 0.27056, + "grad_norm": 1.3262017434061177, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 27056 + }, + { + "epoch": 0.27057, + "grad_norm": 1.5344071000963564, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27057 + }, + { + "epoch": 0.27058, + "grad_norm": 1.4094825263333297, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 27058 + }, + { + "epoch": 0.27059, + "grad_norm": 1.4538020811429941, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27059 + }, + { + "epoch": 0.2706, + "grad_norm": 1.2135593071689947, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 27060 + }, + { + "epoch": 0.27061, + "grad_norm": 1.540307710805703, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 27061 + }, + { + "epoch": 0.27062, + "grad_norm": 1.2748812746733778, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 27062 + }, + { + "epoch": 0.27063, + "grad_norm": 1.3731366140164474, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 27063 + }, + { + "epoch": 0.27064, + "grad_norm": 1.2198690920299886, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 27064 + }, + { + "epoch": 0.27065, + "grad_norm": 1.3030463315661243, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 27065 + }, + { + "epoch": 0.27066, + "grad_norm": 1.345729718144233, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 27066 + }, + { + "epoch": 0.27067, + "grad_norm": 1.3859077534645892, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 27067 + }, + { + "epoch": 0.27068, + "grad_norm": 1.2829382290772628, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 27068 + }, + { + "epoch": 0.27069, + "grad_norm": 1.3516535972622805, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27069 + }, + { + "epoch": 0.2707, + "grad_norm": 1.2407428707144308, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 27070 + }, + { + "epoch": 0.27071, + "grad_norm": 1.533143915358823, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 27071 + }, + { + "epoch": 0.27072, + "grad_norm": 1.3371451288561327, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 27072 + }, + { + "epoch": 0.27073, + "grad_norm": 1.348418093586684, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 27073 + }, + { + "epoch": 0.27074, + "grad_norm": 1.556384500713179, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27074 + }, + { + "epoch": 0.27075, + "grad_norm": 1.2681211326496216, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27075 + }, + { + "epoch": 0.27076, + "grad_norm": 1.5859670767070932, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27076 + }, + { + "epoch": 0.27077, + "grad_norm": 1.5512928556473078, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 27077 + }, + { + "epoch": 0.27078, + "grad_norm": 1.68222908245622, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 27078 + }, + { + "epoch": 0.27079, + "grad_norm": 1.2228128367258795, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 27079 + }, + { + "epoch": 0.2708, + "grad_norm": 1.1620315094952194, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 27080 + }, + { + "epoch": 0.27081, + "grad_norm": 1.4663521590599138, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 27081 + }, + { + "epoch": 0.27082, + "grad_norm": 1.42550118225606, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 27082 + }, + { + "epoch": 0.27083, + "grad_norm": 1.4795583311302396, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27083 + }, + { + "epoch": 0.27084, + "grad_norm": 1.3639315372234706, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 27084 + }, + { + "epoch": 0.27085, + "grad_norm": 1.4981682640378926, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27085 + }, + { + "epoch": 0.27086, + "grad_norm": 1.3801246775423244, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27086 + }, + { + "epoch": 0.27087, + "grad_norm": 1.4880469202764184, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27087 + }, + { + "epoch": 0.27088, + "grad_norm": 1.2824978287801372, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 27088 + }, + { + "epoch": 0.27089, + "grad_norm": 1.5918878918580246, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 27089 + }, + { + "epoch": 0.2709, + "grad_norm": 1.2230943741347018, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 27090 + }, + { + "epoch": 0.27091, + "grad_norm": 1.346930444885163, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 27091 + }, + { + "epoch": 0.27092, + "grad_norm": 1.2866945503318021, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 27092 + }, + { + "epoch": 0.27093, + "grad_norm": 1.3223293217555507, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 27093 + }, + { + "epoch": 0.27094, + "grad_norm": 1.632809998598338, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 27094 + }, + { + "epoch": 0.27095, + "grad_norm": 1.5121758103920764, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27095 + }, + { + "epoch": 0.27096, + "grad_norm": 1.2327607046660043, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27096 + }, + { + "epoch": 0.27097, + "grad_norm": 1.648898230083887, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 27097 + }, + { + "epoch": 0.27098, + "grad_norm": 1.2834744875577775, + "learning_rate": 0.003, + "loss": 4.046, + "step": 27098 + }, + { + "epoch": 0.27099, + "grad_norm": 1.9466056426372496, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 27099 + }, + { + "epoch": 0.271, + "grad_norm": 1.350439436223932, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 27100 + }, + { + "epoch": 0.27101, + "grad_norm": 1.5110752464142043, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 27101 + }, + { + "epoch": 0.27102, + "grad_norm": 1.413752752896247, + "learning_rate": 0.003, + "loss": 4.037, + "step": 27102 + }, + { + "epoch": 0.27103, + "grad_norm": 1.614083980796784, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27103 + }, + { + "epoch": 0.27104, + "grad_norm": 1.1675871967285292, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 27104 + }, + { + "epoch": 0.27105, + "grad_norm": 1.5741413069893073, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 27105 + }, + { + "epoch": 0.27106, + "grad_norm": 1.3829972403290043, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27106 + }, + { + "epoch": 0.27107, + "grad_norm": 1.5180850383953755, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 27107 + }, + { + "epoch": 0.27108, + "grad_norm": 1.265687708436714, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 27108 + }, + { + "epoch": 0.27109, + "grad_norm": 1.7012454621611963, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 27109 + }, + { + "epoch": 0.2711, + "grad_norm": 1.337350924497909, + "learning_rate": 0.003, + "loss": 4.069, + "step": 27110 + }, + { + "epoch": 0.27111, + "grad_norm": 1.5876284478896456, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 27111 + }, + { + "epoch": 0.27112, + "grad_norm": 1.1792009381214559, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 27112 + }, + { + "epoch": 0.27113, + "grad_norm": 1.5260530973396818, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27113 + }, + { + "epoch": 0.27114, + "grad_norm": 1.1895102085523364, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 27114 + }, + { + "epoch": 0.27115, + "grad_norm": 1.6781207958893618, + "learning_rate": 0.003, + "loss": 4.06, + "step": 27115 + }, + { + "epoch": 0.27116, + "grad_norm": 1.3134908475286398, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27116 + }, + { + "epoch": 0.27117, + "grad_norm": 1.5003823095031434, + "learning_rate": 0.003, + "loss": 4.061, + "step": 27117 + }, + { + "epoch": 0.27118, + "grad_norm": 1.3847557815064635, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27118 + }, + { + "epoch": 0.27119, + "grad_norm": 1.2636137602019564, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27119 + }, + { + "epoch": 0.2712, + "grad_norm": 1.3031098356546473, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 27120 + }, + { + "epoch": 0.27121, + "grad_norm": 1.4496636360890205, + "learning_rate": 0.003, + "loss": 4.03, + "step": 27121 + }, + { + "epoch": 0.27122, + "grad_norm": 1.3867218433981028, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 27122 + }, + { + "epoch": 0.27123, + "grad_norm": 1.526834185716347, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 27123 + }, + { + "epoch": 0.27124, + "grad_norm": 1.093894569588041, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27124 + }, + { + "epoch": 0.27125, + "grad_norm": 1.5526749510803455, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 27125 + }, + { + "epoch": 0.27126, + "grad_norm": 1.0969468340258923, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 27126 + }, + { + "epoch": 0.27127, + "grad_norm": 1.6935012911489578, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 27127 + }, + { + "epoch": 0.27128, + "grad_norm": 1.4351298498143739, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 27128 + }, + { + "epoch": 0.27129, + "grad_norm": 1.4888315594338815, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 27129 + }, + { + "epoch": 0.2713, + "grad_norm": 1.4896096403159622, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 27130 + }, + { + "epoch": 0.27131, + "grad_norm": 1.7246888453382525, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 27131 + }, + { + "epoch": 0.27132, + "grad_norm": 1.3949331119944772, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 27132 + }, + { + "epoch": 0.27133, + "grad_norm": 1.5035758292112225, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27133 + }, + { + "epoch": 0.27134, + "grad_norm": 1.3685010786786926, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27134 + }, + { + "epoch": 0.27135, + "grad_norm": 1.6872342231226318, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 27135 + }, + { + "epoch": 0.27136, + "grad_norm": 1.0686872695315939, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 27136 + }, + { + "epoch": 0.27137, + "grad_norm": 1.579170037447245, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27137 + }, + { + "epoch": 0.27138, + "grad_norm": 1.4335586076384188, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27138 + }, + { + "epoch": 0.27139, + "grad_norm": 1.6372305863254273, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 27139 + }, + { + "epoch": 0.2714, + "grad_norm": 1.273265627749544, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 27140 + }, + { + "epoch": 0.27141, + "grad_norm": 1.5764750520993596, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 27141 + }, + { + "epoch": 0.27142, + "grad_norm": 1.330181537460064, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 27142 + }, + { + "epoch": 0.27143, + "grad_norm": 1.750265152576352, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 27143 + }, + { + "epoch": 0.27144, + "grad_norm": 1.0294273871606157, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 27144 + }, + { + "epoch": 0.27145, + "grad_norm": 1.5301939997488445, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 27145 + }, + { + "epoch": 0.27146, + "grad_norm": 1.2450681160589856, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 27146 + }, + { + "epoch": 0.27147, + "grad_norm": 1.4212423513626637, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27147 + }, + { + "epoch": 0.27148, + "grad_norm": 1.2258667653157886, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27148 + }, + { + "epoch": 0.27149, + "grad_norm": 1.6237833974694051, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 27149 + }, + { + "epoch": 0.2715, + "grad_norm": 1.3614819663003723, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27150 + }, + { + "epoch": 0.27151, + "grad_norm": 1.6870660564251712, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 27151 + }, + { + "epoch": 0.27152, + "grad_norm": 1.4056300076391282, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 27152 + }, + { + "epoch": 0.27153, + "grad_norm": 1.5896583064301748, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27153 + }, + { + "epoch": 0.27154, + "grad_norm": 1.3533235877681065, + "learning_rate": 0.003, + "loss": 4.079, + "step": 27154 + }, + { + "epoch": 0.27155, + "grad_norm": 1.475301983027714, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 27155 + }, + { + "epoch": 0.27156, + "grad_norm": 1.2693501790604502, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 27156 + }, + { + "epoch": 0.27157, + "grad_norm": 1.478213736637982, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 27157 + }, + { + "epoch": 0.27158, + "grad_norm": 1.3753767132508734, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27158 + }, + { + "epoch": 0.27159, + "grad_norm": 1.3704536951093826, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 27159 + }, + { + "epoch": 0.2716, + "grad_norm": 1.6906386623380711, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 27160 + }, + { + "epoch": 0.27161, + "grad_norm": 1.4162849969698132, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 27161 + }, + { + "epoch": 0.27162, + "grad_norm": 1.2993185823389635, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 27162 + }, + { + "epoch": 0.27163, + "grad_norm": 1.624413177053113, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 27163 + }, + { + "epoch": 0.27164, + "grad_norm": 1.357416371191483, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27164 + }, + { + "epoch": 0.27165, + "grad_norm": 1.3125697115921524, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27165 + }, + { + "epoch": 0.27166, + "grad_norm": 1.4744475385126714, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 27166 + }, + { + "epoch": 0.27167, + "grad_norm": 1.446053233830993, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 27167 + }, + { + "epoch": 0.27168, + "grad_norm": 1.5983129432057959, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27168 + }, + { + "epoch": 0.27169, + "grad_norm": 0.9779159477080758, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27169 + }, + { + "epoch": 0.2717, + "grad_norm": 1.8132499514083769, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 27170 + }, + { + "epoch": 0.27171, + "grad_norm": 1.2318231442021703, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27171 + }, + { + "epoch": 0.27172, + "grad_norm": 1.4280179920043925, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 27172 + }, + { + "epoch": 0.27173, + "grad_norm": 1.5179798229726857, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 27173 + }, + { + "epoch": 0.27174, + "grad_norm": 1.1701702098665498, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27174 + }, + { + "epoch": 0.27175, + "grad_norm": 1.689758019483156, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 27175 + }, + { + "epoch": 0.27176, + "grad_norm": 1.8715778351864893, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 27176 + }, + { + "epoch": 0.27177, + "grad_norm": 1.2349901852530836, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 27177 + }, + { + "epoch": 0.27178, + "grad_norm": 1.6852271948366062, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 27178 + }, + { + "epoch": 0.27179, + "grad_norm": 1.550065193336806, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 27179 + }, + { + "epoch": 0.2718, + "grad_norm": 1.168115376420016, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 27180 + }, + { + "epoch": 0.27181, + "grad_norm": 1.4221799313019878, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 27181 + }, + { + "epoch": 0.27182, + "grad_norm": 1.3262416574174951, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27182 + }, + { + "epoch": 0.27183, + "grad_norm": 1.2858175462320707, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 27183 + }, + { + "epoch": 0.27184, + "grad_norm": 1.3434271065938637, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 27184 + }, + { + "epoch": 0.27185, + "grad_norm": 1.2301467408152482, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 27185 + }, + { + "epoch": 0.27186, + "grad_norm": 1.4399379554117426, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 27186 + }, + { + "epoch": 0.27187, + "grad_norm": 1.41584360047186, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 27187 + }, + { + "epoch": 0.27188, + "grad_norm": 1.3113516046344833, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27188 + }, + { + "epoch": 0.27189, + "grad_norm": 1.488424525497477, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27189 + }, + { + "epoch": 0.2719, + "grad_norm": 1.4802543052632096, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27190 + }, + { + "epoch": 0.27191, + "grad_norm": 1.478385565829742, + "learning_rate": 0.003, + "loss": 4.044, + "step": 27191 + }, + { + "epoch": 0.27192, + "grad_norm": 1.2239928364530541, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 27192 + }, + { + "epoch": 0.27193, + "grad_norm": 1.6241747943870226, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 27193 + }, + { + "epoch": 0.27194, + "grad_norm": 1.2625503195909824, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 27194 + }, + { + "epoch": 0.27195, + "grad_norm": 1.357941645670838, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 27195 + }, + { + "epoch": 0.27196, + "grad_norm": 1.3694038670351592, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 27196 + }, + { + "epoch": 0.27197, + "grad_norm": 1.431329817492108, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 27197 + }, + { + "epoch": 0.27198, + "grad_norm": 1.3212926017567994, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27198 + }, + { + "epoch": 0.27199, + "grad_norm": 1.3536577504962533, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27199 + }, + { + "epoch": 0.272, + "grad_norm": 1.4469725433013567, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 27200 + }, + { + "epoch": 0.27201, + "grad_norm": 1.3278032149444634, + "learning_rate": 0.003, + "loss": 4.045, + "step": 27201 + }, + { + "epoch": 0.27202, + "grad_norm": 1.3834526593886252, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 27202 + }, + { + "epoch": 0.27203, + "grad_norm": 1.1770538782282398, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 27203 + }, + { + "epoch": 0.27204, + "grad_norm": 1.9762871509917797, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 27204 + }, + { + "epoch": 0.27205, + "grad_norm": 1.2506275328638383, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27205 + }, + { + "epoch": 0.27206, + "grad_norm": 1.2908752967402042, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27206 + }, + { + "epoch": 0.27207, + "grad_norm": 1.756272081943779, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 27207 + }, + { + "epoch": 0.27208, + "grad_norm": 1.4600059530143494, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 27208 + }, + { + "epoch": 0.27209, + "grad_norm": 1.3489148340741934, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27209 + }, + { + "epoch": 0.2721, + "grad_norm": 1.3402422096449238, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 27210 + }, + { + "epoch": 0.27211, + "grad_norm": 1.5901163365728526, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 27211 + }, + { + "epoch": 0.27212, + "grad_norm": 1.2046224414770978, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 27212 + }, + { + "epoch": 0.27213, + "grad_norm": 1.508147854251928, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27213 + }, + { + "epoch": 0.27214, + "grad_norm": 1.4438949044728795, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 27214 + }, + { + "epoch": 0.27215, + "grad_norm": 1.3328764227220158, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 27215 + }, + { + "epoch": 0.27216, + "grad_norm": 1.4079962828518326, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 27216 + }, + { + "epoch": 0.27217, + "grad_norm": 1.3544183797766685, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 27217 + }, + { + "epoch": 0.27218, + "grad_norm": 2.060725869415366, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27218 + }, + { + "epoch": 0.27219, + "grad_norm": 1.3141735163680741, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27219 + }, + { + "epoch": 0.2722, + "grad_norm": 1.3326007379478548, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27220 + }, + { + "epoch": 0.27221, + "grad_norm": 1.6762846613197064, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 27221 + }, + { + "epoch": 0.27222, + "grad_norm": 1.2599759106091708, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 27222 + }, + { + "epoch": 0.27223, + "grad_norm": 1.747136209111518, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 27223 + }, + { + "epoch": 0.27224, + "grad_norm": 1.1768704882528986, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27224 + }, + { + "epoch": 0.27225, + "grad_norm": 1.460338454223819, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 27225 + }, + { + "epoch": 0.27226, + "grad_norm": 1.3518718030513337, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27226 + }, + { + "epoch": 0.27227, + "grad_norm": 1.2735606219842277, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27227 + }, + { + "epoch": 0.27228, + "grad_norm": 1.5796774524474804, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 27228 + }, + { + "epoch": 0.27229, + "grad_norm": 1.3651862817350302, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 27229 + }, + { + "epoch": 0.2723, + "grad_norm": 1.5715752791983386, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 27230 + }, + { + "epoch": 0.27231, + "grad_norm": 1.4599478268830202, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27231 + }, + { + "epoch": 0.27232, + "grad_norm": 1.4065256348440203, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27232 + }, + { + "epoch": 0.27233, + "grad_norm": 1.5308328066922656, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27233 + }, + { + "epoch": 0.27234, + "grad_norm": 1.3747055372190413, + "learning_rate": 0.003, + "loss": 4.036, + "step": 27234 + }, + { + "epoch": 0.27235, + "grad_norm": 1.4394760731227698, + "learning_rate": 0.003, + "loss": 4.022, + "step": 27235 + }, + { + "epoch": 0.27236, + "grad_norm": 1.4995482568679233, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27236 + }, + { + "epoch": 0.27237, + "grad_norm": 1.2646558297460742, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 27237 + }, + { + "epoch": 0.27238, + "grad_norm": 1.5676630173236865, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 27238 + }, + { + "epoch": 0.27239, + "grad_norm": 1.5221458640971095, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 27239 + }, + { + "epoch": 0.2724, + "grad_norm": 1.2759249985800576, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27240 + }, + { + "epoch": 0.27241, + "grad_norm": 1.4087172123365368, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 27241 + }, + { + "epoch": 0.27242, + "grad_norm": 1.4321628692489834, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27242 + }, + { + "epoch": 0.27243, + "grad_norm": 1.4031438174006965, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27243 + }, + { + "epoch": 0.27244, + "grad_norm": 1.447834503124191, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 27244 + }, + { + "epoch": 0.27245, + "grad_norm": 1.2105288731763346, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 27245 + }, + { + "epoch": 0.27246, + "grad_norm": 1.456475057265555, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 27246 + }, + { + "epoch": 0.27247, + "grad_norm": 1.3171490733464788, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 27247 + }, + { + "epoch": 0.27248, + "grad_norm": 1.6076687332859219, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 27248 + }, + { + "epoch": 0.27249, + "grad_norm": 1.2321979365600506, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 27249 + }, + { + "epoch": 0.2725, + "grad_norm": 1.5577793082018239, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27250 + }, + { + "epoch": 0.27251, + "grad_norm": 1.2939579197605702, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 27251 + }, + { + "epoch": 0.27252, + "grad_norm": 1.590994139973501, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 27252 + }, + { + "epoch": 0.27253, + "grad_norm": 1.191844518065357, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27253 + }, + { + "epoch": 0.27254, + "grad_norm": 1.44812991065868, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 27254 + }, + { + "epoch": 0.27255, + "grad_norm": 1.4311435224625388, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27255 + }, + { + "epoch": 0.27256, + "grad_norm": 1.5728213757106058, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 27256 + }, + { + "epoch": 0.27257, + "grad_norm": 1.5729127596185024, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 27257 + }, + { + "epoch": 0.27258, + "grad_norm": 1.5764189228676369, + "learning_rate": 0.003, + "loss": 4.046, + "step": 27258 + }, + { + "epoch": 0.27259, + "grad_norm": 1.455783566480044, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27259 + }, + { + "epoch": 0.2726, + "grad_norm": 1.3757718701494885, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 27260 + }, + { + "epoch": 0.27261, + "grad_norm": 1.490159537637259, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27261 + }, + { + "epoch": 0.27262, + "grad_norm": 1.3049953995706551, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 27262 + }, + { + "epoch": 0.27263, + "grad_norm": 1.4925680726151052, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 27263 + }, + { + "epoch": 0.27264, + "grad_norm": 1.2166214914146738, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 27264 + }, + { + "epoch": 0.27265, + "grad_norm": 1.727206170242687, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 27265 + }, + { + "epoch": 0.27266, + "grad_norm": 1.4938862054626616, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 27266 + }, + { + "epoch": 0.27267, + "grad_norm": 1.2423388569675637, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27267 + }, + { + "epoch": 0.27268, + "grad_norm": 1.419022405389692, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 27268 + }, + { + "epoch": 0.27269, + "grad_norm": 1.4844355081422644, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 27269 + }, + { + "epoch": 0.2727, + "grad_norm": 1.226168733031999, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 27270 + }, + { + "epoch": 0.27271, + "grad_norm": 1.4601077679659176, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 27271 + }, + { + "epoch": 0.27272, + "grad_norm": 1.305979428106994, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27272 + }, + { + "epoch": 0.27273, + "grad_norm": 1.352592671062095, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 27273 + }, + { + "epoch": 0.27274, + "grad_norm": 1.3403721354634313, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 27274 + }, + { + "epoch": 0.27275, + "grad_norm": 1.559384245501926, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 27275 + }, + { + "epoch": 0.27276, + "grad_norm": 1.4885323468549345, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 27276 + }, + { + "epoch": 0.27277, + "grad_norm": 1.2406390441535569, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 27277 + }, + { + "epoch": 0.27278, + "grad_norm": 1.526067976296285, + "learning_rate": 0.003, + "loss": 4.024, + "step": 27278 + }, + { + "epoch": 0.27279, + "grad_norm": 1.453965252792814, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 27279 + }, + { + "epoch": 0.2728, + "grad_norm": 1.2017585121982337, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27280 + }, + { + "epoch": 0.27281, + "grad_norm": 1.3881618001776235, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 27281 + }, + { + "epoch": 0.27282, + "grad_norm": 1.388499821736868, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 27282 + }, + { + "epoch": 0.27283, + "grad_norm": 1.3042508633060137, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 27283 + }, + { + "epoch": 0.27284, + "grad_norm": 1.6987924011605202, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 27284 + }, + { + "epoch": 0.27285, + "grad_norm": 1.1548912204292454, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 27285 + }, + { + "epoch": 0.27286, + "grad_norm": 1.5147205443453398, + "learning_rate": 0.003, + "loss": 4.06, + "step": 27286 + }, + { + "epoch": 0.27287, + "grad_norm": 1.420761728847727, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27287 + }, + { + "epoch": 0.27288, + "grad_norm": 1.6828833475089953, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27288 + }, + { + "epoch": 0.27289, + "grad_norm": 1.2507533287957298, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27289 + }, + { + "epoch": 0.2729, + "grad_norm": 1.4908493069274626, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 27290 + }, + { + "epoch": 0.27291, + "grad_norm": 1.615683421253535, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 27291 + }, + { + "epoch": 0.27292, + "grad_norm": 1.366009642135794, + "learning_rate": 0.003, + "loss": 4.017, + "step": 27292 + }, + { + "epoch": 0.27293, + "grad_norm": 1.0808462935270462, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 27293 + }, + { + "epoch": 0.27294, + "grad_norm": 1.8260248440578517, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 27294 + }, + { + "epoch": 0.27295, + "grad_norm": 1.0664851259128345, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 27295 + }, + { + "epoch": 0.27296, + "grad_norm": 1.4791565072151576, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27296 + }, + { + "epoch": 0.27297, + "grad_norm": 1.3200306138114566, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27297 + }, + { + "epoch": 0.27298, + "grad_norm": 1.1501793496882364, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 27298 + }, + { + "epoch": 0.27299, + "grad_norm": 1.5624584251026161, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 27299 + }, + { + "epoch": 0.273, + "grad_norm": 1.3406447896470206, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 27300 + }, + { + "epoch": 0.27301, + "grad_norm": 1.5115438895033653, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 27301 + }, + { + "epoch": 0.27302, + "grad_norm": 1.590271098340533, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 27302 + }, + { + "epoch": 0.27303, + "grad_norm": 1.6209456935755695, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27303 + }, + { + "epoch": 0.27304, + "grad_norm": 1.6226857682036804, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 27304 + }, + { + "epoch": 0.27305, + "grad_norm": 1.3955978509008178, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27305 + }, + { + "epoch": 0.27306, + "grad_norm": 1.4793517665724751, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 27306 + }, + { + "epoch": 0.27307, + "grad_norm": 1.5505081527268303, + "learning_rate": 0.003, + "loss": 4.038, + "step": 27307 + }, + { + "epoch": 0.27308, + "grad_norm": 1.2011967200442395, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27308 + }, + { + "epoch": 0.27309, + "grad_norm": 1.5794835791042594, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 27309 + }, + { + "epoch": 0.2731, + "grad_norm": 1.580456029151171, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 27310 + }, + { + "epoch": 0.27311, + "grad_norm": 1.0912144454747874, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 27311 + }, + { + "epoch": 0.27312, + "grad_norm": 1.60448997133867, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 27312 + }, + { + "epoch": 0.27313, + "grad_norm": 1.1658836790633778, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 27313 + }, + { + "epoch": 0.27314, + "grad_norm": 1.489817437131182, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 27314 + }, + { + "epoch": 0.27315, + "grad_norm": 1.4334362118617296, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 27315 + }, + { + "epoch": 0.27316, + "grad_norm": 1.273419960162539, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 27316 + }, + { + "epoch": 0.27317, + "grad_norm": 1.4119024514851788, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27317 + }, + { + "epoch": 0.27318, + "grad_norm": 1.3528722826684572, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27318 + }, + { + "epoch": 0.27319, + "grad_norm": 1.5213709363364403, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27319 + }, + { + "epoch": 0.2732, + "grad_norm": 1.3079293627303514, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27320 + }, + { + "epoch": 0.27321, + "grad_norm": 1.6253251704194658, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27321 + }, + { + "epoch": 0.27322, + "grad_norm": 1.5535612826021512, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 27322 + }, + { + "epoch": 0.27323, + "grad_norm": 1.355065370777839, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 27323 + }, + { + "epoch": 0.27324, + "grad_norm": 1.4855906197375466, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27324 + }, + { + "epoch": 0.27325, + "grad_norm": 1.4391881767818742, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 27325 + }, + { + "epoch": 0.27326, + "grad_norm": 1.3997008611498123, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 27326 + }, + { + "epoch": 0.27327, + "grad_norm": 1.3947752731683727, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 27327 + }, + { + "epoch": 0.27328, + "grad_norm": 1.378918358624421, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 27328 + }, + { + "epoch": 0.27329, + "grad_norm": 1.2688865390747435, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 27329 + }, + { + "epoch": 0.2733, + "grad_norm": 1.3177001749181665, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27330 + }, + { + "epoch": 0.27331, + "grad_norm": 1.329148710769667, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 27331 + }, + { + "epoch": 0.27332, + "grad_norm": 1.4200990386782435, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 27332 + }, + { + "epoch": 0.27333, + "grad_norm": 1.5183361510061426, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27333 + }, + { + "epoch": 0.27334, + "grad_norm": 1.391749904667538, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 27334 + }, + { + "epoch": 0.27335, + "grad_norm": 1.7924788367909557, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27335 + }, + { + "epoch": 0.27336, + "grad_norm": 1.2584321730360595, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 27336 + }, + { + "epoch": 0.27337, + "grad_norm": 1.4899868031093342, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 27337 + }, + { + "epoch": 0.27338, + "grad_norm": 1.3512376983366687, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27338 + }, + { + "epoch": 0.27339, + "grad_norm": 1.3730104465466368, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 27339 + }, + { + "epoch": 0.2734, + "grad_norm": 1.3721592290722002, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27340 + }, + { + "epoch": 0.27341, + "grad_norm": 1.4493081667768688, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 27341 + }, + { + "epoch": 0.27342, + "grad_norm": 1.3011538877832773, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 27342 + }, + { + "epoch": 0.27343, + "grad_norm": 1.5002244936229452, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 27343 + }, + { + "epoch": 0.27344, + "grad_norm": 1.3036183717543037, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 27344 + }, + { + "epoch": 0.27345, + "grad_norm": 1.2626422231097942, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 27345 + }, + { + "epoch": 0.27346, + "grad_norm": 1.5634450927529782, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 27346 + }, + { + "epoch": 0.27347, + "grad_norm": 1.4510458669918826, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 27347 + }, + { + "epoch": 0.27348, + "grad_norm": 1.65595686536117, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 27348 + }, + { + "epoch": 0.27349, + "grad_norm": 1.3411876177820887, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27349 + }, + { + "epoch": 0.2735, + "grad_norm": 1.5727496910865475, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 27350 + }, + { + "epoch": 0.27351, + "grad_norm": 1.206154633665334, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 27351 + }, + { + "epoch": 0.27352, + "grad_norm": 1.7554036633899717, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 27352 + }, + { + "epoch": 0.27353, + "grad_norm": 1.1273933593000114, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27353 + }, + { + "epoch": 0.27354, + "grad_norm": 1.6281809385218209, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 27354 + }, + { + "epoch": 0.27355, + "grad_norm": 1.445717878300674, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 27355 + }, + { + "epoch": 0.27356, + "grad_norm": 1.3458946331419783, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 27356 + }, + { + "epoch": 0.27357, + "grad_norm": 1.3927814902958244, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 27357 + }, + { + "epoch": 0.27358, + "grad_norm": 1.2203339449467665, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 27358 + }, + { + "epoch": 0.27359, + "grad_norm": 1.6656612136953715, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27359 + }, + { + "epoch": 0.2736, + "grad_norm": 1.1860098952227451, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27360 + }, + { + "epoch": 0.27361, + "grad_norm": 1.5909643499402832, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 27361 + }, + { + "epoch": 0.27362, + "grad_norm": 1.3341259905531284, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27362 + }, + { + "epoch": 0.27363, + "grad_norm": 1.4561954057899649, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 27363 + }, + { + "epoch": 0.27364, + "grad_norm": 1.343000039060517, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 27364 + }, + { + "epoch": 0.27365, + "grad_norm": 1.5173894765956049, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27365 + }, + { + "epoch": 0.27366, + "grad_norm": 1.235039765511595, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27366 + }, + { + "epoch": 0.27367, + "grad_norm": 1.763305175126938, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27367 + }, + { + "epoch": 0.27368, + "grad_norm": 1.1460290238666704, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27368 + }, + { + "epoch": 0.27369, + "grad_norm": 1.3731135804817358, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 27369 + }, + { + "epoch": 0.2737, + "grad_norm": 1.4552121586793954, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 27370 + }, + { + "epoch": 0.27371, + "grad_norm": 1.3016109744827051, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 27371 + }, + { + "epoch": 0.27372, + "grad_norm": 1.202836797979589, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 27372 + }, + { + "epoch": 0.27373, + "grad_norm": 1.456968336235871, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27373 + }, + { + "epoch": 0.27374, + "grad_norm": 1.5846594377379062, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27374 + }, + { + "epoch": 0.27375, + "grad_norm": 1.1494615811461593, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 27375 + }, + { + "epoch": 0.27376, + "grad_norm": 1.5225015047083212, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 27376 + }, + { + "epoch": 0.27377, + "grad_norm": 1.6085622597476303, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 27377 + }, + { + "epoch": 0.27378, + "grad_norm": 1.9860424968071888, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 27378 + }, + { + "epoch": 0.27379, + "grad_norm": 1.042711992006096, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 27379 + }, + { + "epoch": 0.2738, + "grad_norm": 1.4843959397946231, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 27380 + }, + { + "epoch": 0.27381, + "grad_norm": 1.3050843123518148, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27381 + }, + { + "epoch": 0.27382, + "grad_norm": 1.7094381364481162, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 27382 + }, + { + "epoch": 0.27383, + "grad_norm": 1.4232386620958601, + "learning_rate": 0.003, + "loss": 4.056, + "step": 27383 + }, + { + "epoch": 0.27384, + "grad_norm": 1.8055157068868775, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 27384 + }, + { + "epoch": 0.27385, + "grad_norm": 1.3860541597735365, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 27385 + }, + { + "epoch": 0.27386, + "grad_norm": 1.1740207486444005, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 27386 + }, + { + "epoch": 0.27387, + "grad_norm": 1.7608472610199402, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27387 + }, + { + "epoch": 0.27388, + "grad_norm": 1.1510193942411064, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 27388 + }, + { + "epoch": 0.27389, + "grad_norm": 1.578300798674232, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 27389 + }, + { + "epoch": 0.2739, + "grad_norm": 1.2939621089651496, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 27390 + }, + { + "epoch": 0.27391, + "grad_norm": 1.2683962957271686, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 27391 + }, + { + "epoch": 0.27392, + "grad_norm": 1.3963909978410545, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 27392 + }, + { + "epoch": 0.27393, + "grad_norm": 1.4707810759140567, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 27393 + }, + { + "epoch": 0.27394, + "grad_norm": 1.5926713552369909, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 27394 + }, + { + "epoch": 0.27395, + "grad_norm": 1.5731612535762785, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27395 + }, + { + "epoch": 0.27396, + "grad_norm": 1.3473084354425668, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27396 + }, + { + "epoch": 0.27397, + "grad_norm": 1.3370134323791147, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 27397 + }, + { + "epoch": 0.27398, + "grad_norm": 1.2617820812488503, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 27398 + }, + { + "epoch": 0.27399, + "grad_norm": 1.564847828652029, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 27399 + }, + { + "epoch": 0.274, + "grad_norm": 1.3906863106566616, + "learning_rate": 0.003, + "loss": 4.067, + "step": 27400 + }, + { + "epoch": 0.27401, + "grad_norm": 1.4121978939677584, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 27401 + }, + { + "epoch": 0.27402, + "grad_norm": 1.3396821909274215, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 27402 + }, + { + "epoch": 0.27403, + "grad_norm": 1.4514599175915988, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27403 + }, + { + "epoch": 0.27404, + "grad_norm": 1.6619655949474221, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 27404 + }, + { + "epoch": 0.27405, + "grad_norm": 1.2158841631257267, + "learning_rate": 0.003, + "loss": 4.028, + "step": 27405 + }, + { + "epoch": 0.27406, + "grad_norm": 1.5646777605170605, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 27406 + }, + { + "epoch": 0.27407, + "grad_norm": 1.3785054557977512, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 27407 + }, + { + "epoch": 0.27408, + "grad_norm": 1.4283755293785478, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 27408 + }, + { + "epoch": 0.27409, + "grad_norm": 1.3659664392832949, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27409 + }, + { + "epoch": 0.2741, + "grad_norm": 1.4475826286043763, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 27410 + }, + { + "epoch": 0.27411, + "grad_norm": 1.2427426470519394, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 27411 + }, + { + "epoch": 0.27412, + "grad_norm": 1.428633929810291, + "learning_rate": 0.003, + "loss": 4.033, + "step": 27412 + }, + { + "epoch": 0.27413, + "grad_norm": 1.6429880667022014, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 27413 + }, + { + "epoch": 0.27414, + "grad_norm": 1.7496051409067972, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 27414 + }, + { + "epoch": 0.27415, + "grad_norm": 1.1691952686842264, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 27415 + }, + { + "epoch": 0.27416, + "grad_norm": 1.3787532310446977, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 27416 + }, + { + "epoch": 0.27417, + "grad_norm": 1.6672827707747817, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27417 + }, + { + "epoch": 0.27418, + "grad_norm": 1.3189845901204067, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 27418 + }, + { + "epoch": 0.27419, + "grad_norm": 1.4293493547231173, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 27419 + }, + { + "epoch": 0.2742, + "grad_norm": 1.252411872048423, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27420 + }, + { + "epoch": 0.27421, + "grad_norm": 1.6303453062019482, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 27421 + }, + { + "epoch": 0.27422, + "grad_norm": 1.3529817232998518, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27422 + }, + { + "epoch": 0.27423, + "grad_norm": 1.6062110750482663, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 27423 + }, + { + "epoch": 0.27424, + "grad_norm": 1.1958448611588512, + "learning_rate": 0.003, + "loss": 4.068, + "step": 27424 + }, + { + "epoch": 0.27425, + "grad_norm": 1.681088527686211, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 27425 + }, + { + "epoch": 0.27426, + "grad_norm": 1.1862868894368535, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 27426 + }, + { + "epoch": 0.27427, + "grad_norm": 1.3501877756441167, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 27427 + }, + { + "epoch": 0.27428, + "grad_norm": 1.3390208093767064, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 27428 + }, + { + "epoch": 0.27429, + "grad_norm": 1.4955882914136998, + "learning_rate": 0.003, + "loss": 4.064, + "step": 27429 + }, + { + "epoch": 0.2743, + "grad_norm": 1.233449487222695, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 27430 + }, + { + "epoch": 0.27431, + "grad_norm": 1.8227428960247256, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 27431 + }, + { + "epoch": 0.27432, + "grad_norm": 1.0994748284564608, + "learning_rate": 0.003, + "loss": 4.042, + "step": 27432 + }, + { + "epoch": 0.27433, + "grad_norm": 1.697063049276606, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 27433 + }, + { + "epoch": 0.27434, + "grad_norm": 1.3517935750468684, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 27434 + }, + { + "epoch": 0.27435, + "grad_norm": 1.6587805744263135, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 27435 + }, + { + "epoch": 0.27436, + "grad_norm": 1.5105890819923737, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 27436 + }, + { + "epoch": 0.27437, + "grad_norm": 1.2458264493461038, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 27437 + }, + { + "epoch": 0.27438, + "grad_norm": 1.9552272127796062, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 27438 + }, + { + "epoch": 0.27439, + "grad_norm": 1.4445220367521716, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 27439 + }, + { + "epoch": 0.2744, + "grad_norm": 1.426391142272775, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 27440 + }, + { + "epoch": 0.27441, + "grad_norm": 1.4984822991766495, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27441 + }, + { + "epoch": 0.27442, + "grad_norm": 1.3728001271991361, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 27442 + }, + { + "epoch": 0.27443, + "grad_norm": 1.3387491726184995, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 27443 + }, + { + "epoch": 0.27444, + "grad_norm": 1.4191664425243913, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27444 + }, + { + "epoch": 0.27445, + "grad_norm": 1.3482622041129984, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27445 + }, + { + "epoch": 0.27446, + "grad_norm": 1.3703886041901459, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 27446 + }, + { + "epoch": 0.27447, + "grad_norm": 1.4141906820319516, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27447 + }, + { + "epoch": 0.27448, + "grad_norm": 1.4688404798033685, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 27448 + }, + { + "epoch": 0.27449, + "grad_norm": 1.4195589439348253, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27449 + }, + { + "epoch": 0.2745, + "grad_norm": 1.3851454103881238, + "learning_rate": 0.003, + "loss": 4.05, + "step": 27450 + }, + { + "epoch": 0.27451, + "grad_norm": 1.1817474459771076, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 27451 + }, + { + "epoch": 0.27452, + "grad_norm": 1.4107049905171176, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 27452 + }, + { + "epoch": 0.27453, + "grad_norm": 1.3868709850887786, + "learning_rate": 0.003, + "loss": 4.065, + "step": 27453 + }, + { + "epoch": 0.27454, + "grad_norm": 1.4100691346280503, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 27454 + }, + { + "epoch": 0.27455, + "grad_norm": 1.2106047875784571, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 27455 + }, + { + "epoch": 0.27456, + "grad_norm": 1.5777530549448768, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27456 + }, + { + "epoch": 0.27457, + "grad_norm": 1.2694885319854097, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 27457 + }, + { + "epoch": 0.27458, + "grad_norm": 1.6254937890858445, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 27458 + }, + { + "epoch": 0.27459, + "grad_norm": 1.5019759168401845, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 27459 + }, + { + "epoch": 0.2746, + "grad_norm": 1.2666420549248483, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27460 + }, + { + "epoch": 0.27461, + "grad_norm": 1.4527943087124362, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27461 + }, + { + "epoch": 0.27462, + "grad_norm": 1.2681623700533164, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27462 + }, + { + "epoch": 0.27463, + "grad_norm": 1.6001670316386283, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27463 + }, + { + "epoch": 0.27464, + "grad_norm": 1.20225513703101, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27464 + }, + { + "epoch": 0.27465, + "grad_norm": 1.376385995076624, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 27465 + }, + { + "epoch": 0.27466, + "grad_norm": 1.289648313694689, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 27466 + }, + { + "epoch": 0.27467, + "grad_norm": 1.8586215118053004, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27467 + }, + { + "epoch": 0.27468, + "grad_norm": 1.321656433352933, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 27468 + }, + { + "epoch": 0.27469, + "grad_norm": 1.4788367853934283, + "learning_rate": 0.003, + "loss": 4.06, + "step": 27469 + }, + { + "epoch": 0.2747, + "grad_norm": 1.2427744144109085, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 27470 + }, + { + "epoch": 0.27471, + "grad_norm": 1.7548396752489166, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 27471 + }, + { + "epoch": 0.27472, + "grad_norm": 1.2377937587058196, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 27472 + }, + { + "epoch": 0.27473, + "grad_norm": 1.769448026603225, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27473 + }, + { + "epoch": 0.27474, + "grad_norm": 1.0314470883051146, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 27474 + }, + { + "epoch": 0.27475, + "grad_norm": 1.537587245460602, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27475 + }, + { + "epoch": 0.27476, + "grad_norm": 1.539490480954806, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 27476 + }, + { + "epoch": 0.27477, + "grad_norm": 1.4966354149031873, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 27477 + }, + { + "epoch": 0.27478, + "grad_norm": 1.165963644247902, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 27478 + }, + { + "epoch": 0.27479, + "grad_norm": 1.472174031704091, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 27479 + }, + { + "epoch": 0.2748, + "grad_norm": 1.2391346417319085, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 27480 + }, + { + "epoch": 0.27481, + "grad_norm": 1.537265358501281, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 27481 + }, + { + "epoch": 0.27482, + "grad_norm": 1.2491536320359868, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 27482 + }, + { + "epoch": 0.27483, + "grad_norm": 1.632956351844012, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 27483 + }, + { + "epoch": 0.27484, + "grad_norm": 1.3852569091394433, + "learning_rate": 0.003, + "loss": 4.048, + "step": 27484 + }, + { + "epoch": 0.27485, + "grad_norm": 1.7089335621728075, + "learning_rate": 0.003, + "loss": 4.061, + "step": 27485 + }, + { + "epoch": 0.27486, + "grad_norm": 1.3322594623471742, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 27486 + }, + { + "epoch": 0.27487, + "grad_norm": 1.5552532758058148, + "learning_rate": 0.003, + "loss": 4.057, + "step": 27487 + }, + { + "epoch": 0.27488, + "grad_norm": 1.640917843256971, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 27488 + }, + { + "epoch": 0.27489, + "grad_norm": 1.2351444064271402, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27489 + }, + { + "epoch": 0.2749, + "grad_norm": 1.404112295456891, + "learning_rate": 0.003, + "loss": 4.035, + "step": 27490 + }, + { + "epoch": 0.27491, + "grad_norm": 1.1793546517146904, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 27491 + }, + { + "epoch": 0.27492, + "grad_norm": 1.5487232096345758, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 27492 + }, + { + "epoch": 0.27493, + "grad_norm": 1.2256826590606213, + "learning_rate": 0.003, + "loss": 4.054, + "step": 27493 + }, + { + "epoch": 0.27494, + "grad_norm": 1.8187867119297587, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27494 + }, + { + "epoch": 0.27495, + "grad_norm": 1.4742917051606674, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27495 + }, + { + "epoch": 0.27496, + "grad_norm": 1.3789173385327724, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27496 + }, + { + "epoch": 0.27497, + "grad_norm": 1.362378430910307, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27497 + }, + { + "epoch": 0.27498, + "grad_norm": 1.532366134665585, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27498 + }, + { + "epoch": 0.27499, + "grad_norm": 1.177110290399895, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 27499 + }, + { + "epoch": 0.275, + "grad_norm": 1.3948762271912836, + "learning_rate": 0.003, + "loss": 4.032, + "step": 27500 + }, + { + "epoch": 0.27501, + "grad_norm": 1.6546020904326997, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 27501 + }, + { + "epoch": 0.27502, + "grad_norm": 1.1764059153690418, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27502 + }, + { + "epoch": 0.27503, + "grad_norm": 1.4253347319732779, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 27503 + }, + { + "epoch": 0.27504, + "grad_norm": 1.437783975382049, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 27504 + }, + { + "epoch": 0.27505, + "grad_norm": 1.3286430693689446, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 27505 + }, + { + "epoch": 0.27506, + "grad_norm": 1.2446692587534691, + "learning_rate": 0.003, + "loss": 4.052, + "step": 27506 + }, + { + "epoch": 0.27507, + "grad_norm": 1.5531804006497383, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 27507 + }, + { + "epoch": 0.27508, + "grad_norm": 1.1481308833285804, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 27508 + }, + { + "epoch": 0.27509, + "grad_norm": 1.613417513159795, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 27509 + }, + { + "epoch": 0.2751, + "grad_norm": 1.4588839689813127, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 27510 + }, + { + "epoch": 0.27511, + "grad_norm": 1.3402313245605353, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 27511 + }, + { + "epoch": 0.27512, + "grad_norm": 1.7049346506897969, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27512 + }, + { + "epoch": 0.27513, + "grad_norm": 1.5312181132011853, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 27513 + }, + { + "epoch": 0.27514, + "grad_norm": 1.7972094771205123, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27514 + }, + { + "epoch": 0.27515, + "grad_norm": 1.3303303250765033, + "learning_rate": 0.003, + "loss": 4.032, + "step": 27515 + }, + { + "epoch": 0.27516, + "grad_norm": 1.5253781528493677, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 27516 + }, + { + "epoch": 0.27517, + "grad_norm": 1.5033191410220255, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27517 + }, + { + "epoch": 0.27518, + "grad_norm": 1.2076355461848696, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 27518 + }, + { + "epoch": 0.27519, + "grad_norm": 1.6301062257270231, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27519 + }, + { + "epoch": 0.2752, + "grad_norm": 1.0833711586160946, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 27520 + }, + { + "epoch": 0.27521, + "grad_norm": 1.7119347416144248, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 27521 + }, + { + "epoch": 0.27522, + "grad_norm": 1.265114308772316, + "learning_rate": 0.003, + "loss": 4.054, + "step": 27522 + }, + { + "epoch": 0.27523, + "grad_norm": 1.7303878373488404, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 27523 + }, + { + "epoch": 0.27524, + "grad_norm": 1.28122315662023, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 27524 + }, + { + "epoch": 0.27525, + "grad_norm": 1.3335766547366914, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 27525 + }, + { + "epoch": 0.27526, + "grad_norm": 1.4974287261180717, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 27526 + }, + { + "epoch": 0.27527, + "grad_norm": 1.2837661511383636, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 27527 + }, + { + "epoch": 0.27528, + "grad_norm": 1.3971020609400309, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27528 + }, + { + "epoch": 0.27529, + "grad_norm": 1.2950092715337167, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 27529 + }, + { + "epoch": 0.2753, + "grad_norm": 1.6568106275559724, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 27530 + }, + { + "epoch": 0.27531, + "grad_norm": 1.4575567317630211, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 27531 + }, + { + "epoch": 0.27532, + "grad_norm": 1.3441520713818964, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 27532 + }, + { + "epoch": 0.27533, + "grad_norm": 1.5497928550785547, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 27533 + }, + { + "epoch": 0.27534, + "grad_norm": 1.2572865167058436, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 27534 + }, + { + "epoch": 0.27535, + "grad_norm": 1.6517805355116824, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 27535 + }, + { + "epoch": 0.27536, + "grad_norm": 1.1205310624929177, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 27536 + }, + { + "epoch": 0.27537, + "grad_norm": 1.5603711319746172, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 27537 + }, + { + "epoch": 0.27538, + "grad_norm": 1.2027387939832366, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 27538 + }, + { + "epoch": 0.27539, + "grad_norm": 1.339843570879642, + "learning_rate": 0.003, + "loss": 4.044, + "step": 27539 + }, + { + "epoch": 0.2754, + "grad_norm": 1.370802715550018, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27540 + }, + { + "epoch": 0.27541, + "grad_norm": 1.459981228984006, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27541 + }, + { + "epoch": 0.27542, + "grad_norm": 1.8172828744969483, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 27542 + }, + { + "epoch": 0.27543, + "grad_norm": 1.1900425635350445, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27543 + }, + { + "epoch": 0.27544, + "grad_norm": 1.5236772984229334, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27544 + }, + { + "epoch": 0.27545, + "grad_norm": 1.5834338497099394, + "learning_rate": 0.003, + "loss": 4.026, + "step": 27545 + }, + { + "epoch": 0.27546, + "grad_norm": 1.4381728611672735, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 27546 + }, + { + "epoch": 0.27547, + "grad_norm": 1.307262915818506, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27547 + }, + { + "epoch": 0.27548, + "grad_norm": 1.5497164690159941, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27548 + }, + { + "epoch": 0.27549, + "grad_norm": 1.2064092701980635, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27549 + }, + { + "epoch": 0.2755, + "grad_norm": 1.3855063035832833, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27550 + }, + { + "epoch": 0.27551, + "grad_norm": 1.3487008318432208, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 27551 + }, + { + "epoch": 0.27552, + "grad_norm": 1.4740738109850313, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 27552 + }, + { + "epoch": 0.27553, + "grad_norm": 1.5677403607957054, + "learning_rate": 0.003, + "loss": 4.017, + "step": 27553 + }, + { + "epoch": 0.27554, + "grad_norm": 1.337905832370948, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 27554 + }, + { + "epoch": 0.27555, + "grad_norm": 1.453154537735689, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 27555 + }, + { + "epoch": 0.27556, + "grad_norm": 1.3626759262113541, + "learning_rate": 0.003, + "loss": 4.034, + "step": 27556 + }, + { + "epoch": 0.27557, + "grad_norm": 1.4960302212594259, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 27557 + }, + { + "epoch": 0.27558, + "grad_norm": 1.3044656045932743, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 27558 + }, + { + "epoch": 0.27559, + "grad_norm": 1.5267167870422722, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 27559 + }, + { + "epoch": 0.2756, + "grad_norm": 1.2583000277611296, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 27560 + }, + { + "epoch": 0.27561, + "grad_norm": 1.577054617636473, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 27561 + }, + { + "epoch": 0.27562, + "grad_norm": 1.3835956668379812, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 27562 + }, + { + "epoch": 0.27563, + "grad_norm": 1.2948996580811576, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 27563 + }, + { + "epoch": 0.27564, + "grad_norm": 1.4500999968278394, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 27564 + }, + { + "epoch": 0.27565, + "grad_norm": 1.2478420008710143, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 27565 + }, + { + "epoch": 0.27566, + "grad_norm": 1.3611818591525588, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 27566 + }, + { + "epoch": 0.27567, + "grad_norm": 1.4027150129116284, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 27567 + }, + { + "epoch": 0.27568, + "grad_norm": 1.506447599589195, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27568 + }, + { + "epoch": 0.27569, + "grad_norm": 1.4912910772849397, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27569 + }, + { + "epoch": 0.2757, + "grad_norm": 1.3145970207462314, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 27570 + }, + { + "epoch": 0.27571, + "grad_norm": 1.5768047942883323, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27571 + }, + { + "epoch": 0.27572, + "grad_norm": 1.6331469235833393, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 27572 + }, + { + "epoch": 0.27573, + "grad_norm": 1.119211759839972, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 27573 + }, + { + "epoch": 0.27574, + "grad_norm": 1.6866961145947506, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 27574 + }, + { + "epoch": 0.27575, + "grad_norm": 1.3007201357575238, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27575 + }, + { + "epoch": 0.27576, + "grad_norm": 1.7696309798960612, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 27576 + }, + { + "epoch": 0.27577, + "grad_norm": 1.1362719699281025, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27577 + }, + { + "epoch": 0.27578, + "grad_norm": 1.4281784742792762, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27578 + }, + { + "epoch": 0.27579, + "grad_norm": 1.5417452248531953, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 27579 + }, + { + "epoch": 0.2758, + "grad_norm": 1.517979796606554, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 27580 + }, + { + "epoch": 0.27581, + "grad_norm": 1.3546947764540616, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 27581 + }, + { + "epoch": 0.27582, + "grad_norm": 1.4403156471986407, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 27582 + }, + { + "epoch": 0.27583, + "grad_norm": 1.3077158746226591, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 27583 + }, + { + "epoch": 0.27584, + "grad_norm": 1.5806526885828671, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 27584 + }, + { + "epoch": 0.27585, + "grad_norm": 1.3828224515694316, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 27585 + }, + { + "epoch": 0.27586, + "grad_norm": 1.449882695085182, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27586 + }, + { + "epoch": 0.27587, + "grad_norm": 1.3419193034676502, + "learning_rate": 0.003, + "loss": 4.045, + "step": 27587 + }, + { + "epoch": 0.27588, + "grad_norm": 1.751247428713455, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 27588 + }, + { + "epoch": 0.27589, + "grad_norm": 1.4273264096998195, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 27589 + }, + { + "epoch": 0.2759, + "grad_norm": 1.5351169931671504, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 27590 + }, + { + "epoch": 0.27591, + "grad_norm": 1.3798939400854549, + "learning_rate": 0.003, + "loss": 4.063, + "step": 27591 + }, + { + "epoch": 0.27592, + "grad_norm": 1.4686876830186417, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 27592 + }, + { + "epoch": 0.27593, + "grad_norm": 1.3766453986709053, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 27593 + }, + { + "epoch": 0.27594, + "grad_norm": 1.5825784340924116, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 27594 + }, + { + "epoch": 0.27595, + "grad_norm": 1.287224790993353, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 27595 + }, + { + "epoch": 0.27596, + "grad_norm": 1.5335425791185153, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 27596 + }, + { + "epoch": 0.27597, + "grad_norm": 1.4310966567451835, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 27597 + }, + { + "epoch": 0.27598, + "grad_norm": 1.3524971788328082, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 27598 + }, + { + "epoch": 0.27599, + "grad_norm": 1.153660799068126, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 27599 + }, + { + "epoch": 0.276, + "grad_norm": 1.4732737572913575, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 27600 + }, + { + "epoch": 0.27601, + "grad_norm": 1.3597925512952205, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 27601 + }, + { + "epoch": 0.27602, + "grad_norm": 1.599170916451973, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 27602 + }, + { + "epoch": 0.27603, + "grad_norm": 1.5319048411058651, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 27603 + }, + { + "epoch": 0.27604, + "grad_norm": 1.3632531365053602, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27604 + }, + { + "epoch": 0.27605, + "grad_norm": 1.7280208944039224, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 27605 + }, + { + "epoch": 0.27606, + "grad_norm": 1.3070696446469252, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27606 + }, + { + "epoch": 0.27607, + "grad_norm": 1.4097092265149347, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 27607 + }, + { + "epoch": 0.27608, + "grad_norm": 1.732382800420744, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 27608 + }, + { + "epoch": 0.27609, + "grad_norm": 1.1993289862997278, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27609 + }, + { + "epoch": 0.2761, + "grad_norm": 1.4544338084888853, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27610 + }, + { + "epoch": 0.27611, + "grad_norm": 1.6133433498243652, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 27611 + }, + { + "epoch": 0.27612, + "grad_norm": 1.4058179441316454, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 27612 + }, + { + "epoch": 0.27613, + "grad_norm": 1.449696772505723, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27613 + }, + { + "epoch": 0.27614, + "grad_norm": 1.5666001905046332, + "learning_rate": 0.003, + "loss": 4.062, + "step": 27614 + }, + { + "epoch": 0.27615, + "grad_norm": 1.2456815926955982, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27615 + }, + { + "epoch": 0.27616, + "grad_norm": 1.5066842633212278, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 27616 + }, + { + "epoch": 0.27617, + "grad_norm": 1.2012396149975102, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 27617 + }, + { + "epoch": 0.27618, + "grad_norm": 1.7283392183920614, + "learning_rate": 0.003, + "loss": 4.06, + "step": 27618 + }, + { + "epoch": 0.27619, + "grad_norm": 1.3295013315876978, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 27619 + }, + { + "epoch": 0.2762, + "grad_norm": 1.7362217519139214, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27620 + }, + { + "epoch": 0.27621, + "grad_norm": 1.2986221835640646, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27621 + }, + { + "epoch": 0.27622, + "grad_norm": 1.2820460817612338, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 27622 + }, + { + "epoch": 0.27623, + "grad_norm": 1.2781798048940032, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 27623 + }, + { + "epoch": 0.27624, + "grad_norm": 1.6089428958388294, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 27624 + }, + { + "epoch": 0.27625, + "grad_norm": 1.30948295516169, + "learning_rate": 0.003, + "loss": 4.03, + "step": 27625 + }, + { + "epoch": 0.27626, + "grad_norm": 1.375885646099852, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 27626 + }, + { + "epoch": 0.27627, + "grad_norm": 1.4683326276319324, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 27627 + }, + { + "epoch": 0.27628, + "grad_norm": 1.3225993706967563, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 27628 + }, + { + "epoch": 0.27629, + "grad_norm": 1.3049429047791232, + "learning_rate": 0.003, + "loss": 4.073, + "step": 27629 + }, + { + "epoch": 0.2763, + "grad_norm": 1.5420114646451113, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 27630 + }, + { + "epoch": 0.27631, + "grad_norm": 1.5288171935848882, + "learning_rate": 0.003, + "loss": 4.068, + "step": 27631 + }, + { + "epoch": 0.27632, + "grad_norm": 1.3520861227376026, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 27632 + }, + { + "epoch": 0.27633, + "grad_norm": 1.488529949936612, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 27633 + }, + { + "epoch": 0.27634, + "grad_norm": 1.3037525365964733, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 27634 + }, + { + "epoch": 0.27635, + "grad_norm": 1.415402797613495, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 27635 + }, + { + "epoch": 0.27636, + "grad_norm": 1.199954383031244, + "learning_rate": 0.003, + "loss": 4.047, + "step": 27636 + }, + { + "epoch": 0.27637, + "grad_norm": 1.7504273332635194, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 27637 + }, + { + "epoch": 0.27638, + "grad_norm": 1.1365688236773284, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 27638 + }, + { + "epoch": 0.27639, + "grad_norm": 1.7959301405673336, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 27639 + }, + { + "epoch": 0.2764, + "grad_norm": 1.2578194191862053, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 27640 + }, + { + "epoch": 0.27641, + "grad_norm": 1.5527701615850065, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 27641 + }, + { + "epoch": 0.27642, + "grad_norm": 1.510483029038351, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 27642 + }, + { + "epoch": 0.27643, + "grad_norm": 1.3434581092953704, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27643 + }, + { + "epoch": 0.27644, + "grad_norm": 1.4092816523577727, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27644 + }, + { + "epoch": 0.27645, + "grad_norm": 1.6106634868173944, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 27645 + }, + { + "epoch": 0.27646, + "grad_norm": 1.244362798094309, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 27646 + }, + { + "epoch": 0.27647, + "grad_norm": 1.432384242556579, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27647 + }, + { + "epoch": 0.27648, + "grad_norm": 1.4759337092255405, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 27648 + }, + { + "epoch": 0.27649, + "grad_norm": 1.3388567128727689, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 27649 + }, + { + "epoch": 0.2765, + "grad_norm": 1.5069317511779643, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27650 + }, + { + "epoch": 0.27651, + "grad_norm": 1.5262243739513548, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 27651 + }, + { + "epoch": 0.27652, + "grad_norm": 1.1936632670498435, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 27652 + }, + { + "epoch": 0.27653, + "grad_norm": 1.7979502192044021, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 27653 + }, + { + "epoch": 0.27654, + "grad_norm": 1.4753419270244492, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 27654 + }, + { + "epoch": 0.27655, + "grad_norm": 1.4385895891684404, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 27655 + }, + { + "epoch": 0.27656, + "grad_norm": 1.6031871871548589, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 27656 + }, + { + "epoch": 0.27657, + "grad_norm": 1.694703453320441, + "learning_rate": 0.003, + "loss": 4.044, + "step": 27657 + }, + { + "epoch": 0.27658, + "grad_norm": 1.1606393641193236, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27658 + }, + { + "epoch": 0.27659, + "grad_norm": 1.6281321120122683, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 27659 + }, + { + "epoch": 0.2766, + "grad_norm": 1.5662026403222151, + "learning_rate": 0.003, + "loss": 4.049, + "step": 27660 + }, + { + "epoch": 0.27661, + "grad_norm": 1.2060888396272442, + "learning_rate": 0.003, + "loss": 4.061, + "step": 27661 + }, + { + "epoch": 0.27662, + "grad_norm": 1.47847564952024, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 27662 + }, + { + "epoch": 0.27663, + "grad_norm": 1.3802084982765006, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 27663 + }, + { + "epoch": 0.27664, + "grad_norm": 1.1904576153948796, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 27664 + }, + { + "epoch": 0.27665, + "grad_norm": 1.7991837131969544, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 27665 + }, + { + "epoch": 0.27666, + "grad_norm": 1.2018632069409954, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 27666 + }, + { + "epoch": 0.27667, + "grad_norm": 1.8043670060853298, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27667 + }, + { + "epoch": 0.27668, + "grad_norm": 1.547361084232157, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27668 + }, + { + "epoch": 0.27669, + "grad_norm": 1.3303784896887147, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 27669 + }, + { + "epoch": 0.2767, + "grad_norm": 1.541476516086168, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 27670 + }, + { + "epoch": 0.27671, + "grad_norm": 1.2380687261046037, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 27671 + }, + { + "epoch": 0.27672, + "grad_norm": 1.6421378788258285, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 27672 + }, + { + "epoch": 0.27673, + "grad_norm": 1.0794389322052504, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 27673 + }, + { + "epoch": 0.27674, + "grad_norm": 1.5675979682120027, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 27674 + }, + { + "epoch": 0.27675, + "grad_norm": 1.0966925988354215, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 27675 + }, + { + "epoch": 0.27676, + "grad_norm": 1.5183358924416182, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 27676 + }, + { + "epoch": 0.27677, + "grad_norm": 1.3583624326264176, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 27677 + }, + { + "epoch": 0.27678, + "grad_norm": 1.444446710172524, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 27678 + }, + { + "epoch": 0.27679, + "grad_norm": 1.5852250176181413, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 27679 + }, + { + "epoch": 0.2768, + "grad_norm": 1.2176443732976026, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 27680 + }, + { + "epoch": 0.27681, + "grad_norm": 1.6831684087050773, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27681 + }, + { + "epoch": 0.27682, + "grad_norm": 1.4363564900224683, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27682 + }, + { + "epoch": 0.27683, + "grad_norm": 1.3870232741654926, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27683 + }, + { + "epoch": 0.27684, + "grad_norm": 1.3297917475616683, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27684 + }, + { + "epoch": 0.27685, + "grad_norm": 1.504038268410943, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 27685 + }, + { + "epoch": 0.27686, + "grad_norm": 1.2979401778343098, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27686 + }, + { + "epoch": 0.27687, + "grad_norm": 1.4315343157509444, + "learning_rate": 0.003, + "loss": 4.031, + "step": 27687 + }, + { + "epoch": 0.27688, + "grad_norm": 1.277866953379071, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27688 + }, + { + "epoch": 0.27689, + "grad_norm": 1.352814397889457, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 27689 + }, + { + "epoch": 0.2769, + "grad_norm": 1.4138884416323338, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 27690 + }, + { + "epoch": 0.27691, + "grad_norm": 1.5876047887288274, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 27691 + }, + { + "epoch": 0.27692, + "grad_norm": 1.628992969565625, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 27692 + }, + { + "epoch": 0.27693, + "grad_norm": 1.3034667872181571, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27693 + }, + { + "epoch": 0.27694, + "grad_norm": 1.397497782867376, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27694 + }, + { + "epoch": 0.27695, + "grad_norm": 1.2483133604895729, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27695 + }, + { + "epoch": 0.27696, + "grad_norm": 1.602559142568136, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 27696 + }, + { + "epoch": 0.27697, + "grad_norm": 1.2029154672854405, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 27697 + }, + { + "epoch": 0.27698, + "grad_norm": 1.4330864474669485, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27698 + }, + { + "epoch": 0.27699, + "grad_norm": 1.373276890699816, + "learning_rate": 0.003, + "loss": 4.038, + "step": 27699 + }, + { + "epoch": 0.277, + "grad_norm": 1.2580157244871588, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27700 + }, + { + "epoch": 0.27701, + "grad_norm": 1.7978816997501963, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 27701 + }, + { + "epoch": 0.27702, + "grad_norm": 1.437765548990988, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 27702 + }, + { + "epoch": 0.27703, + "grad_norm": 1.6032881254006637, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 27703 + }, + { + "epoch": 0.27704, + "grad_norm": 1.429246327185285, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 27704 + }, + { + "epoch": 0.27705, + "grad_norm": 1.2075152914814555, + "learning_rate": 0.003, + "loss": 4.069, + "step": 27705 + }, + { + "epoch": 0.27706, + "grad_norm": 1.330194694548333, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27706 + }, + { + "epoch": 0.27707, + "grad_norm": 1.6327016591249688, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 27707 + }, + { + "epoch": 0.27708, + "grad_norm": 1.4289816356087275, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 27708 + }, + { + "epoch": 0.27709, + "grad_norm": 1.4888650859988872, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 27709 + }, + { + "epoch": 0.2771, + "grad_norm": 1.2478132337651335, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 27710 + }, + { + "epoch": 0.27711, + "grad_norm": 1.2797969147633683, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 27711 + }, + { + "epoch": 0.27712, + "grad_norm": 1.3224288677058833, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 27712 + }, + { + "epoch": 0.27713, + "grad_norm": 1.575341778505956, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 27713 + }, + { + "epoch": 0.27714, + "grad_norm": 1.368676949535633, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 27714 + }, + { + "epoch": 0.27715, + "grad_norm": 1.4462488758986867, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 27715 + }, + { + "epoch": 0.27716, + "grad_norm": 1.3994441160395428, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 27716 + }, + { + "epoch": 0.27717, + "grad_norm": 1.2623236954356534, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27717 + }, + { + "epoch": 0.27718, + "grad_norm": 1.4944290730547323, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 27718 + }, + { + "epoch": 0.27719, + "grad_norm": 1.5902289660241944, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 27719 + }, + { + "epoch": 0.2772, + "grad_norm": 1.2426233926193357, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 27720 + }, + { + "epoch": 0.27721, + "grad_norm": 1.6021372052734841, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 27721 + }, + { + "epoch": 0.27722, + "grad_norm": 1.1639820419057634, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 27722 + }, + { + "epoch": 0.27723, + "grad_norm": 1.4851707185014023, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 27723 + }, + { + "epoch": 0.27724, + "grad_norm": 1.3784072032628707, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27724 + }, + { + "epoch": 0.27725, + "grad_norm": 1.4950771540656735, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 27725 + }, + { + "epoch": 0.27726, + "grad_norm": 1.3448365360948729, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 27726 + }, + { + "epoch": 0.27727, + "grad_norm": 1.5130680874913072, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 27727 + }, + { + "epoch": 0.27728, + "grad_norm": 1.4488949968007285, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 27728 + }, + { + "epoch": 0.27729, + "grad_norm": 1.4872807992172274, + "learning_rate": 0.003, + "loss": 4.072, + "step": 27729 + }, + { + "epoch": 0.2773, + "grad_norm": 1.5599257353598148, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 27730 + }, + { + "epoch": 0.27731, + "grad_norm": 1.53934408484739, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 27731 + }, + { + "epoch": 0.27732, + "grad_norm": 1.4642872817394113, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 27732 + }, + { + "epoch": 0.27733, + "grad_norm": 1.3641277972461643, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 27733 + }, + { + "epoch": 0.27734, + "grad_norm": 1.341239437394763, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 27734 + }, + { + "epoch": 0.27735, + "grad_norm": 1.3816594641981501, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 27735 + }, + { + "epoch": 0.27736, + "grad_norm": 1.3666533870138122, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27736 + }, + { + "epoch": 0.27737, + "grad_norm": 1.4839349686375747, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 27737 + }, + { + "epoch": 0.27738, + "grad_norm": 1.3772021745362193, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 27738 + }, + { + "epoch": 0.27739, + "grad_norm": 1.4683687836876431, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 27739 + }, + { + "epoch": 0.2774, + "grad_norm": 1.2674508684027703, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 27740 + }, + { + "epoch": 0.27741, + "grad_norm": 1.6124090455850275, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 27741 + }, + { + "epoch": 0.27742, + "grad_norm": 1.5629326983889664, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 27742 + }, + { + "epoch": 0.27743, + "grad_norm": 1.6725224421199354, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27743 + }, + { + "epoch": 0.27744, + "grad_norm": 1.1980973105736286, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 27744 + }, + { + "epoch": 0.27745, + "grad_norm": 1.678462222663154, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 27745 + }, + { + "epoch": 0.27746, + "grad_norm": 1.1288427012547761, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 27746 + }, + { + "epoch": 0.27747, + "grad_norm": 1.8141197389439345, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 27747 + }, + { + "epoch": 0.27748, + "grad_norm": 1.1785286112904438, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 27748 + }, + { + "epoch": 0.27749, + "grad_norm": 1.4344839161438765, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 27749 + }, + { + "epoch": 0.2775, + "grad_norm": 1.8239116562003315, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 27750 + }, + { + "epoch": 0.27751, + "grad_norm": 1.5119610556044438, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 27751 + }, + { + "epoch": 0.27752, + "grad_norm": 1.3113847472027882, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 27752 + }, + { + "epoch": 0.27753, + "grad_norm": 1.5625368670505968, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 27753 + }, + { + "epoch": 0.27754, + "grad_norm": 1.2731289408415485, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 27754 + }, + { + "epoch": 0.27755, + "grad_norm": 1.4601926832163845, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27755 + }, + { + "epoch": 0.27756, + "grad_norm": 1.345047104987338, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 27756 + }, + { + "epoch": 0.27757, + "grad_norm": 1.2720084971735814, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 27757 + }, + { + "epoch": 0.27758, + "grad_norm": 1.5089481678213499, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 27758 + }, + { + "epoch": 0.27759, + "grad_norm": 1.494589862189271, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 27759 + }, + { + "epoch": 0.2776, + "grad_norm": 1.6838309288971673, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 27760 + }, + { + "epoch": 0.27761, + "grad_norm": 1.354267001744471, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 27761 + }, + { + "epoch": 0.27762, + "grad_norm": 1.666452453691561, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 27762 + }, + { + "epoch": 0.27763, + "grad_norm": 1.558569672347428, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 27763 + }, + { + "epoch": 0.27764, + "grad_norm": 1.053274083598646, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 27764 + }, + { + "epoch": 0.27765, + "grad_norm": 1.7388403344624992, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 27765 + }, + { + "epoch": 0.27766, + "grad_norm": 1.154645228469187, + "learning_rate": 0.003, + "loss": 4.026, + "step": 27766 + }, + { + "epoch": 0.27767, + "grad_norm": 1.3080865200520975, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 27767 + }, + { + "epoch": 0.27768, + "grad_norm": 1.4077369736851504, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27768 + }, + { + "epoch": 0.27769, + "grad_norm": 1.7300128198170752, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 27769 + }, + { + "epoch": 0.2777, + "grad_norm": 1.2736234776268915, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 27770 + }, + { + "epoch": 0.27771, + "grad_norm": 1.520910302119796, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 27771 + }, + { + "epoch": 0.27772, + "grad_norm": 1.2558059046156278, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 27772 + }, + { + "epoch": 0.27773, + "grad_norm": 1.498153703170163, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 27773 + }, + { + "epoch": 0.27774, + "grad_norm": 1.351263810929352, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 27774 + }, + { + "epoch": 0.27775, + "grad_norm": 1.5009284683649284, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 27775 + }, + { + "epoch": 0.27776, + "grad_norm": 1.3920240877638417, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 27776 + }, + { + "epoch": 0.27777, + "grad_norm": 1.612770951059502, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 27777 + }, + { + "epoch": 0.27778, + "grad_norm": 1.627333052280118, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 27778 + }, + { + "epoch": 0.27779, + "grad_norm": 1.5407454522884396, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 27779 + }, + { + "epoch": 0.2778, + "grad_norm": 1.3618274392137104, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 27780 + }, + { + "epoch": 0.27781, + "grad_norm": 1.1941631038011065, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 27781 + }, + { + "epoch": 0.27782, + "grad_norm": 1.609680824013597, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 27782 + }, + { + "epoch": 0.27783, + "grad_norm": 1.5277073332098083, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 27783 + }, + { + "epoch": 0.27784, + "grad_norm": 1.1654421179087928, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 27784 + }, + { + "epoch": 0.27785, + "grad_norm": 1.4915820397346455, + "learning_rate": 0.003, + "loss": 4.056, + "step": 27785 + }, + { + "epoch": 0.27786, + "grad_norm": 1.2977385097066843, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 27786 + }, + { + "epoch": 0.27787, + "grad_norm": 1.5970239181414148, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 27787 + }, + { + "epoch": 0.27788, + "grad_norm": 1.257764722560255, + "learning_rate": 0.003, + "loss": 4.062, + "step": 27788 + }, + { + "epoch": 0.27789, + "grad_norm": 1.5029065161648847, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 27789 + }, + { + "epoch": 0.2779, + "grad_norm": 1.3448247220327745, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 27790 + }, + { + "epoch": 0.27791, + "grad_norm": 1.6044102561604823, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 27791 + }, + { + "epoch": 0.27792, + "grad_norm": 1.2700982924839561, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 27792 + }, + { + "epoch": 0.27793, + "grad_norm": 1.6233295576243167, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 27793 + }, + { + "epoch": 0.27794, + "grad_norm": 1.2428364790143314, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 27794 + }, + { + "epoch": 0.27795, + "grad_norm": 1.2145679770788553, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27795 + }, + { + "epoch": 0.27796, + "grad_norm": 1.7008769917119289, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 27796 + }, + { + "epoch": 0.27797, + "grad_norm": 0.993159775952936, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 27797 + }, + { + "epoch": 0.27798, + "grad_norm": 1.7697873576960674, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 27798 + }, + { + "epoch": 0.27799, + "grad_norm": 1.283310632205693, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 27799 + }, + { + "epoch": 0.278, + "grad_norm": 1.4902295421070038, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27800 + }, + { + "epoch": 0.27801, + "grad_norm": 1.4379539777068495, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 27801 + }, + { + "epoch": 0.27802, + "grad_norm": 1.573139981129907, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 27802 + }, + { + "epoch": 0.27803, + "grad_norm": 1.412600636097537, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 27803 + }, + { + "epoch": 0.27804, + "grad_norm": 1.6090663882194722, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 27804 + }, + { + "epoch": 0.27805, + "grad_norm": 1.4225757867701376, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 27805 + }, + { + "epoch": 0.27806, + "grad_norm": 1.439998829108452, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 27806 + }, + { + "epoch": 0.27807, + "grad_norm": 1.4290468400870904, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 27807 + }, + { + "epoch": 0.27808, + "grad_norm": 1.4307673036439685, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 27808 + }, + { + "epoch": 0.27809, + "grad_norm": 1.3807653186999678, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 27809 + }, + { + "epoch": 0.2781, + "grad_norm": 1.7096469276233144, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 27810 + }, + { + "epoch": 0.27811, + "grad_norm": 1.443818371970015, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 27811 + }, + { + "epoch": 0.27812, + "grad_norm": 1.2990286579488526, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 27812 + }, + { + "epoch": 0.27813, + "grad_norm": 1.586165878130133, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 27813 + }, + { + "epoch": 0.27814, + "grad_norm": 1.4509507641062855, + "learning_rate": 0.003, + "loss": 4.078, + "step": 27814 + }, + { + "epoch": 0.27815, + "grad_norm": 1.1177466393517985, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 27815 + }, + { + "epoch": 0.27816, + "grad_norm": 1.5857018350818985, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 27816 + }, + { + "epoch": 0.27817, + "grad_norm": 1.1531126020134905, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 27817 + }, + { + "epoch": 0.27818, + "grad_norm": 1.6716006473416922, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 27818 + }, + { + "epoch": 0.27819, + "grad_norm": 1.3221642818001973, + "learning_rate": 0.003, + "loss": 4.04, + "step": 27819 + }, + { + "epoch": 0.2782, + "grad_norm": 1.8588332881703102, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 27820 + }, + { + "epoch": 0.27821, + "grad_norm": 1.2688593547340399, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 27821 + }, + { + "epoch": 0.27822, + "grad_norm": 1.5073568111471123, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 27822 + }, + { + "epoch": 0.27823, + "grad_norm": 1.6610120119029177, + "learning_rate": 0.003, + "loss": 4.07, + "step": 27823 + }, + { + "epoch": 0.27824, + "grad_norm": 1.174828621401209, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27824 + }, + { + "epoch": 0.27825, + "grad_norm": 1.6641237607208248, + "learning_rate": 0.003, + "loss": 4.081, + "step": 27825 + }, + { + "epoch": 0.27826, + "grad_norm": 1.5241578366991109, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 27826 + }, + { + "epoch": 0.27827, + "grad_norm": 1.1822413919346058, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 27827 + }, + { + "epoch": 0.27828, + "grad_norm": 1.5699411040217999, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 27828 + }, + { + "epoch": 0.27829, + "grad_norm": 1.3478193710722601, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 27829 + }, + { + "epoch": 0.2783, + "grad_norm": 1.5141432110967796, + "learning_rate": 0.003, + "loss": 4.058, + "step": 27830 + }, + { + "epoch": 0.27831, + "grad_norm": 1.1875750855521194, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 27831 + }, + { + "epoch": 0.27832, + "grad_norm": 1.3571492956200275, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 27832 + }, + { + "epoch": 0.27833, + "grad_norm": 1.296660014244527, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 27833 + }, + { + "epoch": 0.27834, + "grad_norm": 1.3126729410147915, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27834 + }, + { + "epoch": 0.27835, + "grad_norm": 1.6165160965293095, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27835 + }, + { + "epoch": 0.27836, + "grad_norm": 1.280261657082215, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 27836 + }, + { + "epoch": 0.27837, + "grad_norm": 1.428670718252667, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 27837 + }, + { + "epoch": 0.27838, + "grad_norm": 1.1971910613405097, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 27838 + }, + { + "epoch": 0.27839, + "grad_norm": 1.68038634050312, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 27839 + }, + { + "epoch": 0.2784, + "grad_norm": 1.2850201654509281, + "learning_rate": 0.003, + "loss": 4.068, + "step": 27840 + }, + { + "epoch": 0.27841, + "grad_norm": 1.3701480150375187, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 27841 + }, + { + "epoch": 0.27842, + "grad_norm": 1.303940028671007, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27842 + }, + { + "epoch": 0.27843, + "grad_norm": 1.5158019846383994, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 27843 + }, + { + "epoch": 0.27844, + "grad_norm": 1.4127253194482847, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 27844 + }, + { + "epoch": 0.27845, + "grad_norm": 1.4552701333359734, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 27845 + }, + { + "epoch": 0.27846, + "grad_norm": 1.2526271665854096, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 27846 + }, + { + "epoch": 0.27847, + "grad_norm": 1.7147803606171852, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 27847 + }, + { + "epoch": 0.27848, + "grad_norm": 0.9591479051061004, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 27848 + }, + { + "epoch": 0.27849, + "grad_norm": 1.6801033262717169, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 27849 + }, + { + "epoch": 0.2785, + "grad_norm": 1.4924732818155255, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 27850 + }, + { + "epoch": 0.27851, + "grad_norm": 1.5837150022323137, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27851 + }, + { + "epoch": 0.27852, + "grad_norm": 1.186140762169877, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 27852 + }, + { + "epoch": 0.27853, + "grad_norm": 1.2802571367263922, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 27853 + }, + { + "epoch": 0.27854, + "grad_norm": 1.6886268789745973, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 27854 + }, + { + "epoch": 0.27855, + "grad_norm": 1.3363086431619078, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27855 + }, + { + "epoch": 0.27856, + "grad_norm": 1.4939178683420327, + "learning_rate": 0.003, + "loss": 4.067, + "step": 27856 + }, + { + "epoch": 0.27857, + "grad_norm": 1.50578725873819, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 27857 + }, + { + "epoch": 0.27858, + "grad_norm": 1.4772361269307492, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 27858 + }, + { + "epoch": 0.27859, + "grad_norm": 1.3550243214979791, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 27859 + }, + { + "epoch": 0.2786, + "grad_norm": 1.5297825293891432, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 27860 + }, + { + "epoch": 0.27861, + "grad_norm": 1.3144158871243175, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 27861 + }, + { + "epoch": 0.27862, + "grad_norm": 1.3101812579046161, + "learning_rate": 0.003, + "loss": 4.038, + "step": 27862 + }, + { + "epoch": 0.27863, + "grad_norm": 1.4323626866187202, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 27863 + }, + { + "epoch": 0.27864, + "grad_norm": 1.5071765961186165, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 27864 + }, + { + "epoch": 0.27865, + "grad_norm": 1.401040966583306, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 27865 + }, + { + "epoch": 0.27866, + "grad_norm": 1.495048113864856, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 27866 + }, + { + "epoch": 0.27867, + "grad_norm": 1.4064084963494055, + "learning_rate": 0.003, + "loss": 4.039, + "step": 27867 + }, + { + "epoch": 0.27868, + "grad_norm": 1.559326584791895, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 27868 + }, + { + "epoch": 0.27869, + "grad_norm": 1.1643224903435427, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 27869 + }, + { + "epoch": 0.2787, + "grad_norm": 1.7408296398116208, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 27870 + }, + { + "epoch": 0.27871, + "grad_norm": 1.3636834546554506, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 27871 + }, + { + "epoch": 0.27872, + "grad_norm": 1.3080009463787823, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 27872 + }, + { + "epoch": 0.27873, + "grad_norm": 1.537278193823686, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27873 + }, + { + "epoch": 0.27874, + "grad_norm": 1.16407900396194, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 27874 + }, + { + "epoch": 0.27875, + "grad_norm": 1.6919376234828667, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 27875 + }, + { + "epoch": 0.27876, + "grad_norm": 1.0713839307071182, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 27876 + }, + { + "epoch": 0.27877, + "grad_norm": 1.7394346676940537, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 27877 + }, + { + "epoch": 0.27878, + "grad_norm": 1.18512851950238, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 27878 + }, + { + "epoch": 0.27879, + "grad_norm": 1.5841117638392817, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 27879 + }, + { + "epoch": 0.2788, + "grad_norm": 1.502170161858935, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 27880 + }, + { + "epoch": 0.27881, + "grad_norm": 1.3479214651720703, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 27881 + }, + { + "epoch": 0.27882, + "grad_norm": 1.6264010819672292, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 27882 + }, + { + "epoch": 0.27883, + "grad_norm": 1.4026530707653029, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 27883 + }, + { + "epoch": 0.27884, + "grad_norm": 1.5884104764237255, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 27884 + }, + { + "epoch": 0.27885, + "grad_norm": 1.3792998617653653, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 27885 + }, + { + "epoch": 0.27886, + "grad_norm": 1.4939591919186848, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 27886 + }, + { + "epoch": 0.27887, + "grad_norm": 1.2704234276590978, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27887 + }, + { + "epoch": 0.27888, + "grad_norm": 1.4549653391396915, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 27888 + }, + { + "epoch": 0.27889, + "grad_norm": 1.4155840406888174, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27889 + }, + { + "epoch": 0.2789, + "grad_norm": 1.8011380630634692, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27890 + }, + { + "epoch": 0.27891, + "grad_norm": 1.6343514927272953, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 27891 + }, + { + "epoch": 0.27892, + "grad_norm": 1.1841461765467138, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 27892 + }, + { + "epoch": 0.27893, + "grad_norm": 1.712514493187982, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 27893 + }, + { + "epoch": 0.27894, + "grad_norm": 1.2762918634384144, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 27894 + }, + { + "epoch": 0.27895, + "grad_norm": 1.5035377588676369, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 27895 + }, + { + "epoch": 0.27896, + "grad_norm": 1.3473151828495122, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 27896 + }, + { + "epoch": 0.27897, + "grad_norm": 1.524695641636234, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 27897 + }, + { + "epoch": 0.27898, + "grad_norm": 1.3675130760323044, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 27898 + }, + { + "epoch": 0.27899, + "grad_norm": 1.4413703326389489, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 27899 + }, + { + "epoch": 0.279, + "grad_norm": 1.2630429455660526, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 27900 + }, + { + "epoch": 0.27901, + "grad_norm": 1.522728319037244, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27901 + }, + { + "epoch": 0.27902, + "grad_norm": 1.2994197735530777, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 27902 + }, + { + "epoch": 0.27903, + "grad_norm": 1.3373585661904024, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 27903 + }, + { + "epoch": 0.27904, + "grad_norm": 1.6450571427256644, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 27904 + }, + { + "epoch": 0.27905, + "grad_norm": 1.7410389012618426, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 27905 + }, + { + "epoch": 0.27906, + "grad_norm": 1.232471370521989, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 27906 + }, + { + "epoch": 0.27907, + "grad_norm": 1.606933249024544, + "learning_rate": 0.003, + "loss": 4.052, + "step": 27907 + }, + { + "epoch": 0.27908, + "grad_norm": 1.1832863297724772, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 27908 + }, + { + "epoch": 0.27909, + "grad_norm": 1.4878427740327442, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 27909 + }, + { + "epoch": 0.2791, + "grad_norm": 1.5126673576444003, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 27910 + }, + { + "epoch": 0.27911, + "grad_norm": 1.2819286800481546, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27911 + }, + { + "epoch": 0.27912, + "grad_norm": 1.3828400422302813, + "learning_rate": 0.003, + "loss": 4.008, + "step": 27912 + }, + { + "epoch": 0.27913, + "grad_norm": 1.27763456501995, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 27913 + }, + { + "epoch": 0.27914, + "grad_norm": 1.4429937583634431, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 27914 + }, + { + "epoch": 0.27915, + "grad_norm": 1.3203192234244907, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 27915 + }, + { + "epoch": 0.27916, + "grad_norm": 1.2887328443607522, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 27916 + }, + { + "epoch": 0.27917, + "grad_norm": 1.3301811334973468, + "learning_rate": 0.003, + "loss": 4.035, + "step": 27917 + }, + { + "epoch": 0.27918, + "grad_norm": 1.861056841092642, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 27918 + }, + { + "epoch": 0.27919, + "grad_norm": 1.3732598228668593, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 27919 + }, + { + "epoch": 0.2792, + "grad_norm": 1.3311278027275661, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 27920 + }, + { + "epoch": 0.27921, + "grad_norm": 1.6766691384499708, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 27921 + }, + { + "epoch": 0.27922, + "grad_norm": 1.1569939493072154, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 27922 + }, + { + "epoch": 0.27923, + "grad_norm": 1.820948197027293, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 27923 + }, + { + "epoch": 0.27924, + "grad_norm": 1.2768881318735548, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27924 + }, + { + "epoch": 0.27925, + "grad_norm": 1.665192022306569, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 27925 + }, + { + "epoch": 0.27926, + "grad_norm": 1.5742439359416904, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 27926 + }, + { + "epoch": 0.27927, + "grad_norm": 1.3194988353161607, + "learning_rate": 0.003, + "loss": 4.051, + "step": 27927 + }, + { + "epoch": 0.27928, + "grad_norm": 1.3945527884781903, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 27928 + }, + { + "epoch": 0.27929, + "grad_norm": 1.3008603815310773, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 27929 + }, + { + "epoch": 0.2793, + "grad_norm": 1.2229915244772052, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 27930 + }, + { + "epoch": 0.27931, + "grad_norm": 1.6684809531509222, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 27931 + }, + { + "epoch": 0.27932, + "grad_norm": 1.1917605508137092, + "learning_rate": 0.003, + "loss": 4.072, + "step": 27932 + }, + { + "epoch": 0.27933, + "grad_norm": 1.3104847364485503, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27933 + }, + { + "epoch": 0.27934, + "grad_norm": 1.4460956418490507, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 27934 + }, + { + "epoch": 0.27935, + "grad_norm": 1.6053525851866146, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 27935 + }, + { + "epoch": 0.27936, + "grad_norm": 1.7568053941136592, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 27936 + }, + { + "epoch": 0.27937, + "grad_norm": 1.4146769896833138, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 27937 + }, + { + "epoch": 0.27938, + "grad_norm": 1.5254066854198827, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27938 + }, + { + "epoch": 0.27939, + "grad_norm": 1.3894197360509761, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 27939 + }, + { + "epoch": 0.2794, + "grad_norm": 1.4870643129660657, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 27940 + }, + { + "epoch": 0.27941, + "grad_norm": 1.329706192271137, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 27941 + }, + { + "epoch": 0.27942, + "grad_norm": 1.3828891014754834, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 27942 + }, + { + "epoch": 0.27943, + "grad_norm": 1.415657765113758, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 27943 + }, + { + "epoch": 0.27944, + "grad_norm": 1.3949648600297269, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 27944 + }, + { + "epoch": 0.27945, + "grad_norm": 1.1135577817010904, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 27945 + }, + { + "epoch": 0.27946, + "grad_norm": 1.757730434200193, + "learning_rate": 0.003, + "loss": 4.055, + "step": 27946 + }, + { + "epoch": 0.27947, + "grad_norm": 1.218207806147363, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 27947 + }, + { + "epoch": 0.27948, + "grad_norm": 2.0905808084381365, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 27948 + }, + { + "epoch": 0.27949, + "grad_norm": 1.133164159065009, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 27949 + }, + { + "epoch": 0.2795, + "grad_norm": 1.2073339477074838, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 27950 + }, + { + "epoch": 0.27951, + "grad_norm": 1.4222841771964407, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 27951 + }, + { + "epoch": 0.27952, + "grad_norm": 1.2576282451080545, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 27952 + }, + { + "epoch": 0.27953, + "grad_norm": 1.947570568028415, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 27953 + }, + { + "epoch": 0.27954, + "grad_norm": 1.5018050312033153, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 27954 + }, + { + "epoch": 0.27955, + "grad_norm": 1.299756545532794, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 27955 + }, + { + "epoch": 0.27956, + "grad_norm": 1.5909883597244263, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 27956 + }, + { + "epoch": 0.27957, + "grad_norm": 1.444812303684301, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 27957 + }, + { + "epoch": 0.27958, + "grad_norm": 1.4980718163808817, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 27958 + }, + { + "epoch": 0.27959, + "grad_norm": 1.1799228479809993, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 27959 + }, + { + "epoch": 0.2796, + "grad_norm": 1.557251942657385, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 27960 + }, + { + "epoch": 0.27961, + "grad_norm": 1.2025583832665545, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 27961 + }, + { + "epoch": 0.27962, + "grad_norm": 1.6820736400448593, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 27962 + }, + { + "epoch": 0.27963, + "grad_norm": 1.2394097843696588, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 27963 + }, + { + "epoch": 0.27964, + "grad_norm": 1.484931885816524, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 27964 + }, + { + "epoch": 0.27965, + "grad_norm": 1.8108185510223056, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 27965 + }, + { + "epoch": 0.27966, + "grad_norm": 1.338034107635084, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 27966 + }, + { + "epoch": 0.27967, + "grad_norm": 1.3561019637280318, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 27967 + }, + { + "epoch": 0.27968, + "grad_norm": 1.4038058702915799, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 27968 + }, + { + "epoch": 0.27969, + "grad_norm": 1.3471666531693007, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 27969 + }, + { + "epoch": 0.2797, + "grad_norm": 1.4897198090338724, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 27970 + }, + { + "epoch": 0.27971, + "grad_norm": 1.1261806677418562, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 27971 + }, + { + "epoch": 0.27972, + "grad_norm": 1.560461175491009, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 27972 + }, + { + "epoch": 0.27973, + "grad_norm": 1.3919278108886974, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 27973 + }, + { + "epoch": 0.27974, + "grad_norm": 1.3355292353233565, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 27974 + }, + { + "epoch": 0.27975, + "grad_norm": 1.7278223461740063, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 27975 + }, + { + "epoch": 0.27976, + "grad_norm": 1.342519275937547, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 27976 + }, + { + "epoch": 0.27977, + "grad_norm": 1.344754814836903, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 27977 + }, + { + "epoch": 0.27978, + "grad_norm": 1.3952707474682757, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 27978 + }, + { + "epoch": 0.27979, + "grad_norm": 1.4172599526004412, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 27979 + }, + { + "epoch": 0.2798, + "grad_norm": 1.4606292419771179, + "learning_rate": 0.003, + "loss": 4.053, + "step": 27980 + }, + { + "epoch": 0.27981, + "grad_norm": 1.811171448285547, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 27981 + }, + { + "epoch": 0.27982, + "grad_norm": 1.7697611786974516, + "learning_rate": 0.003, + "loss": 4.032, + "step": 27982 + }, + { + "epoch": 0.27983, + "grad_norm": 1.2356396270978498, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 27983 + }, + { + "epoch": 0.27984, + "grad_norm": 1.3716965223798923, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 27984 + }, + { + "epoch": 0.27985, + "grad_norm": 1.5755989398732009, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 27985 + }, + { + "epoch": 0.27986, + "grad_norm": 1.2197797711153024, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 27986 + }, + { + "epoch": 0.27987, + "grad_norm": 1.5017060752347513, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 27987 + }, + { + "epoch": 0.27988, + "grad_norm": 1.1448663526372072, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 27988 + }, + { + "epoch": 0.27989, + "grad_norm": 1.4760755176205622, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 27989 + }, + { + "epoch": 0.2799, + "grad_norm": 1.2272554533410318, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 27990 + }, + { + "epoch": 0.27991, + "grad_norm": 1.757752584024771, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 27991 + }, + { + "epoch": 0.27992, + "grad_norm": 1.4180884065285877, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 27992 + }, + { + "epoch": 0.27993, + "grad_norm": 1.6221471650852317, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 27993 + }, + { + "epoch": 0.27994, + "grad_norm": 1.3694552503451287, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 27994 + }, + { + "epoch": 0.27995, + "grad_norm": 1.6761768636337488, + "learning_rate": 0.003, + "loss": 4.107, + "step": 27995 + }, + { + "epoch": 0.27996, + "grad_norm": 1.1521283179079935, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 27996 + }, + { + "epoch": 0.27997, + "grad_norm": 1.5028222136375287, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 27997 + }, + { + "epoch": 0.27998, + "grad_norm": 1.2132961470272585, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 27998 + }, + { + "epoch": 0.27999, + "grad_norm": 1.2852100745837023, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 27999 + }, + { + "epoch": 0.28, + "grad_norm": 1.4956746137193153, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 28000 + }, + { + "epoch": 0.28001, + "grad_norm": 1.3804756168252021, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 28001 + }, + { + "epoch": 0.28002, + "grad_norm": 1.6546825659794868, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 28002 + }, + { + "epoch": 0.28003, + "grad_norm": 1.4670828910159788, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 28003 + }, + { + "epoch": 0.28004, + "grad_norm": 1.432928103940398, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 28004 + }, + { + "epoch": 0.28005, + "grad_norm": 1.4307517199518711, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 28005 + }, + { + "epoch": 0.28006, + "grad_norm": 1.4099072981688836, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28006 + }, + { + "epoch": 0.28007, + "grad_norm": 1.6742789551357562, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 28007 + }, + { + "epoch": 0.28008, + "grad_norm": 1.558926425051046, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 28008 + }, + { + "epoch": 0.28009, + "grad_norm": 1.5230196330292298, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28009 + }, + { + "epoch": 0.2801, + "grad_norm": 1.3848997811151915, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 28010 + }, + { + "epoch": 0.28011, + "grad_norm": 1.398188047816684, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 28011 + }, + { + "epoch": 0.28012, + "grad_norm": 1.3690226756413342, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28012 + }, + { + "epoch": 0.28013, + "grad_norm": 1.3915964279234811, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 28013 + }, + { + "epoch": 0.28014, + "grad_norm": 1.5539997458459063, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 28014 + }, + { + "epoch": 0.28015, + "grad_norm": 1.3990468432342174, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28015 + }, + { + "epoch": 0.28016, + "grad_norm": 1.703170353042076, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28016 + }, + { + "epoch": 0.28017, + "grad_norm": 1.322174488201851, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 28017 + }, + { + "epoch": 0.28018, + "grad_norm": 1.5040093217993442, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 28018 + }, + { + "epoch": 0.28019, + "grad_norm": 1.324495848915324, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 28019 + }, + { + "epoch": 0.2802, + "grad_norm": 1.6371519077675933, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28020 + }, + { + "epoch": 0.28021, + "grad_norm": 1.324224522515258, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 28021 + }, + { + "epoch": 0.28022, + "grad_norm": 1.342194535950628, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 28022 + }, + { + "epoch": 0.28023, + "grad_norm": 1.5855099710276526, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 28023 + }, + { + "epoch": 0.28024, + "grad_norm": 1.0849542716837126, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 28024 + }, + { + "epoch": 0.28025, + "grad_norm": 1.710932438301282, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 28025 + }, + { + "epoch": 0.28026, + "grad_norm": 1.114872350654895, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28026 + }, + { + "epoch": 0.28027, + "grad_norm": 1.3710911021074739, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 28027 + }, + { + "epoch": 0.28028, + "grad_norm": 1.2939593384707795, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 28028 + }, + { + "epoch": 0.28029, + "grad_norm": 1.535958227810724, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28029 + }, + { + "epoch": 0.2803, + "grad_norm": 1.2890201385181335, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 28030 + }, + { + "epoch": 0.28031, + "grad_norm": 1.4233581975180798, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28031 + }, + { + "epoch": 0.28032, + "grad_norm": 1.250406890813262, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28032 + }, + { + "epoch": 0.28033, + "grad_norm": 1.3705467457167049, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 28033 + }, + { + "epoch": 0.28034, + "grad_norm": 1.5499920452110711, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 28034 + }, + { + "epoch": 0.28035, + "grad_norm": 1.5298260059675142, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28035 + }, + { + "epoch": 0.28036, + "grad_norm": 1.7838146928784466, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28036 + }, + { + "epoch": 0.28037, + "grad_norm": 1.4982549875187978, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 28037 + }, + { + "epoch": 0.28038, + "grad_norm": 1.4342096037855423, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28038 + }, + { + "epoch": 0.28039, + "grad_norm": 1.358984734106416, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 28039 + }, + { + "epoch": 0.2804, + "grad_norm": 1.4466244793159933, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 28040 + }, + { + "epoch": 0.28041, + "grad_norm": 1.527206738002683, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 28041 + }, + { + "epoch": 0.28042, + "grad_norm": 1.4111544463930097, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 28042 + }, + { + "epoch": 0.28043, + "grad_norm": 1.435057353837599, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 28043 + }, + { + "epoch": 0.28044, + "grad_norm": 1.4906973010863491, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 28044 + }, + { + "epoch": 0.28045, + "grad_norm": 1.3723125715928468, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 28045 + }, + { + "epoch": 0.28046, + "grad_norm": 1.3314548143144955, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 28046 + }, + { + "epoch": 0.28047, + "grad_norm": 1.3281370879649248, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 28047 + }, + { + "epoch": 0.28048, + "grad_norm": 1.381053997430662, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28048 + }, + { + "epoch": 0.28049, + "grad_norm": 1.3991769538388725, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 28049 + }, + { + "epoch": 0.2805, + "grad_norm": 1.2903398677707505, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 28050 + }, + { + "epoch": 0.28051, + "grad_norm": 1.6212572465345498, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 28051 + }, + { + "epoch": 0.28052, + "grad_norm": 1.3201888760274292, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28052 + }, + { + "epoch": 0.28053, + "grad_norm": 1.7693317841897231, + "learning_rate": 0.003, + "loss": 4.059, + "step": 28053 + }, + { + "epoch": 0.28054, + "grad_norm": 1.501544724487128, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 28054 + }, + { + "epoch": 0.28055, + "grad_norm": 1.78450530494913, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 28055 + }, + { + "epoch": 0.28056, + "grad_norm": 1.23087963900349, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 28056 + }, + { + "epoch": 0.28057, + "grad_norm": 1.7015768852140118, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 28057 + }, + { + "epoch": 0.28058, + "grad_norm": 1.237684156368919, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28058 + }, + { + "epoch": 0.28059, + "grad_norm": 1.6081261629259893, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28059 + }, + { + "epoch": 0.2806, + "grad_norm": 1.1942257234981466, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 28060 + }, + { + "epoch": 0.28061, + "grad_norm": 1.7653186337114772, + "learning_rate": 0.003, + "loss": 4.077, + "step": 28061 + }, + { + "epoch": 0.28062, + "grad_norm": 1.130318819579011, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 28062 + }, + { + "epoch": 0.28063, + "grad_norm": 1.6793471020980426, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 28063 + }, + { + "epoch": 0.28064, + "grad_norm": 1.4280078316052212, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28064 + }, + { + "epoch": 0.28065, + "grad_norm": 1.6356137736209948, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 28065 + }, + { + "epoch": 0.28066, + "grad_norm": 1.3994956120810886, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28066 + }, + { + "epoch": 0.28067, + "grad_norm": 1.2289202061726094, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 28067 + }, + { + "epoch": 0.28068, + "grad_norm": 1.307063970284685, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28068 + }, + { + "epoch": 0.28069, + "grad_norm": 1.273492010226247, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 28069 + }, + { + "epoch": 0.2807, + "grad_norm": 1.442204764596742, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 28070 + }, + { + "epoch": 0.28071, + "grad_norm": 1.2776056489275278, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 28071 + }, + { + "epoch": 0.28072, + "grad_norm": 1.3700118780558441, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28072 + }, + { + "epoch": 0.28073, + "grad_norm": 1.3770186140465195, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 28073 + }, + { + "epoch": 0.28074, + "grad_norm": 1.421088192008363, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 28074 + }, + { + "epoch": 0.28075, + "grad_norm": 1.8512626556194027, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 28075 + }, + { + "epoch": 0.28076, + "grad_norm": 1.325813498662282, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 28076 + }, + { + "epoch": 0.28077, + "grad_norm": 1.532665485612633, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 28077 + }, + { + "epoch": 0.28078, + "grad_norm": 1.304241413078253, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 28078 + }, + { + "epoch": 0.28079, + "grad_norm": 1.5182069883535134, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28079 + }, + { + "epoch": 0.2808, + "grad_norm": 1.146116060220775, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 28080 + }, + { + "epoch": 0.28081, + "grad_norm": 1.7537883855594985, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28081 + }, + { + "epoch": 0.28082, + "grad_norm": 1.1922237723412377, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 28082 + }, + { + "epoch": 0.28083, + "grad_norm": 1.6648884145062282, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 28083 + }, + { + "epoch": 0.28084, + "grad_norm": 1.3763195423425343, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 28084 + }, + { + "epoch": 0.28085, + "grad_norm": 1.5233752906677427, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28085 + }, + { + "epoch": 0.28086, + "grad_norm": 1.3837598501216062, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 28086 + }, + { + "epoch": 0.28087, + "grad_norm": 1.413823466065853, + "learning_rate": 0.003, + "loss": 4.071, + "step": 28087 + }, + { + "epoch": 0.28088, + "grad_norm": 1.5271110038525275, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28088 + }, + { + "epoch": 0.28089, + "grad_norm": 1.6754065302791639, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 28089 + }, + { + "epoch": 0.2809, + "grad_norm": 1.2873500151898656, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 28090 + }, + { + "epoch": 0.28091, + "grad_norm": 1.410188205131897, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 28091 + }, + { + "epoch": 0.28092, + "grad_norm": 1.506659449072457, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 28092 + }, + { + "epoch": 0.28093, + "grad_norm": 1.3195967981836867, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28093 + }, + { + "epoch": 0.28094, + "grad_norm": 1.3139609150159994, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 28094 + }, + { + "epoch": 0.28095, + "grad_norm": 1.35297428619044, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28095 + }, + { + "epoch": 0.28096, + "grad_norm": 1.2752358278864764, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28096 + }, + { + "epoch": 0.28097, + "grad_norm": 1.594730846666583, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 28097 + }, + { + "epoch": 0.28098, + "grad_norm": 1.6462531255102957, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 28098 + }, + { + "epoch": 0.28099, + "grad_norm": 1.6689290644389172, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 28099 + }, + { + "epoch": 0.281, + "grad_norm": 1.504799128678178, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28100 + }, + { + "epoch": 0.28101, + "grad_norm": 1.6519505907212517, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28101 + }, + { + "epoch": 0.28102, + "grad_norm": 1.189613179408905, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 28102 + }, + { + "epoch": 0.28103, + "grad_norm": 1.6143631975193051, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 28103 + }, + { + "epoch": 0.28104, + "grad_norm": 1.2591835713288635, + "learning_rate": 0.003, + "loss": 4.058, + "step": 28104 + }, + { + "epoch": 0.28105, + "grad_norm": 1.5931109560803387, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28105 + }, + { + "epoch": 0.28106, + "grad_norm": 1.3007183519493517, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 28106 + }, + { + "epoch": 0.28107, + "grad_norm": 1.5314383439577974, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 28107 + }, + { + "epoch": 0.28108, + "grad_norm": 1.3602668718327944, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28108 + }, + { + "epoch": 0.28109, + "grad_norm": 1.355829688167184, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 28109 + }, + { + "epoch": 0.2811, + "grad_norm": 1.4830022364886921, + "learning_rate": 0.003, + "loss": 4.034, + "step": 28110 + }, + { + "epoch": 0.28111, + "grad_norm": 1.3392886261761234, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 28111 + }, + { + "epoch": 0.28112, + "grad_norm": 1.5544126856774205, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28112 + }, + { + "epoch": 0.28113, + "grad_norm": 1.1953966902378643, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 28113 + }, + { + "epoch": 0.28114, + "grad_norm": 1.6165098515519318, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 28114 + }, + { + "epoch": 0.28115, + "grad_norm": 1.3183966751564369, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 28115 + }, + { + "epoch": 0.28116, + "grad_norm": 1.255673023813345, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 28116 + }, + { + "epoch": 0.28117, + "grad_norm": 1.3764616725437633, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 28117 + }, + { + "epoch": 0.28118, + "grad_norm": 1.2530077403331252, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28118 + }, + { + "epoch": 0.28119, + "grad_norm": 1.4051570051261377, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 28119 + }, + { + "epoch": 0.2812, + "grad_norm": 1.34439643904691, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 28120 + }, + { + "epoch": 0.28121, + "grad_norm": 1.776905555707211, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 28121 + }, + { + "epoch": 0.28122, + "grad_norm": 1.8821238726186151, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 28122 + }, + { + "epoch": 0.28123, + "grad_norm": 1.179782646796394, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28123 + }, + { + "epoch": 0.28124, + "grad_norm": 1.5735462463795213, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 28124 + }, + { + "epoch": 0.28125, + "grad_norm": 1.5121490925123597, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 28125 + }, + { + "epoch": 0.28126, + "grad_norm": 1.3337757991358494, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28126 + }, + { + "epoch": 0.28127, + "grad_norm": 1.4872249608872965, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28127 + }, + { + "epoch": 0.28128, + "grad_norm": 1.6327896499076555, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 28128 + }, + { + "epoch": 0.28129, + "grad_norm": 1.5028555240134382, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 28129 + }, + { + "epoch": 0.2813, + "grad_norm": 1.3189292496739016, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28130 + }, + { + "epoch": 0.28131, + "grad_norm": 1.4379015775414343, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 28131 + }, + { + "epoch": 0.28132, + "grad_norm": 1.4028422519973027, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 28132 + }, + { + "epoch": 0.28133, + "grad_norm": 1.5271342836430801, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 28133 + }, + { + "epoch": 0.28134, + "grad_norm": 1.352456375939028, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28134 + }, + { + "epoch": 0.28135, + "grad_norm": 1.3936563104569952, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 28135 + }, + { + "epoch": 0.28136, + "grad_norm": 1.3680961380640062, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 28136 + }, + { + "epoch": 0.28137, + "grad_norm": 1.222759176886141, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 28137 + }, + { + "epoch": 0.28138, + "grad_norm": 1.4533806177109763, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28138 + }, + { + "epoch": 0.28139, + "grad_norm": 1.2276140349916742, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28139 + }, + { + "epoch": 0.2814, + "grad_norm": 1.8785453185649164, + "learning_rate": 0.003, + "loss": 4.094, + "step": 28140 + }, + { + "epoch": 0.28141, + "grad_norm": 1.5637090984351747, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 28141 + }, + { + "epoch": 0.28142, + "grad_norm": 1.4757686963844643, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 28142 + }, + { + "epoch": 0.28143, + "grad_norm": 1.5557196003927498, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 28143 + }, + { + "epoch": 0.28144, + "grad_norm": 1.102786997715197, + "learning_rate": 0.003, + "loss": 4.0995, + "step": 28144 + }, + { + "epoch": 0.28145, + "grad_norm": 1.6202232425951337, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 28145 + }, + { + "epoch": 0.28146, + "grad_norm": 1.327729278688086, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 28146 + }, + { + "epoch": 0.28147, + "grad_norm": 1.5130132522565867, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 28147 + }, + { + "epoch": 0.28148, + "grad_norm": 1.2465021689177174, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 28148 + }, + { + "epoch": 0.28149, + "grad_norm": 1.8543602482451436, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28149 + }, + { + "epoch": 0.2815, + "grad_norm": 1.175303980941688, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 28150 + }, + { + "epoch": 0.28151, + "grad_norm": 1.5187108005363832, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 28151 + }, + { + "epoch": 0.28152, + "grad_norm": 1.6531605345970484, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 28152 + }, + { + "epoch": 0.28153, + "grad_norm": 1.3085611458253656, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 28153 + }, + { + "epoch": 0.28154, + "grad_norm": 1.3861058540608064, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28154 + }, + { + "epoch": 0.28155, + "grad_norm": 1.3581601710276205, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 28155 + }, + { + "epoch": 0.28156, + "grad_norm": 1.484641346876184, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 28156 + }, + { + "epoch": 0.28157, + "grad_norm": 1.4277690663668574, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28157 + }, + { + "epoch": 0.28158, + "grad_norm": 1.5880897259078397, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 28158 + }, + { + "epoch": 0.28159, + "grad_norm": 1.604010605472175, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 28159 + }, + { + "epoch": 0.2816, + "grad_norm": 1.055692215899844, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 28160 + }, + { + "epoch": 0.28161, + "grad_norm": 1.9533047975703055, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28161 + }, + { + "epoch": 0.28162, + "grad_norm": 1.511887189218972, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28162 + }, + { + "epoch": 0.28163, + "grad_norm": 1.3553033152600686, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28163 + }, + { + "epoch": 0.28164, + "grad_norm": 1.645209622080125, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 28164 + }, + { + "epoch": 0.28165, + "grad_norm": 1.257135251991267, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 28165 + }, + { + "epoch": 0.28166, + "grad_norm": 1.5314999032212009, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 28166 + }, + { + "epoch": 0.28167, + "grad_norm": 1.2609848003488957, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 28167 + }, + { + "epoch": 0.28168, + "grad_norm": 1.3976383483347945, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 28168 + }, + { + "epoch": 0.28169, + "grad_norm": 1.3235251576818408, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28169 + }, + { + "epoch": 0.2817, + "grad_norm": 1.5296765356468105, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 28170 + }, + { + "epoch": 0.28171, + "grad_norm": 1.5277882043357218, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 28171 + }, + { + "epoch": 0.28172, + "grad_norm": 1.1499319645528618, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28172 + }, + { + "epoch": 0.28173, + "grad_norm": 1.786755358846477, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 28173 + }, + { + "epoch": 0.28174, + "grad_norm": 1.4858587800473086, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28174 + }, + { + "epoch": 0.28175, + "grad_norm": 1.3582304141812593, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28175 + }, + { + "epoch": 0.28176, + "grad_norm": 1.3830143016137246, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 28176 + }, + { + "epoch": 0.28177, + "grad_norm": 1.1573247147516612, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28177 + }, + { + "epoch": 0.28178, + "grad_norm": 1.5566599990574053, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 28178 + }, + { + "epoch": 0.28179, + "grad_norm": 1.1947950644198504, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 28179 + }, + { + "epoch": 0.2818, + "grad_norm": 1.848421880046737, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28180 + }, + { + "epoch": 0.28181, + "grad_norm": 1.5866276900614273, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 28181 + }, + { + "epoch": 0.28182, + "grad_norm": 1.2825306556616765, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28182 + }, + { + "epoch": 0.28183, + "grad_norm": 1.2814407845859437, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 28183 + }, + { + "epoch": 0.28184, + "grad_norm": 1.4213431709464535, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 28184 + }, + { + "epoch": 0.28185, + "grad_norm": 1.3467453315176776, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 28185 + }, + { + "epoch": 0.28186, + "grad_norm": 1.5757023825150214, + "learning_rate": 0.003, + "loss": 4.043, + "step": 28186 + }, + { + "epoch": 0.28187, + "grad_norm": 1.2524598485986305, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28187 + }, + { + "epoch": 0.28188, + "grad_norm": 1.594159482705982, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28188 + }, + { + "epoch": 0.28189, + "grad_norm": 1.0768423065397121, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 28189 + }, + { + "epoch": 0.2819, + "grad_norm": 1.7912399792150582, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 28190 + }, + { + "epoch": 0.28191, + "grad_norm": 1.426410637896209, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 28191 + }, + { + "epoch": 0.28192, + "grad_norm": 1.736420283230235, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 28192 + }, + { + "epoch": 0.28193, + "grad_norm": 1.3241932161066012, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 28193 + }, + { + "epoch": 0.28194, + "grad_norm": 1.3839362936227906, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 28194 + }, + { + "epoch": 0.28195, + "grad_norm": 1.2899723373789698, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 28195 + }, + { + "epoch": 0.28196, + "grad_norm": 1.626022651931185, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 28196 + }, + { + "epoch": 0.28197, + "grad_norm": 1.2641817482135012, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 28197 + }, + { + "epoch": 0.28198, + "grad_norm": 1.3794874803523336, + "learning_rate": 0.003, + "loss": 4.079, + "step": 28198 + }, + { + "epoch": 0.28199, + "grad_norm": 1.4924975740940736, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 28199 + }, + { + "epoch": 0.282, + "grad_norm": 1.4673957237140205, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28200 + }, + { + "epoch": 0.28201, + "grad_norm": 1.2621907030776494, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28201 + }, + { + "epoch": 0.28202, + "grad_norm": 1.5022679077960042, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 28202 + }, + { + "epoch": 0.28203, + "grad_norm": 1.1557553808131376, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 28203 + }, + { + "epoch": 0.28204, + "grad_norm": 1.5877206179968484, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28204 + }, + { + "epoch": 0.28205, + "grad_norm": 1.2723770107464154, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 28205 + }, + { + "epoch": 0.28206, + "grad_norm": 1.2006848931108327, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 28206 + }, + { + "epoch": 0.28207, + "grad_norm": 1.5170506595129083, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28207 + }, + { + "epoch": 0.28208, + "grad_norm": 1.7266939455402863, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 28208 + }, + { + "epoch": 0.28209, + "grad_norm": 1.531839328269756, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 28209 + }, + { + "epoch": 0.2821, + "grad_norm": 1.4379499066062114, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28210 + }, + { + "epoch": 0.28211, + "grad_norm": 1.547196240180348, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 28211 + }, + { + "epoch": 0.28212, + "grad_norm": 1.3492039145457078, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 28212 + }, + { + "epoch": 0.28213, + "grad_norm": 1.2917430542927375, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28213 + }, + { + "epoch": 0.28214, + "grad_norm": 1.5250773002540878, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28214 + }, + { + "epoch": 0.28215, + "grad_norm": 1.1464518485428474, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 28215 + }, + { + "epoch": 0.28216, + "grad_norm": 1.9199226428576597, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 28216 + }, + { + "epoch": 0.28217, + "grad_norm": 1.1934498825357167, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 28217 + }, + { + "epoch": 0.28218, + "grad_norm": 1.7017768193849903, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 28218 + }, + { + "epoch": 0.28219, + "grad_norm": 1.4269337741436552, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 28219 + }, + { + "epoch": 0.2822, + "grad_norm": 1.871172968069226, + "learning_rate": 0.003, + "loss": 4.034, + "step": 28220 + }, + { + "epoch": 0.28221, + "grad_norm": 1.3056577739156743, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 28221 + }, + { + "epoch": 0.28222, + "grad_norm": 1.4920441036524001, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 28222 + }, + { + "epoch": 0.28223, + "grad_norm": 1.3980824102837253, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 28223 + }, + { + "epoch": 0.28224, + "grad_norm": 1.55500162657575, + "learning_rate": 0.003, + "loss": 4.068, + "step": 28224 + }, + { + "epoch": 0.28225, + "grad_norm": 1.2487786009814215, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 28225 + }, + { + "epoch": 0.28226, + "grad_norm": 1.3638305578316654, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 28226 + }, + { + "epoch": 0.28227, + "grad_norm": 1.2832399655826032, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28227 + }, + { + "epoch": 0.28228, + "grad_norm": 1.1749255698263608, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 28228 + }, + { + "epoch": 0.28229, + "grad_norm": 1.554107184643021, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 28229 + }, + { + "epoch": 0.2823, + "grad_norm": 1.142164236930944, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 28230 + }, + { + "epoch": 0.28231, + "grad_norm": 1.4120558983106895, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28231 + }, + { + "epoch": 0.28232, + "grad_norm": 1.375454208411112, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28232 + }, + { + "epoch": 0.28233, + "grad_norm": 1.5001929102346685, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28233 + }, + { + "epoch": 0.28234, + "grad_norm": 1.4062423331742935, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28234 + }, + { + "epoch": 0.28235, + "grad_norm": 1.6194972908446676, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28235 + }, + { + "epoch": 0.28236, + "grad_norm": 1.575605770701147, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28236 + }, + { + "epoch": 0.28237, + "grad_norm": 1.4124866673310534, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28237 + }, + { + "epoch": 0.28238, + "grad_norm": 1.483479544691791, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 28238 + }, + { + "epoch": 0.28239, + "grad_norm": 1.6395797178431013, + "learning_rate": 0.003, + "loss": 4.061, + "step": 28239 + }, + { + "epoch": 0.2824, + "grad_norm": 1.133226640201686, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 28240 + }, + { + "epoch": 0.28241, + "grad_norm": 1.4878034219809047, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 28241 + }, + { + "epoch": 0.28242, + "grad_norm": 1.3567489696742854, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 28242 + }, + { + "epoch": 0.28243, + "grad_norm": 1.3432026158594408, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28243 + }, + { + "epoch": 0.28244, + "grad_norm": 1.425058841222319, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 28244 + }, + { + "epoch": 0.28245, + "grad_norm": 1.5449618524106492, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 28245 + }, + { + "epoch": 0.28246, + "grad_norm": 1.402305186913856, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28246 + }, + { + "epoch": 0.28247, + "grad_norm": 1.4249620229755198, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28247 + }, + { + "epoch": 0.28248, + "grad_norm": 1.3634774577241124, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 28248 + }, + { + "epoch": 0.28249, + "grad_norm": 1.6135527770024956, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 28249 + }, + { + "epoch": 0.2825, + "grad_norm": 1.398435498477259, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 28250 + }, + { + "epoch": 0.28251, + "grad_norm": 1.3891133562516977, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28251 + }, + { + "epoch": 0.28252, + "grad_norm": 1.3555591600264663, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 28252 + }, + { + "epoch": 0.28253, + "grad_norm": 1.438433497807603, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28253 + }, + { + "epoch": 0.28254, + "grad_norm": 1.2962110687633612, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28254 + }, + { + "epoch": 0.28255, + "grad_norm": 1.5270549675142746, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 28255 + }, + { + "epoch": 0.28256, + "grad_norm": 1.6817811230452795, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 28256 + }, + { + "epoch": 0.28257, + "grad_norm": 1.5328665663630938, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 28257 + }, + { + "epoch": 0.28258, + "grad_norm": 1.213327517401531, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 28258 + }, + { + "epoch": 0.28259, + "grad_norm": 1.3630093071289375, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 28259 + }, + { + "epoch": 0.2826, + "grad_norm": 1.3535000540914286, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28260 + }, + { + "epoch": 0.28261, + "grad_norm": 1.3903640711578449, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 28261 + }, + { + "epoch": 0.28262, + "grad_norm": 1.427573545119795, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 28262 + }, + { + "epoch": 0.28263, + "grad_norm": 1.4367571207829286, + "learning_rate": 0.003, + "loss": 4.012, + "step": 28263 + }, + { + "epoch": 0.28264, + "grad_norm": 1.4415165141477726, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 28264 + }, + { + "epoch": 0.28265, + "grad_norm": 1.3124498077770972, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 28265 + }, + { + "epoch": 0.28266, + "grad_norm": 1.5632095710955431, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 28266 + }, + { + "epoch": 0.28267, + "grad_norm": 1.446289704971485, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 28267 + }, + { + "epoch": 0.28268, + "grad_norm": 1.4817618861346584, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 28268 + }, + { + "epoch": 0.28269, + "grad_norm": 1.5046238375100796, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28269 + }, + { + "epoch": 0.2827, + "grad_norm": 1.467770595774335, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28270 + }, + { + "epoch": 0.28271, + "grad_norm": 1.577243871357258, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 28271 + }, + { + "epoch": 0.28272, + "grad_norm": 1.3329597723481246, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 28272 + }, + { + "epoch": 0.28273, + "grad_norm": 1.48196258008277, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28273 + }, + { + "epoch": 0.28274, + "grad_norm": 1.391254181420815, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28274 + }, + { + "epoch": 0.28275, + "grad_norm": 1.3440681467348172, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28275 + }, + { + "epoch": 0.28276, + "grad_norm": 1.3738946639275094, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28276 + }, + { + "epoch": 0.28277, + "grad_norm": 1.5264153316108218, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 28277 + }, + { + "epoch": 0.28278, + "grad_norm": 1.340004795756959, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 28278 + }, + { + "epoch": 0.28279, + "grad_norm": 1.440912179075416, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 28279 + }, + { + "epoch": 0.2828, + "grad_norm": 1.3823220276742216, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 28280 + }, + { + "epoch": 0.28281, + "grad_norm": 1.572389205568554, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28281 + }, + { + "epoch": 0.28282, + "grad_norm": 1.2500232174121628, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 28282 + }, + { + "epoch": 0.28283, + "grad_norm": 1.623768853755122, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 28283 + }, + { + "epoch": 0.28284, + "grad_norm": 1.4321743411760186, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 28284 + }, + { + "epoch": 0.28285, + "grad_norm": 2.244655752094662, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 28285 + }, + { + "epoch": 0.28286, + "grad_norm": 1.2635885634512158, + "learning_rate": 0.003, + "loss": 4.064, + "step": 28286 + }, + { + "epoch": 0.28287, + "grad_norm": 1.8035966996054713, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 28287 + }, + { + "epoch": 0.28288, + "grad_norm": 1.5054699643759732, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 28288 + }, + { + "epoch": 0.28289, + "grad_norm": 1.4208601855482428, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28289 + }, + { + "epoch": 0.2829, + "grad_norm": 1.3646489962231574, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 28290 + }, + { + "epoch": 0.28291, + "grad_norm": 1.3253439056267866, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 28291 + }, + { + "epoch": 0.28292, + "grad_norm": 1.3060875130948588, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 28292 + }, + { + "epoch": 0.28293, + "grad_norm": 1.405292653487793, + "learning_rate": 0.003, + "loss": 4.043, + "step": 28293 + }, + { + "epoch": 0.28294, + "grad_norm": 1.2569222066702368, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28294 + }, + { + "epoch": 0.28295, + "grad_norm": 1.5485027495419743, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28295 + }, + { + "epoch": 0.28296, + "grad_norm": 1.2344572286811952, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 28296 + }, + { + "epoch": 0.28297, + "grad_norm": 1.5494729284898892, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 28297 + }, + { + "epoch": 0.28298, + "grad_norm": 1.2272089876647239, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28298 + }, + { + "epoch": 0.28299, + "grad_norm": 1.7188703902009392, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 28299 + }, + { + "epoch": 0.283, + "grad_norm": 1.4690105359441603, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28300 + }, + { + "epoch": 0.28301, + "grad_norm": 1.2403548064538852, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28301 + }, + { + "epoch": 0.28302, + "grad_norm": 1.7968033120042142, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 28302 + }, + { + "epoch": 0.28303, + "grad_norm": 1.1903127395931932, + "learning_rate": 0.003, + "loss": 4.018, + "step": 28303 + }, + { + "epoch": 0.28304, + "grad_norm": 1.5364808954206892, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 28304 + }, + { + "epoch": 0.28305, + "grad_norm": 1.4469688307969597, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 28305 + }, + { + "epoch": 0.28306, + "grad_norm": 1.3633265011928755, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 28306 + }, + { + "epoch": 0.28307, + "grad_norm": 1.3830196038496425, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28307 + }, + { + "epoch": 0.28308, + "grad_norm": 1.2778464258778466, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28308 + }, + { + "epoch": 0.28309, + "grad_norm": 1.866470355166943, + "learning_rate": 0.003, + "loss": 4.075, + "step": 28309 + }, + { + "epoch": 0.2831, + "grad_norm": 1.3099545394602994, + "learning_rate": 0.003, + "loss": 4.044, + "step": 28310 + }, + { + "epoch": 0.28311, + "grad_norm": 1.5857349356401895, + "learning_rate": 0.003, + "loss": 4.035, + "step": 28311 + }, + { + "epoch": 0.28312, + "grad_norm": 1.533032413393502, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 28312 + }, + { + "epoch": 0.28313, + "grad_norm": 1.106742150621182, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 28313 + }, + { + "epoch": 0.28314, + "grad_norm": 1.5902572841373173, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 28314 + }, + { + "epoch": 0.28315, + "grad_norm": 1.4110877239025592, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 28315 + }, + { + "epoch": 0.28316, + "grad_norm": 1.4421218565792537, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28316 + }, + { + "epoch": 0.28317, + "grad_norm": 1.3535549880501272, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28317 + }, + { + "epoch": 0.28318, + "grad_norm": 1.3920300276937112, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28318 + }, + { + "epoch": 0.28319, + "grad_norm": 1.3986120188325786, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28319 + }, + { + "epoch": 0.2832, + "grad_norm": 1.2652046085862585, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 28320 + }, + { + "epoch": 0.28321, + "grad_norm": 1.321405445389234, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 28321 + }, + { + "epoch": 0.28322, + "grad_norm": 1.1417113024408434, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 28322 + }, + { + "epoch": 0.28323, + "grad_norm": 1.6481788120783851, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28323 + }, + { + "epoch": 0.28324, + "grad_norm": 1.3681957211490328, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 28324 + }, + { + "epoch": 0.28325, + "grad_norm": 1.4761021104039833, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 28325 + }, + { + "epoch": 0.28326, + "grad_norm": 1.5933445483160555, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 28326 + }, + { + "epoch": 0.28327, + "grad_norm": 1.2536338267680442, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28327 + }, + { + "epoch": 0.28328, + "grad_norm": 1.4839238102676036, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28328 + }, + { + "epoch": 0.28329, + "grad_norm": 1.2715869802787838, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28329 + }, + { + "epoch": 0.2833, + "grad_norm": 1.7060403752795528, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 28330 + }, + { + "epoch": 0.28331, + "grad_norm": 1.1707519526119614, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28331 + }, + { + "epoch": 0.28332, + "grad_norm": 1.4845421142770863, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28332 + }, + { + "epoch": 0.28333, + "grad_norm": 1.1993957680085747, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 28333 + }, + { + "epoch": 0.28334, + "grad_norm": 1.6985827514601377, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 28334 + }, + { + "epoch": 0.28335, + "grad_norm": 1.5352913097118968, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 28335 + }, + { + "epoch": 0.28336, + "grad_norm": 1.5985770929672762, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 28336 + }, + { + "epoch": 0.28337, + "grad_norm": 1.432231516836623, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 28337 + }, + { + "epoch": 0.28338, + "grad_norm": 1.544519496875898, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28338 + }, + { + "epoch": 0.28339, + "grad_norm": 1.2170347179331886, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28339 + }, + { + "epoch": 0.2834, + "grad_norm": 1.5597301657285587, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28340 + }, + { + "epoch": 0.28341, + "grad_norm": 1.3210802384154727, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 28341 + }, + { + "epoch": 0.28342, + "grad_norm": 1.4601306109951118, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 28342 + }, + { + "epoch": 0.28343, + "grad_norm": 1.1421420279783696, + "learning_rate": 0.003, + "loss": 4.065, + "step": 28343 + }, + { + "epoch": 0.28344, + "grad_norm": 1.3532898317622277, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 28344 + }, + { + "epoch": 0.28345, + "grad_norm": 1.4418319694996844, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28345 + }, + { + "epoch": 0.28346, + "grad_norm": 1.7919464103308087, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 28346 + }, + { + "epoch": 0.28347, + "grad_norm": 1.2733745862271963, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 28347 + }, + { + "epoch": 0.28348, + "grad_norm": 1.493898226684768, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28348 + }, + { + "epoch": 0.28349, + "grad_norm": 1.4386342629263957, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 28349 + }, + { + "epoch": 0.2835, + "grad_norm": 1.4320137431219637, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 28350 + }, + { + "epoch": 0.28351, + "grad_norm": 1.44183697507066, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 28351 + }, + { + "epoch": 0.28352, + "grad_norm": 1.84368294662621, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 28352 + }, + { + "epoch": 0.28353, + "grad_norm": 1.365070732311333, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 28353 + }, + { + "epoch": 0.28354, + "grad_norm": 1.541687964235497, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28354 + }, + { + "epoch": 0.28355, + "grad_norm": 1.1508405970708069, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28355 + }, + { + "epoch": 0.28356, + "grad_norm": 1.6535740500416978, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 28356 + }, + { + "epoch": 0.28357, + "grad_norm": 1.3658616986654706, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 28357 + }, + { + "epoch": 0.28358, + "grad_norm": 1.1928346210659264, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 28358 + }, + { + "epoch": 0.28359, + "grad_norm": 1.7010556538737027, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 28359 + }, + { + "epoch": 0.2836, + "grad_norm": 1.2277612608594468, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 28360 + }, + { + "epoch": 0.28361, + "grad_norm": 1.725551030533697, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 28361 + }, + { + "epoch": 0.28362, + "grad_norm": 1.0291718686364348, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 28362 + }, + { + "epoch": 0.28363, + "grad_norm": 1.548693268740041, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28363 + }, + { + "epoch": 0.28364, + "grad_norm": 1.4309128125529267, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28364 + }, + { + "epoch": 0.28365, + "grad_norm": 1.3975513186689261, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 28365 + }, + { + "epoch": 0.28366, + "grad_norm": 1.4130412355594095, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 28366 + }, + { + "epoch": 0.28367, + "grad_norm": 1.640283631246849, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28367 + }, + { + "epoch": 0.28368, + "grad_norm": 1.278787210945701, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 28368 + }, + { + "epoch": 0.28369, + "grad_norm": 1.5474410845554027, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 28369 + }, + { + "epoch": 0.2837, + "grad_norm": 1.4105382697152047, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28370 + }, + { + "epoch": 0.28371, + "grad_norm": 1.376745143689581, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28371 + }, + { + "epoch": 0.28372, + "grad_norm": 1.3061486232932378, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 28372 + }, + { + "epoch": 0.28373, + "grad_norm": 1.3011533125178099, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 28373 + }, + { + "epoch": 0.28374, + "grad_norm": 1.3663690493054308, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 28374 + }, + { + "epoch": 0.28375, + "grad_norm": 1.7139749053621354, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 28375 + }, + { + "epoch": 0.28376, + "grad_norm": 1.1721011045607777, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28376 + }, + { + "epoch": 0.28377, + "grad_norm": 1.7467477457671967, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28377 + }, + { + "epoch": 0.28378, + "grad_norm": 1.5090301053002986, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 28378 + }, + { + "epoch": 0.28379, + "grad_norm": 1.4676533650031012, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 28379 + }, + { + "epoch": 0.2838, + "grad_norm": 1.572574026293832, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 28380 + }, + { + "epoch": 0.28381, + "grad_norm": 1.4186735034667206, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 28381 + }, + { + "epoch": 0.28382, + "grad_norm": 1.22974012621706, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 28382 + }, + { + "epoch": 0.28383, + "grad_norm": 1.4521982979966417, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 28383 + }, + { + "epoch": 0.28384, + "grad_norm": 1.240272292445812, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28384 + }, + { + "epoch": 0.28385, + "grad_norm": 1.6050404580828421, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28385 + }, + { + "epoch": 0.28386, + "grad_norm": 1.4142242114078525, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 28386 + }, + { + "epoch": 0.28387, + "grad_norm": 1.716317617041716, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 28387 + }, + { + "epoch": 0.28388, + "grad_norm": 1.4497964277392628, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28388 + }, + { + "epoch": 0.28389, + "grad_norm": 1.415909221486241, + "learning_rate": 0.003, + "loss": 4.059, + "step": 28389 + }, + { + "epoch": 0.2839, + "grad_norm": 1.6235575698672624, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 28390 + }, + { + "epoch": 0.28391, + "grad_norm": 1.05037832362125, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 28391 + }, + { + "epoch": 0.28392, + "grad_norm": 1.7111035149665006, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 28392 + }, + { + "epoch": 0.28393, + "grad_norm": 1.25591032193965, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 28393 + }, + { + "epoch": 0.28394, + "grad_norm": 1.3580812875195678, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 28394 + }, + { + "epoch": 0.28395, + "grad_norm": 1.3325593198147836, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 28395 + }, + { + "epoch": 0.28396, + "grad_norm": 1.4931900144739771, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 28396 + }, + { + "epoch": 0.28397, + "grad_norm": 1.4865609248191864, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28397 + }, + { + "epoch": 0.28398, + "grad_norm": 1.4908683943192291, + "learning_rate": 0.003, + "loss": 4.056, + "step": 28398 + }, + { + "epoch": 0.28399, + "grad_norm": 1.7201323114133873, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 28399 + }, + { + "epoch": 0.284, + "grad_norm": 1.227030046506881, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 28400 + }, + { + "epoch": 0.28401, + "grad_norm": 1.3821433402607208, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 28401 + }, + { + "epoch": 0.28402, + "grad_norm": 1.5645498213843447, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28402 + }, + { + "epoch": 0.28403, + "grad_norm": 1.6805853284630818, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 28403 + }, + { + "epoch": 0.28404, + "grad_norm": 1.2106558898178057, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 28404 + }, + { + "epoch": 0.28405, + "grad_norm": 1.232821864088713, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 28405 + }, + { + "epoch": 0.28406, + "grad_norm": 1.7318593101941435, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 28406 + }, + { + "epoch": 0.28407, + "grad_norm": 1.0780768933366118, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 28407 + }, + { + "epoch": 0.28408, + "grad_norm": 1.6332819243669392, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28408 + }, + { + "epoch": 0.28409, + "grad_norm": 1.2742444941643756, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28409 + }, + { + "epoch": 0.2841, + "grad_norm": 1.4192045799157091, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28410 + }, + { + "epoch": 0.28411, + "grad_norm": 1.5071564488558022, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 28411 + }, + { + "epoch": 0.28412, + "grad_norm": 1.28184215887409, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 28412 + }, + { + "epoch": 0.28413, + "grad_norm": 1.4696199267586283, + "learning_rate": 0.003, + "loss": 4.049, + "step": 28413 + }, + { + "epoch": 0.28414, + "grad_norm": 1.236489337498277, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 28414 + }, + { + "epoch": 0.28415, + "grad_norm": 1.52271634775828, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28415 + }, + { + "epoch": 0.28416, + "grad_norm": 1.1380228890046873, + "learning_rate": 0.003, + "loss": 4.045, + "step": 28416 + }, + { + "epoch": 0.28417, + "grad_norm": 1.6992000707489885, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 28417 + }, + { + "epoch": 0.28418, + "grad_norm": 1.1133228753416005, + "learning_rate": 0.003, + "loss": 4.09, + "step": 28418 + }, + { + "epoch": 0.28419, + "grad_norm": 1.6261068623265897, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 28419 + }, + { + "epoch": 0.2842, + "grad_norm": 1.2406897004424537, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 28420 + }, + { + "epoch": 0.28421, + "grad_norm": 1.7914406137602386, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28421 + }, + { + "epoch": 0.28422, + "grad_norm": 1.3909186835012055, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 28422 + }, + { + "epoch": 0.28423, + "grad_norm": 1.353162507599526, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 28423 + }, + { + "epoch": 0.28424, + "grad_norm": 1.6470497445305083, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28424 + }, + { + "epoch": 0.28425, + "grad_norm": 1.1746477222138436, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 28425 + }, + { + "epoch": 0.28426, + "grad_norm": 1.5469871904443766, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 28426 + }, + { + "epoch": 0.28427, + "grad_norm": 1.2728502611589136, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 28427 + }, + { + "epoch": 0.28428, + "grad_norm": 1.8862216853559728, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 28428 + }, + { + "epoch": 0.28429, + "grad_norm": 1.111432079550401, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 28429 + }, + { + "epoch": 0.2843, + "grad_norm": 1.5761197446868314, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 28430 + }, + { + "epoch": 0.28431, + "grad_norm": 1.355653115353838, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 28431 + }, + { + "epoch": 0.28432, + "grad_norm": 1.4815837669425533, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 28432 + }, + { + "epoch": 0.28433, + "grad_norm": 1.8320422079566916, + "learning_rate": 0.003, + "loss": 4.058, + "step": 28433 + }, + { + "epoch": 0.28434, + "grad_norm": 1.1921107470272438, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28434 + }, + { + "epoch": 0.28435, + "grad_norm": 1.5292372124018476, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 28435 + }, + { + "epoch": 0.28436, + "grad_norm": 1.4706100857206361, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 28436 + }, + { + "epoch": 0.28437, + "grad_norm": 1.2759918455981099, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 28437 + }, + { + "epoch": 0.28438, + "grad_norm": 1.4302238539730283, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 28438 + }, + { + "epoch": 0.28439, + "grad_norm": 1.4155282653628845, + "learning_rate": 0.003, + "loss": 4.064, + "step": 28439 + }, + { + "epoch": 0.2844, + "grad_norm": 1.452152521980391, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 28440 + }, + { + "epoch": 0.28441, + "grad_norm": 1.346444154522411, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28441 + }, + { + "epoch": 0.28442, + "grad_norm": 1.4640242093532279, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28442 + }, + { + "epoch": 0.28443, + "grad_norm": 1.5013173642035724, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 28443 + }, + { + "epoch": 0.28444, + "grad_norm": 1.3195583002158637, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28444 + }, + { + "epoch": 0.28445, + "grad_norm": 1.6991930728925562, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 28445 + }, + { + "epoch": 0.28446, + "grad_norm": 1.3141843883164201, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 28446 + }, + { + "epoch": 0.28447, + "grad_norm": 1.5081324430892442, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 28447 + }, + { + "epoch": 0.28448, + "grad_norm": 1.4934082528076664, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 28448 + }, + { + "epoch": 0.28449, + "grad_norm": 1.3318961799529008, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 28449 + }, + { + "epoch": 0.2845, + "grad_norm": 1.3624889690071849, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 28450 + }, + { + "epoch": 0.28451, + "grad_norm": 1.296613580394641, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 28451 + }, + { + "epoch": 0.28452, + "grad_norm": 1.309121334762252, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28452 + }, + { + "epoch": 0.28453, + "grad_norm": 1.3611694511277441, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28453 + }, + { + "epoch": 0.28454, + "grad_norm": 1.4829594918828812, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28454 + }, + { + "epoch": 0.28455, + "grad_norm": 1.215546773800007, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 28455 + }, + { + "epoch": 0.28456, + "grad_norm": 1.8866965237603528, + "learning_rate": 0.003, + "loss": 4.048, + "step": 28456 + }, + { + "epoch": 0.28457, + "grad_norm": 1.4847551431911872, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 28457 + }, + { + "epoch": 0.28458, + "grad_norm": 1.8111507856234883, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 28458 + }, + { + "epoch": 0.28459, + "grad_norm": 1.2034643938431917, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 28459 + }, + { + "epoch": 0.2846, + "grad_norm": 1.3918643196304725, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 28460 + }, + { + "epoch": 0.28461, + "grad_norm": 1.4986541753883533, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 28461 + }, + { + "epoch": 0.28462, + "grad_norm": 1.630927431537442, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 28462 + }, + { + "epoch": 0.28463, + "grad_norm": 1.5130182466397852, + "learning_rate": 0.003, + "loss": 4.028, + "step": 28463 + }, + { + "epoch": 0.28464, + "grad_norm": 1.276759382082326, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 28464 + }, + { + "epoch": 0.28465, + "grad_norm": 1.6122696111808585, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 28465 + }, + { + "epoch": 0.28466, + "grad_norm": 1.2538105422601387, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 28466 + }, + { + "epoch": 0.28467, + "grad_norm": 1.615024359460331, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28467 + }, + { + "epoch": 0.28468, + "grad_norm": 1.190607551491421, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 28468 + }, + { + "epoch": 0.28469, + "grad_norm": 1.6816446235926956, + "learning_rate": 0.003, + "loss": 4.027, + "step": 28469 + }, + { + "epoch": 0.2847, + "grad_norm": 1.2712561087449235, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 28470 + }, + { + "epoch": 0.28471, + "grad_norm": 1.381610187061336, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 28471 + }, + { + "epoch": 0.28472, + "grad_norm": 1.3311151686794656, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 28472 + }, + { + "epoch": 0.28473, + "grad_norm": 1.4632854239900808, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28473 + }, + { + "epoch": 0.28474, + "grad_norm": 1.390755809906238, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28474 + }, + { + "epoch": 0.28475, + "grad_norm": 1.6315198089819118, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 28475 + }, + { + "epoch": 0.28476, + "grad_norm": 1.4630255548770528, + "learning_rate": 0.003, + "loss": 4.043, + "step": 28476 + }, + { + "epoch": 0.28477, + "grad_norm": 1.505366685370031, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 28477 + }, + { + "epoch": 0.28478, + "grad_norm": 1.3383941633647074, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 28478 + }, + { + "epoch": 0.28479, + "grad_norm": 1.4265885365104842, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 28479 + }, + { + "epoch": 0.2848, + "grad_norm": 1.214707817310605, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 28480 + }, + { + "epoch": 0.28481, + "grad_norm": 1.56625070791729, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 28481 + }, + { + "epoch": 0.28482, + "grad_norm": 1.4155458845421929, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 28482 + }, + { + "epoch": 0.28483, + "grad_norm": 1.446621939560558, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28483 + }, + { + "epoch": 0.28484, + "grad_norm": 1.341928053234143, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28484 + }, + { + "epoch": 0.28485, + "grad_norm": 1.536378541356296, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 28485 + }, + { + "epoch": 0.28486, + "grad_norm": 1.339050626152325, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 28486 + }, + { + "epoch": 0.28487, + "grad_norm": 1.599677301168698, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 28487 + }, + { + "epoch": 0.28488, + "grad_norm": 1.5168128940702168, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 28488 + }, + { + "epoch": 0.28489, + "grad_norm": 1.3993123462379335, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28489 + }, + { + "epoch": 0.2849, + "grad_norm": 1.3386792329727681, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28490 + }, + { + "epoch": 0.28491, + "grad_norm": 1.4789498544471147, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 28491 + }, + { + "epoch": 0.28492, + "grad_norm": 1.0583236692548652, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 28492 + }, + { + "epoch": 0.28493, + "grad_norm": 1.6718673141925822, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 28493 + }, + { + "epoch": 0.28494, + "grad_norm": 1.5297102642327054, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 28494 + }, + { + "epoch": 0.28495, + "grad_norm": 1.3058340967795652, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 28495 + }, + { + "epoch": 0.28496, + "grad_norm": 1.2377768960121154, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 28496 + }, + { + "epoch": 0.28497, + "grad_norm": 1.6033974639248756, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 28497 + }, + { + "epoch": 0.28498, + "grad_norm": 1.3717589438501936, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 28498 + }, + { + "epoch": 0.28499, + "grad_norm": 1.5836571418644048, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 28499 + }, + { + "epoch": 0.285, + "grad_norm": 1.3936922297865229, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 28500 + }, + { + "epoch": 0.28501, + "grad_norm": 1.6776279095530382, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 28501 + }, + { + "epoch": 0.28502, + "grad_norm": 1.4341278757009417, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28502 + }, + { + "epoch": 0.28503, + "grad_norm": 2.0199447655484613, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 28503 + }, + { + "epoch": 0.28504, + "grad_norm": 1.5019852993991938, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 28504 + }, + { + "epoch": 0.28505, + "grad_norm": 1.1559888817461215, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 28505 + }, + { + "epoch": 0.28506, + "grad_norm": 1.5963045203527984, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28506 + }, + { + "epoch": 0.28507, + "grad_norm": 1.3250117894150448, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 28507 + }, + { + "epoch": 0.28508, + "grad_norm": 1.2413992315748508, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 28508 + }, + { + "epoch": 0.28509, + "grad_norm": 1.7359542127412444, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 28509 + }, + { + "epoch": 0.2851, + "grad_norm": 1.3483876901681986, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 28510 + }, + { + "epoch": 0.28511, + "grad_norm": 1.5146267706435483, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 28511 + }, + { + "epoch": 0.28512, + "grad_norm": 1.3128195467749888, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 28512 + }, + { + "epoch": 0.28513, + "grad_norm": 1.500225081702982, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28513 + }, + { + "epoch": 0.28514, + "grad_norm": 1.4295000969940344, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 28514 + }, + { + "epoch": 0.28515, + "grad_norm": 1.408978506648195, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 28515 + }, + { + "epoch": 0.28516, + "grad_norm": 1.5817252510856057, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 28516 + }, + { + "epoch": 0.28517, + "grad_norm": 1.2419166210041264, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 28517 + }, + { + "epoch": 0.28518, + "grad_norm": 1.479316196061784, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 28518 + }, + { + "epoch": 0.28519, + "grad_norm": 1.2769382463035746, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 28519 + }, + { + "epoch": 0.2852, + "grad_norm": 1.1483592845857908, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28520 + }, + { + "epoch": 0.28521, + "grad_norm": 1.813155011844083, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 28521 + }, + { + "epoch": 0.28522, + "grad_norm": 1.656941496599643, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28522 + }, + { + "epoch": 0.28523, + "grad_norm": 1.3447217152010331, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 28523 + }, + { + "epoch": 0.28524, + "grad_norm": 1.9324327852461156, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 28524 + }, + { + "epoch": 0.28525, + "grad_norm": 1.1580475794158924, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 28525 + }, + { + "epoch": 0.28526, + "grad_norm": 1.8404143924920862, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28526 + }, + { + "epoch": 0.28527, + "grad_norm": 1.3237864496289198, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 28527 + }, + { + "epoch": 0.28528, + "grad_norm": 1.5589608499421355, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 28528 + }, + { + "epoch": 0.28529, + "grad_norm": 1.3735416334083808, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 28529 + }, + { + "epoch": 0.2853, + "grad_norm": 1.5095241619323354, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28530 + }, + { + "epoch": 0.28531, + "grad_norm": 1.2434588056179077, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 28531 + }, + { + "epoch": 0.28532, + "grad_norm": 1.469791127707363, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 28532 + }, + { + "epoch": 0.28533, + "grad_norm": 1.4833464430615733, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 28533 + }, + { + "epoch": 0.28534, + "grad_norm": 1.392791450946078, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 28534 + }, + { + "epoch": 0.28535, + "grad_norm": 1.4995080443785558, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 28535 + }, + { + "epoch": 0.28536, + "grad_norm": 1.5494245609934516, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 28536 + }, + { + "epoch": 0.28537, + "grad_norm": 1.1946364413120574, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 28537 + }, + { + "epoch": 0.28538, + "grad_norm": 1.6117495082789655, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 28538 + }, + { + "epoch": 0.28539, + "grad_norm": 1.4018868437957246, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28539 + }, + { + "epoch": 0.2854, + "grad_norm": 1.422804970448077, + "learning_rate": 0.003, + "loss": 4.049, + "step": 28540 + }, + { + "epoch": 0.28541, + "grad_norm": 1.4941542329604338, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 28541 + }, + { + "epoch": 0.28542, + "grad_norm": 1.0781721827591855, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28542 + }, + { + "epoch": 0.28543, + "grad_norm": 1.5838917940878074, + "learning_rate": 0.003, + "loss": 4.066, + "step": 28543 + }, + { + "epoch": 0.28544, + "grad_norm": 1.216741706603079, + "learning_rate": 0.003, + "loss": 4.051, + "step": 28544 + }, + { + "epoch": 0.28545, + "grad_norm": 1.6095025097641433, + "learning_rate": 0.003, + "loss": 4.056, + "step": 28545 + }, + { + "epoch": 0.28546, + "grad_norm": 1.4887671213107323, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28546 + }, + { + "epoch": 0.28547, + "grad_norm": 1.295551974445617, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 28547 + }, + { + "epoch": 0.28548, + "grad_norm": 1.4054606783513615, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 28548 + }, + { + "epoch": 0.28549, + "grad_norm": 1.3189682845518989, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 28549 + }, + { + "epoch": 0.2855, + "grad_norm": 1.3273247022289598, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 28550 + }, + { + "epoch": 0.28551, + "grad_norm": 1.529482762935456, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28551 + }, + { + "epoch": 0.28552, + "grad_norm": 1.3993724578304805, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 28552 + }, + { + "epoch": 0.28553, + "grad_norm": 1.2870503764599177, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 28553 + }, + { + "epoch": 0.28554, + "grad_norm": 1.532081458500832, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 28554 + }, + { + "epoch": 0.28555, + "grad_norm": 1.3226888845225935, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28555 + }, + { + "epoch": 0.28556, + "grad_norm": 1.6652718237835702, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 28556 + }, + { + "epoch": 0.28557, + "grad_norm": 1.4842308689719053, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28557 + }, + { + "epoch": 0.28558, + "grad_norm": 1.7071712224060862, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28558 + }, + { + "epoch": 0.28559, + "grad_norm": 1.3970897045095265, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 28559 + }, + { + "epoch": 0.2856, + "grad_norm": 1.5985052038465093, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28560 + }, + { + "epoch": 0.28561, + "grad_norm": 1.2837836974051042, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 28561 + }, + { + "epoch": 0.28562, + "grad_norm": 1.6910492592579178, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28562 + }, + { + "epoch": 0.28563, + "grad_norm": 1.3222977506593254, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28563 + }, + { + "epoch": 0.28564, + "grad_norm": 1.460095250076089, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28564 + }, + { + "epoch": 0.28565, + "grad_norm": 1.3140937534741497, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 28565 + }, + { + "epoch": 0.28566, + "grad_norm": 1.3461340857950526, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 28566 + }, + { + "epoch": 0.28567, + "grad_norm": 1.4488262796589773, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 28567 + }, + { + "epoch": 0.28568, + "grad_norm": 1.3647010140103029, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28568 + }, + { + "epoch": 0.28569, + "grad_norm": 1.4135983015919227, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 28569 + }, + { + "epoch": 0.2857, + "grad_norm": 1.3058164297662085, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28570 + }, + { + "epoch": 0.28571, + "grad_norm": 1.6748798136584804, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 28571 + }, + { + "epoch": 0.28572, + "grad_norm": 1.0805361810037497, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28572 + }, + { + "epoch": 0.28573, + "grad_norm": 1.5266690797488038, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 28573 + }, + { + "epoch": 0.28574, + "grad_norm": 1.3911097660718705, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 28574 + }, + { + "epoch": 0.28575, + "grad_norm": 1.1609447113966898, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 28575 + }, + { + "epoch": 0.28576, + "grad_norm": 1.5916651674463018, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28576 + }, + { + "epoch": 0.28577, + "grad_norm": 1.1788051845890168, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 28577 + }, + { + "epoch": 0.28578, + "grad_norm": 1.6498187812331018, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 28578 + }, + { + "epoch": 0.28579, + "grad_norm": 1.2518361995837206, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 28579 + }, + { + "epoch": 0.2858, + "grad_norm": 1.7848715555135197, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 28580 + }, + { + "epoch": 0.28581, + "grad_norm": 1.3056819944067417, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 28581 + }, + { + "epoch": 0.28582, + "grad_norm": 1.5378529866978023, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 28582 + }, + { + "epoch": 0.28583, + "grad_norm": 1.316509864429928, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 28583 + }, + { + "epoch": 0.28584, + "grad_norm": 1.4541553627546877, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 28584 + }, + { + "epoch": 0.28585, + "grad_norm": 1.3011484945692864, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28585 + }, + { + "epoch": 0.28586, + "grad_norm": 1.5269842551954684, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 28586 + }, + { + "epoch": 0.28587, + "grad_norm": 1.4470811899304001, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28587 + }, + { + "epoch": 0.28588, + "grad_norm": 1.6303343560014427, + "learning_rate": 0.003, + "loss": 4.041, + "step": 28588 + }, + { + "epoch": 0.28589, + "grad_norm": 1.3109616976882155, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 28589 + }, + { + "epoch": 0.2859, + "grad_norm": 1.514016956336778, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 28590 + }, + { + "epoch": 0.28591, + "grad_norm": 1.360280971704477, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28591 + }, + { + "epoch": 0.28592, + "grad_norm": 1.5456497794640875, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 28592 + }, + { + "epoch": 0.28593, + "grad_norm": 1.1946753143313622, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28593 + }, + { + "epoch": 0.28594, + "grad_norm": 1.5094978470599765, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 28594 + }, + { + "epoch": 0.28595, + "grad_norm": 1.4033698631260234, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 28595 + }, + { + "epoch": 0.28596, + "grad_norm": 1.3810078237260124, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 28596 + }, + { + "epoch": 0.28597, + "grad_norm": 1.4782523712968452, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 28597 + }, + { + "epoch": 0.28598, + "grad_norm": 1.3069810583917705, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28598 + }, + { + "epoch": 0.28599, + "grad_norm": 1.4839805149432772, + "learning_rate": 0.003, + "loss": 4.022, + "step": 28599 + }, + { + "epoch": 0.286, + "grad_norm": 1.744030454788103, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 28600 + }, + { + "epoch": 0.28601, + "grad_norm": 1.3044957506931976, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 28601 + }, + { + "epoch": 0.28602, + "grad_norm": 1.5508652567677712, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 28602 + }, + { + "epoch": 0.28603, + "grad_norm": 1.3840374148526744, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 28603 + }, + { + "epoch": 0.28604, + "grad_norm": 1.2604465618391616, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 28604 + }, + { + "epoch": 0.28605, + "grad_norm": 1.508120971205373, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 28605 + }, + { + "epoch": 0.28606, + "grad_norm": 1.2635458270527637, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 28606 + }, + { + "epoch": 0.28607, + "grad_norm": 1.4217353112631201, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 28607 + }, + { + "epoch": 0.28608, + "grad_norm": 1.4851241049915962, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 28608 + }, + { + "epoch": 0.28609, + "grad_norm": 1.296721128940789, + "learning_rate": 0.003, + "loss": 4.055, + "step": 28609 + }, + { + "epoch": 0.2861, + "grad_norm": 1.3291411590831286, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28610 + }, + { + "epoch": 0.28611, + "grad_norm": 1.2655191443836327, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28611 + }, + { + "epoch": 0.28612, + "grad_norm": 1.6373362806814638, + "learning_rate": 0.003, + "loss": 4.062, + "step": 28612 + }, + { + "epoch": 0.28613, + "grad_norm": 1.804729227923244, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 28613 + }, + { + "epoch": 0.28614, + "grad_norm": 1.3494427768373223, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28614 + }, + { + "epoch": 0.28615, + "grad_norm": 1.6109884769536549, + "learning_rate": 0.003, + "loss": 4.039, + "step": 28615 + }, + { + "epoch": 0.28616, + "grad_norm": 1.4533734668489664, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 28616 + }, + { + "epoch": 0.28617, + "grad_norm": 1.2310849750760027, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 28617 + }, + { + "epoch": 0.28618, + "grad_norm": 1.5471430866155227, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28618 + }, + { + "epoch": 0.28619, + "grad_norm": 1.5418969267328984, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 28619 + }, + { + "epoch": 0.2862, + "grad_norm": 1.2397374076789205, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 28620 + }, + { + "epoch": 0.28621, + "grad_norm": 1.472477794900468, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 28621 + }, + { + "epoch": 0.28622, + "grad_norm": 1.3023718271428277, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 28622 + }, + { + "epoch": 0.28623, + "grad_norm": 1.6931091104461524, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 28623 + }, + { + "epoch": 0.28624, + "grad_norm": 1.1774920514461633, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 28624 + }, + { + "epoch": 0.28625, + "grad_norm": 1.4103367705510736, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 28625 + }, + { + "epoch": 0.28626, + "grad_norm": 1.6002049767747277, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 28626 + }, + { + "epoch": 0.28627, + "grad_norm": 1.4756349412173464, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 28627 + }, + { + "epoch": 0.28628, + "grad_norm": 1.2198243674574893, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28628 + }, + { + "epoch": 0.28629, + "grad_norm": 1.7292853774807015, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 28629 + }, + { + "epoch": 0.2863, + "grad_norm": 1.3028303276019446, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 28630 + }, + { + "epoch": 0.28631, + "grad_norm": 1.5070208055599235, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28631 + }, + { + "epoch": 0.28632, + "grad_norm": 1.3373226535043168, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 28632 + }, + { + "epoch": 0.28633, + "grad_norm": 1.646618762133711, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 28633 + }, + { + "epoch": 0.28634, + "grad_norm": 1.1101392539081913, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 28634 + }, + { + "epoch": 0.28635, + "grad_norm": 1.4371472893361419, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 28635 + }, + { + "epoch": 0.28636, + "grad_norm": 1.4021302199193482, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28636 + }, + { + "epoch": 0.28637, + "grad_norm": 1.2869113876607294, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28637 + }, + { + "epoch": 0.28638, + "grad_norm": 1.4376883255120136, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 28638 + }, + { + "epoch": 0.28639, + "grad_norm": 1.517684277836809, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 28639 + }, + { + "epoch": 0.2864, + "grad_norm": 1.4166257838327008, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28640 + }, + { + "epoch": 0.28641, + "grad_norm": 1.5003746016889403, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 28641 + }, + { + "epoch": 0.28642, + "grad_norm": 1.4358134447438822, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 28642 + }, + { + "epoch": 0.28643, + "grad_norm": 1.3923448361901443, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 28643 + }, + { + "epoch": 0.28644, + "grad_norm": 1.3752905440174295, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 28644 + }, + { + "epoch": 0.28645, + "grad_norm": 1.4494590232855353, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 28645 + }, + { + "epoch": 0.28646, + "grad_norm": 1.2482868263732882, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 28646 + }, + { + "epoch": 0.28647, + "grad_norm": 1.554943282583762, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28647 + }, + { + "epoch": 0.28648, + "grad_norm": 1.1598389217891678, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 28648 + }, + { + "epoch": 0.28649, + "grad_norm": 1.7914029551435393, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 28649 + }, + { + "epoch": 0.2865, + "grad_norm": 1.1697390330050597, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 28650 + }, + { + "epoch": 0.28651, + "grad_norm": 1.665894273622613, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28651 + }, + { + "epoch": 0.28652, + "grad_norm": 1.2787014474854037, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28652 + }, + { + "epoch": 0.28653, + "grad_norm": 1.798912111966233, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 28653 + }, + { + "epoch": 0.28654, + "grad_norm": 1.3259160287244391, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 28654 + }, + { + "epoch": 0.28655, + "grad_norm": 1.4455347833363354, + "learning_rate": 0.003, + "loss": 4.031, + "step": 28655 + }, + { + "epoch": 0.28656, + "grad_norm": 1.193792395558412, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 28656 + }, + { + "epoch": 0.28657, + "grad_norm": 1.7032008651679347, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28657 + }, + { + "epoch": 0.28658, + "grad_norm": 1.4224688689926326, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 28658 + }, + { + "epoch": 0.28659, + "grad_norm": 1.6491248857915795, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 28659 + }, + { + "epoch": 0.2866, + "grad_norm": 1.431188607066839, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 28660 + }, + { + "epoch": 0.28661, + "grad_norm": 1.4131847950981462, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 28661 + }, + { + "epoch": 0.28662, + "grad_norm": 1.5728033069266454, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 28662 + }, + { + "epoch": 0.28663, + "grad_norm": 1.4139485752828358, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 28663 + }, + { + "epoch": 0.28664, + "grad_norm": 1.36127736540429, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 28664 + }, + { + "epoch": 0.28665, + "grad_norm": 1.6869789315200818, + "learning_rate": 0.003, + "loss": 4.03, + "step": 28665 + }, + { + "epoch": 0.28666, + "grad_norm": 1.2411299499595683, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 28666 + }, + { + "epoch": 0.28667, + "grad_norm": 1.2799004109244037, + "learning_rate": 0.003, + "loss": 4.015, + "step": 28667 + }, + { + "epoch": 0.28668, + "grad_norm": 1.4192967204411646, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 28668 + }, + { + "epoch": 0.28669, + "grad_norm": 1.289787193809837, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 28669 + }, + { + "epoch": 0.2867, + "grad_norm": 1.321988250306071, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28670 + }, + { + "epoch": 0.28671, + "grad_norm": 1.2319787689874042, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 28671 + }, + { + "epoch": 0.28672, + "grad_norm": 1.5568950090240274, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 28672 + }, + { + "epoch": 0.28673, + "grad_norm": 1.5807578046850213, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 28673 + }, + { + "epoch": 0.28674, + "grad_norm": 1.4656360428943953, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 28674 + }, + { + "epoch": 0.28675, + "grad_norm": 1.4516560268211511, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 28675 + }, + { + "epoch": 0.28676, + "grad_norm": 1.2121030159529538, + "learning_rate": 0.003, + "loss": 4.022, + "step": 28676 + }, + { + "epoch": 0.28677, + "grad_norm": 1.5902245722428017, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 28677 + }, + { + "epoch": 0.28678, + "grad_norm": 1.6530843583119257, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28678 + }, + { + "epoch": 0.28679, + "grad_norm": 1.327056842061068, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 28679 + }, + { + "epoch": 0.2868, + "grad_norm": 1.7606444191061303, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 28680 + }, + { + "epoch": 0.28681, + "grad_norm": 1.3483977143338854, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 28681 + }, + { + "epoch": 0.28682, + "grad_norm": 1.3693366424005775, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 28682 + }, + { + "epoch": 0.28683, + "grad_norm": 1.602272890065639, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 28683 + }, + { + "epoch": 0.28684, + "grad_norm": 1.367076627168425, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 28684 + }, + { + "epoch": 0.28685, + "grad_norm": 1.423760253976167, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 28685 + }, + { + "epoch": 0.28686, + "grad_norm": 1.5155486223395447, + "learning_rate": 0.003, + "loss": 4.019, + "step": 28686 + }, + { + "epoch": 0.28687, + "grad_norm": 1.3796751531764144, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 28687 + }, + { + "epoch": 0.28688, + "grad_norm": 1.557065478519261, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 28688 + }, + { + "epoch": 0.28689, + "grad_norm": 1.5224663257341087, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 28689 + }, + { + "epoch": 0.2869, + "grad_norm": 1.5436095634045126, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 28690 + }, + { + "epoch": 0.28691, + "grad_norm": 1.216508522653653, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 28691 + }, + { + "epoch": 0.28692, + "grad_norm": 1.4391509573825927, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 28692 + }, + { + "epoch": 0.28693, + "grad_norm": 1.2826782926891716, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 28693 + }, + { + "epoch": 0.28694, + "grad_norm": 1.7792556332462455, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 28694 + }, + { + "epoch": 0.28695, + "grad_norm": 1.403380553059504, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28695 + }, + { + "epoch": 0.28696, + "grad_norm": 1.6505183617460568, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 28696 + }, + { + "epoch": 0.28697, + "grad_norm": 1.251369409114038, + "learning_rate": 0.003, + "loss": 4.029, + "step": 28697 + }, + { + "epoch": 0.28698, + "grad_norm": 1.69703247389684, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28698 + }, + { + "epoch": 0.28699, + "grad_norm": 1.4049729857454265, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 28699 + }, + { + "epoch": 0.287, + "grad_norm": 1.6352715878640571, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28700 + }, + { + "epoch": 0.28701, + "grad_norm": 1.1862133008379014, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28701 + }, + { + "epoch": 0.28702, + "grad_norm": 1.3626594153668952, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 28702 + }, + { + "epoch": 0.28703, + "grad_norm": 1.362135824576494, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 28703 + }, + { + "epoch": 0.28704, + "grad_norm": 1.3087040635836826, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28704 + }, + { + "epoch": 0.28705, + "grad_norm": 1.3807861722876906, + "learning_rate": 0.003, + "loss": 4.063, + "step": 28705 + }, + { + "epoch": 0.28706, + "grad_norm": 1.3531683515852089, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 28706 + }, + { + "epoch": 0.28707, + "grad_norm": 1.5838930257761903, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 28707 + }, + { + "epoch": 0.28708, + "grad_norm": 1.4285763735714654, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 28708 + }, + { + "epoch": 0.28709, + "grad_norm": 1.360344913723792, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 28709 + }, + { + "epoch": 0.2871, + "grad_norm": 1.593109682225453, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 28710 + }, + { + "epoch": 0.28711, + "grad_norm": 1.3343084826902745, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 28711 + }, + { + "epoch": 0.28712, + "grad_norm": 1.759722128071524, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 28712 + }, + { + "epoch": 0.28713, + "grad_norm": 1.3396616803739312, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 28713 + }, + { + "epoch": 0.28714, + "grad_norm": 1.3851132159719548, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 28714 + }, + { + "epoch": 0.28715, + "grad_norm": 1.6220566717456786, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 28715 + }, + { + "epoch": 0.28716, + "grad_norm": 1.5356173577299168, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 28716 + }, + { + "epoch": 0.28717, + "grad_norm": 1.4095538591437102, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 28717 + }, + { + "epoch": 0.28718, + "grad_norm": 1.3492345357941888, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28718 + }, + { + "epoch": 0.28719, + "grad_norm": 1.423808791584855, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 28719 + }, + { + "epoch": 0.2872, + "grad_norm": 1.4185467888566181, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 28720 + }, + { + "epoch": 0.28721, + "grad_norm": 1.2593543499089002, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 28721 + }, + { + "epoch": 0.28722, + "grad_norm": 1.5911468250659049, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 28722 + }, + { + "epoch": 0.28723, + "grad_norm": 1.2488097245106384, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 28723 + }, + { + "epoch": 0.28724, + "grad_norm": 1.798716899538055, + "learning_rate": 0.003, + "loss": 4.05, + "step": 28724 + }, + { + "epoch": 0.28725, + "grad_norm": 1.4069223432149272, + "learning_rate": 0.003, + "loss": 4.033, + "step": 28725 + }, + { + "epoch": 0.28726, + "grad_norm": 1.8511653774835417, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 28726 + }, + { + "epoch": 0.28727, + "grad_norm": 1.3439015128716358, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 28727 + }, + { + "epoch": 0.28728, + "grad_norm": 1.421331854123661, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28728 + }, + { + "epoch": 0.28729, + "grad_norm": 1.715888007346362, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 28729 + }, + { + "epoch": 0.2873, + "grad_norm": 1.471207591988034, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 28730 + }, + { + "epoch": 0.28731, + "grad_norm": 1.2350199125152486, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 28731 + }, + { + "epoch": 0.28732, + "grad_norm": 1.4259278195543597, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 28732 + }, + { + "epoch": 0.28733, + "grad_norm": 1.486161005620763, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 28733 + }, + { + "epoch": 0.28734, + "grad_norm": 1.2711669269892052, + "learning_rate": 0.003, + "loss": 4.037, + "step": 28734 + }, + { + "epoch": 0.28735, + "grad_norm": 1.3124333824747234, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 28735 + }, + { + "epoch": 0.28736, + "grad_norm": 1.325355716511573, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 28736 + }, + { + "epoch": 0.28737, + "grad_norm": 1.3443420328368918, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 28737 + }, + { + "epoch": 0.28738, + "grad_norm": 1.4104873255592887, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28738 + }, + { + "epoch": 0.28739, + "grad_norm": 1.5047761070246746, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 28739 + }, + { + "epoch": 0.2874, + "grad_norm": 1.5696019829283183, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 28740 + }, + { + "epoch": 0.28741, + "grad_norm": 1.5192821544947805, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 28741 + }, + { + "epoch": 0.28742, + "grad_norm": 1.4678269170025584, + "learning_rate": 0.003, + "loss": 4.038, + "step": 28742 + }, + { + "epoch": 0.28743, + "grad_norm": 1.8850952459935504, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 28743 + }, + { + "epoch": 0.28744, + "grad_norm": 0.9537239771262839, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 28744 + }, + { + "epoch": 0.28745, + "grad_norm": 1.4229170306003067, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28745 + }, + { + "epoch": 0.28746, + "grad_norm": 1.5264567340966526, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28746 + }, + { + "epoch": 0.28747, + "grad_norm": 1.2217932029456677, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 28747 + }, + { + "epoch": 0.28748, + "grad_norm": 1.5294093362568235, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28748 + }, + { + "epoch": 0.28749, + "grad_norm": 1.4903685168457024, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 28749 + }, + { + "epoch": 0.2875, + "grad_norm": 1.630854012553363, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28750 + }, + { + "epoch": 0.28751, + "grad_norm": 1.3151234735118484, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 28751 + }, + { + "epoch": 0.28752, + "grad_norm": 1.4369955176660116, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 28752 + }, + { + "epoch": 0.28753, + "grad_norm": 1.6194524338490488, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 28753 + }, + { + "epoch": 0.28754, + "grad_norm": 1.5809222156700529, + "learning_rate": 0.003, + "loss": 4.051, + "step": 28754 + }, + { + "epoch": 0.28755, + "grad_norm": 1.3562107102702445, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28755 + }, + { + "epoch": 0.28756, + "grad_norm": 1.4948295417953112, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 28756 + }, + { + "epoch": 0.28757, + "grad_norm": 1.3037134089654157, + "learning_rate": 0.003, + "loss": 4.023, + "step": 28757 + }, + { + "epoch": 0.28758, + "grad_norm": 1.4055553380793184, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28758 + }, + { + "epoch": 0.28759, + "grad_norm": 1.612576111626021, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28759 + }, + { + "epoch": 0.2876, + "grad_norm": 1.2008569321500087, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 28760 + }, + { + "epoch": 0.28761, + "grad_norm": 1.6977839140796485, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 28761 + }, + { + "epoch": 0.28762, + "grad_norm": 1.3216031480739037, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 28762 + }, + { + "epoch": 0.28763, + "grad_norm": 1.3716682789787085, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 28763 + }, + { + "epoch": 0.28764, + "grad_norm": 1.5003909038193104, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 28764 + }, + { + "epoch": 0.28765, + "grad_norm": 1.124602976740587, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 28765 + }, + { + "epoch": 0.28766, + "grad_norm": 1.4734684050733624, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 28766 + }, + { + "epoch": 0.28767, + "grad_norm": 1.3656612321469919, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 28767 + }, + { + "epoch": 0.28768, + "grad_norm": 1.584244689925028, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 28768 + }, + { + "epoch": 0.28769, + "grad_norm": 1.2551991100643325, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28769 + }, + { + "epoch": 0.2877, + "grad_norm": 1.6828531049479458, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 28770 + }, + { + "epoch": 0.28771, + "grad_norm": 1.6224217784676633, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 28771 + }, + { + "epoch": 0.28772, + "grad_norm": 1.7850172295661384, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 28772 + }, + { + "epoch": 0.28773, + "grad_norm": 1.4617724790462578, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 28773 + }, + { + "epoch": 0.28774, + "grad_norm": 1.2314320132713186, + "learning_rate": 0.003, + "loss": 4.051, + "step": 28774 + }, + { + "epoch": 0.28775, + "grad_norm": 1.5840819199926572, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28775 + }, + { + "epoch": 0.28776, + "grad_norm": 1.262499344267113, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 28776 + }, + { + "epoch": 0.28777, + "grad_norm": 1.3405243519451175, + "learning_rate": 0.003, + "loss": 4.024, + "step": 28777 + }, + { + "epoch": 0.28778, + "grad_norm": 1.3945613859535941, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 28778 + }, + { + "epoch": 0.28779, + "grad_norm": 1.6362647401717672, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 28779 + }, + { + "epoch": 0.2878, + "grad_norm": 1.449800833426115, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28780 + }, + { + "epoch": 0.28781, + "grad_norm": 1.2415535827596258, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28781 + }, + { + "epoch": 0.28782, + "grad_norm": 1.521646353787801, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 28782 + }, + { + "epoch": 0.28783, + "grad_norm": 1.2735495840825226, + "learning_rate": 0.003, + "loss": 4.027, + "step": 28783 + }, + { + "epoch": 0.28784, + "grad_norm": 1.7847693078595501, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 28784 + }, + { + "epoch": 0.28785, + "grad_norm": 1.3521972484623463, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28785 + }, + { + "epoch": 0.28786, + "grad_norm": 1.5680524917428422, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 28786 + }, + { + "epoch": 0.28787, + "grad_norm": 1.4635063845790155, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 28787 + }, + { + "epoch": 0.28788, + "grad_norm": 1.245150251600708, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 28788 + }, + { + "epoch": 0.28789, + "grad_norm": 1.4250989438551047, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 28789 + }, + { + "epoch": 0.2879, + "grad_norm": 1.2968917852781725, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 28790 + }, + { + "epoch": 0.28791, + "grad_norm": 1.5958748264748708, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 28791 + }, + { + "epoch": 0.28792, + "grad_norm": 0.9821328946220279, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28792 + }, + { + "epoch": 0.28793, + "grad_norm": 1.7070663748211514, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 28793 + }, + { + "epoch": 0.28794, + "grad_norm": 1.2095234050061783, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 28794 + }, + { + "epoch": 0.28795, + "grad_norm": 1.5886017466553106, + "learning_rate": 0.003, + "loss": 4.038, + "step": 28795 + }, + { + "epoch": 0.28796, + "grad_norm": 1.3994033269442914, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28796 + }, + { + "epoch": 0.28797, + "grad_norm": 1.3417287460821838, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 28797 + }, + { + "epoch": 0.28798, + "grad_norm": 1.5167152025571071, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 28798 + }, + { + "epoch": 0.28799, + "grad_norm": 1.71952756766421, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 28799 + }, + { + "epoch": 0.288, + "grad_norm": 2.127786291861802, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 28800 + }, + { + "epoch": 0.28801, + "grad_norm": 1.0225574687641712, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28801 + }, + { + "epoch": 0.28802, + "grad_norm": 1.6749487010957511, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 28802 + }, + { + "epoch": 0.28803, + "grad_norm": 1.2499163663349346, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 28803 + }, + { + "epoch": 0.28804, + "grad_norm": 1.6020988894152568, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 28804 + }, + { + "epoch": 0.28805, + "grad_norm": 1.171212211229508, + "learning_rate": 0.003, + "loss": 4.054, + "step": 28805 + }, + { + "epoch": 0.28806, + "grad_norm": 1.516033864856683, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 28806 + }, + { + "epoch": 0.28807, + "grad_norm": 1.5224691795657581, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 28807 + }, + { + "epoch": 0.28808, + "grad_norm": 1.833587794664921, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 28808 + }, + { + "epoch": 0.28809, + "grad_norm": 1.3790193237096746, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 28809 + }, + { + "epoch": 0.2881, + "grad_norm": 1.3591996668409094, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28810 + }, + { + "epoch": 0.28811, + "grad_norm": 1.4652685016145957, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 28811 + }, + { + "epoch": 0.28812, + "grad_norm": 1.677261851378081, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 28812 + }, + { + "epoch": 0.28813, + "grad_norm": 1.4607619663917992, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 28813 + }, + { + "epoch": 0.28814, + "grad_norm": 1.3070941437037464, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 28814 + }, + { + "epoch": 0.28815, + "grad_norm": 1.4284864359039107, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 28815 + }, + { + "epoch": 0.28816, + "grad_norm": 1.3940370145047578, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 28816 + }, + { + "epoch": 0.28817, + "grad_norm": 1.2238137952508135, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 28817 + }, + { + "epoch": 0.28818, + "grad_norm": 1.6223491496140399, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 28818 + }, + { + "epoch": 0.28819, + "grad_norm": 1.2600586822356665, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 28819 + }, + { + "epoch": 0.2882, + "grad_norm": 1.4896481379268536, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 28820 + }, + { + "epoch": 0.28821, + "grad_norm": 1.2886495476092874, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 28821 + }, + { + "epoch": 0.28822, + "grad_norm": 1.828218169277914, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 28822 + }, + { + "epoch": 0.28823, + "grad_norm": 1.311525335767891, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 28823 + }, + { + "epoch": 0.28824, + "grad_norm": 1.6579300975089541, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 28824 + }, + { + "epoch": 0.28825, + "grad_norm": 1.2066976960578315, + "learning_rate": 0.003, + "loss": 4.053, + "step": 28825 + }, + { + "epoch": 0.28826, + "grad_norm": 1.7327580624163064, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 28826 + }, + { + "epoch": 0.28827, + "grad_norm": 1.2355667249252817, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 28827 + }, + { + "epoch": 0.28828, + "grad_norm": 1.4895633943537705, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 28828 + }, + { + "epoch": 0.28829, + "grad_norm": 1.5198121782517768, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 28829 + }, + { + "epoch": 0.2883, + "grad_norm": 1.4559693080715865, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 28830 + }, + { + "epoch": 0.28831, + "grad_norm": 1.2415758670976433, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28831 + }, + { + "epoch": 0.28832, + "grad_norm": 1.6119841538976996, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28832 + }, + { + "epoch": 0.28833, + "grad_norm": 1.2367667956724449, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 28833 + }, + { + "epoch": 0.28834, + "grad_norm": 1.4521041904431695, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 28834 + }, + { + "epoch": 0.28835, + "grad_norm": 1.3655779742049738, + "learning_rate": 0.003, + "loss": 4.085, + "step": 28835 + }, + { + "epoch": 0.28836, + "grad_norm": 1.2830393261093334, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 28836 + }, + { + "epoch": 0.28837, + "grad_norm": 1.5913295390477484, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28837 + }, + { + "epoch": 0.28838, + "grad_norm": 1.3574681175141403, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 28838 + }, + { + "epoch": 0.28839, + "grad_norm": 1.7601455166062776, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 28839 + }, + { + "epoch": 0.2884, + "grad_norm": 1.2727796157085034, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 28840 + }, + { + "epoch": 0.28841, + "grad_norm": 1.888516399349873, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28841 + }, + { + "epoch": 0.28842, + "grad_norm": 1.0859214743869117, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 28842 + }, + { + "epoch": 0.28843, + "grad_norm": 1.5102464916165097, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28843 + }, + { + "epoch": 0.28844, + "grad_norm": 1.2682153280416422, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 28844 + }, + { + "epoch": 0.28845, + "grad_norm": 1.6814784223359345, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 28845 + }, + { + "epoch": 0.28846, + "grad_norm": 1.2225429341954546, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 28846 + }, + { + "epoch": 0.28847, + "grad_norm": 1.5649470166731871, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 28847 + }, + { + "epoch": 0.28848, + "grad_norm": 1.3683011775587122, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 28848 + }, + { + "epoch": 0.28849, + "grad_norm": 1.331846262082091, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 28849 + }, + { + "epoch": 0.2885, + "grad_norm": 1.5484562311759518, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 28850 + }, + { + "epoch": 0.28851, + "grad_norm": 1.637890072193845, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 28851 + }, + { + "epoch": 0.28852, + "grad_norm": 1.509924203534104, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 28852 + }, + { + "epoch": 0.28853, + "grad_norm": 1.4940222356572925, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 28853 + }, + { + "epoch": 0.28854, + "grad_norm": 1.4114081589427252, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 28854 + }, + { + "epoch": 0.28855, + "grad_norm": 1.3651577531317665, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 28855 + }, + { + "epoch": 0.28856, + "grad_norm": 1.444650479430926, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 28856 + }, + { + "epoch": 0.28857, + "grad_norm": 1.4132803304731714, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 28857 + }, + { + "epoch": 0.28858, + "grad_norm": 1.5174652596667082, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 28858 + }, + { + "epoch": 0.28859, + "grad_norm": 1.4506927389892437, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28859 + }, + { + "epoch": 0.2886, + "grad_norm": 1.3472875996246982, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 28860 + }, + { + "epoch": 0.28861, + "grad_norm": 1.4249997557794551, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 28861 + }, + { + "epoch": 0.28862, + "grad_norm": 1.321040560103138, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 28862 + }, + { + "epoch": 0.28863, + "grad_norm": 1.2876089865763745, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 28863 + }, + { + "epoch": 0.28864, + "grad_norm": 1.2546806626752252, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 28864 + }, + { + "epoch": 0.28865, + "grad_norm": 1.3460673583119565, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28865 + }, + { + "epoch": 0.28866, + "grad_norm": 1.2656641949841494, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 28866 + }, + { + "epoch": 0.28867, + "grad_norm": 1.3532690912978995, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 28867 + }, + { + "epoch": 0.28868, + "grad_norm": 1.4984512021570195, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 28868 + }, + { + "epoch": 0.28869, + "grad_norm": 1.4176564634169044, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 28869 + }, + { + "epoch": 0.2887, + "grad_norm": 1.5451545166253349, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28870 + }, + { + "epoch": 0.28871, + "grad_norm": 1.8804225867572577, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28871 + }, + { + "epoch": 0.28872, + "grad_norm": 1.0868245360960531, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 28872 + }, + { + "epoch": 0.28873, + "grad_norm": 1.424050859945354, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 28873 + }, + { + "epoch": 0.28874, + "grad_norm": 1.2986418157162898, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 28874 + }, + { + "epoch": 0.28875, + "grad_norm": 1.4368494018218163, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 28875 + }, + { + "epoch": 0.28876, + "grad_norm": 1.435296056753356, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 28876 + }, + { + "epoch": 0.28877, + "grad_norm": 1.65529924358904, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 28877 + }, + { + "epoch": 0.28878, + "grad_norm": 1.5214144713602056, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 28878 + }, + { + "epoch": 0.28879, + "grad_norm": 1.4371182023927809, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28879 + }, + { + "epoch": 0.2888, + "grad_norm": 1.4789282959481247, + "learning_rate": 0.003, + "loss": 4.067, + "step": 28880 + }, + { + "epoch": 0.28881, + "grad_norm": 1.3365824755701088, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 28881 + }, + { + "epoch": 0.28882, + "grad_norm": 1.481959730733225, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 28882 + }, + { + "epoch": 0.28883, + "grad_norm": 1.2911958396900456, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28883 + }, + { + "epoch": 0.28884, + "grad_norm": 1.6546265934712774, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 28884 + }, + { + "epoch": 0.28885, + "grad_norm": 1.5522203669203145, + "learning_rate": 0.003, + "loss": 4.052, + "step": 28885 + }, + { + "epoch": 0.28886, + "grad_norm": 1.6265786217773097, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 28886 + }, + { + "epoch": 0.28887, + "grad_norm": 1.2517377126439295, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 28887 + }, + { + "epoch": 0.28888, + "grad_norm": 1.956258743349752, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 28888 + }, + { + "epoch": 0.28889, + "grad_norm": 1.2112680767206345, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 28889 + }, + { + "epoch": 0.2889, + "grad_norm": 1.569622942748367, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 28890 + }, + { + "epoch": 0.28891, + "grad_norm": 1.5543277394195478, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 28891 + }, + { + "epoch": 0.28892, + "grad_norm": 1.3813379350089345, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 28892 + }, + { + "epoch": 0.28893, + "grad_norm": 1.2411676221321781, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 28893 + }, + { + "epoch": 0.28894, + "grad_norm": 1.4848497339894466, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 28894 + }, + { + "epoch": 0.28895, + "grad_norm": 1.2975648165198899, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 28895 + }, + { + "epoch": 0.28896, + "grad_norm": 1.3518462642208235, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 28896 + }, + { + "epoch": 0.28897, + "grad_norm": 1.3012074042453827, + "learning_rate": 0.003, + "loss": 4.046, + "step": 28897 + }, + { + "epoch": 0.28898, + "grad_norm": 1.3277911082154013, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 28898 + }, + { + "epoch": 0.28899, + "grad_norm": 1.33678045940474, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 28899 + }, + { + "epoch": 0.289, + "grad_norm": 1.3383571812950033, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 28900 + }, + { + "epoch": 0.28901, + "grad_norm": 1.323853413512312, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 28901 + }, + { + "epoch": 0.28902, + "grad_norm": 1.439413799450553, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 28902 + }, + { + "epoch": 0.28903, + "grad_norm": 1.5855925679823788, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 28903 + }, + { + "epoch": 0.28904, + "grad_norm": 1.5566167639058013, + "learning_rate": 0.003, + "loss": 4.039, + "step": 28904 + }, + { + "epoch": 0.28905, + "grad_norm": 1.9122243687146092, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 28905 + }, + { + "epoch": 0.28906, + "grad_norm": 1.0743459792488148, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 28906 + }, + { + "epoch": 0.28907, + "grad_norm": 1.3549351744047426, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 28907 + }, + { + "epoch": 0.28908, + "grad_norm": 1.6284135877948673, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 28908 + }, + { + "epoch": 0.28909, + "grad_norm": 1.3522432358064762, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 28909 + }, + { + "epoch": 0.2891, + "grad_norm": 1.6901939842138176, + "learning_rate": 0.003, + "loss": 4.037, + "step": 28910 + }, + { + "epoch": 0.28911, + "grad_norm": 1.3355250649786337, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 28911 + }, + { + "epoch": 0.28912, + "grad_norm": 1.597807608733278, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 28912 + }, + { + "epoch": 0.28913, + "grad_norm": 1.319806302999126, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 28913 + }, + { + "epoch": 0.28914, + "grad_norm": 1.3188860611136748, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 28914 + }, + { + "epoch": 0.28915, + "grad_norm": 1.466714339639197, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 28915 + }, + { + "epoch": 0.28916, + "grad_norm": 1.304901024046907, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 28916 + }, + { + "epoch": 0.28917, + "grad_norm": 1.345314254838694, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 28917 + }, + { + "epoch": 0.28918, + "grad_norm": 1.5083743179759865, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 28918 + }, + { + "epoch": 0.28919, + "grad_norm": 1.3556918651877874, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 28919 + }, + { + "epoch": 0.2892, + "grad_norm": 1.3409315609994605, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28920 + }, + { + "epoch": 0.28921, + "grad_norm": 1.5715707850232332, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 28921 + }, + { + "epoch": 0.28922, + "grad_norm": 1.4972752373703981, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 28922 + }, + { + "epoch": 0.28923, + "grad_norm": 1.2791740229149011, + "learning_rate": 0.003, + "loss": 4.018, + "step": 28923 + }, + { + "epoch": 0.28924, + "grad_norm": 1.566234547531159, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28924 + }, + { + "epoch": 0.28925, + "grad_norm": 1.42261950944752, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28925 + }, + { + "epoch": 0.28926, + "grad_norm": 1.530344707447041, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28926 + }, + { + "epoch": 0.28927, + "grad_norm": 1.3387966948835905, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28927 + }, + { + "epoch": 0.28928, + "grad_norm": 1.5420470145307472, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 28928 + }, + { + "epoch": 0.28929, + "grad_norm": 1.3743797044019535, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 28929 + }, + { + "epoch": 0.2893, + "grad_norm": 1.3478847974873849, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 28930 + }, + { + "epoch": 0.28931, + "grad_norm": 1.6844010511638703, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 28931 + }, + { + "epoch": 0.28932, + "grad_norm": 1.1637487020789825, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 28932 + }, + { + "epoch": 0.28933, + "grad_norm": 1.7943779521039425, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 28933 + }, + { + "epoch": 0.28934, + "grad_norm": 1.4018514123429802, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 28934 + }, + { + "epoch": 0.28935, + "grad_norm": 1.3779233415827727, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 28935 + }, + { + "epoch": 0.28936, + "grad_norm": 1.4298750762456083, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28936 + }, + { + "epoch": 0.28937, + "grad_norm": 1.3902287759920087, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 28937 + }, + { + "epoch": 0.28938, + "grad_norm": 1.5386298933900961, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 28938 + }, + { + "epoch": 0.28939, + "grad_norm": 1.2767443749166796, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 28939 + }, + { + "epoch": 0.2894, + "grad_norm": 1.4642856247384488, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28940 + }, + { + "epoch": 0.28941, + "grad_norm": 1.413728361881422, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 28941 + }, + { + "epoch": 0.28942, + "grad_norm": 1.5826678870110644, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 28942 + }, + { + "epoch": 0.28943, + "grad_norm": 1.359802024942231, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 28943 + }, + { + "epoch": 0.28944, + "grad_norm": 1.3945686970400157, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 28944 + }, + { + "epoch": 0.28945, + "grad_norm": 1.3882314284431216, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 28945 + }, + { + "epoch": 0.28946, + "grad_norm": 1.2529215568269743, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28946 + }, + { + "epoch": 0.28947, + "grad_norm": 1.3868535152562569, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 28947 + }, + { + "epoch": 0.28948, + "grad_norm": 1.3135285076382959, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 28948 + }, + { + "epoch": 0.28949, + "grad_norm": 1.3569911044416947, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 28949 + }, + { + "epoch": 0.2895, + "grad_norm": 1.4114327701353988, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 28950 + }, + { + "epoch": 0.28951, + "grad_norm": 1.392247342984486, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 28951 + }, + { + "epoch": 0.28952, + "grad_norm": 1.4950813817242772, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 28952 + }, + { + "epoch": 0.28953, + "grad_norm": 1.2683969061708686, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 28953 + }, + { + "epoch": 0.28954, + "grad_norm": 1.5583799720984444, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 28954 + }, + { + "epoch": 0.28955, + "grad_norm": 1.365905471062512, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 28955 + }, + { + "epoch": 0.28956, + "grad_norm": 1.37723280372274, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 28956 + }, + { + "epoch": 0.28957, + "grad_norm": 1.7616520370309348, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 28957 + }, + { + "epoch": 0.28958, + "grad_norm": 1.499662721652096, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 28958 + }, + { + "epoch": 0.28959, + "grad_norm": 1.3495614039502664, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 28959 + }, + { + "epoch": 0.2896, + "grad_norm": 1.4845214785565177, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 28960 + }, + { + "epoch": 0.28961, + "grad_norm": 1.5426990576829978, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 28961 + }, + { + "epoch": 0.28962, + "grad_norm": 1.4865126464619722, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 28962 + }, + { + "epoch": 0.28963, + "grad_norm": 1.301049919597042, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 28963 + }, + { + "epoch": 0.28964, + "grad_norm": 1.529292591399979, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 28964 + }, + { + "epoch": 0.28965, + "grad_norm": 1.1971037259309973, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 28965 + }, + { + "epoch": 0.28966, + "grad_norm": 1.5559573156262432, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 28966 + }, + { + "epoch": 0.28967, + "grad_norm": 1.3411950731239324, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28967 + }, + { + "epoch": 0.28968, + "grad_norm": 1.4980305769134368, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 28968 + }, + { + "epoch": 0.28969, + "grad_norm": 1.5646465963442908, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 28969 + }, + { + "epoch": 0.2897, + "grad_norm": 1.7760063850800323, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 28970 + }, + { + "epoch": 0.28971, + "grad_norm": 1.33291664124801, + "learning_rate": 0.003, + "loss": 4.061, + "step": 28971 + }, + { + "epoch": 0.28972, + "grad_norm": 1.5769197985666463, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28972 + }, + { + "epoch": 0.28973, + "grad_norm": 1.3903166407845922, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 28973 + }, + { + "epoch": 0.28974, + "grad_norm": 1.2306616191917736, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 28974 + }, + { + "epoch": 0.28975, + "grad_norm": 1.5542590743554496, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 28975 + }, + { + "epoch": 0.28976, + "grad_norm": 1.2069105213561828, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 28976 + }, + { + "epoch": 0.28977, + "grad_norm": 1.6780754868220336, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 28977 + }, + { + "epoch": 0.28978, + "grad_norm": 1.4547166043006323, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 28978 + }, + { + "epoch": 0.28979, + "grad_norm": 1.4020918940915232, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 28979 + }, + { + "epoch": 0.2898, + "grad_norm": 1.5738844964754353, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28980 + }, + { + "epoch": 0.28981, + "grad_norm": 1.2632591997437783, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 28981 + }, + { + "epoch": 0.28982, + "grad_norm": 1.2743856861983411, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 28982 + }, + { + "epoch": 0.28983, + "grad_norm": 1.507469097023979, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 28983 + }, + { + "epoch": 0.28984, + "grad_norm": 1.2527084576145429, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 28984 + }, + { + "epoch": 0.28985, + "grad_norm": 1.473701436217265, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 28985 + }, + { + "epoch": 0.28986, + "grad_norm": 1.349013397008001, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 28986 + }, + { + "epoch": 0.28987, + "grad_norm": 1.4918964967258264, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 28987 + }, + { + "epoch": 0.28988, + "grad_norm": 1.4664313101364448, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28988 + }, + { + "epoch": 0.28989, + "grad_norm": 1.578078919667235, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 28989 + }, + { + "epoch": 0.2899, + "grad_norm": 1.3677018652102884, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 28990 + }, + { + "epoch": 0.28991, + "grad_norm": 1.4782845087015777, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 28991 + }, + { + "epoch": 0.28992, + "grad_norm": 1.3476999276138517, + "learning_rate": 0.003, + "loss": 4.057, + "step": 28992 + }, + { + "epoch": 0.28993, + "grad_norm": 1.5234044733662664, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28993 + }, + { + "epoch": 0.28994, + "grad_norm": 1.1316057564615376, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 28994 + }, + { + "epoch": 0.28995, + "grad_norm": 1.6543391057249897, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 28995 + }, + { + "epoch": 0.28996, + "grad_norm": 1.3625270168772567, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 28996 + }, + { + "epoch": 0.28997, + "grad_norm": 1.6512520360487193, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 28997 + }, + { + "epoch": 0.28998, + "grad_norm": 1.3298459725862253, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 28998 + }, + { + "epoch": 0.28999, + "grad_norm": 1.5226531890032038, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 28999 + }, + { + "epoch": 0.29, + "grad_norm": 1.4304909311379896, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29000 + }, + { + "epoch": 0.29001, + "grad_norm": 1.528937487144021, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 29001 + }, + { + "epoch": 0.29002, + "grad_norm": 1.2899448441957753, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 29002 + }, + { + "epoch": 0.29003, + "grad_norm": 1.7151652793580388, + "learning_rate": 0.003, + "loss": 4.038, + "step": 29003 + }, + { + "epoch": 0.29004, + "grad_norm": 1.3836938626456252, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29004 + }, + { + "epoch": 0.29005, + "grad_norm": 1.3223841165417813, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 29005 + }, + { + "epoch": 0.29006, + "grad_norm": 1.515032486664928, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 29006 + }, + { + "epoch": 0.29007, + "grad_norm": 1.310015299110832, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 29007 + }, + { + "epoch": 0.29008, + "grad_norm": 1.4373369218079608, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29008 + }, + { + "epoch": 0.29009, + "grad_norm": 1.4366406842084105, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 29009 + }, + { + "epoch": 0.2901, + "grad_norm": 1.5176499704522561, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 29010 + }, + { + "epoch": 0.29011, + "grad_norm": 1.5691560662619404, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29011 + }, + { + "epoch": 0.29012, + "grad_norm": 1.3281197999203564, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 29012 + }, + { + "epoch": 0.29013, + "grad_norm": 1.3622782689300663, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29013 + }, + { + "epoch": 0.29014, + "grad_norm": 1.4090455186544242, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 29014 + }, + { + "epoch": 0.29015, + "grad_norm": 1.3056676168655994, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 29015 + }, + { + "epoch": 0.29016, + "grad_norm": 1.6407946221327931, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29016 + }, + { + "epoch": 0.29017, + "grad_norm": 1.3672438940776872, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29017 + }, + { + "epoch": 0.29018, + "grad_norm": 1.5037648563969876, + "learning_rate": 0.003, + "loss": 4.033, + "step": 29018 + }, + { + "epoch": 0.29019, + "grad_norm": 1.2587844875996934, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 29019 + }, + { + "epoch": 0.2902, + "grad_norm": 1.5713270457437452, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 29020 + }, + { + "epoch": 0.29021, + "grad_norm": 1.3733083800289922, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 29021 + }, + { + "epoch": 0.29022, + "grad_norm": 1.6290796920230852, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 29022 + }, + { + "epoch": 0.29023, + "grad_norm": 1.2955119807180406, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 29023 + }, + { + "epoch": 0.29024, + "grad_norm": 1.2671200311649493, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29024 + }, + { + "epoch": 0.29025, + "grad_norm": 1.6010729689009553, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29025 + }, + { + "epoch": 0.29026, + "grad_norm": 1.1558421653541773, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29026 + }, + { + "epoch": 0.29027, + "grad_norm": 1.5526105340282617, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 29027 + }, + { + "epoch": 0.29028, + "grad_norm": 1.4225969840489554, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 29028 + }, + { + "epoch": 0.29029, + "grad_norm": 1.4818298681299353, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 29029 + }, + { + "epoch": 0.2903, + "grad_norm": 1.4338584638087397, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29030 + }, + { + "epoch": 0.29031, + "grad_norm": 1.6046713670061263, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 29031 + }, + { + "epoch": 0.29032, + "grad_norm": 1.3123132536384643, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 29032 + }, + { + "epoch": 0.29033, + "grad_norm": 1.712399336284405, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 29033 + }, + { + "epoch": 0.29034, + "grad_norm": 1.37425331724171, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 29034 + }, + { + "epoch": 0.29035, + "grad_norm": 1.5937878870973574, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29035 + }, + { + "epoch": 0.29036, + "grad_norm": 1.204977391611658, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29036 + }, + { + "epoch": 0.29037, + "grad_norm": 1.311487305952911, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 29037 + }, + { + "epoch": 0.29038, + "grad_norm": 1.9893631673016672, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 29038 + }, + { + "epoch": 0.29039, + "grad_norm": 1.4462043662475816, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 29039 + }, + { + "epoch": 0.2904, + "grad_norm": 1.4749966591883759, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 29040 + }, + { + "epoch": 0.29041, + "grad_norm": 1.4068209085122128, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 29041 + }, + { + "epoch": 0.29042, + "grad_norm": 1.415650644235765, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 29042 + }, + { + "epoch": 0.29043, + "grad_norm": 1.4268637160977122, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 29043 + }, + { + "epoch": 0.29044, + "grad_norm": 1.354901685763752, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 29044 + }, + { + "epoch": 0.29045, + "grad_norm": 1.4621528035746845, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29045 + }, + { + "epoch": 0.29046, + "grad_norm": 1.560964805817243, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29046 + }, + { + "epoch": 0.29047, + "grad_norm": 1.7003725233496854, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 29047 + }, + { + "epoch": 0.29048, + "grad_norm": 1.5647538987723772, + "learning_rate": 0.003, + "loss": 4.064, + "step": 29048 + }, + { + "epoch": 0.29049, + "grad_norm": 1.382938934933747, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 29049 + }, + { + "epoch": 0.2905, + "grad_norm": 1.6924384629906915, + "learning_rate": 0.003, + "loss": 4.084, + "step": 29050 + }, + { + "epoch": 0.29051, + "grad_norm": 1.481601980928585, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29051 + }, + { + "epoch": 0.29052, + "grad_norm": 1.4518607353658501, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29052 + }, + { + "epoch": 0.29053, + "grad_norm": 1.5030369125650878, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 29053 + }, + { + "epoch": 0.29054, + "grad_norm": 1.187915013356685, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 29054 + }, + { + "epoch": 0.29055, + "grad_norm": 1.4105459872294703, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 29055 + }, + { + "epoch": 0.29056, + "grad_norm": 1.3850926433115658, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 29056 + }, + { + "epoch": 0.29057, + "grad_norm": 1.5375803272447097, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 29057 + }, + { + "epoch": 0.29058, + "grad_norm": 1.591284347582257, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 29058 + }, + { + "epoch": 0.29059, + "grad_norm": 1.2415342477089353, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 29059 + }, + { + "epoch": 0.2906, + "grad_norm": 1.6462862634593147, + "learning_rate": 0.003, + "loss": 4.023, + "step": 29060 + }, + { + "epoch": 0.29061, + "grad_norm": 1.289670762660021, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29061 + }, + { + "epoch": 0.29062, + "grad_norm": 1.5040858642777417, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 29062 + }, + { + "epoch": 0.29063, + "grad_norm": 1.2237942617874145, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 29063 + }, + { + "epoch": 0.29064, + "grad_norm": 1.6478428702913688, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29064 + }, + { + "epoch": 0.29065, + "grad_norm": 1.4784314052326142, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 29065 + }, + { + "epoch": 0.29066, + "grad_norm": 1.4456741560421627, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 29066 + }, + { + "epoch": 0.29067, + "grad_norm": 1.5532782295548864, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 29067 + }, + { + "epoch": 0.29068, + "grad_norm": 1.5578702868392185, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 29068 + }, + { + "epoch": 0.29069, + "grad_norm": 1.21209443494854, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 29069 + }, + { + "epoch": 0.2907, + "grad_norm": 1.4964154976185173, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29070 + }, + { + "epoch": 0.29071, + "grad_norm": 1.4493024910285408, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29071 + }, + { + "epoch": 0.29072, + "grad_norm": 1.435891513224366, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29072 + }, + { + "epoch": 0.29073, + "grad_norm": 1.3040856441633415, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29073 + }, + { + "epoch": 0.29074, + "grad_norm": 1.3913605650106222, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29074 + }, + { + "epoch": 0.29075, + "grad_norm": 1.3325758434203288, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 29075 + }, + { + "epoch": 0.29076, + "grad_norm": 1.3564824728209517, + "learning_rate": 0.003, + "loss": 4.053, + "step": 29076 + }, + { + "epoch": 0.29077, + "grad_norm": 1.3209970128321953, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29077 + }, + { + "epoch": 0.29078, + "grad_norm": 1.3440312038505038, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 29078 + }, + { + "epoch": 0.29079, + "grad_norm": 1.5743450030685682, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29079 + }, + { + "epoch": 0.2908, + "grad_norm": 1.3239415102753909, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 29080 + }, + { + "epoch": 0.29081, + "grad_norm": 1.559729880427107, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 29081 + }, + { + "epoch": 0.29082, + "grad_norm": 1.5736461588916608, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 29082 + }, + { + "epoch": 0.29083, + "grad_norm": 1.5192078628768217, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 29083 + }, + { + "epoch": 0.29084, + "grad_norm": 1.2540065303283823, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 29084 + }, + { + "epoch": 0.29085, + "grad_norm": 1.6239725123573567, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29085 + }, + { + "epoch": 0.29086, + "grad_norm": 1.39532971000719, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29086 + }, + { + "epoch": 0.29087, + "grad_norm": 1.5029759717447226, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29087 + }, + { + "epoch": 0.29088, + "grad_norm": 1.4156487329901313, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 29088 + }, + { + "epoch": 0.29089, + "grad_norm": 1.3128575309887225, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 29089 + }, + { + "epoch": 0.2909, + "grad_norm": 1.2764304445107852, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29090 + }, + { + "epoch": 0.29091, + "grad_norm": 1.2784093924297166, + "learning_rate": 0.003, + "loss": 4.032, + "step": 29091 + }, + { + "epoch": 0.29092, + "grad_norm": 1.457159153647996, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29092 + }, + { + "epoch": 0.29093, + "grad_norm": 1.4443951203012113, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 29093 + }, + { + "epoch": 0.29094, + "grad_norm": 1.508215652047, + "learning_rate": 0.003, + "loss": 4.038, + "step": 29094 + }, + { + "epoch": 0.29095, + "grad_norm": 1.5120702041857133, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29095 + }, + { + "epoch": 0.29096, + "grad_norm": 1.3206133447819606, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29096 + }, + { + "epoch": 0.29097, + "grad_norm": 1.6542624706079627, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 29097 + }, + { + "epoch": 0.29098, + "grad_norm": 1.4943009261988613, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 29098 + }, + { + "epoch": 0.29099, + "grad_norm": 1.3510735480323721, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29099 + }, + { + "epoch": 0.291, + "grad_norm": 1.6679866735827624, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 29100 + }, + { + "epoch": 0.29101, + "grad_norm": 1.1998744355517914, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29101 + }, + { + "epoch": 0.29102, + "grad_norm": 1.3382062811763336, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 29102 + }, + { + "epoch": 0.29103, + "grad_norm": 1.4194216176636374, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 29103 + }, + { + "epoch": 0.29104, + "grad_norm": 1.3755636730325655, + "learning_rate": 0.003, + "loss": 4.014, + "step": 29104 + }, + { + "epoch": 0.29105, + "grad_norm": 1.503608025659624, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 29105 + }, + { + "epoch": 0.29106, + "grad_norm": 1.4544838389436114, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 29106 + }, + { + "epoch": 0.29107, + "grad_norm": 1.9294287383221764, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 29107 + }, + { + "epoch": 0.29108, + "grad_norm": 1.3237338978913413, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 29108 + }, + { + "epoch": 0.29109, + "grad_norm": 1.7971760611955454, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 29109 + }, + { + "epoch": 0.2911, + "grad_norm": 1.2503951598004432, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 29110 + }, + { + "epoch": 0.29111, + "grad_norm": 1.5572079438199895, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 29111 + }, + { + "epoch": 0.29112, + "grad_norm": 1.4536745097691313, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 29112 + }, + { + "epoch": 0.29113, + "grad_norm": 1.6003210451186805, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 29113 + }, + { + "epoch": 0.29114, + "grad_norm": 1.4975689601771516, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 29114 + }, + { + "epoch": 0.29115, + "grad_norm": 1.5419445943135595, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29115 + }, + { + "epoch": 0.29116, + "grad_norm": 1.2845563751430098, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 29116 + }, + { + "epoch": 0.29117, + "grad_norm": 1.3841262048965504, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 29117 + }, + { + "epoch": 0.29118, + "grad_norm": 1.1894378135583639, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29118 + }, + { + "epoch": 0.29119, + "grad_norm": 1.9565641076369094, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 29119 + }, + { + "epoch": 0.2912, + "grad_norm": 1.1755511481898906, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 29120 + }, + { + "epoch": 0.29121, + "grad_norm": 1.7654791216227546, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 29121 + }, + { + "epoch": 0.29122, + "grad_norm": 1.7572928375965988, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29122 + }, + { + "epoch": 0.29123, + "grad_norm": 1.5047711447625207, + "learning_rate": 0.003, + "loss": 4.039, + "step": 29123 + }, + { + "epoch": 0.29124, + "grad_norm": 1.4372296579320274, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29124 + }, + { + "epoch": 0.29125, + "grad_norm": 1.530419521862659, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 29125 + }, + { + "epoch": 0.29126, + "grad_norm": 1.4489897327748396, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29126 + }, + { + "epoch": 0.29127, + "grad_norm": 1.312143785682127, + "learning_rate": 0.003, + "loss": 4.021, + "step": 29127 + }, + { + "epoch": 0.29128, + "grad_norm": 1.3701581264968232, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 29128 + }, + { + "epoch": 0.29129, + "grad_norm": 1.344006778478838, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 29129 + }, + { + "epoch": 0.2913, + "grad_norm": 1.51775721195218, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29130 + }, + { + "epoch": 0.29131, + "grad_norm": 1.4264855312764932, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 29131 + }, + { + "epoch": 0.29132, + "grad_norm": 1.3174660831889227, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 29132 + }, + { + "epoch": 0.29133, + "grad_norm": 1.4207337342520643, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 29133 + }, + { + "epoch": 0.29134, + "grad_norm": 1.3473633528970832, + "learning_rate": 0.003, + "loss": 4.026, + "step": 29134 + }, + { + "epoch": 0.29135, + "grad_norm": 1.6011987042189988, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 29135 + }, + { + "epoch": 0.29136, + "grad_norm": 1.2831047135644653, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 29136 + }, + { + "epoch": 0.29137, + "grad_norm": 1.2874523420341897, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 29137 + }, + { + "epoch": 0.29138, + "grad_norm": 1.616861299652945, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29138 + }, + { + "epoch": 0.29139, + "grad_norm": 1.1998988155641073, + "learning_rate": 0.003, + "loss": 4.059, + "step": 29139 + }, + { + "epoch": 0.2914, + "grad_norm": 1.7494081003709236, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29140 + }, + { + "epoch": 0.29141, + "grad_norm": 1.2790012909146455, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 29141 + }, + { + "epoch": 0.29142, + "grad_norm": 1.9341776449828083, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 29142 + }, + { + "epoch": 0.29143, + "grad_norm": 1.5758579742126078, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29143 + }, + { + "epoch": 0.29144, + "grad_norm": 1.5046727796313175, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 29144 + }, + { + "epoch": 0.29145, + "grad_norm": 1.5702769613932126, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 29145 + }, + { + "epoch": 0.29146, + "grad_norm": 1.183133543199987, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 29146 + }, + { + "epoch": 0.29147, + "grad_norm": 1.7077359348920662, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 29147 + }, + { + "epoch": 0.29148, + "grad_norm": 1.2424389377528142, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 29148 + }, + { + "epoch": 0.29149, + "grad_norm": 1.6231836868656575, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29149 + }, + { + "epoch": 0.2915, + "grad_norm": 1.5602813082066433, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29150 + }, + { + "epoch": 0.29151, + "grad_norm": 1.4683627516896005, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29151 + }, + { + "epoch": 0.29152, + "grad_norm": 1.7394646779194385, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 29152 + }, + { + "epoch": 0.29153, + "grad_norm": 1.1667235960246314, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 29153 + }, + { + "epoch": 0.29154, + "grad_norm": 1.5646393387585849, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 29154 + }, + { + "epoch": 0.29155, + "grad_norm": 1.4900740924314724, + "learning_rate": 0.003, + "loss": 4.037, + "step": 29155 + }, + { + "epoch": 0.29156, + "grad_norm": 1.361609493476467, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 29156 + }, + { + "epoch": 0.29157, + "grad_norm": 1.454952668624488, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29157 + }, + { + "epoch": 0.29158, + "grad_norm": 1.2718989958758724, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 29158 + }, + { + "epoch": 0.29159, + "grad_norm": 1.322743074384142, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29159 + }, + { + "epoch": 0.2916, + "grad_norm": 1.3049085011436306, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 29160 + }, + { + "epoch": 0.29161, + "grad_norm": 1.3427915127141246, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29161 + }, + { + "epoch": 0.29162, + "grad_norm": 1.9572049335491446, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29162 + }, + { + "epoch": 0.29163, + "grad_norm": 1.75494349631373, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29163 + }, + { + "epoch": 0.29164, + "grad_norm": 1.4851848548134488, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 29164 + }, + { + "epoch": 0.29165, + "grad_norm": 1.2804118739232184, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 29165 + }, + { + "epoch": 0.29166, + "grad_norm": 1.8749606743057439, + "learning_rate": 0.003, + "loss": 4.04, + "step": 29166 + }, + { + "epoch": 0.29167, + "grad_norm": 1.062894307336015, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29167 + }, + { + "epoch": 0.29168, + "grad_norm": 1.4699425260986139, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29168 + }, + { + "epoch": 0.29169, + "grad_norm": 1.6615047936135274, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29169 + }, + { + "epoch": 0.2917, + "grad_norm": 1.4668099647081343, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29170 + }, + { + "epoch": 0.29171, + "grad_norm": 1.635968595982996, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29171 + }, + { + "epoch": 0.29172, + "grad_norm": 1.2175991056049043, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 29172 + }, + { + "epoch": 0.29173, + "grad_norm": 1.6179914440041148, + "learning_rate": 0.003, + "loss": 4.029, + "step": 29173 + }, + { + "epoch": 0.29174, + "grad_norm": 1.1980290209935658, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 29174 + }, + { + "epoch": 0.29175, + "grad_norm": 1.488389987498707, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 29175 + }, + { + "epoch": 0.29176, + "grad_norm": 1.2865169870645383, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 29176 + }, + { + "epoch": 0.29177, + "grad_norm": 1.4257416657646222, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 29177 + }, + { + "epoch": 0.29178, + "grad_norm": 1.3564135234383436, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 29178 + }, + { + "epoch": 0.29179, + "grad_norm": 1.42061586805152, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 29179 + }, + { + "epoch": 0.2918, + "grad_norm": 1.7755729823327908, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 29180 + }, + { + "epoch": 0.29181, + "grad_norm": 1.4275675169847282, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 29181 + }, + { + "epoch": 0.29182, + "grad_norm": 1.2771521954035525, + "learning_rate": 0.003, + "loss": 4.045, + "step": 29182 + }, + { + "epoch": 0.29183, + "grad_norm": 1.6218509419892164, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29183 + }, + { + "epoch": 0.29184, + "grad_norm": 1.1634960617158523, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29184 + }, + { + "epoch": 0.29185, + "grad_norm": 1.8517016020738972, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 29185 + }, + { + "epoch": 0.29186, + "grad_norm": 1.2489925294148212, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 29186 + }, + { + "epoch": 0.29187, + "grad_norm": 1.4358558681014166, + "learning_rate": 0.003, + "loss": 4.035, + "step": 29187 + }, + { + "epoch": 0.29188, + "grad_norm": 1.8744106456560374, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 29188 + }, + { + "epoch": 0.29189, + "grad_norm": 1.2620880960696803, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 29189 + }, + { + "epoch": 0.2919, + "grad_norm": 1.48918441808212, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 29190 + }, + { + "epoch": 0.29191, + "grad_norm": 1.4938975502058292, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 29191 + }, + { + "epoch": 0.29192, + "grad_norm": 1.40717773440232, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 29192 + }, + { + "epoch": 0.29193, + "grad_norm": 1.4513410974015684, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29193 + }, + { + "epoch": 0.29194, + "grad_norm": 1.1798005128253686, + "learning_rate": 0.003, + "loss": 4.061, + "step": 29194 + }, + { + "epoch": 0.29195, + "grad_norm": 1.7276209426237885, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 29195 + }, + { + "epoch": 0.29196, + "grad_norm": 1.126750121074111, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 29196 + }, + { + "epoch": 0.29197, + "grad_norm": 1.7759648184630616, + "learning_rate": 0.003, + "loss": 4.059, + "step": 29197 + }, + { + "epoch": 0.29198, + "grad_norm": 1.4473974985237656, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 29198 + }, + { + "epoch": 0.29199, + "grad_norm": 1.5305936984557265, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 29199 + }, + { + "epoch": 0.292, + "grad_norm": 1.6585410146738535, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 29200 + }, + { + "epoch": 0.29201, + "grad_norm": 1.310684060795603, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29201 + }, + { + "epoch": 0.29202, + "grad_norm": 1.6332391200467846, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29202 + }, + { + "epoch": 0.29203, + "grad_norm": 1.1597070557726756, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 29203 + }, + { + "epoch": 0.29204, + "grad_norm": 1.440638395347902, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 29204 + }, + { + "epoch": 0.29205, + "grad_norm": 1.3372459215835815, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29205 + }, + { + "epoch": 0.29206, + "grad_norm": 1.7423556574690864, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 29206 + }, + { + "epoch": 0.29207, + "grad_norm": 1.3066694067867068, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 29207 + }, + { + "epoch": 0.29208, + "grad_norm": 1.659140990322641, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 29208 + }, + { + "epoch": 0.29209, + "grad_norm": 1.2991073254919963, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 29209 + }, + { + "epoch": 0.2921, + "grad_norm": 1.5663158143113427, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29210 + }, + { + "epoch": 0.29211, + "grad_norm": 1.2006043788284337, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 29211 + }, + { + "epoch": 0.29212, + "grad_norm": 1.4484281339161376, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29212 + }, + { + "epoch": 0.29213, + "grad_norm": 1.5022731728346461, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 29213 + }, + { + "epoch": 0.29214, + "grad_norm": 1.1541757643639574, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29214 + }, + { + "epoch": 0.29215, + "grad_norm": 1.489687187475689, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29215 + }, + { + "epoch": 0.29216, + "grad_norm": 1.1158827720476927, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29216 + }, + { + "epoch": 0.29217, + "grad_norm": 1.4508673644996093, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29217 + }, + { + "epoch": 0.29218, + "grad_norm": 1.4144406398032792, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29218 + }, + { + "epoch": 0.29219, + "grad_norm": 1.5546071668042993, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 29219 + }, + { + "epoch": 0.2922, + "grad_norm": 1.6244138633290048, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29220 + }, + { + "epoch": 0.29221, + "grad_norm": 1.4857575334034165, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 29221 + }, + { + "epoch": 0.29222, + "grad_norm": 1.296102528210188, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 29222 + }, + { + "epoch": 0.29223, + "grad_norm": 1.666239501583979, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29223 + }, + { + "epoch": 0.29224, + "grad_norm": 1.308193197276401, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29224 + }, + { + "epoch": 0.29225, + "grad_norm": 1.3923112490058975, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 29225 + }, + { + "epoch": 0.29226, + "grad_norm": 1.3600630713971753, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 29226 + }, + { + "epoch": 0.29227, + "grad_norm": 1.4372881307156415, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 29227 + }, + { + "epoch": 0.29228, + "grad_norm": 1.5903390752452236, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 29228 + }, + { + "epoch": 0.29229, + "grad_norm": 1.305302692771258, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 29229 + }, + { + "epoch": 0.2923, + "grad_norm": 1.639977704015093, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 29230 + }, + { + "epoch": 0.29231, + "grad_norm": 1.2623566275177114, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 29231 + }, + { + "epoch": 0.29232, + "grad_norm": 1.5689139794477556, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 29232 + }, + { + "epoch": 0.29233, + "grad_norm": 1.4070825051564528, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 29233 + }, + { + "epoch": 0.29234, + "grad_norm": 1.4240124681113204, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 29234 + }, + { + "epoch": 0.29235, + "grad_norm": 1.407749094640242, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 29235 + }, + { + "epoch": 0.29236, + "grad_norm": 1.5693620279510192, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 29236 + }, + { + "epoch": 0.29237, + "grad_norm": 1.1930059771745443, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29237 + }, + { + "epoch": 0.29238, + "grad_norm": 1.4026086624294178, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 29238 + }, + { + "epoch": 0.29239, + "grad_norm": 1.493642274251611, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29239 + }, + { + "epoch": 0.2924, + "grad_norm": 1.2262752826851822, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 29240 + }, + { + "epoch": 0.29241, + "grad_norm": 1.8466655662187352, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29241 + }, + { + "epoch": 0.29242, + "grad_norm": 1.39791219065393, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 29242 + }, + { + "epoch": 0.29243, + "grad_norm": 1.3545064495001546, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 29243 + }, + { + "epoch": 0.29244, + "grad_norm": 1.4704698668511602, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 29244 + }, + { + "epoch": 0.29245, + "grad_norm": 1.2122350383541847, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 29245 + }, + { + "epoch": 0.29246, + "grad_norm": 1.4433771257769041, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 29246 + }, + { + "epoch": 0.29247, + "grad_norm": 1.2772487052636459, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29247 + }, + { + "epoch": 0.29248, + "grad_norm": 1.2540916856396938, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 29248 + }, + { + "epoch": 0.29249, + "grad_norm": 1.2894624783850974, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29249 + }, + { + "epoch": 0.2925, + "grad_norm": 1.745564623799202, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 29250 + }, + { + "epoch": 0.29251, + "grad_norm": 1.3845444476175777, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29251 + }, + { + "epoch": 0.29252, + "grad_norm": 1.9445459398650955, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 29252 + }, + { + "epoch": 0.29253, + "grad_norm": 1.3601057174888596, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29253 + }, + { + "epoch": 0.29254, + "grad_norm": 1.435081258557918, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 29254 + }, + { + "epoch": 0.29255, + "grad_norm": 1.676154037270746, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29255 + }, + { + "epoch": 0.29256, + "grad_norm": 1.1982168800276212, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 29256 + }, + { + "epoch": 0.29257, + "grad_norm": 1.4663181822279088, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29257 + }, + { + "epoch": 0.29258, + "grad_norm": 1.5203706257308593, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 29258 + }, + { + "epoch": 0.29259, + "grad_norm": 1.332766810741733, + "learning_rate": 0.003, + "loss": 4.032, + "step": 29259 + }, + { + "epoch": 0.2926, + "grad_norm": 1.4301658484976703, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 29260 + }, + { + "epoch": 0.29261, + "grad_norm": 1.499452698372927, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 29261 + }, + { + "epoch": 0.29262, + "grad_norm": 1.2867708416541854, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29262 + }, + { + "epoch": 0.29263, + "grad_norm": 1.4935071367538166, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 29263 + }, + { + "epoch": 0.29264, + "grad_norm": 1.5315682637244175, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 29264 + }, + { + "epoch": 0.29265, + "grad_norm": 1.4249473380031907, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 29265 + }, + { + "epoch": 0.29266, + "grad_norm": 1.3510188192069663, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 29266 + }, + { + "epoch": 0.29267, + "grad_norm": 1.6425164987196328, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29267 + }, + { + "epoch": 0.29268, + "grad_norm": 1.3059557809577855, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 29268 + }, + { + "epoch": 0.29269, + "grad_norm": 1.5847063227866531, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29269 + }, + { + "epoch": 0.2927, + "grad_norm": 1.1813879099277036, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 29270 + }, + { + "epoch": 0.29271, + "grad_norm": 1.4691828385204093, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 29271 + }, + { + "epoch": 0.29272, + "grad_norm": 1.1205454115193005, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 29272 + }, + { + "epoch": 0.29273, + "grad_norm": 1.5781283638366754, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 29273 + }, + { + "epoch": 0.29274, + "grad_norm": 1.3935987520308184, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29274 + }, + { + "epoch": 0.29275, + "grad_norm": 1.669681362243707, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 29275 + }, + { + "epoch": 0.29276, + "grad_norm": 1.728589785076805, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 29276 + }, + { + "epoch": 0.29277, + "grad_norm": 1.1292550653657023, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 29277 + }, + { + "epoch": 0.29278, + "grad_norm": 1.9144082607150523, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 29278 + }, + { + "epoch": 0.29279, + "grad_norm": 1.087607065845994, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 29279 + }, + { + "epoch": 0.2928, + "grad_norm": 1.7508645817128887, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29280 + }, + { + "epoch": 0.29281, + "grad_norm": 1.291596993310055, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 29281 + }, + { + "epoch": 0.29282, + "grad_norm": 1.5916890588217705, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 29282 + }, + { + "epoch": 0.29283, + "grad_norm": 1.700788814432171, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 29283 + }, + { + "epoch": 0.29284, + "grad_norm": 1.209788591256133, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29284 + }, + { + "epoch": 0.29285, + "grad_norm": 1.3193623436034945, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 29285 + }, + { + "epoch": 0.29286, + "grad_norm": 1.6780059816966864, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 29286 + }, + { + "epoch": 0.29287, + "grad_norm": 1.5096764378998047, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 29287 + }, + { + "epoch": 0.29288, + "grad_norm": 1.6190121788147296, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 29288 + }, + { + "epoch": 0.29289, + "grad_norm": 1.3362904597329404, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 29289 + }, + { + "epoch": 0.2929, + "grad_norm": 1.6961118184215926, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 29290 + }, + { + "epoch": 0.29291, + "grad_norm": 1.266509439735306, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29291 + }, + { + "epoch": 0.29292, + "grad_norm": 1.6801418134024306, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 29292 + }, + { + "epoch": 0.29293, + "grad_norm": 1.367816417108794, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 29293 + }, + { + "epoch": 0.29294, + "grad_norm": 1.310067191996904, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 29294 + }, + { + "epoch": 0.29295, + "grad_norm": 1.4371432032409925, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29295 + }, + { + "epoch": 0.29296, + "grad_norm": 1.49926103371534, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 29296 + }, + { + "epoch": 0.29297, + "grad_norm": 1.681262614648091, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 29297 + }, + { + "epoch": 0.29298, + "grad_norm": 1.2267494056381478, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 29298 + }, + { + "epoch": 0.29299, + "grad_norm": 1.1654320399298548, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29299 + }, + { + "epoch": 0.293, + "grad_norm": 1.5087385768271127, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 29300 + }, + { + "epoch": 0.29301, + "grad_norm": 1.5215242803355542, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 29301 + }, + { + "epoch": 0.29302, + "grad_norm": 1.2663355091330775, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 29302 + }, + { + "epoch": 0.29303, + "grad_norm": 1.7129686003615945, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29303 + }, + { + "epoch": 0.29304, + "grad_norm": 1.1152996831144082, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29304 + }, + { + "epoch": 0.29305, + "grad_norm": 1.5051418605429132, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 29305 + }, + { + "epoch": 0.29306, + "grad_norm": 1.5089448841133626, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 29306 + }, + { + "epoch": 0.29307, + "grad_norm": 1.2617245315859948, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 29307 + }, + { + "epoch": 0.29308, + "grad_norm": 1.4272425332732293, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 29308 + }, + { + "epoch": 0.29309, + "grad_norm": 1.510125154352793, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 29309 + }, + { + "epoch": 0.2931, + "grad_norm": 1.6090885420211791, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29310 + }, + { + "epoch": 0.29311, + "grad_norm": 1.5424844995159905, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 29311 + }, + { + "epoch": 0.29312, + "grad_norm": 1.6165180866982052, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29312 + }, + { + "epoch": 0.29313, + "grad_norm": 1.5263141124260309, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29313 + }, + { + "epoch": 0.29314, + "grad_norm": 1.6323973776556095, + "learning_rate": 0.003, + "loss": 4.053, + "step": 29314 + }, + { + "epoch": 0.29315, + "grad_norm": 1.1959135368005651, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 29315 + }, + { + "epoch": 0.29316, + "grad_norm": 1.6189852729951417, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 29316 + }, + { + "epoch": 0.29317, + "grad_norm": 1.059494861781781, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 29317 + }, + { + "epoch": 0.29318, + "grad_norm": 1.5518203363148864, + "learning_rate": 0.003, + "loss": 4.076, + "step": 29318 + }, + { + "epoch": 0.29319, + "grad_norm": 1.1632196183257169, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 29319 + }, + { + "epoch": 0.2932, + "grad_norm": 1.76627591981066, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29320 + }, + { + "epoch": 0.29321, + "grad_norm": 1.4375553851944864, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 29321 + }, + { + "epoch": 0.29322, + "grad_norm": 1.326747869017298, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 29322 + }, + { + "epoch": 0.29323, + "grad_norm": 1.5185822962684152, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29323 + }, + { + "epoch": 0.29324, + "grad_norm": 1.2749884387081611, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 29324 + }, + { + "epoch": 0.29325, + "grad_norm": 1.5796188684960493, + "learning_rate": 0.003, + "loss": 4.049, + "step": 29325 + }, + { + "epoch": 0.29326, + "grad_norm": 1.3627270423491595, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 29326 + }, + { + "epoch": 0.29327, + "grad_norm": 1.2924881523549643, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 29327 + }, + { + "epoch": 0.29328, + "grad_norm": 1.599406068594436, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 29328 + }, + { + "epoch": 0.29329, + "grad_norm": 1.2672815651298874, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29329 + }, + { + "epoch": 0.2933, + "grad_norm": 1.4398137065586663, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29330 + }, + { + "epoch": 0.29331, + "grad_norm": 1.566361661775085, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29331 + }, + { + "epoch": 0.29332, + "grad_norm": 1.4678069417647706, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 29332 + }, + { + "epoch": 0.29333, + "grad_norm": 1.4020440303635275, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 29333 + }, + { + "epoch": 0.29334, + "grad_norm": 1.3642560393879808, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29334 + }, + { + "epoch": 0.29335, + "grad_norm": 1.7270935115519281, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 29335 + }, + { + "epoch": 0.29336, + "grad_norm": 1.287727658975346, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29336 + }, + { + "epoch": 0.29337, + "grad_norm": 1.6962666813239102, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 29337 + }, + { + "epoch": 0.29338, + "grad_norm": 1.4858193812868288, + "learning_rate": 0.003, + "loss": 4.058, + "step": 29338 + }, + { + "epoch": 0.29339, + "grad_norm": 1.3231092135050488, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 29339 + }, + { + "epoch": 0.2934, + "grad_norm": 1.4317420219428327, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29340 + }, + { + "epoch": 0.29341, + "grad_norm": 1.1842114003303594, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 29341 + }, + { + "epoch": 0.29342, + "grad_norm": 2.077527282173171, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 29342 + }, + { + "epoch": 0.29343, + "grad_norm": 1.4626711201852676, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 29343 + }, + { + "epoch": 0.29344, + "grad_norm": 1.3291467742099305, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 29344 + }, + { + "epoch": 0.29345, + "grad_norm": 1.5981832030445215, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 29345 + }, + { + "epoch": 0.29346, + "grad_norm": 1.316523353756106, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 29346 + }, + { + "epoch": 0.29347, + "grad_norm": 1.355738278897056, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 29347 + }, + { + "epoch": 0.29348, + "grad_norm": 1.3890052449758543, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29348 + }, + { + "epoch": 0.29349, + "grad_norm": 1.450610760157984, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 29349 + }, + { + "epoch": 0.2935, + "grad_norm": 1.3720236804533106, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 29350 + }, + { + "epoch": 0.29351, + "grad_norm": 1.656292069241058, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 29351 + }, + { + "epoch": 0.29352, + "grad_norm": 1.523769918462285, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 29352 + }, + { + "epoch": 0.29353, + "grad_norm": 1.358786599342174, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 29353 + }, + { + "epoch": 0.29354, + "grad_norm": 1.652197403006986, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 29354 + }, + { + "epoch": 0.29355, + "grad_norm": 1.2210017203023957, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29355 + }, + { + "epoch": 0.29356, + "grad_norm": 1.4819501016590073, + "learning_rate": 0.003, + "loss": 4.064, + "step": 29356 + }, + { + "epoch": 0.29357, + "grad_norm": 1.0991716613282572, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 29357 + }, + { + "epoch": 0.29358, + "grad_norm": 1.772499737230505, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29358 + }, + { + "epoch": 0.29359, + "grad_norm": 1.185979130614648, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 29359 + }, + { + "epoch": 0.2936, + "grad_norm": 1.778147018141483, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 29360 + }, + { + "epoch": 0.29361, + "grad_norm": 1.20806734436739, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 29361 + }, + { + "epoch": 0.29362, + "grad_norm": 1.423171905381718, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 29362 + }, + { + "epoch": 0.29363, + "grad_norm": 1.3960780109890747, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 29363 + }, + { + "epoch": 0.29364, + "grad_norm": 1.5089516482526764, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 29364 + }, + { + "epoch": 0.29365, + "grad_norm": 1.5517961908697215, + "learning_rate": 0.003, + "loss": 4.049, + "step": 29365 + }, + { + "epoch": 0.29366, + "grad_norm": 1.2919370273261621, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 29366 + }, + { + "epoch": 0.29367, + "grad_norm": 1.6866835966002416, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 29367 + }, + { + "epoch": 0.29368, + "grad_norm": 1.2683745800462114, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 29368 + }, + { + "epoch": 0.29369, + "grad_norm": 1.5202827758950215, + "learning_rate": 0.003, + "loss": 4.076, + "step": 29369 + }, + { + "epoch": 0.2937, + "grad_norm": 1.2644008662537969, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 29370 + }, + { + "epoch": 0.29371, + "grad_norm": 1.675061148778537, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29371 + }, + { + "epoch": 0.29372, + "grad_norm": 1.6350517152180248, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 29372 + }, + { + "epoch": 0.29373, + "grad_norm": 2.0239295289589716, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 29373 + }, + { + "epoch": 0.29374, + "grad_norm": 1.1593476881031233, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29374 + }, + { + "epoch": 0.29375, + "grad_norm": 1.5960734525441895, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 29375 + }, + { + "epoch": 0.29376, + "grad_norm": 1.3396938271507275, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29376 + }, + { + "epoch": 0.29377, + "grad_norm": 1.6190000597680316, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 29377 + }, + { + "epoch": 0.29378, + "grad_norm": 1.4108123167479543, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 29378 + }, + { + "epoch": 0.29379, + "grad_norm": 1.3863592542464591, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 29379 + }, + { + "epoch": 0.2938, + "grad_norm": 1.3901911197815706, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 29380 + }, + { + "epoch": 0.29381, + "grad_norm": 1.6091058979880397, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 29381 + }, + { + "epoch": 0.29382, + "grad_norm": 1.2043831859311658, + "learning_rate": 0.003, + "loss": 4.05, + "step": 29382 + }, + { + "epoch": 0.29383, + "grad_norm": 1.5882011448563838, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 29383 + }, + { + "epoch": 0.29384, + "grad_norm": 1.2214359189232047, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 29384 + }, + { + "epoch": 0.29385, + "grad_norm": 1.81612576984766, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 29385 + }, + { + "epoch": 0.29386, + "grad_norm": 1.3289998154917537, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 29386 + }, + { + "epoch": 0.29387, + "grad_norm": 1.4530919536329103, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29387 + }, + { + "epoch": 0.29388, + "grad_norm": 1.4898698336045213, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 29388 + }, + { + "epoch": 0.29389, + "grad_norm": 1.3009326958810712, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 29389 + }, + { + "epoch": 0.2939, + "grad_norm": 1.5282585823838948, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 29390 + }, + { + "epoch": 0.29391, + "grad_norm": 1.445488470696079, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29391 + }, + { + "epoch": 0.29392, + "grad_norm": 1.287267622730324, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 29392 + }, + { + "epoch": 0.29393, + "grad_norm": 1.4504525489977738, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 29393 + }, + { + "epoch": 0.29394, + "grad_norm": 1.5093670457441573, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 29394 + }, + { + "epoch": 0.29395, + "grad_norm": 1.3778424278173935, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29395 + }, + { + "epoch": 0.29396, + "grad_norm": 1.2729009279883023, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 29396 + }, + { + "epoch": 0.29397, + "grad_norm": 1.3872351028236343, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 29397 + }, + { + "epoch": 0.29398, + "grad_norm": 1.5730313442607025, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29398 + }, + { + "epoch": 0.29399, + "grad_norm": 1.2421864588961402, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 29399 + }, + { + "epoch": 0.294, + "grad_norm": 1.6483713690273314, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 29400 + }, + { + "epoch": 0.29401, + "grad_norm": 1.26423522639112, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 29401 + }, + { + "epoch": 0.29402, + "grad_norm": 1.6149603615406183, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 29402 + }, + { + "epoch": 0.29403, + "grad_norm": 1.3952485157438874, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 29403 + }, + { + "epoch": 0.29404, + "grad_norm": 1.585612494790328, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 29404 + }, + { + "epoch": 0.29405, + "grad_norm": 1.1334359520170354, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 29405 + }, + { + "epoch": 0.29406, + "grad_norm": 1.6045038786759696, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 29406 + }, + { + "epoch": 0.29407, + "grad_norm": 1.4032494677494909, + "learning_rate": 0.003, + "loss": 4.04, + "step": 29407 + }, + { + "epoch": 0.29408, + "grad_norm": 1.4658264544296067, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 29408 + }, + { + "epoch": 0.29409, + "grad_norm": 1.3814464562565825, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 29409 + }, + { + "epoch": 0.2941, + "grad_norm": 1.6155195300021752, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 29410 + }, + { + "epoch": 0.29411, + "grad_norm": 1.271958208556957, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 29411 + }, + { + "epoch": 0.29412, + "grad_norm": 1.612739658336224, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 29412 + }, + { + "epoch": 0.29413, + "grad_norm": 1.2356872416489155, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 29413 + }, + { + "epoch": 0.29414, + "grad_norm": 1.7263840435392848, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29414 + }, + { + "epoch": 0.29415, + "grad_norm": 1.4450568050921697, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 29415 + }, + { + "epoch": 0.29416, + "grad_norm": 1.8327570151489885, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 29416 + }, + { + "epoch": 0.29417, + "grad_norm": 1.2521312472138966, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 29417 + }, + { + "epoch": 0.29418, + "grad_norm": 1.3528807427037979, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29418 + }, + { + "epoch": 0.29419, + "grad_norm": 1.658581974232522, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 29419 + }, + { + "epoch": 0.2942, + "grad_norm": 1.1539247819950178, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 29420 + }, + { + "epoch": 0.29421, + "grad_norm": 1.6008378314347098, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29421 + }, + { + "epoch": 0.29422, + "grad_norm": 1.3941196540249752, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 29422 + }, + { + "epoch": 0.29423, + "grad_norm": 1.6890236915296446, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 29423 + }, + { + "epoch": 0.29424, + "grad_norm": 1.522814594309457, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 29424 + }, + { + "epoch": 0.29425, + "grad_norm": 1.2821775498468408, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 29425 + }, + { + "epoch": 0.29426, + "grad_norm": 1.495423750645436, + "learning_rate": 0.003, + "loss": 4.044, + "step": 29426 + }, + { + "epoch": 0.29427, + "grad_norm": 1.3804207034492957, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 29427 + }, + { + "epoch": 0.29428, + "grad_norm": 1.5181202227507828, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 29428 + }, + { + "epoch": 0.29429, + "grad_norm": 1.211766958280413, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 29429 + }, + { + "epoch": 0.2943, + "grad_norm": 1.6321724814631033, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29430 + }, + { + "epoch": 0.29431, + "grad_norm": 1.4229281087298498, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 29431 + }, + { + "epoch": 0.29432, + "grad_norm": 1.3986300152089768, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 29432 + }, + { + "epoch": 0.29433, + "grad_norm": 1.3940416433436509, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29433 + }, + { + "epoch": 0.29434, + "grad_norm": 1.6604814464812323, + "learning_rate": 0.003, + "loss": 4.058, + "step": 29434 + }, + { + "epoch": 0.29435, + "grad_norm": 1.2427260845084842, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 29435 + }, + { + "epoch": 0.29436, + "grad_norm": 1.63438340156479, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29436 + }, + { + "epoch": 0.29437, + "grad_norm": 1.1588372198051036, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29437 + }, + { + "epoch": 0.29438, + "grad_norm": 1.3712882779326327, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 29438 + }, + { + "epoch": 0.29439, + "grad_norm": 1.167024019352043, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 29439 + }, + { + "epoch": 0.2944, + "grad_norm": 1.6301578411179352, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 29440 + }, + { + "epoch": 0.29441, + "grad_norm": 1.5909912144377922, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29441 + }, + { + "epoch": 0.29442, + "grad_norm": 1.33822256365621, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 29442 + }, + { + "epoch": 0.29443, + "grad_norm": 1.5190980818642605, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29443 + }, + { + "epoch": 0.29444, + "grad_norm": 1.2981474257904482, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 29444 + }, + { + "epoch": 0.29445, + "grad_norm": 1.5333220742868277, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 29445 + }, + { + "epoch": 0.29446, + "grad_norm": 1.2929875289171653, + "learning_rate": 0.003, + "loss": 4.043, + "step": 29446 + }, + { + "epoch": 0.29447, + "grad_norm": 1.6015268084021137, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 29447 + }, + { + "epoch": 0.29448, + "grad_norm": 1.2465941707519643, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29448 + }, + { + "epoch": 0.29449, + "grad_norm": 1.946467204969171, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 29449 + }, + { + "epoch": 0.2945, + "grad_norm": 1.1736015902804267, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29450 + }, + { + "epoch": 0.29451, + "grad_norm": 1.8871387789604783, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 29451 + }, + { + "epoch": 0.29452, + "grad_norm": 1.1686980312586202, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 29452 + }, + { + "epoch": 0.29453, + "grad_norm": 1.988262355183983, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 29453 + }, + { + "epoch": 0.29454, + "grad_norm": 1.2132998101355863, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 29454 + }, + { + "epoch": 0.29455, + "grad_norm": 1.1761410187327577, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 29455 + }, + { + "epoch": 0.29456, + "grad_norm": 1.564908882058278, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 29456 + }, + { + "epoch": 0.29457, + "grad_norm": 1.2798923171310757, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29457 + }, + { + "epoch": 0.29458, + "grad_norm": 1.8292630899552835, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29458 + }, + { + "epoch": 0.29459, + "grad_norm": 1.2985512923276556, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 29459 + }, + { + "epoch": 0.2946, + "grad_norm": 1.7049418519055513, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 29460 + }, + { + "epoch": 0.29461, + "grad_norm": 1.4288823637585657, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 29461 + }, + { + "epoch": 0.29462, + "grad_norm": 1.1176470679751518, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 29462 + }, + { + "epoch": 0.29463, + "grad_norm": 1.5530361771712557, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29463 + }, + { + "epoch": 0.29464, + "grad_norm": 1.4615923191361506, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29464 + }, + { + "epoch": 0.29465, + "grad_norm": 1.4317048929084162, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29465 + }, + { + "epoch": 0.29466, + "grad_norm": 1.6200597770263392, + "learning_rate": 0.003, + "loss": 4.046, + "step": 29466 + }, + { + "epoch": 0.29467, + "grad_norm": 1.4015689602597654, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 29467 + }, + { + "epoch": 0.29468, + "grad_norm": 1.3322476180724683, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 29468 + }, + { + "epoch": 0.29469, + "grad_norm": 1.5384827185835102, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29469 + }, + { + "epoch": 0.2947, + "grad_norm": 1.0834229748912159, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29470 + }, + { + "epoch": 0.29471, + "grad_norm": 1.6218961440459359, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 29471 + }, + { + "epoch": 0.29472, + "grad_norm": 1.3227929442225794, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 29472 + }, + { + "epoch": 0.29473, + "grad_norm": 1.8552332217649876, + "learning_rate": 0.003, + "loss": 4.068, + "step": 29473 + }, + { + "epoch": 0.29474, + "grad_norm": 1.5461600406605769, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 29474 + }, + { + "epoch": 0.29475, + "grad_norm": 1.3427916137862272, + "learning_rate": 0.003, + "loss": 4.055, + "step": 29475 + }, + { + "epoch": 0.29476, + "grad_norm": 1.4847018083907315, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 29476 + }, + { + "epoch": 0.29477, + "grad_norm": 1.5096956010623435, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29477 + }, + { + "epoch": 0.29478, + "grad_norm": 1.2293837102072191, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 29478 + }, + { + "epoch": 0.29479, + "grad_norm": 1.5623033676653277, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 29479 + }, + { + "epoch": 0.2948, + "grad_norm": 1.33390986782351, + "learning_rate": 0.003, + "loss": 4.036, + "step": 29480 + }, + { + "epoch": 0.29481, + "grad_norm": 1.331113145405095, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 29481 + }, + { + "epoch": 0.29482, + "grad_norm": 1.1749981097012636, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 29482 + }, + { + "epoch": 0.29483, + "grad_norm": 1.6940590815869974, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29483 + }, + { + "epoch": 0.29484, + "grad_norm": 1.5640108371725128, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 29484 + }, + { + "epoch": 0.29485, + "grad_norm": 1.9079335804741604, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 29485 + }, + { + "epoch": 0.29486, + "grad_norm": 1.2327424331998704, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29486 + }, + { + "epoch": 0.29487, + "grad_norm": 1.3485539595345761, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29487 + }, + { + "epoch": 0.29488, + "grad_norm": 1.689875397857719, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 29488 + }, + { + "epoch": 0.29489, + "grad_norm": 1.5004274420943315, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 29489 + }, + { + "epoch": 0.2949, + "grad_norm": 1.4154211943108546, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29490 + }, + { + "epoch": 0.29491, + "grad_norm": 1.4476367570452546, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 29491 + }, + { + "epoch": 0.29492, + "grad_norm": 1.34596989752265, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29492 + }, + { + "epoch": 0.29493, + "grad_norm": 1.32036707383905, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 29493 + }, + { + "epoch": 0.29494, + "grad_norm": 1.3166233463680952, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 29494 + }, + { + "epoch": 0.29495, + "grad_norm": 1.203451120579011, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29495 + }, + { + "epoch": 0.29496, + "grad_norm": 1.5873518784314988, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 29496 + }, + { + "epoch": 0.29497, + "grad_norm": 1.4448899131788917, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 29497 + }, + { + "epoch": 0.29498, + "grad_norm": 1.6828599243906937, + "learning_rate": 0.003, + "loss": 4.047, + "step": 29498 + }, + { + "epoch": 0.29499, + "grad_norm": 1.35270215408051, + "learning_rate": 0.003, + "loss": 4.029, + "step": 29499 + }, + { + "epoch": 0.295, + "grad_norm": 1.7006673250740023, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 29500 + }, + { + "epoch": 0.29501, + "grad_norm": 1.2419373791561557, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 29501 + }, + { + "epoch": 0.29502, + "grad_norm": 1.9791923638701046, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 29502 + }, + { + "epoch": 0.29503, + "grad_norm": 1.4128339383016388, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 29503 + }, + { + "epoch": 0.29504, + "grad_norm": 1.1440385604471317, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 29504 + }, + { + "epoch": 0.29505, + "grad_norm": 1.9450602260228906, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 29505 + }, + { + "epoch": 0.29506, + "grad_norm": 1.9546007768247158, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29506 + }, + { + "epoch": 0.29507, + "grad_norm": 2.463158565466238, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 29507 + }, + { + "epoch": 0.29508, + "grad_norm": 1.5401770000252428, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 29508 + }, + { + "epoch": 0.29509, + "grad_norm": 1.864436233564162, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 29509 + }, + { + "epoch": 0.2951, + "grad_norm": 2.907109277222743, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 29510 + }, + { + "epoch": 0.29511, + "grad_norm": 2.054676421092645, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 29511 + }, + { + "epoch": 0.29512, + "grad_norm": 1.4825126523943408, + "learning_rate": 0.003, + "loss": 4.084, + "step": 29512 + }, + { + "epoch": 0.29513, + "grad_norm": 1.532973503600126, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 29513 + }, + { + "epoch": 0.29514, + "grad_norm": 1.4179527719945462, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 29514 + }, + { + "epoch": 0.29515, + "grad_norm": 1.5441591805037527, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 29515 + }, + { + "epoch": 0.29516, + "grad_norm": 1.4188405992209667, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 29516 + }, + { + "epoch": 0.29517, + "grad_norm": 1.810441082596753, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 29517 + }, + { + "epoch": 0.29518, + "grad_norm": 1.5407284977188422, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 29518 + }, + { + "epoch": 0.29519, + "grad_norm": 1.6552524546881293, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 29519 + }, + { + "epoch": 0.2952, + "grad_norm": 1.4686900837510646, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 29520 + }, + { + "epoch": 0.29521, + "grad_norm": 1.5853284976470046, + "learning_rate": 0.003, + "loss": 4.045, + "step": 29521 + }, + { + "epoch": 0.29522, + "grad_norm": 1.8125035518588881, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 29522 + }, + { + "epoch": 0.29523, + "grad_norm": 1.7941608673980132, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 29523 + }, + { + "epoch": 0.29524, + "grad_norm": 1.8877091782035464, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 29524 + }, + { + "epoch": 0.29525, + "grad_norm": 1.3071343693669437, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 29525 + }, + { + "epoch": 0.29526, + "grad_norm": 1.3285362169903332, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29526 + }, + { + "epoch": 0.29527, + "grad_norm": 1.6687663970061366, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 29527 + }, + { + "epoch": 0.29528, + "grad_norm": 1.1697312441695027, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 29528 + }, + { + "epoch": 0.29529, + "grad_norm": 1.4032240524272124, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 29529 + }, + { + "epoch": 0.2953, + "grad_norm": 1.652589840747388, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 29530 + }, + { + "epoch": 0.29531, + "grad_norm": 1.3286950119269714, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 29531 + }, + { + "epoch": 0.29532, + "grad_norm": 1.715722659030729, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 29532 + }, + { + "epoch": 0.29533, + "grad_norm": 1.2053043460669615, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 29533 + }, + { + "epoch": 0.29534, + "grad_norm": 1.3260534982646222, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 29534 + }, + { + "epoch": 0.29535, + "grad_norm": 1.1752733940071336, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 29535 + }, + { + "epoch": 0.29536, + "grad_norm": 1.5674376304715354, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29536 + }, + { + "epoch": 0.29537, + "grad_norm": 1.4200892750322616, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 29537 + }, + { + "epoch": 0.29538, + "grad_norm": 1.9617945018572442, + "learning_rate": 0.003, + "loss": 4.042, + "step": 29538 + }, + { + "epoch": 0.29539, + "grad_norm": 1.3255903368238975, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29539 + }, + { + "epoch": 0.2954, + "grad_norm": 1.4846611905905969, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29540 + }, + { + "epoch": 0.29541, + "grad_norm": 1.505795505286239, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 29541 + }, + { + "epoch": 0.29542, + "grad_norm": 1.584860208921482, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29542 + }, + { + "epoch": 0.29543, + "grad_norm": 1.4457260599275668, + "learning_rate": 0.003, + "loss": 4.033, + "step": 29543 + }, + { + "epoch": 0.29544, + "grad_norm": 1.5749432318844432, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 29544 + }, + { + "epoch": 0.29545, + "grad_norm": 1.4913056135881853, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 29545 + }, + { + "epoch": 0.29546, + "grad_norm": 1.2763620983884196, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 29546 + }, + { + "epoch": 0.29547, + "grad_norm": 1.5912304097061392, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 29547 + }, + { + "epoch": 0.29548, + "grad_norm": 2.0763331025371325, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 29548 + }, + { + "epoch": 0.29549, + "grad_norm": 1.3562307123900088, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 29549 + }, + { + "epoch": 0.2955, + "grad_norm": 1.3324877748410955, + "learning_rate": 0.003, + "loss": 4.088, + "step": 29550 + }, + { + "epoch": 0.29551, + "grad_norm": 1.4441082115016586, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 29551 + }, + { + "epoch": 0.29552, + "grad_norm": 1.417879085045667, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 29552 + }, + { + "epoch": 0.29553, + "grad_norm": 1.3467761017654436, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29553 + }, + { + "epoch": 0.29554, + "grad_norm": 1.4414248421012987, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 29554 + }, + { + "epoch": 0.29555, + "grad_norm": 1.6899517722611928, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29555 + }, + { + "epoch": 0.29556, + "grad_norm": 1.2873232167467472, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29556 + }, + { + "epoch": 0.29557, + "grad_norm": 1.653178226709529, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 29557 + }, + { + "epoch": 0.29558, + "grad_norm": 1.238818262704665, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29558 + }, + { + "epoch": 0.29559, + "grad_norm": 1.5887610452671799, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 29559 + }, + { + "epoch": 0.2956, + "grad_norm": 1.1864140711145585, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29560 + }, + { + "epoch": 0.29561, + "grad_norm": 1.8000031676066184, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 29561 + }, + { + "epoch": 0.29562, + "grad_norm": 1.2081934938642918, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 29562 + }, + { + "epoch": 0.29563, + "grad_norm": 1.321349306049731, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 29563 + }, + { + "epoch": 0.29564, + "grad_norm": 1.612273370776185, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 29564 + }, + { + "epoch": 0.29565, + "grad_norm": 1.433090033339689, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29565 + }, + { + "epoch": 0.29566, + "grad_norm": 1.317069942368862, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 29566 + }, + { + "epoch": 0.29567, + "grad_norm": 1.5976128203764532, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29567 + }, + { + "epoch": 0.29568, + "grad_norm": 1.5878362438823914, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 29568 + }, + { + "epoch": 0.29569, + "grad_norm": 1.2443181211590866, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 29569 + }, + { + "epoch": 0.2957, + "grad_norm": 1.6832400859496306, + "learning_rate": 0.003, + "loss": 4.002, + "step": 29570 + }, + { + "epoch": 0.29571, + "grad_norm": 1.2656703069013058, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 29571 + }, + { + "epoch": 0.29572, + "grad_norm": 1.7220359729682229, + "learning_rate": 0.003, + "loss": 4.069, + "step": 29572 + }, + { + "epoch": 0.29573, + "grad_norm": 1.1400061813740794, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29573 + }, + { + "epoch": 0.29574, + "grad_norm": 1.778638202301005, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29574 + }, + { + "epoch": 0.29575, + "grad_norm": 1.2733629212645075, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29575 + }, + { + "epoch": 0.29576, + "grad_norm": 1.7331461049444195, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 29576 + }, + { + "epoch": 0.29577, + "grad_norm": 1.4165268422713067, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 29577 + }, + { + "epoch": 0.29578, + "grad_norm": 1.2570789882731355, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 29578 + }, + { + "epoch": 0.29579, + "grad_norm": 1.7753785073518755, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 29579 + }, + { + "epoch": 0.2958, + "grad_norm": 1.3908722177158905, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29580 + }, + { + "epoch": 0.29581, + "grad_norm": 1.4057144806580635, + "learning_rate": 0.003, + "loss": 4.055, + "step": 29581 + }, + { + "epoch": 0.29582, + "grad_norm": 1.5936028788046048, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 29582 + }, + { + "epoch": 0.29583, + "grad_norm": 1.4170960120959504, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 29583 + }, + { + "epoch": 0.29584, + "grad_norm": 1.2472623336893094, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29584 + }, + { + "epoch": 0.29585, + "grad_norm": 1.390096921133506, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29585 + }, + { + "epoch": 0.29586, + "grad_norm": 1.3018480052019632, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 29586 + }, + { + "epoch": 0.29587, + "grad_norm": 1.5089685274267743, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29587 + }, + { + "epoch": 0.29588, + "grad_norm": 1.3960212127275022, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 29588 + }, + { + "epoch": 0.29589, + "grad_norm": 1.53372050491043, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 29589 + }, + { + "epoch": 0.2959, + "grad_norm": 1.3528168515162273, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 29590 + }, + { + "epoch": 0.29591, + "grad_norm": 1.4454788787368122, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 29591 + }, + { + "epoch": 0.29592, + "grad_norm": 1.2141993821183437, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29592 + }, + { + "epoch": 0.29593, + "grad_norm": 1.433059301292696, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 29593 + }, + { + "epoch": 0.29594, + "grad_norm": 1.4394495501146913, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29594 + }, + { + "epoch": 0.29595, + "grad_norm": 1.346594141944316, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29595 + }, + { + "epoch": 0.29596, + "grad_norm": 1.4504049770335217, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29596 + }, + { + "epoch": 0.29597, + "grad_norm": 1.5551368097879674, + "learning_rate": 0.003, + "loss": 4.048, + "step": 29597 + }, + { + "epoch": 0.29598, + "grad_norm": 1.4762822224340535, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 29598 + }, + { + "epoch": 0.29599, + "grad_norm": 1.3586855494618892, + "learning_rate": 0.003, + "loss": 4.03, + "step": 29599 + }, + { + "epoch": 0.296, + "grad_norm": 1.4903377260188757, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29600 + }, + { + "epoch": 0.29601, + "grad_norm": 1.5082641960681722, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 29601 + }, + { + "epoch": 0.29602, + "grad_norm": 1.6575221819461898, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 29602 + }, + { + "epoch": 0.29603, + "grad_norm": 1.119885275235292, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29603 + }, + { + "epoch": 0.29604, + "grad_norm": 1.857422566793844, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 29604 + }, + { + "epoch": 0.29605, + "grad_norm": 1.3356461695649762, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29605 + }, + { + "epoch": 0.29606, + "grad_norm": 1.3521815087137699, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 29606 + }, + { + "epoch": 0.29607, + "grad_norm": 1.2256582688505853, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 29607 + }, + { + "epoch": 0.29608, + "grad_norm": 1.3781523077151305, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 29608 + }, + { + "epoch": 0.29609, + "grad_norm": 1.5976175787502203, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 29609 + }, + { + "epoch": 0.2961, + "grad_norm": 1.3795386093058946, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 29610 + }, + { + "epoch": 0.29611, + "grad_norm": 1.4819643591770433, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29611 + }, + { + "epoch": 0.29612, + "grad_norm": 1.577290289318663, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 29612 + }, + { + "epoch": 0.29613, + "grad_norm": 1.2149843327611436, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 29613 + }, + { + "epoch": 0.29614, + "grad_norm": 1.3293183652956135, + "learning_rate": 0.003, + "loss": 4.056, + "step": 29614 + }, + { + "epoch": 0.29615, + "grad_norm": 1.3030044913705126, + "learning_rate": 0.003, + "loss": 4.016, + "step": 29615 + }, + { + "epoch": 0.29616, + "grad_norm": 1.359752114963903, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 29616 + }, + { + "epoch": 0.29617, + "grad_norm": 1.353249342121318, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 29617 + }, + { + "epoch": 0.29618, + "grad_norm": 1.4941679648040758, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 29618 + }, + { + "epoch": 0.29619, + "grad_norm": 1.2972150207336481, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29619 + }, + { + "epoch": 0.2962, + "grad_norm": 1.5770225149737511, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29620 + }, + { + "epoch": 0.29621, + "grad_norm": 1.2584832407147915, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29621 + }, + { + "epoch": 0.29622, + "grad_norm": 1.5254250868200876, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29622 + }, + { + "epoch": 0.29623, + "grad_norm": 1.8340455369993915, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 29623 + }, + { + "epoch": 0.29624, + "grad_norm": 1.1559429408985935, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29624 + }, + { + "epoch": 0.29625, + "grad_norm": 1.703131557685693, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29625 + }, + { + "epoch": 0.29626, + "grad_norm": 1.3468173730954578, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 29626 + }, + { + "epoch": 0.29627, + "grad_norm": 1.4777867488205094, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 29627 + }, + { + "epoch": 0.29628, + "grad_norm": 1.438400211083588, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 29628 + }, + { + "epoch": 0.29629, + "grad_norm": 1.493649155368632, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 29629 + }, + { + "epoch": 0.2963, + "grad_norm": 1.3059240725862309, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29630 + }, + { + "epoch": 0.29631, + "grad_norm": 1.5322743614104064, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 29631 + }, + { + "epoch": 0.29632, + "grad_norm": 1.3569670476953113, + "learning_rate": 0.003, + "loss": 4.012, + "step": 29632 + }, + { + "epoch": 0.29633, + "grad_norm": 1.5435150723463762, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29633 + }, + { + "epoch": 0.29634, + "grad_norm": 1.302001633505215, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 29634 + }, + { + "epoch": 0.29635, + "grad_norm": 1.626363707904047, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 29635 + }, + { + "epoch": 0.29636, + "grad_norm": 1.1827301323799169, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29636 + }, + { + "epoch": 0.29637, + "grad_norm": 1.3800558169516528, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 29637 + }, + { + "epoch": 0.29638, + "grad_norm": 1.3330304072259815, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 29638 + }, + { + "epoch": 0.29639, + "grad_norm": 1.9369633857302044, + "learning_rate": 0.003, + "loss": 4.069, + "step": 29639 + }, + { + "epoch": 0.2964, + "grad_norm": 1.1189310607160143, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 29640 + }, + { + "epoch": 0.29641, + "grad_norm": 1.7249238557214277, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29641 + }, + { + "epoch": 0.29642, + "grad_norm": 1.8748392366054623, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29642 + }, + { + "epoch": 0.29643, + "grad_norm": 1.3210258639577208, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 29643 + }, + { + "epoch": 0.29644, + "grad_norm": 1.3737626200635709, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29644 + }, + { + "epoch": 0.29645, + "grad_norm": 1.4663630255475282, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 29645 + }, + { + "epoch": 0.29646, + "grad_norm": 1.2397411212684595, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 29646 + }, + { + "epoch": 0.29647, + "grad_norm": 1.6155691753451964, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 29647 + }, + { + "epoch": 0.29648, + "grad_norm": 1.3849435856302506, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 29648 + }, + { + "epoch": 0.29649, + "grad_norm": 1.5933111799436335, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 29649 + }, + { + "epoch": 0.2965, + "grad_norm": 1.4131956365486875, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 29650 + }, + { + "epoch": 0.29651, + "grad_norm": 1.4030172024664984, + "learning_rate": 0.003, + "loss": 4.031, + "step": 29651 + }, + { + "epoch": 0.29652, + "grad_norm": 1.5493069694814634, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29652 + }, + { + "epoch": 0.29653, + "grad_norm": 1.110111883659851, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 29653 + }, + { + "epoch": 0.29654, + "grad_norm": 1.6694266924836891, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 29654 + }, + { + "epoch": 0.29655, + "grad_norm": 1.2065163067987221, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 29655 + }, + { + "epoch": 0.29656, + "grad_norm": 1.4189322252267795, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 29656 + }, + { + "epoch": 0.29657, + "grad_norm": 1.4231761380233425, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 29657 + }, + { + "epoch": 0.29658, + "grad_norm": 1.7402954960630874, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29658 + }, + { + "epoch": 0.29659, + "grad_norm": 1.50278562256387, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 29659 + }, + { + "epoch": 0.2966, + "grad_norm": 1.5765847356062612, + "learning_rate": 0.003, + "loss": 4.065, + "step": 29660 + }, + { + "epoch": 0.29661, + "grad_norm": 1.2962447011527627, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29661 + }, + { + "epoch": 0.29662, + "grad_norm": 1.553187824241197, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 29662 + }, + { + "epoch": 0.29663, + "grad_norm": 1.4833361979268078, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 29663 + }, + { + "epoch": 0.29664, + "grad_norm": 1.4094928153484663, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29664 + }, + { + "epoch": 0.29665, + "grad_norm": 1.3523609347198133, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 29665 + }, + { + "epoch": 0.29666, + "grad_norm": 1.3664819855196202, + "learning_rate": 0.003, + "loss": 4.055, + "step": 29666 + }, + { + "epoch": 0.29667, + "grad_norm": 1.3815151022171708, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 29667 + }, + { + "epoch": 0.29668, + "grad_norm": 1.2223146316595501, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 29668 + }, + { + "epoch": 0.29669, + "grad_norm": 1.515758859128414, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 29669 + }, + { + "epoch": 0.2967, + "grad_norm": 1.1643400960355768, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 29670 + }, + { + "epoch": 0.29671, + "grad_norm": 1.5618218885689892, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 29671 + }, + { + "epoch": 0.29672, + "grad_norm": 1.1736038845243697, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 29672 + }, + { + "epoch": 0.29673, + "grad_norm": 1.7721954566653404, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 29673 + }, + { + "epoch": 0.29674, + "grad_norm": 1.2108216727240393, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 29674 + }, + { + "epoch": 0.29675, + "grad_norm": 1.6391411933719555, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 29675 + }, + { + "epoch": 0.29676, + "grad_norm": 1.3772772520091177, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 29676 + }, + { + "epoch": 0.29677, + "grad_norm": 1.521698385308192, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 29677 + }, + { + "epoch": 0.29678, + "grad_norm": 1.5483108405973751, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 29678 + }, + { + "epoch": 0.29679, + "grad_norm": 1.639275720843659, + "learning_rate": 0.003, + "loss": 4.023, + "step": 29679 + }, + { + "epoch": 0.2968, + "grad_norm": 1.2441942384675482, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 29680 + }, + { + "epoch": 0.29681, + "grad_norm": 1.8930008486885521, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 29681 + }, + { + "epoch": 0.29682, + "grad_norm": 1.183784404454133, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 29682 + }, + { + "epoch": 0.29683, + "grad_norm": 1.6881732458786358, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 29683 + }, + { + "epoch": 0.29684, + "grad_norm": 1.74138554545244, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 29684 + }, + { + "epoch": 0.29685, + "grad_norm": 1.2968691002934825, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 29685 + }, + { + "epoch": 0.29686, + "grad_norm": 1.2711722861979586, + "learning_rate": 0.003, + "loss": 4.053, + "step": 29686 + }, + { + "epoch": 0.29687, + "grad_norm": 1.8820821737282007, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29687 + }, + { + "epoch": 0.29688, + "grad_norm": 1.0772760153853942, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 29688 + }, + { + "epoch": 0.29689, + "grad_norm": 1.592341143171055, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 29689 + }, + { + "epoch": 0.2969, + "grad_norm": 1.2780609634129447, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 29690 + }, + { + "epoch": 0.29691, + "grad_norm": 1.3350203621341141, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29691 + }, + { + "epoch": 0.29692, + "grad_norm": 1.446055550336389, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 29692 + }, + { + "epoch": 0.29693, + "grad_norm": 1.3196244503352264, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 29693 + }, + { + "epoch": 0.29694, + "grad_norm": 1.615137667038377, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 29694 + }, + { + "epoch": 0.29695, + "grad_norm": 1.3039674226635576, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29695 + }, + { + "epoch": 0.29696, + "grad_norm": 1.8585032730258921, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29696 + }, + { + "epoch": 0.29697, + "grad_norm": 1.2784075256286886, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 29697 + }, + { + "epoch": 0.29698, + "grad_norm": 1.3193958455436805, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 29698 + }, + { + "epoch": 0.29699, + "grad_norm": 1.6054727993151936, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 29699 + }, + { + "epoch": 0.297, + "grad_norm": 1.4554293586467384, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 29700 + }, + { + "epoch": 0.29701, + "grad_norm": 1.2619055151770793, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 29701 + }, + { + "epoch": 0.29702, + "grad_norm": 1.4254049450525466, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 29702 + }, + { + "epoch": 0.29703, + "grad_norm": 1.3271195157778501, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 29703 + }, + { + "epoch": 0.29704, + "grad_norm": 1.4636602817146251, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 29704 + }, + { + "epoch": 0.29705, + "grad_norm": 1.3817944458058087, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 29705 + }, + { + "epoch": 0.29706, + "grad_norm": 1.4824955866413272, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 29706 + }, + { + "epoch": 0.29707, + "grad_norm": 1.0447718842788054, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 29707 + }, + { + "epoch": 0.29708, + "grad_norm": 1.523350726186231, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 29708 + }, + { + "epoch": 0.29709, + "grad_norm": 1.218353835787855, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 29709 + }, + { + "epoch": 0.2971, + "grad_norm": 1.9899772420695083, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 29710 + }, + { + "epoch": 0.29711, + "grad_norm": 1.5620405316935226, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 29711 + }, + { + "epoch": 0.29712, + "grad_norm": 1.3377093241819513, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 29712 + }, + { + "epoch": 0.29713, + "grad_norm": 1.5641395841049273, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 29713 + }, + { + "epoch": 0.29714, + "grad_norm": 1.6368166746695767, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 29714 + }, + { + "epoch": 0.29715, + "grad_norm": 1.3191713623533745, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 29715 + }, + { + "epoch": 0.29716, + "grad_norm": 1.4190700057582735, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 29716 + }, + { + "epoch": 0.29717, + "grad_norm": 1.5496028134141882, + "learning_rate": 0.003, + "loss": 4.053, + "step": 29717 + }, + { + "epoch": 0.29718, + "grad_norm": 1.1677942218464579, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 29718 + }, + { + "epoch": 0.29719, + "grad_norm": 1.5987870652902887, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 29719 + }, + { + "epoch": 0.2972, + "grad_norm": 1.461229583653366, + "learning_rate": 0.003, + "loss": 4.043, + "step": 29720 + }, + { + "epoch": 0.29721, + "grad_norm": 1.5279371673726818, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 29721 + }, + { + "epoch": 0.29722, + "grad_norm": 1.3260010557969888, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 29722 + }, + { + "epoch": 0.29723, + "grad_norm": 1.6523266368341627, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 29723 + }, + { + "epoch": 0.29724, + "grad_norm": 1.0931314811535107, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 29724 + }, + { + "epoch": 0.29725, + "grad_norm": 1.7502690160851728, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 29725 + }, + { + "epoch": 0.29726, + "grad_norm": 1.3651195392893234, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 29726 + }, + { + "epoch": 0.29727, + "grad_norm": 1.5443828908255757, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 29727 + }, + { + "epoch": 0.29728, + "grad_norm": 1.231978491753288, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 29728 + }, + { + "epoch": 0.29729, + "grad_norm": 1.5115334219025218, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 29729 + }, + { + "epoch": 0.2973, + "grad_norm": 1.6680418634723304, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 29730 + }, + { + "epoch": 0.29731, + "grad_norm": 1.6187789446420877, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 29731 + }, + { + "epoch": 0.29732, + "grad_norm": 1.3183659389682174, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29732 + }, + { + "epoch": 0.29733, + "grad_norm": 1.4762898660013326, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 29733 + }, + { + "epoch": 0.29734, + "grad_norm": 1.2564411301819496, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29734 + }, + { + "epoch": 0.29735, + "grad_norm": 1.65828634350145, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 29735 + }, + { + "epoch": 0.29736, + "grad_norm": 1.2819386135922708, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 29736 + }, + { + "epoch": 0.29737, + "grad_norm": 1.619445632349337, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 29737 + }, + { + "epoch": 0.29738, + "grad_norm": 1.3655194751311677, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 29738 + }, + { + "epoch": 0.29739, + "grad_norm": 1.4745425003369854, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 29739 + }, + { + "epoch": 0.2974, + "grad_norm": 1.480747282899522, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 29740 + }, + { + "epoch": 0.29741, + "grad_norm": 1.4181026892737834, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29741 + }, + { + "epoch": 0.29742, + "grad_norm": 1.765718375627461, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 29742 + }, + { + "epoch": 0.29743, + "grad_norm": 1.3603512618481926, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 29743 + }, + { + "epoch": 0.29744, + "grad_norm": 1.4085673527871279, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 29744 + }, + { + "epoch": 0.29745, + "grad_norm": 1.337484447018076, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 29745 + }, + { + "epoch": 0.29746, + "grad_norm": 1.4423567718938732, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 29746 + }, + { + "epoch": 0.29747, + "grad_norm": 1.5867948330857258, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 29747 + }, + { + "epoch": 0.29748, + "grad_norm": 1.6957315317100394, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 29748 + }, + { + "epoch": 0.29749, + "grad_norm": 1.1689135184259658, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 29749 + }, + { + "epoch": 0.2975, + "grad_norm": 1.5595359019344701, + "learning_rate": 0.003, + "loss": 4.072, + "step": 29750 + }, + { + "epoch": 0.29751, + "grad_norm": 1.2286572448152018, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 29751 + }, + { + "epoch": 0.29752, + "grad_norm": 1.2513581722112204, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 29752 + }, + { + "epoch": 0.29753, + "grad_norm": 1.575967595096816, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 29753 + }, + { + "epoch": 0.29754, + "grad_norm": 1.5315104039876368, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 29754 + }, + { + "epoch": 0.29755, + "grad_norm": 1.705444905957616, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29755 + }, + { + "epoch": 0.29756, + "grad_norm": 1.5685798235478612, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 29756 + }, + { + "epoch": 0.29757, + "grad_norm": 1.2892473961782613, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 29757 + }, + { + "epoch": 0.29758, + "grad_norm": 1.5361218047943723, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 29758 + }, + { + "epoch": 0.29759, + "grad_norm": 1.450528588309077, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 29759 + }, + { + "epoch": 0.2976, + "grad_norm": 1.5097587127627259, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 29760 + }, + { + "epoch": 0.29761, + "grad_norm": 1.3462532251965578, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 29761 + }, + { + "epoch": 0.29762, + "grad_norm": 1.369615957374208, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29762 + }, + { + "epoch": 0.29763, + "grad_norm": 1.4242003480679755, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29763 + }, + { + "epoch": 0.29764, + "grad_norm": 1.406968264507757, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 29764 + }, + { + "epoch": 0.29765, + "grad_norm": 1.4073835204077054, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 29765 + }, + { + "epoch": 0.29766, + "grad_norm": 1.469120890260429, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29766 + }, + { + "epoch": 0.29767, + "grad_norm": 1.5871790938417385, + "learning_rate": 0.003, + "loss": 4.055, + "step": 29767 + }, + { + "epoch": 0.29768, + "grad_norm": 1.47040281054975, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 29768 + }, + { + "epoch": 0.29769, + "grad_norm": 1.4691048129564754, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 29769 + }, + { + "epoch": 0.2977, + "grad_norm": 1.58580639929341, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 29770 + }, + { + "epoch": 0.29771, + "grad_norm": 1.1805066043437302, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 29771 + }, + { + "epoch": 0.29772, + "grad_norm": 1.6311752552094194, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 29772 + }, + { + "epoch": 0.29773, + "grad_norm": 1.3520253888738787, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 29773 + }, + { + "epoch": 0.29774, + "grad_norm": 1.4167699149287036, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 29774 + }, + { + "epoch": 0.29775, + "grad_norm": 1.3881068544421047, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 29775 + }, + { + "epoch": 0.29776, + "grad_norm": 1.4090161817183138, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 29776 + }, + { + "epoch": 0.29777, + "grad_norm": 1.499740418580938, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 29777 + }, + { + "epoch": 0.29778, + "grad_norm": 1.352467263079265, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 29778 + }, + { + "epoch": 0.29779, + "grad_norm": 1.5237659512526311, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 29779 + }, + { + "epoch": 0.2978, + "grad_norm": 1.2430306493479109, + "learning_rate": 0.003, + "loss": 4.024, + "step": 29780 + }, + { + "epoch": 0.29781, + "grad_norm": 1.7081985098525676, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 29781 + }, + { + "epoch": 0.29782, + "grad_norm": 1.2825164210414333, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 29782 + }, + { + "epoch": 0.29783, + "grad_norm": 1.862139369218944, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 29783 + }, + { + "epoch": 0.29784, + "grad_norm": 1.2070280253656374, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 29784 + }, + { + "epoch": 0.29785, + "grad_norm": 1.6906374493032834, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 29785 + }, + { + "epoch": 0.29786, + "grad_norm": 1.4914107606210782, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 29786 + }, + { + "epoch": 0.29787, + "grad_norm": 1.3515734844540417, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 29787 + }, + { + "epoch": 0.29788, + "grad_norm": 1.3254108945832006, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 29788 + }, + { + "epoch": 0.29789, + "grad_norm": 1.4639659380189054, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 29789 + }, + { + "epoch": 0.2979, + "grad_norm": 1.3316126147215688, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 29790 + }, + { + "epoch": 0.29791, + "grad_norm": 1.3220311219391556, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 29791 + }, + { + "epoch": 0.29792, + "grad_norm": 1.3133946227323277, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 29792 + }, + { + "epoch": 0.29793, + "grad_norm": 1.6906151627275294, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 29793 + }, + { + "epoch": 0.29794, + "grad_norm": 1.3488719970882173, + "learning_rate": 0.003, + "loss": 4.064, + "step": 29794 + }, + { + "epoch": 0.29795, + "grad_norm": 1.572423195445399, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 29795 + }, + { + "epoch": 0.29796, + "grad_norm": 1.7723987620192916, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 29796 + }, + { + "epoch": 0.29797, + "grad_norm": 1.5586546913774186, + "learning_rate": 0.003, + "loss": 4.048, + "step": 29797 + }, + { + "epoch": 0.29798, + "grad_norm": 1.615677076609843, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 29798 + }, + { + "epoch": 0.29799, + "grad_norm": 1.3227912931583834, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 29799 + }, + { + "epoch": 0.298, + "grad_norm": 1.365656687576291, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 29800 + }, + { + "epoch": 0.29801, + "grad_norm": 1.6626287536503004, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 29801 + }, + { + "epoch": 0.29802, + "grad_norm": 1.2371357255090514, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29802 + }, + { + "epoch": 0.29803, + "grad_norm": 1.6874501116009588, + "learning_rate": 0.003, + "loss": 4.037, + "step": 29803 + }, + { + "epoch": 0.29804, + "grad_norm": 1.1839728168999317, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 29804 + }, + { + "epoch": 0.29805, + "grad_norm": 1.552460864659453, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 29805 + }, + { + "epoch": 0.29806, + "grad_norm": 1.231086963323609, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 29806 + }, + { + "epoch": 0.29807, + "grad_norm": 1.5378581485279348, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 29807 + }, + { + "epoch": 0.29808, + "grad_norm": 1.1895673544814995, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 29808 + }, + { + "epoch": 0.29809, + "grad_norm": 1.2668306045704065, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 29809 + }, + { + "epoch": 0.2981, + "grad_norm": 1.4897315489705347, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 29810 + }, + { + "epoch": 0.29811, + "grad_norm": 1.376356832036761, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 29811 + }, + { + "epoch": 0.29812, + "grad_norm": 1.5118876086596662, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 29812 + }, + { + "epoch": 0.29813, + "grad_norm": 1.6419898062264136, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 29813 + }, + { + "epoch": 0.29814, + "grad_norm": 1.5469288531095668, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 29814 + }, + { + "epoch": 0.29815, + "grad_norm": 1.3875742082212243, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 29815 + }, + { + "epoch": 0.29816, + "grad_norm": 1.3522799427260768, + "learning_rate": 0.003, + "loss": 4.054, + "step": 29816 + }, + { + "epoch": 0.29817, + "grad_norm": 1.6002712629164033, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 29817 + }, + { + "epoch": 0.29818, + "grad_norm": 1.3161542447814663, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 29818 + }, + { + "epoch": 0.29819, + "grad_norm": 1.4278849013308776, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 29819 + }, + { + "epoch": 0.2982, + "grad_norm": 1.5134786267574845, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 29820 + }, + { + "epoch": 0.29821, + "grad_norm": 1.272824908047317, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 29821 + }, + { + "epoch": 0.29822, + "grad_norm": 1.407350022881553, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 29822 + }, + { + "epoch": 0.29823, + "grad_norm": 1.64065425383268, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 29823 + }, + { + "epoch": 0.29824, + "grad_norm": 1.695560597172358, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29824 + }, + { + "epoch": 0.29825, + "grad_norm": 1.2184929072177104, + "learning_rate": 0.003, + "loss": 4.024, + "step": 29825 + }, + { + "epoch": 0.29826, + "grad_norm": 1.5399194088395518, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 29826 + }, + { + "epoch": 0.29827, + "grad_norm": 1.556705891763179, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 29827 + }, + { + "epoch": 0.29828, + "grad_norm": 1.1781229951927157, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 29828 + }, + { + "epoch": 0.29829, + "grad_norm": 1.4698042913950633, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 29829 + }, + { + "epoch": 0.2983, + "grad_norm": 1.4361035746419482, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 29830 + }, + { + "epoch": 0.29831, + "grad_norm": 1.5510524524272893, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 29831 + }, + { + "epoch": 0.29832, + "grad_norm": 1.0979182542183161, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 29832 + }, + { + "epoch": 0.29833, + "grad_norm": 1.505872234781082, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29833 + }, + { + "epoch": 0.29834, + "grad_norm": 1.255859612596669, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 29834 + }, + { + "epoch": 0.29835, + "grad_norm": 1.5998790892446972, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 29835 + }, + { + "epoch": 0.29836, + "grad_norm": 1.3044092197475923, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 29836 + }, + { + "epoch": 0.29837, + "grad_norm": 1.5706801421450445, + "learning_rate": 0.003, + "loss": 4.048, + "step": 29837 + }, + { + "epoch": 0.29838, + "grad_norm": 1.253858642801113, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 29838 + }, + { + "epoch": 0.29839, + "grad_norm": 1.5608621609315658, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 29839 + }, + { + "epoch": 0.2984, + "grad_norm": 1.4417315162955702, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 29840 + }, + { + "epoch": 0.29841, + "grad_norm": 1.4498023829711506, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 29841 + }, + { + "epoch": 0.29842, + "grad_norm": 1.465790193759315, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 29842 + }, + { + "epoch": 0.29843, + "grad_norm": 1.3986571714266878, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 29843 + }, + { + "epoch": 0.29844, + "grad_norm": 1.4035847816362108, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29844 + }, + { + "epoch": 0.29845, + "grad_norm": 1.3298028393629286, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 29845 + }, + { + "epoch": 0.29846, + "grad_norm": 1.6149973260364836, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 29846 + }, + { + "epoch": 0.29847, + "grad_norm": 1.169967842098827, + "learning_rate": 0.003, + "loss": 4.057, + "step": 29847 + }, + { + "epoch": 0.29848, + "grad_norm": 1.642275812399359, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 29848 + }, + { + "epoch": 0.29849, + "grad_norm": 1.037205055510388, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 29849 + }, + { + "epoch": 0.2985, + "grad_norm": 1.7182603873013793, + "learning_rate": 0.003, + "loss": 4.051, + "step": 29850 + }, + { + "epoch": 0.29851, + "grad_norm": 1.5373984327884005, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 29851 + }, + { + "epoch": 0.29852, + "grad_norm": 1.5170450368091923, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 29852 + }, + { + "epoch": 0.29853, + "grad_norm": 1.8644739445533944, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 29853 + }, + { + "epoch": 0.29854, + "grad_norm": 1.4024590056800232, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29854 + }, + { + "epoch": 0.29855, + "grad_norm": 1.522600657724739, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 29855 + }, + { + "epoch": 0.29856, + "grad_norm": 1.382691482349311, + "learning_rate": 0.003, + "loss": 4.085, + "step": 29856 + }, + { + "epoch": 0.29857, + "grad_norm": 1.539194653783551, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 29857 + }, + { + "epoch": 0.29858, + "grad_norm": 1.2036865914789912, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 29858 + }, + { + "epoch": 0.29859, + "grad_norm": 1.3994114821603287, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29859 + }, + { + "epoch": 0.2986, + "grad_norm": 1.4328609248732296, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 29860 + }, + { + "epoch": 0.29861, + "grad_norm": 1.6309026900992087, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 29861 + }, + { + "epoch": 0.29862, + "grad_norm": 1.124859451545883, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 29862 + }, + { + "epoch": 0.29863, + "grad_norm": 1.4710987120858492, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 29863 + }, + { + "epoch": 0.29864, + "grad_norm": 1.430892833340299, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29864 + }, + { + "epoch": 0.29865, + "grad_norm": 1.676789821305284, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 29865 + }, + { + "epoch": 0.29866, + "grad_norm": 1.5019862186511412, + "learning_rate": 0.003, + "loss": 4.073, + "step": 29866 + }, + { + "epoch": 0.29867, + "grad_norm": 1.4898773291956573, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 29867 + }, + { + "epoch": 0.29868, + "grad_norm": 1.2140762251958113, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 29868 + }, + { + "epoch": 0.29869, + "grad_norm": 1.6216736668841458, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 29869 + }, + { + "epoch": 0.2987, + "grad_norm": 1.249231257597406, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 29870 + }, + { + "epoch": 0.29871, + "grad_norm": 1.7172459980411128, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 29871 + }, + { + "epoch": 0.29872, + "grad_norm": 1.3413922464149035, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 29872 + }, + { + "epoch": 0.29873, + "grad_norm": 1.499854568165322, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29873 + }, + { + "epoch": 0.29874, + "grad_norm": 1.4864011783394007, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 29874 + }, + { + "epoch": 0.29875, + "grad_norm": 1.31764964655554, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 29875 + }, + { + "epoch": 0.29876, + "grad_norm": 1.4364473490724712, + "learning_rate": 0.003, + "loss": 4.06, + "step": 29876 + }, + { + "epoch": 0.29877, + "grad_norm": 1.451249209156402, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 29877 + }, + { + "epoch": 0.29878, + "grad_norm": 1.645680308593471, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 29878 + }, + { + "epoch": 0.29879, + "grad_norm": 1.3071630832560144, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 29879 + }, + { + "epoch": 0.2988, + "grad_norm": 1.767913551497664, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 29880 + }, + { + "epoch": 0.29881, + "grad_norm": 1.2209836544804158, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 29881 + }, + { + "epoch": 0.29882, + "grad_norm": 1.6703772660534129, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 29882 + }, + { + "epoch": 0.29883, + "grad_norm": 1.5212806304335673, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 29883 + }, + { + "epoch": 0.29884, + "grad_norm": 1.444574358355813, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 29884 + }, + { + "epoch": 0.29885, + "grad_norm": 1.314467337065314, + "learning_rate": 0.003, + "loss": 4.029, + "step": 29885 + }, + { + "epoch": 0.29886, + "grad_norm": 1.5420538321553805, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 29886 + }, + { + "epoch": 0.29887, + "grad_norm": 1.3979861362912758, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 29887 + }, + { + "epoch": 0.29888, + "grad_norm": 1.6496484052398297, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 29888 + }, + { + "epoch": 0.29889, + "grad_norm": 1.4332294953917788, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 29889 + }, + { + "epoch": 0.2989, + "grad_norm": 1.5747351898777417, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 29890 + }, + { + "epoch": 0.29891, + "grad_norm": 1.335295039532714, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 29891 + }, + { + "epoch": 0.29892, + "grad_norm": 1.1376292416280758, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 29892 + }, + { + "epoch": 0.29893, + "grad_norm": 1.3086986578019058, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 29893 + }, + { + "epoch": 0.29894, + "grad_norm": 1.3944241058439844, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29894 + }, + { + "epoch": 0.29895, + "grad_norm": 1.4512355868509057, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 29895 + }, + { + "epoch": 0.29896, + "grad_norm": 1.6639320439903376, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 29896 + }, + { + "epoch": 0.29897, + "grad_norm": 1.3453733464088429, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 29897 + }, + { + "epoch": 0.29898, + "grad_norm": 1.427627487703273, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29898 + }, + { + "epoch": 0.29899, + "grad_norm": 1.301659533053862, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29899 + }, + { + "epoch": 0.299, + "grad_norm": 1.5407076192363653, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 29900 + }, + { + "epoch": 0.29901, + "grad_norm": 1.3095577090218384, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 29901 + }, + { + "epoch": 0.29902, + "grad_norm": 1.6911837742970903, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29902 + }, + { + "epoch": 0.29903, + "grad_norm": 1.208256063353247, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 29903 + }, + { + "epoch": 0.29904, + "grad_norm": 1.6722395045196825, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 29904 + }, + { + "epoch": 0.29905, + "grad_norm": 1.3363136364402588, + "learning_rate": 0.003, + "loss": 4.05, + "step": 29905 + }, + { + "epoch": 0.29906, + "grad_norm": 1.4868831381509156, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 29906 + }, + { + "epoch": 0.29907, + "grad_norm": 1.4051522096803613, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 29907 + }, + { + "epoch": 0.29908, + "grad_norm": 1.4579570014847432, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 29908 + }, + { + "epoch": 0.29909, + "grad_norm": 1.431971577685788, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 29909 + }, + { + "epoch": 0.2991, + "grad_norm": 1.3709480877586322, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 29910 + }, + { + "epoch": 0.29911, + "grad_norm": 1.4217996866086142, + "learning_rate": 0.003, + "loss": 4.035, + "step": 29911 + }, + { + "epoch": 0.29912, + "grad_norm": 1.5844444184055748, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 29912 + }, + { + "epoch": 0.29913, + "grad_norm": 1.3292028653418058, + "learning_rate": 0.003, + "loss": 4.05, + "step": 29913 + }, + { + "epoch": 0.29914, + "grad_norm": 1.6131062340519604, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 29914 + }, + { + "epoch": 0.29915, + "grad_norm": 1.2667741534528416, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 29915 + }, + { + "epoch": 0.29916, + "grad_norm": 1.465481915200225, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29916 + }, + { + "epoch": 0.29917, + "grad_norm": 1.79116304264325, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 29917 + }, + { + "epoch": 0.29918, + "grad_norm": 1.3951659731182506, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 29918 + }, + { + "epoch": 0.29919, + "grad_norm": 1.5294154913184146, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 29919 + }, + { + "epoch": 0.2992, + "grad_norm": 1.2457464148935378, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 29920 + }, + { + "epoch": 0.29921, + "grad_norm": 1.6237123146067418, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 29921 + }, + { + "epoch": 0.29922, + "grad_norm": 1.3315536920848159, + "learning_rate": 0.003, + "loss": 4.036, + "step": 29922 + }, + { + "epoch": 0.29923, + "grad_norm": 1.3949708809202683, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 29923 + }, + { + "epoch": 0.29924, + "grad_norm": 1.7382930300103172, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 29924 + }, + { + "epoch": 0.29925, + "grad_norm": 1.257172836273197, + "learning_rate": 0.003, + "loss": 4.053, + "step": 29925 + }, + { + "epoch": 0.29926, + "grad_norm": 1.5584515470514664, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 29926 + }, + { + "epoch": 0.29927, + "grad_norm": 1.4691072735907962, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 29927 + }, + { + "epoch": 0.29928, + "grad_norm": 1.453319639955991, + "learning_rate": 0.003, + "loss": 4.003, + "step": 29928 + }, + { + "epoch": 0.29929, + "grad_norm": 1.3623300184739702, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 29929 + }, + { + "epoch": 0.2993, + "grad_norm": 1.4058602906602584, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 29930 + }, + { + "epoch": 0.29931, + "grad_norm": 1.4077542228580715, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 29931 + }, + { + "epoch": 0.29932, + "grad_norm": 1.377782846068373, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 29932 + }, + { + "epoch": 0.29933, + "grad_norm": 1.2305470575496547, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 29933 + }, + { + "epoch": 0.29934, + "grad_norm": 1.5801064913219518, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 29934 + }, + { + "epoch": 0.29935, + "grad_norm": 1.4399248955487527, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 29935 + }, + { + "epoch": 0.29936, + "grad_norm": 1.3771628554309612, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 29936 + }, + { + "epoch": 0.29937, + "grad_norm": 1.4410651297440757, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29937 + }, + { + "epoch": 0.29938, + "grad_norm": 1.3501458817510021, + "learning_rate": 0.003, + "loss": 4.01, + "step": 29938 + }, + { + "epoch": 0.29939, + "grad_norm": 1.6098284806728036, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 29939 + }, + { + "epoch": 0.2994, + "grad_norm": 1.2823430031869558, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 29940 + }, + { + "epoch": 0.29941, + "grad_norm": 1.510874620133719, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 29941 + }, + { + "epoch": 0.29942, + "grad_norm": 1.6732723387130586, + "learning_rate": 0.003, + "loss": 4.057, + "step": 29942 + }, + { + "epoch": 0.29943, + "grad_norm": 1.4751544803988448, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29943 + }, + { + "epoch": 0.29944, + "grad_norm": 1.484408961183351, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 29944 + }, + { + "epoch": 0.29945, + "grad_norm": 1.5137611903027162, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 29945 + }, + { + "epoch": 0.29946, + "grad_norm": 1.345074088250087, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 29946 + }, + { + "epoch": 0.29947, + "grad_norm": 1.5013140467159323, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 29947 + }, + { + "epoch": 0.29948, + "grad_norm": 1.4338500421504101, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 29948 + }, + { + "epoch": 0.29949, + "grad_norm": 1.6976614180947804, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 29949 + }, + { + "epoch": 0.2995, + "grad_norm": 1.4965825059377724, + "learning_rate": 0.003, + "loss": 4.059, + "step": 29950 + }, + { + "epoch": 0.29951, + "grad_norm": 1.3722482132395621, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 29951 + }, + { + "epoch": 0.29952, + "grad_norm": 1.3108364944620055, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 29952 + }, + { + "epoch": 0.29953, + "grad_norm": 1.6312645435135598, + "learning_rate": 0.003, + "loss": 4.03, + "step": 29953 + }, + { + "epoch": 0.29954, + "grad_norm": 1.492939327938371, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 29954 + }, + { + "epoch": 0.29955, + "grad_norm": 1.525923179985819, + "learning_rate": 0.003, + "loss": 4.037, + "step": 29955 + }, + { + "epoch": 0.29956, + "grad_norm": 1.534541786074435, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 29956 + }, + { + "epoch": 0.29957, + "grad_norm": 1.4585743720367044, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 29957 + }, + { + "epoch": 0.29958, + "grad_norm": 1.4172849035702064, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 29958 + }, + { + "epoch": 0.29959, + "grad_norm": 1.3851251878989825, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 29959 + }, + { + "epoch": 0.2996, + "grad_norm": 1.3520292998809036, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 29960 + }, + { + "epoch": 0.29961, + "grad_norm": 1.3538073862472382, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 29961 + }, + { + "epoch": 0.29962, + "grad_norm": 1.3274641414949486, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 29962 + }, + { + "epoch": 0.29963, + "grad_norm": 1.5193996173541795, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 29963 + }, + { + "epoch": 0.29964, + "grad_norm": 1.3195041635329703, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 29964 + }, + { + "epoch": 0.29965, + "grad_norm": 1.4418222659660227, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 29965 + }, + { + "epoch": 0.29966, + "grad_norm": 1.50050249201258, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29966 + }, + { + "epoch": 0.29967, + "grad_norm": 1.5077605887396142, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 29967 + }, + { + "epoch": 0.29968, + "grad_norm": 1.5647398742838736, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 29968 + }, + { + "epoch": 0.29969, + "grad_norm": 1.241027919115298, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29969 + }, + { + "epoch": 0.2997, + "grad_norm": 1.7922485466768519, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 29970 + }, + { + "epoch": 0.29971, + "grad_norm": 1.1095327483644768, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 29971 + }, + { + "epoch": 0.29972, + "grad_norm": 1.6433872396873737, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 29972 + }, + { + "epoch": 0.29973, + "grad_norm": 1.2396671388270264, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 29973 + }, + { + "epoch": 0.29974, + "grad_norm": 1.3979175712825365, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 29974 + }, + { + "epoch": 0.29975, + "grad_norm": 1.5209400791058094, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 29975 + }, + { + "epoch": 0.29976, + "grad_norm": 1.651020275792951, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 29976 + }, + { + "epoch": 0.29977, + "grad_norm": 1.3964920738633264, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 29977 + }, + { + "epoch": 0.29978, + "grad_norm": 1.6658494231510486, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 29978 + }, + { + "epoch": 0.29979, + "grad_norm": 1.4159036369095106, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 29979 + }, + { + "epoch": 0.2998, + "grad_norm": 1.5258483045451985, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 29980 + }, + { + "epoch": 0.29981, + "grad_norm": 1.412023355563673, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 29981 + }, + { + "epoch": 0.29982, + "grad_norm": 1.4602284210019096, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 29982 + }, + { + "epoch": 0.29983, + "grad_norm": 1.5093387441002497, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 29983 + }, + { + "epoch": 0.29984, + "grad_norm": 1.3156505559000122, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29984 + }, + { + "epoch": 0.29985, + "grad_norm": 1.496119587239554, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 29985 + }, + { + "epoch": 0.29986, + "grad_norm": 1.563883603635088, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 29986 + }, + { + "epoch": 0.29987, + "grad_norm": 1.4925443745482736, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 29987 + }, + { + "epoch": 0.29988, + "grad_norm": 1.4423546918460755, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 29988 + }, + { + "epoch": 0.29989, + "grad_norm": 1.5909761776867442, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 29989 + }, + { + "epoch": 0.2999, + "grad_norm": 1.2284140422552634, + "learning_rate": 0.003, + "loss": 4.044, + "step": 29990 + }, + { + "epoch": 0.29991, + "grad_norm": 1.4002451389775301, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 29991 + }, + { + "epoch": 0.29992, + "grad_norm": 1.461079053456257, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 29992 + }, + { + "epoch": 0.29993, + "grad_norm": 1.4468207618402549, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 29993 + }, + { + "epoch": 0.29994, + "grad_norm": 1.4184278432189459, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 29994 + }, + { + "epoch": 0.29995, + "grad_norm": 1.4756563769497941, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29995 + }, + { + "epoch": 0.29996, + "grad_norm": 1.2653463702990144, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 29996 + }, + { + "epoch": 0.29997, + "grad_norm": 1.5218981859904634, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 29997 + }, + { + "epoch": 0.29998, + "grad_norm": 1.2751238060853234, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 29998 + }, + { + "epoch": 0.29999, + "grad_norm": 1.49790038782773, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 29999 + }, + { + "epoch": 0.3, + "grad_norm": 1.268639544600666, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 30000 + }, + { + "epoch": 0.30001, + "grad_norm": 1.4464600118100175, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30001 + }, + { + "epoch": 0.30002, + "grad_norm": 1.447269808600227, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 30002 + }, + { + "epoch": 0.30003, + "grad_norm": 1.2853772434124984, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 30003 + }, + { + "epoch": 0.30004, + "grad_norm": 1.5824693222217188, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 30004 + }, + { + "epoch": 0.30005, + "grad_norm": 1.2466849542205265, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 30005 + }, + { + "epoch": 0.30006, + "grad_norm": 1.7732816779572402, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 30006 + }, + { + "epoch": 0.30007, + "grad_norm": 1.3686586374676348, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 30007 + }, + { + "epoch": 0.30008, + "grad_norm": 1.3722623437752608, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30008 + }, + { + "epoch": 0.30009, + "grad_norm": 1.4918739219248973, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30009 + }, + { + "epoch": 0.3001, + "grad_norm": 1.4378134268344462, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 30010 + }, + { + "epoch": 0.30011, + "grad_norm": 1.4563785881082498, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 30011 + }, + { + "epoch": 0.30012, + "grad_norm": 1.4253366431830023, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 30012 + }, + { + "epoch": 0.30013, + "grad_norm": 1.556253221263865, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30013 + }, + { + "epoch": 0.30014, + "grad_norm": 1.7463141957639319, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 30014 + }, + { + "epoch": 0.30015, + "grad_norm": 1.3702637421594217, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30015 + }, + { + "epoch": 0.30016, + "grad_norm": 1.506313563059671, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 30016 + }, + { + "epoch": 0.30017, + "grad_norm": 1.449972571485044, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 30017 + }, + { + "epoch": 0.30018, + "grad_norm": 1.3939852007429177, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 30018 + }, + { + "epoch": 0.30019, + "grad_norm": 1.2223045234557661, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 30019 + }, + { + "epoch": 0.3002, + "grad_norm": 1.5784230677204911, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 30020 + }, + { + "epoch": 0.30021, + "grad_norm": 1.4101347344643278, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 30021 + }, + { + "epoch": 0.30022, + "grad_norm": 1.5528520128765784, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 30022 + }, + { + "epoch": 0.30023, + "grad_norm": 1.249693488757426, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30023 + }, + { + "epoch": 0.30024, + "grad_norm": 1.7584203943854801, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 30024 + }, + { + "epoch": 0.30025, + "grad_norm": 1.3737839334849518, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 30025 + }, + { + "epoch": 0.30026, + "grad_norm": 1.9712939303785006, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 30026 + }, + { + "epoch": 0.30027, + "grad_norm": 1.0539386453456214, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 30027 + }, + { + "epoch": 0.30028, + "grad_norm": 1.650495945799643, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 30028 + }, + { + "epoch": 0.30029, + "grad_norm": 1.516381613339983, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30029 + }, + { + "epoch": 0.3003, + "grad_norm": 1.4846323449395824, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30030 + }, + { + "epoch": 0.30031, + "grad_norm": 1.47913201476335, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 30031 + }, + { + "epoch": 0.30032, + "grad_norm": 1.7107120947730616, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30032 + }, + { + "epoch": 0.30033, + "grad_norm": 1.2587634515770725, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30033 + }, + { + "epoch": 0.30034, + "grad_norm": 1.4080835771811901, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 30034 + }, + { + "epoch": 0.30035, + "grad_norm": 1.464964424083474, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 30035 + }, + { + "epoch": 0.30036, + "grad_norm": 1.3069249727134276, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 30036 + }, + { + "epoch": 0.30037, + "grad_norm": 1.515406016126011, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30037 + }, + { + "epoch": 0.30038, + "grad_norm": 1.5183602397630094, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 30038 + }, + { + "epoch": 0.30039, + "grad_norm": 1.464151748629389, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 30039 + }, + { + "epoch": 0.3004, + "grad_norm": 1.3593896123430032, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30040 + }, + { + "epoch": 0.30041, + "grad_norm": 1.436735988343557, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30041 + }, + { + "epoch": 0.30042, + "grad_norm": 1.342176917870834, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30042 + }, + { + "epoch": 0.30043, + "grad_norm": 1.818855188063568, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 30043 + }, + { + "epoch": 0.30044, + "grad_norm": 1.2864656099269638, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 30044 + }, + { + "epoch": 0.30045, + "grad_norm": 1.4357104775207346, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 30045 + }, + { + "epoch": 0.30046, + "grad_norm": 1.416595997707662, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30046 + }, + { + "epoch": 0.30047, + "grad_norm": 1.8618400149950118, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 30047 + }, + { + "epoch": 0.30048, + "grad_norm": 0.9673536292793447, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 30048 + }, + { + "epoch": 0.30049, + "grad_norm": 1.8023908240048674, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 30049 + }, + { + "epoch": 0.3005, + "grad_norm": 1.2695805657878076, + "learning_rate": 0.003, + "loss": 4.033, + "step": 30050 + }, + { + "epoch": 0.30051, + "grad_norm": 1.6963638992316141, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 30051 + }, + { + "epoch": 0.30052, + "grad_norm": 1.3844075036149157, + "learning_rate": 0.003, + "loss": 4.055, + "step": 30052 + }, + { + "epoch": 0.30053, + "grad_norm": 1.5037187288353855, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 30053 + }, + { + "epoch": 0.30054, + "grad_norm": 1.3333178077161916, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 30054 + }, + { + "epoch": 0.30055, + "grad_norm": 1.5199544488756482, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30055 + }, + { + "epoch": 0.30056, + "grad_norm": 1.3951100784504555, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 30056 + }, + { + "epoch": 0.30057, + "grad_norm": 1.687689105857649, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30057 + }, + { + "epoch": 0.30058, + "grad_norm": 1.475251292442554, + "learning_rate": 0.003, + "loss": 4.035, + "step": 30058 + }, + { + "epoch": 0.30059, + "grad_norm": 1.3838673255224205, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 30059 + }, + { + "epoch": 0.3006, + "grad_norm": 1.5478331771484752, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 30060 + }, + { + "epoch": 0.30061, + "grad_norm": 1.2851212421951506, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30061 + }, + { + "epoch": 0.30062, + "grad_norm": 1.4812988957942654, + "learning_rate": 0.003, + "loss": 4.035, + "step": 30062 + }, + { + "epoch": 0.30063, + "grad_norm": 1.1048758469025326, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30063 + }, + { + "epoch": 0.30064, + "grad_norm": 1.3270799804348772, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 30064 + }, + { + "epoch": 0.30065, + "grad_norm": 1.2667052566669161, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 30065 + }, + { + "epoch": 0.30066, + "grad_norm": 1.755938332174933, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 30066 + }, + { + "epoch": 0.30067, + "grad_norm": 1.3716704311626753, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30067 + }, + { + "epoch": 0.30068, + "grad_norm": 1.7047761642672097, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 30068 + }, + { + "epoch": 0.30069, + "grad_norm": 1.3892162796020822, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 30069 + }, + { + "epoch": 0.3007, + "grad_norm": 1.488721241211426, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 30070 + }, + { + "epoch": 0.30071, + "grad_norm": 1.3924981879258704, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 30071 + }, + { + "epoch": 0.30072, + "grad_norm": 1.325299976293918, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 30072 + }, + { + "epoch": 0.30073, + "grad_norm": 1.6627044391373647, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30073 + }, + { + "epoch": 0.30074, + "grad_norm": 1.2810451664359859, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 30074 + }, + { + "epoch": 0.30075, + "grad_norm": 1.2964420933912137, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 30075 + }, + { + "epoch": 0.30076, + "grad_norm": 1.642776601849019, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 30076 + }, + { + "epoch": 0.30077, + "grad_norm": 1.4946087601830766, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 30077 + }, + { + "epoch": 0.30078, + "grad_norm": 1.505033676093016, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 30078 + }, + { + "epoch": 0.30079, + "grad_norm": 1.5031854461420795, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 30079 + }, + { + "epoch": 0.3008, + "grad_norm": 1.173192191232246, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 30080 + }, + { + "epoch": 0.30081, + "grad_norm": 1.507528158509832, + "learning_rate": 0.003, + "loss": 4.056, + "step": 30081 + }, + { + "epoch": 0.30082, + "grad_norm": 1.2530647139205886, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 30082 + }, + { + "epoch": 0.30083, + "grad_norm": 1.5870431794380175, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 30083 + }, + { + "epoch": 0.30084, + "grad_norm": 1.918100511752767, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 30084 + }, + { + "epoch": 0.30085, + "grad_norm": 1.2538327966573255, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30085 + }, + { + "epoch": 0.30086, + "grad_norm": 1.7489994098695858, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30086 + }, + { + "epoch": 0.30087, + "grad_norm": 1.41722664821913, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30087 + }, + { + "epoch": 0.30088, + "grad_norm": 1.3646267319458363, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 30088 + }, + { + "epoch": 0.30089, + "grad_norm": 1.600114299276762, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 30089 + }, + { + "epoch": 0.3009, + "grad_norm": 1.2804396522653063, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 30090 + }, + { + "epoch": 0.30091, + "grad_norm": 1.4934114168013248, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 30091 + }, + { + "epoch": 0.30092, + "grad_norm": 1.3237805490384864, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 30092 + }, + { + "epoch": 0.30093, + "grad_norm": 1.4541866084044226, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 30093 + }, + { + "epoch": 0.30094, + "grad_norm": 1.1093436570358657, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 30094 + }, + { + "epoch": 0.30095, + "grad_norm": 1.7158537011151813, + "learning_rate": 0.003, + "loss": 4.034, + "step": 30095 + }, + { + "epoch": 0.30096, + "grad_norm": 1.2477105885472466, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30096 + }, + { + "epoch": 0.30097, + "grad_norm": 1.506518287070167, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 30097 + }, + { + "epoch": 0.30098, + "grad_norm": 1.4016275324021412, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 30098 + }, + { + "epoch": 0.30099, + "grad_norm": 1.3512940151779267, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 30099 + }, + { + "epoch": 0.301, + "grad_norm": 1.907469125943842, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 30100 + }, + { + "epoch": 0.30101, + "grad_norm": 1.1135538437213301, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30101 + }, + { + "epoch": 0.30102, + "grad_norm": 1.6250485244489878, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 30102 + }, + { + "epoch": 0.30103, + "grad_norm": 1.6789171410996548, + "learning_rate": 0.003, + "loss": 4.075, + "step": 30103 + }, + { + "epoch": 0.30104, + "grad_norm": 1.5997931889009904, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30104 + }, + { + "epoch": 0.30105, + "grad_norm": 1.4141974184118842, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 30105 + }, + { + "epoch": 0.30106, + "grad_norm": 1.4485086260342592, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30106 + }, + { + "epoch": 0.30107, + "grad_norm": 1.3019824156825381, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30107 + }, + { + "epoch": 0.30108, + "grad_norm": 1.5093914297501092, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 30108 + }, + { + "epoch": 0.30109, + "grad_norm": 1.3983504470317611, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 30109 + }, + { + "epoch": 0.3011, + "grad_norm": 1.7039877780231725, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 30110 + }, + { + "epoch": 0.30111, + "grad_norm": 1.6028528317344972, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30111 + }, + { + "epoch": 0.30112, + "grad_norm": 1.6235733709077491, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30112 + }, + { + "epoch": 0.30113, + "grad_norm": 1.3603118230375757, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 30113 + }, + { + "epoch": 0.30114, + "grad_norm": 1.1186098209877922, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30114 + }, + { + "epoch": 0.30115, + "grad_norm": 1.6447078493930374, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30115 + }, + { + "epoch": 0.30116, + "grad_norm": 1.2586540314737915, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 30116 + }, + { + "epoch": 0.30117, + "grad_norm": 1.4992858146018706, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 30117 + }, + { + "epoch": 0.30118, + "grad_norm": 1.4829970903223166, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 30118 + }, + { + "epoch": 0.30119, + "grad_norm": 1.299865463632515, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 30119 + }, + { + "epoch": 0.3012, + "grad_norm": 1.6118789671520592, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 30120 + }, + { + "epoch": 0.30121, + "grad_norm": 1.3747267385823325, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 30121 + }, + { + "epoch": 0.30122, + "grad_norm": 1.318267298504243, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 30122 + }, + { + "epoch": 0.30123, + "grad_norm": 1.5022610383114596, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 30123 + }, + { + "epoch": 0.30124, + "grad_norm": 1.3500838410319835, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 30124 + }, + { + "epoch": 0.30125, + "grad_norm": 1.7054427189244752, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30125 + }, + { + "epoch": 0.30126, + "grad_norm": 1.3435037998486576, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 30126 + }, + { + "epoch": 0.30127, + "grad_norm": 1.670710535087531, + "learning_rate": 0.003, + "loss": 4.023, + "step": 30127 + }, + { + "epoch": 0.30128, + "grad_norm": 1.371820498135048, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 30128 + }, + { + "epoch": 0.30129, + "grad_norm": 1.426131534666377, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 30129 + }, + { + "epoch": 0.3013, + "grad_norm": 1.3322395164884522, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 30130 + }, + { + "epoch": 0.30131, + "grad_norm": 1.4457910558029692, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 30131 + }, + { + "epoch": 0.30132, + "grad_norm": 1.4389572423058536, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 30132 + }, + { + "epoch": 0.30133, + "grad_norm": 1.3254467737448132, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 30133 + }, + { + "epoch": 0.30134, + "grad_norm": 1.5135724064485094, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 30134 + }, + { + "epoch": 0.30135, + "grad_norm": 1.2985719911273241, + "learning_rate": 0.003, + "loss": 4.063, + "step": 30135 + }, + { + "epoch": 0.30136, + "grad_norm": 1.2424180918446146, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30136 + }, + { + "epoch": 0.30137, + "grad_norm": 1.3552434816172345, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 30137 + }, + { + "epoch": 0.30138, + "grad_norm": 1.800716611753137, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 30138 + }, + { + "epoch": 0.30139, + "grad_norm": 1.421922252339153, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 30139 + }, + { + "epoch": 0.3014, + "grad_norm": 1.4075714247146056, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 30140 + }, + { + "epoch": 0.30141, + "grad_norm": 1.4117098128900794, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 30141 + }, + { + "epoch": 0.30142, + "grad_norm": 1.4742638884104091, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 30142 + }, + { + "epoch": 0.30143, + "grad_norm": 1.5669752083410435, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 30143 + }, + { + "epoch": 0.30144, + "grad_norm": 1.2002729611779976, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 30144 + }, + { + "epoch": 0.30145, + "grad_norm": 1.5064014476073084, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 30145 + }, + { + "epoch": 0.30146, + "grad_norm": 1.3573409500729299, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 30146 + }, + { + "epoch": 0.30147, + "grad_norm": 1.5600965563608857, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 30147 + }, + { + "epoch": 0.30148, + "grad_norm": 1.7206283058122067, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 30148 + }, + { + "epoch": 0.30149, + "grad_norm": 1.3268929527581999, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 30149 + }, + { + "epoch": 0.3015, + "grad_norm": 1.6111913365022004, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 30150 + }, + { + "epoch": 0.30151, + "grad_norm": 1.3952957740208611, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30151 + }, + { + "epoch": 0.30152, + "grad_norm": 1.227130876432038, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 30152 + }, + { + "epoch": 0.30153, + "grad_norm": 1.4251495550174882, + "learning_rate": 0.003, + "loss": 4.066, + "step": 30153 + }, + { + "epoch": 0.30154, + "grad_norm": 1.607063882428113, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30154 + }, + { + "epoch": 0.30155, + "grad_norm": 1.2521102543406066, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 30155 + }, + { + "epoch": 0.30156, + "grad_norm": 1.5787322512135986, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30156 + }, + { + "epoch": 0.30157, + "grad_norm": 1.4880433642743764, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 30157 + }, + { + "epoch": 0.30158, + "grad_norm": 1.2493579606755623, + "learning_rate": 0.003, + "loss": 4.067, + "step": 30158 + }, + { + "epoch": 0.30159, + "grad_norm": 1.4396485159993084, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 30159 + }, + { + "epoch": 0.3016, + "grad_norm": 1.133065962706903, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 30160 + }, + { + "epoch": 0.30161, + "grad_norm": 1.6115045674273698, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30161 + }, + { + "epoch": 0.30162, + "grad_norm": 1.306316872589409, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30162 + }, + { + "epoch": 0.30163, + "grad_norm": 1.5831761058348788, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30163 + }, + { + "epoch": 0.30164, + "grad_norm": 1.5227003044436649, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 30164 + }, + { + "epoch": 0.30165, + "grad_norm": 1.5499349932517617, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 30165 + }, + { + "epoch": 0.30166, + "grad_norm": 1.3757292197884572, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 30166 + }, + { + "epoch": 0.30167, + "grad_norm": 1.7270044102356672, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30167 + }, + { + "epoch": 0.30168, + "grad_norm": 1.180517326905304, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30168 + }, + { + "epoch": 0.30169, + "grad_norm": 1.6106198588550549, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 30169 + }, + { + "epoch": 0.3017, + "grad_norm": 1.4220083065748022, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30170 + }, + { + "epoch": 0.30171, + "grad_norm": 1.6439283616907436, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30171 + }, + { + "epoch": 0.30172, + "grad_norm": 1.2918045249756571, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30172 + }, + { + "epoch": 0.30173, + "grad_norm": 1.5099161622893393, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 30173 + }, + { + "epoch": 0.30174, + "grad_norm": 1.2157060945324614, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 30174 + }, + { + "epoch": 0.30175, + "grad_norm": 2.0946632146710362, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 30175 + }, + { + "epoch": 0.30176, + "grad_norm": 0.9553384360496355, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 30176 + }, + { + "epoch": 0.30177, + "grad_norm": 1.4528810979794977, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 30177 + }, + { + "epoch": 0.30178, + "grad_norm": 1.4676247614801485, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30178 + }, + { + "epoch": 0.30179, + "grad_norm": 1.152497541361644, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30179 + }, + { + "epoch": 0.3018, + "grad_norm": 1.8370478112417445, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 30180 + }, + { + "epoch": 0.30181, + "grad_norm": 1.291476320655636, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30181 + }, + { + "epoch": 0.30182, + "grad_norm": 1.7107169389944052, + "learning_rate": 0.003, + "loss": 4.085, + "step": 30182 + }, + { + "epoch": 0.30183, + "grad_norm": 1.2908363438035644, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30183 + }, + { + "epoch": 0.30184, + "grad_norm": 1.4715400515695383, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 30184 + }, + { + "epoch": 0.30185, + "grad_norm": 1.4672700933889473, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 30185 + }, + { + "epoch": 0.30186, + "grad_norm": 1.6914850225594256, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 30186 + }, + { + "epoch": 0.30187, + "grad_norm": 1.2823288639121704, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 30187 + }, + { + "epoch": 0.30188, + "grad_norm": 1.5940094069470583, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 30188 + }, + { + "epoch": 0.30189, + "grad_norm": 1.1599934600682207, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30189 + }, + { + "epoch": 0.3019, + "grad_norm": 1.3941793943117857, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 30190 + }, + { + "epoch": 0.30191, + "grad_norm": 1.477164199772091, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 30191 + }, + { + "epoch": 0.30192, + "grad_norm": 1.3934479514215186, + "learning_rate": 0.003, + "loss": 4.049, + "step": 30192 + }, + { + "epoch": 0.30193, + "grad_norm": 1.4062986267734048, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 30193 + }, + { + "epoch": 0.30194, + "grad_norm": 1.6258856976765013, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 30194 + }, + { + "epoch": 0.30195, + "grad_norm": 1.276427091952291, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 30195 + }, + { + "epoch": 0.30196, + "grad_norm": 1.452645326221105, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 30196 + }, + { + "epoch": 0.30197, + "grad_norm": 1.5832189030156703, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30197 + }, + { + "epoch": 0.30198, + "grad_norm": 1.397883514574077, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 30198 + }, + { + "epoch": 0.30199, + "grad_norm": 1.50951184046399, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30199 + }, + { + "epoch": 0.302, + "grad_norm": 1.3424500793824283, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 30200 + }, + { + "epoch": 0.30201, + "grad_norm": 1.4159473828605453, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 30201 + }, + { + "epoch": 0.30202, + "grad_norm": 1.4970712701052553, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30202 + }, + { + "epoch": 0.30203, + "grad_norm": 1.5918693540252886, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30203 + }, + { + "epoch": 0.30204, + "grad_norm": 1.6379982424051003, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30204 + }, + { + "epoch": 0.30205, + "grad_norm": 1.3235433221213366, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30205 + }, + { + "epoch": 0.30206, + "grad_norm": 1.6037039242233988, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 30206 + }, + { + "epoch": 0.30207, + "grad_norm": 1.439324849213266, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 30207 + }, + { + "epoch": 0.30208, + "grad_norm": 1.341991005280487, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 30208 + }, + { + "epoch": 0.30209, + "grad_norm": 1.8879141363880168, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 30209 + }, + { + "epoch": 0.3021, + "grad_norm": 1.3818560191170095, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30210 + }, + { + "epoch": 0.30211, + "grad_norm": 1.7280650090102745, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 30211 + }, + { + "epoch": 0.30212, + "grad_norm": 1.2110513602849107, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 30212 + }, + { + "epoch": 0.30213, + "grad_norm": 1.5664923673351319, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 30213 + }, + { + "epoch": 0.30214, + "grad_norm": 1.475036333431804, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 30214 + }, + { + "epoch": 0.30215, + "grad_norm": 1.4524769988721817, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 30215 + }, + { + "epoch": 0.30216, + "grad_norm": 1.334007726164355, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30216 + }, + { + "epoch": 0.30217, + "grad_norm": 1.3279547598080088, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30217 + }, + { + "epoch": 0.30218, + "grad_norm": 1.6774576803379422, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30218 + }, + { + "epoch": 0.30219, + "grad_norm": 1.2036017547033369, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 30219 + }, + { + "epoch": 0.3022, + "grad_norm": 1.5807924677500877, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 30220 + }, + { + "epoch": 0.30221, + "grad_norm": 1.20120660757266, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 30221 + }, + { + "epoch": 0.30222, + "grad_norm": 1.7517770618888016, + "learning_rate": 0.003, + "loss": 4.081, + "step": 30222 + }, + { + "epoch": 0.30223, + "grad_norm": 1.381016542976933, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 30223 + }, + { + "epoch": 0.30224, + "grad_norm": 1.4518500071977054, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 30224 + }, + { + "epoch": 0.30225, + "grad_norm": 1.6738698923239979, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30225 + }, + { + "epoch": 0.30226, + "grad_norm": 1.4229199366494825, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 30226 + }, + { + "epoch": 0.30227, + "grad_norm": 1.4703101757785355, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 30227 + }, + { + "epoch": 0.30228, + "grad_norm": 1.2950617736798196, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 30228 + }, + { + "epoch": 0.30229, + "grad_norm": 1.5071860400034602, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 30229 + }, + { + "epoch": 0.3023, + "grad_norm": 1.5422967054311474, + "learning_rate": 0.003, + "loss": 4.065, + "step": 30230 + }, + { + "epoch": 0.30231, + "grad_norm": 1.4641272961589025, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 30231 + }, + { + "epoch": 0.30232, + "grad_norm": 1.4284185044604476, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 30232 + }, + { + "epoch": 0.30233, + "grad_norm": 1.410404402987444, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30233 + }, + { + "epoch": 0.30234, + "grad_norm": 1.2563711254189287, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30234 + }, + { + "epoch": 0.30235, + "grad_norm": 1.5483267644323087, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30235 + }, + { + "epoch": 0.30236, + "grad_norm": 1.183157285048396, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30236 + }, + { + "epoch": 0.30237, + "grad_norm": 1.8071798929368532, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 30237 + }, + { + "epoch": 0.30238, + "grad_norm": 1.0577693104146986, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30238 + }, + { + "epoch": 0.30239, + "grad_norm": 1.9183481208484416, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30239 + }, + { + "epoch": 0.3024, + "grad_norm": 1.4081887447509205, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 30240 + }, + { + "epoch": 0.30241, + "grad_norm": 1.5215195930068095, + "learning_rate": 0.003, + "loss": 4.062, + "step": 30241 + }, + { + "epoch": 0.30242, + "grad_norm": 1.560427014162768, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30242 + }, + { + "epoch": 0.30243, + "grad_norm": 1.210785324848156, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 30243 + }, + { + "epoch": 0.30244, + "grad_norm": 1.4475484588279914, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 30244 + }, + { + "epoch": 0.30245, + "grad_norm": 1.4855656460609077, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 30245 + }, + { + "epoch": 0.30246, + "grad_norm": 1.7559074352695134, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 30246 + }, + { + "epoch": 0.30247, + "grad_norm": 1.2618239407368785, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30247 + }, + { + "epoch": 0.30248, + "grad_norm": 1.5136180163474222, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 30248 + }, + { + "epoch": 0.30249, + "grad_norm": 1.4151925846114979, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 30249 + }, + { + "epoch": 0.3025, + "grad_norm": 1.6725587280496843, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 30250 + }, + { + "epoch": 0.30251, + "grad_norm": 1.3053630941989949, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 30251 + }, + { + "epoch": 0.30252, + "grad_norm": 1.6833687627567835, + "learning_rate": 0.003, + "loss": 4.05, + "step": 30252 + }, + { + "epoch": 0.30253, + "grad_norm": 0.97324072161328, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 30253 + }, + { + "epoch": 0.30254, + "grad_norm": 1.7113311436620502, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30254 + }, + { + "epoch": 0.30255, + "grad_norm": 1.3265617542031605, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 30255 + }, + { + "epoch": 0.30256, + "grad_norm": 1.642063766382081, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 30256 + }, + { + "epoch": 0.30257, + "grad_norm": 1.1899185520960514, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30257 + }, + { + "epoch": 0.30258, + "grad_norm": 1.5335391200524342, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 30258 + }, + { + "epoch": 0.30259, + "grad_norm": 1.3302718218715055, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30259 + }, + { + "epoch": 0.3026, + "grad_norm": 2.0700962288970732, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 30260 + }, + { + "epoch": 0.30261, + "grad_norm": 1.2637527779359072, + "learning_rate": 0.003, + "loss": 4.062, + "step": 30261 + }, + { + "epoch": 0.30262, + "grad_norm": 1.8613771664806777, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30262 + }, + { + "epoch": 0.30263, + "grad_norm": 1.2119250982008212, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30263 + }, + { + "epoch": 0.30264, + "grad_norm": 1.337950318020059, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30264 + }, + { + "epoch": 0.30265, + "grad_norm": 1.6287677609902396, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 30265 + }, + { + "epoch": 0.30266, + "grad_norm": 1.2803987975366313, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 30266 + }, + { + "epoch": 0.30267, + "grad_norm": 1.5242152520630365, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 30267 + }, + { + "epoch": 0.30268, + "grad_norm": 1.385303424935767, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 30268 + }, + { + "epoch": 0.30269, + "grad_norm": 1.4405352818287556, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30269 + }, + { + "epoch": 0.3027, + "grad_norm": 1.3340189335140673, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 30270 + }, + { + "epoch": 0.30271, + "grad_norm": 1.3632384003356302, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 30271 + }, + { + "epoch": 0.30272, + "grad_norm": 1.4204238564284013, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 30272 + }, + { + "epoch": 0.30273, + "grad_norm": 1.761631468662593, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 30273 + }, + { + "epoch": 0.30274, + "grad_norm": 1.160779816236195, + "learning_rate": 0.003, + "loss": 4.089, + "step": 30274 + }, + { + "epoch": 0.30275, + "grad_norm": 1.538999756881304, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30275 + }, + { + "epoch": 0.30276, + "grad_norm": 1.6466283836014814, + "learning_rate": 0.003, + "loss": 4.061, + "step": 30276 + }, + { + "epoch": 0.30277, + "grad_norm": 1.4156447791567413, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30277 + }, + { + "epoch": 0.30278, + "grad_norm": 1.5391529743279624, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 30278 + }, + { + "epoch": 0.30279, + "grad_norm": 1.3457529936245616, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 30279 + }, + { + "epoch": 0.3028, + "grad_norm": 1.3153481056055913, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 30280 + }, + { + "epoch": 0.30281, + "grad_norm": 1.738815347618948, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 30281 + }, + { + "epoch": 0.30282, + "grad_norm": 1.586594591296667, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 30282 + }, + { + "epoch": 0.30283, + "grad_norm": 1.604319759353266, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 30283 + }, + { + "epoch": 0.30284, + "grad_norm": 1.9878541643851455, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30284 + }, + { + "epoch": 0.30285, + "grad_norm": 1.1103362843404219, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 30285 + }, + { + "epoch": 0.30286, + "grad_norm": 1.455212424304886, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 30286 + }, + { + "epoch": 0.30287, + "grad_norm": 1.6604943660285718, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 30287 + }, + { + "epoch": 0.30288, + "grad_norm": 1.7291202481737544, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 30288 + }, + { + "epoch": 0.30289, + "grad_norm": 1.058807721944661, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 30289 + }, + { + "epoch": 0.3029, + "grad_norm": 1.6096928507871824, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 30290 + }, + { + "epoch": 0.30291, + "grad_norm": 1.5114798834306282, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 30291 + }, + { + "epoch": 0.30292, + "grad_norm": 1.5704685148950173, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 30292 + }, + { + "epoch": 0.30293, + "grad_norm": 1.655380273809948, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 30293 + }, + { + "epoch": 0.30294, + "grad_norm": 1.568343638138509, + "learning_rate": 0.003, + "loss": 4.08, + "step": 30294 + }, + { + "epoch": 0.30295, + "grad_norm": 1.845873655199643, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30295 + }, + { + "epoch": 0.30296, + "grad_norm": 1.498072374538838, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30296 + }, + { + "epoch": 0.30297, + "grad_norm": 1.305678975025732, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30297 + }, + { + "epoch": 0.30298, + "grad_norm": 1.6056549084525968, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 30298 + }, + { + "epoch": 0.30299, + "grad_norm": 1.3057213377684518, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30299 + }, + { + "epoch": 0.303, + "grad_norm": 1.6367638422110538, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 30300 + }, + { + "epoch": 0.30301, + "grad_norm": 1.1810834292544194, + "learning_rate": 0.003, + "loss": 4.033, + "step": 30301 + }, + { + "epoch": 0.30302, + "grad_norm": 1.3575119693602795, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 30302 + }, + { + "epoch": 0.30303, + "grad_norm": 1.30635562645203, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 30303 + }, + { + "epoch": 0.30304, + "grad_norm": 1.8015776483134782, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 30304 + }, + { + "epoch": 0.30305, + "grad_norm": 1.2491067429545792, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 30305 + }, + { + "epoch": 0.30306, + "grad_norm": 1.6490856250428727, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 30306 + }, + { + "epoch": 0.30307, + "grad_norm": 1.768138032516728, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 30307 + }, + { + "epoch": 0.30308, + "grad_norm": 1.2911212285302702, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30308 + }, + { + "epoch": 0.30309, + "grad_norm": 1.5187651531885555, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 30309 + }, + { + "epoch": 0.3031, + "grad_norm": 1.4409179200224154, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 30310 + }, + { + "epoch": 0.30311, + "grad_norm": 1.5559345956729205, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 30311 + }, + { + "epoch": 0.30312, + "grad_norm": 1.1606494352016776, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 30312 + }, + { + "epoch": 0.30313, + "grad_norm": 1.5695436479431042, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30313 + }, + { + "epoch": 0.30314, + "grad_norm": 1.2610956205629864, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 30314 + }, + { + "epoch": 0.30315, + "grad_norm": 1.485646461524864, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 30315 + }, + { + "epoch": 0.30316, + "grad_norm": 1.258730555931977, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 30316 + }, + { + "epoch": 0.30317, + "grad_norm": 1.3288623817570246, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 30317 + }, + { + "epoch": 0.30318, + "grad_norm": 1.5500078897192822, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 30318 + }, + { + "epoch": 0.30319, + "grad_norm": 1.3928953485117466, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30319 + }, + { + "epoch": 0.3032, + "grad_norm": 1.801377465646676, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30320 + }, + { + "epoch": 0.30321, + "grad_norm": 1.2326960630881083, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30321 + }, + { + "epoch": 0.30322, + "grad_norm": 1.6866069307434186, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 30322 + }, + { + "epoch": 0.30323, + "grad_norm": 1.3553614424555258, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30323 + }, + { + "epoch": 0.30324, + "grad_norm": 1.7505848411010538, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 30324 + }, + { + "epoch": 0.30325, + "grad_norm": 1.1547048478323834, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 30325 + }, + { + "epoch": 0.30326, + "grad_norm": 1.5598873204907029, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 30326 + }, + { + "epoch": 0.30327, + "grad_norm": 1.4522182227068798, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30327 + }, + { + "epoch": 0.30328, + "grad_norm": 1.347987660041549, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 30328 + }, + { + "epoch": 0.30329, + "grad_norm": 1.4698006785587407, + "learning_rate": 0.003, + "loss": 4.05, + "step": 30329 + }, + { + "epoch": 0.3033, + "grad_norm": 1.438486574993766, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 30330 + }, + { + "epoch": 0.30331, + "grad_norm": 1.4108323043275373, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 30331 + }, + { + "epoch": 0.30332, + "grad_norm": 1.3235809968263428, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 30332 + }, + { + "epoch": 0.30333, + "grad_norm": 1.5376189047167599, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 30333 + }, + { + "epoch": 0.30334, + "grad_norm": 1.4655037876747683, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 30334 + }, + { + "epoch": 0.30335, + "grad_norm": 1.6011809970397093, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30335 + }, + { + "epoch": 0.30336, + "grad_norm": 1.3332428120009376, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 30336 + }, + { + "epoch": 0.30337, + "grad_norm": 1.5662187439301571, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 30337 + }, + { + "epoch": 0.30338, + "grad_norm": 1.5041603377588042, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 30338 + }, + { + "epoch": 0.30339, + "grad_norm": 1.169023436678524, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 30339 + }, + { + "epoch": 0.3034, + "grad_norm": 1.6997380409462863, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 30340 + }, + { + "epoch": 0.30341, + "grad_norm": 1.127976138157665, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 30341 + }, + { + "epoch": 0.30342, + "grad_norm": 1.8893453753665364, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 30342 + }, + { + "epoch": 0.30343, + "grad_norm": 1.1838171124820378, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 30343 + }, + { + "epoch": 0.30344, + "grad_norm": 1.5412257801147773, + "learning_rate": 0.003, + "loss": 4.056, + "step": 30344 + }, + { + "epoch": 0.30345, + "grad_norm": 1.4964645500011204, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30345 + }, + { + "epoch": 0.30346, + "grad_norm": 1.3891050988932316, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 30346 + }, + { + "epoch": 0.30347, + "grad_norm": 1.2691753316277083, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 30347 + }, + { + "epoch": 0.30348, + "grad_norm": 1.4400089522204993, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 30348 + }, + { + "epoch": 0.30349, + "grad_norm": 1.2084840930174943, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 30349 + }, + { + "epoch": 0.3035, + "grad_norm": 1.521699881068546, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 30350 + }, + { + "epoch": 0.30351, + "grad_norm": 1.4320046229476224, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 30351 + }, + { + "epoch": 0.30352, + "grad_norm": 1.515284715814854, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30352 + }, + { + "epoch": 0.30353, + "grad_norm": 1.2401001632303486, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 30353 + }, + { + "epoch": 0.30354, + "grad_norm": 1.5258765140190993, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 30354 + }, + { + "epoch": 0.30355, + "grad_norm": 1.4138217337206886, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 30355 + }, + { + "epoch": 0.30356, + "grad_norm": 1.8358618163944154, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30356 + }, + { + "epoch": 0.30357, + "grad_norm": 1.248297137613215, + "learning_rate": 0.003, + "loss": 4.052, + "step": 30357 + }, + { + "epoch": 0.30358, + "grad_norm": 1.5114555347951282, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 30358 + }, + { + "epoch": 0.30359, + "grad_norm": 1.4945954581597058, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30359 + }, + { + "epoch": 0.3036, + "grad_norm": 1.2190281593586108, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30360 + }, + { + "epoch": 0.30361, + "grad_norm": 1.228213490662456, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30361 + }, + { + "epoch": 0.30362, + "grad_norm": 1.3370784866786352, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 30362 + }, + { + "epoch": 0.30363, + "grad_norm": 1.5414033193079308, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30363 + }, + { + "epoch": 0.30364, + "grad_norm": 1.407809806359045, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30364 + }, + { + "epoch": 0.30365, + "grad_norm": 1.5749747449625784, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30365 + }, + { + "epoch": 0.30366, + "grad_norm": 1.1246551184893494, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30366 + }, + { + "epoch": 0.30367, + "grad_norm": 1.4879390842549667, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30367 + }, + { + "epoch": 0.30368, + "grad_norm": 1.5441841665397953, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30368 + }, + { + "epoch": 0.30369, + "grad_norm": 1.7811727550546863, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30369 + }, + { + "epoch": 0.3037, + "grad_norm": 1.2308897774639689, + "learning_rate": 0.003, + "loss": 4.057, + "step": 30370 + }, + { + "epoch": 0.30371, + "grad_norm": 1.5466894758758087, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 30371 + }, + { + "epoch": 0.30372, + "grad_norm": 1.6448782961567916, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 30372 + }, + { + "epoch": 0.30373, + "grad_norm": 1.1401804161434301, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30373 + }, + { + "epoch": 0.30374, + "grad_norm": 1.702446216439237, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 30374 + }, + { + "epoch": 0.30375, + "grad_norm": 1.257792893718816, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 30375 + }, + { + "epoch": 0.30376, + "grad_norm": 1.5136944166003086, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 30376 + }, + { + "epoch": 0.30377, + "grad_norm": 1.3813292895157743, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 30377 + }, + { + "epoch": 0.30378, + "grad_norm": 1.402451736689553, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 30378 + }, + { + "epoch": 0.30379, + "grad_norm": 1.3664496469045935, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 30379 + }, + { + "epoch": 0.3038, + "grad_norm": 1.5381361731701657, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30380 + }, + { + "epoch": 0.30381, + "grad_norm": 1.3815275060252938, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 30381 + }, + { + "epoch": 0.30382, + "grad_norm": 2.0401806241654254, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 30382 + }, + { + "epoch": 0.30383, + "grad_norm": 1.2776188794517225, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 30383 + }, + { + "epoch": 0.30384, + "grad_norm": 1.3009832164041801, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30384 + }, + { + "epoch": 0.30385, + "grad_norm": 1.7275300760679537, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 30385 + }, + { + "epoch": 0.30386, + "grad_norm": 1.4441785643058165, + "learning_rate": 0.003, + "loss": 4.065, + "step": 30386 + }, + { + "epoch": 0.30387, + "grad_norm": 1.3725229018110499, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 30387 + }, + { + "epoch": 0.30388, + "grad_norm": 1.62562652586988, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 30388 + }, + { + "epoch": 0.30389, + "grad_norm": 1.227314713509042, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30389 + }, + { + "epoch": 0.3039, + "grad_norm": 1.3763290133895845, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 30390 + }, + { + "epoch": 0.30391, + "grad_norm": 1.4883059254426922, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30391 + }, + { + "epoch": 0.30392, + "grad_norm": 1.5166273664790901, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 30392 + }, + { + "epoch": 0.30393, + "grad_norm": 1.2205705632685926, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 30393 + }, + { + "epoch": 0.30394, + "grad_norm": 1.7562818927618606, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 30394 + }, + { + "epoch": 0.30395, + "grad_norm": 1.181112828550476, + "learning_rate": 0.003, + "loss": 4.041, + "step": 30395 + }, + { + "epoch": 0.30396, + "grad_norm": 1.5834400562381103, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30396 + }, + { + "epoch": 0.30397, + "grad_norm": 1.2194182259864021, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 30397 + }, + { + "epoch": 0.30398, + "grad_norm": 1.6065551110314789, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30398 + }, + { + "epoch": 0.30399, + "grad_norm": 1.0783465816646598, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 30399 + }, + { + "epoch": 0.304, + "grad_norm": 1.8583528183990607, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30400 + }, + { + "epoch": 0.30401, + "grad_norm": 1.2386524500125902, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30401 + }, + { + "epoch": 0.30402, + "grad_norm": 1.5214292236413451, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30402 + }, + { + "epoch": 0.30403, + "grad_norm": 1.2326404185831101, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 30403 + }, + { + "epoch": 0.30404, + "grad_norm": 1.7178789214049277, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 30404 + }, + { + "epoch": 0.30405, + "grad_norm": 1.2550087177828821, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 30405 + }, + { + "epoch": 0.30406, + "grad_norm": 1.5491020630680192, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 30406 + }, + { + "epoch": 0.30407, + "grad_norm": 1.6033155417822622, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30407 + }, + { + "epoch": 0.30408, + "grad_norm": 1.526235953294545, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 30408 + }, + { + "epoch": 0.30409, + "grad_norm": 1.2398194619465892, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 30409 + }, + { + "epoch": 0.3041, + "grad_norm": 1.5937016083092204, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 30410 + }, + { + "epoch": 0.30411, + "grad_norm": 1.5972460016837653, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 30411 + }, + { + "epoch": 0.30412, + "grad_norm": 1.493489254172365, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 30412 + }, + { + "epoch": 0.30413, + "grad_norm": 1.6059028358121774, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30413 + }, + { + "epoch": 0.30414, + "grad_norm": 1.2288200605925588, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 30414 + }, + { + "epoch": 0.30415, + "grad_norm": 1.5926651694127885, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30415 + }, + { + "epoch": 0.30416, + "grad_norm": 1.205937252204648, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 30416 + }, + { + "epoch": 0.30417, + "grad_norm": 1.688044577497037, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 30417 + }, + { + "epoch": 0.30418, + "grad_norm": 1.4695989052375669, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30418 + }, + { + "epoch": 0.30419, + "grad_norm": 1.514790668376153, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 30419 + }, + { + "epoch": 0.3042, + "grad_norm": 1.3076180114451956, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 30420 + }, + { + "epoch": 0.30421, + "grad_norm": 1.4661228288186239, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 30421 + }, + { + "epoch": 0.30422, + "grad_norm": 1.2636618845725824, + "learning_rate": 0.003, + "loss": 4.044, + "step": 30422 + }, + { + "epoch": 0.30423, + "grad_norm": 1.3352630301040542, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 30423 + }, + { + "epoch": 0.30424, + "grad_norm": 1.3085403053420994, + "learning_rate": 0.003, + "loss": 4.013, + "step": 30424 + }, + { + "epoch": 0.30425, + "grad_norm": 1.6718464011984415, + "learning_rate": 0.003, + "loss": 4.033, + "step": 30425 + }, + { + "epoch": 0.30426, + "grad_norm": 1.657757118727031, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30426 + }, + { + "epoch": 0.30427, + "grad_norm": 1.386832208225331, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 30427 + }, + { + "epoch": 0.30428, + "grad_norm": 1.3917529780460132, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 30428 + }, + { + "epoch": 0.30429, + "grad_norm": 1.5396030669517922, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30429 + }, + { + "epoch": 0.3043, + "grad_norm": 1.2646698345914005, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 30430 + }, + { + "epoch": 0.30431, + "grad_norm": 1.6431706344073904, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30431 + }, + { + "epoch": 0.30432, + "grad_norm": 1.2768102901799296, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 30432 + }, + { + "epoch": 0.30433, + "grad_norm": 1.4348899597843967, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30433 + }, + { + "epoch": 0.30434, + "grad_norm": 1.4904954234729613, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 30434 + }, + { + "epoch": 0.30435, + "grad_norm": 1.3668687686909582, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30435 + }, + { + "epoch": 0.30436, + "grad_norm": 1.2891201665295742, + "learning_rate": 0.003, + "loss": 4.024, + "step": 30436 + }, + { + "epoch": 0.30437, + "grad_norm": 1.6051989742385069, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 30437 + }, + { + "epoch": 0.30438, + "grad_norm": 1.3626160029671346, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 30438 + }, + { + "epoch": 0.30439, + "grad_norm": 1.386876161272624, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 30439 + }, + { + "epoch": 0.3044, + "grad_norm": 1.4324581279838093, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30440 + }, + { + "epoch": 0.30441, + "grad_norm": 1.431869830429368, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30441 + }, + { + "epoch": 0.30442, + "grad_norm": 1.8765974633083025, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 30442 + }, + { + "epoch": 0.30443, + "grad_norm": 1.3169794422149956, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 30443 + }, + { + "epoch": 0.30444, + "grad_norm": 1.4508087376104744, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 30444 + }, + { + "epoch": 0.30445, + "grad_norm": 1.4980634147025185, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 30445 + }, + { + "epoch": 0.30446, + "grad_norm": 1.2585414455259614, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 30446 + }, + { + "epoch": 0.30447, + "grad_norm": 1.498511707619422, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30447 + }, + { + "epoch": 0.30448, + "grad_norm": 1.278553213779063, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 30448 + }, + { + "epoch": 0.30449, + "grad_norm": 1.5073101255465906, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 30449 + }, + { + "epoch": 0.3045, + "grad_norm": 1.4017298268062957, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30450 + }, + { + "epoch": 0.30451, + "grad_norm": 1.5033766640533817, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 30451 + }, + { + "epoch": 0.30452, + "grad_norm": 1.3521191051814194, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 30452 + }, + { + "epoch": 0.30453, + "grad_norm": 1.4584866442058149, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30453 + }, + { + "epoch": 0.30454, + "grad_norm": 1.5431563977394667, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 30454 + }, + { + "epoch": 0.30455, + "grad_norm": 1.9865784788551588, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 30455 + }, + { + "epoch": 0.30456, + "grad_norm": 1.259259023899781, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30456 + }, + { + "epoch": 0.30457, + "grad_norm": 1.708645230289961, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 30457 + }, + { + "epoch": 0.30458, + "grad_norm": 1.4825578652676747, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 30458 + }, + { + "epoch": 0.30459, + "grad_norm": 1.5133071667055407, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 30459 + }, + { + "epoch": 0.3046, + "grad_norm": 1.4224390128884425, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 30460 + }, + { + "epoch": 0.30461, + "grad_norm": 1.5482786850133572, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 30461 + }, + { + "epoch": 0.30462, + "grad_norm": 1.289100823845669, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30462 + }, + { + "epoch": 0.30463, + "grad_norm": 1.615961862430366, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 30463 + }, + { + "epoch": 0.30464, + "grad_norm": 1.4395619365514305, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 30464 + }, + { + "epoch": 0.30465, + "grad_norm": 1.5554530274281146, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 30465 + }, + { + "epoch": 0.30466, + "grad_norm": 1.2109080311665048, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 30466 + }, + { + "epoch": 0.30467, + "grad_norm": 1.3684846257121697, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 30467 + }, + { + "epoch": 0.30468, + "grad_norm": 1.6259587597266536, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 30468 + }, + { + "epoch": 0.30469, + "grad_norm": 1.413666669551902, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 30469 + }, + { + "epoch": 0.3047, + "grad_norm": 1.544053198052247, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30470 + }, + { + "epoch": 0.30471, + "grad_norm": 1.3090412742335247, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 30471 + }, + { + "epoch": 0.30472, + "grad_norm": 1.3427020083528713, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 30472 + }, + { + "epoch": 0.30473, + "grad_norm": 1.4680556766958786, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30473 + }, + { + "epoch": 0.30474, + "grad_norm": 1.2941108768616274, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 30474 + }, + { + "epoch": 0.30475, + "grad_norm": 1.5731868694414763, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 30475 + }, + { + "epoch": 0.30476, + "grad_norm": 1.2031465688367424, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 30476 + }, + { + "epoch": 0.30477, + "grad_norm": 1.857853162945015, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 30477 + }, + { + "epoch": 0.30478, + "grad_norm": 1.0659078029071107, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 30478 + }, + { + "epoch": 0.30479, + "grad_norm": 1.6701987195645205, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 30479 + }, + { + "epoch": 0.3048, + "grad_norm": 1.3530277505484398, + "learning_rate": 0.003, + "loss": 4.054, + "step": 30480 + }, + { + "epoch": 0.30481, + "grad_norm": 1.6107587877211924, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 30481 + }, + { + "epoch": 0.30482, + "grad_norm": 1.5460749459091276, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 30482 + }, + { + "epoch": 0.30483, + "grad_norm": 1.3301671556052619, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30483 + }, + { + "epoch": 0.30484, + "grad_norm": 1.6036343749763444, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 30484 + }, + { + "epoch": 0.30485, + "grad_norm": 1.390112321465274, + "learning_rate": 0.003, + "loss": 4.056, + "step": 30485 + }, + { + "epoch": 0.30486, + "grad_norm": 1.1459098333247573, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30486 + }, + { + "epoch": 0.30487, + "grad_norm": 1.5719728804091389, + "learning_rate": 0.003, + "loss": 4.041, + "step": 30487 + }, + { + "epoch": 0.30488, + "grad_norm": 1.5180071831525175, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 30488 + }, + { + "epoch": 0.30489, + "grad_norm": 1.3109169189225356, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30489 + }, + { + "epoch": 0.3049, + "grad_norm": 1.4289107014548397, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 30490 + }, + { + "epoch": 0.30491, + "grad_norm": 1.4625048245319447, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 30491 + }, + { + "epoch": 0.30492, + "grad_norm": 1.4810115754574877, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 30492 + }, + { + "epoch": 0.30493, + "grad_norm": 1.2130969181188187, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 30493 + }, + { + "epoch": 0.30494, + "grad_norm": 1.8941569328965318, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 30494 + }, + { + "epoch": 0.30495, + "grad_norm": 1.619968394931619, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 30495 + }, + { + "epoch": 0.30496, + "grad_norm": 1.437397341354001, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 30496 + }, + { + "epoch": 0.30497, + "grad_norm": 1.6195075218559254, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 30497 + }, + { + "epoch": 0.30498, + "grad_norm": 1.5001282712652864, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30498 + }, + { + "epoch": 0.30499, + "grad_norm": 1.2548152739485008, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30499 + }, + { + "epoch": 0.305, + "grad_norm": 1.4889925986077643, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 30500 + }, + { + "epoch": 0.30501, + "grad_norm": 1.755113610376422, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 30501 + }, + { + "epoch": 0.30502, + "grad_norm": 1.1440039249694052, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 30502 + }, + { + "epoch": 0.30503, + "grad_norm": 1.6013091590775375, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 30503 + }, + { + "epoch": 0.30504, + "grad_norm": 1.0484484374556249, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 30504 + }, + { + "epoch": 0.30505, + "grad_norm": 1.796779517580672, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 30505 + }, + { + "epoch": 0.30506, + "grad_norm": 1.1490799424743336, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 30506 + }, + { + "epoch": 0.30507, + "grad_norm": 1.7325399384861346, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 30507 + }, + { + "epoch": 0.30508, + "grad_norm": 1.3944934450764592, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 30508 + }, + { + "epoch": 0.30509, + "grad_norm": 1.3340392240289287, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 30509 + }, + { + "epoch": 0.3051, + "grad_norm": 1.3731716079089438, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 30510 + }, + { + "epoch": 0.30511, + "grad_norm": 1.3340224432730083, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30511 + }, + { + "epoch": 0.30512, + "grad_norm": 1.3826047603362908, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 30512 + }, + { + "epoch": 0.30513, + "grad_norm": 1.4680403393117147, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30513 + }, + { + "epoch": 0.30514, + "grad_norm": 1.534166295758493, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30514 + }, + { + "epoch": 0.30515, + "grad_norm": 1.3849356128902202, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30515 + }, + { + "epoch": 0.30516, + "grad_norm": 1.2251161568903433, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 30516 + }, + { + "epoch": 0.30517, + "grad_norm": 1.8530526715598472, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 30517 + }, + { + "epoch": 0.30518, + "grad_norm": 1.5724350951875987, + "learning_rate": 0.003, + "loss": 4.072, + "step": 30518 + }, + { + "epoch": 0.30519, + "grad_norm": 1.2917561913115718, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30519 + }, + { + "epoch": 0.3052, + "grad_norm": 1.558185015835392, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 30520 + }, + { + "epoch": 0.30521, + "grad_norm": 1.9103560384083174, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 30521 + }, + { + "epoch": 0.30522, + "grad_norm": 1.0788478500485095, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30522 + }, + { + "epoch": 0.30523, + "grad_norm": 1.766019266114419, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 30523 + }, + { + "epoch": 0.30524, + "grad_norm": 1.387330369116967, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 30524 + }, + { + "epoch": 0.30525, + "grad_norm": 1.389084919001246, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 30525 + }, + { + "epoch": 0.30526, + "grad_norm": 1.234144780027068, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 30526 + }, + { + "epoch": 0.30527, + "grad_norm": 1.5494172261627686, + "learning_rate": 0.003, + "loss": 4.01, + "step": 30527 + }, + { + "epoch": 0.30528, + "grad_norm": 1.306352567463229, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 30528 + }, + { + "epoch": 0.30529, + "grad_norm": 1.6477925814639216, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 30529 + }, + { + "epoch": 0.3053, + "grad_norm": 1.4002053311800937, + "learning_rate": 0.003, + "loss": 4.061, + "step": 30530 + }, + { + "epoch": 0.30531, + "grad_norm": 1.6198648200289676, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 30531 + }, + { + "epoch": 0.30532, + "grad_norm": 1.3593775875364078, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30532 + }, + { + "epoch": 0.30533, + "grad_norm": 1.6378090399987502, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 30533 + }, + { + "epoch": 0.30534, + "grad_norm": 1.4456952561087366, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 30534 + }, + { + "epoch": 0.30535, + "grad_norm": 1.3108326154754029, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30535 + }, + { + "epoch": 0.30536, + "grad_norm": 1.4919818746318985, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30536 + }, + { + "epoch": 0.30537, + "grad_norm": 1.2949604471315985, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 30537 + }, + { + "epoch": 0.30538, + "grad_norm": 1.444953269083929, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 30538 + }, + { + "epoch": 0.30539, + "grad_norm": 1.1758125259638126, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 30539 + }, + { + "epoch": 0.3054, + "grad_norm": 1.8164180201835014, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 30540 + }, + { + "epoch": 0.30541, + "grad_norm": 1.0339981922539805, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30541 + }, + { + "epoch": 0.30542, + "grad_norm": 1.7623366189184122, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 30542 + }, + { + "epoch": 0.30543, + "grad_norm": 1.152970396706828, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 30543 + }, + { + "epoch": 0.30544, + "grad_norm": 1.6476241382224892, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 30544 + }, + { + "epoch": 0.30545, + "grad_norm": 1.4381588825028044, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30545 + }, + { + "epoch": 0.30546, + "grad_norm": 1.21422782761376, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30546 + }, + { + "epoch": 0.30547, + "grad_norm": 1.393059057444036, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 30547 + }, + { + "epoch": 0.30548, + "grad_norm": 1.3638539226627522, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30548 + }, + { + "epoch": 0.30549, + "grad_norm": 1.5576133528293818, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 30549 + }, + { + "epoch": 0.3055, + "grad_norm": 1.8015230892954666, + "learning_rate": 0.003, + "loss": 4.068, + "step": 30550 + }, + { + "epoch": 0.30551, + "grad_norm": 1.5564418654612961, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 30551 + }, + { + "epoch": 0.30552, + "grad_norm": 1.610495160805222, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 30552 + }, + { + "epoch": 0.30553, + "grad_norm": 1.6716005825320344, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30553 + }, + { + "epoch": 0.30554, + "grad_norm": 1.1134051419689093, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 30554 + }, + { + "epoch": 0.30555, + "grad_norm": 1.625306863130142, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30555 + }, + { + "epoch": 0.30556, + "grad_norm": 1.285125337071396, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 30556 + }, + { + "epoch": 0.30557, + "grad_norm": 1.328844251668422, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30557 + }, + { + "epoch": 0.30558, + "grad_norm": 1.4403469182943403, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 30558 + }, + { + "epoch": 0.30559, + "grad_norm": 1.2643341331202127, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 30559 + }, + { + "epoch": 0.3056, + "grad_norm": 1.5101374154214324, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 30560 + }, + { + "epoch": 0.30561, + "grad_norm": 1.172199717100432, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 30561 + }, + { + "epoch": 0.30562, + "grad_norm": 1.7072495700448478, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 30562 + }, + { + "epoch": 0.30563, + "grad_norm": 1.325346982237122, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 30563 + }, + { + "epoch": 0.30564, + "grad_norm": 1.688256937372465, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 30564 + }, + { + "epoch": 0.30565, + "grad_norm": 1.3351229152864905, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 30565 + }, + { + "epoch": 0.30566, + "grad_norm": 1.587853058326247, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30566 + }, + { + "epoch": 0.30567, + "grad_norm": 1.289315133813073, + "learning_rate": 0.003, + "loss": 4.051, + "step": 30567 + }, + { + "epoch": 0.30568, + "grad_norm": 1.493617721866168, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30568 + }, + { + "epoch": 0.30569, + "grad_norm": 1.4141892991010738, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30569 + }, + { + "epoch": 0.3057, + "grad_norm": 1.3425668408508054, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 30570 + }, + { + "epoch": 0.30571, + "grad_norm": 1.3387826336251587, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30571 + }, + { + "epoch": 0.30572, + "grad_norm": 1.4893317972671887, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 30572 + }, + { + "epoch": 0.30573, + "grad_norm": 1.3616655830127171, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 30573 + }, + { + "epoch": 0.30574, + "grad_norm": 1.7594529469714042, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 30574 + }, + { + "epoch": 0.30575, + "grad_norm": 1.422878861220597, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30575 + }, + { + "epoch": 0.30576, + "grad_norm": 1.6231102060990794, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30576 + }, + { + "epoch": 0.30577, + "grad_norm": 1.4240058478660407, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 30577 + }, + { + "epoch": 0.30578, + "grad_norm": 1.3781910223417015, + "learning_rate": 0.003, + "loss": 4.041, + "step": 30578 + }, + { + "epoch": 0.30579, + "grad_norm": 1.2779073039918056, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 30579 + }, + { + "epoch": 0.3058, + "grad_norm": 1.6755216527215178, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 30580 + }, + { + "epoch": 0.30581, + "grad_norm": 1.5955701935962912, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 30581 + }, + { + "epoch": 0.30582, + "grad_norm": 1.602701429528181, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 30582 + }, + { + "epoch": 0.30583, + "grad_norm": 1.3188752357997195, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 30583 + }, + { + "epoch": 0.30584, + "grad_norm": 1.4503871209976116, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30584 + }, + { + "epoch": 0.30585, + "grad_norm": 1.6141224725232683, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30585 + }, + { + "epoch": 0.30586, + "grad_norm": 1.3940061533612351, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 30586 + }, + { + "epoch": 0.30587, + "grad_norm": 1.5135291588865842, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 30587 + }, + { + "epoch": 0.30588, + "grad_norm": 1.6400684223312747, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 30588 + }, + { + "epoch": 0.30589, + "grad_norm": 1.1950397796372245, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 30589 + }, + { + "epoch": 0.3059, + "grad_norm": 1.404429276077388, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 30590 + }, + { + "epoch": 0.30591, + "grad_norm": 1.3067441404633753, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 30591 + }, + { + "epoch": 0.30592, + "grad_norm": 1.5758303316649633, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30592 + }, + { + "epoch": 0.30593, + "grad_norm": 1.7259331337679527, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 30593 + }, + { + "epoch": 0.30594, + "grad_norm": 1.3404225458627426, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 30594 + }, + { + "epoch": 0.30595, + "grad_norm": 1.4813119822070406, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30595 + }, + { + "epoch": 0.30596, + "grad_norm": 1.6065324575501947, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 30596 + }, + { + "epoch": 0.30597, + "grad_norm": 1.5139748788564067, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30597 + }, + { + "epoch": 0.30598, + "grad_norm": 1.2592284191674323, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30598 + }, + { + "epoch": 0.30599, + "grad_norm": 1.5460599366557928, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30599 + }, + { + "epoch": 0.306, + "grad_norm": 1.551895757101562, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 30600 + }, + { + "epoch": 0.30601, + "grad_norm": 1.2207458058317704, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 30601 + }, + { + "epoch": 0.30602, + "grad_norm": 1.6427877126222896, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30602 + }, + { + "epoch": 0.30603, + "grad_norm": 1.183751642119822, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 30603 + }, + { + "epoch": 0.30604, + "grad_norm": 1.7354260583349446, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 30604 + }, + { + "epoch": 0.30605, + "grad_norm": 1.3566119834901438, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 30605 + }, + { + "epoch": 0.30606, + "grad_norm": 1.5712428756994024, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 30606 + }, + { + "epoch": 0.30607, + "grad_norm": 1.3891530190721175, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30607 + }, + { + "epoch": 0.30608, + "grad_norm": 1.7965658592711768, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30608 + }, + { + "epoch": 0.30609, + "grad_norm": 1.7599924188213212, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 30609 + }, + { + "epoch": 0.3061, + "grad_norm": 1.2864190771167758, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 30610 + }, + { + "epoch": 0.30611, + "grad_norm": 1.6176269057068318, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30611 + }, + { + "epoch": 0.30612, + "grad_norm": 1.4959901612784379, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 30612 + }, + { + "epoch": 0.30613, + "grad_norm": 1.3802739065597767, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 30613 + }, + { + "epoch": 0.30614, + "grad_norm": 1.4252960309327793, + "learning_rate": 0.003, + "loss": 4.013, + "step": 30614 + }, + { + "epoch": 0.30615, + "grad_norm": 1.428984776181122, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 30615 + }, + { + "epoch": 0.30616, + "grad_norm": 1.5016901897646056, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30616 + }, + { + "epoch": 0.30617, + "grad_norm": 1.51674152245335, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30617 + }, + { + "epoch": 0.30618, + "grad_norm": 1.3314323896361546, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 30618 + }, + { + "epoch": 0.30619, + "grad_norm": 1.5411218626776457, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 30619 + }, + { + "epoch": 0.3062, + "grad_norm": 1.2174723191204053, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 30620 + }, + { + "epoch": 0.30621, + "grad_norm": 1.5976805722714518, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30621 + }, + { + "epoch": 0.30622, + "grad_norm": 1.4213601348509506, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30622 + }, + { + "epoch": 0.30623, + "grad_norm": 1.7044958298417767, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 30623 + }, + { + "epoch": 0.30624, + "grad_norm": 1.4737648373864038, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 30624 + }, + { + "epoch": 0.30625, + "grad_norm": 1.4421118500182348, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 30625 + }, + { + "epoch": 0.30626, + "grad_norm": 1.5182348993397452, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30626 + }, + { + "epoch": 0.30627, + "grad_norm": 1.4862290890544252, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 30627 + }, + { + "epoch": 0.30628, + "grad_norm": 1.3470297200342107, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30628 + }, + { + "epoch": 0.30629, + "grad_norm": 1.4005869964763884, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30629 + }, + { + "epoch": 0.3063, + "grad_norm": 1.3112499944916234, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 30630 + }, + { + "epoch": 0.30631, + "grad_norm": 1.7297816223419713, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 30631 + }, + { + "epoch": 0.30632, + "grad_norm": 1.017941047054487, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30632 + }, + { + "epoch": 0.30633, + "grad_norm": 1.8679138247524223, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 30633 + }, + { + "epoch": 0.30634, + "grad_norm": 1.5276295584781565, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30634 + }, + { + "epoch": 0.30635, + "grad_norm": 1.418683076962579, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 30635 + }, + { + "epoch": 0.30636, + "grad_norm": 1.7067992005177854, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 30636 + }, + { + "epoch": 0.30637, + "grad_norm": 1.431392214563957, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 30637 + }, + { + "epoch": 0.30638, + "grad_norm": 1.7273110084630972, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 30638 + }, + { + "epoch": 0.30639, + "grad_norm": 1.395160244254288, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30639 + }, + { + "epoch": 0.3064, + "grad_norm": 1.4610108802086512, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 30640 + }, + { + "epoch": 0.30641, + "grad_norm": 1.7872412848741899, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 30641 + }, + { + "epoch": 0.30642, + "grad_norm": 1.1125114294016583, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30642 + }, + { + "epoch": 0.30643, + "grad_norm": 1.4904589593433633, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 30643 + }, + { + "epoch": 0.30644, + "grad_norm": 1.2766006987583456, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30644 + }, + { + "epoch": 0.30645, + "grad_norm": 1.3088340161855072, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 30645 + }, + { + "epoch": 0.30646, + "grad_norm": 1.3036187201411835, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 30646 + }, + { + "epoch": 0.30647, + "grad_norm": 1.669951890434241, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30647 + }, + { + "epoch": 0.30648, + "grad_norm": 1.7304532334148013, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30648 + }, + { + "epoch": 0.30649, + "grad_norm": 1.3690118406565026, + "learning_rate": 0.003, + "loss": 4.064, + "step": 30649 + }, + { + "epoch": 0.3065, + "grad_norm": 1.715208466059429, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 30650 + }, + { + "epoch": 0.30651, + "grad_norm": 1.232262927653259, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 30651 + }, + { + "epoch": 0.30652, + "grad_norm": 1.4107033620407343, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 30652 + }, + { + "epoch": 0.30653, + "grad_norm": 1.469692626327123, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 30653 + }, + { + "epoch": 0.30654, + "grad_norm": 1.2286638967021424, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 30654 + }, + { + "epoch": 0.30655, + "grad_norm": 1.8002551421903406, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30655 + }, + { + "epoch": 0.30656, + "grad_norm": 1.1682973781091543, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 30656 + }, + { + "epoch": 0.30657, + "grad_norm": 1.9332569185431694, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30657 + }, + { + "epoch": 0.30658, + "grad_norm": 1.3142793748375847, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 30658 + }, + { + "epoch": 0.30659, + "grad_norm": 1.5061384968519036, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 30659 + }, + { + "epoch": 0.3066, + "grad_norm": 1.352087769893709, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 30660 + }, + { + "epoch": 0.30661, + "grad_norm": 1.2467843542515067, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 30661 + }, + { + "epoch": 0.30662, + "grad_norm": 1.3448329245734758, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30662 + }, + { + "epoch": 0.30663, + "grad_norm": 1.3957810958044765, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 30663 + }, + { + "epoch": 0.30664, + "grad_norm": 1.6567586853440528, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 30664 + }, + { + "epoch": 0.30665, + "grad_norm": 1.506852996771263, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 30665 + }, + { + "epoch": 0.30666, + "grad_norm": 2.0745521723877633, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 30666 + }, + { + "epoch": 0.30667, + "grad_norm": 1.302959181527717, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 30667 + }, + { + "epoch": 0.30668, + "grad_norm": 1.4892122993758354, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 30668 + }, + { + "epoch": 0.30669, + "grad_norm": 1.1040921671072168, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 30669 + }, + { + "epoch": 0.3067, + "grad_norm": 1.6972605655752802, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 30670 + }, + { + "epoch": 0.30671, + "grad_norm": 1.451263475587716, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30671 + }, + { + "epoch": 0.30672, + "grad_norm": 1.5196804253896015, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 30672 + }, + { + "epoch": 0.30673, + "grad_norm": 1.3469670890475345, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 30673 + }, + { + "epoch": 0.30674, + "grad_norm": 1.377683719848872, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 30674 + }, + { + "epoch": 0.30675, + "grad_norm": 1.3376275244778975, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 30675 + }, + { + "epoch": 0.30676, + "grad_norm": 1.4980329272537531, + "learning_rate": 0.003, + "loss": 4.05, + "step": 30676 + }, + { + "epoch": 0.30677, + "grad_norm": 1.3426619435487313, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 30677 + }, + { + "epoch": 0.30678, + "grad_norm": 1.3687116441856342, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30678 + }, + { + "epoch": 0.30679, + "grad_norm": 1.4014634709639417, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 30679 + }, + { + "epoch": 0.3068, + "grad_norm": 1.2329884180867163, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 30680 + }, + { + "epoch": 0.30681, + "grad_norm": 1.5664521614638194, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 30681 + }, + { + "epoch": 0.30682, + "grad_norm": 1.2429113888566135, + "learning_rate": 0.003, + "loss": 4.067, + "step": 30682 + }, + { + "epoch": 0.30683, + "grad_norm": 1.8206682868751587, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 30683 + }, + { + "epoch": 0.30684, + "grad_norm": 1.1186884904942216, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 30684 + }, + { + "epoch": 0.30685, + "grad_norm": 1.6185502848482018, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 30685 + }, + { + "epoch": 0.30686, + "grad_norm": 1.376249429068746, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 30686 + }, + { + "epoch": 0.30687, + "grad_norm": 1.6815165800176577, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30687 + }, + { + "epoch": 0.30688, + "grad_norm": 1.4402070422195847, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 30688 + }, + { + "epoch": 0.30689, + "grad_norm": 1.419859678383165, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30689 + }, + { + "epoch": 0.3069, + "grad_norm": 1.6572356569268867, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30690 + }, + { + "epoch": 0.30691, + "grad_norm": 1.1589155382223726, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 30691 + }, + { + "epoch": 0.30692, + "grad_norm": 2.1183888817483187, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 30692 + }, + { + "epoch": 0.30693, + "grad_norm": 1.3638548626222255, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 30693 + }, + { + "epoch": 0.30694, + "grad_norm": 1.8179702845840906, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 30694 + }, + { + "epoch": 0.30695, + "grad_norm": 1.4293591666296166, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30695 + }, + { + "epoch": 0.30696, + "grad_norm": 1.5773753027505797, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 30696 + }, + { + "epoch": 0.30697, + "grad_norm": 1.6022991244048221, + "learning_rate": 0.003, + "loss": 4.065, + "step": 30697 + }, + { + "epoch": 0.30698, + "grad_norm": 1.3108413549610598, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30698 + }, + { + "epoch": 0.30699, + "grad_norm": 1.1619778735347701, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 30699 + }, + { + "epoch": 0.307, + "grad_norm": 1.4168532160862806, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 30700 + }, + { + "epoch": 0.30701, + "grad_norm": 1.5905666003872732, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 30701 + }, + { + "epoch": 0.30702, + "grad_norm": 1.6590193722713948, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 30702 + }, + { + "epoch": 0.30703, + "grad_norm": 1.39353406599917, + "learning_rate": 0.003, + "loss": 4.062, + "step": 30703 + }, + { + "epoch": 0.30704, + "grad_norm": 1.5928577887899364, + "learning_rate": 0.003, + "loss": 4.033, + "step": 30704 + }, + { + "epoch": 0.30705, + "grad_norm": 1.539122780620397, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 30705 + }, + { + "epoch": 0.30706, + "grad_norm": 1.465639142756281, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 30706 + }, + { + "epoch": 0.30707, + "grad_norm": 1.1599315073961274, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 30707 + }, + { + "epoch": 0.30708, + "grad_norm": 1.467151061220592, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30708 + }, + { + "epoch": 0.30709, + "grad_norm": 1.3023501986968442, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 30709 + }, + { + "epoch": 0.3071, + "grad_norm": 1.5340608542354106, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30710 + }, + { + "epoch": 0.30711, + "grad_norm": 1.2302850566812602, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 30711 + }, + { + "epoch": 0.30712, + "grad_norm": 1.6350425571193519, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30712 + }, + { + "epoch": 0.30713, + "grad_norm": 1.4040937755830223, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 30713 + }, + { + "epoch": 0.30714, + "grad_norm": 1.3992292408670555, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 30714 + }, + { + "epoch": 0.30715, + "grad_norm": 1.4066787700161956, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30715 + }, + { + "epoch": 0.30716, + "grad_norm": 1.366547086852613, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 30716 + }, + { + "epoch": 0.30717, + "grad_norm": 1.3308936120936925, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30717 + }, + { + "epoch": 0.30718, + "grad_norm": 1.5837766621262466, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 30718 + }, + { + "epoch": 0.30719, + "grad_norm": 1.556959386098069, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30719 + }, + { + "epoch": 0.3072, + "grad_norm": 1.478546500165145, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 30720 + }, + { + "epoch": 0.30721, + "grad_norm": 1.347731376132108, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30721 + }, + { + "epoch": 0.30722, + "grad_norm": 1.7063226835076555, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 30722 + }, + { + "epoch": 0.30723, + "grad_norm": 1.2603993597491163, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 30723 + }, + { + "epoch": 0.30724, + "grad_norm": 1.4604595450189417, + "learning_rate": 0.003, + "loss": 4.047, + "step": 30724 + }, + { + "epoch": 0.30725, + "grad_norm": 1.6019030768830362, + "learning_rate": 0.003, + "loss": 4.054, + "step": 30725 + }, + { + "epoch": 0.30726, + "grad_norm": 1.3225209347435707, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 30726 + }, + { + "epoch": 0.30727, + "grad_norm": 1.6050973017379433, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 30727 + }, + { + "epoch": 0.30728, + "grad_norm": 1.6130439023804795, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30728 + }, + { + "epoch": 0.30729, + "grad_norm": 1.3480488653456764, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 30729 + }, + { + "epoch": 0.3073, + "grad_norm": 1.4101774457711571, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 30730 + }, + { + "epoch": 0.30731, + "grad_norm": 1.2713671608507575, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30731 + }, + { + "epoch": 0.30732, + "grad_norm": 1.5250569143698864, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30732 + }, + { + "epoch": 0.30733, + "grad_norm": 1.281725400649618, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30733 + }, + { + "epoch": 0.30734, + "grad_norm": 1.626828667449555, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 30734 + }, + { + "epoch": 0.30735, + "grad_norm": 1.412427238748577, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 30735 + }, + { + "epoch": 0.30736, + "grad_norm": 1.6514242191913984, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 30736 + }, + { + "epoch": 0.30737, + "grad_norm": 1.2565662865184324, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 30737 + }, + { + "epoch": 0.30738, + "grad_norm": 1.5411056199615218, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 30738 + }, + { + "epoch": 0.30739, + "grad_norm": 1.2683848541283942, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 30739 + }, + { + "epoch": 0.3074, + "grad_norm": 1.5733001426331328, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 30740 + }, + { + "epoch": 0.30741, + "grad_norm": 1.53136226349309, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 30741 + }, + { + "epoch": 0.30742, + "grad_norm": 1.3472238768832057, + "learning_rate": 0.003, + "loss": 4.043, + "step": 30742 + }, + { + "epoch": 0.30743, + "grad_norm": 1.5166788477448294, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 30743 + }, + { + "epoch": 0.30744, + "grad_norm": 1.3032811119740857, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 30744 + }, + { + "epoch": 0.30745, + "grad_norm": 1.5283193142642129, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 30745 + }, + { + "epoch": 0.30746, + "grad_norm": 1.354857267778998, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30746 + }, + { + "epoch": 0.30747, + "grad_norm": 1.2516490645669465, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 30747 + }, + { + "epoch": 0.30748, + "grad_norm": 1.3987175262464517, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 30748 + }, + { + "epoch": 0.30749, + "grad_norm": 1.8157155475028024, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30749 + }, + { + "epoch": 0.3075, + "grad_norm": 1.6221285434084867, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30750 + }, + { + "epoch": 0.30751, + "grad_norm": 1.3753768054402935, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 30751 + }, + { + "epoch": 0.30752, + "grad_norm": 1.7758132973568472, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30752 + }, + { + "epoch": 0.30753, + "grad_norm": 1.1482604265333431, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 30753 + }, + { + "epoch": 0.30754, + "grad_norm": 1.7490114641101788, + "learning_rate": 0.003, + "loss": 4.025, + "step": 30754 + }, + { + "epoch": 0.30755, + "grad_norm": 1.2845767838703008, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 30755 + }, + { + "epoch": 0.30756, + "grad_norm": 1.4802002102848129, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 30756 + }, + { + "epoch": 0.30757, + "grad_norm": 1.4391984471943564, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30757 + }, + { + "epoch": 0.30758, + "grad_norm": 1.5104514058467924, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 30758 + }, + { + "epoch": 0.30759, + "grad_norm": 1.2332198075807055, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 30759 + }, + { + "epoch": 0.3076, + "grad_norm": 1.6084399786937973, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 30760 + }, + { + "epoch": 0.30761, + "grad_norm": 1.5113231233497177, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 30761 + }, + { + "epoch": 0.30762, + "grad_norm": 1.3767147166632734, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30762 + }, + { + "epoch": 0.30763, + "grad_norm": 1.6724846282508639, + "learning_rate": 0.003, + "loss": 4.06, + "step": 30763 + }, + { + "epoch": 0.30764, + "grad_norm": 1.119157599302319, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30764 + }, + { + "epoch": 0.30765, + "grad_norm": 1.6999974099814605, + "learning_rate": 0.003, + "loss": 4.036, + "step": 30765 + }, + { + "epoch": 0.30766, + "grad_norm": 1.50733255828746, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30766 + }, + { + "epoch": 0.30767, + "grad_norm": 1.5097928977363233, + "learning_rate": 0.003, + "loss": 4.084, + "step": 30767 + }, + { + "epoch": 0.30768, + "grad_norm": 1.3952716748814653, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 30768 + }, + { + "epoch": 0.30769, + "grad_norm": 1.4858901438162093, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 30769 + }, + { + "epoch": 0.3077, + "grad_norm": 1.659997899323977, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 30770 + }, + { + "epoch": 0.30771, + "grad_norm": 1.4804743606052702, + "learning_rate": 0.003, + "loss": 4.013, + "step": 30771 + }, + { + "epoch": 0.30772, + "grad_norm": 1.589289664666177, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 30772 + }, + { + "epoch": 0.30773, + "grad_norm": 1.355679449224217, + "learning_rate": 0.003, + "loss": 4.037, + "step": 30773 + }, + { + "epoch": 0.30774, + "grad_norm": 1.5426789847597988, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 30774 + }, + { + "epoch": 0.30775, + "grad_norm": 1.6764263908991246, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30775 + }, + { + "epoch": 0.30776, + "grad_norm": 1.0338054799935417, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 30776 + }, + { + "epoch": 0.30777, + "grad_norm": 1.883846101768387, + "learning_rate": 0.003, + "loss": 4.041, + "step": 30777 + }, + { + "epoch": 0.30778, + "grad_norm": 1.153721076352165, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 30778 + }, + { + "epoch": 0.30779, + "grad_norm": 1.774426096914145, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 30779 + }, + { + "epoch": 0.3078, + "grad_norm": 1.4798557631411364, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30780 + }, + { + "epoch": 0.30781, + "grad_norm": 1.456109041171837, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 30781 + }, + { + "epoch": 0.30782, + "grad_norm": 1.4213066774494072, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 30782 + }, + { + "epoch": 0.30783, + "grad_norm": 1.3622676115403967, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 30783 + }, + { + "epoch": 0.30784, + "grad_norm": 1.3288125807598548, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 30784 + }, + { + "epoch": 0.30785, + "grad_norm": 1.3087455658881535, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 30785 + }, + { + "epoch": 0.30786, + "grad_norm": 1.484181294790671, + "learning_rate": 0.003, + "loss": 4.06, + "step": 30786 + }, + { + "epoch": 0.30787, + "grad_norm": 1.2305415096851118, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 30787 + }, + { + "epoch": 0.30788, + "grad_norm": 1.546833541148759, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 30788 + }, + { + "epoch": 0.30789, + "grad_norm": 1.2095132913944924, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 30789 + }, + { + "epoch": 0.3079, + "grad_norm": 1.6074375767579396, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 30790 + }, + { + "epoch": 0.30791, + "grad_norm": 1.3868298766719234, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 30791 + }, + { + "epoch": 0.30792, + "grad_norm": 1.485325466016135, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 30792 + }, + { + "epoch": 0.30793, + "grad_norm": 1.3721231869436494, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 30793 + }, + { + "epoch": 0.30794, + "grad_norm": 1.4170513199704478, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30794 + }, + { + "epoch": 0.30795, + "grad_norm": 1.372951711611916, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 30795 + }, + { + "epoch": 0.30796, + "grad_norm": 1.52746725620711, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 30796 + }, + { + "epoch": 0.30797, + "grad_norm": 1.5155874837357275, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 30797 + }, + { + "epoch": 0.30798, + "grad_norm": 1.3940465701278362, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 30798 + }, + { + "epoch": 0.30799, + "grad_norm": 1.523788808320051, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 30799 + }, + { + "epoch": 0.308, + "grad_norm": 1.6622861053510598, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 30800 + }, + { + "epoch": 0.30801, + "grad_norm": 1.1122838827393213, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30801 + }, + { + "epoch": 0.30802, + "grad_norm": 1.567818556343114, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 30802 + }, + { + "epoch": 0.30803, + "grad_norm": 1.43548035775555, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 30803 + }, + { + "epoch": 0.30804, + "grad_norm": 1.2472387822802684, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 30804 + }, + { + "epoch": 0.30805, + "grad_norm": 1.5831413791120001, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30805 + }, + { + "epoch": 0.30806, + "grad_norm": 1.3752683184153054, + "learning_rate": 0.003, + "loss": 4.045, + "step": 30806 + }, + { + "epoch": 0.30807, + "grad_norm": 1.661436019636817, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 30807 + }, + { + "epoch": 0.30808, + "grad_norm": 1.4250294373978454, + "learning_rate": 0.003, + "loss": 4.023, + "step": 30808 + }, + { + "epoch": 0.30809, + "grad_norm": 1.6127096188382228, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 30809 + }, + { + "epoch": 0.3081, + "grad_norm": 1.312369191361544, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30810 + }, + { + "epoch": 0.30811, + "grad_norm": 1.7695461519153282, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30811 + }, + { + "epoch": 0.30812, + "grad_norm": 1.3798949062337673, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 30812 + }, + { + "epoch": 0.30813, + "grad_norm": 1.3774310286323768, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30813 + }, + { + "epoch": 0.30814, + "grad_norm": 1.2097704960187636, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 30814 + }, + { + "epoch": 0.30815, + "grad_norm": 1.8686454069810805, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 30815 + }, + { + "epoch": 0.30816, + "grad_norm": 1.474268943507354, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 30816 + }, + { + "epoch": 0.30817, + "grad_norm": 1.2850669230345488, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 30817 + }, + { + "epoch": 0.30818, + "grad_norm": 1.4243293125342298, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 30818 + }, + { + "epoch": 0.30819, + "grad_norm": 1.3565108586156096, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 30819 + }, + { + "epoch": 0.3082, + "grad_norm": 1.6791256756040416, + "learning_rate": 0.003, + "loss": 4.059, + "step": 30820 + }, + { + "epoch": 0.30821, + "grad_norm": 1.3842056741052475, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 30821 + }, + { + "epoch": 0.30822, + "grad_norm": 1.4872250636714175, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 30822 + }, + { + "epoch": 0.30823, + "grad_norm": 1.4815694078607018, + "learning_rate": 0.003, + "loss": 4.044, + "step": 30823 + }, + { + "epoch": 0.30824, + "grad_norm": 1.4516751771529022, + "learning_rate": 0.003, + "loss": 4.034, + "step": 30824 + }, + { + "epoch": 0.30825, + "grad_norm": 1.5822022002954097, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30825 + }, + { + "epoch": 0.30826, + "grad_norm": 1.5289353199137063, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 30826 + }, + { + "epoch": 0.30827, + "grad_norm": 1.3953618689424252, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30827 + }, + { + "epoch": 0.30828, + "grad_norm": 1.4713997492266249, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30828 + }, + { + "epoch": 0.30829, + "grad_norm": 1.3252710161019732, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 30829 + }, + { + "epoch": 0.3083, + "grad_norm": 1.7169193795359738, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 30830 + }, + { + "epoch": 0.30831, + "grad_norm": 1.299640433540319, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 30831 + }, + { + "epoch": 0.30832, + "grad_norm": 1.5257509876477158, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 30832 + }, + { + "epoch": 0.30833, + "grad_norm": 1.5516909222026625, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 30833 + }, + { + "epoch": 0.30834, + "grad_norm": 1.4679685843424948, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 30834 + }, + { + "epoch": 0.30835, + "grad_norm": 1.6297564355552185, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 30835 + }, + { + "epoch": 0.30836, + "grad_norm": 1.1987747066344672, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 30836 + }, + { + "epoch": 0.30837, + "grad_norm": 1.5615291387130916, + "learning_rate": 0.003, + "loss": 4.073, + "step": 30837 + }, + { + "epoch": 0.30838, + "grad_norm": 1.4459333953343618, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 30838 + }, + { + "epoch": 0.30839, + "grad_norm": 1.3217076890999766, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 30839 + }, + { + "epoch": 0.3084, + "grad_norm": 1.5390438202668573, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30840 + }, + { + "epoch": 0.30841, + "grad_norm": 1.571552149566802, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 30841 + }, + { + "epoch": 0.30842, + "grad_norm": 1.5132392820920566, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 30842 + }, + { + "epoch": 0.30843, + "grad_norm": 1.307773764559888, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 30843 + }, + { + "epoch": 0.30844, + "grad_norm": 1.6079892935170903, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30844 + }, + { + "epoch": 0.30845, + "grad_norm": 1.435921885007686, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 30845 + }, + { + "epoch": 0.30846, + "grad_norm": 1.4747423880458823, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 30846 + }, + { + "epoch": 0.30847, + "grad_norm": 1.330857544346197, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 30847 + }, + { + "epoch": 0.30848, + "grad_norm": 1.5140022526201693, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 30848 + }, + { + "epoch": 0.30849, + "grad_norm": 1.1089617960793208, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 30849 + }, + { + "epoch": 0.3085, + "grad_norm": 1.6973131869395024, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 30850 + }, + { + "epoch": 0.30851, + "grad_norm": 1.3039408316925316, + "learning_rate": 0.003, + "loss": 4.039, + "step": 30851 + }, + { + "epoch": 0.30852, + "grad_norm": 1.8963674351540474, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 30852 + }, + { + "epoch": 0.30853, + "grad_norm": 1.3276136763223572, + "learning_rate": 0.003, + "loss": 4.04, + "step": 30853 + }, + { + "epoch": 0.30854, + "grad_norm": 1.4788165952563956, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 30854 + }, + { + "epoch": 0.30855, + "grad_norm": 1.7517033244363247, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30855 + }, + { + "epoch": 0.30856, + "grad_norm": 1.196346650422106, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 30856 + }, + { + "epoch": 0.30857, + "grad_norm": 1.5342204976244733, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 30857 + }, + { + "epoch": 0.30858, + "grad_norm": 1.3922795514133426, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 30858 + }, + { + "epoch": 0.30859, + "grad_norm": 1.6090984461031606, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 30859 + }, + { + "epoch": 0.3086, + "grad_norm": 1.3866973455532563, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 30860 + }, + { + "epoch": 0.30861, + "grad_norm": 1.8284299469055219, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30861 + }, + { + "epoch": 0.30862, + "grad_norm": 1.2714683678393925, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30862 + }, + { + "epoch": 0.30863, + "grad_norm": 1.5008949426036617, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 30863 + }, + { + "epoch": 0.30864, + "grad_norm": 1.5680050689304332, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30864 + }, + { + "epoch": 0.30865, + "grad_norm": 1.216833453941707, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 30865 + }, + { + "epoch": 0.30866, + "grad_norm": 1.6822594978227439, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 30866 + }, + { + "epoch": 0.30867, + "grad_norm": 1.0901348736994232, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 30867 + }, + { + "epoch": 0.30868, + "grad_norm": 1.686542091727339, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 30868 + }, + { + "epoch": 0.30869, + "grad_norm": 1.5095522884467358, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 30869 + }, + { + "epoch": 0.3087, + "grad_norm": 1.3656979012549073, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 30870 + }, + { + "epoch": 0.30871, + "grad_norm": 1.4755212157700195, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 30871 + }, + { + "epoch": 0.30872, + "grad_norm": 1.4341826678136218, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 30872 + }, + { + "epoch": 0.30873, + "grad_norm": 1.4070687612891295, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 30873 + }, + { + "epoch": 0.30874, + "grad_norm": 1.415912728416533, + "learning_rate": 0.003, + "loss": 4.053, + "step": 30874 + }, + { + "epoch": 0.30875, + "grad_norm": 1.4261533226941634, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 30875 + }, + { + "epoch": 0.30876, + "grad_norm": 1.3738800438272838, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 30876 + }, + { + "epoch": 0.30877, + "grad_norm": 1.4282565501667133, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 30877 + }, + { + "epoch": 0.30878, + "grad_norm": 1.6003097129187536, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 30878 + }, + { + "epoch": 0.30879, + "grad_norm": 1.4111985892597931, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 30879 + }, + { + "epoch": 0.3088, + "grad_norm": 1.3511147194177207, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 30880 + }, + { + "epoch": 0.30881, + "grad_norm": 1.283017113207975, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30881 + }, + { + "epoch": 0.30882, + "grad_norm": 1.5357939120787303, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 30882 + }, + { + "epoch": 0.30883, + "grad_norm": 1.2670373876153969, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 30883 + }, + { + "epoch": 0.30884, + "grad_norm": 1.6303750198498204, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 30884 + }, + { + "epoch": 0.30885, + "grad_norm": 1.4556826227598354, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 30885 + }, + { + "epoch": 0.30886, + "grad_norm": 1.4565603441707662, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30886 + }, + { + "epoch": 0.30887, + "grad_norm": 1.5176851024918656, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 30887 + }, + { + "epoch": 0.30888, + "grad_norm": 1.690380515924541, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 30888 + }, + { + "epoch": 0.30889, + "grad_norm": 1.375058451999136, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 30889 + }, + { + "epoch": 0.3089, + "grad_norm": 1.5906161950854385, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 30890 + }, + { + "epoch": 0.30891, + "grad_norm": 1.4334434847934798, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 30891 + }, + { + "epoch": 0.30892, + "grad_norm": 1.148329594558648, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 30892 + }, + { + "epoch": 0.30893, + "grad_norm": 1.6988659255741965, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 30893 + }, + { + "epoch": 0.30894, + "grad_norm": 1.5321521277581567, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 30894 + }, + { + "epoch": 0.30895, + "grad_norm": 1.6776633112144292, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 30895 + }, + { + "epoch": 0.30896, + "grad_norm": 1.2689508667186233, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 30896 + }, + { + "epoch": 0.30897, + "grad_norm": 1.4208106909728222, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 30897 + }, + { + "epoch": 0.30898, + "grad_norm": 1.7651395011899969, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 30898 + }, + { + "epoch": 0.30899, + "grad_norm": 1.3870755920042988, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30899 + }, + { + "epoch": 0.309, + "grad_norm": 1.5201746394165685, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 30900 + }, + { + "epoch": 0.30901, + "grad_norm": 1.3708611355788005, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 30901 + }, + { + "epoch": 0.30902, + "grad_norm": 1.475544917384829, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 30902 + }, + { + "epoch": 0.30903, + "grad_norm": 1.0933231727227666, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 30903 + }, + { + "epoch": 0.30904, + "grad_norm": 1.4928295733478765, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 30904 + }, + { + "epoch": 0.30905, + "grad_norm": 1.4599456654077512, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30905 + }, + { + "epoch": 0.30906, + "grad_norm": 1.3722224103387752, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 30906 + }, + { + "epoch": 0.30907, + "grad_norm": 1.6945305816966862, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 30907 + }, + { + "epoch": 0.30908, + "grad_norm": 1.3365855692203805, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 30908 + }, + { + "epoch": 0.30909, + "grad_norm": 1.229442720649027, + "learning_rate": 0.003, + "loss": 4.014, + "step": 30909 + }, + { + "epoch": 0.3091, + "grad_norm": 1.786851837529431, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 30910 + }, + { + "epoch": 0.30911, + "grad_norm": 1.1313060894055815, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 30911 + }, + { + "epoch": 0.30912, + "grad_norm": 1.5524174316141373, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 30912 + }, + { + "epoch": 0.30913, + "grad_norm": 1.300662566685414, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 30913 + }, + { + "epoch": 0.30914, + "grad_norm": 1.328113940140739, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30914 + }, + { + "epoch": 0.30915, + "grad_norm": 1.5059617080765313, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 30915 + }, + { + "epoch": 0.30916, + "grad_norm": 1.342649807162857, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 30916 + }, + { + "epoch": 0.30917, + "grad_norm": 1.757183008834262, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 30917 + }, + { + "epoch": 0.30918, + "grad_norm": 1.7390543167248038, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 30918 + }, + { + "epoch": 0.30919, + "grad_norm": 1.4655806186088784, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 30919 + }, + { + "epoch": 0.3092, + "grad_norm": 1.5724556515488073, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 30920 + }, + { + "epoch": 0.30921, + "grad_norm": 1.5549494527768835, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 30921 + }, + { + "epoch": 0.30922, + "grad_norm": 1.6070114697128395, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 30922 + }, + { + "epoch": 0.30923, + "grad_norm": 1.4292698134597144, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 30923 + }, + { + "epoch": 0.30924, + "grad_norm": 1.3956282856741435, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 30924 + }, + { + "epoch": 0.30925, + "grad_norm": 1.6727658887656014, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 30925 + }, + { + "epoch": 0.30926, + "grad_norm": 1.2515297714085691, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 30926 + }, + { + "epoch": 0.30927, + "grad_norm": 2.168734739948335, + "learning_rate": 0.003, + "loss": 4.073, + "step": 30927 + }, + { + "epoch": 0.30928, + "grad_norm": 1.1656994982939872, + "learning_rate": 0.003, + "loss": 4.051, + "step": 30928 + }, + { + "epoch": 0.30929, + "grad_norm": 1.7976713634145198, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30929 + }, + { + "epoch": 0.3093, + "grad_norm": 1.5549826024285207, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 30930 + }, + { + "epoch": 0.30931, + "grad_norm": 1.591439271606837, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 30931 + }, + { + "epoch": 0.30932, + "grad_norm": 1.2619056163962719, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 30932 + }, + { + "epoch": 0.30933, + "grad_norm": 1.358900865224621, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 30933 + }, + { + "epoch": 0.30934, + "grad_norm": 1.4752153814241176, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 30934 + }, + { + "epoch": 0.30935, + "grad_norm": 1.216485854480269, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 30935 + }, + { + "epoch": 0.30936, + "grad_norm": 2.0595779964589673, + "learning_rate": 0.003, + "loss": 4.088, + "step": 30936 + }, + { + "epoch": 0.30937, + "grad_norm": 1.2875871289106193, + "learning_rate": 0.003, + "loss": 4.085, + "step": 30937 + }, + { + "epoch": 0.30938, + "grad_norm": 1.6821869746725262, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 30938 + }, + { + "epoch": 0.30939, + "grad_norm": 1.2175154211700372, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 30939 + }, + { + "epoch": 0.3094, + "grad_norm": 1.6400751476494864, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 30940 + }, + { + "epoch": 0.30941, + "grad_norm": 1.6695649826396735, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 30941 + }, + { + "epoch": 0.30942, + "grad_norm": 1.7361788763471164, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30942 + }, + { + "epoch": 0.30943, + "grad_norm": 1.6187931132957654, + "learning_rate": 0.003, + "loss": 4.054, + "step": 30943 + }, + { + "epoch": 0.30944, + "grad_norm": 1.3292904821086045, + "learning_rate": 0.003, + "loss": 4.084, + "step": 30944 + }, + { + "epoch": 0.30945, + "grad_norm": 1.4318510817574908, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 30945 + }, + { + "epoch": 0.30946, + "grad_norm": 1.36213688611055, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 30946 + }, + { + "epoch": 0.30947, + "grad_norm": 1.48777205067439, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 30947 + }, + { + "epoch": 0.30948, + "grad_norm": 1.7467072870662754, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 30948 + }, + { + "epoch": 0.30949, + "grad_norm": 1.6188011628673933, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 30949 + }, + { + "epoch": 0.3095, + "grad_norm": 1.6982423625863912, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 30950 + }, + { + "epoch": 0.30951, + "grad_norm": 1.4209337020046862, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 30951 + }, + { + "epoch": 0.30952, + "grad_norm": 1.3867790812305194, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 30952 + }, + { + "epoch": 0.30953, + "grad_norm": 1.5985219011320615, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 30953 + }, + { + "epoch": 0.30954, + "grad_norm": 1.2692177264724114, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 30954 + }, + { + "epoch": 0.30955, + "grad_norm": 1.3985894792852755, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 30955 + }, + { + "epoch": 0.30956, + "grad_norm": 1.3187609286903643, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30956 + }, + { + "epoch": 0.30957, + "grad_norm": 1.2830976847047264, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 30957 + }, + { + "epoch": 0.30958, + "grad_norm": 1.4970873485067622, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 30958 + }, + { + "epoch": 0.30959, + "grad_norm": 1.602865566780008, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 30959 + }, + { + "epoch": 0.3096, + "grad_norm": 1.7614292874055537, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 30960 + }, + { + "epoch": 0.30961, + "grad_norm": 1.4903100880436426, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 30961 + }, + { + "epoch": 0.30962, + "grad_norm": 1.6982693181614201, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 30962 + }, + { + "epoch": 0.30963, + "grad_norm": 1.5038469619004609, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 30963 + }, + { + "epoch": 0.30964, + "grad_norm": 1.5624799327060466, + "learning_rate": 0.003, + "loss": 4.034, + "step": 30964 + }, + { + "epoch": 0.30965, + "grad_norm": 1.3355303739891211, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 30965 + }, + { + "epoch": 0.30966, + "grad_norm": 1.491050135518759, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 30966 + }, + { + "epoch": 0.30967, + "grad_norm": 1.4480598660909294, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 30967 + }, + { + "epoch": 0.30968, + "grad_norm": 1.5005180634955186, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 30968 + }, + { + "epoch": 0.30969, + "grad_norm": 1.3261503053249803, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 30969 + }, + { + "epoch": 0.3097, + "grad_norm": 1.5051438595362323, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 30970 + }, + { + "epoch": 0.30971, + "grad_norm": 1.2888507258178543, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 30971 + }, + { + "epoch": 0.30972, + "grad_norm": 1.4918108220017112, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 30972 + }, + { + "epoch": 0.30973, + "grad_norm": 1.3754132642239971, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 30973 + }, + { + "epoch": 0.30974, + "grad_norm": 1.3357848216557013, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 30974 + }, + { + "epoch": 0.30975, + "grad_norm": 1.5698742193796555, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30975 + }, + { + "epoch": 0.30976, + "grad_norm": 1.2623369226579972, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 30976 + }, + { + "epoch": 0.30977, + "grad_norm": 1.5198906980912037, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 30977 + }, + { + "epoch": 0.30978, + "grad_norm": 1.1442628840224238, + "learning_rate": 0.003, + "loss": 4.022, + "step": 30978 + }, + { + "epoch": 0.30979, + "grad_norm": 2.0807054428820186, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 30979 + }, + { + "epoch": 0.3098, + "grad_norm": 1.1903666683588403, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 30980 + }, + { + "epoch": 0.30981, + "grad_norm": 1.4965632952110228, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 30981 + }, + { + "epoch": 0.30982, + "grad_norm": 1.3176804103583069, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 30982 + }, + { + "epoch": 0.30983, + "grad_norm": 1.2395654332129502, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 30983 + }, + { + "epoch": 0.30984, + "grad_norm": 1.529564732446051, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 30984 + }, + { + "epoch": 0.30985, + "grad_norm": 1.739477183850989, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 30985 + }, + { + "epoch": 0.30986, + "grad_norm": 1.756437413778274, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 30986 + }, + { + "epoch": 0.30987, + "grad_norm": 1.4571045674698384, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 30987 + }, + { + "epoch": 0.30988, + "grad_norm": 1.4630077820877678, + "learning_rate": 0.003, + "loss": 4.071, + "step": 30988 + }, + { + "epoch": 0.30989, + "grad_norm": 1.5862066805653863, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 30989 + }, + { + "epoch": 0.3099, + "grad_norm": 1.4338099500988677, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 30990 + }, + { + "epoch": 0.30991, + "grad_norm": 1.5521233080043495, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 30991 + }, + { + "epoch": 0.30992, + "grad_norm": 1.509753615358063, + "learning_rate": 0.003, + "loss": 4.048, + "step": 30992 + }, + { + "epoch": 0.30993, + "grad_norm": 1.562160444864314, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 30993 + }, + { + "epoch": 0.30994, + "grad_norm": 1.6623374737563674, + "learning_rate": 0.003, + "loss": 4.06, + "step": 30994 + }, + { + "epoch": 0.30995, + "grad_norm": 1.3059965876367856, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 30995 + }, + { + "epoch": 0.30996, + "grad_norm": 1.4899047949770898, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 30996 + }, + { + "epoch": 0.30997, + "grad_norm": 1.2368029550821864, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 30997 + }, + { + "epoch": 0.30998, + "grad_norm": 1.4156752013191904, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 30998 + }, + { + "epoch": 0.30999, + "grad_norm": 1.3573917627679748, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 30999 + }, + { + "epoch": 0.31, + "grad_norm": 1.8179902904784533, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 31000 + }, + { + "epoch": 0.31001, + "grad_norm": 1.5562660949389528, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 31001 + }, + { + "epoch": 0.31002, + "grad_norm": 1.5158493357231626, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 31002 + }, + { + "epoch": 0.31003, + "grad_norm": 1.5669957513068382, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 31003 + }, + { + "epoch": 0.31004, + "grad_norm": 1.3915327259907357, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31004 + }, + { + "epoch": 0.31005, + "grad_norm": 1.4221423325942166, + "learning_rate": 0.003, + "loss": 4.043, + "step": 31005 + }, + { + "epoch": 0.31006, + "grad_norm": 1.1883709532743212, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 31006 + }, + { + "epoch": 0.31007, + "grad_norm": 1.7669895058102099, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31007 + }, + { + "epoch": 0.31008, + "grad_norm": 1.0113605550932006, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 31008 + }, + { + "epoch": 0.31009, + "grad_norm": 1.6532523918026254, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 31009 + }, + { + "epoch": 0.3101, + "grad_norm": 1.2874396238292791, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31010 + }, + { + "epoch": 0.31011, + "grad_norm": 1.6570380588630245, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31011 + }, + { + "epoch": 0.31012, + "grad_norm": 1.3628670221752714, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 31012 + }, + { + "epoch": 0.31013, + "grad_norm": 1.6649531591252864, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 31013 + }, + { + "epoch": 0.31014, + "grad_norm": 2.1308584847148313, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 31014 + }, + { + "epoch": 0.31015, + "grad_norm": 1.545420950813347, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 31015 + }, + { + "epoch": 0.31016, + "grad_norm": 1.549824906748769, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 31016 + }, + { + "epoch": 0.31017, + "grad_norm": 1.2909905654945304, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 31017 + }, + { + "epoch": 0.31018, + "grad_norm": 1.5604194177485624, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 31018 + }, + { + "epoch": 0.31019, + "grad_norm": 1.3341150218534534, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 31019 + }, + { + "epoch": 0.3102, + "grad_norm": 1.49763080116329, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 31020 + }, + { + "epoch": 0.31021, + "grad_norm": 1.5443396827583982, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31021 + }, + { + "epoch": 0.31022, + "grad_norm": 1.6086653254423036, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31022 + }, + { + "epoch": 0.31023, + "grad_norm": 1.5786900564943003, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 31023 + }, + { + "epoch": 0.31024, + "grad_norm": 1.8964225464642583, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31024 + }, + { + "epoch": 0.31025, + "grad_norm": 1.6428073048721636, + "learning_rate": 0.003, + "loss": 4.043, + "step": 31025 + }, + { + "epoch": 0.31026, + "grad_norm": 1.4105164988562653, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31026 + }, + { + "epoch": 0.31027, + "grad_norm": 1.2315550376383615, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 31027 + }, + { + "epoch": 0.31028, + "grad_norm": 1.3162748154982442, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31028 + }, + { + "epoch": 0.31029, + "grad_norm": 1.7317604572808611, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 31029 + }, + { + "epoch": 0.3103, + "grad_norm": 1.2808536456688142, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 31030 + }, + { + "epoch": 0.31031, + "grad_norm": 1.4901932264722564, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 31031 + }, + { + "epoch": 0.31032, + "grad_norm": 1.381881361584601, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 31032 + }, + { + "epoch": 0.31033, + "grad_norm": 1.580224534060841, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 31033 + }, + { + "epoch": 0.31034, + "grad_norm": 1.4136266599930978, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 31034 + }, + { + "epoch": 0.31035, + "grad_norm": 1.6715115243686276, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 31035 + }, + { + "epoch": 0.31036, + "grad_norm": 1.468527313231214, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 31036 + }, + { + "epoch": 0.31037, + "grad_norm": 1.6494903952747115, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 31037 + }, + { + "epoch": 0.31038, + "grad_norm": 1.411818762562179, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 31038 + }, + { + "epoch": 0.31039, + "grad_norm": 1.5887162253736737, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 31039 + }, + { + "epoch": 0.3104, + "grad_norm": 1.25293660128397, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 31040 + }, + { + "epoch": 0.31041, + "grad_norm": 1.6218729108237548, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 31041 + }, + { + "epoch": 0.31042, + "grad_norm": 1.823872445048236, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31042 + }, + { + "epoch": 0.31043, + "grad_norm": 1.2569963690156976, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31043 + }, + { + "epoch": 0.31044, + "grad_norm": 1.6976883116183534, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 31044 + }, + { + "epoch": 0.31045, + "grad_norm": 1.575664433432658, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 31045 + }, + { + "epoch": 0.31046, + "grad_norm": 1.0487359289517952, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 31046 + }, + { + "epoch": 0.31047, + "grad_norm": 1.7539044501352787, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31047 + }, + { + "epoch": 0.31048, + "grad_norm": 1.2015996203100827, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31048 + }, + { + "epoch": 0.31049, + "grad_norm": 1.4728685973974434, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31049 + }, + { + "epoch": 0.3105, + "grad_norm": 1.3813660308812516, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 31050 + }, + { + "epoch": 0.31051, + "grad_norm": 1.4966025143666035, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 31051 + }, + { + "epoch": 0.31052, + "grad_norm": 1.3900643884434114, + "learning_rate": 0.003, + "loss": 4.068, + "step": 31052 + }, + { + "epoch": 0.31053, + "grad_norm": 1.6763402235595415, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 31053 + }, + { + "epoch": 0.31054, + "grad_norm": 1.3650715645340257, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 31054 + }, + { + "epoch": 0.31055, + "grad_norm": 1.5674354258924714, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 31055 + }, + { + "epoch": 0.31056, + "grad_norm": 1.2422332296703866, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 31056 + }, + { + "epoch": 0.31057, + "grad_norm": 1.5765258256956576, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 31057 + }, + { + "epoch": 0.31058, + "grad_norm": 1.2520503160220207, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31058 + }, + { + "epoch": 0.31059, + "grad_norm": 1.7176546234785226, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 31059 + }, + { + "epoch": 0.3106, + "grad_norm": 1.157638927286062, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 31060 + }, + { + "epoch": 0.31061, + "grad_norm": 1.571807533160437, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 31061 + }, + { + "epoch": 0.31062, + "grad_norm": 1.306976849964148, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 31062 + }, + { + "epoch": 0.31063, + "grad_norm": 1.3548078013015519, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 31063 + }, + { + "epoch": 0.31064, + "grad_norm": 1.650746066189786, + "learning_rate": 0.003, + "loss": 4.052, + "step": 31064 + }, + { + "epoch": 0.31065, + "grad_norm": 1.1676545335712059, + "learning_rate": 0.003, + "loss": 4.057, + "step": 31065 + }, + { + "epoch": 0.31066, + "grad_norm": 1.492797531640834, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 31066 + }, + { + "epoch": 0.31067, + "grad_norm": 1.4424697987084718, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 31067 + }, + { + "epoch": 0.31068, + "grad_norm": 1.5090350251139997, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 31068 + }, + { + "epoch": 0.31069, + "grad_norm": 1.484062258596378, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31069 + }, + { + "epoch": 0.3107, + "grad_norm": 1.3607267779060335, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 31070 + }, + { + "epoch": 0.31071, + "grad_norm": 1.5576417261788724, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 31071 + }, + { + "epoch": 0.31072, + "grad_norm": 1.161615590647574, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 31072 + }, + { + "epoch": 0.31073, + "grad_norm": 1.5349337260513551, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 31073 + }, + { + "epoch": 0.31074, + "grad_norm": 1.3611358685817125, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31074 + }, + { + "epoch": 0.31075, + "grad_norm": 1.2303078461007524, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31075 + }, + { + "epoch": 0.31076, + "grad_norm": 1.3574700998888098, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 31076 + }, + { + "epoch": 0.31077, + "grad_norm": 1.598501749045473, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 31077 + }, + { + "epoch": 0.31078, + "grad_norm": 1.2408016808794695, + "learning_rate": 0.003, + "loss": 4.049, + "step": 31078 + }, + { + "epoch": 0.31079, + "grad_norm": 1.4779059592510806, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 31079 + }, + { + "epoch": 0.3108, + "grad_norm": 1.9191528495902985, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 31080 + }, + { + "epoch": 0.31081, + "grad_norm": 1.7582704090463386, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 31081 + }, + { + "epoch": 0.31082, + "grad_norm": 1.570051712127788, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31082 + }, + { + "epoch": 0.31083, + "grad_norm": 1.3409622606049083, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 31083 + }, + { + "epoch": 0.31084, + "grad_norm": 1.9030503618379582, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31084 + }, + { + "epoch": 0.31085, + "grad_norm": 1.0509681312914134, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 31085 + }, + { + "epoch": 0.31086, + "grad_norm": 1.6391659791365025, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 31086 + }, + { + "epoch": 0.31087, + "grad_norm": 1.2650814904325618, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 31087 + }, + { + "epoch": 0.31088, + "grad_norm": 1.313545706836239, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 31088 + }, + { + "epoch": 0.31089, + "grad_norm": 1.324787369300705, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31089 + }, + { + "epoch": 0.3109, + "grad_norm": 1.6406632130354906, + "learning_rate": 0.003, + "loss": 4.067, + "step": 31090 + }, + { + "epoch": 0.31091, + "grad_norm": 1.6231418928218446, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 31091 + }, + { + "epoch": 0.31092, + "grad_norm": 1.2868982265065942, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31092 + }, + { + "epoch": 0.31093, + "grad_norm": 1.2759645601362861, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 31093 + }, + { + "epoch": 0.31094, + "grad_norm": 1.6165202672332681, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31094 + }, + { + "epoch": 0.31095, + "grad_norm": 1.3041446999756607, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31095 + }, + { + "epoch": 0.31096, + "grad_norm": 1.5939777730377334, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 31096 + }, + { + "epoch": 0.31097, + "grad_norm": 1.588030321249607, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 31097 + }, + { + "epoch": 0.31098, + "grad_norm": 1.6145099481171998, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 31098 + }, + { + "epoch": 0.31099, + "grad_norm": 1.2103742465264213, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 31099 + }, + { + "epoch": 0.311, + "grad_norm": 1.6620699753595622, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31100 + }, + { + "epoch": 0.31101, + "grad_norm": 1.114375234143686, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31101 + }, + { + "epoch": 0.31102, + "grad_norm": 1.7989827963007465, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31102 + }, + { + "epoch": 0.31103, + "grad_norm": 1.4072891645142531, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 31103 + }, + { + "epoch": 0.31104, + "grad_norm": 1.5431567908002042, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 31104 + }, + { + "epoch": 0.31105, + "grad_norm": 1.5057230244363327, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 31105 + }, + { + "epoch": 0.31106, + "grad_norm": 1.3248370951477848, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 31106 + }, + { + "epoch": 0.31107, + "grad_norm": 1.2726012155315403, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 31107 + }, + { + "epoch": 0.31108, + "grad_norm": 1.4001742800483803, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 31108 + }, + { + "epoch": 0.31109, + "grad_norm": 1.5801123034974542, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31109 + }, + { + "epoch": 0.3111, + "grad_norm": 1.211988168578211, + "learning_rate": 0.003, + "loss": 4.024, + "step": 31110 + }, + { + "epoch": 0.31111, + "grad_norm": 1.5840890951185258, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31111 + }, + { + "epoch": 0.31112, + "grad_norm": 1.3620278233231444, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 31112 + }, + { + "epoch": 0.31113, + "grad_norm": 1.6701507902725377, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 31113 + }, + { + "epoch": 0.31114, + "grad_norm": 1.3216935689226135, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 31114 + }, + { + "epoch": 0.31115, + "grad_norm": 1.6407644837104627, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 31115 + }, + { + "epoch": 0.31116, + "grad_norm": 1.1383750341207168, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31116 + }, + { + "epoch": 0.31117, + "grad_norm": 1.55671970826668, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31117 + }, + { + "epoch": 0.31118, + "grad_norm": 1.3272412033337748, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 31118 + }, + { + "epoch": 0.31119, + "grad_norm": 1.6140744871029926, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 31119 + }, + { + "epoch": 0.3112, + "grad_norm": 1.5018400644862182, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31120 + }, + { + "epoch": 0.31121, + "grad_norm": 1.499059963724105, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 31121 + }, + { + "epoch": 0.31122, + "grad_norm": 1.159943661685609, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 31122 + }, + { + "epoch": 0.31123, + "grad_norm": 1.4934426762056423, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 31123 + }, + { + "epoch": 0.31124, + "grad_norm": 1.7067894928736946, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31124 + }, + { + "epoch": 0.31125, + "grad_norm": 1.3245496484663455, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31125 + }, + { + "epoch": 0.31126, + "grad_norm": 1.4607502543344986, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 31126 + }, + { + "epoch": 0.31127, + "grad_norm": 1.3448582024741245, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 31127 + }, + { + "epoch": 0.31128, + "grad_norm": 1.4720651147333101, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 31128 + }, + { + "epoch": 0.31129, + "grad_norm": 1.2343888145964863, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31129 + }, + { + "epoch": 0.3113, + "grad_norm": 1.399977603654607, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31130 + }, + { + "epoch": 0.31131, + "grad_norm": 1.3347117434477584, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 31131 + }, + { + "epoch": 0.31132, + "grad_norm": 1.3864265381313294, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 31132 + }, + { + "epoch": 0.31133, + "grad_norm": 1.8135802578185316, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 31133 + }, + { + "epoch": 0.31134, + "grad_norm": 1.4063734070018408, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 31134 + }, + { + "epoch": 0.31135, + "grad_norm": 1.4432916535894518, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 31135 + }, + { + "epoch": 0.31136, + "grad_norm": 1.312800913139961, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 31136 + }, + { + "epoch": 0.31137, + "grad_norm": 1.811303612430373, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31137 + }, + { + "epoch": 0.31138, + "grad_norm": 1.2160000306560297, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31138 + }, + { + "epoch": 0.31139, + "grad_norm": 1.5911541792231023, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31139 + }, + { + "epoch": 0.3114, + "grad_norm": 1.6243513790928148, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 31140 + }, + { + "epoch": 0.31141, + "grad_norm": 1.3938107447108028, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 31141 + }, + { + "epoch": 0.31142, + "grad_norm": 1.4615398513551912, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31142 + }, + { + "epoch": 0.31143, + "grad_norm": 1.255863194836435, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 31143 + }, + { + "epoch": 0.31144, + "grad_norm": 1.449336768123671, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31144 + }, + { + "epoch": 0.31145, + "grad_norm": 1.365567674188652, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31145 + }, + { + "epoch": 0.31146, + "grad_norm": 1.609054121601752, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31146 + }, + { + "epoch": 0.31147, + "grad_norm": 1.5379061379096204, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31147 + }, + { + "epoch": 0.31148, + "grad_norm": 1.6108844671245584, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31148 + }, + { + "epoch": 0.31149, + "grad_norm": 1.4081396241206086, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31149 + }, + { + "epoch": 0.3115, + "grad_norm": 1.5498532636087647, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 31150 + }, + { + "epoch": 0.31151, + "grad_norm": 1.125684032747048, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 31151 + }, + { + "epoch": 0.31152, + "grad_norm": 1.6136967187873672, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31152 + }, + { + "epoch": 0.31153, + "grad_norm": 1.3740201616703178, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 31153 + }, + { + "epoch": 0.31154, + "grad_norm": 1.247670061164326, + "learning_rate": 0.003, + "loss": 4.025, + "step": 31154 + }, + { + "epoch": 0.31155, + "grad_norm": 1.560649082055222, + "learning_rate": 0.003, + "loss": 4.056, + "step": 31155 + }, + { + "epoch": 0.31156, + "grad_norm": 1.383250725895758, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 31156 + }, + { + "epoch": 0.31157, + "grad_norm": 1.4815776225669468, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 31157 + }, + { + "epoch": 0.31158, + "grad_norm": 1.599832768232451, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 31158 + }, + { + "epoch": 0.31159, + "grad_norm": 1.5005720225521546, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31159 + }, + { + "epoch": 0.3116, + "grad_norm": 1.3761165897440453, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31160 + }, + { + "epoch": 0.31161, + "grad_norm": 1.4229948179532461, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 31161 + }, + { + "epoch": 0.31162, + "grad_norm": 1.5036271479443861, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 31162 + }, + { + "epoch": 0.31163, + "grad_norm": 1.3527273540496882, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31163 + }, + { + "epoch": 0.31164, + "grad_norm": 1.6528131187676574, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 31164 + }, + { + "epoch": 0.31165, + "grad_norm": 1.3437682610888675, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 31165 + }, + { + "epoch": 0.31166, + "grad_norm": 1.6067879279082782, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31166 + }, + { + "epoch": 0.31167, + "grad_norm": 1.4418312964192113, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 31167 + }, + { + "epoch": 0.31168, + "grad_norm": 1.649816510914663, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 31168 + }, + { + "epoch": 0.31169, + "grad_norm": 1.6169160055270695, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 31169 + }, + { + "epoch": 0.3117, + "grad_norm": 1.1078147931926707, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 31170 + }, + { + "epoch": 0.31171, + "grad_norm": 1.5576416614472974, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 31171 + }, + { + "epoch": 0.31172, + "grad_norm": 1.1429472547954875, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 31172 + }, + { + "epoch": 0.31173, + "grad_norm": 1.5793309554973962, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 31173 + }, + { + "epoch": 0.31174, + "grad_norm": 1.342231568399624, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 31174 + }, + { + "epoch": 0.31175, + "grad_norm": 1.4663662230490264, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 31175 + }, + { + "epoch": 0.31176, + "grad_norm": 1.4089585806975227, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31176 + }, + { + "epoch": 0.31177, + "grad_norm": 1.292170820771109, + "learning_rate": 0.003, + "loss": 4.057, + "step": 31177 + }, + { + "epoch": 0.31178, + "grad_norm": 1.4316289048022632, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 31178 + }, + { + "epoch": 0.31179, + "grad_norm": 1.4027202480816874, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 31179 + }, + { + "epoch": 0.3118, + "grad_norm": 1.597284854707112, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 31180 + }, + { + "epoch": 0.31181, + "grad_norm": 1.814323885562034, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 31181 + }, + { + "epoch": 0.31182, + "grad_norm": 1.6244560083255053, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 31182 + }, + { + "epoch": 0.31183, + "grad_norm": 1.30192097144574, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 31183 + }, + { + "epoch": 0.31184, + "grad_norm": 1.4956677314296367, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31184 + }, + { + "epoch": 0.31185, + "grad_norm": 1.5978901784581436, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 31185 + }, + { + "epoch": 0.31186, + "grad_norm": 1.4789472649295732, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 31186 + }, + { + "epoch": 0.31187, + "grad_norm": 1.5058598869702369, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 31187 + }, + { + "epoch": 0.31188, + "grad_norm": 1.4470236987853975, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31188 + }, + { + "epoch": 0.31189, + "grad_norm": 1.4141187916939728, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 31189 + }, + { + "epoch": 0.3119, + "grad_norm": 1.5117267730432469, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31190 + }, + { + "epoch": 0.31191, + "grad_norm": 1.4224482877174305, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31191 + }, + { + "epoch": 0.31192, + "grad_norm": 1.4549915996738303, + "learning_rate": 0.003, + "loss": 4.04, + "step": 31192 + }, + { + "epoch": 0.31193, + "grad_norm": 1.3199030738722517, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 31193 + }, + { + "epoch": 0.31194, + "grad_norm": 1.286831431777756, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 31194 + }, + { + "epoch": 0.31195, + "grad_norm": 1.4006697937452965, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 31195 + }, + { + "epoch": 0.31196, + "grad_norm": 1.2034693127749951, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31196 + }, + { + "epoch": 0.31197, + "grad_norm": 1.5332894901597818, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 31197 + }, + { + "epoch": 0.31198, + "grad_norm": 1.4803987697648626, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 31198 + }, + { + "epoch": 0.31199, + "grad_norm": 1.5432620206952867, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31199 + }, + { + "epoch": 0.312, + "grad_norm": 1.4167409837323677, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31200 + }, + { + "epoch": 0.31201, + "grad_norm": 1.5881161847093892, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 31201 + }, + { + "epoch": 0.31202, + "grad_norm": 1.3196658001555521, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31202 + }, + { + "epoch": 0.31203, + "grad_norm": 2.451497109818437, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 31203 + }, + { + "epoch": 0.31204, + "grad_norm": 1.0405773201739483, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 31204 + }, + { + "epoch": 0.31205, + "grad_norm": 1.5989781727193308, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 31205 + }, + { + "epoch": 0.31206, + "grad_norm": 1.2500443774970478, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 31206 + }, + { + "epoch": 0.31207, + "grad_norm": 1.7457871160068759, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 31207 + }, + { + "epoch": 0.31208, + "grad_norm": 1.293415033820103, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 31208 + }, + { + "epoch": 0.31209, + "grad_norm": 1.6633779707143201, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 31209 + }, + { + "epoch": 0.3121, + "grad_norm": 1.4031116587141375, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31210 + }, + { + "epoch": 0.31211, + "grad_norm": 1.658110180953978, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31211 + }, + { + "epoch": 0.31212, + "grad_norm": 1.8416010764207875, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31212 + }, + { + "epoch": 0.31213, + "grad_norm": 1.6752989149615058, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 31213 + }, + { + "epoch": 0.31214, + "grad_norm": 1.63218571958245, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 31214 + }, + { + "epoch": 0.31215, + "grad_norm": 1.2743775173822505, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 31215 + }, + { + "epoch": 0.31216, + "grad_norm": 1.6694617802560348, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31216 + }, + { + "epoch": 0.31217, + "grad_norm": 1.479727060409624, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31217 + }, + { + "epoch": 0.31218, + "grad_norm": 1.375805383116969, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 31218 + }, + { + "epoch": 0.31219, + "grad_norm": 1.4253766605091527, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 31219 + }, + { + "epoch": 0.3122, + "grad_norm": 1.4390486360824644, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 31220 + }, + { + "epoch": 0.31221, + "grad_norm": 1.6488402056286846, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 31221 + }, + { + "epoch": 0.31222, + "grad_norm": 1.2995110016775049, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 31222 + }, + { + "epoch": 0.31223, + "grad_norm": 1.3614118293398794, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 31223 + }, + { + "epoch": 0.31224, + "grad_norm": 1.2053981537657206, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 31224 + }, + { + "epoch": 0.31225, + "grad_norm": 1.7241079713944896, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 31225 + }, + { + "epoch": 0.31226, + "grad_norm": 1.109353476, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31226 + }, + { + "epoch": 0.31227, + "grad_norm": 1.6553203840178954, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 31227 + }, + { + "epoch": 0.31228, + "grad_norm": 1.191118500979816, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 31228 + }, + { + "epoch": 0.31229, + "grad_norm": 1.44726852679446, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 31229 + }, + { + "epoch": 0.3123, + "grad_norm": 1.2658330200524488, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 31230 + }, + { + "epoch": 0.31231, + "grad_norm": 1.6714041041655388, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 31231 + }, + { + "epoch": 0.31232, + "grad_norm": 1.1746694453981332, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 31232 + }, + { + "epoch": 0.31233, + "grad_norm": 1.5152995499328454, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 31233 + }, + { + "epoch": 0.31234, + "grad_norm": 1.1888247407314092, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 31234 + }, + { + "epoch": 0.31235, + "grad_norm": 1.3479604376107146, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 31235 + }, + { + "epoch": 0.31236, + "grad_norm": 1.3906623834588907, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31236 + }, + { + "epoch": 0.31237, + "grad_norm": 1.3389112608075475, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31237 + }, + { + "epoch": 0.31238, + "grad_norm": 1.4920017425376473, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 31238 + }, + { + "epoch": 0.31239, + "grad_norm": 1.4452980602512298, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 31239 + }, + { + "epoch": 0.3124, + "grad_norm": 1.8617717151942834, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 31240 + }, + { + "epoch": 0.31241, + "grad_norm": 1.374411020715944, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 31241 + }, + { + "epoch": 0.31242, + "grad_norm": 1.895831224501949, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 31242 + }, + { + "epoch": 0.31243, + "grad_norm": 1.5372139408863101, + "learning_rate": 0.003, + "loss": 4.038, + "step": 31243 + }, + { + "epoch": 0.31244, + "grad_norm": 1.2670236699489021, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31244 + }, + { + "epoch": 0.31245, + "grad_norm": 1.4392214449158565, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 31245 + }, + { + "epoch": 0.31246, + "grad_norm": 1.4443018205164504, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31246 + }, + { + "epoch": 0.31247, + "grad_norm": 1.4663551667170494, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 31247 + }, + { + "epoch": 0.31248, + "grad_norm": 1.7112324236460537, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 31248 + }, + { + "epoch": 0.31249, + "grad_norm": 1.2653574749155594, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31249 + }, + { + "epoch": 0.3125, + "grad_norm": 1.4708738277411237, + "learning_rate": 0.003, + "loss": 4.051, + "step": 31250 + }, + { + "epoch": 0.31251, + "grad_norm": 1.4359716641081164, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 31251 + }, + { + "epoch": 0.31252, + "grad_norm": 1.507489911844483, + "learning_rate": 0.003, + "loss": 4.051, + "step": 31252 + }, + { + "epoch": 0.31253, + "grad_norm": 1.2257096747090428, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 31253 + }, + { + "epoch": 0.31254, + "grad_norm": 1.6609205578563202, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31254 + }, + { + "epoch": 0.31255, + "grad_norm": 1.27642757508778, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31255 + }, + { + "epoch": 0.31256, + "grad_norm": 1.439724998638068, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 31256 + }, + { + "epoch": 0.31257, + "grad_norm": 1.447915414221077, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 31257 + }, + { + "epoch": 0.31258, + "grad_norm": 1.3323666135402596, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 31258 + }, + { + "epoch": 0.31259, + "grad_norm": 1.3470939463432436, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 31259 + }, + { + "epoch": 0.3126, + "grad_norm": 1.6324710360799992, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 31260 + }, + { + "epoch": 0.31261, + "grad_norm": 1.766618291175467, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 31261 + }, + { + "epoch": 0.31262, + "grad_norm": 1.463823494718476, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 31262 + }, + { + "epoch": 0.31263, + "grad_norm": 1.2983533544191597, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31263 + }, + { + "epoch": 0.31264, + "grad_norm": 1.4937533822778386, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31264 + }, + { + "epoch": 0.31265, + "grad_norm": 1.447426308251477, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31265 + }, + { + "epoch": 0.31266, + "grad_norm": 1.349729741171138, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 31266 + }, + { + "epoch": 0.31267, + "grad_norm": 1.6356557955087823, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31267 + }, + { + "epoch": 0.31268, + "grad_norm": 1.2912454744393085, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 31268 + }, + { + "epoch": 0.31269, + "grad_norm": 1.597531715701171, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 31269 + }, + { + "epoch": 0.3127, + "grad_norm": 1.336072867028712, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 31270 + }, + { + "epoch": 0.31271, + "grad_norm": 1.4441089470271609, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 31271 + }, + { + "epoch": 0.31272, + "grad_norm": 1.4146986285199088, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31272 + }, + { + "epoch": 0.31273, + "grad_norm": 1.7473633365045957, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 31273 + }, + { + "epoch": 0.31274, + "grad_norm": 1.3236664930126723, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 31274 + }, + { + "epoch": 0.31275, + "grad_norm": 1.5892919231728873, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 31275 + }, + { + "epoch": 0.31276, + "grad_norm": 1.62457773363959, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 31276 + }, + { + "epoch": 0.31277, + "grad_norm": 1.3715459692609555, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 31277 + }, + { + "epoch": 0.31278, + "grad_norm": 1.5355119121284362, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31278 + }, + { + "epoch": 0.31279, + "grad_norm": 1.3490998403438528, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 31279 + }, + { + "epoch": 0.3128, + "grad_norm": 1.2659664027663482, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 31280 + }, + { + "epoch": 0.31281, + "grad_norm": 1.5815380124480738, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31281 + }, + { + "epoch": 0.31282, + "grad_norm": 1.3783581235105917, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 31282 + }, + { + "epoch": 0.31283, + "grad_norm": 1.407551568981695, + "learning_rate": 0.003, + "loss": 4.017, + "step": 31283 + }, + { + "epoch": 0.31284, + "grad_norm": 1.7112540651203787, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 31284 + }, + { + "epoch": 0.31285, + "grad_norm": 1.4478246863577635, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 31285 + }, + { + "epoch": 0.31286, + "grad_norm": 1.5602808414002935, + "learning_rate": 0.003, + "loss": 4.021, + "step": 31286 + }, + { + "epoch": 0.31287, + "grad_norm": 1.2502389220660615, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 31287 + }, + { + "epoch": 0.31288, + "grad_norm": 1.7777163283241362, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 31288 + }, + { + "epoch": 0.31289, + "grad_norm": 1.3094689123035945, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 31289 + }, + { + "epoch": 0.3129, + "grad_norm": 1.4783225165736311, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 31290 + }, + { + "epoch": 0.31291, + "grad_norm": 1.417946283658721, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31291 + }, + { + "epoch": 0.31292, + "grad_norm": 1.570296753171117, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 31292 + }, + { + "epoch": 0.31293, + "grad_norm": 1.4036533357699295, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 31293 + }, + { + "epoch": 0.31294, + "grad_norm": 1.4196275938850498, + "learning_rate": 0.003, + "loss": 4.079, + "step": 31294 + }, + { + "epoch": 0.31295, + "grad_norm": 1.2440216790894112, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31295 + }, + { + "epoch": 0.31296, + "grad_norm": 1.7560020574756123, + "learning_rate": 0.003, + "loss": 4.054, + "step": 31296 + }, + { + "epoch": 0.31297, + "grad_norm": 1.333434031110636, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 31297 + }, + { + "epoch": 0.31298, + "grad_norm": 1.7874825443107778, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 31298 + }, + { + "epoch": 0.31299, + "grad_norm": 1.4541636123791268, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 31299 + }, + { + "epoch": 0.313, + "grad_norm": 1.311932224248857, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31300 + }, + { + "epoch": 0.31301, + "grad_norm": 1.285120562067753, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 31301 + }, + { + "epoch": 0.31302, + "grad_norm": 1.7777528893237637, + "learning_rate": 0.003, + "loss": 4.071, + "step": 31302 + }, + { + "epoch": 0.31303, + "grad_norm": 1.1035161477697975, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 31303 + }, + { + "epoch": 0.31304, + "grad_norm": 1.9054094991292974, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 31304 + }, + { + "epoch": 0.31305, + "grad_norm": 1.3196360594028074, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31305 + }, + { + "epoch": 0.31306, + "grad_norm": 1.5325374507658092, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31306 + }, + { + "epoch": 0.31307, + "grad_norm": 1.275144635393454, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 31307 + }, + { + "epoch": 0.31308, + "grad_norm": 1.5218215617659303, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 31308 + }, + { + "epoch": 0.31309, + "grad_norm": 1.4350110738584212, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 31309 + }, + { + "epoch": 0.3131, + "grad_norm": 1.3867960040803913, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 31310 + }, + { + "epoch": 0.31311, + "grad_norm": 1.303184298130891, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31311 + }, + { + "epoch": 0.31312, + "grad_norm": 1.523606701687459, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 31312 + }, + { + "epoch": 0.31313, + "grad_norm": 1.3485920652523231, + "learning_rate": 0.003, + "loss": 4.043, + "step": 31313 + }, + { + "epoch": 0.31314, + "grad_norm": 1.6307714562536875, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 31314 + }, + { + "epoch": 0.31315, + "grad_norm": 1.721471013429387, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 31315 + }, + { + "epoch": 0.31316, + "grad_norm": 1.539206621751055, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 31316 + }, + { + "epoch": 0.31317, + "grad_norm": 1.4615412256620022, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31317 + }, + { + "epoch": 0.31318, + "grad_norm": 1.708719353939305, + "learning_rate": 0.003, + "loss": 4.05, + "step": 31318 + }, + { + "epoch": 0.31319, + "grad_norm": 1.0222209170698073, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 31319 + }, + { + "epoch": 0.3132, + "grad_norm": 1.714559372546531, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 31320 + }, + { + "epoch": 0.31321, + "grad_norm": 1.10312273039813, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31321 + }, + { + "epoch": 0.31322, + "grad_norm": 1.5379327828675606, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 31322 + }, + { + "epoch": 0.31323, + "grad_norm": 1.6731674421863545, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 31323 + }, + { + "epoch": 0.31324, + "grad_norm": 1.5557428849504964, + "learning_rate": 0.003, + "loss": 4.05, + "step": 31324 + }, + { + "epoch": 0.31325, + "grad_norm": 1.3656661476016492, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 31325 + }, + { + "epoch": 0.31326, + "grad_norm": 1.5882538366130214, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 31326 + }, + { + "epoch": 0.31327, + "grad_norm": 1.2170439785932152, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 31327 + }, + { + "epoch": 0.31328, + "grad_norm": 1.4520987751322625, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 31328 + }, + { + "epoch": 0.31329, + "grad_norm": 1.5818569607005075, + "learning_rate": 0.003, + "loss": 4.036, + "step": 31329 + }, + { + "epoch": 0.3133, + "grad_norm": 1.3122132421208619, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 31330 + }, + { + "epoch": 0.31331, + "grad_norm": 1.3453954044865537, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31331 + }, + { + "epoch": 0.31332, + "grad_norm": 1.5251494406592554, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 31332 + }, + { + "epoch": 0.31333, + "grad_norm": 1.363883216221518, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31333 + }, + { + "epoch": 0.31334, + "grad_norm": 1.6934757687622146, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31334 + }, + { + "epoch": 0.31335, + "grad_norm": 1.2581923323813338, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 31335 + }, + { + "epoch": 0.31336, + "grad_norm": 1.4888763637791627, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 31336 + }, + { + "epoch": 0.31337, + "grad_norm": 1.3058413909107776, + "learning_rate": 0.003, + "loss": 4.039, + "step": 31337 + }, + { + "epoch": 0.31338, + "grad_norm": 1.5006714154575562, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31338 + }, + { + "epoch": 0.31339, + "grad_norm": 1.661168022938388, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 31339 + }, + { + "epoch": 0.3134, + "grad_norm": 1.5628102579264884, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 31340 + }, + { + "epoch": 0.31341, + "grad_norm": 1.6311747611398346, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31341 + }, + { + "epoch": 0.31342, + "grad_norm": 1.2664245639383453, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 31342 + }, + { + "epoch": 0.31343, + "grad_norm": 1.611231749428014, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 31343 + }, + { + "epoch": 0.31344, + "grad_norm": 1.3746327290426597, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31344 + }, + { + "epoch": 0.31345, + "grad_norm": 1.6597168808729368, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 31345 + }, + { + "epoch": 0.31346, + "grad_norm": 1.3530380968162534, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 31346 + }, + { + "epoch": 0.31347, + "grad_norm": 1.53588528534034, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 31347 + }, + { + "epoch": 0.31348, + "grad_norm": 1.4637611079349717, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 31348 + }, + { + "epoch": 0.31349, + "grad_norm": 1.216801764117153, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 31349 + }, + { + "epoch": 0.3135, + "grad_norm": 1.5723865762366285, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 31350 + }, + { + "epoch": 0.31351, + "grad_norm": 1.469728989585046, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 31351 + }, + { + "epoch": 0.31352, + "grad_norm": 1.4779749487165337, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31352 + }, + { + "epoch": 0.31353, + "grad_norm": 1.374146692570771, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31353 + }, + { + "epoch": 0.31354, + "grad_norm": 1.4767045051682208, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 31354 + }, + { + "epoch": 0.31355, + "grad_norm": 1.5342102172130603, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 31355 + }, + { + "epoch": 0.31356, + "grad_norm": 1.3703270211867384, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 31356 + }, + { + "epoch": 0.31357, + "grad_norm": 1.4162517263703183, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31357 + }, + { + "epoch": 0.31358, + "grad_norm": 1.4027809101321351, + "learning_rate": 0.003, + "loss": 4.055, + "step": 31358 + }, + { + "epoch": 0.31359, + "grad_norm": 1.4293985966026888, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 31359 + }, + { + "epoch": 0.3136, + "grad_norm": 1.285633189376554, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 31360 + }, + { + "epoch": 0.31361, + "grad_norm": 1.6731685781912327, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31361 + }, + { + "epoch": 0.31362, + "grad_norm": 1.3984843701798793, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31362 + }, + { + "epoch": 0.31363, + "grad_norm": 1.6908246557994293, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 31363 + }, + { + "epoch": 0.31364, + "grad_norm": 1.7096657496335117, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 31364 + }, + { + "epoch": 0.31365, + "grad_norm": 1.598908285446209, + "learning_rate": 0.003, + "loss": 4.045, + "step": 31365 + }, + { + "epoch": 0.31366, + "grad_norm": 1.2704372766395664, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 31366 + }, + { + "epoch": 0.31367, + "grad_norm": 1.3774493401691326, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31367 + }, + { + "epoch": 0.31368, + "grad_norm": 1.4543152960800418, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 31368 + }, + { + "epoch": 0.31369, + "grad_norm": 1.502021506254183, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31369 + }, + { + "epoch": 0.3137, + "grad_norm": 1.787502800390195, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 31370 + }, + { + "epoch": 0.31371, + "grad_norm": 1.3790980361938114, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 31371 + }, + { + "epoch": 0.31372, + "grad_norm": 1.5045676808436395, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 31372 + }, + { + "epoch": 0.31373, + "grad_norm": 1.2915877009673198, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31373 + }, + { + "epoch": 0.31374, + "grad_norm": 1.6689453815583226, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31374 + }, + { + "epoch": 0.31375, + "grad_norm": 1.4893597945756372, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 31375 + }, + { + "epoch": 0.31376, + "grad_norm": 1.4640124019157863, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 31376 + }, + { + "epoch": 0.31377, + "grad_norm": 1.6246204260314472, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31377 + }, + { + "epoch": 0.31378, + "grad_norm": 1.2336310314650076, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 31378 + }, + { + "epoch": 0.31379, + "grad_norm": 1.4268867685885496, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31379 + }, + { + "epoch": 0.3138, + "grad_norm": 1.3714707200413871, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 31380 + }, + { + "epoch": 0.31381, + "grad_norm": 1.6477485218236216, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 31381 + }, + { + "epoch": 0.31382, + "grad_norm": 1.2988394652056636, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 31382 + }, + { + "epoch": 0.31383, + "grad_norm": 1.4935521948423585, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31383 + }, + { + "epoch": 0.31384, + "grad_norm": 1.3589059918162942, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 31384 + }, + { + "epoch": 0.31385, + "grad_norm": 1.4890924200110733, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 31385 + }, + { + "epoch": 0.31386, + "grad_norm": 1.424829585133665, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31386 + }, + { + "epoch": 0.31387, + "grad_norm": 1.56585326487416, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31387 + }, + { + "epoch": 0.31388, + "grad_norm": 1.530635931624457, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 31388 + }, + { + "epoch": 0.31389, + "grad_norm": 1.2382979387790356, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 31389 + }, + { + "epoch": 0.3139, + "grad_norm": 1.7760343117351245, + "learning_rate": 0.003, + "loss": 4.063, + "step": 31390 + }, + { + "epoch": 0.31391, + "grad_norm": 1.1415309148008839, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 31391 + }, + { + "epoch": 0.31392, + "grad_norm": 1.5318120162057176, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 31392 + }, + { + "epoch": 0.31393, + "grad_norm": 1.498602470952099, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 31393 + }, + { + "epoch": 0.31394, + "grad_norm": 1.415167425536915, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 31394 + }, + { + "epoch": 0.31395, + "grad_norm": 1.301039845192683, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 31395 + }, + { + "epoch": 0.31396, + "grad_norm": 1.5067371388095359, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 31396 + }, + { + "epoch": 0.31397, + "grad_norm": 1.4427344697142717, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 31397 + }, + { + "epoch": 0.31398, + "grad_norm": 1.53816558694614, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31398 + }, + { + "epoch": 0.31399, + "grad_norm": 1.6350082610279957, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31399 + }, + { + "epoch": 0.314, + "grad_norm": 1.5313345260904867, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 31400 + }, + { + "epoch": 0.31401, + "grad_norm": 1.6669355774038344, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 31401 + }, + { + "epoch": 0.31402, + "grad_norm": 1.2957119015891008, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 31402 + }, + { + "epoch": 0.31403, + "grad_norm": 1.5817356806255773, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 31403 + }, + { + "epoch": 0.31404, + "grad_norm": 1.2674265902067692, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 31404 + }, + { + "epoch": 0.31405, + "grad_norm": 1.4104290927640988, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 31405 + }, + { + "epoch": 0.31406, + "grad_norm": 1.3904461218839559, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 31406 + }, + { + "epoch": 0.31407, + "grad_norm": 1.590748198642492, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31407 + }, + { + "epoch": 0.31408, + "grad_norm": 1.939264177531695, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 31408 + }, + { + "epoch": 0.31409, + "grad_norm": 1.146278669978246, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 31409 + }, + { + "epoch": 0.3141, + "grad_norm": 1.6529249304200306, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31410 + }, + { + "epoch": 0.31411, + "grad_norm": 1.5591134876360413, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 31411 + }, + { + "epoch": 0.31412, + "grad_norm": 1.376394201876212, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 31412 + }, + { + "epoch": 0.31413, + "grad_norm": 1.6458580394348643, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31413 + }, + { + "epoch": 0.31414, + "grad_norm": 1.3989203538880974, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 31414 + }, + { + "epoch": 0.31415, + "grad_norm": 1.5087641110954382, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 31415 + }, + { + "epoch": 0.31416, + "grad_norm": 1.2215278333909203, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 31416 + }, + { + "epoch": 0.31417, + "grad_norm": 1.4880163220130653, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31417 + }, + { + "epoch": 0.31418, + "grad_norm": 1.4409901579179047, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 31418 + }, + { + "epoch": 0.31419, + "grad_norm": 1.3899258869692985, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 31419 + }, + { + "epoch": 0.3142, + "grad_norm": 1.4806814749696988, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31420 + }, + { + "epoch": 0.31421, + "grad_norm": 1.2843291247284825, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 31421 + }, + { + "epoch": 0.31422, + "grad_norm": 1.7083921826841308, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31422 + }, + { + "epoch": 0.31423, + "grad_norm": 1.505151126223095, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 31423 + }, + { + "epoch": 0.31424, + "grad_norm": 1.6551750581157174, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 31424 + }, + { + "epoch": 0.31425, + "grad_norm": 1.4443248044531785, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 31425 + }, + { + "epoch": 0.31426, + "grad_norm": 1.4209992451989488, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 31426 + }, + { + "epoch": 0.31427, + "grad_norm": 1.7896325690923567, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 31427 + }, + { + "epoch": 0.31428, + "grad_norm": 1.4561682415465433, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 31428 + }, + { + "epoch": 0.31429, + "grad_norm": 1.308170757852549, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31429 + }, + { + "epoch": 0.3143, + "grad_norm": 1.4106403059949706, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 31430 + }, + { + "epoch": 0.31431, + "grad_norm": 1.354236495297284, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 31431 + }, + { + "epoch": 0.31432, + "grad_norm": 1.30907045939332, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31432 + }, + { + "epoch": 0.31433, + "grad_norm": 1.391687942887727, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 31433 + }, + { + "epoch": 0.31434, + "grad_norm": 1.4128723688094227, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 31434 + }, + { + "epoch": 0.31435, + "grad_norm": 1.3797962070910894, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31435 + }, + { + "epoch": 0.31436, + "grad_norm": 1.775484773593172, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 31436 + }, + { + "epoch": 0.31437, + "grad_norm": 1.2057183552555955, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31437 + }, + { + "epoch": 0.31438, + "grad_norm": 1.7555234880181123, + "learning_rate": 0.003, + "loss": 4.026, + "step": 31438 + }, + { + "epoch": 0.31439, + "grad_norm": 1.6416287097083067, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 31439 + }, + { + "epoch": 0.3144, + "grad_norm": 1.2753955154111747, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 31440 + }, + { + "epoch": 0.31441, + "grad_norm": 1.530830725497449, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 31441 + }, + { + "epoch": 0.31442, + "grad_norm": 1.4993260570273315, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31442 + }, + { + "epoch": 0.31443, + "grad_norm": 1.714757286204135, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31443 + }, + { + "epoch": 0.31444, + "grad_norm": 1.3399920094302271, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31444 + }, + { + "epoch": 0.31445, + "grad_norm": 1.3713593861338527, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 31445 + }, + { + "epoch": 0.31446, + "grad_norm": 1.6133017114911, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31446 + }, + { + "epoch": 0.31447, + "grad_norm": 1.3572174864169986, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 31447 + }, + { + "epoch": 0.31448, + "grad_norm": 1.433445390863004, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 31448 + }, + { + "epoch": 0.31449, + "grad_norm": 1.3864090936059892, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31449 + }, + { + "epoch": 0.3145, + "grad_norm": 1.4703732752611227, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 31450 + }, + { + "epoch": 0.31451, + "grad_norm": 1.7186809643021, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 31451 + }, + { + "epoch": 0.31452, + "grad_norm": 1.2738232870888493, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 31452 + }, + { + "epoch": 0.31453, + "grad_norm": 1.551022374803525, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 31453 + }, + { + "epoch": 0.31454, + "grad_norm": 1.135855264866556, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 31454 + }, + { + "epoch": 0.31455, + "grad_norm": 1.4784941095011654, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31455 + }, + { + "epoch": 0.31456, + "grad_norm": 1.3941224730946187, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 31456 + }, + { + "epoch": 0.31457, + "grad_norm": 1.936680110430944, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 31457 + }, + { + "epoch": 0.31458, + "grad_norm": 1.165534848228759, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 31458 + }, + { + "epoch": 0.31459, + "grad_norm": 1.4700443771500615, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 31459 + }, + { + "epoch": 0.3146, + "grad_norm": 1.5503323053850975, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31460 + }, + { + "epoch": 0.31461, + "grad_norm": 1.7142486558000818, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 31461 + }, + { + "epoch": 0.31462, + "grad_norm": 1.4942975279093091, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 31462 + }, + { + "epoch": 0.31463, + "grad_norm": 1.5636182446052997, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 31463 + }, + { + "epoch": 0.31464, + "grad_norm": 1.6002528902902795, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 31464 + }, + { + "epoch": 0.31465, + "grad_norm": 1.1835723825790474, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 31465 + }, + { + "epoch": 0.31466, + "grad_norm": 1.5432092475209482, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31466 + }, + { + "epoch": 0.31467, + "grad_norm": 1.3876750335034145, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31467 + }, + { + "epoch": 0.31468, + "grad_norm": 1.4656966107086151, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 31468 + }, + { + "epoch": 0.31469, + "grad_norm": 1.562999110962671, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 31469 + }, + { + "epoch": 0.3147, + "grad_norm": 1.4116315371009065, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 31470 + }, + { + "epoch": 0.31471, + "grad_norm": 1.5361125837480494, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 31471 + }, + { + "epoch": 0.31472, + "grad_norm": 1.6604520190083318, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31472 + }, + { + "epoch": 0.31473, + "grad_norm": 1.3099135895053784, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 31473 + }, + { + "epoch": 0.31474, + "grad_norm": 1.41960975236322, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 31474 + }, + { + "epoch": 0.31475, + "grad_norm": 1.3815643026002162, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 31475 + }, + { + "epoch": 0.31476, + "grad_norm": 1.574014932419494, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 31476 + }, + { + "epoch": 0.31477, + "grad_norm": 1.2923264188185815, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 31477 + }, + { + "epoch": 0.31478, + "grad_norm": 1.8769984335404755, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31478 + }, + { + "epoch": 0.31479, + "grad_norm": 1.4726535316154536, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 31479 + }, + { + "epoch": 0.3148, + "grad_norm": 1.5722181608754675, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 31480 + }, + { + "epoch": 0.31481, + "grad_norm": 1.4013262587295814, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 31481 + }, + { + "epoch": 0.31482, + "grad_norm": 1.4804922404224052, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 31482 + }, + { + "epoch": 0.31483, + "grad_norm": 1.4282942060662214, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 31483 + }, + { + "epoch": 0.31484, + "grad_norm": 1.58798307001112, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 31484 + }, + { + "epoch": 0.31485, + "grad_norm": 1.40982121491937, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31485 + }, + { + "epoch": 0.31486, + "grad_norm": 1.605301291458939, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 31486 + }, + { + "epoch": 0.31487, + "grad_norm": 1.1455532210763921, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31487 + }, + { + "epoch": 0.31488, + "grad_norm": 1.7155989111828291, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31488 + }, + { + "epoch": 0.31489, + "grad_norm": 1.0923329626601896, + "learning_rate": 0.003, + "loss": 4.037, + "step": 31489 + }, + { + "epoch": 0.3149, + "grad_norm": 1.6864191891176794, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 31490 + }, + { + "epoch": 0.31491, + "grad_norm": 1.481774456231294, + "learning_rate": 0.003, + "loss": 4.058, + "step": 31491 + }, + { + "epoch": 0.31492, + "grad_norm": 1.526379932179963, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 31492 + }, + { + "epoch": 0.31493, + "grad_norm": 1.3738966407587605, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31493 + }, + { + "epoch": 0.31494, + "grad_norm": 1.447358399015166, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 31494 + }, + { + "epoch": 0.31495, + "grad_norm": 1.5272168062874534, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 31495 + }, + { + "epoch": 0.31496, + "grad_norm": 1.246396470069403, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 31496 + }, + { + "epoch": 0.31497, + "grad_norm": 1.647037473154409, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31497 + }, + { + "epoch": 0.31498, + "grad_norm": 1.3130736022885172, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 31498 + }, + { + "epoch": 0.31499, + "grad_norm": 1.451257076192651, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 31499 + }, + { + "epoch": 0.315, + "grad_norm": 1.4364394162456051, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31500 + }, + { + "epoch": 0.31501, + "grad_norm": 1.3864520761697858, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31501 + }, + { + "epoch": 0.31502, + "grad_norm": 1.6269439899173563, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 31502 + }, + { + "epoch": 0.31503, + "grad_norm": 1.6020821132875005, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 31503 + }, + { + "epoch": 0.31504, + "grad_norm": 1.8026337441767697, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31504 + }, + { + "epoch": 0.31505, + "grad_norm": 1.4301875383842746, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 31505 + }, + { + "epoch": 0.31506, + "grad_norm": 1.4543107343425619, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 31506 + }, + { + "epoch": 0.31507, + "grad_norm": 1.4265005022358568, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 31507 + }, + { + "epoch": 0.31508, + "grad_norm": 1.4652216051825044, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31508 + }, + { + "epoch": 0.31509, + "grad_norm": 1.3604200974595633, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31509 + }, + { + "epoch": 0.3151, + "grad_norm": 1.7464805102855365, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31510 + }, + { + "epoch": 0.31511, + "grad_norm": 1.1586231693095117, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 31511 + }, + { + "epoch": 0.31512, + "grad_norm": 1.7888157382454217, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 31512 + }, + { + "epoch": 0.31513, + "grad_norm": 1.2170714000116047, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 31513 + }, + { + "epoch": 0.31514, + "grad_norm": 1.5657413971181857, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31514 + }, + { + "epoch": 0.31515, + "grad_norm": 1.3166987222905293, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 31515 + }, + { + "epoch": 0.31516, + "grad_norm": 1.3794297540605065, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 31516 + }, + { + "epoch": 0.31517, + "grad_norm": 1.897835538683095, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 31517 + }, + { + "epoch": 0.31518, + "grad_norm": 1.308097084969569, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 31518 + }, + { + "epoch": 0.31519, + "grad_norm": 2.1217537685660504, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 31519 + }, + { + "epoch": 0.3152, + "grad_norm": 1.1207537409130393, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31520 + }, + { + "epoch": 0.31521, + "grad_norm": 1.5661398305333487, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31521 + }, + { + "epoch": 0.31522, + "grad_norm": 1.488158259873089, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 31522 + }, + { + "epoch": 0.31523, + "grad_norm": 1.740151491006189, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 31523 + }, + { + "epoch": 0.31524, + "grad_norm": 1.2039394193277067, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 31524 + }, + { + "epoch": 0.31525, + "grad_norm": 1.6141709649488931, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 31525 + }, + { + "epoch": 0.31526, + "grad_norm": 1.1284749787327482, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31526 + }, + { + "epoch": 0.31527, + "grad_norm": 1.4089419243426669, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 31527 + }, + { + "epoch": 0.31528, + "grad_norm": 1.5056493947574523, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31528 + }, + { + "epoch": 0.31529, + "grad_norm": 1.4271903628472864, + "learning_rate": 0.003, + "loss": 4.044, + "step": 31529 + }, + { + "epoch": 0.3153, + "grad_norm": 1.4815882861660281, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 31530 + }, + { + "epoch": 0.31531, + "grad_norm": 1.3641020259612282, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 31531 + }, + { + "epoch": 0.31532, + "grad_norm": 1.3968577244072102, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 31532 + }, + { + "epoch": 0.31533, + "grad_norm": 1.4579018855426231, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31533 + }, + { + "epoch": 0.31534, + "grad_norm": 1.469581036869854, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 31534 + }, + { + "epoch": 0.31535, + "grad_norm": 1.242314459660967, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 31535 + }, + { + "epoch": 0.31536, + "grad_norm": 1.7607237325756695, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 31536 + }, + { + "epoch": 0.31537, + "grad_norm": 1.3039406024657578, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 31537 + }, + { + "epoch": 0.31538, + "grad_norm": 1.3794093454186043, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31538 + }, + { + "epoch": 0.31539, + "grad_norm": 1.568254605694663, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 31539 + }, + { + "epoch": 0.3154, + "grad_norm": 1.292988623589103, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31540 + }, + { + "epoch": 0.31541, + "grad_norm": 1.5466302866287298, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 31541 + }, + { + "epoch": 0.31542, + "grad_norm": 1.275758237303295, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31542 + }, + { + "epoch": 0.31543, + "grad_norm": 1.5276016024692765, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 31543 + }, + { + "epoch": 0.31544, + "grad_norm": 1.5945740924674685, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 31544 + }, + { + "epoch": 0.31545, + "grad_norm": 1.4737102741368358, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31545 + }, + { + "epoch": 0.31546, + "grad_norm": 1.5801881591731919, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31546 + }, + { + "epoch": 0.31547, + "grad_norm": 1.071552306389584, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 31547 + }, + { + "epoch": 0.31548, + "grad_norm": 1.707454778699416, + "learning_rate": 0.003, + "loss": 4.02, + "step": 31548 + }, + { + "epoch": 0.31549, + "grad_norm": 1.3168469855883964, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 31549 + }, + { + "epoch": 0.3155, + "grad_norm": 1.63421032569156, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 31550 + }, + { + "epoch": 0.31551, + "grad_norm": 1.3339250538281535, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 31551 + }, + { + "epoch": 0.31552, + "grad_norm": 1.5174035999046547, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31552 + }, + { + "epoch": 0.31553, + "grad_norm": 1.4718105790927343, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31553 + }, + { + "epoch": 0.31554, + "grad_norm": 1.4538579978809913, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 31554 + }, + { + "epoch": 0.31555, + "grad_norm": 1.4035181286926177, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 31555 + }, + { + "epoch": 0.31556, + "grad_norm": 1.5911751232284095, + "learning_rate": 0.003, + "loss": 4.068, + "step": 31556 + }, + { + "epoch": 0.31557, + "grad_norm": 1.7870980097129352, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 31557 + }, + { + "epoch": 0.31558, + "grad_norm": 1.6062451047904904, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 31558 + }, + { + "epoch": 0.31559, + "grad_norm": 1.5286288475241037, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31559 + }, + { + "epoch": 0.3156, + "grad_norm": 1.6059931358636241, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31560 + }, + { + "epoch": 0.31561, + "grad_norm": 1.274004393487784, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 31561 + }, + { + "epoch": 0.31562, + "grad_norm": 1.7862351157247844, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 31562 + }, + { + "epoch": 0.31563, + "grad_norm": 1.2170113909693852, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 31563 + }, + { + "epoch": 0.31564, + "grad_norm": 1.6208226180407206, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31564 + }, + { + "epoch": 0.31565, + "grad_norm": 1.5749342021012989, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 31565 + }, + { + "epoch": 0.31566, + "grad_norm": 1.353010935797751, + "learning_rate": 0.003, + "loss": 4.028, + "step": 31566 + }, + { + "epoch": 0.31567, + "grad_norm": 1.3347316307510781, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31567 + }, + { + "epoch": 0.31568, + "grad_norm": 1.4331689590233185, + "learning_rate": 0.003, + "loss": 4.056, + "step": 31568 + }, + { + "epoch": 0.31569, + "grad_norm": 1.3901495128155723, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31569 + }, + { + "epoch": 0.3157, + "grad_norm": 1.278376295751996, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 31570 + }, + { + "epoch": 0.31571, + "grad_norm": 1.6365718510401674, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 31571 + }, + { + "epoch": 0.31572, + "grad_norm": 1.4161657388066822, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 31572 + }, + { + "epoch": 0.31573, + "grad_norm": 1.740169560602989, + "learning_rate": 0.003, + "loss": 4.023, + "step": 31573 + }, + { + "epoch": 0.31574, + "grad_norm": 1.5888334426492972, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 31574 + }, + { + "epoch": 0.31575, + "grad_norm": 1.582387684303911, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 31575 + }, + { + "epoch": 0.31576, + "grad_norm": 1.7963350104740712, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 31576 + }, + { + "epoch": 0.31577, + "grad_norm": 1.2355453490211616, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31577 + }, + { + "epoch": 0.31578, + "grad_norm": 1.398072439237864, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 31578 + }, + { + "epoch": 0.31579, + "grad_norm": 1.2216418810173124, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 31579 + }, + { + "epoch": 0.3158, + "grad_norm": 1.8114781181352908, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 31580 + }, + { + "epoch": 0.31581, + "grad_norm": 1.2742142879359364, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31581 + }, + { + "epoch": 0.31582, + "grad_norm": 1.8035311086543704, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31582 + }, + { + "epoch": 0.31583, + "grad_norm": 1.313711787417442, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31583 + }, + { + "epoch": 0.31584, + "grad_norm": 1.3942250586593792, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 31584 + }, + { + "epoch": 0.31585, + "grad_norm": 1.5089348003404786, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 31585 + }, + { + "epoch": 0.31586, + "grad_norm": 1.2231826082589665, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31586 + }, + { + "epoch": 0.31587, + "grad_norm": 1.487264425451158, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 31587 + }, + { + "epoch": 0.31588, + "grad_norm": 1.4360524095343912, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 31588 + }, + { + "epoch": 0.31589, + "grad_norm": 1.6939170875462592, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 31589 + }, + { + "epoch": 0.3159, + "grad_norm": 1.3718781649736735, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31590 + }, + { + "epoch": 0.31591, + "grad_norm": 1.3682487699953287, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 31591 + }, + { + "epoch": 0.31592, + "grad_norm": 1.669437766597689, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 31592 + }, + { + "epoch": 0.31593, + "grad_norm": 1.363068487151912, + "learning_rate": 0.003, + "loss": 4.032, + "step": 31593 + }, + { + "epoch": 0.31594, + "grad_norm": 1.5964507695737535, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 31594 + }, + { + "epoch": 0.31595, + "grad_norm": 1.4912797022517845, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 31595 + }, + { + "epoch": 0.31596, + "grad_norm": 1.2707464466776544, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 31596 + }, + { + "epoch": 0.31597, + "grad_norm": 1.6164848409243675, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 31597 + }, + { + "epoch": 0.31598, + "grad_norm": 1.163057056327509, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 31598 + }, + { + "epoch": 0.31599, + "grad_norm": 1.7447249825089997, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31599 + }, + { + "epoch": 0.316, + "grad_norm": 1.0056973932287445, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 31600 + }, + { + "epoch": 0.31601, + "grad_norm": 1.752827794037149, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 31601 + }, + { + "epoch": 0.31602, + "grad_norm": 1.264814474433306, + "learning_rate": 0.003, + "loss": 4.068, + "step": 31602 + }, + { + "epoch": 0.31603, + "grad_norm": 1.921082401934801, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 31603 + }, + { + "epoch": 0.31604, + "grad_norm": 1.6035615792460596, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31604 + }, + { + "epoch": 0.31605, + "grad_norm": 1.699129857604046, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 31605 + }, + { + "epoch": 0.31606, + "grad_norm": 1.3043519860201274, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 31606 + }, + { + "epoch": 0.31607, + "grad_norm": 1.3663167816179504, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31607 + }, + { + "epoch": 0.31608, + "grad_norm": 1.4169061129483018, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31608 + }, + { + "epoch": 0.31609, + "grad_norm": 1.4834977512646612, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 31609 + }, + { + "epoch": 0.3161, + "grad_norm": 1.3114843606597755, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 31610 + }, + { + "epoch": 0.31611, + "grad_norm": 1.5978058577917753, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 31611 + }, + { + "epoch": 0.31612, + "grad_norm": 1.271103270349528, + "learning_rate": 0.003, + "loss": 4.082, + "step": 31612 + }, + { + "epoch": 0.31613, + "grad_norm": 1.5028398747733056, + "learning_rate": 0.003, + "loss": 4.051, + "step": 31613 + }, + { + "epoch": 0.31614, + "grad_norm": 1.4048928328285024, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 31614 + }, + { + "epoch": 0.31615, + "grad_norm": 1.905314541621983, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 31615 + }, + { + "epoch": 0.31616, + "grad_norm": 1.548764162647382, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 31616 + }, + { + "epoch": 0.31617, + "grad_norm": 1.3157198940097055, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 31617 + }, + { + "epoch": 0.31618, + "grad_norm": 1.577588759153033, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31618 + }, + { + "epoch": 0.31619, + "grad_norm": 1.362493783900725, + "learning_rate": 0.003, + "loss": 4.059, + "step": 31619 + }, + { + "epoch": 0.3162, + "grad_norm": 1.4768250590179022, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31620 + }, + { + "epoch": 0.31621, + "grad_norm": 1.44169322976058, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31621 + }, + { + "epoch": 0.31622, + "grad_norm": 1.6473099819627086, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31622 + }, + { + "epoch": 0.31623, + "grad_norm": 1.3112375511965824, + "learning_rate": 0.003, + "loss": 4.021, + "step": 31623 + }, + { + "epoch": 0.31624, + "grad_norm": 1.4757340443799825, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 31624 + }, + { + "epoch": 0.31625, + "grad_norm": 1.4258659120094104, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 31625 + }, + { + "epoch": 0.31626, + "grad_norm": 1.239889827199201, + "learning_rate": 0.003, + "loss": 4.045, + "step": 31626 + }, + { + "epoch": 0.31627, + "grad_norm": 1.4418665143835763, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 31627 + }, + { + "epoch": 0.31628, + "grad_norm": 1.3891325836720605, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 31628 + }, + { + "epoch": 0.31629, + "grad_norm": 1.5175054891564652, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31629 + }, + { + "epoch": 0.3163, + "grad_norm": 1.424588903224038, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 31630 + }, + { + "epoch": 0.31631, + "grad_norm": 1.734623473089667, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 31631 + }, + { + "epoch": 0.31632, + "grad_norm": 1.36890079125464, + "learning_rate": 0.003, + "loss": 4.067, + "step": 31632 + }, + { + "epoch": 0.31633, + "grad_norm": 1.693207013923669, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31633 + }, + { + "epoch": 0.31634, + "grad_norm": 1.3123367777480406, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 31634 + }, + { + "epoch": 0.31635, + "grad_norm": 1.5245391013626344, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 31635 + }, + { + "epoch": 0.31636, + "grad_norm": 1.661043875546939, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 31636 + }, + { + "epoch": 0.31637, + "grad_norm": 1.5155800400352224, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 31637 + }, + { + "epoch": 0.31638, + "grad_norm": 1.5624930222281246, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 31638 + }, + { + "epoch": 0.31639, + "grad_norm": 1.2073612537156755, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 31639 + }, + { + "epoch": 0.3164, + "grad_norm": 1.5870923245723212, + "learning_rate": 0.003, + "loss": 4.023, + "step": 31640 + }, + { + "epoch": 0.31641, + "grad_norm": 1.3835699117558538, + "learning_rate": 0.003, + "loss": 4.018, + "step": 31641 + }, + { + "epoch": 0.31642, + "grad_norm": 1.7510001333043788, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 31642 + }, + { + "epoch": 0.31643, + "grad_norm": 1.62232173211784, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 31643 + }, + { + "epoch": 0.31644, + "grad_norm": 1.2661322738142926, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31644 + }, + { + "epoch": 0.31645, + "grad_norm": 1.524136387469188, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 31645 + }, + { + "epoch": 0.31646, + "grad_norm": 1.6757132320170345, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 31646 + }, + { + "epoch": 0.31647, + "grad_norm": 1.3025076697004108, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 31647 + }, + { + "epoch": 0.31648, + "grad_norm": 1.581241654873218, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 31648 + }, + { + "epoch": 0.31649, + "grad_norm": 1.312820592696001, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31649 + }, + { + "epoch": 0.3165, + "grad_norm": 1.4237908382207511, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 31650 + }, + { + "epoch": 0.31651, + "grad_norm": 1.146343948366054, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 31651 + }, + { + "epoch": 0.31652, + "grad_norm": 1.6453070776790346, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 31652 + }, + { + "epoch": 0.31653, + "grad_norm": 1.438640704065881, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 31653 + }, + { + "epoch": 0.31654, + "grad_norm": 1.4284416470187564, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 31654 + }, + { + "epoch": 0.31655, + "grad_norm": 1.5513125132623162, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 31655 + }, + { + "epoch": 0.31656, + "grad_norm": 1.6340933728831737, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 31656 + }, + { + "epoch": 0.31657, + "grad_norm": 1.748454153438018, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 31657 + }, + { + "epoch": 0.31658, + "grad_norm": 1.2715172157705492, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 31658 + }, + { + "epoch": 0.31659, + "grad_norm": 1.606400602616717, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 31659 + }, + { + "epoch": 0.3166, + "grad_norm": 1.686132364938198, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 31660 + }, + { + "epoch": 0.31661, + "grad_norm": 1.4859940172362731, + "learning_rate": 0.003, + "loss": 4.018, + "step": 31661 + }, + { + "epoch": 0.31662, + "grad_norm": 1.4747396956404988, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31662 + }, + { + "epoch": 0.31663, + "grad_norm": 1.250075893879771, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 31663 + }, + { + "epoch": 0.31664, + "grad_norm": 1.620662240293303, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 31664 + }, + { + "epoch": 0.31665, + "grad_norm": 1.2690881573781077, + "learning_rate": 0.003, + "loss": 4.035, + "step": 31665 + }, + { + "epoch": 0.31666, + "grad_norm": 1.465201371058871, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 31666 + }, + { + "epoch": 0.31667, + "grad_norm": 1.3011078965028449, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31667 + }, + { + "epoch": 0.31668, + "grad_norm": 1.5891352006246413, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31668 + }, + { + "epoch": 0.31669, + "grad_norm": 1.6892603590445976, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 31669 + }, + { + "epoch": 0.3167, + "grad_norm": 1.1621989190270703, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 31670 + }, + { + "epoch": 0.31671, + "grad_norm": 1.8192982650922513, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 31671 + }, + { + "epoch": 0.31672, + "grad_norm": 1.5970810178537218, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 31672 + }, + { + "epoch": 0.31673, + "grad_norm": 1.3437172866838676, + "learning_rate": 0.003, + "loss": 4.061, + "step": 31673 + }, + { + "epoch": 0.31674, + "grad_norm": 1.5920919250897243, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 31674 + }, + { + "epoch": 0.31675, + "grad_norm": 1.2152532316731584, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 31675 + }, + { + "epoch": 0.31676, + "grad_norm": 1.5881247509473009, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 31676 + }, + { + "epoch": 0.31677, + "grad_norm": 1.322774533413701, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31677 + }, + { + "epoch": 0.31678, + "grad_norm": 1.4042915687495117, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 31678 + }, + { + "epoch": 0.31679, + "grad_norm": 1.8816855784924422, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 31679 + }, + { + "epoch": 0.3168, + "grad_norm": 1.4112181935230583, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 31680 + }, + { + "epoch": 0.31681, + "grad_norm": 1.407450946339393, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31681 + }, + { + "epoch": 0.31682, + "grad_norm": 1.616611455666668, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 31682 + }, + { + "epoch": 0.31683, + "grad_norm": 1.1555214775688687, + "learning_rate": 0.003, + "loss": 4.028, + "step": 31683 + }, + { + "epoch": 0.31684, + "grad_norm": 1.3965584979789896, + "learning_rate": 0.003, + "loss": 4.013, + "step": 31684 + }, + { + "epoch": 0.31685, + "grad_norm": 1.3816757744682289, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31685 + }, + { + "epoch": 0.31686, + "grad_norm": 1.3865431189674076, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 31686 + }, + { + "epoch": 0.31687, + "grad_norm": 1.6072584576830862, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 31687 + }, + { + "epoch": 0.31688, + "grad_norm": 1.8001060167201146, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 31688 + }, + { + "epoch": 0.31689, + "grad_norm": 1.2432255606406877, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 31689 + }, + { + "epoch": 0.3169, + "grad_norm": 1.4495185023351291, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 31690 + }, + { + "epoch": 0.31691, + "grad_norm": 1.6075360350647, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 31691 + }, + { + "epoch": 0.31692, + "grad_norm": 1.272101751696692, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 31692 + }, + { + "epoch": 0.31693, + "grad_norm": 1.7493622514800824, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31693 + }, + { + "epoch": 0.31694, + "grad_norm": 1.3397117934742282, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 31694 + }, + { + "epoch": 0.31695, + "grad_norm": 1.2147306716781177, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31695 + }, + { + "epoch": 0.31696, + "grad_norm": 1.483387587330142, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 31696 + }, + { + "epoch": 0.31697, + "grad_norm": 1.3630098640272137, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 31697 + }, + { + "epoch": 0.31698, + "grad_norm": 1.4426804488897202, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 31698 + }, + { + "epoch": 0.31699, + "grad_norm": 1.6776077206699411, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 31699 + }, + { + "epoch": 0.317, + "grad_norm": 1.1587123395330716, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 31700 + }, + { + "epoch": 0.31701, + "grad_norm": 1.5521240032713235, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 31701 + }, + { + "epoch": 0.31702, + "grad_norm": 1.267870429821238, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 31702 + }, + { + "epoch": 0.31703, + "grad_norm": 1.592071358904762, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31703 + }, + { + "epoch": 0.31704, + "grad_norm": 1.38226903141223, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31704 + }, + { + "epoch": 0.31705, + "grad_norm": 1.98802952904572, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 31705 + }, + { + "epoch": 0.31706, + "grad_norm": 1.2836735642691472, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 31706 + }, + { + "epoch": 0.31707, + "grad_norm": 1.3432088202169281, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 31707 + }, + { + "epoch": 0.31708, + "grad_norm": 1.508573445473065, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 31708 + }, + { + "epoch": 0.31709, + "grad_norm": 1.8186421556596113, + "learning_rate": 0.003, + "loss": 4.063, + "step": 31709 + }, + { + "epoch": 0.3171, + "grad_norm": 1.281259647847336, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31710 + }, + { + "epoch": 0.31711, + "grad_norm": 1.365002032565585, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31711 + }, + { + "epoch": 0.31712, + "grad_norm": 1.597678510982602, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31712 + }, + { + "epoch": 0.31713, + "grad_norm": 1.2650505674407957, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 31713 + }, + { + "epoch": 0.31714, + "grad_norm": 1.4140755997624934, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 31714 + }, + { + "epoch": 0.31715, + "grad_norm": 1.457636879212183, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 31715 + }, + { + "epoch": 0.31716, + "grad_norm": 1.3402856241497079, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 31716 + }, + { + "epoch": 0.31717, + "grad_norm": 1.5513148648141701, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31717 + }, + { + "epoch": 0.31718, + "grad_norm": 1.5833063773263245, + "learning_rate": 0.003, + "loss": 4.046, + "step": 31718 + }, + { + "epoch": 0.31719, + "grad_norm": 1.3502076509725196, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31719 + }, + { + "epoch": 0.3172, + "grad_norm": 1.491303560515791, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 31720 + }, + { + "epoch": 0.31721, + "grad_norm": 1.3526755182752717, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 31721 + }, + { + "epoch": 0.31722, + "grad_norm": 1.5774831362869632, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 31722 + }, + { + "epoch": 0.31723, + "grad_norm": 1.4404851821360232, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31723 + }, + { + "epoch": 0.31724, + "grad_norm": 1.5613119447317645, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 31724 + }, + { + "epoch": 0.31725, + "grad_norm": 1.292102639754055, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 31725 + }, + { + "epoch": 0.31726, + "grad_norm": 1.5124665683992766, + "learning_rate": 0.003, + "loss": 4.071, + "step": 31726 + }, + { + "epoch": 0.31727, + "grad_norm": 1.6798834758207755, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31727 + }, + { + "epoch": 0.31728, + "grad_norm": 1.5968798058803118, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 31728 + }, + { + "epoch": 0.31729, + "grad_norm": 1.5247197606260043, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 31729 + }, + { + "epoch": 0.3173, + "grad_norm": 1.4258606126475393, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 31730 + }, + { + "epoch": 0.31731, + "grad_norm": 1.44386416631015, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 31731 + }, + { + "epoch": 0.31732, + "grad_norm": 1.568130058358322, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31732 + }, + { + "epoch": 0.31733, + "grad_norm": 1.5887825367050306, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 31733 + }, + { + "epoch": 0.31734, + "grad_norm": 1.50333725351725, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 31734 + }, + { + "epoch": 0.31735, + "grad_norm": 1.4095323120992067, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31735 + }, + { + "epoch": 0.31736, + "grad_norm": 1.4402349038905053, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31736 + }, + { + "epoch": 0.31737, + "grad_norm": 1.5011472336911138, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 31737 + }, + { + "epoch": 0.31738, + "grad_norm": 1.5108459263164509, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 31738 + }, + { + "epoch": 0.31739, + "grad_norm": 1.1972652244518562, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 31739 + }, + { + "epoch": 0.3174, + "grad_norm": 1.8261165747545658, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 31740 + }, + { + "epoch": 0.31741, + "grad_norm": 1.1044771977438284, + "learning_rate": 0.003, + "loss": 4.037, + "step": 31741 + }, + { + "epoch": 0.31742, + "grad_norm": 1.5805010031039386, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 31742 + }, + { + "epoch": 0.31743, + "grad_norm": 1.3894084817737793, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31743 + }, + { + "epoch": 0.31744, + "grad_norm": 1.600071985671135, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31744 + }, + { + "epoch": 0.31745, + "grad_norm": 1.445500498132966, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 31745 + }, + { + "epoch": 0.31746, + "grad_norm": 1.2833744108620488, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 31746 + }, + { + "epoch": 0.31747, + "grad_norm": 1.3839548624687332, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31747 + }, + { + "epoch": 0.31748, + "grad_norm": 1.3636033852901892, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 31748 + }, + { + "epoch": 0.31749, + "grad_norm": 1.5827903213627554, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 31749 + }, + { + "epoch": 0.3175, + "grad_norm": 1.5621379626987992, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 31750 + }, + { + "epoch": 0.31751, + "grad_norm": 1.4641260022783458, + "learning_rate": 0.003, + "loss": 4.056, + "step": 31751 + }, + { + "epoch": 0.31752, + "grad_norm": 1.6130406458553765, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31752 + }, + { + "epoch": 0.31753, + "grad_norm": 1.4563165991929612, + "learning_rate": 0.003, + "loss": 4.057, + "step": 31753 + }, + { + "epoch": 0.31754, + "grad_norm": 1.6246244708756379, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31754 + }, + { + "epoch": 0.31755, + "grad_norm": 1.5935400598895113, + "learning_rate": 0.003, + "loss": 4.059, + "step": 31755 + }, + { + "epoch": 0.31756, + "grad_norm": 1.230686933691166, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 31756 + }, + { + "epoch": 0.31757, + "grad_norm": 1.501897133897795, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31757 + }, + { + "epoch": 0.31758, + "grad_norm": 1.2471836528056428, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 31758 + }, + { + "epoch": 0.31759, + "grad_norm": 1.4953615162721083, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 31759 + }, + { + "epoch": 0.3176, + "grad_norm": 1.79326004902132, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 31760 + }, + { + "epoch": 0.31761, + "grad_norm": 1.4174534867319646, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 31761 + }, + { + "epoch": 0.31762, + "grad_norm": 1.33318256492651, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 31762 + }, + { + "epoch": 0.31763, + "grad_norm": 1.423487210476404, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 31763 + }, + { + "epoch": 0.31764, + "grad_norm": 1.4257636177049313, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 31764 + }, + { + "epoch": 0.31765, + "grad_norm": 1.2757503191428223, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31765 + }, + { + "epoch": 0.31766, + "grad_norm": 1.53509080579218, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 31766 + }, + { + "epoch": 0.31767, + "grad_norm": 1.3186149738558834, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31767 + }, + { + "epoch": 0.31768, + "grad_norm": 1.4155812479202572, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 31768 + }, + { + "epoch": 0.31769, + "grad_norm": 1.4350133043197206, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 31769 + }, + { + "epoch": 0.3177, + "grad_norm": 1.8988860451940728, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 31770 + }, + { + "epoch": 0.31771, + "grad_norm": 1.5524956659514708, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 31771 + }, + { + "epoch": 0.31772, + "grad_norm": 1.668207538007233, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 31772 + }, + { + "epoch": 0.31773, + "grad_norm": 1.5709470098086065, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 31773 + }, + { + "epoch": 0.31774, + "grad_norm": 1.5966524986314738, + "learning_rate": 0.003, + "loss": 4.073, + "step": 31774 + }, + { + "epoch": 0.31775, + "grad_norm": 1.3648876735866056, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 31775 + }, + { + "epoch": 0.31776, + "grad_norm": 1.4476306642301469, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 31776 + }, + { + "epoch": 0.31777, + "grad_norm": 1.3968563508014338, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31777 + }, + { + "epoch": 0.31778, + "grad_norm": 1.498812158241786, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 31778 + }, + { + "epoch": 0.31779, + "grad_norm": 1.4488232138949546, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 31779 + }, + { + "epoch": 0.3178, + "grad_norm": 1.5383352633661915, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 31780 + }, + { + "epoch": 0.31781, + "grad_norm": 1.3610544882366253, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 31781 + }, + { + "epoch": 0.31782, + "grad_norm": 1.6116227029977221, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31782 + }, + { + "epoch": 0.31783, + "grad_norm": 1.523391515932369, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 31783 + }, + { + "epoch": 0.31784, + "grad_norm": 1.2147210050243913, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 31784 + }, + { + "epoch": 0.31785, + "grad_norm": 1.7621281361859942, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 31785 + }, + { + "epoch": 0.31786, + "grad_norm": 1.2301729290512615, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 31786 + }, + { + "epoch": 0.31787, + "grad_norm": 1.5147746927320194, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 31787 + }, + { + "epoch": 0.31788, + "grad_norm": 1.4371629844613811, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 31788 + }, + { + "epoch": 0.31789, + "grad_norm": 1.6072699483368924, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31789 + }, + { + "epoch": 0.3179, + "grad_norm": 1.5075405788350962, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 31790 + }, + { + "epoch": 0.31791, + "grad_norm": 1.5542151547081247, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 31791 + }, + { + "epoch": 0.31792, + "grad_norm": 1.4745279317672257, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 31792 + }, + { + "epoch": 0.31793, + "grad_norm": 1.518581069501834, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 31793 + }, + { + "epoch": 0.31794, + "grad_norm": 1.2738806666320646, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 31794 + }, + { + "epoch": 0.31795, + "grad_norm": 1.6490043575167062, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 31795 + }, + { + "epoch": 0.31796, + "grad_norm": 1.2956983181340989, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 31796 + }, + { + "epoch": 0.31797, + "grad_norm": 1.4445787016008393, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 31797 + }, + { + "epoch": 0.31798, + "grad_norm": 1.327346830748614, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 31798 + }, + { + "epoch": 0.31799, + "grad_norm": 1.627770871833538, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 31799 + }, + { + "epoch": 0.318, + "grad_norm": 1.5006666463389786, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 31800 + }, + { + "epoch": 0.31801, + "grad_norm": 1.6857428722187002, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31801 + }, + { + "epoch": 0.31802, + "grad_norm": 1.868136731902181, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 31802 + }, + { + "epoch": 0.31803, + "grad_norm": 0.9847767569951981, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 31803 + }, + { + "epoch": 0.31804, + "grad_norm": 1.4244941100434287, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31804 + }, + { + "epoch": 0.31805, + "grad_norm": 1.2670541615987694, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 31805 + }, + { + "epoch": 0.31806, + "grad_norm": 1.732574006474403, + "learning_rate": 0.003, + "loss": 4.059, + "step": 31806 + }, + { + "epoch": 0.31807, + "grad_norm": 1.3812514545609933, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 31807 + }, + { + "epoch": 0.31808, + "grad_norm": 1.7119330476623769, + "learning_rate": 0.003, + "loss": 4.048, + "step": 31808 + }, + { + "epoch": 0.31809, + "grad_norm": 1.4309780530188563, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31809 + }, + { + "epoch": 0.3181, + "grad_norm": 1.5707304466389076, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 31810 + }, + { + "epoch": 0.31811, + "grad_norm": 1.3190524577110787, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 31811 + }, + { + "epoch": 0.31812, + "grad_norm": 1.348389410701029, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 31812 + }, + { + "epoch": 0.31813, + "grad_norm": 1.6490895945601054, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 31813 + }, + { + "epoch": 0.31814, + "grad_norm": 1.3311029703205848, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31814 + }, + { + "epoch": 0.31815, + "grad_norm": 1.5622968072528058, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 31815 + }, + { + "epoch": 0.31816, + "grad_norm": 1.5753711358177118, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 31816 + }, + { + "epoch": 0.31817, + "grad_norm": 1.3392052104101906, + "learning_rate": 0.003, + "loss": 4.05, + "step": 31817 + }, + { + "epoch": 0.31818, + "grad_norm": 1.4738291787095488, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 31818 + }, + { + "epoch": 0.31819, + "grad_norm": 1.3662900231538577, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 31819 + }, + { + "epoch": 0.3182, + "grad_norm": 1.2583140298472315, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 31820 + }, + { + "epoch": 0.31821, + "grad_norm": 1.4142993521505354, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 31821 + }, + { + "epoch": 0.31822, + "grad_norm": 1.4491619513847154, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 31822 + }, + { + "epoch": 0.31823, + "grad_norm": 1.5837012881511323, + "learning_rate": 0.003, + "loss": 4.034, + "step": 31823 + }, + { + "epoch": 0.31824, + "grad_norm": 1.6931575688812326, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 31824 + }, + { + "epoch": 0.31825, + "grad_norm": 1.6801525866574791, + "learning_rate": 0.003, + "loss": 4.04, + "step": 31825 + }, + { + "epoch": 0.31826, + "grad_norm": 1.2857889113286467, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 31826 + }, + { + "epoch": 0.31827, + "grad_norm": 1.858510020521556, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 31827 + }, + { + "epoch": 0.31828, + "grad_norm": 1.4675310803247021, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 31828 + }, + { + "epoch": 0.31829, + "grad_norm": 1.5063222132838856, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 31829 + }, + { + "epoch": 0.3183, + "grad_norm": 1.5136580493986125, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 31830 + }, + { + "epoch": 0.31831, + "grad_norm": 1.5318221558003284, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 31831 + }, + { + "epoch": 0.31832, + "grad_norm": 1.0799223429677822, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31832 + }, + { + "epoch": 0.31833, + "grad_norm": 1.6664203656649588, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 31833 + }, + { + "epoch": 0.31834, + "grad_norm": 1.3391007342009509, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 31834 + }, + { + "epoch": 0.31835, + "grad_norm": 1.4553956670192492, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 31835 + }, + { + "epoch": 0.31836, + "grad_norm": 1.4098527795600397, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 31836 + }, + { + "epoch": 0.31837, + "grad_norm": 1.4606110690935745, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 31837 + }, + { + "epoch": 0.31838, + "grad_norm": 1.7341322099310938, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 31838 + }, + { + "epoch": 0.31839, + "grad_norm": 1.2845491400805984, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 31839 + }, + { + "epoch": 0.3184, + "grad_norm": 1.5964765117143742, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 31840 + }, + { + "epoch": 0.31841, + "grad_norm": 1.4361246665590661, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 31841 + }, + { + "epoch": 0.31842, + "grad_norm": 1.3711638044571055, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 31842 + }, + { + "epoch": 0.31843, + "grad_norm": 1.6856273234679255, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 31843 + }, + { + "epoch": 0.31844, + "grad_norm": 1.4721192278307957, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 31844 + }, + { + "epoch": 0.31845, + "grad_norm": 1.3997527899239923, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 31845 + }, + { + "epoch": 0.31846, + "grad_norm": 1.5238879628792736, + "learning_rate": 0.003, + "loss": 4.03, + "step": 31846 + }, + { + "epoch": 0.31847, + "grad_norm": 1.4601846443365505, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31847 + }, + { + "epoch": 0.31848, + "grad_norm": 1.4622840808993396, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 31848 + }, + { + "epoch": 0.31849, + "grad_norm": 1.7091653296549891, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 31849 + }, + { + "epoch": 0.3185, + "grad_norm": 1.289069273384199, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 31850 + }, + { + "epoch": 0.31851, + "grad_norm": 1.6054055825618165, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 31851 + }, + { + "epoch": 0.31852, + "grad_norm": 1.2509491470432217, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 31852 + }, + { + "epoch": 0.31853, + "grad_norm": 1.4479134961856392, + "learning_rate": 0.003, + "loss": 4.071, + "step": 31853 + }, + { + "epoch": 0.31854, + "grad_norm": 1.4117979429623115, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 31854 + }, + { + "epoch": 0.31855, + "grad_norm": 1.7300006661155014, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 31855 + }, + { + "epoch": 0.31856, + "grad_norm": 1.5786169295615862, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 31856 + }, + { + "epoch": 0.31857, + "grad_norm": 1.5601959837586092, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 31857 + }, + { + "epoch": 0.31858, + "grad_norm": 1.5479888239691282, + "learning_rate": 0.003, + "loss": 4.047, + "step": 31858 + }, + { + "epoch": 0.31859, + "grad_norm": 1.3653100556733282, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31859 + }, + { + "epoch": 0.3186, + "grad_norm": 1.4675225015550446, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 31860 + }, + { + "epoch": 0.31861, + "grad_norm": 1.4338709771139468, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31861 + }, + { + "epoch": 0.31862, + "grad_norm": 1.533059387439679, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 31862 + }, + { + "epoch": 0.31863, + "grad_norm": 1.2294925051902033, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 31863 + }, + { + "epoch": 0.31864, + "grad_norm": 1.468513114827876, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 31864 + }, + { + "epoch": 0.31865, + "grad_norm": 1.3108296947482279, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 31865 + }, + { + "epoch": 0.31866, + "grad_norm": 1.5960704455236856, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 31866 + }, + { + "epoch": 0.31867, + "grad_norm": 1.1230861114662838, + "learning_rate": 0.003, + "loss": 4.062, + "step": 31867 + }, + { + "epoch": 0.31868, + "grad_norm": 1.5969288531316588, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 31868 + }, + { + "epoch": 0.31869, + "grad_norm": 1.2643446375781586, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31869 + }, + { + "epoch": 0.3187, + "grad_norm": 1.8351292293717845, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31870 + }, + { + "epoch": 0.31871, + "grad_norm": 1.3156629229691177, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 31871 + }, + { + "epoch": 0.31872, + "grad_norm": 1.4968054191046412, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 31872 + }, + { + "epoch": 0.31873, + "grad_norm": 1.6487646877931457, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 31873 + }, + { + "epoch": 0.31874, + "grad_norm": 1.6160783010022066, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 31874 + }, + { + "epoch": 0.31875, + "grad_norm": 1.7701573352439268, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 31875 + }, + { + "epoch": 0.31876, + "grad_norm": 1.466852420959085, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 31876 + }, + { + "epoch": 0.31877, + "grad_norm": 1.4179746133439328, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 31877 + }, + { + "epoch": 0.31878, + "grad_norm": 1.4233248458534393, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 31878 + }, + { + "epoch": 0.31879, + "grad_norm": 1.5726526486037764, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31879 + }, + { + "epoch": 0.3188, + "grad_norm": 1.2984896444732879, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 31880 + }, + { + "epoch": 0.31881, + "grad_norm": 1.5175148727136152, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 31881 + }, + { + "epoch": 0.31882, + "grad_norm": 1.5337847661417547, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 31882 + }, + { + "epoch": 0.31883, + "grad_norm": 1.4094513229934373, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 31883 + }, + { + "epoch": 0.31884, + "grad_norm": 1.4930714689583517, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31884 + }, + { + "epoch": 0.31885, + "grad_norm": 1.4396656677062374, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 31885 + }, + { + "epoch": 0.31886, + "grad_norm": 1.518938494654217, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 31886 + }, + { + "epoch": 0.31887, + "grad_norm": 1.2329448624953399, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 31887 + }, + { + "epoch": 0.31888, + "grad_norm": 1.427380716348042, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 31888 + }, + { + "epoch": 0.31889, + "grad_norm": 1.558470030699737, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 31889 + }, + { + "epoch": 0.3189, + "grad_norm": 1.4776587989480938, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 31890 + }, + { + "epoch": 0.31891, + "grad_norm": 1.6590157379235793, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 31891 + }, + { + "epoch": 0.31892, + "grad_norm": 1.2280724255757494, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 31892 + }, + { + "epoch": 0.31893, + "grad_norm": 1.7325345411811117, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 31893 + }, + { + "epoch": 0.31894, + "grad_norm": 1.2721589756428837, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 31894 + }, + { + "epoch": 0.31895, + "grad_norm": 2.1376864798721233, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 31895 + }, + { + "epoch": 0.31896, + "grad_norm": 1.6319535125547142, + "learning_rate": 0.003, + "loss": 4.059, + "step": 31896 + }, + { + "epoch": 0.31897, + "grad_norm": 1.520407063848585, + "learning_rate": 0.003, + "loss": 4.068, + "step": 31897 + }, + { + "epoch": 0.31898, + "grad_norm": 1.329714077751962, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31898 + }, + { + "epoch": 0.31899, + "grad_norm": 1.5817661051988532, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 31899 + }, + { + "epoch": 0.319, + "grad_norm": 1.5908854814584021, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 31900 + }, + { + "epoch": 0.31901, + "grad_norm": 1.1824653203521736, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 31901 + }, + { + "epoch": 0.31902, + "grad_norm": 1.8185775777405886, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 31902 + }, + { + "epoch": 0.31903, + "grad_norm": 1.5048790961297893, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 31903 + }, + { + "epoch": 0.31904, + "grad_norm": 1.3437183393482994, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 31904 + }, + { + "epoch": 0.31905, + "grad_norm": 1.5182361551104722, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 31905 + }, + { + "epoch": 0.31906, + "grad_norm": 1.3015519541375307, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 31906 + }, + { + "epoch": 0.31907, + "grad_norm": 1.3978372841527584, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 31907 + }, + { + "epoch": 0.31908, + "grad_norm": 1.6022209290136313, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 31908 + }, + { + "epoch": 0.31909, + "grad_norm": 1.3280791480172878, + "learning_rate": 0.003, + "loss": 4.033, + "step": 31909 + }, + { + "epoch": 0.3191, + "grad_norm": 1.876187224815218, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 31910 + }, + { + "epoch": 0.31911, + "grad_norm": 1.3639959314475716, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31911 + }, + { + "epoch": 0.31912, + "grad_norm": 1.3816627956237906, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 31912 + }, + { + "epoch": 0.31913, + "grad_norm": 1.6588020893846738, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 31913 + }, + { + "epoch": 0.31914, + "grad_norm": 1.2793271034179972, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 31914 + }, + { + "epoch": 0.31915, + "grad_norm": 1.3636189040488675, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 31915 + }, + { + "epoch": 0.31916, + "grad_norm": 1.6826835976490602, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 31916 + }, + { + "epoch": 0.31917, + "grad_norm": 1.4770713758154852, + "learning_rate": 0.003, + "loss": 4.063, + "step": 31917 + }, + { + "epoch": 0.31918, + "grad_norm": 1.3285927429449969, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 31918 + }, + { + "epoch": 0.31919, + "grad_norm": 1.3753766477936504, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 31919 + }, + { + "epoch": 0.3192, + "grad_norm": 1.505595699472246, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 31920 + }, + { + "epoch": 0.31921, + "grad_norm": 1.5167929645254132, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31921 + }, + { + "epoch": 0.31922, + "grad_norm": 1.527115718447532, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31922 + }, + { + "epoch": 0.31923, + "grad_norm": 1.3858681178138468, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 31923 + }, + { + "epoch": 0.31924, + "grad_norm": 1.6060185101804125, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31924 + }, + { + "epoch": 0.31925, + "grad_norm": 1.2669725771559894, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 31925 + }, + { + "epoch": 0.31926, + "grad_norm": 1.5196099804997611, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 31926 + }, + { + "epoch": 0.31927, + "grad_norm": 1.4192988576536916, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 31927 + }, + { + "epoch": 0.31928, + "grad_norm": 1.3994589824047647, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 31928 + }, + { + "epoch": 0.31929, + "grad_norm": 1.5384087537935693, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 31929 + }, + { + "epoch": 0.3193, + "grad_norm": 1.580455999294843, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 31930 + }, + { + "epoch": 0.31931, + "grad_norm": 1.5634440976547488, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 31931 + }, + { + "epoch": 0.31932, + "grad_norm": 1.4072298654805104, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 31932 + }, + { + "epoch": 0.31933, + "grad_norm": 1.4400884932848113, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 31933 + }, + { + "epoch": 0.31934, + "grad_norm": 1.5138637101362744, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 31934 + }, + { + "epoch": 0.31935, + "grad_norm": 1.5376102035457666, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 31935 + }, + { + "epoch": 0.31936, + "grad_norm": 1.4440547121857812, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 31936 + }, + { + "epoch": 0.31937, + "grad_norm": 1.5733744672636276, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 31937 + }, + { + "epoch": 0.31938, + "grad_norm": 1.2204490441275677, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 31938 + }, + { + "epoch": 0.31939, + "grad_norm": 1.543944291618139, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 31939 + }, + { + "epoch": 0.3194, + "grad_norm": 1.1247717875638144, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 31940 + }, + { + "epoch": 0.31941, + "grad_norm": 1.613478910905346, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 31941 + }, + { + "epoch": 0.31942, + "grad_norm": 1.4799191385163346, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 31942 + }, + { + "epoch": 0.31943, + "grad_norm": 1.4258715190741718, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 31943 + }, + { + "epoch": 0.31944, + "grad_norm": 1.5397822299902695, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 31944 + }, + { + "epoch": 0.31945, + "grad_norm": 1.4715779010360672, + "learning_rate": 0.003, + "loss": 4.042, + "step": 31945 + }, + { + "epoch": 0.31946, + "grad_norm": 1.4293359401298469, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31946 + }, + { + "epoch": 0.31947, + "grad_norm": 1.4525278285894931, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 31947 + }, + { + "epoch": 0.31948, + "grad_norm": 1.3639997100531291, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 31948 + }, + { + "epoch": 0.31949, + "grad_norm": 1.4209939271561294, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 31949 + }, + { + "epoch": 0.3195, + "grad_norm": 1.5949020039521862, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31950 + }, + { + "epoch": 0.31951, + "grad_norm": 1.5757782043399333, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 31951 + }, + { + "epoch": 0.31952, + "grad_norm": 1.6444518922629952, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31952 + }, + { + "epoch": 0.31953, + "grad_norm": 1.6978841287599042, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 31953 + }, + { + "epoch": 0.31954, + "grad_norm": 1.2133387618335083, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 31954 + }, + { + "epoch": 0.31955, + "grad_norm": 1.4823129422055075, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 31955 + }, + { + "epoch": 0.31956, + "grad_norm": 1.557260147768072, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 31956 + }, + { + "epoch": 0.31957, + "grad_norm": 1.481798515808229, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 31957 + }, + { + "epoch": 0.31958, + "grad_norm": 1.5901018980792452, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 31958 + }, + { + "epoch": 0.31959, + "grad_norm": 1.2349660601504628, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 31959 + }, + { + "epoch": 0.3196, + "grad_norm": 1.5842514275769262, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 31960 + }, + { + "epoch": 0.31961, + "grad_norm": 1.3136636779726956, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 31961 + }, + { + "epoch": 0.31962, + "grad_norm": 1.4338181294470167, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 31962 + }, + { + "epoch": 0.31963, + "grad_norm": 1.3570869012566877, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 31963 + }, + { + "epoch": 0.31964, + "grad_norm": 1.3061343082971582, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 31964 + }, + { + "epoch": 0.31965, + "grad_norm": 1.5535375467958934, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 31965 + }, + { + "epoch": 0.31966, + "grad_norm": 1.3854348935264387, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 31966 + }, + { + "epoch": 0.31967, + "grad_norm": 1.4803812584611467, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 31967 + }, + { + "epoch": 0.31968, + "grad_norm": 1.5645311345174955, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 31968 + }, + { + "epoch": 0.31969, + "grad_norm": 1.8023624486059049, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 31969 + }, + { + "epoch": 0.3197, + "grad_norm": 1.6821478931930924, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31970 + }, + { + "epoch": 0.31971, + "grad_norm": 1.635832166195344, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 31971 + }, + { + "epoch": 0.31972, + "grad_norm": 1.346661089866376, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 31972 + }, + { + "epoch": 0.31973, + "grad_norm": 1.6883333068868642, + "learning_rate": 0.003, + "loss": 4.073, + "step": 31973 + }, + { + "epoch": 0.31974, + "grad_norm": 1.3300340798725265, + "learning_rate": 0.003, + "loss": 4.047, + "step": 31974 + }, + { + "epoch": 0.31975, + "grad_norm": 1.3503438759932134, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31975 + }, + { + "epoch": 0.31976, + "grad_norm": 1.5702791840982937, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 31976 + }, + { + "epoch": 0.31977, + "grad_norm": 1.3464208789349807, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 31977 + }, + { + "epoch": 0.31978, + "grad_norm": 1.6043746253211038, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 31978 + }, + { + "epoch": 0.31979, + "grad_norm": 1.1832751309849414, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 31979 + }, + { + "epoch": 0.3198, + "grad_norm": 1.6068871896343477, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 31980 + }, + { + "epoch": 0.31981, + "grad_norm": 1.287027772775131, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 31981 + }, + { + "epoch": 0.31982, + "grad_norm": 1.4246817786013481, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 31982 + }, + { + "epoch": 0.31983, + "grad_norm": 1.2720922939035573, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 31983 + }, + { + "epoch": 0.31984, + "grad_norm": 1.5788590104026279, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 31984 + }, + { + "epoch": 0.31985, + "grad_norm": 1.411864452509951, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 31985 + }, + { + "epoch": 0.31986, + "grad_norm": 1.6373681050607582, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 31986 + }, + { + "epoch": 0.31987, + "grad_norm": 1.5691199329464538, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 31987 + }, + { + "epoch": 0.31988, + "grad_norm": 1.8399671037562886, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 31988 + }, + { + "epoch": 0.31989, + "grad_norm": 1.4446802236975345, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 31989 + }, + { + "epoch": 0.3199, + "grad_norm": 1.2513879262918535, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 31990 + }, + { + "epoch": 0.31991, + "grad_norm": 2.214219000526846, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 31991 + }, + { + "epoch": 0.31992, + "grad_norm": 1.0355288776565839, + "learning_rate": 0.003, + "loss": 4.046, + "step": 31992 + }, + { + "epoch": 0.31993, + "grad_norm": 1.6739640463949133, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 31993 + }, + { + "epoch": 0.31994, + "grad_norm": 1.2938406675692866, + "learning_rate": 0.003, + "loss": 4.065, + "step": 31994 + }, + { + "epoch": 0.31995, + "grad_norm": 1.5247493681095148, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 31995 + }, + { + "epoch": 0.31996, + "grad_norm": 1.4224059924331698, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 31996 + }, + { + "epoch": 0.31997, + "grad_norm": 1.7887938805631087, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 31997 + }, + { + "epoch": 0.31998, + "grad_norm": 1.3997561462694024, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 31998 + }, + { + "epoch": 0.31999, + "grad_norm": 1.4820382888488528, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 31999 + }, + { + "epoch": 0.32, + "grad_norm": 1.5836285879698344, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 32000 + }, + { + "epoch": 0.32001, + "grad_norm": 1.5282941563826555, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32001 + }, + { + "epoch": 0.32002, + "grad_norm": 1.4974381581506382, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32002 + }, + { + "epoch": 0.32003, + "grad_norm": 1.4674635129044364, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 32003 + }, + { + "epoch": 0.32004, + "grad_norm": 1.5645148666350626, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 32004 + }, + { + "epoch": 0.32005, + "grad_norm": 1.2819002272897704, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 32005 + }, + { + "epoch": 0.32006, + "grad_norm": 1.5179328027232248, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 32006 + }, + { + "epoch": 0.32007, + "grad_norm": 1.185424970899993, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32007 + }, + { + "epoch": 0.32008, + "grad_norm": 1.4696946433188733, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 32008 + }, + { + "epoch": 0.32009, + "grad_norm": 1.3895947510143374, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 32009 + }, + { + "epoch": 0.3201, + "grad_norm": 1.6577761675866307, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 32010 + }, + { + "epoch": 0.32011, + "grad_norm": 1.153406936328362, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 32011 + }, + { + "epoch": 0.32012, + "grad_norm": 1.8611387128236323, + "learning_rate": 0.003, + "loss": 4.073, + "step": 32012 + }, + { + "epoch": 0.32013, + "grad_norm": 1.3741360621330312, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32013 + }, + { + "epoch": 0.32014, + "grad_norm": 1.8049866131378596, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 32014 + }, + { + "epoch": 0.32015, + "grad_norm": 1.4784636349384899, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32015 + }, + { + "epoch": 0.32016, + "grad_norm": 1.1880751695179668, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 32016 + }, + { + "epoch": 0.32017, + "grad_norm": 1.560473748213602, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 32017 + }, + { + "epoch": 0.32018, + "grad_norm": 1.188524403941292, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 32018 + }, + { + "epoch": 0.32019, + "grad_norm": 1.4289969696637592, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32019 + }, + { + "epoch": 0.3202, + "grad_norm": 1.554510953000664, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 32020 + }, + { + "epoch": 0.32021, + "grad_norm": 1.750537061676101, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 32021 + }, + { + "epoch": 0.32022, + "grad_norm": 1.2791305830647526, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32022 + }, + { + "epoch": 0.32023, + "grad_norm": 1.3564351570355864, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32023 + }, + { + "epoch": 0.32024, + "grad_norm": 1.5088956361784611, + "learning_rate": 0.003, + "loss": 4.025, + "step": 32024 + }, + { + "epoch": 0.32025, + "grad_norm": 1.3124269575491774, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32025 + }, + { + "epoch": 0.32026, + "grad_norm": 1.4725818506583208, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 32026 + }, + { + "epoch": 0.32027, + "grad_norm": 1.8268820641111694, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 32027 + }, + { + "epoch": 0.32028, + "grad_norm": 1.1236082279365893, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32028 + }, + { + "epoch": 0.32029, + "grad_norm": 1.6244880672860884, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32029 + }, + { + "epoch": 0.3203, + "grad_norm": 1.803310950339708, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 32030 + }, + { + "epoch": 0.32031, + "grad_norm": 1.3058148585603722, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 32031 + }, + { + "epoch": 0.32032, + "grad_norm": 1.4745831947109638, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 32032 + }, + { + "epoch": 0.32033, + "grad_norm": 1.4368886997736299, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 32033 + }, + { + "epoch": 0.32034, + "grad_norm": 1.2971819042950379, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 32034 + }, + { + "epoch": 0.32035, + "grad_norm": 1.5623693125214317, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 32035 + }, + { + "epoch": 0.32036, + "grad_norm": 1.4747007810054855, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32036 + }, + { + "epoch": 0.32037, + "grad_norm": 1.3898496878317739, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 32037 + }, + { + "epoch": 0.32038, + "grad_norm": 1.5324478770458312, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 32038 + }, + { + "epoch": 0.32039, + "grad_norm": 1.948344430703949, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32039 + }, + { + "epoch": 0.3204, + "grad_norm": 1.1969276846095496, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 32040 + }, + { + "epoch": 0.32041, + "grad_norm": 1.4109878068530766, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 32041 + }, + { + "epoch": 0.32042, + "grad_norm": 1.4973727003821224, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32042 + }, + { + "epoch": 0.32043, + "grad_norm": 1.6318030381217798, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 32043 + }, + { + "epoch": 0.32044, + "grad_norm": 1.4856867732980226, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 32044 + }, + { + "epoch": 0.32045, + "grad_norm": 1.6707752043467925, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 32045 + }, + { + "epoch": 0.32046, + "grad_norm": 1.3210350289077628, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 32046 + }, + { + "epoch": 0.32047, + "grad_norm": 1.1737335562022055, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 32047 + }, + { + "epoch": 0.32048, + "grad_norm": 1.5953041651785693, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32048 + }, + { + "epoch": 0.32049, + "grad_norm": 1.5816895561135405, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32049 + }, + { + "epoch": 0.3205, + "grad_norm": 1.6560102953572435, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 32050 + }, + { + "epoch": 0.32051, + "grad_norm": 1.3737508630416815, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32051 + }, + { + "epoch": 0.32052, + "grad_norm": 1.5047015977488716, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 32052 + }, + { + "epoch": 0.32053, + "grad_norm": 1.4699720099805766, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 32053 + }, + { + "epoch": 0.32054, + "grad_norm": 1.5188718100527396, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 32054 + }, + { + "epoch": 0.32055, + "grad_norm": 1.1965917116321083, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 32055 + }, + { + "epoch": 0.32056, + "grad_norm": 1.4714060184869555, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 32056 + }, + { + "epoch": 0.32057, + "grad_norm": 1.4285691596331906, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 32057 + }, + { + "epoch": 0.32058, + "grad_norm": 1.2945243193963665, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 32058 + }, + { + "epoch": 0.32059, + "grad_norm": 1.4950550371816114, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 32059 + }, + { + "epoch": 0.3206, + "grad_norm": 1.4244083936082486, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32060 + }, + { + "epoch": 0.32061, + "grad_norm": 1.7182258682223286, + "learning_rate": 0.003, + "loss": 4.047, + "step": 32061 + }, + { + "epoch": 0.32062, + "grad_norm": 1.296234522977695, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32062 + }, + { + "epoch": 0.32063, + "grad_norm": 1.6615714513885913, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32063 + }, + { + "epoch": 0.32064, + "grad_norm": 1.433662051787677, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 32064 + }, + { + "epoch": 0.32065, + "grad_norm": 1.5265378806670789, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 32065 + }, + { + "epoch": 0.32066, + "grad_norm": 1.5975764703032629, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32066 + }, + { + "epoch": 0.32067, + "grad_norm": 1.641172200487926, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32067 + }, + { + "epoch": 0.32068, + "grad_norm": 1.3858330816027087, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 32068 + }, + { + "epoch": 0.32069, + "grad_norm": 1.8129596987917531, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32069 + }, + { + "epoch": 0.3207, + "grad_norm": 1.0964532148071826, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32070 + }, + { + "epoch": 0.32071, + "grad_norm": 1.7010046892186943, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32071 + }, + { + "epoch": 0.32072, + "grad_norm": 1.343326862375331, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 32072 + }, + { + "epoch": 0.32073, + "grad_norm": 1.7650403429469055, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 32073 + }, + { + "epoch": 0.32074, + "grad_norm": 1.4068677794005118, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 32074 + }, + { + "epoch": 0.32075, + "grad_norm": 1.5458646948264279, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 32075 + }, + { + "epoch": 0.32076, + "grad_norm": 1.5814007796535081, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 32076 + }, + { + "epoch": 0.32077, + "grad_norm": 1.406636451698401, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32077 + }, + { + "epoch": 0.32078, + "grad_norm": 1.5886513722450508, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 32078 + }, + { + "epoch": 0.32079, + "grad_norm": 1.491082745643525, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32079 + }, + { + "epoch": 0.3208, + "grad_norm": 1.3340203353613214, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 32080 + }, + { + "epoch": 0.32081, + "grad_norm": 1.7744961749432202, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32081 + }, + { + "epoch": 0.32082, + "grad_norm": 1.2057334414491154, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32082 + }, + { + "epoch": 0.32083, + "grad_norm": 1.5564089470820948, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32083 + }, + { + "epoch": 0.32084, + "grad_norm": 1.1582775923301996, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 32084 + }, + { + "epoch": 0.32085, + "grad_norm": 1.8301500675747888, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 32085 + }, + { + "epoch": 0.32086, + "grad_norm": 1.3193500664472617, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 32086 + }, + { + "epoch": 0.32087, + "grad_norm": 1.6890733141047112, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 32087 + }, + { + "epoch": 0.32088, + "grad_norm": 1.5425006990341479, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 32088 + }, + { + "epoch": 0.32089, + "grad_norm": 1.6717910478690854, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 32089 + }, + { + "epoch": 0.3209, + "grad_norm": 1.1981045494074603, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 32090 + }, + { + "epoch": 0.32091, + "grad_norm": 1.5733541031160252, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 32091 + }, + { + "epoch": 0.32092, + "grad_norm": 1.5412299096067352, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 32092 + }, + { + "epoch": 0.32093, + "grad_norm": 1.5736209720222667, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 32093 + }, + { + "epoch": 0.32094, + "grad_norm": 1.5826690853159224, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 32094 + }, + { + "epoch": 0.32095, + "grad_norm": 1.4209779919091048, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 32095 + }, + { + "epoch": 0.32096, + "grad_norm": 1.3673066974840575, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 32096 + }, + { + "epoch": 0.32097, + "grad_norm": 1.388983424677327, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 32097 + }, + { + "epoch": 0.32098, + "grad_norm": 1.3559777326041471, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32098 + }, + { + "epoch": 0.32099, + "grad_norm": 1.64215209166924, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 32099 + }, + { + "epoch": 0.321, + "grad_norm": 1.3707842652867606, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32100 + }, + { + "epoch": 0.32101, + "grad_norm": 1.3855632181424542, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32101 + }, + { + "epoch": 0.32102, + "grad_norm": 1.245955598952426, + "learning_rate": 0.003, + "loss": 3.999, + "step": 32102 + }, + { + "epoch": 0.32103, + "grad_norm": 1.5840030595429195, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 32103 + }, + { + "epoch": 0.32104, + "grad_norm": 1.3929412943864605, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 32104 + }, + { + "epoch": 0.32105, + "grad_norm": 1.1778599042042757, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 32105 + }, + { + "epoch": 0.32106, + "grad_norm": 1.6160909765230735, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32106 + }, + { + "epoch": 0.32107, + "grad_norm": 1.3461141818160884, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 32107 + }, + { + "epoch": 0.32108, + "grad_norm": 1.6356159948700582, + "learning_rate": 0.003, + "loss": 4.052, + "step": 32108 + }, + { + "epoch": 0.32109, + "grad_norm": 1.4809152215041848, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 32109 + }, + { + "epoch": 0.3211, + "grad_norm": 1.7827643822471866, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 32110 + }, + { + "epoch": 0.32111, + "grad_norm": 1.405747560731545, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32111 + }, + { + "epoch": 0.32112, + "grad_norm": 1.4144837636941492, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 32112 + }, + { + "epoch": 0.32113, + "grad_norm": 1.3483250307944907, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 32113 + }, + { + "epoch": 0.32114, + "grad_norm": 1.5174213681791047, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 32114 + }, + { + "epoch": 0.32115, + "grad_norm": 1.20682870946543, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32115 + }, + { + "epoch": 0.32116, + "grad_norm": 1.4524347343249175, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32116 + }, + { + "epoch": 0.32117, + "grad_norm": 1.656029887570055, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 32117 + }, + { + "epoch": 0.32118, + "grad_norm": 1.444322581599768, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32118 + }, + { + "epoch": 0.32119, + "grad_norm": 1.3429256368110487, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 32119 + }, + { + "epoch": 0.3212, + "grad_norm": 1.5929659353894527, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 32120 + }, + { + "epoch": 0.32121, + "grad_norm": 1.434509956733265, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32121 + }, + { + "epoch": 0.32122, + "grad_norm": 1.6791487558868734, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 32122 + }, + { + "epoch": 0.32123, + "grad_norm": 1.585518626957845, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 32123 + }, + { + "epoch": 0.32124, + "grad_norm": 1.8962513206807243, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 32124 + }, + { + "epoch": 0.32125, + "grad_norm": 1.228012281382326, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 32125 + }, + { + "epoch": 0.32126, + "grad_norm": 1.5169006128673224, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 32126 + }, + { + "epoch": 0.32127, + "grad_norm": 1.828774312630365, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 32127 + }, + { + "epoch": 0.32128, + "grad_norm": 1.2794134550910925, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 32128 + }, + { + "epoch": 0.32129, + "grad_norm": 1.7122155480106287, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 32129 + }, + { + "epoch": 0.3213, + "grad_norm": 1.2278158739861813, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32130 + }, + { + "epoch": 0.32131, + "grad_norm": 1.391308058371101, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32131 + }, + { + "epoch": 0.32132, + "grad_norm": 1.387320537299021, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32132 + }, + { + "epoch": 0.32133, + "grad_norm": 1.48318841223393, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 32133 + }, + { + "epoch": 0.32134, + "grad_norm": 1.774480354195924, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32134 + }, + { + "epoch": 0.32135, + "grad_norm": 1.443202688202521, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 32135 + }, + { + "epoch": 0.32136, + "grad_norm": 1.98884043918733, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32136 + }, + { + "epoch": 0.32137, + "grad_norm": 1.3143511357528734, + "learning_rate": 0.003, + "loss": 4.039, + "step": 32137 + }, + { + "epoch": 0.32138, + "grad_norm": 1.4070268298332371, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 32138 + }, + { + "epoch": 0.32139, + "grad_norm": 1.5789559765115961, + "learning_rate": 0.003, + "loss": 4.024, + "step": 32139 + }, + { + "epoch": 0.3214, + "grad_norm": 1.3898590505009467, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32140 + }, + { + "epoch": 0.32141, + "grad_norm": 1.370151464807648, + "learning_rate": 0.003, + "loss": 4.065, + "step": 32141 + }, + { + "epoch": 0.32142, + "grad_norm": 1.6359592649277026, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 32142 + }, + { + "epoch": 0.32143, + "grad_norm": 1.2795884806671756, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32143 + }, + { + "epoch": 0.32144, + "grad_norm": 1.2465783941738915, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 32144 + }, + { + "epoch": 0.32145, + "grad_norm": 1.5441429564936722, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32145 + }, + { + "epoch": 0.32146, + "grad_norm": 1.2131391523823132, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 32146 + }, + { + "epoch": 0.32147, + "grad_norm": 1.889283923919957, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 32147 + }, + { + "epoch": 0.32148, + "grad_norm": 1.590760431712892, + "learning_rate": 0.003, + "loss": 4.01, + "step": 32148 + }, + { + "epoch": 0.32149, + "grad_norm": 1.33139243955704, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32149 + }, + { + "epoch": 0.3215, + "grad_norm": 1.1988368478456475, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 32150 + }, + { + "epoch": 0.32151, + "grad_norm": 1.4218998438899728, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 32151 + }, + { + "epoch": 0.32152, + "grad_norm": 1.4313511314590175, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 32152 + }, + { + "epoch": 0.32153, + "grad_norm": 1.3691458937574699, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 32153 + }, + { + "epoch": 0.32154, + "grad_norm": 1.6605629307899925, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 32154 + }, + { + "epoch": 0.32155, + "grad_norm": 1.4280478781776618, + "learning_rate": 0.003, + "loss": 4.061, + "step": 32155 + }, + { + "epoch": 0.32156, + "grad_norm": 1.5090126001354507, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 32156 + }, + { + "epoch": 0.32157, + "grad_norm": 1.357720018117069, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 32157 + }, + { + "epoch": 0.32158, + "grad_norm": 1.5122017580370617, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 32158 + }, + { + "epoch": 0.32159, + "grad_norm": 1.268666491214585, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 32159 + }, + { + "epoch": 0.3216, + "grad_norm": 1.3339793064983008, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 32160 + }, + { + "epoch": 0.32161, + "grad_norm": 1.7195684911801825, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 32161 + }, + { + "epoch": 0.32162, + "grad_norm": 1.8495485959665732, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 32162 + }, + { + "epoch": 0.32163, + "grad_norm": 1.4926765033763114, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 32163 + }, + { + "epoch": 0.32164, + "grad_norm": 1.6048774309524887, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 32164 + }, + { + "epoch": 0.32165, + "grad_norm": 1.2840012040498014, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 32165 + }, + { + "epoch": 0.32166, + "grad_norm": 1.6506368298398704, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32166 + }, + { + "epoch": 0.32167, + "grad_norm": 1.4645166467668043, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 32167 + }, + { + "epoch": 0.32168, + "grad_norm": 1.5336660727208844, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32168 + }, + { + "epoch": 0.32169, + "grad_norm": 1.3702621185257997, + "learning_rate": 0.003, + "loss": 4.074, + "step": 32169 + }, + { + "epoch": 0.3217, + "grad_norm": 1.566513829262213, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 32170 + }, + { + "epoch": 0.32171, + "grad_norm": 1.3492588138257238, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 32171 + }, + { + "epoch": 0.32172, + "grad_norm": 1.3938190668388286, + "learning_rate": 0.003, + "loss": 4.065, + "step": 32172 + }, + { + "epoch": 0.32173, + "grad_norm": 1.5023631853727164, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 32173 + }, + { + "epoch": 0.32174, + "grad_norm": 1.5280084888117011, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32174 + }, + { + "epoch": 0.32175, + "grad_norm": 1.5879192084641522, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32175 + }, + { + "epoch": 0.32176, + "grad_norm": 1.414094611617056, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32176 + }, + { + "epoch": 0.32177, + "grad_norm": 1.5071182768930012, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 32177 + }, + { + "epoch": 0.32178, + "grad_norm": 1.329119436617175, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32178 + }, + { + "epoch": 0.32179, + "grad_norm": 1.3430664321944068, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 32179 + }, + { + "epoch": 0.3218, + "grad_norm": 1.4287505553823692, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32180 + }, + { + "epoch": 0.32181, + "grad_norm": 1.5779990917385391, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 32181 + }, + { + "epoch": 0.32182, + "grad_norm": 1.624532570090565, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 32182 + }, + { + "epoch": 0.32183, + "grad_norm": 1.561975700009567, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32183 + }, + { + "epoch": 0.32184, + "grad_norm": 1.6268485907666457, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 32184 + }, + { + "epoch": 0.32185, + "grad_norm": 1.3385218050059655, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 32185 + }, + { + "epoch": 0.32186, + "grad_norm": 1.6920145042465322, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 32186 + }, + { + "epoch": 0.32187, + "grad_norm": 1.6002911160517936, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 32187 + }, + { + "epoch": 0.32188, + "grad_norm": 1.3937410326844963, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 32188 + }, + { + "epoch": 0.32189, + "grad_norm": 1.5906465961866116, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 32189 + }, + { + "epoch": 0.3219, + "grad_norm": 1.3176509684436224, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 32190 + }, + { + "epoch": 0.32191, + "grad_norm": 1.5527899932640323, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 32191 + }, + { + "epoch": 0.32192, + "grad_norm": 1.4001360699222865, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 32192 + }, + { + "epoch": 0.32193, + "grad_norm": 1.6392043946912074, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32193 + }, + { + "epoch": 0.32194, + "grad_norm": 1.4632593831376532, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 32194 + }, + { + "epoch": 0.32195, + "grad_norm": 1.6304604034548953, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 32195 + }, + { + "epoch": 0.32196, + "grad_norm": 1.5184863178317196, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 32196 + }, + { + "epoch": 0.32197, + "grad_norm": 1.3925052361659997, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 32197 + }, + { + "epoch": 0.32198, + "grad_norm": 1.289707924977062, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 32198 + }, + { + "epoch": 0.32199, + "grad_norm": 1.3937129908511636, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 32199 + }, + { + "epoch": 0.322, + "grad_norm": 1.2219752678684763, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32200 + }, + { + "epoch": 0.32201, + "grad_norm": 1.766259672408021, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32201 + }, + { + "epoch": 0.32202, + "grad_norm": 1.6003253838313085, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 32202 + }, + { + "epoch": 0.32203, + "grad_norm": 1.3912500415400024, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 32203 + }, + { + "epoch": 0.32204, + "grad_norm": 1.718731146774283, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 32204 + }, + { + "epoch": 0.32205, + "grad_norm": 1.4059703798257643, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 32205 + }, + { + "epoch": 0.32206, + "grad_norm": 1.7910518466773744, + "learning_rate": 0.003, + "loss": 4.066, + "step": 32206 + }, + { + "epoch": 0.32207, + "grad_norm": 1.09326412648228, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 32207 + }, + { + "epoch": 0.32208, + "grad_norm": 1.6586232917101307, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 32208 + }, + { + "epoch": 0.32209, + "grad_norm": 1.4450338425177462, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32209 + }, + { + "epoch": 0.3221, + "grad_norm": 1.4456768274747533, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 32210 + }, + { + "epoch": 0.32211, + "grad_norm": 1.4433999074849346, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 32211 + }, + { + "epoch": 0.32212, + "grad_norm": 1.465251362118771, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 32212 + }, + { + "epoch": 0.32213, + "grad_norm": 1.6987278645684296, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 32213 + }, + { + "epoch": 0.32214, + "grad_norm": 1.4864940695773272, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32214 + }, + { + "epoch": 0.32215, + "grad_norm": 1.3991110878743251, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 32215 + }, + { + "epoch": 0.32216, + "grad_norm": 1.6551957484234174, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32216 + }, + { + "epoch": 0.32217, + "grad_norm": 1.2498885236133255, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 32217 + }, + { + "epoch": 0.32218, + "grad_norm": 1.4547937096722707, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32218 + }, + { + "epoch": 0.32219, + "grad_norm": 1.3155194528131138, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 32219 + }, + { + "epoch": 0.3222, + "grad_norm": 1.8454976518254553, + "learning_rate": 0.003, + "loss": 4.031, + "step": 32220 + }, + { + "epoch": 0.32221, + "grad_norm": 1.2214753976003754, + "learning_rate": 0.003, + "loss": 4.031, + "step": 32221 + }, + { + "epoch": 0.32222, + "grad_norm": 1.486833306958884, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32222 + }, + { + "epoch": 0.32223, + "grad_norm": 1.8594979973150756, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 32223 + }, + { + "epoch": 0.32224, + "grad_norm": 1.291313942906528, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 32224 + }, + { + "epoch": 0.32225, + "grad_norm": 1.5238912614822828, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32225 + }, + { + "epoch": 0.32226, + "grad_norm": 1.6130097658760338, + "learning_rate": 0.003, + "loss": 4.027, + "step": 32226 + }, + { + "epoch": 0.32227, + "grad_norm": 1.3171116449577989, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32227 + }, + { + "epoch": 0.32228, + "grad_norm": 1.6493412897821518, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32228 + }, + { + "epoch": 0.32229, + "grad_norm": 1.3101875049531604, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32229 + }, + { + "epoch": 0.3223, + "grad_norm": 1.658807253729349, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32230 + }, + { + "epoch": 0.32231, + "grad_norm": 1.3333983505762605, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 32231 + }, + { + "epoch": 0.32232, + "grad_norm": 1.5794896320679537, + "learning_rate": 0.003, + "loss": 4.064, + "step": 32232 + }, + { + "epoch": 0.32233, + "grad_norm": 1.3392110483256419, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32233 + }, + { + "epoch": 0.32234, + "grad_norm": 2.1082066319517216, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 32234 + }, + { + "epoch": 0.32235, + "grad_norm": 1.4244236772865504, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 32235 + }, + { + "epoch": 0.32236, + "grad_norm": 1.539348391150737, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 32236 + }, + { + "epoch": 0.32237, + "grad_norm": 1.6423124122413637, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 32237 + }, + { + "epoch": 0.32238, + "grad_norm": 1.3476633677999001, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 32238 + }, + { + "epoch": 0.32239, + "grad_norm": 1.423238020867462, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 32239 + }, + { + "epoch": 0.3224, + "grad_norm": 1.4662074855163192, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 32240 + }, + { + "epoch": 0.32241, + "grad_norm": 1.4595481113289337, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 32241 + }, + { + "epoch": 0.32242, + "grad_norm": 1.458257114316648, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32242 + }, + { + "epoch": 0.32243, + "grad_norm": 1.3486951050632112, + "learning_rate": 0.003, + "loss": 4.069, + "step": 32243 + }, + { + "epoch": 0.32244, + "grad_norm": 1.6474742318656428, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 32244 + }, + { + "epoch": 0.32245, + "grad_norm": 1.5481694511455637, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 32245 + }, + { + "epoch": 0.32246, + "grad_norm": 1.5450131757713705, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 32246 + }, + { + "epoch": 0.32247, + "grad_norm": 1.6188660741339147, + "learning_rate": 0.003, + "loss": 4.05, + "step": 32247 + }, + { + "epoch": 0.32248, + "grad_norm": 1.3179869679144147, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32248 + }, + { + "epoch": 0.32249, + "grad_norm": 1.1247315952184165, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32249 + }, + { + "epoch": 0.3225, + "grad_norm": 1.6909743718118397, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 32250 + }, + { + "epoch": 0.32251, + "grad_norm": 1.3858562429481538, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32251 + }, + { + "epoch": 0.32252, + "grad_norm": 1.754104771253988, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 32252 + }, + { + "epoch": 0.32253, + "grad_norm": 1.4718738399641729, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 32253 + }, + { + "epoch": 0.32254, + "grad_norm": 1.1356504597661765, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32254 + }, + { + "epoch": 0.32255, + "grad_norm": 1.46843104404379, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 32255 + }, + { + "epoch": 0.32256, + "grad_norm": 1.3540035250208515, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 32256 + }, + { + "epoch": 0.32257, + "grad_norm": 1.3307411813509071, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 32257 + }, + { + "epoch": 0.32258, + "grad_norm": 1.4549348221688239, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32258 + }, + { + "epoch": 0.32259, + "grad_norm": 1.6835772950278094, + "learning_rate": 0.003, + "loss": 4.024, + "step": 32259 + }, + { + "epoch": 0.3226, + "grad_norm": 1.3061677034311656, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32260 + }, + { + "epoch": 0.32261, + "grad_norm": 1.6139264832609388, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 32261 + }, + { + "epoch": 0.32262, + "grad_norm": 1.1564870866394479, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 32262 + }, + { + "epoch": 0.32263, + "grad_norm": 1.8761433206199505, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 32263 + }, + { + "epoch": 0.32264, + "grad_norm": 1.2278232308174533, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 32264 + }, + { + "epoch": 0.32265, + "grad_norm": 1.610742059501103, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32265 + }, + { + "epoch": 0.32266, + "grad_norm": 1.4278723951980588, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 32266 + }, + { + "epoch": 0.32267, + "grad_norm": 1.5796712967495243, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 32267 + }, + { + "epoch": 0.32268, + "grad_norm": 1.7516217647691397, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 32268 + }, + { + "epoch": 0.32269, + "grad_norm": 1.3857136675779247, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 32269 + }, + { + "epoch": 0.3227, + "grad_norm": 1.553246472743542, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 32270 + }, + { + "epoch": 0.32271, + "grad_norm": 1.1249578584080933, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32271 + }, + { + "epoch": 0.32272, + "grad_norm": 1.6686795201600184, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32272 + }, + { + "epoch": 0.32273, + "grad_norm": 1.2286428014732473, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32273 + }, + { + "epoch": 0.32274, + "grad_norm": 1.631985082734547, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 32274 + }, + { + "epoch": 0.32275, + "grad_norm": 1.0818282572934834, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 32275 + }, + { + "epoch": 0.32276, + "grad_norm": 1.6322960776400284, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 32276 + }, + { + "epoch": 0.32277, + "grad_norm": 1.1493376918746319, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 32277 + }, + { + "epoch": 0.32278, + "grad_norm": 1.7767718331751106, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32278 + }, + { + "epoch": 0.32279, + "grad_norm": 1.7063599956758677, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 32279 + }, + { + "epoch": 0.3228, + "grad_norm": 1.4342123499724317, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32280 + }, + { + "epoch": 0.32281, + "grad_norm": 1.5812541124734047, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 32281 + }, + { + "epoch": 0.32282, + "grad_norm": 1.4978177243447117, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 32282 + }, + { + "epoch": 0.32283, + "grad_norm": 1.4349071909695155, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 32283 + }, + { + "epoch": 0.32284, + "grad_norm": 1.5741152516974182, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32284 + }, + { + "epoch": 0.32285, + "grad_norm": 1.38104500447656, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32285 + }, + { + "epoch": 0.32286, + "grad_norm": 1.603885852583158, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32286 + }, + { + "epoch": 0.32287, + "grad_norm": 1.441367986264136, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 32287 + }, + { + "epoch": 0.32288, + "grad_norm": 1.2344115580326367, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 32288 + }, + { + "epoch": 0.32289, + "grad_norm": 1.4498566473653145, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 32289 + }, + { + "epoch": 0.3229, + "grad_norm": 1.5354081738870407, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32290 + }, + { + "epoch": 0.32291, + "grad_norm": 1.483093290657086, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32291 + }, + { + "epoch": 0.32292, + "grad_norm": 1.2946014649791924, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32292 + }, + { + "epoch": 0.32293, + "grad_norm": 1.4459196299154828, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32293 + }, + { + "epoch": 0.32294, + "grad_norm": 1.347287518759165, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32294 + }, + { + "epoch": 0.32295, + "grad_norm": 1.5866010190283175, + "learning_rate": 0.003, + "loss": 4.054, + "step": 32295 + }, + { + "epoch": 0.32296, + "grad_norm": 1.2968618213678256, + "learning_rate": 0.003, + "loss": 4.05, + "step": 32296 + }, + { + "epoch": 0.32297, + "grad_norm": 1.6677000467738685, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 32297 + }, + { + "epoch": 0.32298, + "grad_norm": 1.0034214571592814, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32298 + }, + { + "epoch": 0.32299, + "grad_norm": 1.753171042152029, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 32299 + }, + { + "epoch": 0.323, + "grad_norm": 1.5373572536018074, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 32300 + }, + { + "epoch": 0.32301, + "grad_norm": 1.4940783885617772, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 32301 + }, + { + "epoch": 0.32302, + "grad_norm": 1.5259093378631277, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32302 + }, + { + "epoch": 0.32303, + "grad_norm": 1.5378876284796381, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 32303 + }, + { + "epoch": 0.32304, + "grad_norm": 1.3019056022495643, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32304 + }, + { + "epoch": 0.32305, + "grad_norm": 1.58836521611446, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 32305 + }, + { + "epoch": 0.32306, + "grad_norm": 1.5916087387122482, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 32306 + }, + { + "epoch": 0.32307, + "grad_norm": 1.2429180952499677, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32307 + }, + { + "epoch": 0.32308, + "grad_norm": 1.616312162295466, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32308 + }, + { + "epoch": 0.32309, + "grad_norm": 1.4292879502901632, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 32309 + }, + { + "epoch": 0.3231, + "grad_norm": 1.4934822476611527, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 32310 + }, + { + "epoch": 0.32311, + "grad_norm": 1.1537118537858868, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 32311 + }, + { + "epoch": 0.32312, + "grad_norm": 1.7731427168866039, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 32312 + }, + { + "epoch": 0.32313, + "grad_norm": 1.357744928853238, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 32313 + }, + { + "epoch": 0.32314, + "grad_norm": 1.8523181954722323, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 32314 + }, + { + "epoch": 0.32315, + "grad_norm": 1.2755082019878845, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 32315 + }, + { + "epoch": 0.32316, + "grad_norm": 1.6347586517010861, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 32316 + }, + { + "epoch": 0.32317, + "grad_norm": 1.933982297642474, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 32317 + }, + { + "epoch": 0.32318, + "grad_norm": 1.5732216758311663, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 32318 + }, + { + "epoch": 0.32319, + "grad_norm": 1.4312751428509172, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32319 + }, + { + "epoch": 0.3232, + "grad_norm": 1.3918345391827995, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 32320 + }, + { + "epoch": 0.32321, + "grad_norm": 1.678808172742936, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 32321 + }, + { + "epoch": 0.32322, + "grad_norm": 1.3256579167060192, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 32322 + }, + { + "epoch": 0.32323, + "grad_norm": 1.4840666492982721, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32323 + }, + { + "epoch": 0.32324, + "grad_norm": 1.4479804920677455, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 32324 + }, + { + "epoch": 0.32325, + "grad_norm": 1.2629902403550513, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 32325 + }, + { + "epoch": 0.32326, + "grad_norm": 1.4907373152273466, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32326 + }, + { + "epoch": 0.32327, + "grad_norm": 1.3848681267254566, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 32327 + }, + { + "epoch": 0.32328, + "grad_norm": 1.5722545698019075, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 32328 + }, + { + "epoch": 0.32329, + "grad_norm": 1.4403522759973797, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 32329 + }, + { + "epoch": 0.3233, + "grad_norm": 1.4098154619704895, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 32330 + }, + { + "epoch": 0.32331, + "grad_norm": 1.5384714415376088, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32331 + }, + { + "epoch": 0.32332, + "grad_norm": 1.4722174221648154, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 32332 + }, + { + "epoch": 0.32333, + "grad_norm": 1.5167082877002513, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 32333 + }, + { + "epoch": 0.32334, + "grad_norm": 1.2507255484800928, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 32334 + }, + { + "epoch": 0.32335, + "grad_norm": 1.5418775512524516, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32335 + }, + { + "epoch": 0.32336, + "grad_norm": 1.4619334827773607, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 32336 + }, + { + "epoch": 0.32337, + "grad_norm": 1.6197793076532248, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 32337 + }, + { + "epoch": 0.32338, + "grad_norm": 1.7496464621435628, + "learning_rate": 0.003, + "loss": 4.052, + "step": 32338 + }, + { + "epoch": 0.32339, + "grad_norm": 1.4303793994682428, + "learning_rate": 0.003, + "loss": 4.017, + "step": 32339 + }, + { + "epoch": 0.3234, + "grad_norm": 1.3593162369748, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32340 + }, + { + "epoch": 0.32341, + "grad_norm": 1.7986181695284, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 32341 + }, + { + "epoch": 0.32342, + "grad_norm": 1.2503838848341042, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32342 + }, + { + "epoch": 0.32343, + "grad_norm": 1.5404159510289266, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 32343 + }, + { + "epoch": 0.32344, + "grad_norm": 1.2794452902939883, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32344 + }, + { + "epoch": 0.32345, + "grad_norm": 1.5674903570888645, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32345 + }, + { + "epoch": 0.32346, + "grad_norm": 1.63353755286058, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 32346 + }, + { + "epoch": 0.32347, + "grad_norm": 1.911111724932433, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 32347 + }, + { + "epoch": 0.32348, + "grad_norm": 1.1426192349114133, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 32348 + }, + { + "epoch": 0.32349, + "grad_norm": 1.8006057928137533, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32349 + }, + { + "epoch": 0.3235, + "grad_norm": 1.126602834883946, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 32350 + }, + { + "epoch": 0.32351, + "grad_norm": 1.627346832379279, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 32351 + }, + { + "epoch": 0.32352, + "grad_norm": 1.3242503181925047, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32352 + }, + { + "epoch": 0.32353, + "grad_norm": 1.4053636175797206, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32353 + }, + { + "epoch": 0.32354, + "grad_norm": 1.5021441563218472, + "learning_rate": 0.003, + "loss": 4.061, + "step": 32354 + }, + { + "epoch": 0.32355, + "grad_norm": 1.4297027006771168, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 32355 + }, + { + "epoch": 0.32356, + "grad_norm": 1.427099377175084, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 32356 + }, + { + "epoch": 0.32357, + "grad_norm": 1.222413930082904, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 32357 + }, + { + "epoch": 0.32358, + "grad_norm": 1.7997910741074534, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 32358 + }, + { + "epoch": 0.32359, + "grad_norm": 1.3704487931404616, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 32359 + }, + { + "epoch": 0.3236, + "grad_norm": 1.5433580598466552, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32360 + }, + { + "epoch": 0.32361, + "grad_norm": 1.4195388710932817, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32361 + }, + { + "epoch": 0.32362, + "grad_norm": 1.841739866616879, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 32362 + }, + { + "epoch": 0.32363, + "grad_norm": 1.2847811553057984, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 32363 + }, + { + "epoch": 0.32364, + "grad_norm": 1.5837087875908142, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 32364 + }, + { + "epoch": 0.32365, + "grad_norm": 1.3408299420627248, + "learning_rate": 0.003, + "loss": 4.025, + "step": 32365 + }, + { + "epoch": 0.32366, + "grad_norm": 1.486540168451555, + "learning_rate": 0.003, + "loss": 4.07, + "step": 32366 + }, + { + "epoch": 0.32367, + "grad_norm": 1.731938718611717, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 32367 + }, + { + "epoch": 0.32368, + "grad_norm": 1.0919272650439067, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 32368 + }, + { + "epoch": 0.32369, + "grad_norm": 1.6491608077209856, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 32369 + }, + { + "epoch": 0.3237, + "grad_norm": 1.4232380938410627, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32370 + }, + { + "epoch": 0.32371, + "grad_norm": 1.8476909884065993, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 32371 + }, + { + "epoch": 0.32372, + "grad_norm": 1.368334562028884, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 32372 + }, + { + "epoch": 0.32373, + "grad_norm": 1.4727390142769152, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 32373 + }, + { + "epoch": 0.32374, + "grad_norm": 1.5762034980137272, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 32374 + }, + { + "epoch": 0.32375, + "grad_norm": 1.6583268875138713, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32375 + }, + { + "epoch": 0.32376, + "grad_norm": 1.4556592158997044, + "learning_rate": 0.003, + "loss": 4.023, + "step": 32376 + }, + { + "epoch": 0.32377, + "grad_norm": 1.2947648193298185, + "learning_rate": 0.003, + "loss": 4.066, + "step": 32377 + }, + { + "epoch": 0.32378, + "grad_norm": 1.6940336097354287, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32378 + }, + { + "epoch": 0.32379, + "grad_norm": 1.2041452651990112, + "learning_rate": 0.003, + "loss": 4.029, + "step": 32379 + }, + { + "epoch": 0.3238, + "grad_norm": 1.522007534663757, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 32380 + }, + { + "epoch": 0.32381, + "grad_norm": 1.4304598210554809, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 32381 + }, + { + "epoch": 0.32382, + "grad_norm": 1.5015649139784593, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 32382 + }, + { + "epoch": 0.32383, + "grad_norm": 1.2516551671294567, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 32383 + }, + { + "epoch": 0.32384, + "grad_norm": 1.419920655175393, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 32384 + }, + { + "epoch": 0.32385, + "grad_norm": 1.383291378141968, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 32385 + }, + { + "epoch": 0.32386, + "grad_norm": 1.54802499688294, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 32386 + }, + { + "epoch": 0.32387, + "grad_norm": 1.503223324050554, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 32387 + }, + { + "epoch": 0.32388, + "grad_norm": 1.3889146741204075, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32388 + }, + { + "epoch": 0.32389, + "grad_norm": 1.3406894996694136, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 32389 + }, + { + "epoch": 0.3239, + "grad_norm": 1.5106835481497556, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 32390 + }, + { + "epoch": 0.32391, + "grad_norm": 1.3584489676177012, + "learning_rate": 0.003, + "loss": 4.013, + "step": 32391 + }, + { + "epoch": 0.32392, + "grad_norm": 1.4395938739949712, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 32392 + }, + { + "epoch": 0.32393, + "grad_norm": 1.50521520472078, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32393 + }, + { + "epoch": 0.32394, + "grad_norm": 1.4282385824857946, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 32394 + }, + { + "epoch": 0.32395, + "grad_norm": 1.4626141851980543, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 32395 + }, + { + "epoch": 0.32396, + "grad_norm": 1.4897928756141756, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32396 + }, + { + "epoch": 0.32397, + "grad_norm": 1.6048401899764129, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32397 + }, + { + "epoch": 0.32398, + "grad_norm": 1.3735812955675109, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 32398 + }, + { + "epoch": 0.32399, + "grad_norm": 1.4169910880703511, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32399 + }, + { + "epoch": 0.324, + "grad_norm": 1.6493650551703585, + "learning_rate": 0.003, + "loss": 4.037, + "step": 32400 + }, + { + "epoch": 0.32401, + "grad_norm": 1.6286863344349556, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 32401 + }, + { + "epoch": 0.32402, + "grad_norm": 1.279711983792975, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 32402 + }, + { + "epoch": 0.32403, + "grad_norm": 1.4121410293584755, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 32403 + }, + { + "epoch": 0.32404, + "grad_norm": 1.4667307629601454, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 32404 + }, + { + "epoch": 0.32405, + "grad_norm": 1.7858363172403628, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32405 + }, + { + "epoch": 0.32406, + "grad_norm": 1.5707240657353723, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 32406 + }, + { + "epoch": 0.32407, + "grad_norm": 1.6046776226508548, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32407 + }, + { + "epoch": 0.32408, + "grad_norm": 1.4443119144364611, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32408 + }, + { + "epoch": 0.32409, + "grad_norm": 1.2514817944023975, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 32409 + }, + { + "epoch": 0.3241, + "grad_norm": 1.5050136469266608, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 32410 + }, + { + "epoch": 0.32411, + "grad_norm": 1.5357310878357842, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32411 + }, + { + "epoch": 0.32412, + "grad_norm": 1.4710324041205662, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 32412 + }, + { + "epoch": 0.32413, + "grad_norm": 1.2291753362569349, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 32413 + }, + { + "epoch": 0.32414, + "grad_norm": 1.5824895457810348, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 32414 + }, + { + "epoch": 0.32415, + "grad_norm": 1.4351036049765138, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 32415 + }, + { + "epoch": 0.32416, + "grad_norm": 1.30334303671683, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 32416 + }, + { + "epoch": 0.32417, + "grad_norm": 1.5702390375395392, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32417 + }, + { + "epoch": 0.32418, + "grad_norm": 1.5631053489008293, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32418 + }, + { + "epoch": 0.32419, + "grad_norm": 1.3030858259568443, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 32419 + }, + { + "epoch": 0.3242, + "grad_norm": 1.6029376581151074, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32420 + }, + { + "epoch": 0.32421, + "grad_norm": 1.2817149398468655, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 32421 + }, + { + "epoch": 0.32422, + "grad_norm": 1.6385699041798103, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32422 + }, + { + "epoch": 0.32423, + "grad_norm": 1.555274851367309, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 32423 + }, + { + "epoch": 0.32424, + "grad_norm": 1.9306411066319153, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32424 + }, + { + "epoch": 0.32425, + "grad_norm": 1.4373570248435998, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 32425 + }, + { + "epoch": 0.32426, + "grad_norm": 1.4714189463685816, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32426 + }, + { + "epoch": 0.32427, + "grad_norm": 1.385720240649821, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 32427 + }, + { + "epoch": 0.32428, + "grad_norm": 1.7709763607441642, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 32428 + }, + { + "epoch": 0.32429, + "grad_norm": 1.3614004216602769, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 32429 + }, + { + "epoch": 0.3243, + "grad_norm": 1.6209833394094066, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 32430 + }, + { + "epoch": 0.32431, + "grad_norm": 1.3575550935658598, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32431 + }, + { + "epoch": 0.32432, + "grad_norm": 1.624674417373739, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 32432 + }, + { + "epoch": 0.32433, + "grad_norm": 1.3012720686829278, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 32433 + }, + { + "epoch": 0.32434, + "grad_norm": 1.2633739936833719, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32434 + }, + { + "epoch": 0.32435, + "grad_norm": 1.9159126782684155, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32435 + }, + { + "epoch": 0.32436, + "grad_norm": 1.118253789561597, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 32436 + }, + { + "epoch": 0.32437, + "grad_norm": 1.832015213192375, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32437 + }, + { + "epoch": 0.32438, + "grad_norm": 1.2329106461969375, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32438 + }, + { + "epoch": 0.32439, + "grad_norm": 1.7050713045815924, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 32439 + }, + { + "epoch": 0.3244, + "grad_norm": 1.428050878020372, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 32440 + }, + { + "epoch": 0.32441, + "grad_norm": 1.3929637101473415, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 32441 + }, + { + "epoch": 0.32442, + "grad_norm": 1.6027481419756537, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32442 + }, + { + "epoch": 0.32443, + "grad_norm": 1.5443291136579889, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 32443 + }, + { + "epoch": 0.32444, + "grad_norm": 1.4008687205700072, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 32444 + }, + { + "epoch": 0.32445, + "grad_norm": 1.460589665583895, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 32445 + }, + { + "epoch": 0.32446, + "grad_norm": 1.4667410889369483, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32446 + }, + { + "epoch": 0.32447, + "grad_norm": 1.3057113859001306, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 32447 + }, + { + "epoch": 0.32448, + "grad_norm": 1.4620555459154467, + "learning_rate": 0.003, + "loss": 4.05, + "step": 32448 + }, + { + "epoch": 0.32449, + "grad_norm": 1.3249449121649037, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32449 + }, + { + "epoch": 0.3245, + "grad_norm": 1.3431441001548656, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 32450 + }, + { + "epoch": 0.32451, + "grad_norm": 1.5373590034195417, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32451 + }, + { + "epoch": 0.32452, + "grad_norm": 1.4329914784245077, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 32452 + }, + { + "epoch": 0.32453, + "grad_norm": 1.556186384743845, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 32453 + }, + { + "epoch": 0.32454, + "grad_norm": 1.5227988674762494, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32454 + }, + { + "epoch": 0.32455, + "grad_norm": 1.4343745422370417, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 32455 + }, + { + "epoch": 0.32456, + "grad_norm": 1.7339243065650074, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32456 + }, + { + "epoch": 0.32457, + "grad_norm": 1.463532131279998, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 32457 + }, + { + "epoch": 0.32458, + "grad_norm": 1.939991750123029, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 32458 + }, + { + "epoch": 0.32459, + "grad_norm": 1.1935351907696203, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 32459 + }, + { + "epoch": 0.3246, + "grad_norm": 1.9570838842436549, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 32460 + }, + { + "epoch": 0.32461, + "grad_norm": 1.321820439765612, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 32461 + }, + { + "epoch": 0.32462, + "grad_norm": 1.6879827765459485, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 32462 + }, + { + "epoch": 0.32463, + "grad_norm": 1.5516684138538168, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32463 + }, + { + "epoch": 0.32464, + "grad_norm": 1.422678867771485, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 32464 + }, + { + "epoch": 0.32465, + "grad_norm": 1.4300241662021551, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 32465 + }, + { + "epoch": 0.32466, + "grad_norm": 1.402544914373237, + "learning_rate": 0.003, + "loss": 4.031, + "step": 32466 + }, + { + "epoch": 0.32467, + "grad_norm": 1.4579715468243897, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 32467 + }, + { + "epoch": 0.32468, + "grad_norm": 1.4457200262270788, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 32468 + }, + { + "epoch": 0.32469, + "grad_norm": 1.2236736077990886, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32469 + }, + { + "epoch": 0.3247, + "grad_norm": 1.8325995882465396, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 32470 + }, + { + "epoch": 0.32471, + "grad_norm": 1.4317530723823175, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 32471 + }, + { + "epoch": 0.32472, + "grad_norm": 1.6171840898166185, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 32472 + }, + { + "epoch": 0.32473, + "grad_norm": 1.5031466145333985, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32473 + }, + { + "epoch": 0.32474, + "grad_norm": 1.4933606739229963, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32474 + }, + { + "epoch": 0.32475, + "grad_norm": 1.1440108966724294, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 32475 + }, + { + "epoch": 0.32476, + "grad_norm": 1.7454841932184915, + "learning_rate": 0.003, + "loss": 4.044, + "step": 32476 + }, + { + "epoch": 0.32477, + "grad_norm": 1.4504921465243599, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 32477 + }, + { + "epoch": 0.32478, + "grad_norm": 1.7460613037861237, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32478 + }, + { + "epoch": 0.32479, + "grad_norm": 1.371368288274377, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32479 + }, + { + "epoch": 0.3248, + "grad_norm": 1.4972323759314092, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 32480 + }, + { + "epoch": 0.32481, + "grad_norm": 1.1936625384439574, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32481 + }, + { + "epoch": 0.32482, + "grad_norm": 1.6816772183481379, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32482 + }, + { + "epoch": 0.32483, + "grad_norm": 1.202732339030746, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 32483 + }, + { + "epoch": 0.32484, + "grad_norm": 1.390340468457136, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32484 + }, + { + "epoch": 0.32485, + "grad_norm": 1.6324322245235192, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 32485 + }, + { + "epoch": 0.32486, + "grad_norm": 1.1611162390781717, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 32486 + }, + { + "epoch": 0.32487, + "grad_norm": 1.4698317973231996, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 32487 + }, + { + "epoch": 0.32488, + "grad_norm": 1.217108659389562, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 32488 + }, + { + "epoch": 0.32489, + "grad_norm": 1.4601186329276015, + "learning_rate": 0.003, + "loss": 4.071, + "step": 32489 + }, + { + "epoch": 0.3249, + "grad_norm": 1.4145230244063036, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32490 + }, + { + "epoch": 0.32491, + "grad_norm": 1.5565311787178875, + "learning_rate": 0.003, + "loss": 4.086, + "step": 32491 + }, + { + "epoch": 0.32492, + "grad_norm": 1.3459254015598565, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32492 + }, + { + "epoch": 0.32493, + "grad_norm": 1.5366488274782188, + "learning_rate": 0.003, + "loss": 4.049, + "step": 32493 + }, + { + "epoch": 0.32494, + "grad_norm": 1.503601907623824, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 32494 + }, + { + "epoch": 0.32495, + "grad_norm": 1.6127412409972701, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 32495 + }, + { + "epoch": 0.32496, + "grad_norm": 1.5866901819050339, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 32496 + }, + { + "epoch": 0.32497, + "grad_norm": 1.5999378197388345, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 32497 + }, + { + "epoch": 0.32498, + "grad_norm": 1.3901129401971966, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 32498 + }, + { + "epoch": 0.32499, + "grad_norm": 1.563383085802422, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32499 + }, + { + "epoch": 0.325, + "grad_norm": 1.2924234929374097, + "learning_rate": 0.003, + "loss": 4.024, + "step": 32500 + }, + { + "epoch": 0.32501, + "grad_norm": 1.512371361709129, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32501 + }, + { + "epoch": 0.32502, + "grad_norm": 1.3703268665759865, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 32502 + }, + { + "epoch": 0.32503, + "grad_norm": 1.4937807027907084, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32503 + }, + { + "epoch": 0.32504, + "grad_norm": 1.4016135607458862, + "learning_rate": 0.003, + "loss": 4.027, + "step": 32504 + }, + { + "epoch": 0.32505, + "grad_norm": 1.4823749807887345, + "learning_rate": 0.003, + "loss": 4.034, + "step": 32505 + }, + { + "epoch": 0.32506, + "grad_norm": 1.4749912497312854, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 32506 + }, + { + "epoch": 0.32507, + "grad_norm": 1.665308721586397, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 32507 + }, + { + "epoch": 0.32508, + "grad_norm": 1.7256862714798233, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 32508 + }, + { + "epoch": 0.32509, + "grad_norm": 1.4818500997237682, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 32509 + }, + { + "epoch": 0.3251, + "grad_norm": 1.270475876593696, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 32510 + }, + { + "epoch": 0.32511, + "grad_norm": 1.718300675052822, + "learning_rate": 0.003, + "loss": 4.054, + "step": 32511 + }, + { + "epoch": 0.32512, + "grad_norm": 1.3883478021190514, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 32512 + }, + { + "epoch": 0.32513, + "grad_norm": 1.6562188613620161, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 32513 + }, + { + "epoch": 0.32514, + "grad_norm": 1.299566203669316, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32514 + }, + { + "epoch": 0.32515, + "grad_norm": 1.3959565670785108, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 32515 + }, + { + "epoch": 0.32516, + "grad_norm": 1.4032912963905604, + "learning_rate": 0.003, + "loss": 4.034, + "step": 32516 + }, + { + "epoch": 0.32517, + "grad_norm": 1.4582890862068068, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32517 + }, + { + "epoch": 0.32518, + "grad_norm": 1.530071187777648, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 32518 + }, + { + "epoch": 0.32519, + "grad_norm": 1.5055159766130732, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 32519 + }, + { + "epoch": 0.3252, + "grad_norm": 1.5789824368894674, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 32520 + }, + { + "epoch": 0.32521, + "grad_norm": 1.350130904349119, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 32521 + }, + { + "epoch": 0.32522, + "grad_norm": 1.7647902389320658, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32522 + }, + { + "epoch": 0.32523, + "grad_norm": 1.5017352890136773, + "learning_rate": 0.003, + "loss": 4.03, + "step": 32523 + }, + { + "epoch": 0.32524, + "grad_norm": 1.5509557573678556, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 32524 + }, + { + "epoch": 0.32525, + "grad_norm": 1.480113322083759, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 32525 + }, + { + "epoch": 0.32526, + "grad_norm": 1.453285495549427, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32526 + }, + { + "epoch": 0.32527, + "grad_norm": 1.477271340366542, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32527 + }, + { + "epoch": 0.32528, + "grad_norm": 1.3587491737689235, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 32528 + }, + { + "epoch": 0.32529, + "grad_norm": 1.8878485256938962, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32529 + }, + { + "epoch": 0.3253, + "grad_norm": 1.0053836381786105, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 32530 + }, + { + "epoch": 0.32531, + "grad_norm": 1.8388925075063547, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 32531 + }, + { + "epoch": 0.32532, + "grad_norm": 1.0272906060877325, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 32532 + }, + { + "epoch": 0.32533, + "grad_norm": 1.7155343246120953, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 32533 + }, + { + "epoch": 0.32534, + "grad_norm": 1.4452660715850951, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32534 + }, + { + "epoch": 0.32535, + "grad_norm": 1.3536090476259854, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 32535 + }, + { + "epoch": 0.32536, + "grad_norm": 1.472946328103338, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 32536 + }, + { + "epoch": 0.32537, + "grad_norm": 2.034924465835072, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 32537 + }, + { + "epoch": 0.32538, + "grad_norm": 1.6190093601920568, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 32538 + }, + { + "epoch": 0.32539, + "grad_norm": 1.5283510737303172, + "learning_rate": 0.003, + "loss": 4.073, + "step": 32539 + }, + { + "epoch": 0.3254, + "grad_norm": 1.4838300711804089, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32540 + }, + { + "epoch": 0.32541, + "grad_norm": 1.2334900365537973, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 32541 + }, + { + "epoch": 0.32542, + "grad_norm": 1.6001610696604385, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 32542 + }, + { + "epoch": 0.32543, + "grad_norm": 1.5031471246831545, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 32543 + }, + { + "epoch": 0.32544, + "grad_norm": 1.3294116521284003, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 32544 + }, + { + "epoch": 0.32545, + "grad_norm": 1.527028720376124, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 32545 + }, + { + "epoch": 0.32546, + "grad_norm": 1.6122840823564533, + "learning_rate": 0.003, + "loss": 4.072, + "step": 32546 + }, + { + "epoch": 0.32547, + "grad_norm": 1.4855081659717988, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 32547 + }, + { + "epoch": 0.32548, + "grad_norm": 1.5842868209502037, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 32548 + }, + { + "epoch": 0.32549, + "grad_norm": 1.2010520001518226, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 32549 + }, + { + "epoch": 0.3255, + "grad_norm": 1.7337343862374832, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 32550 + }, + { + "epoch": 0.32551, + "grad_norm": 1.636856429655589, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32551 + }, + { + "epoch": 0.32552, + "grad_norm": 1.3979692906572427, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 32552 + }, + { + "epoch": 0.32553, + "grad_norm": 1.4082389478634427, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 32553 + }, + { + "epoch": 0.32554, + "grad_norm": 1.7264405927740534, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 32554 + }, + { + "epoch": 0.32555, + "grad_norm": 1.1911131545830058, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32555 + }, + { + "epoch": 0.32556, + "grad_norm": 1.757527814569614, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 32556 + }, + { + "epoch": 0.32557, + "grad_norm": 1.3039450185327044, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 32557 + }, + { + "epoch": 0.32558, + "grad_norm": 1.4620660202666185, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32558 + }, + { + "epoch": 0.32559, + "grad_norm": 1.37277868569918, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 32559 + }, + { + "epoch": 0.3256, + "grad_norm": 1.5512011478464547, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 32560 + }, + { + "epoch": 0.32561, + "grad_norm": 1.4729683755362915, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32561 + }, + { + "epoch": 0.32562, + "grad_norm": 1.5690251243858158, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 32562 + }, + { + "epoch": 0.32563, + "grad_norm": 1.2515809471682664, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32563 + }, + { + "epoch": 0.32564, + "grad_norm": 1.342058136603553, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32564 + }, + { + "epoch": 0.32565, + "grad_norm": 1.493766269120391, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32565 + }, + { + "epoch": 0.32566, + "grad_norm": 1.3828715307628845, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 32566 + }, + { + "epoch": 0.32567, + "grad_norm": 1.5559936197004953, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32567 + }, + { + "epoch": 0.32568, + "grad_norm": 1.5005705745181641, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 32568 + }, + { + "epoch": 0.32569, + "grad_norm": 1.4459895159078877, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 32569 + }, + { + "epoch": 0.3257, + "grad_norm": 1.7132341876625667, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 32570 + }, + { + "epoch": 0.32571, + "grad_norm": 1.2149506461782238, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 32571 + }, + { + "epoch": 0.32572, + "grad_norm": 1.538838606746498, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 32572 + }, + { + "epoch": 0.32573, + "grad_norm": 1.466741281352767, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 32573 + }, + { + "epoch": 0.32574, + "grad_norm": 1.5189965433958204, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 32574 + }, + { + "epoch": 0.32575, + "grad_norm": 1.3520713553642745, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 32575 + }, + { + "epoch": 0.32576, + "grad_norm": 1.6394558703780595, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 32576 + }, + { + "epoch": 0.32577, + "grad_norm": 1.3541906367992165, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32577 + }, + { + "epoch": 0.32578, + "grad_norm": 1.1402710191613068, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 32578 + }, + { + "epoch": 0.32579, + "grad_norm": 1.7355141519158561, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 32579 + }, + { + "epoch": 0.3258, + "grad_norm": 1.258421753324234, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 32580 + }, + { + "epoch": 0.32581, + "grad_norm": 1.6726267742227232, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 32581 + }, + { + "epoch": 0.32582, + "grad_norm": 1.3163675257361933, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 32582 + }, + { + "epoch": 0.32583, + "grad_norm": 1.6730549969545294, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32583 + }, + { + "epoch": 0.32584, + "grad_norm": 1.3555258268991763, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 32584 + }, + { + "epoch": 0.32585, + "grad_norm": 1.6098080865714666, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32585 + }, + { + "epoch": 0.32586, + "grad_norm": 1.5919699589141207, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32586 + }, + { + "epoch": 0.32587, + "grad_norm": 1.3106983862021622, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32587 + }, + { + "epoch": 0.32588, + "grad_norm": 1.3736354675181508, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 32588 + }, + { + "epoch": 0.32589, + "grad_norm": 1.6112697255145856, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 32589 + }, + { + "epoch": 0.3259, + "grad_norm": 1.2000548366187611, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 32590 + }, + { + "epoch": 0.32591, + "grad_norm": 1.4633347214702286, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 32591 + }, + { + "epoch": 0.32592, + "grad_norm": 1.1839822863225395, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 32592 + }, + { + "epoch": 0.32593, + "grad_norm": 1.7507848704503648, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32593 + }, + { + "epoch": 0.32594, + "grad_norm": 1.9634186167698622, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 32594 + }, + { + "epoch": 0.32595, + "grad_norm": 1.3183002615508579, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 32595 + }, + { + "epoch": 0.32596, + "grad_norm": 1.6091267185654956, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 32596 + }, + { + "epoch": 0.32597, + "grad_norm": 1.572490558964845, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 32597 + }, + { + "epoch": 0.32598, + "grad_norm": 1.3298394023627202, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 32598 + }, + { + "epoch": 0.32599, + "grad_norm": 1.3337712092725238, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32599 + }, + { + "epoch": 0.326, + "grad_norm": 1.4926973991135766, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 32600 + }, + { + "epoch": 0.32601, + "grad_norm": 1.3551636880020017, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 32601 + }, + { + "epoch": 0.32602, + "grad_norm": 1.853182933036962, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 32602 + }, + { + "epoch": 0.32603, + "grad_norm": 1.4568252169998093, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32603 + }, + { + "epoch": 0.32604, + "grad_norm": 1.5059522760252195, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32604 + }, + { + "epoch": 0.32605, + "grad_norm": 1.4393393422711729, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 32605 + }, + { + "epoch": 0.32606, + "grad_norm": 1.4359812102099467, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32606 + }, + { + "epoch": 0.32607, + "grad_norm": 1.454602865432233, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 32607 + }, + { + "epoch": 0.32608, + "grad_norm": 1.466490654509849, + "learning_rate": 0.003, + "loss": 4.047, + "step": 32608 + }, + { + "epoch": 0.32609, + "grad_norm": 1.217517519272598, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 32609 + }, + { + "epoch": 0.3261, + "grad_norm": 1.5727968762417963, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 32610 + }, + { + "epoch": 0.32611, + "grad_norm": 1.44396416565061, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32611 + }, + { + "epoch": 0.32612, + "grad_norm": 1.4992563101818066, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 32612 + }, + { + "epoch": 0.32613, + "grad_norm": 1.4332843111733045, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 32613 + }, + { + "epoch": 0.32614, + "grad_norm": 1.5146446021752724, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 32614 + }, + { + "epoch": 0.32615, + "grad_norm": 1.5972077108467473, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32615 + }, + { + "epoch": 0.32616, + "grad_norm": 1.457080532184677, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 32616 + }, + { + "epoch": 0.32617, + "grad_norm": 1.2685901275288298, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 32617 + }, + { + "epoch": 0.32618, + "grad_norm": 1.574465575299868, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32618 + }, + { + "epoch": 0.32619, + "grad_norm": 1.0905207111037523, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 32619 + }, + { + "epoch": 0.3262, + "grad_norm": 1.543355366151098, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32620 + }, + { + "epoch": 0.32621, + "grad_norm": 1.2939527080814812, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32621 + }, + { + "epoch": 0.32622, + "grad_norm": 1.6387263038184412, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 32622 + }, + { + "epoch": 0.32623, + "grad_norm": 1.3822144870327786, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 32623 + }, + { + "epoch": 0.32624, + "grad_norm": 1.8457035329319509, + "learning_rate": 0.003, + "loss": 4.075, + "step": 32624 + }, + { + "epoch": 0.32625, + "grad_norm": 1.4544379448469424, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 32625 + }, + { + "epoch": 0.32626, + "grad_norm": 1.677830445839783, + "learning_rate": 0.003, + "loss": 4.033, + "step": 32626 + }, + { + "epoch": 0.32627, + "grad_norm": 1.600543881633963, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 32627 + }, + { + "epoch": 0.32628, + "grad_norm": 1.3966660096629702, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 32628 + }, + { + "epoch": 0.32629, + "grad_norm": 1.6908160881507854, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 32629 + }, + { + "epoch": 0.3263, + "grad_norm": 1.2834138171519105, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 32630 + }, + { + "epoch": 0.32631, + "grad_norm": 1.6235900322227346, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 32631 + }, + { + "epoch": 0.32632, + "grad_norm": 1.3143120004870972, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32632 + }, + { + "epoch": 0.32633, + "grad_norm": 1.8120471017165998, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 32633 + }, + { + "epoch": 0.32634, + "grad_norm": 1.408705825807223, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32634 + }, + { + "epoch": 0.32635, + "grad_norm": 1.487139070585365, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 32635 + }, + { + "epoch": 0.32636, + "grad_norm": 1.4024511977488945, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 32636 + }, + { + "epoch": 0.32637, + "grad_norm": 1.4616031946393193, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 32637 + }, + { + "epoch": 0.32638, + "grad_norm": 1.3724402226529848, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 32638 + }, + { + "epoch": 0.32639, + "grad_norm": 1.5820354726281523, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 32639 + }, + { + "epoch": 0.3264, + "grad_norm": 1.4491879288561944, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32640 + }, + { + "epoch": 0.32641, + "grad_norm": 1.4921065926672725, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 32641 + }, + { + "epoch": 0.32642, + "grad_norm": 1.8335764167163702, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 32642 + }, + { + "epoch": 0.32643, + "grad_norm": 1.3637619782491526, + "learning_rate": 0.003, + "loss": 4.068, + "step": 32643 + }, + { + "epoch": 0.32644, + "grad_norm": 1.7105956182148967, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 32644 + }, + { + "epoch": 0.32645, + "grad_norm": 1.4531197234915634, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 32645 + }, + { + "epoch": 0.32646, + "grad_norm": 1.2992103024936352, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 32646 + }, + { + "epoch": 0.32647, + "grad_norm": 1.5461997269764371, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 32647 + }, + { + "epoch": 0.32648, + "grad_norm": 1.3815855666889756, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32648 + }, + { + "epoch": 0.32649, + "grad_norm": 1.676630681090202, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 32649 + }, + { + "epoch": 0.3265, + "grad_norm": 1.0726544244252934, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 32650 + }, + { + "epoch": 0.32651, + "grad_norm": 1.5072738483230805, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 32651 + }, + { + "epoch": 0.32652, + "grad_norm": 1.643731751881673, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32652 + }, + { + "epoch": 0.32653, + "grad_norm": 1.4153830616200178, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 32653 + }, + { + "epoch": 0.32654, + "grad_norm": 1.5204207948309472, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 32654 + }, + { + "epoch": 0.32655, + "grad_norm": 1.4971065043796394, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32655 + }, + { + "epoch": 0.32656, + "grad_norm": 1.4654278297735992, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 32656 + }, + { + "epoch": 0.32657, + "grad_norm": 1.4630868218556756, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 32657 + }, + { + "epoch": 0.32658, + "grad_norm": 1.6946710469581456, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 32658 + }, + { + "epoch": 0.32659, + "grad_norm": 1.081439782361858, + "learning_rate": 0.003, + "loss": 4.048, + "step": 32659 + }, + { + "epoch": 0.3266, + "grad_norm": 1.841883024846223, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32660 + }, + { + "epoch": 0.32661, + "grad_norm": 1.5975093735869408, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32661 + }, + { + "epoch": 0.32662, + "grad_norm": 1.4760598481496918, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 32662 + }, + { + "epoch": 0.32663, + "grad_norm": 1.5775873213946725, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 32663 + }, + { + "epoch": 0.32664, + "grad_norm": 1.6291388328414325, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 32664 + }, + { + "epoch": 0.32665, + "grad_norm": 1.3660043638455615, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 32665 + }, + { + "epoch": 0.32666, + "grad_norm": 1.2933985511751716, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 32666 + }, + { + "epoch": 0.32667, + "grad_norm": 1.494492325606787, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 32667 + }, + { + "epoch": 0.32668, + "grad_norm": 1.4291547079467022, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 32668 + }, + { + "epoch": 0.32669, + "grad_norm": 1.648644002476279, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 32669 + }, + { + "epoch": 0.3267, + "grad_norm": 1.7145528671711991, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 32670 + }, + { + "epoch": 0.32671, + "grad_norm": 1.3655355564937854, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 32671 + }, + { + "epoch": 0.32672, + "grad_norm": 1.4928984321337706, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32672 + }, + { + "epoch": 0.32673, + "grad_norm": 1.456610288804189, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 32673 + }, + { + "epoch": 0.32674, + "grad_norm": 1.3857927851523884, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 32674 + }, + { + "epoch": 0.32675, + "grad_norm": 1.5045227562568098, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32675 + }, + { + "epoch": 0.32676, + "grad_norm": 1.5478473544992175, + "learning_rate": 0.003, + "loss": 4.028, + "step": 32676 + }, + { + "epoch": 0.32677, + "grad_norm": 1.4379395179583363, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 32677 + }, + { + "epoch": 0.32678, + "grad_norm": 1.5620599276306955, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32678 + }, + { + "epoch": 0.32679, + "grad_norm": 1.440811304800231, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32679 + }, + { + "epoch": 0.3268, + "grad_norm": 1.3223523081332065, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 32680 + }, + { + "epoch": 0.32681, + "grad_norm": 1.6190205628710477, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 32681 + }, + { + "epoch": 0.32682, + "grad_norm": 1.2706259227048091, + "learning_rate": 0.003, + "loss": 4.033, + "step": 32682 + }, + { + "epoch": 0.32683, + "grad_norm": 1.675893040317369, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 32683 + }, + { + "epoch": 0.32684, + "grad_norm": 1.168167300149475, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 32684 + }, + { + "epoch": 0.32685, + "grad_norm": 1.8895196918981727, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 32685 + }, + { + "epoch": 0.32686, + "grad_norm": 1.184273733798758, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 32686 + }, + { + "epoch": 0.32687, + "grad_norm": 1.6448565700059967, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 32687 + }, + { + "epoch": 0.32688, + "grad_norm": 1.5917084173392497, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 32688 + }, + { + "epoch": 0.32689, + "grad_norm": 1.4603909754110809, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 32689 + }, + { + "epoch": 0.3269, + "grad_norm": 1.4272786855566904, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 32690 + }, + { + "epoch": 0.32691, + "grad_norm": 1.659050407210462, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 32691 + }, + { + "epoch": 0.32692, + "grad_norm": 1.29927311387899, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 32692 + }, + { + "epoch": 0.32693, + "grad_norm": 1.559306622971788, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 32693 + }, + { + "epoch": 0.32694, + "grad_norm": 1.422744331639828, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 32694 + }, + { + "epoch": 0.32695, + "grad_norm": 1.5844115143549118, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 32695 + }, + { + "epoch": 0.32696, + "grad_norm": 1.1863471537261137, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 32696 + }, + { + "epoch": 0.32697, + "grad_norm": 1.8509687073697876, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 32697 + }, + { + "epoch": 0.32698, + "grad_norm": 1.6116185274400359, + "learning_rate": 0.003, + "loss": 4.078, + "step": 32698 + }, + { + "epoch": 0.32699, + "grad_norm": 1.398672705403487, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 32699 + }, + { + "epoch": 0.327, + "grad_norm": 1.436584491494178, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 32700 + }, + { + "epoch": 0.32701, + "grad_norm": 1.4567274245878974, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32701 + }, + { + "epoch": 0.32702, + "grad_norm": 1.3562556331789364, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 32702 + }, + { + "epoch": 0.32703, + "grad_norm": 1.3184266216580305, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 32703 + }, + { + "epoch": 0.32704, + "grad_norm": 1.343531894367954, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 32704 + }, + { + "epoch": 0.32705, + "grad_norm": 1.2986908728893714, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 32705 + }, + { + "epoch": 0.32706, + "grad_norm": 1.5712646706425824, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32706 + }, + { + "epoch": 0.32707, + "grad_norm": 1.2664288190577377, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32707 + }, + { + "epoch": 0.32708, + "grad_norm": 2.0528439100177245, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 32708 + }, + { + "epoch": 0.32709, + "grad_norm": 1.1829598883719745, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 32709 + }, + { + "epoch": 0.3271, + "grad_norm": 1.753474068633984, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32710 + }, + { + "epoch": 0.32711, + "grad_norm": 1.294587160624768, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 32711 + }, + { + "epoch": 0.32712, + "grad_norm": 1.7090845536232828, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 32712 + }, + { + "epoch": 0.32713, + "grad_norm": 1.425028206238877, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 32713 + }, + { + "epoch": 0.32714, + "grad_norm": 1.6192733354328654, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32714 + }, + { + "epoch": 0.32715, + "grad_norm": 1.4276341336023601, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 32715 + }, + { + "epoch": 0.32716, + "grad_norm": 1.6447339837513806, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 32716 + }, + { + "epoch": 0.32717, + "grad_norm": 1.7926665743907406, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 32717 + }, + { + "epoch": 0.32718, + "grad_norm": 1.3867004987219147, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 32718 + }, + { + "epoch": 0.32719, + "grad_norm": 1.6744734625402053, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 32719 + }, + { + "epoch": 0.3272, + "grad_norm": 1.2519456328238405, + "learning_rate": 0.003, + "loss": 4.057, + "step": 32720 + }, + { + "epoch": 0.32721, + "grad_norm": 1.6575716778386507, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 32721 + }, + { + "epoch": 0.32722, + "grad_norm": 1.286627358323194, + "learning_rate": 0.003, + "loss": 4.02, + "step": 32722 + }, + { + "epoch": 0.32723, + "grad_norm": 1.6963484342659516, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32723 + }, + { + "epoch": 0.32724, + "grad_norm": 1.3013046408352014, + "learning_rate": 0.003, + "loss": 4.054, + "step": 32724 + }, + { + "epoch": 0.32725, + "grad_norm": 1.598315678421, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 32725 + }, + { + "epoch": 0.32726, + "grad_norm": 1.362109273370826, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 32726 + }, + { + "epoch": 0.32727, + "grad_norm": 1.4729277748389697, + "learning_rate": 0.003, + "loss": 4.053, + "step": 32727 + }, + { + "epoch": 0.32728, + "grad_norm": 1.529277044368224, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 32728 + }, + { + "epoch": 0.32729, + "grad_norm": 1.3609553649616022, + "learning_rate": 0.003, + "loss": 4.054, + "step": 32729 + }, + { + "epoch": 0.3273, + "grad_norm": 1.3372784672142626, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 32730 + }, + { + "epoch": 0.32731, + "grad_norm": 1.2673467720152116, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 32731 + }, + { + "epoch": 0.32732, + "grad_norm": 1.730981431610937, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32732 + }, + { + "epoch": 0.32733, + "grad_norm": 1.3800799797180021, + "learning_rate": 0.003, + "loss": 4.061, + "step": 32733 + }, + { + "epoch": 0.32734, + "grad_norm": 1.8297007115129487, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32734 + }, + { + "epoch": 0.32735, + "grad_norm": 1.6070413701966135, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 32735 + }, + { + "epoch": 0.32736, + "grad_norm": 1.195999723515905, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 32736 + }, + { + "epoch": 0.32737, + "grad_norm": 1.631985903807904, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 32737 + }, + { + "epoch": 0.32738, + "grad_norm": 1.588702267700524, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32738 + }, + { + "epoch": 0.32739, + "grad_norm": 1.2224012825603139, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32739 + }, + { + "epoch": 0.3274, + "grad_norm": 1.554148271784534, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 32740 + }, + { + "epoch": 0.32741, + "grad_norm": 1.4941778448695158, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 32741 + }, + { + "epoch": 0.32742, + "grad_norm": 1.486980176587374, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 32742 + }, + { + "epoch": 0.32743, + "grad_norm": 1.4855545533783048, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 32743 + }, + { + "epoch": 0.32744, + "grad_norm": 1.1714558300883877, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32744 + }, + { + "epoch": 0.32745, + "grad_norm": 1.386244357607021, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32745 + }, + { + "epoch": 0.32746, + "grad_norm": 1.3589368763537553, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32746 + }, + { + "epoch": 0.32747, + "grad_norm": 1.4379805917265074, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 32747 + }, + { + "epoch": 0.32748, + "grad_norm": 1.3453095990382415, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32748 + }, + { + "epoch": 0.32749, + "grad_norm": 1.4869855899361253, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 32749 + }, + { + "epoch": 0.3275, + "grad_norm": 1.4686229283279844, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 32750 + }, + { + "epoch": 0.32751, + "grad_norm": 1.3143081346421739, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 32751 + }, + { + "epoch": 0.32752, + "grad_norm": 1.593176219826704, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 32752 + }, + { + "epoch": 0.32753, + "grad_norm": 1.2182108370468414, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 32753 + }, + { + "epoch": 0.32754, + "grad_norm": 1.559697015927542, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32754 + }, + { + "epoch": 0.32755, + "grad_norm": 1.4459048808143125, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 32755 + }, + { + "epoch": 0.32756, + "grad_norm": 1.8583734896202952, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 32756 + }, + { + "epoch": 0.32757, + "grad_norm": 1.6744319453951562, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 32757 + }, + { + "epoch": 0.32758, + "grad_norm": 1.6491558120295253, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 32758 + }, + { + "epoch": 0.32759, + "grad_norm": 1.2095980690985155, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 32759 + }, + { + "epoch": 0.3276, + "grad_norm": 1.5695335692691454, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 32760 + }, + { + "epoch": 0.32761, + "grad_norm": 1.3203981899548354, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 32761 + }, + { + "epoch": 0.32762, + "grad_norm": 1.5780031549077596, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 32762 + }, + { + "epoch": 0.32763, + "grad_norm": 1.4273406025187787, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 32763 + }, + { + "epoch": 0.32764, + "grad_norm": 1.331908266395152, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32764 + }, + { + "epoch": 0.32765, + "grad_norm": 1.4801090525482068, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 32765 + }, + { + "epoch": 0.32766, + "grad_norm": 1.3510993911474192, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 32766 + }, + { + "epoch": 0.32767, + "grad_norm": 1.4771742625490123, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 32767 + }, + { + "epoch": 0.32768, + "grad_norm": 1.3685181816087446, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 32768 + }, + { + "epoch": 0.32769, + "grad_norm": 1.671239731367002, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32769 + }, + { + "epoch": 0.3277, + "grad_norm": 1.4435434937658291, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 32770 + }, + { + "epoch": 0.32771, + "grad_norm": 1.485562304677857, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 32771 + }, + { + "epoch": 0.32772, + "grad_norm": 1.4629368681374837, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 32772 + }, + { + "epoch": 0.32773, + "grad_norm": 1.8552288071572747, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 32773 + }, + { + "epoch": 0.32774, + "grad_norm": 1.4859309586464706, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 32774 + }, + { + "epoch": 0.32775, + "grad_norm": 1.7047519283755599, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 32775 + }, + { + "epoch": 0.32776, + "grad_norm": 1.4385073138159814, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32776 + }, + { + "epoch": 0.32777, + "grad_norm": 1.422876182214273, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 32777 + }, + { + "epoch": 0.32778, + "grad_norm": 1.418457292878246, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32778 + }, + { + "epoch": 0.32779, + "grad_norm": 1.4015493636689722, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 32779 + }, + { + "epoch": 0.3278, + "grad_norm": 1.4306761897980664, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 32780 + }, + { + "epoch": 0.32781, + "grad_norm": 1.5293610413069574, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 32781 + }, + { + "epoch": 0.32782, + "grad_norm": 1.537890694980871, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 32782 + }, + { + "epoch": 0.32783, + "grad_norm": 1.3306275817119833, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 32783 + }, + { + "epoch": 0.32784, + "grad_norm": 1.7055056288946884, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 32784 + }, + { + "epoch": 0.32785, + "grad_norm": 1.2796994034762477, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 32785 + }, + { + "epoch": 0.32786, + "grad_norm": 1.5069584405091367, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 32786 + }, + { + "epoch": 0.32787, + "grad_norm": 1.363749834021279, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 32787 + }, + { + "epoch": 0.32788, + "grad_norm": 1.582012054609873, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 32788 + }, + { + "epoch": 0.32789, + "grad_norm": 1.5009829319980224, + "learning_rate": 0.003, + "loss": 4.052, + "step": 32789 + }, + { + "epoch": 0.3279, + "grad_norm": 1.4300656151045663, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 32790 + }, + { + "epoch": 0.32791, + "grad_norm": 1.4691723182967587, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 32791 + }, + { + "epoch": 0.32792, + "grad_norm": 1.498269891234717, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32792 + }, + { + "epoch": 0.32793, + "grad_norm": 1.493323209793951, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32793 + }, + { + "epoch": 0.32794, + "grad_norm": 1.4846160326440812, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32794 + }, + { + "epoch": 0.32795, + "grad_norm": 1.47390197033521, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 32795 + }, + { + "epoch": 0.32796, + "grad_norm": 1.6188709176434675, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 32796 + }, + { + "epoch": 0.32797, + "grad_norm": 1.4474048686278251, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32797 + }, + { + "epoch": 0.32798, + "grad_norm": 1.3675891821245914, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 32798 + }, + { + "epoch": 0.32799, + "grad_norm": 1.7161956352929384, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32799 + }, + { + "epoch": 0.328, + "grad_norm": 1.342668696249903, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 32800 + }, + { + "epoch": 0.32801, + "grad_norm": 1.25653302805468, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 32801 + }, + { + "epoch": 0.32802, + "grad_norm": 1.502877938972115, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 32802 + }, + { + "epoch": 0.32803, + "grad_norm": 1.2243208822041391, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 32803 + }, + { + "epoch": 0.32804, + "grad_norm": 1.772027364593043, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32804 + }, + { + "epoch": 0.32805, + "grad_norm": 1.2941295601919607, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32805 + }, + { + "epoch": 0.32806, + "grad_norm": 1.9825973794562557, + "learning_rate": 0.003, + "loss": 4.065, + "step": 32806 + }, + { + "epoch": 0.32807, + "grad_norm": 1.4064581781927952, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 32807 + }, + { + "epoch": 0.32808, + "grad_norm": 1.525942843711547, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 32808 + }, + { + "epoch": 0.32809, + "grad_norm": 1.4367896769539015, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 32809 + }, + { + "epoch": 0.3281, + "grad_norm": 1.393460184944441, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 32810 + }, + { + "epoch": 0.32811, + "grad_norm": 1.5148569903083362, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 32811 + }, + { + "epoch": 0.32812, + "grad_norm": 1.6770886384042198, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 32812 + }, + { + "epoch": 0.32813, + "grad_norm": 1.20290679404939, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 32813 + }, + { + "epoch": 0.32814, + "grad_norm": 2.0035868276792237, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32814 + }, + { + "epoch": 0.32815, + "grad_norm": 1.3034741102756313, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 32815 + }, + { + "epoch": 0.32816, + "grad_norm": 1.5783637809111093, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 32816 + }, + { + "epoch": 0.32817, + "grad_norm": 1.454398424715327, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32817 + }, + { + "epoch": 0.32818, + "grad_norm": 1.5402080955258346, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 32818 + }, + { + "epoch": 0.32819, + "grad_norm": 1.2942846563610586, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 32819 + }, + { + "epoch": 0.3282, + "grad_norm": 1.8115219412295898, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 32820 + }, + { + "epoch": 0.32821, + "grad_norm": 1.3308185531863654, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 32821 + }, + { + "epoch": 0.32822, + "grad_norm": 1.4715558292623558, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 32822 + }, + { + "epoch": 0.32823, + "grad_norm": 1.702366912898513, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 32823 + }, + { + "epoch": 0.32824, + "grad_norm": 1.5370157090124899, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 32824 + }, + { + "epoch": 0.32825, + "grad_norm": 1.3923880092976815, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 32825 + }, + { + "epoch": 0.32826, + "grad_norm": 1.431000102333161, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 32826 + }, + { + "epoch": 0.32827, + "grad_norm": 1.506131716287829, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 32827 + }, + { + "epoch": 0.32828, + "grad_norm": 1.3724286631652634, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 32828 + }, + { + "epoch": 0.32829, + "grad_norm": 1.6812968716438215, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32829 + }, + { + "epoch": 0.3283, + "grad_norm": 1.168430151766161, + "learning_rate": 0.003, + "loss": 4.063, + "step": 32830 + }, + { + "epoch": 0.32831, + "grad_norm": 1.65178720063156, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32831 + }, + { + "epoch": 0.32832, + "grad_norm": 1.271227400356102, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32832 + }, + { + "epoch": 0.32833, + "grad_norm": 1.7245425563872254, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 32833 + }, + { + "epoch": 0.32834, + "grad_norm": 1.4730698692944695, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 32834 + }, + { + "epoch": 0.32835, + "grad_norm": 1.3104676532089012, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 32835 + }, + { + "epoch": 0.32836, + "grad_norm": 1.8279480739513985, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 32836 + }, + { + "epoch": 0.32837, + "grad_norm": 1.7167311739643831, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 32837 + }, + { + "epoch": 0.32838, + "grad_norm": 1.2807026797334007, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 32838 + }, + { + "epoch": 0.32839, + "grad_norm": 1.6360539882681504, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 32839 + }, + { + "epoch": 0.3284, + "grad_norm": 1.4421804178037783, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 32840 + }, + { + "epoch": 0.32841, + "grad_norm": 1.3848166456158995, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 32841 + }, + { + "epoch": 0.32842, + "grad_norm": 1.4358185018398704, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 32842 + }, + { + "epoch": 0.32843, + "grad_norm": 1.28616210436744, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 32843 + }, + { + "epoch": 0.32844, + "grad_norm": 1.494820547791264, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 32844 + }, + { + "epoch": 0.32845, + "grad_norm": 1.4856153697180339, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32845 + }, + { + "epoch": 0.32846, + "grad_norm": 1.8061920233509683, + "learning_rate": 0.003, + "loss": 4.029, + "step": 32846 + }, + { + "epoch": 0.32847, + "grad_norm": 1.5311379034230514, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 32847 + }, + { + "epoch": 0.32848, + "grad_norm": 1.4501994667536864, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 32848 + }, + { + "epoch": 0.32849, + "grad_norm": 1.5743459287985908, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32849 + }, + { + "epoch": 0.3285, + "grad_norm": 1.61131034108662, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32850 + }, + { + "epoch": 0.32851, + "grad_norm": 1.165691814320377, + "learning_rate": 0.003, + "loss": 4.059, + "step": 32851 + }, + { + "epoch": 0.32852, + "grad_norm": 1.5654007424340932, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 32852 + }, + { + "epoch": 0.32853, + "grad_norm": 1.211937587954622, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 32853 + }, + { + "epoch": 0.32854, + "grad_norm": 1.7211506585968912, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 32854 + }, + { + "epoch": 0.32855, + "grad_norm": 1.282061940791079, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 32855 + }, + { + "epoch": 0.32856, + "grad_norm": 1.8245396034138213, + "learning_rate": 0.003, + "loss": 4.046, + "step": 32856 + }, + { + "epoch": 0.32857, + "grad_norm": 1.950109622555555, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 32857 + }, + { + "epoch": 0.32858, + "grad_norm": 1.0760703612287976, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 32858 + }, + { + "epoch": 0.32859, + "grad_norm": 1.8316279620646638, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 32859 + }, + { + "epoch": 0.3286, + "grad_norm": 1.3340648508811428, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 32860 + }, + { + "epoch": 0.32861, + "grad_norm": 1.770175527890696, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 32861 + }, + { + "epoch": 0.32862, + "grad_norm": 1.3233459589099628, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 32862 + }, + { + "epoch": 0.32863, + "grad_norm": 1.5684345263258994, + "learning_rate": 0.003, + "loss": 4.023, + "step": 32863 + }, + { + "epoch": 0.32864, + "grad_norm": 1.3646737173057513, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 32864 + }, + { + "epoch": 0.32865, + "grad_norm": 1.41008445740309, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 32865 + }, + { + "epoch": 0.32866, + "grad_norm": 1.3249516918226647, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 32866 + }, + { + "epoch": 0.32867, + "grad_norm": 1.521518803386459, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 32867 + }, + { + "epoch": 0.32868, + "grad_norm": 1.7283775978169755, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 32868 + }, + { + "epoch": 0.32869, + "grad_norm": 1.316988249464705, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 32869 + }, + { + "epoch": 0.3287, + "grad_norm": 1.680646181727678, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 32870 + }, + { + "epoch": 0.32871, + "grad_norm": 1.4449764111616537, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 32871 + }, + { + "epoch": 0.32872, + "grad_norm": 1.231843179451984, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 32872 + }, + { + "epoch": 0.32873, + "grad_norm": 1.666801770124811, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 32873 + }, + { + "epoch": 0.32874, + "grad_norm": 1.361351346262077, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 32874 + }, + { + "epoch": 0.32875, + "grad_norm": 1.5752942325711194, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 32875 + }, + { + "epoch": 0.32876, + "grad_norm": 1.3835712983934116, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32876 + }, + { + "epoch": 0.32877, + "grad_norm": 1.3277397548742695, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 32877 + }, + { + "epoch": 0.32878, + "grad_norm": 1.5891152239515556, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 32878 + }, + { + "epoch": 0.32879, + "grad_norm": 1.1798926436772936, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 32879 + }, + { + "epoch": 0.3288, + "grad_norm": 1.7061256170183863, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 32880 + }, + { + "epoch": 0.32881, + "grad_norm": 1.2045338206249838, + "learning_rate": 0.003, + "loss": 4.038, + "step": 32881 + }, + { + "epoch": 0.32882, + "grad_norm": 1.7063306560556484, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 32882 + }, + { + "epoch": 0.32883, + "grad_norm": 1.7724149329407028, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 32883 + }, + { + "epoch": 0.32884, + "grad_norm": 1.290147173647931, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 32884 + }, + { + "epoch": 0.32885, + "grad_norm": 1.592576772555407, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 32885 + }, + { + "epoch": 0.32886, + "grad_norm": 1.5776398392062694, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 32886 + }, + { + "epoch": 0.32887, + "grad_norm": 1.6058154542550953, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 32887 + }, + { + "epoch": 0.32888, + "grad_norm": 1.4000684244483657, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 32888 + }, + { + "epoch": 0.32889, + "grad_norm": 1.58607254810144, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 32889 + }, + { + "epoch": 0.3289, + "grad_norm": 1.1661167075793835, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 32890 + }, + { + "epoch": 0.32891, + "grad_norm": 1.741262108823949, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 32891 + }, + { + "epoch": 0.32892, + "grad_norm": 1.2653017971567502, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 32892 + }, + { + "epoch": 0.32893, + "grad_norm": 1.6164375915377511, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 32893 + }, + { + "epoch": 0.32894, + "grad_norm": 1.5459635394945124, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 32894 + }, + { + "epoch": 0.32895, + "grad_norm": 1.86514263601676, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 32895 + }, + { + "epoch": 0.32896, + "grad_norm": 1.243214082513604, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 32896 + }, + { + "epoch": 0.32897, + "grad_norm": 1.8483674228273848, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 32897 + }, + { + "epoch": 0.32898, + "grad_norm": 1.3696409558475073, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 32898 + }, + { + "epoch": 0.32899, + "grad_norm": 1.561502083674468, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 32899 + }, + { + "epoch": 0.329, + "grad_norm": 1.1842171793790381, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 32900 + }, + { + "epoch": 0.32901, + "grad_norm": 1.751865227143096, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 32901 + }, + { + "epoch": 0.32902, + "grad_norm": 1.441376611357372, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 32902 + }, + { + "epoch": 0.32903, + "grad_norm": 1.2886327501331394, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 32903 + }, + { + "epoch": 0.32904, + "grad_norm": 1.487407948019109, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32904 + }, + { + "epoch": 0.32905, + "grad_norm": 1.3964391279073594, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 32905 + }, + { + "epoch": 0.32906, + "grad_norm": 1.3378167792774478, + "learning_rate": 0.003, + "loss": 4.051, + "step": 32906 + }, + { + "epoch": 0.32907, + "grad_norm": 1.46291351854592, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 32907 + }, + { + "epoch": 0.32908, + "grad_norm": 1.5913316242758662, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 32908 + }, + { + "epoch": 0.32909, + "grad_norm": 1.3762351694093742, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 32909 + }, + { + "epoch": 0.3291, + "grad_norm": 1.9063185819806598, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 32910 + }, + { + "epoch": 0.32911, + "grad_norm": 1.4311067307886725, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 32911 + }, + { + "epoch": 0.32912, + "grad_norm": 1.6985008600330451, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 32912 + }, + { + "epoch": 0.32913, + "grad_norm": 1.161613125447573, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 32913 + }, + { + "epoch": 0.32914, + "grad_norm": 1.5629280617378938, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 32914 + }, + { + "epoch": 0.32915, + "grad_norm": 1.5708827772220584, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 32915 + }, + { + "epoch": 0.32916, + "grad_norm": 1.4221375765488051, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 32916 + }, + { + "epoch": 0.32917, + "grad_norm": 1.4973362063807638, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 32917 + }, + { + "epoch": 0.32918, + "grad_norm": 1.496768958189968, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 32918 + }, + { + "epoch": 0.32919, + "grad_norm": 1.7044310820716282, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32919 + }, + { + "epoch": 0.3292, + "grad_norm": 1.5231259521776972, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 32920 + }, + { + "epoch": 0.32921, + "grad_norm": 1.5635844945034996, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32921 + }, + { + "epoch": 0.32922, + "grad_norm": 1.2665742209934021, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 32922 + }, + { + "epoch": 0.32923, + "grad_norm": 1.4078435426989508, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 32923 + }, + { + "epoch": 0.32924, + "grad_norm": 1.1644946290106397, + "learning_rate": 0.003, + "loss": 4.004, + "step": 32924 + }, + { + "epoch": 0.32925, + "grad_norm": 1.6236485677754868, + "learning_rate": 0.003, + "loss": 4.024, + "step": 32925 + }, + { + "epoch": 0.32926, + "grad_norm": 1.4119492378427234, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 32926 + }, + { + "epoch": 0.32927, + "grad_norm": 1.8159270233268263, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 32927 + }, + { + "epoch": 0.32928, + "grad_norm": 1.247313854169582, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 32928 + }, + { + "epoch": 0.32929, + "grad_norm": 1.4805643079301132, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 32929 + }, + { + "epoch": 0.3293, + "grad_norm": 1.471104159236813, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 32930 + }, + { + "epoch": 0.32931, + "grad_norm": 1.5535366861664004, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 32931 + }, + { + "epoch": 0.32932, + "grad_norm": 1.5391184195327003, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 32932 + }, + { + "epoch": 0.32933, + "grad_norm": 1.4842368418091312, + "learning_rate": 0.003, + "loss": 4.07, + "step": 32933 + }, + { + "epoch": 0.32934, + "grad_norm": 1.4249213963159517, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 32934 + }, + { + "epoch": 0.32935, + "grad_norm": 1.1797342832091695, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 32935 + }, + { + "epoch": 0.32936, + "grad_norm": 1.7149107558597216, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 32936 + }, + { + "epoch": 0.32937, + "grad_norm": 1.2274052113661789, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 32937 + }, + { + "epoch": 0.32938, + "grad_norm": 1.4324380174946243, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 32938 + }, + { + "epoch": 0.32939, + "grad_norm": 1.5539622957837267, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 32939 + }, + { + "epoch": 0.3294, + "grad_norm": 1.7037157465436337, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 32940 + }, + { + "epoch": 0.32941, + "grad_norm": 1.5145164308826418, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32941 + }, + { + "epoch": 0.32942, + "grad_norm": 1.409307521456281, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 32942 + }, + { + "epoch": 0.32943, + "grad_norm": 1.3420849551164706, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 32943 + }, + { + "epoch": 0.32944, + "grad_norm": 1.4567996679724071, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 32944 + }, + { + "epoch": 0.32945, + "grad_norm": 1.2474287165726166, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 32945 + }, + { + "epoch": 0.32946, + "grad_norm": 1.6897181057974584, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 32946 + }, + { + "epoch": 0.32947, + "grad_norm": 1.3774017417974127, + "learning_rate": 0.003, + "loss": 4.04, + "step": 32947 + }, + { + "epoch": 0.32948, + "grad_norm": 1.5016951578221354, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 32948 + }, + { + "epoch": 0.32949, + "grad_norm": 1.2013091821075679, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 32949 + }, + { + "epoch": 0.3295, + "grad_norm": 1.705587604256363, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 32950 + }, + { + "epoch": 0.32951, + "grad_norm": 1.1921580954069693, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 32951 + }, + { + "epoch": 0.32952, + "grad_norm": 1.5618458680339187, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 32952 + }, + { + "epoch": 0.32953, + "grad_norm": 1.6712749456590368, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 32953 + }, + { + "epoch": 0.32954, + "grad_norm": 1.375979725535691, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32954 + }, + { + "epoch": 0.32955, + "grad_norm": 1.8142314165333502, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 32955 + }, + { + "epoch": 0.32956, + "grad_norm": 1.1454036667779017, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 32956 + }, + { + "epoch": 0.32957, + "grad_norm": 1.8010797205587208, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 32957 + }, + { + "epoch": 0.32958, + "grad_norm": 1.7586570225255231, + "learning_rate": 0.003, + "loss": 4.064, + "step": 32958 + }, + { + "epoch": 0.32959, + "grad_norm": 1.3864189856456004, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32959 + }, + { + "epoch": 0.3296, + "grad_norm": 1.6628647862718358, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 32960 + }, + { + "epoch": 0.32961, + "grad_norm": 1.5907466077151038, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 32961 + }, + { + "epoch": 0.32962, + "grad_norm": 1.6089236373180247, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 32962 + }, + { + "epoch": 0.32963, + "grad_norm": 1.3591445952190708, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 32963 + }, + { + "epoch": 0.32964, + "grad_norm": 1.4108637529631833, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 32964 + }, + { + "epoch": 0.32965, + "grad_norm": 1.6527457928019225, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 32965 + }, + { + "epoch": 0.32966, + "grad_norm": 1.2106461225169156, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 32966 + }, + { + "epoch": 0.32967, + "grad_norm": 1.4881544940883578, + "learning_rate": 0.003, + "loss": 4.045, + "step": 32967 + }, + { + "epoch": 0.32968, + "grad_norm": 1.3713693415859232, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 32968 + }, + { + "epoch": 0.32969, + "grad_norm": 1.6864041440110211, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 32969 + }, + { + "epoch": 0.3297, + "grad_norm": 1.6467192798995163, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 32970 + }, + { + "epoch": 0.32971, + "grad_norm": 1.079042255751393, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 32971 + }, + { + "epoch": 0.32972, + "grad_norm": 2.0894129118002263, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 32972 + }, + { + "epoch": 0.32973, + "grad_norm": 1.2740349245188483, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 32973 + }, + { + "epoch": 0.32974, + "grad_norm": 1.6228580934777737, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 32974 + }, + { + "epoch": 0.32975, + "grad_norm": 1.4146932063566464, + "learning_rate": 0.003, + "loss": 4.06, + "step": 32975 + }, + { + "epoch": 0.32976, + "grad_norm": 1.3902564397653032, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 32976 + }, + { + "epoch": 0.32977, + "grad_norm": 1.5157844157462361, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 32977 + }, + { + "epoch": 0.32978, + "grad_norm": 1.5764805929738175, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 32978 + }, + { + "epoch": 0.32979, + "grad_norm": 1.3944446762151066, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 32979 + }, + { + "epoch": 0.3298, + "grad_norm": 1.6353629879366571, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 32980 + }, + { + "epoch": 0.32981, + "grad_norm": 1.4637284324677757, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 32981 + }, + { + "epoch": 0.32982, + "grad_norm": 1.7679948052223975, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 32982 + }, + { + "epoch": 0.32983, + "grad_norm": 1.643732238069666, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 32983 + }, + { + "epoch": 0.32984, + "grad_norm": 1.612327668598258, + "learning_rate": 0.003, + "loss": 4.058, + "step": 32984 + }, + { + "epoch": 0.32985, + "grad_norm": 1.1396877526285214, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 32985 + }, + { + "epoch": 0.32986, + "grad_norm": 1.553448378393188, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 32986 + }, + { + "epoch": 0.32987, + "grad_norm": 1.201408818370931, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 32987 + }, + { + "epoch": 0.32988, + "grad_norm": 1.567312002109541, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 32988 + }, + { + "epoch": 0.32989, + "grad_norm": 1.0166658733551492, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 32989 + }, + { + "epoch": 0.3299, + "grad_norm": 1.8977142422653765, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 32990 + }, + { + "epoch": 0.32991, + "grad_norm": 1.3110238061317219, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 32991 + }, + { + "epoch": 0.32992, + "grad_norm": 1.7057297957160817, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 32992 + }, + { + "epoch": 0.32993, + "grad_norm": 1.262665504609876, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 32993 + }, + { + "epoch": 0.32994, + "grad_norm": 1.725479407048537, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 32994 + }, + { + "epoch": 0.32995, + "grad_norm": 1.581664956913769, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 32995 + }, + { + "epoch": 0.32996, + "grad_norm": 1.4452527120030327, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 32996 + }, + { + "epoch": 0.32997, + "grad_norm": 1.5451450381175071, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 32997 + }, + { + "epoch": 0.32998, + "grad_norm": 1.5093746255706044, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 32998 + }, + { + "epoch": 0.32999, + "grad_norm": 1.47847526475648, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 32999 + }, + { + "epoch": 0.33, + "grad_norm": 1.4737165452865213, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33000 + }, + { + "epoch": 0.33001, + "grad_norm": 1.4971163906033162, + "learning_rate": 0.003, + "loss": 4.05, + "step": 33001 + }, + { + "epoch": 0.33002, + "grad_norm": 1.4245037432147094, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 33002 + }, + { + "epoch": 0.33003, + "grad_norm": 1.4189073746797782, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 33003 + }, + { + "epoch": 0.33004, + "grad_norm": 1.7117543221628675, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 33004 + }, + { + "epoch": 0.33005, + "grad_norm": 1.3983606420819576, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 33005 + }, + { + "epoch": 0.33006, + "grad_norm": 1.5568638058123774, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 33006 + }, + { + "epoch": 0.33007, + "grad_norm": 1.464056130626441, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 33007 + }, + { + "epoch": 0.33008, + "grad_norm": 1.4267379330320935, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 33008 + }, + { + "epoch": 0.33009, + "grad_norm": 1.4143436811455434, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 33009 + }, + { + "epoch": 0.3301, + "grad_norm": 1.3046787377545175, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 33010 + }, + { + "epoch": 0.33011, + "grad_norm": 1.6410558628296859, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33011 + }, + { + "epoch": 0.33012, + "grad_norm": 1.602506246971521, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33012 + }, + { + "epoch": 0.33013, + "grad_norm": 1.262042833174165, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 33013 + }, + { + "epoch": 0.33014, + "grad_norm": 1.6838993749308948, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 33014 + }, + { + "epoch": 0.33015, + "grad_norm": 1.2536632781601096, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 33015 + }, + { + "epoch": 0.33016, + "grad_norm": 1.6003960509217365, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 33016 + }, + { + "epoch": 0.33017, + "grad_norm": 1.327738611888329, + "learning_rate": 0.003, + "loss": 4.024, + "step": 33017 + }, + { + "epoch": 0.33018, + "grad_norm": 1.604490321309371, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 33018 + }, + { + "epoch": 0.33019, + "grad_norm": 1.4194650845742882, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33019 + }, + { + "epoch": 0.3302, + "grad_norm": 1.312316627461949, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33020 + }, + { + "epoch": 0.33021, + "grad_norm": 1.7670303424897702, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 33021 + }, + { + "epoch": 0.33022, + "grad_norm": 1.1932982755051316, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 33022 + }, + { + "epoch": 0.33023, + "grad_norm": 2.1476415835591136, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 33023 + }, + { + "epoch": 0.33024, + "grad_norm": 1.301765204657242, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33024 + }, + { + "epoch": 0.33025, + "grad_norm": 1.3654717723859366, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 33025 + }, + { + "epoch": 0.33026, + "grad_norm": 1.482464810191581, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 33026 + }, + { + "epoch": 0.33027, + "grad_norm": 1.462364626041765, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 33027 + }, + { + "epoch": 0.33028, + "grad_norm": 1.5825345218138773, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 33028 + }, + { + "epoch": 0.33029, + "grad_norm": 1.5761836453507965, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33029 + }, + { + "epoch": 0.3303, + "grad_norm": 1.4160927786678315, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 33030 + }, + { + "epoch": 0.33031, + "grad_norm": 1.4173255644093183, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33031 + }, + { + "epoch": 0.33032, + "grad_norm": 1.4578083499647247, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 33032 + }, + { + "epoch": 0.33033, + "grad_norm": 1.499603459003407, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 33033 + }, + { + "epoch": 0.33034, + "grad_norm": 1.2363802218973952, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 33034 + }, + { + "epoch": 0.33035, + "grad_norm": 1.3467344256796527, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 33035 + }, + { + "epoch": 0.33036, + "grad_norm": 1.2677162685549312, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 33036 + }, + { + "epoch": 0.33037, + "grad_norm": 1.6637198014651928, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 33037 + }, + { + "epoch": 0.33038, + "grad_norm": 1.6956129754746134, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 33038 + }, + { + "epoch": 0.33039, + "grad_norm": 1.3792791277507297, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33039 + }, + { + "epoch": 0.3304, + "grad_norm": 1.84443075846854, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 33040 + }, + { + "epoch": 0.33041, + "grad_norm": 1.5808884958195764, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 33041 + }, + { + "epoch": 0.33042, + "grad_norm": 1.3241788339137923, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 33042 + }, + { + "epoch": 0.33043, + "grad_norm": 1.6230490650213534, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 33043 + }, + { + "epoch": 0.33044, + "grad_norm": 1.4023901019213194, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 33044 + }, + { + "epoch": 0.33045, + "grad_norm": 1.1669641071641053, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 33045 + }, + { + "epoch": 0.33046, + "grad_norm": 1.5699962515457668, + "learning_rate": 0.003, + "loss": 4.019, + "step": 33046 + }, + { + "epoch": 0.33047, + "grad_norm": 1.6302298878594348, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 33047 + }, + { + "epoch": 0.33048, + "grad_norm": 1.4706521872372222, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33048 + }, + { + "epoch": 0.33049, + "grad_norm": 1.5670892185714178, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 33049 + }, + { + "epoch": 0.3305, + "grad_norm": 1.4929369316071737, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 33050 + }, + { + "epoch": 0.33051, + "grad_norm": 1.182343867329554, + "learning_rate": 0.003, + "loss": 4.005, + "step": 33051 + }, + { + "epoch": 0.33052, + "grad_norm": 1.9708609463064561, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 33052 + }, + { + "epoch": 0.33053, + "grad_norm": 1.1784471138483061, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 33053 + }, + { + "epoch": 0.33054, + "grad_norm": 1.541416647075535, + "learning_rate": 0.003, + "loss": 4.059, + "step": 33054 + }, + { + "epoch": 0.33055, + "grad_norm": 1.606091538443571, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 33055 + }, + { + "epoch": 0.33056, + "grad_norm": 1.7339388187952798, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33056 + }, + { + "epoch": 0.33057, + "grad_norm": 1.3576052420986626, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 33057 + }, + { + "epoch": 0.33058, + "grad_norm": 1.7040005250439265, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 33058 + }, + { + "epoch": 0.33059, + "grad_norm": 1.3061402459645797, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33059 + }, + { + "epoch": 0.3306, + "grad_norm": 1.3455109956059828, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33060 + }, + { + "epoch": 0.33061, + "grad_norm": 1.3910126715814726, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33061 + }, + { + "epoch": 0.33062, + "grad_norm": 1.534600823512123, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33062 + }, + { + "epoch": 0.33063, + "grad_norm": 1.3467852141813859, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33063 + }, + { + "epoch": 0.33064, + "grad_norm": 1.8265264135371793, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 33064 + }, + { + "epoch": 0.33065, + "grad_norm": 1.0898993445927088, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 33065 + }, + { + "epoch": 0.33066, + "grad_norm": 1.8372793932637133, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 33066 + }, + { + "epoch": 0.33067, + "grad_norm": 1.2944399708688257, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 33067 + }, + { + "epoch": 0.33068, + "grad_norm": 1.6847765202207734, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 33068 + }, + { + "epoch": 0.33069, + "grad_norm": 1.2239585355656089, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33069 + }, + { + "epoch": 0.3307, + "grad_norm": 1.6628083094826265, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 33070 + }, + { + "epoch": 0.33071, + "grad_norm": 1.3045777910236929, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33071 + }, + { + "epoch": 0.33072, + "grad_norm": 1.3326824707251468, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 33072 + }, + { + "epoch": 0.33073, + "grad_norm": 1.6591974496379756, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33073 + }, + { + "epoch": 0.33074, + "grad_norm": 1.7311157483840698, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 33074 + }, + { + "epoch": 0.33075, + "grad_norm": 1.3078516935687616, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33075 + }, + { + "epoch": 0.33076, + "grad_norm": 1.7843587275169313, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 33076 + }, + { + "epoch": 0.33077, + "grad_norm": 1.164745045059604, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33077 + }, + { + "epoch": 0.33078, + "grad_norm": 1.8378505872260837, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33078 + }, + { + "epoch": 0.33079, + "grad_norm": 1.3009408600800476, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 33079 + }, + { + "epoch": 0.3308, + "grad_norm": 1.6887292296227772, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 33080 + }, + { + "epoch": 0.33081, + "grad_norm": 1.3798301792025145, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 33081 + }, + { + "epoch": 0.33082, + "grad_norm": 1.6964856212241075, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33082 + }, + { + "epoch": 0.33083, + "grad_norm": 1.4102088785657607, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33083 + }, + { + "epoch": 0.33084, + "grad_norm": 1.4168558373279325, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33084 + }, + { + "epoch": 0.33085, + "grad_norm": 1.4976853334919515, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 33085 + }, + { + "epoch": 0.33086, + "grad_norm": 1.2967838852275948, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 33086 + }, + { + "epoch": 0.33087, + "grad_norm": 1.2158331841233527, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33087 + }, + { + "epoch": 0.33088, + "grad_norm": 1.6189771224221614, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 33088 + }, + { + "epoch": 0.33089, + "grad_norm": 1.38372332872298, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 33089 + }, + { + "epoch": 0.3309, + "grad_norm": 1.7560896873524179, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 33090 + }, + { + "epoch": 0.33091, + "grad_norm": 1.565433657297068, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 33091 + }, + { + "epoch": 0.33092, + "grad_norm": 1.558891183331168, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 33092 + }, + { + "epoch": 0.33093, + "grad_norm": 1.4747886925114246, + "learning_rate": 0.003, + "loss": 4.056, + "step": 33093 + }, + { + "epoch": 0.33094, + "grad_norm": 1.433055576905928, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 33094 + }, + { + "epoch": 0.33095, + "grad_norm": 1.8149161107436058, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 33095 + }, + { + "epoch": 0.33096, + "grad_norm": 1.2292730264766782, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 33096 + }, + { + "epoch": 0.33097, + "grad_norm": 1.7505831173712338, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 33097 + }, + { + "epoch": 0.33098, + "grad_norm": 1.363971150610536, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 33098 + }, + { + "epoch": 0.33099, + "grad_norm": 1.6158326237475467, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 33099 + }, + { + "epoch": 0.331, + "grad_norm": 1.4953508746996127, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 33100 + }, + { + "epoch": 0.33101, + "grad_norm": 1.375560182912709, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33101 + }, + { + "epoch": 0.33102, + "grad_norm": 1.278208930796203, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33102 + }, + { + "epoch": 0.33103, + "grad_norm": 1.4443648003268668, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 33103 + }, + { + "epoch": 0.33104, + "grad_norm": 1.350415974561946, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 33104 + }, + { + "epoch": 0.33105, + "grad_norm": 1.4084810910085823, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 33105 + }, + { + "epoch": 0.33106, + "grad_norm": 1.4834062091724762, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33106 + }, + { + "epoch": 0.33107, + "grad_norm": 1.547416093582877, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 33107 + }, + { + "epoch": 0.33108, + "grad_norm": 1.948544492391234, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 33108 + }, + { + "epoch": 0.33109, + "grad_norm": 1.4688441500189, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33109 + }, + { + "epoch": 0.3311, + "grad_norm": 1.622475665421413, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33110 + }, + { + "epoch": 0.33111, + "grad_norm": 1.740810698871973, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 33111 + }, + { + "epoch": 0.33112, + "grad_norm": 1.17372082720912, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 33112 + }, + { + "epoch": 0.33113, + "grad_norm": 1.6110971736708177, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33113 + }, + { + "epoch": 0.33114, + "grad_norm": 1.1922550585930283, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33114 + }, + { + "epoch": 0.33115, + "grad_norm": 1.862140253660251, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33115 + }, + { + "epoch": 0.33116, + "grad_norm": 1.256231312667658, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 33116 + }, + { + "epoch": 0.33117, + "grad_norm": 1.5608836939991069, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 33117 + }, + { + "epoch": 0.33118, + "grad_norm": 1.2458263348564, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 33118 + }, + { + "epoch": 0.33119, + "grad_norm": 1.50414425961133, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 33119 + }, + { + "epoch": 0.3312, + "grad_norm": 1.5427289338113284, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33120 + }, + { + "epoch": 0.33121, + "grad_norm": 1.4820807820778963, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33121 + }, + { + "epoch": 0.33122, + "grad_norm": 1.5806337233550205, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 33122 + }, + { + "epoch": 0.33123, + "grad_norm": 1.481911112490676, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 33123 + }, + { + "epoch": 0.33124, + "grad_norm": 1.1789778839689617, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 33124 + }, + { + "epoch": 0.33125, + "grad_norm": 1.5063539130937076, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 33125 + }, + { + "epoch": 0.33126, + "grad_norm": 1.317049583453163, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 33126 + }, + { + "epoch": 0.33127, + "grad_norm": 1.3325634863794515, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 33127 + }, + { + "epoch": 0.33128, + "grad_norm": 1.4723511650501182, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 33128 + }, + { + "epoch": 0.33129, + "grad_norm": 1.6343929865243114, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 33129 + }, + { + "epoch": 0.3313, + "grad_norm": 1.7922390875589294, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 33130 + }, + { + "epoch": 0.33131, + "grad_norm": 1.3643904899639052, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33131 + }, + { + "epoch": 0.33132, + "grad_norm": 1.3926146444078298, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 33132 + }, + { + "epoch": 0.33133, + "grad_norm": 1.6744257902750779, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33133 + }, + { + "epoch": 0.33134, + "grad_norm": 1.1405483033283792, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 33134 + }, + { + "epoch": 0.33135, + "grad_norm": 1.889718134577996, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 33135 + }, + { + "epoch": 0.33136, + "grad_norm": 1.219160022706264, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 33136 + }, + { + "epoch": 0.33137, + "grad_norm": 1.5494993711170613, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 33137 + }, + { + "epoch": 0.33138, + "grad_norm": 1.3017622666561393, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 33138 + }, + { + "epoch": 0.33139, + "grad_norm": 1.3668527127417607, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 33139 + }, + { + "epoch": 0.3314, + "grad_norm": 1.4568562938633423, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 33140 + }, + { + "epoch": 0.33141, + "grad_norm": 1.446934424143021, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 33141 + }, + { + "epoch": 0.33142, + "grad_norm": 1.6033496633189321, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 33142 + }, + { + "epoch": 0.33143, + "grad_norm": 1.8227527214953994, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33143 + }, + { + "epoch": 0.33144, + "grad_norm": 1.5655111802721975, + "learning_rate": 0.003, + "loss": 4.059, + "step": 33144 + }, + { + "epoch": 0.33145, + "grad_norm": 1.299568852154415, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 33145 + }, + { + "epoch": 0.33146, + "grad_norm": 1.340737406715517, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33146 + }, + { + "epoch": 0.33147, + "grad_norm": 1.5822443573045182, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 33147 + }, + { + "epoch": 0.33148, + "grad_norm": 1.5074910039194824, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 33148 + }, + { + "epoch": 0.33149, + "grad_norm": 1.2855835467891288, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 33149 + }, + { + "epoch": 0.3315, + "grad_norm": 1.6962035949327992, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 33150 + }, + { + "epoch": 0.33151, + "grad_norm": 1.277921185120321, + "learning_rate": 0.003, + "loss": 4.048, + "step": 33151 + }, + { + "epoch": 0.33152, + "grad_norm": 1.6093428702238437, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33152 + }, + { + "epoch": 0.33153, + "grad_norm": 1.3717007295592434, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33153 + }, + { + "epoch": 0.33154, + "grad_norm": 1.566366015025939, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 33154 + }, + { + "epoch": 0.33155, + "grad_norm": 1.3402737749177254, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 33155 + }, + { + "epoch": 0.33156, + "grad_norm": 1.7733403701859236, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 33156 + }, + { + "epoch": 0.33157, + "grad_norm": 1.3378779820614963, + "learning_rate": 0.003, + "loss": 4.063, + "step": 33157 + }, + { + "epoch": 0.33158, + "grad_norm": 1.5507887765920165, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33158 + }, + { + "epoch": 0.33159, + "grad_norm": 1.4387782562206202, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 33159 + }, + { + "epoch": 0.3316, + "grad_norm": 1.747560218308632, + "learning_rate": 0.003, + "loss": 4.033, + "step": 33160 + }, + { + "epoch": 0.33161, + "grad_norm": 1.8015530337676497, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 33161 + }, + { + "epoch": 0.33162, + "grad_norm": 1.364293154143733, + "learning_rate": 0.003, + "loss": 4.063, + "step": 33162 + }, + { + "epoch": 0.33163, + "grad_norm": 1.6333642978631369, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 33163 + }, + { + "epoch": 0.33164, + "grad_norm": 1.5281786192471438, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33164 + }, + { + "epoch": 0.33165, + "grad_norm": 1.3858592439745785, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 33165 + }, + { + "epoch": 0.33166, + "grad_norm": 1.5393118558061007, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 33166 + }, + { + "epoch": 0.33167, + "grad_norm": 1.418190667032343, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 33167 + }, + { + "epoch": 0.33168, + "grad_norm": 1.411475741791294, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 33168 + }, + { + "epoch": 0.33169, + "grad_norm": 1.4376552175492072, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33169 + }, + { + "epoch": 0.3317, + "grad_norm": 1.5259791744375149, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 33170 + }, + { + "epoch": 0.33171, + "grad_norm": 1.4111349345966426, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 33171 + }, + { + "epoch": 0.33172, + "grad_norm": 1.328395650876137, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33172 + }, + { + "epoch": 0.33173, + "grad_norm": 1.5621617723124075, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 33173 + }, + { + "epoch": 0.33174, + "grad_norm": 1.3046527339688516, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33174 + }, + { + "epoch": 0.33175, + "grad_norm": 1.6136297023018558, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33175 + }, + { + "epoch": 0.33176, + "grad_norm": 1.5571320310743735, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 33176 + }, + { + "epoch": 0.33177, + "grad_norm": 1.5006126539343752, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 33177 + }, + { + "epoch": 0.33178, + "grad_norm": 1.2932275454295112, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 33178 + }, + { + "epoch": 0.33179, + "grad_norm": 1.6420942894616177, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 33179 + }, + { + "epoch": 0.3318, + "grad_norm": 1.340974906769641, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33180 + }, + { + "epoch": 0.33181, + "grad_norm": 1.8361580835946765, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 33181 + }, + { + "epoch": 0.33182, + "grad_norm": 1.3813783828608595, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33182 + }, + { + "epoch": 0.33183, + "grad_norm": 1.3913751378983326, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33183 + }, + { + "epoch": 0.33184, + "grad_norm": 1.733370366338208, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 33184 + }, + { + "epoch": 0.33185, + "grad_norm": 1.2204109306626618, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 33185 + }, + { + "epoch": 0.33186, + "grad_norm": 1.3922908785855002, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 33186 + }, + { + "epoch": 0.33187, + "grad_norm": 1.3742848614749639, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 33187 + }, + { + "epoch": 0.33188, + "grad_norm": 1.139851831216521, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 33188 + }, + { + "epoch": 0.33189, + "grad_norm": 1.8455673466503666, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 33189 + }, + { + "epoch": 0.3319, + "grad_norm": 1.1521556117643914, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 33190 + }, + { + "epoch": 0.33191, + "grad_norm": 1.7920973615715683, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 33191 + }, + { + "epoch": 0.33192, + "grad_norm": 1.2956785277860954, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 33192 + }, + { + "epoch": 0.33193, + "grad_norm": 2.0895464507622, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 33193 + }, + { + "epoch": 0.33194, + "grad_norm": 1.3789847749514592, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 33194 + }, + { + "epoch": 0.33195, + "grad_norm": 1.4118193601516662, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 33195 + }, + { + "epoch": 0.33196, + "grad_norm": 1.698951685306751, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 33196 + }, + { + "epoch": 0.33197, + "grad_norm": 1.506993865713899, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 33197 + }, + { + "epoch": 0.33198, + "grad_norm": 1.4028045930258692, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33198 + }, + { + "epoch": 0.33199, + "grad_norm": 1.5453931835541286, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 33199 + }, + { + "epoch": 0.332, + "grad_norm": 1.4130136156525699, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 33200 + }, + { + "epoch": 0.33201, + "grad_norm": 1.5851545982315707, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 33201 + }, + { + "epoch": 0.33202, + "grad_norm": 1.1700145705248288, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 33202 + }, + { + "epoch": 0.33203, + "grad_norm": 1.6765782855055567, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33203 + }, + { + "epoch": 0.33204, + "grad_norm": 1.3300916272238652, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33204 + }, + { + "epoch": 0.33205, + "grad_norm": 1.675139244750326, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 33205 + }, + { + "epoch": 0.33206, + "grad_norm": 1.4164546609900914, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33206 + }, + { + "epoch": 0.33207, + "grad_norm": 1.3966356321111502, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33207 + }, + { + "epoch": 0.33208, + "grad_norm": 1.3109984170777726, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 33208 + }, + { + "epoch": 0.33209, + "grad_norm": 1.4200043449059931, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 33209 + }, + { + "epoch": 0.3321, + "grad_norm": 1.6452714797705745, + "learning_rate": 0.003, + "loss": 4.035, + "step": 33210 + }, + { + "epoch": 0.33211, + "grad_norm": 1.425821051777862, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 33211 + }, + { + "epoch": 0.33212, + "grad_norm": 1.8334336294788849, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33212 + }, + { + "epoch": 0.33213, + "grad_norm": 1.8225671381783786, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 33213 + }, + { + "epoch": 0.33214, + "grad_norm": 1.079765710881302, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33214 + }, + { + "epoch": 0.33215, + "grad_norm": 1.504436538108778, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 33215 + }, + { + "epoch": 0.33216, + "grad_norm": 1.266482405049146, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33216 + }, + { + "epoch": 0.33217, + "grad_norm": 1.8225519861848456, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33217 + }, + { + "epoch": 0.33218, + "grad_norm": 1.223661446963973, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 33218 + }, + { + "epoch": 0.33219, + "grad_norm": 1.6118177839310663, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 33219 + }, + { + "epoch": 0.3322, + "grad_norm": 1.7100069467897023, + "learning_rate": 0.003, + "loss": 4.069, + "step": 33220 + }, + { + "epoch": 0.33221, + "grad_norm": 1.6050651853386937, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33221 + }, + { + "epoch": 0.33222, + "grad_norm": 1.7099545575179669, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 33222 + }, + { + "epoch": 0.33223, + "grad_norm": 1.2143114523187342, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33223 + }, + { + "epoch": 0.33224, + "grad_norm": 1.6851334825941253, + "learning_rate": 0.003, + "loss": 4.069, + "step": 33224 + }, + { + "epoch": 0.33225, + "grad_norm": 1.3193635486802129, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 33225 + }, + { + "epoch": 0.33226, + "grad_norm": 1.5227943644217306, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 33226 + }, + { + "epoch": 0.33227, + "grad_norm": 1.2411450553693737, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 33227 + }, + { + "epoch": 0.33228, + "grad_norm": 1.7187934720104776, + "learning_rate": 0.003, + "loss": 4.041, + "step": 33228 + }, + { + "epoch": 0.33229, + "grad_norm": 1.4697480712516182, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 33229 + }, + { + "epoch": 0.3323, + "grad_norm": 1.390449311660208, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 33230 + }, + { + "epoch": 0.33231, + "grad_norm": 1.4935079436106289, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 33231 + }, + { + "epoch": 0.33232, + "grad_norm": 1.7022347466936534, + "learning_rate": 0.003, + "loss": 4.077, + "step": 33232 + }, + { + "epoch": 0.33233, + "grad_norm": 1.3243775726386646, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33233 + }, + { + "epoch": 0.33234, + "grad_norm": 1.8174359023751936, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33234 + }, + { + "epoch": 0.33235, + "grad_norm": 1.07484902969692, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33235 + }, + { + "epoch": 0.33236, + "grad_norm": 1.6425221117433273, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 33236 + }, + { + "epoch": 0.33237, + "grad_norm": 1.3788070309454767, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 33237 + }, + { + "epoch": 0.33238, + "grad_norm": 1.7899308883406757, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33238 + }, + { + "epoch": 0.33239, + "grad_norm": 1.4676265001502746, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 33239 + }, + { + "epoch": 0.3324, + "grad_norm": 1.555555488117964, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 33240 + }, + { + "epoch": 0.33241, + "grad_norm": 1.5788040387663864, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 33241 + }, + { + "epoch": 0.33242, + "grad_norm": 1.4759005450928986, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33242 + }, + { + "epoch": 0.33243, + "grad_norm": 1.207480884245126, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33243 + }, + { + "epoch": 0.33244, + "grad_norm": 1.5300964810049915, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33244 + }, + { + "epoch": 0.33245, + "grad_norm": 1.2282614865923327, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 33245 + }, + { + "epoch": 0.33246, + "grad_norm": 1.3456369017819427, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 33246 + }, + { + "epoch": 0.33247, + "grad_norm": 1.5067465091610193, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33247 + }, + { + "epoch": 0.33248, + "grad_norm": 1.5019757747521882, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33248 + }, + { + "epoch": 0.33249, + "grad_norm": 1.6012680456347061, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 33249 + }, + { + "epoch": 0.3325, + "grad_norm": 1.392508470236499, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 33250 + }, + { + "epoch": 0.33251, + "grad_norm": 1.6921432217714125, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 33251 + }, + { + "epoch": 0.33252, + "grad_norm": 1.159019767744898, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33252 + }, + { + "epoch": 0.33253, + "grad_norm": 1.5489101754272365, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 33253 + }, + { + "epoch": 0.33254, + "grad_norm": 1.5178545163779313, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 33254 + }, + { + "epoch": 0.33255, + "grad_norm": 1.3601577058342618, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 33255 + }, + { + "epoch": 0.33256, + "grad_norm": 1.3976813147537914, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33256 + }, + { + "epoch": 0.33257, + "grad_norm": 1.5165637252797193, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 33257 + }, + { + "epoch": 0.33258, + "grad_norm": 1.2828797533209826, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33258 + }, + { + "epoch": 0.33259, + "grad_norm": 1.9542412535031444, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33259 + }, + { + "epoch": 0.3326, + "grad_norm": 1.4099045024140089, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 33260 + }, + { + "epoch": 0.33261, + "grad_norm": 1.7436352000789825, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 33261 + }, + { + "epoch": 0.33262, + "grad_norm": 1.3467818115547723, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 33262 + }, + { + "epoch": 0.33263, + "grad_norm": 1.3209495171360857, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 33263 + }, + { + "epoch": 0.33264, + "grad_norm": 1.6221176254073468, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33264 + }, + { + "epoch": 0.33265, + "grad_norm": 1.4070573514328228, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 33265 + }, + { + "epoch": 0.33266, + "grad_norm": 1.4878337234852657, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 33266 + }, + { + "epoch": 0.33267, + "grad_norm": 1.596508732027358, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33267 + }, + { + "epoch": 0.33268, + "grad_norm": 1.441407418239922, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33268 + }, + { + "epoch": 0.33269, + "grad_norm": 1.3490129516163094, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 33269 + }, + { + "epoch": 0.3327, + "grad_norm": 1.6437374840350585, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 33270 + }, + { + "epoch": 0.33271, + "grad_norm": 1.374782024320917, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 33271 + }, + { + "epoch": 0.33272, + "grad_norm": 1.4907976851783726, + "learning_rate": 0.003, + "loss": 4.044, + "step": 33272 + }, + { + "epoch": 0.33273, + "grad_norm": 1.3977382855820413, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 33273 + }, + { + "epoch": 0.33274, + "grad_norm": 1.5866965851796226, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33274 + }, + { + "epoch": 0.33275, + "grad_norm": 1.3658697987554294, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33275 + }, + { + "epoch": 0.33276, + "grad_norm": 1.4640585644756479, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 33276 + }, + { + "epoch": 0.33277, + "grad_norm": 1.5778719962006131, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33277 + }, + { + "epoch": 0.33278, + "grad_norm": 1.2294616677324488, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 33278 + }, + { + "epoch": 0.33279, + "grad_norm": 1.670596856519899, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 33279 + }, + { + "epoch": 0.3328, + "grad_norm": 1.3394155048693925, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 33280 + }, + { + "epoch": 0.33281, + "grad_norm": 1.4403669510779662, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 33281 + }, + { + "epoch": 0.33282, + "grad_norm": 1.4128168029975656, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 33282 + }, + { + "epoch": 0.33283, + "grad_norm": 1.4097403426067738, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 33283 + }, + { + "epoch": 0.33284, + "grad_norm": 1.5152784946797846, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 33284 + }, + { + "epoch": 0.33285, + "grad_norm": 1.4230045619453642, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33285 + }, + { + "epoch": 0.33286, + "grad_norm": 1.667159624635499, + "learning_rate": 0.003, + "loss": 4.046, + "step": 33286 + }, + { + "epoch": 0.33287, + "grad_norm": 1.5511486196626456, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 33287 + }, + { + "epoch": 0.33288, + "grad_norm": 1.3528545525098519, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33288 + }, + { + "epoch": 0.33289, + "grad_norm": 1.4617147501650027, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33289 + }, + { + "epoch": 0.3329, + "grad_norm": 1.3717835423735167, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 33290 + }, + { + "epoch": 0.33291, + "grad_norm": 1.3797437345271644, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 33291 + }, + { + "epoch": 0.33292, + "grad_norm": 1.5907347787688493, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33292 + }, + { + "epoch": 0.33293, + "grad_norm": 1.3519197840645354, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33293 + }, + { + "epoch": 0.33294, + "grad_norm": 1.5833325889921144, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 33294 + }, + { + "epoch": 0.33295, + "grad_norm": 1.2111463686673338, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 33295 + }, + { + "epoch": 0.33296, + "grad_norm": 1.574058301363517, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33296 + }, + { + "epoch": 0.33297, + "grad_norm": 1.2754685927365972, + "learning_rate": 0.003, + "loss": 4.016, + "step": 33297 + }, + { + "epoch": 0.33298, + "grad_norm": 1.4411418219579737, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 33298 + }, + { + "epoch": 0.33299, + "grad_norm": 1.765599414957739, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33299 + }, + { + "epoch": 0.333, + "grad_norm": 1.4974011944273953, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33300 + }, + { + "epoch": 0.33301, + "grad_norm": 1.8105522501494005, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 33301 + }, + { + "epoch": 0.33302, + "grad_norm": 1.2082825666052832, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33302 + }, + { + "epoch": 0.33303, + "grad_norm": 1.5656357826665668, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 33303 + }, + { + "epoch": 0.33304, + "grad_norm": 1.3220714315205382, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 33304 + }, + { + "epoch": 0.33305, + "grad_norm": 1.7737087622222258, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33305 + }, + { + "epoch": 0.33306, + "grad_norm": 1.4195040552973943, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 33306 + }, + { + "epoch": 0.33307, + "grad_norm": 1.5337128250358656, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 33307 + }, + { + "epoch": 0.33308, + "grad_norm": 1.4444792577269507, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 33308 + }, + { + "epoch": 0.33309, + "grad_norm": 1.4949820135079517, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 33309 + }, + { + "epoch": 0.3331, + "grad_norm": 1.9336877935571606, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 33310 + }, + { + "epoch": 0.33311, + "grad_norm": 1.2302259351401226, + "learning_rate": 0.003, + "loss": 4.055, + "step": 33311 + }, + { + "epoch": 0.33312, + "grad_norm": 1.448005157696977, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 33312 + }, + { + "epoch": 0.33313, + "grad_norm": 1.5808310411953932, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33313 + }, + { + "epoch": 0.33314, + "grad_norm": 1.4616977116463095, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 33314 + }, + { + "epoch": 0.33315, + "grad_norm": 1.3633639204205463, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 33315 + }, + { + "epoch": 0.33316, + "grad_norm": 1.6333978080673766, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 33316 + }, + { + "epoch": 0.33317, + "grad_norm": 1.2092099539089436, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 33317 + }, + { + "epoch": 0.33318, + "grad_norm": 1.561230958636716, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33318 + }, + { + "epoch": 0.33319, + "grad_norm": 1.4074271475816955, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 33319 + }, + { + "epoch": 0.3332, + "grad_norm": 2.0630373553985333, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 33320 + }, + { + "epoch": 0.33321, + "grad_norm": 1.5083404447826916, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33321 + }, + { + "epoch": 0.33322, + "grad_norm": 1.2991848247998583, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33322 + }, + { + "epoch": 0.33323, + "grad_norm": 1.4506213439157938, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 33323 + }, + { + "epoch": 0.33324, + "grad_norm": 1.5848509202205567, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 33324 + }, + { + "epoch": 0.33325, + "grad_norm": 1.2645845548600991, + "learning_rate": 0.003, + "loss": 4.029, + "step": 33325 + }, + { + "epoch": 0.33326, + "grad_norm": 1.5387382944453285, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 33326 + }, + { + "epoch": 0.33327, + "grad_norm": 1.5257302971673157, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 33327 + }, + { + "epoch": 0.33328, + "grad_norm": 1.5383369710722445, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 33328 + }, + { + "epoch": 0.33329, + "grad_norm": 1.5649065547207166, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 33329 + }, + { + "epoch": 0.3333, + "grad_norm": 1.5786051956674727, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 33330 + }, + { + "epoch": 0.33331, + "grad_norm": 1.7891068065131015, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33331 + }, + { + "epoch": 0.33332, + "grad_norm": 1.4327898516877884, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 33332 + }, + { + "epoch": 0.33333, + "grad_norm": 1.2321647092022865, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33333 + }, + { + "epoch": 0.33334, + "grad_norm": 2.065445582064233, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 33334 + }, + { + "epoch": 0.33335, + "grad_norm": 1.2783044426336567, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33335 + }, + { + "epoch": 0.33336, + "grad_norm": 1.8219568964506359, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 33336 + }, + { + "epoch": 0.33337, + "grad_norm": 1.434717854874047, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33337 + }, + { + "epoch": 0.33338, + "grad_norm": 1.53241520084464, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 33338 + }, + { + "epoch": 0.33339, + "grad_norm": 1.240526787394995, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33339 + }, + { + "epoch": 0.3334, + "grad_norm": 1.501091889899396, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 33340 + }, + { + "epoch": 0.33341, + "grad_norm": 1.3030592476907232, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 33341 + }, + { + "epoch": 0.33342, + "grad_norm": 1.6903242372693756, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 33342 + }, + { + "epoch": 0.33343, + "grad_norm": 1.5252541401984756, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 33343 + }, + { + "epoch": 0.33344, + "grad_norm": 1.4427798449162672, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33344 + }, + { + "epoch": 0.33345, + "grad_norm": 1.6791872879138219, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 33345 + }, + { + "epoch": 0.33346, + "grad_norm": 1.6788338368059108, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33346 + }, + { + "epoch": 0.33347, + "grad_norm": 1.3845185065821168, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33347 + }, + { + "epoch": 0.33348, + "grad_norm": 1.4247149615067023, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 33348 + }, + { + "epoch": 0.33349, + "grad_norm": 1.5370488169537266, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33349 + }, + { + "epoch": 0.3335, + "grad_norm": 1.4086726585300615, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33350 + }, + { + "epoch": 0.33351, + "grad_norm": 1.335510749349671, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 33351 + }, + { + "epoch": 0.33352, + "grad_norm": 1.433652024451877, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33352 + }, + { + "epoch": 0.33353, + "grad_norm": 1.4958496117643552, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 33353 + }, + { + "epoch": 0.33354, + "grad_norm": 1.4305101142688812, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 33354 + }, + { + "epoch": 0.33355, + "grad_norm": 1.5413651002243527, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 33355 + }, + { + "epoch": 0.33356, + "grad_norm": 1.5595441882140735, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33356 + }, + { + "epoch": 0.33357, + "grad_norm": 1.390310340067706, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33357 + }, + { + "epoch": 0.33358, + "grad_norm": 1.514211656093097, + "learning_rate": 0.003, + "loss": 4.043, + "step": 33358 + }, + { + "epoch": 0.33359, + "grad_norm": 1.5408715217389959, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 33359 + }, + { + "epoch": 0.3336, + "grad_norm": 1.385733178977651, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 33360 + }, + { + "epoch": 0.33361, + "grad_norm": 1.4764610172166863, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 33361 + }, + { + "epoch": 0.33362, + "grad_norm": 1.4989203762801837, + "learning_rate": 0.003, + "loss": 4.072, + "step": 33362 + }, + { + "epoch": 0.33363, + "grad_norm": 1.3284826206553646, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 33363 + }, + { + "epoch": 0.33364, + "grad_norm": 1.476006436099714, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 33364 + }, + { + "epoch": 0.33365, + "grad_norm": 1.389582624969669, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 33365 + }, + { + "epoch": 0.33366, + "grad_norm": 1.294529514786307, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33366 + }, + { + "epoch": 0.33367, + "grad_norm": 1.6333728603440316, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 33367 + }, + { + "epoch": 0.33368, + "grad_norm": 1.0756950260806784, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 33368 + }, + { + "epoch": 0.33369, + "grad_norm": 1.6724234262728759, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33369 + }, + { + "epoch": 0.3337, + "grad_norm": 1.3923566501758997, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 33370 + }, + { + "epoch": 0.33371, + "grad_norm": 1.6574098536110524, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 33371 + }, + { + "epoch": 0.33372, + "grad_norm": 2.0123713488479913, + "learning_rate": 0.003, + "loss": 4.04, + "step": 33372 + }, + { + "epoch": 0.33373, + "grad_norm": 1.4350345915860818, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 33373 + }, + { + "epoch": 0.33374, + "grad_norm": 1.6120125916776047, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 33374 + }, + { + "epoch": 0.33375, + "grad_norm": 2.0114225729243964, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 33375 + }, + { + "epoch": 0.33376, + "grad_norm": 1.0580851018592612, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 33376 + }, + { + "epoch": 0.33377, + "grad_norm": 1.595608285457347, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 33377 + }, + { + "epoch": 0.33378, + "grad_norm": 1.1364242839322172, + "learning_rate": 0.003, + "loss": 4.046, + "step": 33378 + }, + { + "epoch": 0.33379, + "grad_norm": 1.6229932429951117, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 33379 + }, + { + "epoch": 0.3338, + "grad_norm": 1.2088566243831969, + "learning_rate": 0.003, + "loss": 4.052, + "step": 33380 + }, + { + "epoch": 0.33381, + "grad_norm": 1.7256167775245947, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 33381 + }, + { + "epoch": 0.33382, + "grad_norm": 1.8585767944477365, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 33382 + }, + { + "epoch": 0.33383, + "grad_norm": 1.1768026283713406, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 33383 + }, + { + "epoch": 0.33384, + "grad_norm": 1.627034352401402, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 33384 + }, + { + "epoch": 0.33385, + "grad_norm": 1.4341803475280486, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 33385 + }, + { + "epoch": 0.33386, + "grad_norm": 1.4966842660916415, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 33386 + }, + { + "epoch": 0.33387, + "grad_norm": 1.3945697779107273, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 33387 + }, + { + "epoch": 0.33388, + "grad_norm": 1.6017716728624534, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 33388 + }, + { + "epoch": 0.33389, + "grad_norm": 1.3171203271435368, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33389 + }, + { + "epoch": 0.3339, + "grad_norm": 1.4071709508504273, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 33390 + }, + { + "epoch": 0.33391, + "grad_norm": 1.4989431112811473, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 33391 + }, + { + "epoch": 0.33392, + "grad_norm": 1.3879058204591324, + "learning_rate": 0.003, + "loss": 4.063, + "step": 33392 + }, + { + "epoch": 0.33393, + "grad_norm": 1.5489205812656053, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 33393 + }, + { + "epoch": 0.33394, + "grad_norm": 1.3114191393326011, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 33394 + }, + { + "epoch": 0.33395, + "grad_norm": 1.4881570018578458, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33395 + }, + { + "epoch": 0.33396, + "grad_norm": 1.4063246060666803, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33396 + }, + { + "epoch": 0.33397, + "grad_norm": 1.5609774109622203, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 33397 + }, + { + "epoch": 0.33398, + "grad_norm": 1.650016780484885, + "learning_rate": 0.003, + "loss": 4.031, + "step": 33398 + }, + { + "epoch": 0.33399, + "grad_norm": 1.461348191245201, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33399 + }, + { + "epoch": 0.334, + "grad_norm": 1.8300902501465752, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 33400 + }, + { + "epoch": 0.33401, + "grad_norm": 1.3158177496882162, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 33401 + }, + { + "epoch": 0.33402, + "grad_norm": 1.4041443328691208, + "learning_rate": 0.003, + "loss": 4.045, + "step": 33402 + }, + { + "epoch": 0.33403, + "grad_norm": 1.7192841800750014, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33403 + }, + { + "epoch": 0.33404, + "grad_norm": 1.597322601931171, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 33404 + }, + { + "epoch": 0.33405, + "grad_norm": 1.4782804807710157, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33405 + }, + { + "epoch": 0.33406, + "grad_norm": 1.4542698339698126, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 33406 + }, + { + "epoch": 0.33407, + "grad_norm": 1.5662591816952856, + "learning_rate": 0.003, + "loss": 4.067, + "step": 33407 + }, + { + "epoch": 0.33408, + "grad_norm": 1.4286166798418936, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33408 + }, + { + "epoch": 0.33409, + "grad_norm": 1.2204651040427628, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 33409 + }, + { + "epoch": 0.3341, + "grad_norm": 1.5641203726734874, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 33410 + }, + { + "epoch": 0.33411, + "grad_norm": 1.2580325329679412, + "learning_rate": 0.003, + "loss": 4.02, + "step": 33411 + }, + { + "epoch": 0.33412, + "grad_norm": 1.305478747381783, + "learning_rate": 0.003, + "loss": 4.031, + "step": 33412 + }, + { + "epoch": 0.33413, + "grad_norm": 1.3681176456977389, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33413 + }, + { + "epoch": 0.33414, + "grad_norm": 1.6408545927080525, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 33414 + }, + { + "epoch": 0.33415, + "grad_norm": 1.4281592031530346, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 33415 + }, + { + "epoch": 0.33416, + "grad_norm": 1.6304138028635766, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33416 + }, + { + "epoch": 0.33417, + "grad_norm": 1.5239729182076969, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33417 + }, + { + "epoch": 0.33418, + "grad_norm": 1.6172725743487797, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 33418 + }, + { + "epoch": 0.33419, + "grad_norm": 1.5244133808680993, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 33419 + }, + { + "epoch": 0.3342, + "grad_norm": 1.689879328111604, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 33420 + }, + { + "epoch": 0.33421, + "grad_norm": 1.5036926746941575, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 33421 + }, + { + "epoch": 0.33422, + "grad_norm": 1.4566233134033062, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 33422 + }, + { + "epoch": 0.33423, + "grad_norm": 1.4882095313494712, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 33423 + }, + { + "epoch": 0.33424, + "grad_norm": 1.7659768849614499, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33424 + }, + { + "epoch": 0.33425, + "grad_norm": 1.2397336336881928, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 33425 + }, + { + "epoch": 0.33426, + "grad_norm": 1.421024913279002, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 33426 + }, + { + "epoch": 0.33427, + "grad_norm": 1.5088520180896487, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33427 + }, + { + "epoch": 0.33428, + "grad_norm": 1.6929454815908216, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 33428 + }, + { + "epoch": 0.33429, + "grad_norm": 1.1018723795247045, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 33429 + }, + { + "epoch": 0.3343, + "grad_norm": 1.6150423081988012, + "learning_rate": 0.003, + "loss": 4.073, + "step": 33430 + }, + { + "epoch": 0.33431, + "grad_norm": 1.6054951565381586, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 33431 + }, + { + "epoch": 0.33432, + "grad_norm": 1.2970792838796352, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 33432 + }, + { + "epoch": 0.33433, + "grad_norm": 1.4340213038668859, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33433 + }, + { + "epoch": 0.33434, + "grad_norm": 1.477680205561617, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 33434 + }, + { + "epoch": 0.33435, + "grad_norm": 1.9270045612325208, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 33435 + }, + { + "epoch": 0.33436, + "grad_norm": 1.3465880528857384, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 33436 + }, + { + "epoch": 0.33437, + "grad_norm": 1.5791195315454052, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33437 + }, + { + "epoch": 0.33438, + "grad_norm": 1.5434284548039088, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 33438 + }, + { + "epoch": 0.33439, + "grad_norm": 1.353774742527388, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33439 + }, + { + "epoch": 0.3344, + "grad_norm": 1.5459555233717883, + "learning_rate": 0.003, + "loss": 4.051, + "step": 33440 + }, + { + "epoch": 0.33441, + "grad_norm": 1.1797546156205618, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 33441 + }, + { + "epoch": 0.33442, + "grad_norm": 1.7479429111272518, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 33442 + }, + { + "epoch": 0.33443, + "grad_norm": 1.3526990320000671, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 33443 + }, + { + "epoch": 0.33444, + "grad_norm": 1.7843197509821358, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 33444 + }, + { + "epoch": 0.33445, + "grad_norm": 1.4289540315647147, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33445 + }, + { + "epoch": 0.33446, + "grad_norm": 1.6086703153739186, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33446 + }, + { + "epoch": 0.33447, + "grad_norm": 1.4253369576343358, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 33447 + }, + { + "epoch": 0.33448, + "grad_norm": 1.6063156937388483, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 33448 + }, + { + "epoch": 0.33449, + "grad_norm": 1.656982561857423, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 33449 + }, + { + "epoch": 0.3345, + "grad_norm": 1.5679474134603653, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 33450 + }, + { + "epoch": 0.33451, + "grad_norm": 1.361452211788495, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33451 + }, + { + "epoch": 0.33452, + "grad_norm": 1.589191590638316, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33452 + }, + { + "epoch": 0.33453, + "grad_norm": 1.3613837104369393, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 33453 + }, + { + "epoch": 0.33454, + "grad_norm": 1.4017465605355768, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 33454 + }, + { + "epoch": 0.33455, + "grad_norm": 1.609173109200912, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33455 + }, + { + "epoch": 0.33456, + "grad_norm": 1.489710227527623, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 33456 + }, + { + "epoch": 0.33457, + "grad_norm": 1.7381855802752673, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 33457 + }, + { + "epoch": 0.33458, + "grad_norm": 1.2741126129549372, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33458 + }, + { + "epoch": 0.33459, + "grad_norm": 1.4975010962704067, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 33459 + }, + { + "epoch": 0.3346, + "grad_norm": 1.2097041486547833, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33460 + }, + { + "epoch": 0.33461, + "grad_norm": 1.7701520850181722, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33461 + }, + { + "epoch": 0.33462, + "grad_norm": 1.1233250863288742, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 33462 + }, + { + "epoch": 0.33463, + "grad_norm": 1.6930567119291113, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33463 + }, + { + "epoch": 0.33464, + "grad_norm": 1.3228683423602001, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33464 + }, + { + "epoch": 0.33465, + "grad_norm": 1.5122498927415855, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33465 + }, + { + "epoch": 0.33466, + "grad_norm": 1.4519100798489775, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 33466 + }, + { + "epoch": 0.33467, + "grad_norm": 1.558888828901578, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 33467 + }, + { + "epoch": 0.33468, + "grad_norm": 2.094325079016452, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 33468 + }, + { + "epoch": 0.33469, + "grad_norm": 1.354765535945746, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 33469 + }, + { + "epoch": 0.3347, + "grad_norm": 1.7234295972145877, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 33470 + }, + { + "epoch": 0.33471, + "grad_norm": 1.2980977677069339, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33471 + }, + { + "epoch": 0.33472, + "grad_norm": 1.390269696110806, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 33472 + }, + { + "epoch": 0.33473, + "grad_norm": 1.4961500331722828, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 33473 + }, + { + "epoch": 0.33474, + "grad_norm": 1.620965189437478, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 33474 + }, + { + "epoch": 0.33475, + "grad_norm": 1.3956815099195894, + "learning_rate": 0.003, + "loss": 4.041, + "step": 33475 + }, + { + "epoch": 0.33476, + "grad_norm": 1.5275531545748264, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33476 + }, + { + "epoch": 0.33477, + "grad_norm": 1.9266994035752305, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 33477 + }, + { + "epoch": 0.33478, + "grad_norm": 1.068553367260887, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 33478 + }, + { + "epoch": 0.33479, + "grad_norm": 1.5173629736692034, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 33479 + }, + { + "epoch": 0.3348, + "grad_norm": 1.6996693070364175, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 33480 + }, + { + "epoch": 0.33481, + "grad_norm": 1.5193771710725401, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 33481 + }, + { + "epoch": 0.33482, + "grad_norm": 1.3767623441098054, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 33482 + }, + { + "epoch": 0.33483, + "grad_norm": 1.3604304696006408, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 33483 + }, + { + "epoch": 0.33484, + "grad_norm": 1.8067485242340129, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 33484 + }, + { + "epoch": 0.33485, + "grad_norm": 1.2290062797605044, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 33485 + }, + { + "epoch": 0.33486, + "grad_norm": 1.7731174790722488, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33486 + }, + { + "epoch": 0.33487, + "grad_norm": 1.2512104620470277, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 33487 + }, + { + "epoch": 0.33488, + "grad_norm": 1.5426002342261091, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 33488 + }, + { + "epoch": 0.33489, + "grad_norm": 1.4355636549569253, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 33489 + }, + { + "epoch": 0.3349, + "grad_norm": 1.4475846845376423, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33490 + }, + { + "epoch": 0.33491, + "grad_norm": 1.552422680774398, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33491 + }, + { + "epoch": 0.33492, + "grad_norm": 1.4108437810978935, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 33492 + }, + { + "epoch": 0.33493, + "grad_norm": 1.3596149480692683, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 33493 + }, + { + "epoch": 0.33494, + "grad_norm": 1.4874277162317162, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33494 + }, + { + "epoch": 0.33495, + "grad_norm": 1.3778043231997397, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33495 + }, + { + "epoch": 0.33496, + "grad_norm": 1.4893611324320324, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 33496 + }, + { + "epoch": 0.33497, + "grad_norm": 1.365292810850449, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 33497 + }, + { + "epoch": 0.33498, + "grad_norm": 1.508895166084428, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 33498 + }, + { + "epoch": 0.33499, + "grad_norm": 1.581180436623798, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 33499 + }, + { + "epoch": 0.335, + "grad_norm": 1.5040057640714302, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 33500 + }, + { + "epoch": 0.33501, + "grad_norm": 1.3173643709936762, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 33501 + }, + { + "epoch": 0.33502, + "grad_norm": 1.4280668696220535, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 33502 + }, + { + "epoch": 0.33503, + "grad_norm": 1.454121900484076, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 33503 + }, + { + "epoch": 0.33504, + "grad_norm": 1.6986607605816344, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 33504 + }, + { + "epoch": 0.33505, + "grad_norm": 1.423552940358185, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33505 + }, + { + "epoch": 0.33506, + "grad_norm": 1.523756128488055, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 33506 + }, + { + "epoch": 0.33507, + "grad_norm": 1.3962689202494551, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 33507 + }, + { + "epoch": 0.33508, + "grad_norm": 1.5311731895070042, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 33508 + }, + { + "epoch": 0.33509, + "grad_norm": 1.6404928840866981, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 33509 + }, + { + "epoch": 0.3351, + "grad_norm": 1.2674056436970975, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 33510 + }, + { + "epoch": 0.33511, + "grad_norm": 1.9056904133223538, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 33511 + }, + { + "epoch": 0.33512, + "grad_norm": 1.1905978614296122, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 33512 + }, + { + "epoch": 0.33513, + "grad_norm": 1.7361535550308447, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33513 + }, + { + "epoch": 0.33514, + "grad_norm": 1.3497514518543745, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 33514 + }, + { + "epoch": 0.33515, + "grad_norm": 1.4595996308280823, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 33515 + }, + { + "epoch": 0.33516, + "grad_norm": 1.41056682570424, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33516 + }, + { + "epoch": 0.33517, + "grad_norm": 1.64249709866304, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 33517 + }, + { + "epoch": 0.33518, + "grad_norm": 1.7146629315752704, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 33518 + }, + { + "epoch": 0.33519, + "grad_norm": 1.566416622013584, + "learning_rate": 0.003, + "loss": 4.015, + "step": 33519 + }, + { + "epoch": 0.3352, + "grad_norm": 1.5209539883335148, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33520 + }, + { + "epoch": 0.33521, + "grad_norm": 1.585111999944207, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 33521 + }, + { + "epoch": 0.33522, + "grad_norm": 1.5758857705214973, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 33522 + }, + { + "epoch": 0.33523, + "grad_norm": 1.49240239015169, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33523 + }, + { + "epoch": 0.33524, + "grad_norm": 1.3660999042633208, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33524 + }, + { + "epoch": 0.33525, + "grad_norm": 1.4642303054281633, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33525 + }, + { + "epoch": 0.33526, + "grad_norm": 1.2535548886546661, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 33526 + }, + { + "epoch": 0.33527, + "grad_norm": 1.3059917970164037, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 33527 + }, + { + "epoch": 0.33528, + "grad_norm": 1.2779280905857013, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 33528 + }, + { + "epoch": 0.33529, + "grad_norm": 1.6279461660188668, + "learning_rate": 0.003, + "loss": 4.031, + "step": 33529 + }, + { + "epoch": 0.3353, + "grad_norm": 1.2853379235923383, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 33530 + }, + { + "epoch": 0.33531, + "grad_norm": 1.5499378179487238, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 33531 + }, + { + "epoch": 0.33532, + "grad_norm": 1.817599478788432, + "learning_rate": 0.003, + "loss": 4.047, + "step": 33532 + }, + { + "epoch": 0.33533, + "grad_norm": 1.736379158062309, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33533 + }, + { + "epoch": 0.33534, + "grad_norm": 1.3892372419191277, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 33534 + }, + { + "epoch": 0.33535, + "grad_norm": 1.2587651901350159, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33535 + }, + { + "epoch": 0.33536, + "grad_norm": 1.846211687086233, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 33536 + }, + { + "epoch": 0.33537, + "grad_norm": 1.2668153971041725, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 33537 + }, + { + "epoch": 0.33538, + "grad_norm": 1.6531705855918213, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 33538 + }, + { + "epoch": 0.33539, + "grad_norm": 1.2837741281286532, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33539 + }, + { + "epoch": 0.3354, + "grad_norm": 1.6684936421947645, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33540 + }, + { + "epoch": 0.33541, + "grad_norm": 1.1777652756285417, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33541 + }, + { + "epoch": 0.33542, + "grad_norm": 1.6362795949376427, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 33542 + }, + { + "epoch": 0.33543, + "grad_norm": 1.456965339219611, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 33543 + }, + { + "epoch": 0.33544, + "grad_norm": 1.6067532114464556, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 33544 + }, + { + "epoch": 0.33545, + "grad_norm": 1.434606506465584, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 33545 + }, + { + "epoch": 0.33546, + "grad_norm": 1.5289825454978536, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33546 + }, + { + "epoch": 0.33547, + "grad_norm": 1.287780202383561, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 33547 + }, + { + "epoch": 0.33548, + "grad_norm": 1.5293564504196229, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 33548 + }, + { + "epoch": 0.33549, + "grad_norm": 1.2104852010510423, + "learning_rate": 0.003, + "loss": 4.056, + "step": 33549 + }, + { + "epoch": 0.3355, + "grad_norm": 1.9446389578357746, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33550 + }, + { + "epoch": 0.33551, + "grad_norm": 1.2994255507891432, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 33551 + }, + { + "epoch": 0.33552, + "grad_norm": 1.682845425933207, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 33552 + }, + { + "epoch": 0.33553, + "grad_norm": 1.6374750675869463, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 33553 + }, + { + "epoch": 0.33554, + "grad_norm": 1.4357908687676595, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 33554 + }, + { + "epoch": 0.33555, + "grad_norm": 1.4847135561999996, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33555 + }, + { + "epoch": 0.33556, + "grad_norm": 1.5558449959890441, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33556 + }, + { + "epoch": 0.33557, + "grad_norm": 1.339097883380543, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 33557 + }, + { + "epoch": 0.33558, + "grad_norm": 1.4625459560853786, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33558 + }, + { + "epoch": 0.33559, + "grad_norm": 1.3754312687040522, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33559 + }, + { + "epoch": 0.3356, + "grad_norm": 1.5125156211554194, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33560 + }, + { + "epoch": 0.33561, + "grad_norm": 1.444572593318966, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 33561 + }, + { + "epoch": 0.33562, + "grad_norm": 1.626150674007393, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 33562 + }, + { + "epoch": 0.33563, + "grad_norm": 1.4498475205470946, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 33563 + }, + { + "epoch": 0.33564, + "grad_norm": 1.5053611972439411, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 33564 + }, + { + "epoch": 0.33565, + "grad_norm": 1.5985228975730563, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 33565 + }, + { + "epoch": 0.33566, + "grad_norm": 1.3333950503685967, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33566 + }, + { + "epoch": 0.33567, + "grad_norm": 1.390744167062581, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 33567 + }, + { + "epoch": 0.33568, + "grad_norm": 1.3080895577012508, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33568 + }, + { + "epoch": 0.33569, + "grad_norm": 1.3861707128047274, + "learning_rate": 0.003, + "loss": 4.055, + "step": 33569 + }, + { + "epoch": 0.3357, + "grad_norm": 1.5933924978535963, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 33570 + }, + { + "epoch": 0.33571, + "grad_norm": 1.7589259283501444, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33571 + }, + { + "epoch": 0.33572, + "grad_norm": 1.4134779037665133, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 33572 + }, + { + "epoch": 0.33573, + "grad_norm": 1.7186940460901345, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 33573 + }, + { + "epoch": 0.33574, + "grad_norm": 1.2448997668689035, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 33574 + }, + { + "epoch": 0.33575, + "grad_norm": 1.8332148664386612, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 33575 + }, + { + "epoch": 0.33576, + "grad_norm": 1.159716407146405, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 33576 + }, + { + "epoch": 0.33577, + "grad_norm": 1.6630214978206175, + "learning_rate": 0.003, + "loss": 4.044, + "step": 33577 + }, + { + "epoch": 0.33578, + "grad_norm": 1.4053561039206934, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 33578 + }, + { + "epoch": 0.33579, + "grad_norm": 1.4191996185395996, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 33579 + }, + { + "epoch": 0.3358, + "grad_norm": 1.463852093154677, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 33580 + }, + { + "epoch": 0.33581, + "grad_norm": 1.352697107499805, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 33581 + }, + { + "epoch": 0.33582, + "grad_norm": 1.533913651371521, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 33582 + }, + { + "epoch": 0.33583, + "grad_norm": 1.4312625039289197, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 33583 + }, + { + "epoch": 0.33584, + "grad_norm": 1.5601009871988678, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 33584 + }, + { + "epoch": 0.33585, + "grad_norm": 1.8001069743017846, + "learning_rate": 0.003, + "loss": 4.005, + "step": 33585 + }, + { + "epoch": 0.33586, + "grad_norm": 1.3544531827453736, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33586 + }, + { + "epoch": 0.33587, + "grad_norm": 1.5490774676323844, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 33587 + }, + { + "epoch": 0.33588, + "grad_norm": 1.396511880713688, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33588 + }, + { + "epoch": 0.33589, + "grad_norm": 1.4434353395338138, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33589 + }, + { + "epoch": 0.3359, + "grad_norm": 1.6372491138830318, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 33590 + }, + { + "epoch": 0.33591, + "grad_norm": 1.1752864815064714, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33591 + }, + { + "epoch": 0.33592, + "grad_norm": 1.66467782130713, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 33592 + }, + { + "epoch": 0.33593, + "grad_norm": 1.286672041512522, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33593 + }, + { + "epoch": 0.33594, + "grad_norm": 1.619766620433725, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 33594 + }, + { + "epoch": 0.33595, + "grad_norm": 1.5805799339854087, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 33595 + }, + { + "epoch": 0.33596, + "grad_norm": 1.4525535388815205, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 33596 + }, + { + "epoch": 0.33597, + "grad_norm": 1.4615962424556865, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 33597 + }, + { + "epoch": 0.33598, + "grad_norm": 1.6866632346904409, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 33598 + }, + { + "epoch": 0.33599, + "grad_norm": 1.5321696105138185, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 33599 + }, + { + "epoch": 0.336, + "grad_norm": 1.7058639159945725, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33600 + }, + { + "epoch": 0.33601, + "grad_norm": 1.3386823175564684, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 33601 + }, + { + "epoch": 0.33602, + "grad_norm": 1.5242953392475471, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33602 + }, + { + "epoch": 0.33603, + "grad_norm": 1.591343691766804, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 33603 + }, + { + "epoch": 0.33604, + "grad_norm": 1.7332373082071426, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 33604 + }, + { + "epoch": 0.33605, + "grad_norm": 1.2346086801329184, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 33605 + }, + { + "epoch": 0.33606, + "grad_norm": 1.7871912917680874, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 33606 + }, + { + "epoch": 0.33607, + "grad_norm": 1.357485845671261, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33607 + }, + { + "epoch": 0.33608, + "grad_norm": 1.514948025833918, + "learning_rate": 0.003, + "loss": 4.03, + "step": 33608 + }, + { + "epoch": 0.33609, + "grad_norm": 1.3177326613327573, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 33609 + }, + { + "epoch": 0.3361, + "grad_norm": 1.5818028057671254, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 33610 + }, + { + "epoch": 0.33611, + "grad_norm": 1.3736320632151138, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33611 + }, + { + "epoch": 0.33612, + "grad_norm": 1.3621372166312204, + "learning_rate": 0.003, + "loss": 4.055, + "step": 33612 + }, + { + "epoch": 0.33613, + "grad_norm": 1.3388285054781521, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33613 + }, + { + "epoch": 0.33614, + "grad_norm": 1.4576757588233695, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 33614 + }, + { + "epoch": 0.33615, + "grad_norm": 1.192516286099507, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 33615 + }, + { + "epoch": 0.33616, + "grad_norm": 1.8815691614586265, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 33616 + }, + { + "epoch": 0.33617, + "grad_norm": 1.3312022928293192, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33617 + }, + { + "epoch": 0.33618, + "grad_norm": 1.5629436063262818, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 33618 + }, + { + "epoch": 0.33619, + "grad_norm": 1.525455462679645, + "learning_rate": 0.003, + "loss": 4.052, + "step": 33619 + }, + { + "epoch": 0.3362, + "grad_norm": 1.7763012037759813, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33620 + }, + { + "epoch": 0.33621, + "grad_norm": 1.4245462751770954, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33621 + }, + { + "epoch": 0.33622, + "grad_norm": 1.3307830033868546, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 33622 + }, + { + "epoch": 0.33623, + "grad_norm": 1.523701304652321, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 33623 + }, + { + "epoch": 0.33624, + "grad_norm": 1.6876895540864483, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 33624 + }, + { + "epoch": 0.33625, + "grad_norm": 1.337703304085102, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 33625 + }, + { + "epoch": 0.33626, + "grad_norm": 1.7012151954242332, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 33626 + }, + { + "epoch": 0.33627, + "grad_norm": 1.2815379702514782, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 33627 + }, + { + "epoch": 0.33628, + "grad_norm": 1.5648263854139002, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 33628 + }, + { + "epoch": 0.33629, + "grad_norm": 1.379119582811406, + "learning_rate": 0.003, + "loss": 4.04, + "step": 33629 + }, + { + "epoch": 0.3363, + "grad_norm": 1.6048487598521812, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33630 + }, + { + "epoch": 0.33631, + "grad_norm": 1.3690665550717755, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 33631 + }, + { + "epoch": 0.33632, + "grad_norm": 1.6066814154534403, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33632 + }, + { + "epoch": 0.33633, + "grad_norm": 1.4359473156661644, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 33633 + }, + { + "epoch": 0.33634, + "grad_norm": 1.4972884729260776, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 33634 + }, + { + "epoch": 0.33635, + "grad_norm": 1.7596785874107237, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 33635 + }, + { + "epoch": 0.33636, + "grad_norm": 1.367295403970411, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33636 + }, + { + "epoch": 0.33637, + "grad_norm": 1.5735830596437517, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 33637 + }, + { + "epoch": 0.33638, + "grad_norm": 1.3435477030773515, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33638 + }, + { + "epoch": 0.33639, + "grad_norm": 1.7779550190334372, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 33639 + }, + { + "epoch": 0.3364, + "grad_norm": 1.435530262161628, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 33640 + }, + { + "epoch": 0.33641, + "grad_norm": 1.4943584231848284, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 33641 + }, + { + "epoch": 0.33642, + "grad_norm": 1.6551296532650792, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33642 + }, + { + "epoch": 0.33643, + "grad_norm": 1.255555841927531, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33643 + }, + { + "epoch": 0.33644, + "grad_norm": 1.3361642986038786, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 33644 + }, + { + "epoch": 0.33645, + "grad_norm": 1.2643174298230306, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33645 + }, + { + "epoch": 0.33646, + "grad_norm": 1.602502861975874, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33646 + }, + { + "epoch": 0.33647, + "grad_norm": 1.4509736650721183, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 33647 + }, + { + "epoch": 0.33648, + "grad_norm": 1.950955899409527, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33648 + }, + { + "epoch": 0.33649, + "grad_norm": 1.4891048319387439, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 33649 + }, + { + "epoch": 0.3365, + "grad_norm": 1.561518604072788, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 33650 + }, + { + "epoch": 0.33651, + "grad_norm": 1.374393755468386, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 33651 + }, + { + "epoch": 0.33652, + "grad_norm": 1.3650647752288103, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 33652 + }, + { + "epoch": 0.33653, + "grad_norm": 1.7297633272386839, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 33653 + }, + { + "epoch": 0.33654, + "grad_norm": 1.2062279462100796, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 33654 + }, + { + "epoch": 0.33655, + "grad_norm": 1.638407157434409, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 33655 + }, + { + "epoch": 0.33656, + "grad_norm": 1.4019058010408643, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 33656 + }, + { + "epoch": 0.33657, + "grad_norm": 1.4957356740946912, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33657 + }, + { + "epoch": 0.33658, + "grad_norm": 1.2823979835349257, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 33658 + }, + { + "epoch": 0.33659, + "grad_norm": 1.6274716659817605, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 33659 + }, + { + "epoch": 0.3366, + "grad_norm": 1.1613247008051568, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 33660 + }, + { + "epoch": 0.33661, + "grad_norm": 1.5711254007988242, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 33661 + }, + { + "epoch": 0.33662, + "grad_norm": 1.2747787152293262, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33662 + }, + { + "epoch": 0.33663, + "grad_norm": 1.5790977158808348, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 33663 + }, + { + "epoch": 0.33664, + "grad_norm": 1.4432103264697786, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 33664 + }, + { + "epoch": 0.33665, + "grad_norm": 2.0179723122990065, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 33665 + }, + { + "epoch": 0.33666, + "grad_norm": 1.5705477473484364, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33666 + }, + { + "epoch": 0.33667, + "grad_norm": 1.7562736950013444, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33667 + }, + { + "epoch": 0.33668, + "grad_norm": 1.5093604557624785, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 33668 + }, + { + "epoch": 0.33669, + "grad_norm": 1.4985715841954106, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33669 + }, + { + "epoch": 0.3367, + "grad_norm": 1.5726461581171933, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 33670 + }, + { + "epoch": 0.33671, + "grad_norm": 1.6056812653612547, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 33671 + }, + { + "epoch": 0.33672, + "grad_norm": 1.577282426427472, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 33672 + }, + { + "epoch": 0.33673, + "grad_norm": 1.570910415301165, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 33673 + }, + { + "epoch": 0.33674, + "grad_norm": 1.575792448800653, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 33674 + }, + { + "epoch": 0.33675, + "grad_norm": 1.4217335802056512, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 33675 + }, + { + "epoch": 0.33676, + "grad_norm": 1.4714749642032714, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 33676 + }, + { + "epoch": 0.33677, + "grad_norm": 1.4248344908686323, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 33677 + }, + { + "epoch": 0.33678, + "grad_norm": 1.3329479635208614, + "learning_rate": 0.003, + "loss": 4.044, + "step": 33678 + }, + { + "epoch": 0.33679, + "grad_norm": 1.6336944124262656, + "learning_rate": 0.003, + "loss": 4.046, + "step": 33679 + }, + { + "epoch": 0.3368, + "grad_norm": 1.319666866801993, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33680 + }, + { + "epoch": 0.33681, + "grad_norm": 1.8666806758281187, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33681 + }, + { + "epoch": 0.33682, + "grad_norm": 1.5548746052524736, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 33682 + }, + { + "epoch": 0.33683, + "grad_norm": 1.7577619531657722, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 33683 + }, + { + "epoch": 0.33684, + "grad_norm": 1.196737502260287, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33684 + }, + { + "epoch": 0.33685, + "grad_norm": 1.786309915158464, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 33685 + }, + { + "epoch": 0.33686, + "grad_norm": 1.2286435272700205, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 33686 + }, + { + "epoch": 0.33687, + "grad_norm": 1.577834166209912, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 33687 + }, + { + "epoch": 0.33688, + "grad_norm": 1.4582849704606726, + "learning_rate": 0.003, + "loss": 4.072, + "step": 33688 + }, + { + "epoch": 0.33689, + "grad_norm": 1.2780471067974557, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 33689 + }, + { + "epoch": 0.3369, + "grad_norm": 1.8210934836621548, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 33690 + }, + { + "epoch": 0.33691, + "grad_norm": 1.362481659564622, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 33691 + }, + { + "epoch": 0.33692, + "grad_norm": 1.628133990635747, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 33692 + }, + { + "epoch": 0.33693, + "grad_norm": 1.4421480287678081, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33693 + }, + { + "epoch": 0.33694, + "grad_norm": 1.637620983788089, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 33694 + }, + { + "epoch": 0.33695, + "grad_norm": 1.3531302443679585, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 33695 + }, + { + "epoch": 0.33696, + "grad_norm": 1.676665196357749, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 33696 + }, + { + "epoch": 0.33697, + "grad_norm": 1.341084241780648, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 33697 + }, + { + "epoch": 0.33698, + "grad_norm": 1.4809509367429214, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 33698 + }, + { + "epoch": 0.33699, + "grad_norm": 1.6230166336152454, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 33699 + }, + { + "epoch": 0.337, + "grad_norm": 1.375260327333431, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 33700 + }, + { + "epoch": 0.33701, + "grad_norm": 1.5821834004651185, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 33701 + }, + { + "epoch": 0.33702, + "grad_norm": 1.2602982008694823, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 33702 + }, + { + "epoch": 0.33703, + "grad_norm": 1.5891190618703055, + "learning_rate": 0.003, + "loss": 4.051, + "step": 33703 + }, + { + "epoch": 0.33704, + "grad_norm": 1.4383599262527071, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 33704 + }, + { + "epoch": 0.33705, + "grad_norm": 1.4274904388457044, + "learning_rate": 0.003, + "loss": 4.053, + "step": 33705 + }, + { + "epoch": 0.33706, + "grad_norm": 1.241514789700815, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 33706 + }, + { + "epoch": 0.33707, + "grad_norm": 1.64846832282001, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 33707 + }, + { + "epoch": 0.33708, + "grad_norm": 1.4370219989738533, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 33708 + }, + { + "epoch": 0.33709, + "grad_norm": 1.6960146952391677, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 33709 + }, + { + "epoch": 0.3371, + "grad_norm": 1.3719616598905824, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 33710 + }, + { + "epoch": 0.33711, + "grad_norm": 1.672875682445026, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 33711 + }, + { + "epoch": 0.33712, + "grad_norm": 1.2537484315047778, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 33712 + }, + { + "epoch": 0.33713, + "grad_norm": 1.5972005092084436, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33713 + }, + { + "epoch": 0.33714, + "grad_norm": 1.309328345539681, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 33714 + }, + { + "epoch": 0.33715, + "grad_norm": 1.6400704633418624, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33715 + }, + { + "epoch": 0.33716, + "grad_norm": 1.277806223560873, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33716 + }, + { + "epoch": 0.33717, + "grad_norm": 1.7725299475511376, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 33717 + }, + { + "epoch": 0.33718, + "grad_norm": 1.164963465758545, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 33718 + }, + { + "epoch": 0.33719, + "grad_norm": 2.44501907917649, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33719 + }, + { + "epoch": 0.3372, + "grad_norm": 1.4105915500158381, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 33720 + }, + { + "epoch": 0.33721, + "grad_norm": 1.515754020293978, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33721 + }, + { + "epoch": 0.33722, + "grad_norm": 1.5111039682445286, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33722 + }, + { + "epoch": 0.33723, + "grad_norm": 1.4376658949234247, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 33723 + }, + { + "epoch": 0.33724, + "grad_norm": 1.3044201172959387, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 33724 + }, + { + "epoch": 0.33725, + "grad_norm": 1.4289184659589802, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 33725 + }, + { + "epoch": 0.33726, + "grad_norm": 1.5497097351701143, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33726 + }, + { + "epoch": 0.33727, + "grad_norm": 1.7514428195105542, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 33727 + }, + { + "epoch": 0.33728, + "grad_norm": 1.4811200990539233, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33728 + }, + { + "epoch": 0.33729, + "grad_norm": 1.8668183785731352, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 33729 + }, + { + "epoch": 0.3373, + "grad_norm": 1.3623700638528085, + "learning_rate": 0.003, + "loss": 4.049, + "step": 33730 + }, + { + "epoch": 0.33731, + "grad_norm": 1.5070383659945306, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 33731 + }, + { + "epoch": 0.33732, + "grad_norm": 1.6593206688125581, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 33732 + }, + { + "epoch": 0.33733, + "grad_norm": 1.0827127631526259, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 33733 + }, + { + "epoch": 0.33734, + "grad_norm": 1.7188595506095752, + "learning_rate": 0.003, + "loss": 4.058, + "step": 33734 + }, + { + "epoch": 0.33735, + "grad_norm": 1.3230342673061295, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 33735 + }, + { + "epoch": 0.33736, + "grad_norm": 1.6201840885050296, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33736 + }, + { + "epoch": 0.33737, + "grad_norm": 1.4694167646703447, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 33737 + }, + { + "epoch": 0.33738, + "grad_norm": 1.3256807371193202, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33738 + }, + { + "epoch": 0.33739, + "grad_norm": 1.4938158518140192, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 33739 + }, + { + "epoch": 0.3374, + "grad_norm": 1.420009353522172, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 33740 + }, + { + "epoch": 0.33741, + "grad_norm": 1.4512036557334302, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 33741 + }, + { + "epoch": 0.33742, + "grad_norm": 1.5606135606474785, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 33742 + }, + { + "epoch": 0.33743, + "grad_norm": 1.3635648307369748, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33743 + }, + { + "epoch": 0.33744, + "grad_norm": 1.6905890149530778, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33744 + }, + { + "epoch": 0.33745, + "grad_norm": 1.5425750631599953, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 33745 + }, + { + "epoch": 0.33746, + "grad_norm": 1.525758725354352, + "learning_rate": 0.003, + "loss": 4.039, + "step": 33746 + }, + { + "epoch": 0.33747, + "grad_norm": 1.4376132476559942, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 33747 + }, + { + "epoch": 0.33748, + "grad_norm": 1.669488680747202, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33748 + }, + { + "epoch": 0.33749, + "grad_norm": 1.3525185299901525, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 33749 + }, + { + "epoch": 0.3375, + "grad_norm": 1.5758292465991968, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 33750 + }, + { + "epoch": 0.33751, + "grad_norm": 1.4123997999175313, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 33751 + }, + { + "epoch": 0.33752, + "grad_norm": 1.353939976477273, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 33752 + }, + { + "epoch": 0.33753, + "grad_norm": 1.373397165076527, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 33753 + }, + { + "epoch": 0.33754, + "grad_norm": 1.5885161894085362, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 33754 + }, + { + "epoch": 0.33755, + "grad_norm": 1.3566380153380224, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 33755 + }, + { + "epoch": 0.33756, + "grad_norm": 1.6050315721107908, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 33756 + }, + { + "epoch": 0.33757, + "grad_norm": 1.4363866768500195, + "learning_rate": 0.003, + "loss": 4.042, + "step": 33757 + }, + { + "epoch": 0.33758, + "grad_norm": 1.5437632761060314, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 33758 + }, + { + "epoch": 0.33759, + "grad_norm": 2.0625871528095163, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 33759 + }, + { + "epoch": 0.3376, + "grad_norm": 1.324408085980304, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 33760 + }, + { + "epoch": 0.33761, + "grad_norm": 1.5613775830673913, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 33761 + }, + { + "epoch": 0.33762, + "grad_norm": 1.2892102779189853, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 33762 + }, + { + "epoch": 0.33763, + "grad_norm": 1.6683997733495126, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 33763 + }, + { + "epoch": 0.33764, + "grad_norm": 1.2848396805853766, + "learning_rate": 0.003, + "loss": 4.038, + "step": 33764 + }, + { + "epoch": 0.33765, + "grad_norm": 1.6904087411312974, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 33765 + }, + { + "epoch": 0.33766, + "grad_norm": 1.283501636998295, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 33766 + }, + { + "epoch": 0.33767, + "grad_norm": 1.586692122375109, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 33767 + }, + { + "epoch": 0.33768, + "grad_norm": 1.3686945969992506, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 33768 + }, + { + "epoch": 0.33769, + "grad_norm": 1.401686406351377, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 33769 + }, + { + "epoch": 0.3377, + "grad_norm": 1.3755513769805003, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 33770 + }, + { + "epoch": 0.33771, + "grad_norm": 1.8311694396771006, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 33771 + }, + { + "epoch": 0.33772, + "grad_norm": 1.9458217049269728, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 33772 + }, + { + "epoch": 0.33773, + "grad_norm": 1.2750033795353897, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33773 + }, + { + "epoch": 0.33774, + "grad_norm": 1.3823331265832657, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 33774 + }, + { + "epoch": 0.33775, + "grad_norm": 1.7927585052440895, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 33775 + }, + { + "epoch": 0.33776, + "grad_norm": 1.343756483626176, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 33776 + }, + { + "epoch": 0.33777, + "grad_norm": 1.5109247675803237, + "learning_rate": 0.003, + "loss": 4.033, + "step": 33777 + }, + { + "epoch": 0.33778, + "grad_norm": 1.648858712324945, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 33778 + }, + { + "epoch": 0.33779, + "grad_norm": 1.1834408631625732, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 33779 + }, + { + "epoch": 0.3378, + "grad_norm": 1.687614486354441, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 33780 + }, + { + "epoch": 0.33781, + "grad_norm": 1.2995321558853665, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 33781 + }, + { + "epoch": 0.33782, + "grad_norm": 1.6272636745150137, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33782 + }, + { + "epoch": 0.33783, + "grad_norm": 1.2516179875690336, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 33783 + }, + { + "epoch": 0.33784, + "grad_norm": 1.67393262741159, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 33784 + }, + { + "epoch": 0.33785, + "grad_norm": 1.207384365290344, + "learning_rate": 0.003, + "loss": 4.022, + "step": 33785 + }, + { + "epoch": 0.33786, + "grad_norm": 1.7923009135795294, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 33786 + }, + { + "epoch": 0.33787, + "grad_norm": 1.1770914557558276, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 33787 + }, + { + "epoch": 0.33788, + "grad_norm": 1.6180776565878079, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 33788 + }, + { + "epoch": 0.33789, + "grad_norm": 1.2864923629806448, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 33789 + }, + { + "epoch": 0.3379, + "grad_norm": 1.6701488311101675, + "learning_rate": 0.003, + "loss": 4.05, + "step": 33790 + }, + { + "epoch": 0.33791, + "grad_norm": 1.634626305063256, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 33791 + }, + { + "epoch": 0.33792, + "grad_norm": 1.632571387113249, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 33792 + }, + { + "epoch": 0.33793, + "grad_norm": 1.5354379550674362, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 33793 + }, + { + "epoch": 0.33794, + "grad_norm": 1.4906218819635615, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 33794 + }, + { + "epoch": 0.33795, + "grad_norm": 1.6671784283741156, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 33795 + }, + { + "epoch": 0.33796, + "grad_norm": 1.3599960220058225, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33796 + }, + { + "epoch": 0.33797, + "grad_norm": 1.4388947329712085, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 33797 + }, + { + "epoch": 0.33798, + "grad_norm": 1.2707841342226938, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 33798 + }, + { + "epoch": 0.33799, + "grad_norm": 1.5757474831189144, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 33799 + }, + { + "epoch": 0.338, + "grad_norm": 1.4824027729642397, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 33800 + }, + { + "epoch": 0.33801, + "grad_norm": 1.5218003329540242, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 33801 + }, + { + "epoch": 0.33802, + "grad_norm": 1.3576786368393143, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 33802 + }, + { + "epoch": 0.33803, + "grad_norm": 1.6296154819325164, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 33803 + }, + { + "epoch": 0.33804, + "grad_norm": 1.4714585432724887, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 33804 + }, + { + "epoch": 0.33805, + "grad_norm": 1.6021105511800675, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 33805 + }, + { + "epoch": 0.33806, + "grad_norm": 1.374525408330484, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 33806 + }, + { + "epoch": 0.33807, + "grad_norm": 1.6592444885922064, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 33807 + }, + { + "epoch": 0.33808, + "grad_norm": 1.1830210959220782, + "learning_rate": 0.003, + "loss": 4.038, + "step": 33808 + }, + { + "epoch": 0.33809, + "grad_norm": 1.5360546930685957, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 33809 + }, + { + "epoch": 0.3381, + "grad_norm": 1.2404196504543785, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 33810 + }, + { + "epoch": 0.33811, + "grad_norm": 1.8083898022691012, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 33811 + }, + { + "epoch": 0.33812, + "grad_norm": 1.3103110730374317, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 33812 + }, + { + "epoch": 0.33813, + "grad_norm": 1.3789329077976842, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 33813 + }, + { + "epoch": 0.33814, + "grad_norm": 1.4090405994006958, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33814 + }, + { + "epoch": 0.33815, + "grad_norm": 1.67666173132623, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 33815 + }, + { + "epoch": 0.33816, + "grad_norm": 1.9788415552644862, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 33816 + }, + { + "epoch": 0.33817, + "grad_norm": 1.4430878750424407, + "learning_rate": 0.003, + "loss": 4.06, + "step": 33817 + }, + { + "epoch": 0.33818, + "grad_norm": 1.633906976970602, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 33818 + }, + { + "epoch": 0.33819, + "grad_norm": 1.4182216664597236, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 33819 + }, + { + "epoch": 0.3382, + "grad_norm": 1.3535593914172868, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 33820 + }, + { + "epoch": 0.33821, + "grad_norm": 1.407758252670597, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 33821 + }, + { + "epoch": 0.33822, + "grad_norm": 1.372591951356406, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 33822 + }, + { + "epoch": 0.33823, + "grad_norm": 1.597970942452673, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33823 + }, + { + "epoch": 0.33824, + "grad_norm": 1.4857453334084214, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 33824 + }, + { + "epoch": 0.33825, + "grad_norm": 1.506132131418191, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 33825 + }, + { + "epoch": 0.33826, + "grad_norm": 1.5070885375993013, + "learning_rate": 0.003, + "loss": 4.054, + "step": 33826 + }, + { + "epoch": 0.33827, + "grad_norm": 1.3636065325364743, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 33827 + }, + { + "epoch": 0.33828, + "grad_norm": 1.3687947772830003, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 33828 + }, + { + "epoch": 0.33829, + "grad_norm": 1.6352717873987848, + "learning_rate": 0.003, + "loss": 4.037, + "step": 33829 + }, + { + "epoch": 0.3383, + "grad_norm": 1.5189283724371008, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 33830 + }, + { + "epoch": 0.33831, + "grad_norm": 1.4214558560490853, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 33831 + }, + { + "epoch": 0.33832, + "grad_norm": 1.5422461910413816, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 33832 + }, + { + "epoch": 0.33833, + "grad_norm": 1.2870546120930382, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 33833 + }, + { + "epoch": 0.33834, + "grad_norm": 1.6005665851534479, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33834 + }, + { + "epoch": 0.33835, + "grad_norm": 1.4111604034686014, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 33835 + }, + { + "epoch": 0.33836, + "grad_norm": 1.8219760584123799, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 33836 + }, + { + "epoch": 0.33837, + "grad_norm": 1.044460633783337, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 33837 + }, + { + "epoch": 0.33838, + "grad_norm": 1.719899695177368, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33838 + }, + { + "epoch": 0.33839, + "grad_norm": 1.446182239213308, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 33839 + }, + { + "epoch": 0.3384, + "grad_norm": 1.822294185651693, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 33840 + }, + { + "epoch": 0.33841, + "grad_norm": 1.8026688528854662, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 33841 + }, + { + "epoch": 0.33842, + "grad_norm": 1.2322600946439868, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 33842 + }, + { + "epoch": 0.33843, + "grad_norm": 1.8986580526667123, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 33843 + }, + { + "epoch": 0.33844, + "grad_norm": 1.625122038732714, + "learning_rate": 0.003, + "loss": 4.051, + "step": 33844 + }, + { + "epoch": 0.33845, + "grad_norm": 1.3682615201766994, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33845 + }, + { + "epoch": 0.33846, + "grad_norm": 1.5955612995538786, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 33846 + }, + { + "epoch": 0.33847, + "grad_norm": 1.7329745271862147, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 33847 + }, + { + "epoch": 0.33848, + "grad_norm": 1.2218332541281955, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33848 + }, + { + "epoch": 0.33849, + "grad_norm": 1.6237537058829885, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 33849 + }, + { + "epoch": 0.3385, + "grad_norm": 1.3251348403343461, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 33850 + }, + { + "epoch": 0.33851, + "grad_norm": 1.6858754692225462, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 33851 + }, + { + "epoch": 0.33852, + "grad_norm": 1.368258616016516, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 33852 + }, + { + "epoch": 0.33853, + "grad_norm": 1.9151004562457095, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 33853 + }, + { + "epoch": 0.33854, + "grad_norm": 1.2736302042226686, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33854 + }, + { + "epoch": 0.33855, + "grad_norm": 1.1918021905072134, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 33855 + }, + { + "epoch": 0.33856, + "grad_norm": 1.5409913973714167, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 33856 + }, + { + "epoch": 0.33857, + "grad_norm": 1.3596029941527181, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33857 + }, + { + "epoch": 0.33858, + "grad_norm": 1.7234020178884784, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 33858 + }, + { + "epoch": 0.33859, + "grad_norm": 1.510258999279519, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 33859 + }, + { + "epoch": 0.3386, + "grad_norm": 1.390134371839058, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 33860 + }, + { + "epoch": 0.33861, + "grad_norm": 1.4756239289593915, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 33861 + }, + { + "epoch": 0.33862, + "grad_norm": 1.210484570035534, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 33862 + }, + { + "epoch": 0.33863, + "grad_norm": 1.5866529708557051, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 33863 + }, + { + "epoch": 0.33864, + "grad_norm": 1.1179278071144099, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33864 + }, + { + "epoch": 0.33865, + "grad_norm": 1.8941982385422615, + "learning_rate": 0.003, + "loss": 4.048, + "step": 33865 + }, + { + "epoch": 0.33866, + "grad_norm": 1.3282004940182444, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 33866 + }, + { + "epoch": 0.33867, + "grad_norm": 1.6764739556206791, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 33867 + }, + { + "epoch": 0.33868, + "grad_norm": 1.7006672353388317, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33868 + }, + { + "epoch": 0.33869, + "grad_norm": 1.5897068056941255, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 33869 + }, + { + "epoch": 0.3387, + "grad_norm": 1.2644963385496581, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33870 + }, + { + "epoch": 0.33871, + "grad_norm": 1.5838770165055058, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 33871 + }, + { + "epoch": 0.33872, + "grad_norm": 1.5914925805818527, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 33872 + }, + { + "epoch": 0.33873, + "grad_norm": 1.3958060813994262, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 33873 + }, + { + "epoch": 0.33874, + "grad_norm": 1.5046110555681227, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 33874 + }, + { + "epoch": 0.33875, + "grad_norm": 1.3469676326109594, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 33875 + }, + { + "epoch": 0.33876, + "grad_norm": 1.3916159230604885, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33876 + }, + { + "epoch": 0.33877, + "grad_norm": 1.3140542092062, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33877 + }, + { + "epoch": 0.33878, + "grad_norm": 1.4049567071157012, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 33878 + }, + { + "epoch": 0.33879, + "grad_norm": 1.279730393725256, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 33879 + }, + { + "epoch": 0.3388, + "grad_norm": 1.6851991249237606, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 33880 + }, + { + "epoch": 0.33881, + "grad_norm": 1.528884504610281, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 33881 + }, + { + "epoch": 0.33882, + "grad_norm": 1.5608649569778321, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 33882 + }, + { + "epoch": 0.33883, + "grad_norm": 1.7783125342118618, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 33883 + }, + { + "epoch": 0.33884, + "grad_norm": 1.9330610085999933, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 33884 + }, + { + "epoch": 0.33885, + "grad_norm": 1.1170116647905626, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 33885 + }, + { + "epoch": 0.33886, + "grad_norm": 1.7170876232524548, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 33886 + }, + { + "epoch": 0.33887, + "grad_norm": 1.9170543501142736, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 33887 + }, + { + "epoch": 0.33888, + "grad_norm": 1.69518927858405, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 33888 + }, + { + "epoch": 0.33889, + "grad_norm": 1.4844821988210652, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 33889 + }, + { + "epoch": 0.3389, + "grad_norm": 1.2035856921069452, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 33890 + }, + { + "epoch": 0.33891, + "grad_norm": 1.4707282432352482, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 33891 + }, + { + "epoch": 0.33892, + "grad_norm": 1.4990572374384994, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 33892 + }, + { + "epoch": 0.33893, + "grad_norm": 1.3241241725394708, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 33893 + }, + { + "epoch": 0.33894, + "grad_norm": 1.845679335912263, + "learning_rate": 0.003, + "loss": 4.064, + "step": 33894 + }, + { + "epoch": 0.33895, + "grad_norm": 1.8158343209631302, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 33895 + }, + { + "epoch": 0.33896, + "grad_norm": 1.527096099795052, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 33896 + }, + { + "epoch": 0.33897, + "grad_norm": 1.301124646307603, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 33897 + }, + { + "epoch": 0.33898, + "grad_norm": 1.9755600125265116, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 33898 + }, + { + "epoch": 0.33899, + "grad_norm": 1.0595465471483352, + "learning_rate": 0.003, + "loss": 4.06, + "step": 33899 + }, + { + "epoch": 0.339, + "grad_norm": 1.603382518773895, + "learning_rate": 0.003, + "loss": 4.04, + "step": 33900 + }, + { + "epoch": 0.33901, + "grad_norm": 1.4827541259281087, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33901 + }, + { + "epoch": 0.33902, + "grad_norm": 1.2964433918739509, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 33902 + }, + { + "epoch": 0.33903, + "grad_norm": 1.3008734998097433, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 33903 + }, + { + "epoch": 0.33904, + "grad_norm": 1.3534525079337496, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 33904 + }, + { + "epoch": 0.33905, + "grad_norm": 1.7709474977637858, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 33905 + }, + { + "epoch": 0.33906, + "grad_norm": 1.2446944238327182, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 33906 + }, + { + "epoch": 0.33907, + "grad_norm": 1.8077998372252855, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 33907 + }, + { + "epoch": 0.33908, + "grad_norm": 1.2164577521849023, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 33908 + }, + { + "epoch": 0.33909, + "grad_norm": 1.8533023112197635, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 33909 + }, + { + "epoch": 0.3391, + "grad_norm": 1.2213594911759953, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 33910 + }, + { + "epoch": 0.33911, + "grad_norm": 1.507329441427186, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 33911 + }, + { + "epoch": 0.33912, + "grad_norm": 1.885080979259471, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33912 + }, + { + "epoch": 0.33913, + "grad_norm": 1.6705822494196254, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 33913 + }, + { + "epoch": 0.33914, + "grad_norm": 1.5639964452642825, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 33914 + }, + { + "epoch": 0.33915, + "grad_norm": 1.272602256584958, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 33915 + }, + { + "epoch": 0.33916, + "grad_norm": 1.6164376757950425, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 33916 + }, + { + "epoch": 0.33917, + "grad_norm": 1.5105935141865083, + "learning_rate": 0.003, + "loss": 4.027, + "step": 33917 + }, + { + "epoch": 0.33918, + "grad_norm": 1.6334067966102948, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 33918 + }, + { + "epoch": 0.33919, + "grad_norm": 1.3607772058904177, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 33919 + }, + { + "epoch": 0.3392, + "grad_norm": 1.89674655150953, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 33920 + }, + { + "epoch": 0.33921, + "grad_norm": 0.9374509692570185, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 33921 + }, + { + "epoch": 0.33922, + "grad_norm": 1.891408534914725, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 33922 + }, + { + "epoch": 0.33923, + "grad_norm": 1.2187284452942189, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 33923 + }, + { + "epoch": 0.33924, + "grad_norm": 1.8823145011406168, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 33924 + }, + { + "epoch": 0.33925, + "grad_norm": 1.5990710257816443, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 33925 + }, + { + "epoch": 0.33926, + "grad_norm": 1.4147030023384493, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 33926 + }, + { + "epoch": 0.33927, + "grad_norm": 1.791282405584767, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 33927 + }, + { + "epoch": 0.33928, + "grad_norm": 1.2351872053098423, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 33928 + }, + { + "epoch": 0.33929, + "grad_norm": 1.4636545273189292, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 33929 + }, + { + "epoch": 0.3393, + "grad_norm": 1.4323580323936191, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 33930 + }, + { + "epoch": 0.33931, + "grad_norm": 1.4979543317428619, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 33931 + }, + { + "epoch": 0.33932, + "grad_norm": 1.4808847923176924, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33932 + }, + { + "epoch": 0.33933, + "grad_norm": 1.5299488333230196, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 33933 + }, + { + "epoch": 0.33934, + "grad_norm": 1.5482275170291033, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 33934 + }, + { + "epoch": 0.33935, + "grad_norm": 1.4341720383851915, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 33935 + }, + { + "epoch": 0.33936, + "grad_norm": 1.5210558164246262, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 33936 + }, + { + "epoch": 0.33937, + "grad_norm": 1.3644600670952571, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 33937 + }, + { + "epoch": 0.33938, + "grad_norm": 1.6681525812598574, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 33938 + }, + { + "epoch": 0.33939, + "grad_norm": 1.1226399567930105, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 33939 + }, + { + "epoch": 0.3394, + "grad_norm": 1.6910305098493235, + "learning_rate": 0.003, + "loss": 4.075, + "step": 33940 + }, + { + "epoch": 0.33941, + "grad_norm": 1.317567182988594, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 33941 + }, + { + "epoch": 0.33942, + "grad_norm": 1.6909089686091792, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 33942 + }, + { + "epoch": 0.33943, + "grad_norm": 1.5798486895778905, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 33943 + }, + { + "epoch": 0.33944, + "grad_norm": 1.6709280328517193, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 33944 + }, + { + "epoch": 0.33945, + "grad_norm": 1.39147976269779, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 33945 + }, + { + "epoch": 0.33946, + "grad_norm": 1.5523777683014557, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 33946 + }, + { + "epoch": 0.33947, + "grad_norm": 1.3349229622799366, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 33947 + }, + { + "epoch": 0.33948, + "grad_norm": 1.2489543087772144, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 33948 + }, + { + "epoch": 0.33949, + "grad_norm": 1.5154679133766096, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 33949 + }, + { + "epoch": 0.3395, + "grad_norm": 1.3391744729825126, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 33950 + }, + { + "epoch": 0.33951, + "grad_norm": 1.5148952029309526, + "learning_rate": 0.003, + "loss": 4.062, + "step": 33951 + }, + { + "epoch": 0.33952, + "grad_norm": 1.4219619331754803, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33952 + }, + { + "epoch": 0.33953, + "grad_norm": 1.4621164737985914, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 33953 + }, + { + "epoch": 0.33954, + "grad_norm": 1.6333140519969676, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 33954 + }, + { + "epoch": 0.33955, + "grad_norm": 1.8349030593452234, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 33955 + }, + { + "epoch": 0.33956, + "grad_norm": 1.289529873396207, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 33956 + }, + { + "epoch": 0.33957, + "grad_norm": 1.687020602592991, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 33957 + }, + { + "epoch": 0.33958, + "grad_norm": 1.5958030790180275, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 33958 + }, + { + "epoch": 0.33959, + "grad_norm": 1.6752305467992954, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 33959 + }, + { + "epoch": 0.3396, + "grad_norm": 1.2372576262825854, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 33960 + }, + { + "epoch": 0.33961, + "grad_norm": 1.507280588256371, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 33961 + }, + { + "epoch": 0.33962, + "grad_norm": 1.498513748074373, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 33962 + }, + { + "epoch": 0.33963, + "grad_norm": 1.280171573756553, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 33963 + }, + { + "epoch": 0.33964, + "grad_norm": 1.6208643785702512, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 33964 + }, + { + "epoch": 0.33965, + "grad_norm": 1.4033796721598877, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 33965 + }, + { + "epoch": 0.33966, + "grad_norm": 1.6975532371917423, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 33966 + }, + { + "epoch": 0.33967, + "grad_norm": 1.2274544981811977, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 33967 + }, + { + "epoch": 0.33968, + "grad_norm": 1.8936449697997169, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 33968 + }, + { + "epoch": 0.33969, + "grad_norm": 1.6464370369557706, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 33969 + }, + { + "epoch": 0.3397, + "grad_norm": 1.3882215560083462, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 33970 + }, + { + "epoch": 0.33971, + "grad_norm": 1.5055955585093022, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33971 + }, + { + "epoch": 0.33972, + "grad_norm": 1.6547701975479978, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 33972 + }, + { + "epoch": 0.33973, + "grad_norm": 1.183454533795308, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 33973 + }, + { + "epoch": 0.33974, + "grad_norm": 1.7906246154392877, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 33974 + }, + { + "epoch": 0.33975, + "grad_norm": 1.2601849825698748, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 33975 + }, + { + "epoch": 0.33976, + "grad_norm": 1.6508092533717318, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 33976 + }, + { + "epoch": 0.33977, + "grad_norm": 1.4958093331056928, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 33977 + }, + { + "epoch": 0.33978, + "grad_norm": 1.5996986929377295, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 33978 + }, + { + "epoch": 0.33979, + "grad_norm": 1.5237008152056408, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 33979 + }, + { + "epoch": 0.3398, + "grad_norm": 1.7811436680750452, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 33980 + }, + { + "epoch": 0.33981, + "grad_norm": 1.2688088398163813, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 33981 + }, + { + "epoch": 0.33982, + "grad_norm": 1.4591494153253892, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 33982 + }, + { + "epoch": 0.33983, + "grad_norm": 1.6396306766937423, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 33983 + }, + { + "epoch": 0.33984, + "grad_norm": 1.396649720423154, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 33984 + }, + { + "epoch": 0.33985, + "grad_norm": 1.3845829608256706, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 33985 + }, + { + "epoch": 0.33986, + "grad_norm": 1.5620105621163967, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 33986 + }, + { + "epoch": 0.33987, + "grad_norm": 1.1527093944025644, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33987 + }, + { + "epoch": 0.33988, + "grad_norm": 1.458253334833136, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 33988 + }, + { + "epoch": 0.33989, + "grad_norm": 1.3827646267820413, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 33989 + }, + { + "epoch": 0.3399, + "grad_norm": 1.6903174640701488, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 33990 + }, + { + "epoch": 0.33991, + "grad_norm": 1.7859954386524084, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 33991 + }, + { + "epoch": 0.33992, + "grad_norm": 1.5500593369890479, + "learning_rate": 0.003, + "loss": 4.046, + "step": 33992 + }, + { + "epoch": 0.33993, + "grad_norm": 1.494935842615547, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 33993 + }, + { + "epoch": 0.33994, + "grad_norm": 1.4751944059274833, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 33994 + }, + { + "epoch": 0.33995, + "grad_norm": 1.4534244678486632, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 33995 + }, + { + "epoch": 0.33996, + "grad_norm": 1.439426004571761, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 33996 + }, + { + "epoch": 0.33997, + "grad_norm": 1.4012127278120423, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 33997 + }, + { + "epoch": 0.33998, + "grad_norm": 1.5883268126769006, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 33998 + }, + { + "epoch": 0.33999, + "grad_norm": 1.295550657590464, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 33999 + }, + { + "epoch": 0.34, + "grad_norm": 1.4348179841163597, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 34000 + }, + { + "epoch": 0.34001, + "grad_norm": 1.4495482217292213, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 34001 + }, + { + "epoch": 0.34002, + "grad_norm": 1.6220304536033208, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34002 + }, + { + "epoch": 0.34003, + "grad_norm": 1.2303195274636618, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 34003 + }, + { + "epoch": 0.34004, + "grad_norm": 1.6160107621866224, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 34004 + }, + { + "epoch": 0.34005, + "grad_norm": 1.4913807797666878, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 34005 + }, + { + "epoch": 0.34006, + "grad_norm": 1.5812758287842554, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 34006 + }, + { + "epoch": 0.34007, + "grad_norm": 2.03588443112162, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 34007 + }, + { + "epoch": 0.34008, + "grad_norm": 1.2469161209241524, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34008 + }, + { + "epoch": 0.34009, + "grad_norm": 1.8514153367175479, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 34009 + }, + { + "epoch": 0.3401, + "grad_norm": 1.6694352418675178, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34010 + }, + { + "epoch": 0.34011, + "grad_norm": 1.1674759989970656, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34011 + }, + { + "epoch": 0.34012, + "grad_norm": 1.9400454515750978, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 34012 + }, + { + "epoch": 0.34013, + "grad_norm": 1.1858305714269997, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 34013 + }, + { + "epoch": 0.34014, + "grad_norm": 1.5578125042168425, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34014 + }, + { + "epoch": 0.34015, + "grad_norm": 1.5802294163330635, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 34015 + }, + { + "epoch": 0.34016, + "grad_norm": 1.349361085065218, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34016 + }, + { + "epoch": 0.34017, + "grad_norm": 1.513700540679448, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 34017 + }, + { + "epoch": 0.34018, + "grad_norm": 1.5878711041620353, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 34018 + }, + { + "epoch": 0.34019, + "grad_norm": 1.6040627932875486, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 34019 + }, + { + "epoch": 0.3402, + "grad_norm": 1.8100352557665056, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 34020 + }, + { + "epoch": 0.34021, + "grad_norm": 1.6755988829004664, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 34021 + }, + { + "epoch": 0.34022, + "grad_norm": 1.2500841556432511, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 34022 + }, + { + "epoch": 0.34023, + "grad_norm": 1.4546046826792636, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 34023 + }, + { + "epoch": 0.34024, + "grad_norm": 1.4642910849625272, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34024 + }, + { + "epoch": 0.34025, + "grad_norm": 1.554637993394431, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 34025 + }, + { + "epoch": 0.34026, + "grad_norm": 1.3705622884001945, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 34026 + }, + { + "epoch": 0.34027, + "grad_norm": 1.7034418267424483, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34027 + }, + { + "epoch": 0.34028, + "grad_norm": 1.4495821192938834, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34028 + }, + { + "epoch": 0.34029, + "grad_norm": 1.3159751030274869, + "learning_rate": 0.003, + "loss": 4.014, + "step": 34029 + }, + { + "epoch": 0.3403, + "grad_norm": 1.3092259172877554, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 34030 + }, + { + "epoch": 0.34031, + "grad_norm": 1.8729177297143427, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 34031 + }, + { + "epoch": 0.34032, + "grad_norm": 1.2825285454032085, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 34032 + }, + { + "epoch": 0.34033, + "grad_norm": 1.5794919974614945, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34033 + }, + { + "epoch": 0.34034, + "grad_norm": 1.313001590637404, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 34034 + }, + { + "epoch": 0.34035, + "grad_norm": 1.6395736325579846, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34035 + }, + { + "epoch": 0.34036, + "grad_norm": 1.3577014493588504, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34036 + }, + { + "epoch": 0.34037, + "grad_norm": 1.6457220561474792, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 34037 + }, + { + "epoch": 0.34038, + "grad_norm": 1.8367632489738546, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 34038 + }, + { + "epoch": 0.34039, + "grad_norm": 1.6374131262903384, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 34039 + }, + { + "epoch": 0.3404, + "grad_norm": 1.4089428193669096, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34040 + }, + { + "epoch": 0.34041, + "grad_norm": 1.5561653373380397, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 34041 + }, + { + "epoch": 0.34042, + "grad_norm": 1.5579034134592473, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 34042 + }, + { + "epoch": 0.34043, + "grad_norm": 1.1881235902714231, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34043 + }, + { + "epoch": 0.34044, + "grad_norm": 1.71016904596021, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 34044 + }, + { + "epoch": 0.34045, + "grad_norm": 1.365343899943106, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 34045 + }, + { + "epoch": 0.34046, + "grad_norm": 1.7815565199584538, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 34046 + }, + { + "epoch": 0.34047, + "grad_norm": 1.2874902625445743, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 34047 + }, + { + "epoch": 0.34048, + "grad_norm": 1.5912135818047484, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 34048 + }, + { + "epoch": 0.34049, + "grad_norm": 1.2401326014594831, + "learning_rate": 0.003, + "loss": 4.031, + "step": 34049 + }, + { + "epoch": 0.3405, + "grad_norm": 1.7409636093648646, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 34050 + }, + { + "epoch": 0.34051, + "grad_norm": 1.3628185355743023, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34051 + }, + { + "epoch": 0.34052, + "grad_norm": 1.2393964086788183, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 34052 + }, + { + "epoch": 0.34053, + "grad_norm": 1.7616732826494725, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 34053 + }, + { + "epoch": 0.34054, + "grad_norm": 1.122612263592818, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 34054 + }, + { + "epoch": 0.34055, + "grad_norm": 1.444721512159515, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 34055 + }, + { + "epoch": 0.34056, + "grad_norm": 1.4269571057591546, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 34056 + }, + { + "epoch": 0.34057, + "grad_norm": 1.607870300553402, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34057 + }, + { + "epoch": 0.34058, + "grad_norm": 1.5772915620340926, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 34058 + }, + { + "epoch": 0.34059, + "grad_norm": 1.6415860334455403, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34059 + }, + { + "epoch": 0.3406, + "grad_norm": 1.5117266281428992, + "learning_rate": 0.003, + "loss": 4.015, + "step": 34060 + }, + { + "epoch": 0.34061, + "grad_norm": 1.3186529437979664, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34061 + }, + { + "epoch": 0.34062, + "grad_norm": 1.5696541302838511, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34062 + }, + { + "epoch": 0.34063, + "grad_norm": 1.5286138223187284, + "learning_rate": 0.003, + "loss": 4.062, + "step": 34063 + }, + { + "epoch": 0.34064, + "grad_norm": 1.3955650290163897, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 34064 + }, + { + "epoch": 0.34065, + "grad_norm": 1.6516499502642668, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 34065 + }, + { + "epoch": 0.34066, + "grad_norm": 1.0998143854973923, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34066 + }, + { + "epoch": 0.34067, + "grad_norm": 1.711903875330929, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 34067 + }, + { + "epoch": 0.34068, + "grad_norm": 1.1019892863624445, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 34068 + }, + { + "epoch": 0.34069, + "grad_norm": 1.705805755020341, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 34069 + }, + { + "epoch": 0.3407, + "grad_norm": 1.1312650353498808, + "learning_rate": 0.003, + "loss": 4.064, + "step": 34070 + }, + { + "epoch": 0.34071, + "grad_norm": 1.7530736066120256, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 34071 + }, + { + "epoch": 0.34072, + "grad_norm": 1.337716513603304, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 34072 + }, + { + "epoch": 0.34073, + "grad_norm": 1.776469358112483, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 34073 + }, + { + "epoch": 0.34074, + "grad_norm": 1.9151974630741577, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 34074 + }, + { + "epoch": 0.34075, + "grad_norm": 1.4313911262989258, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 34075 + }, + { + "epoch": 0.34076, + "grad_norm": 1.4963468189045528, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34076 + }, + { + "epoch": 0.34077, + "grad_norm": 1.4800697312841449, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34077 + }, + { + "epoch": 0.34078, + "grad_norm": 1.3905364856437086, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 34078 + }, + { + "epoch": 0.34079, + "grad_norm": 1.5563622264840664, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 34079 + }, + { + "epoch": 0.3408, + "grad_norm": 1.3828924337827708, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 34080 + }, + { + "epoch": 0.34081, + "grad_norm": 1.4357633520715203, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34081 + }, + { + "epoch": 0.34082, + "grad_norm": 1.274422841436797, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34082 + }, + { + "epoch": 0.34083, + "grad_norm": 1.6734783533703141, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 34083 + }, + { + "epoch": 0.34084, + "grad_norm": 1.3040138101351837, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34084 + }, + { + "epoch": 0.34085, + "grad_norm": 1.4713466681496354, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34085 + }, + { + "epoch": 0.34086, + "grad_norm": 1.1614455247646822, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 34086 + }, + { + "epoch": 0.34087, + "grad_norm": 1.7712173387097923, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34087 + }, + { + "epoch": 0.34088, + "grad_norm": 1.6556174285554062, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 34088 + }, + { + "epoch": 0.34089, + "grad_norm": 1.5849097278571767, + "learning_rate": 0.003, + "loss": 4.029, + "step": 34089 + }, + { + "epoch": 0.3409, + "grad_norm": 1.2807160803760573, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 34090 + }, + { + "epoch": 0.34091, + "grad_norm": 1.7104521009799707, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 34091 + }, + { + "epoch": 0.34092, + "grad_norm": 1.3545497332962364, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 34092 + }, + { + "epoch": 0.34093, + "grad_norm": 1.869784451360222, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 34093 + }, + { + "epoch": 0.34094, + "grad_norm": 1.0858586010616715, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34094 + }, + { + "epoch": 0.34095, + "grad_norm": 1.8518604269644925, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 34095 + }, + { + "epoch": 0.34096, + "grad_norm": 1.1399343355364249, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34096 + }, + { + "epoch": 0.34097, + "grad_norm": 1.7535392793851166, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 34097 + }, + { + "epoch": 0.34098, + "grad_norm": 1.4466584108259388, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34098 + }, + { + "epoch": 0.34099, + "grad_norm": 1.5491163678455082, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34099 + }, + { + "epoch": 0.341, + "grad_norm": 1.5105860323100773, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 34100 + }, + { + "epoch": 0.34101, + "grad_norm": 1.504528941067984, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 34101 + }, + { + "epoch": 0.34102, + "grad_norm": 1.3408482382578726, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 34102 + }, + { + "epoch": 0.34103, + "grad_norm": 1.4941300455118929, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34103 + }, + { + "epoch": 0.34104, + "grad_norm": 1.505336763947847, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 34104 + }, + { + "epoch": 0.34105, + "grad_norm": 1.379758651788681, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 34105 + }, + { + "epoch": 0.34106, + "grad_norm": 1.506772487197942, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34106 + }, + { + "epoch": 0.34107, + "grad_norm": 1.2730777583693622, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 34107 + }, + { + "epoch": 0.34108, + "grad_norm": 1.596224283768673, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 34108 + }, + { + "epoch": 0.34109, + "grad_norm": 1.1500061744723933, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 34109 + }, + { + "epoch": 0.3411, + "grad_norm": 1.7602357909133286, + "learning_rate": 0.003, + "loss": 4.063, + "step": 34110 + }, + { + "epoch": 0.34111, + "grad_norm": 1.255776940100232, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34111 + }, + { + "epoch": 0.34112, + "grad_norm": 1.7349055784450174, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 34112 + }, + { + "epoch": 0.34113, + "grad_norm": 1.631916415584338, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 34113 + }, + { + "epoch": 0.34114, + "grad_norm": 1.2871066637962807, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34114 + }, + { + "epoch": 0.34115, + "grad_norm": 1.4520672359364266, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34115 + }, + { + "epoch": 0.34116, + "grad_norm": 1.7997080974599873, + "learning_rate": 0.003, + "loss": 4.068, + "step": 34116 + }, + { + "epoch": 0.34117, + "grad_norm": 1.4721605926905552, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 34117 + }, + { + "epoch": 0.34118, + "grad_norm": 1.4730819065905394, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 34118 + }, + { + "epoch": 0.34119, + "grad_norm": 1.4259620795731625, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 34119 + }, + { + "epoch": 0.3412, + "grad_norm": 1.4065626898062962, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 34120 + }, + { + "epoch": 0.34121, + "grad_norm": 1.397393401931859, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 34121 + }, + { + "epoch": 0.34122, + "grad_norm": 1.6832209839899546, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34122 + }, + { + "epoch": 0.34123, + "grad_norm": 1.1771203488059956, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34123 + }, + { + "epoch": 0.34124, + "grad_norm": 1.4696849298011145, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34124 + }, + { + "epoch": 0.34125, + "grad_norm": 1.6879388967977949, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34125 + }, + { + "epoch": 0.34126, + "grad_norm": 1.5067077066472498, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 34126 + }, + { + "epoch": 0.34127, + "grad_norm": 1.6970732435301326, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 34127 + }, + { + "epoch": 0.34128, + "grad_norm": 1.297610276177003, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 34128 + }, + { + "epoch": 0.34129, + "grad_norm": 1.515787325852398, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34129 + }, + { + "epoch": 0.3413, + "grad_norm": 1.5824320600194648, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 34130 + }, + { + "epoch": 0.34131, + "grad_norm": 1.1809235693818596, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 34131 + }, + { + "epoch": 0.34132, + "grad_norm": 1.5721996149682542, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34132 + }, + { + "epoch": 0.34133, + "grad_norm": 1.585020906591267, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 34133 + }, + { + "epoch": 0.34134, + "grad_norm": 1.4775128582413508, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 34134 + }, + { + "epoch": 0.34135, + "grad_norm": 1.7349017754894425, + "learning_rate": 0.003, + "loss": 4.053, + "step": 34135 + }, + { + "epoch": 0.34136, + "grad_norm": 1.4673345065765924, + "learning_rate": 0.003, + "loss": 4.044, + "step": 34136 + }, + { + "epoch": 0.34137, + "grad_norm": 1.4382278967210371, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 34137 + }, + { + "epoch": 0.34138, + "grad_norm": 1.687569952243861, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 34138 + }, + { + "epoch": 0.34139, + "grad_norm": 1.306711490358578, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 34139 + }, + { + "epoch": 0.3414, + "grad_norm": 1.5288217804656006, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34140 + }, + { + "epoch": 0.34141, + "grad_norm": 1.2571780407185902, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 34141 + }, + { + "epoch": 0.34142, + "grad_norm": 1.7807264513521155, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34142 + }, + { + "epoch": 0.34143, + "grad_norm": 1.4911199066815195, + "learning_rate": 0.003, + "loss": 4.043, + "step": 34143 + }, + { + "epoch": 0.34144, + "grad_norm": 1.7457478898460204, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 34144 + }, + { + "epoch": 0.34145, + "grad_norm": 1.4207280312847164, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34145 + }, + { + "epoch": 0.34146, + "grad_norm": 1.3447044629269185, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 34146 + }, + { + "epoch": 0.34147, + "grad_norm": 1.360034878289368, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34147 + }, + { + "epoch": 0.34148, + "grad_norm": 1.4865843237050418, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34148 + }, + { + "epoch": 0.34149, + "grad_norm": 1.5618010495222339, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 34149 + }, + { + "epoch": 0.3415, + "grad_norm": 1.5129929362959174, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 34150 + }, + { + "epoch": 0.34151, + "grad_norm": 1.544102277539507, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 34151 + }, + { + "epoch": 0.34152, + "grad_norm": 1.449536694839304, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 34152 + }, + { + "epoch": 0.34153, + "grad_norm": 1.2636184733613758, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 34153 + }, + { + "epoch": 0.34154, + "grad_norm": 1.6205457061220796, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 34154 + }, + { + "epoch": 0.34155, + "grad_norm": 1.3542102926467097, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 34155 + }, + { + "epoch": 0.34156, + "grad_norm": 1.5520590785816895, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34156 + }, + { + "epoch": 0.34157, + "grad_norm": 1.267588219426406, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34157 + }, + { + "epoch": 0.34158, + "grad_norm": 1.6743180674629363, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 34158 + }, + { + "epoch": 0.34159, + "grad_norm": 1.385504529553065, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 34159 + }, + { + "epoch": 0.3416, + "grad_norm": 1.389118206159624, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 34160 + }, + { + "epoch": 0.34161, + "grad_norm": 1.8374273174820133, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 34161 + }, + { + "epoch": 0.34162, + "grad_norm": 1.3718214381756053, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 34162 + }, + { + "epoch": 0.34163, + "grad_norm": 1.4984477182403302, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34163 + }, + { + "epoch": 0.34164, + "grad_norm": 1.3194486628469464, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34164 + }, + { + "epoch": 0.34165, + "grad_norm": 1.6666479910565701, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34165 + }, + { + "epoch": 0.34166, + "grad_norm": 1.4698264349143275, + "learning_rate": 0.003, + "loss": 4.015, + "step": 34166 + }, + { + "epoch": 0.34167, + "grad_norm": 1.495051043237457, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 34167 + }, + { + "epoch": 0.34168, + "grad_norm": 1.3943454292308795, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 34168 + }, + { + "epoch": 0.34169, + "grad_norm": 1.7527321139575456, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 34169 + }, + { + "epoch": 0.3417, + "grad_norm": 1.4068930187337398, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 34170 + }, + { + "epoch": 0.34171, + "grad_norm": 1.2870921243364841, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34171 + }, + { + "epoch": 0.34172, + "grad_norm": 1.5292199204325982, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 34172 + }, + { + "epoch": 0.34173, + "grad_norm": 1.5765138041486746, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34173 + }, + { + "epoch": 0.34174, + "grad_norm": 1.5519277461150471, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 34174 + }, + { + "epoch": 0.34175, + "grad_norm": 1.4932027729544834, + "learning_rate": 0.003, + "loss": 4.035, + "step": 34175 + }, + { + "epoch": 0.34176, + "grad_norm": 1.4541203477438245, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 34176 + }, + { + "epoch": 0.34177, + "grad_norm": 1.6083619974141459, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34177 + }, + { + "epoch": 0.34178, + "grad_norm": 1.3877956755885341, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34178 + }, + { + "epoch": 0.34179, + "grad_norm": 1.8023253373030756, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 34179 + }, + { + "epoch": 0.3418, + "grad_norm": 1.710645076350043, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34180 + }, + { + "epoch": 0.34181, + "grad_norm": 1.470627907012399, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34181 + }, + { + "epoch": 0.34182, + "grad_norm": 1.4661620694085657, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 34182 + }, + { + "epoch": 0.34183, + "grad_norm": 1.8646183600125643, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 34183 + }, + { + "epoch": 0.34184, + "grad_norm": 1.2151309342862544, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 34184 + }, + { + "epoch": 0.34185, + "grad_norm": 1.5013119984694168, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34185 + }, + { + "epoch": 0.34186, + "grad_norm": 1.2864493774601462, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34186 + }, + { + "epoch": 0.34187, + "grad_norm": 1.7049691509636056, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 34187 + }, + { + "epoch": 0.34188, + "grad_norm": 1.288704336792899, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 34188 + }, + { + "epoch": 0.34189, + "grad_norm": 1.8880793581851067, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 34189 + }, + { + "epoch": 0.3419, + "grad_norm": 1.8649363097175926, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 34190 + }, + { + "epoch": 0.34191, + "grad_norm": 1.0686747082960009, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 34191 + }, + { + "epoch": 0.34192, + "grad_norm": 1.4685332274347718, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 34192 + }, + { + "epoch": 0.34193, + "grad_norm": 1.5740480948541073, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 34193 + }, + { + "epoch": 0.34194, + "grad_norm": 1.461638751706403, + "learning_rate": 0.003, + "loss": 4.058, + "step": 34194 + }, + { + "epoch": 0.34195, + "grad_norm": 1.3144580203189344, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 34195 + }, + { + "epoch": 0.34196, + "grad_norm": 1.3332953223175237, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 34196 + }, + { + "epoch": 0.34197, + "grad_norm": 1.7498475971534786, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 34197 + }, + { + "epoch": 0.34198, + "grad_norm": 1.4596853683139392, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 34198 + }, + { + "epoch": 0.34199, + "grad_norm": 1.6406517543075159, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 34199 + }, + { + "epoch": 0.342, + "grad_norm": 1.4169586662898346, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34200 + }, + { + "epoch": 0.34201, + "grad_norm": 2.0245263892248775, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 34201 + }, + { + "epoch": 0.34202, + "grad_norm": 1.2297078112501527, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 34202 + }, + { + "epoch": 0.34203, + "grad_norm": 1.277988631834222, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34203 + }, + { + "epoch": 0.34204, + "grad_norm": 1.6020305327786295, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 34204 + }, + { + "epoch": 0.34205, + "grad_norm": 1.3879640290851318, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 34205 + }, + { + "epoch": 0.34206, + "grad_norm": 1.4323879259783256, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34206 + }, + { + "epoch": 0.34207, + "grad_norm": 1.3947664505719302, + "learning_rate": 0.003, + "loss": 4.035, + "step": 34207 + }, + { + "epoch": 0.34208, + "grad_norm": 1.5753689054945872, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 34208 + }, + { + "epoch": 0.34209, + "grad_norm": 1.2936658576870939, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 34209 + }, + { + "epoch": 0.3421, + "grad_norm": 1.5361703313040769, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 34210 + }, + { + "epoch": 0.34211, + "grad_norm": 1.411863447068691, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 34211 + }, + { + "epoch": 0.34212, + "grad_norm": 1.506741667256098, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 34212 + }, + { + "epoch": 0.34213, + "grad_norm": 1.783228299105147, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 34213 + }, + { + "epoch": 0.34214, + "grad_norm": 1.4979662225858856, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34214 + }, + { + "epoch": 0.34215, + "grad_norm": 1.5219429765731625, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 34215 + }, + { + "epoch": 0.34216, + "grad_norm": 1.4088066353618733, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 34216 + }, + { + "epoch": 0.34217, + "grad_norm": 1.4858249421845646, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34217 + }, + { + "epoch": 0.34218, + "grad_norm": 1.3992931130504316, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34218 + }, + { + "epoch": 0.34219, + "grad_norm": 1.5609047002548246, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 34219 + }, + { + "epoch": 0.3422, + "grad_norm": 1.410879165487592, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 34220 + }, + { + "epoch": 0.34221, + "grad_norm": 1.4953913474630998, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 34221 + }, + { + "epoch": 0.34222, + "grad_norm": 1.438406445804973, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34222 + }, + { + "epoch": 0.34223, + "grad_norm": 1.379591812335639, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 34223 + }, + { + "epoch": 0.34224, + "grad_norm": 1.741681646114862, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 34224 + }, + { + "epoch": 0.34225, + "grad_norm": 1.7349282465151925, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 34225 + }, + { + "epoch": 0.34226, + "grad_norm": 1.3424157380911874, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 34226 + }, + { + "epoch": 0.34227, + "grad_norm": 1.6725932770566696, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 34227 + }, + { + "epoch": 0.34228, + "grad_norm": 1.4511834778173738, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 34228 + }, + { + "epoch": 0.34229, + "grad_norm": 1.2811868602915042, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 34229 + }, + { + "epoch": 0.3423, + "grad_norm": 1.460840097638962, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 34230 + }, + { + "epoch": 0.34231, + "grad_norm": 1.3155018673912757, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34231 + }, + { + "epoch": 0.34232, + "grad_norm": 1.7539715774866014, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34232 + }, + { + "epoch": 0.34233, + "grad_norm": 1.1929271053056398, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 34233 + }, + { + "epoch": 0.34234, + "grad_norm": 1.9996138342580432, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34234 + }, + { + "epoch": 0.34235, + "grad_norm": 1.0768147519397862, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 34235 + }, + { + "epoch": 0.34236, + "grad_norm": 1.8262529097637392, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 34236 + }, + { + "epoch": 0.34237, + "grad_norm": 1.4768934441729356, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 34237 + }, + { + "epoch": 0.34238, + "grad_norm": 1.5976493556503408, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 34238 + }, + { + "epoch": 0.34239, + "grad_norm": 1.619364165179621, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 34239 + }, + { + "epoch": 0.3424, + "grad_norm": 1.400897675083427, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34240 + }, + { + "epoch": 0.34241, + "grad_norm": 1.5090193520286155, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34241 + }, + { + "epoch": 0.34242, + "grad_norm": 1.4143637642477114, + "learning_rate": 0.003, + "loss": 4.037, + "step": 34242 + }, + { + "epoch": 0.34243, + "grad_norm": 1.5304283344485512, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 34243 + }, + { + "epoch": 0.34244, + "grad_norm": 1.4329110686043365, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 34244 + }, + { + "epoch": 0.34245, + "grad_norm": 1.7294212226296366, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 34245 + }, + { + "epoch": 0.34246, + "grad_norm": 1.3325947724721445, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 34246 + }, + { + "epoch": 0.34247, + "grad_norm": 1.7714403609510343, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 34247 + }, + { + "epoch": 0.34248, + "grad_norm": 1.366949770318711, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34248 + }, + { + "epoch": 0.34249, + "grad_norm": 1.4514244719714922, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34249 + }, + { + "epoch": 0.3425, + "grad_norm": 1.4253119048422431, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 34250 + }, + { + "epoch": 0.34251, + "grad_norm": 1.550247114719812, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34251 + }, + { + "epoch": 0.34252, + "grad_norm": 1.690442632375625, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34252 + }, + { + "epoch": 0.34253, + "grad_norm": 1.323395523294194, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 34253 + }, + { + "epoch": 0.34254, + "grad_norm": 1.6349721689639607, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 34254 + }, + { + "epoch": 0.34255, + "grad_norm": 1.3187505054945117, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34255 + }, + { + "epoch": 0.34256, + "grad_norm": 1.5767299185289108, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34256 + }, + { + "epoch": 0.34257, + "grad_norm": 1.3610523242286825, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 34257 + }, + { + "epoch": 0.34258, + "grad_norm": 1.801638484674479, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 34258 + }, + { + "epoch": 0.34259, + "grad_norm": 1.4670222036826395, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 34259 + }, + { + "epoch": 0.3426, + "grad_norm": 1.5507151079421408, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 34260 + }, + { + "epoch": 0.34261, + "grad_norm": 1.2350709722998359, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 34261 + }, + { + "epoch": 0.34262, + "grad_norm": 1.6688182327515053, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 34262 + }, + { + "epoch": 0.34263, + "grad_norm": 1.4163065608643852, + "learning_rate": 0.003, + "loss": 4.042, + "step": 34263 + }, + { + "epoch": 0.34264, + "grad_norm": 1.3920364630268847, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 34264 + }, + { + "epoch": 0.34265, + "grad_norm": 1.5247326154540988, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 34265 + }, + { + "epoch": 0.34266, + "grad_norm": 1.5885646075044435, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34266 + }, + { + "epoch": 0.34267, + "grad_norm": 1.3239402475995936, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 34267 + }, + { + "epoch": 0.34268, + "grad_norm": 1.3002324537745134, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 34268 + }, + { + "epoch": 0.34269, + "grad_norm": 1.5252143242082528, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34269 + }, + { + "epoch": 0.3427, + "grad_norm": 1.376804496774604, + "learning_rate": 0.003, + "loss": 4.054, + "step": 34270 + }, + { + "epoch": 0.34271, + "grad_norm": 1.5036456061442613, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34271 + }, + { + "epoch": 0.34272, + "grad_norm": 1.6572059447343317, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 34272 + }, + { + "epoch": 0.34273, + "grad_norm": 1.9390539768403894, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 34273 + }, + { + "epoch": 0.34274, + "grad_norm": 1.4311992862320324, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 34274 + }, + { + "epoch": 0.34275, + "grad_norm": 1.3964959442693272, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 34275 + }, + { + "epoch": 0.34276, + "grad_norm": 1.4133346218388838, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34276 + }, + { + "epoch": 0.34277, + "grad_norm": 1.6135772418902756, + "learning_rate": 0.003, + "loss": 4.015, + "step": 34277 + }, + { + "epoch": 0.34278, + "grad_norm": 1.1636097538904198, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 34278 + }, + { + "epoch": 0.34279, + "grad_norm": 1.630585734262787, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 34279 + }, + { + "epoch": 0.3428, + "grad_norm": 1.1860442376306528, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34280 + }, + { + "epoch": 0.34281, + "grad_norm": 1.6564059227026253, + "learning_rate": 0.003, + "loss": 4.067, + "step": 34281 + }, + { + "epoch": 0.34282, + "grad_norm": 1.3801862857825686, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 34282 + }, + { + "epoch": 0.34283, + "grad_norm": 1.4362381050085848, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 34283 + }, + { + "epoch": 0.34284, + "grad_norm": 1.4630601175736158, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 34284 + }, + { + "epoch": 0.34285, + "grad_norm": 1.564858623682543, + "learning_rate": 0.003, + "loss": 4.027, + "step": 34285 + }, + { + "epoch": 0.34286, + "grad_norm": 1.8966429297013652, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 34286 + }, + { + "epoch": 0.34287, + "grad_norm": 1.2801787134998874, + "learning_rate": 0.003, + "loss": 4.05, + "step": 34287 + }, + { + "epoch": 0.34288, + "grad_norm": 1.5286801096185678, + "learning_rate": 0.003, + "loss": 4.057, + "step": 34288 + }, + { + "epoch": 0.34289, + "grad_norm": 1.6744549333856007, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34289 + }, + { + "epoch": 0.3429, + "grad_norm": 1.1920962584932362, + "learning_rate": 0.003, + "loss": 4.016, + "step": 34290 + }, + { + "epoch": 0.34291, + "grad_norm": 1.6911629267527901, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34291 + }, + { + "epoch": 0.34292, + "grad_norm": 1.3413702555971836, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34292 + }, + { + "epoch": 0.34293, + "grad_norm": 1.7691057179338325, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34293 + }, + { + "epoch": 0.34294, + "grad_norm": 1.284313663017511, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34294 + }, + { + "epoch": 0.34295, + "grad_norm": 1.6913450383939792, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34295 + }, + { + "epoch": 0.34296, + "grad_norm": 1.65420263610679, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34296 + }, + { + "epoch": 0.34297, + "grad_norm": 1.3653586384540373, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 34297 + }, + { + "epoch": 0.34298, + "grad_norm": 1.733393696502836, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 34298 + }, + { + "epoch": 0.34299, + "grad_norm": 1.2826436818678977, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34299 + }, + { + "epoch": 0.343, + "grad_norm": 1.4428409473443715, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 34300 + }, + { + "epoch": 0.34301, + "grad_norm": 1.5560238720292476, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 34301 + }, + { + "epoch": 0.34302, + "grad_norm": 1.6661381680722112, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34302 + }, + { + "epoch": 0.34303, + "grad_norm": 1.7292651713822407, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34303 + }, + { + "epoch": 0.34304, + "grad_norm": 1.548605494641736, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 34304 + }, + { + "epoch": 0.34305, + "grad_norm": 1.411217571628965, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34305 + }, + { + "epoch": 0.34306, + "grad_norm": 1.4421883144027694, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 34306 + }, + { + "epoch": 0.34307, + "grad_norm": 1.5104332985889053, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 34307 + }, + { + "epoch": 0.34308, + "grad_norm": 1.4901183824019433, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 34308 + }, + { + "epoch": 0.34309, + "grad_norm": 1.352121342335828, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 34309 + }, + { + "epoch": 0.3431, + "grad_norm": 1.5347152500978298, + "learning_rate": 0.003, + "loss": 4.05, + "step": 34310 + }, + { + "epoch": 0.34311, + "grad_norm": 1.3363726981661073, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 34311 + }, + { + "epoch": 0.34312, + "grad_norm": 1.5079386056506578, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34312 + }, + { + "epoch": 0.34313, + "grad_norm": 1.3043883509339211, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 34313 + }, + { + "epoch": 0.34314, + "grad_norm": 1.500300211750023, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34314 + }, + { + "epoch": 0.34315, + "grad_norm": 1.2803383694870527, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34315 + }, + { + "epoch": 0.34316, + "grad_norm": 1.4824389619522655, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 34316 + }, + { + "epoch": 0.34317, + "grad_norm": 1.3055789713955193, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 34317 + }, + { + "epoch": 0.34318, + "grad_norm": 1.6938357541368, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 34318 + }, + { + "epoch": 0.34319, + "grad_norm": 1.6862247667062498, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 34319 + }, + { + "epoch": 0.3432, + "grad_norm": 1.559390162066165, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 34320 + }, + { + "epoch": 0.34321, + "grad_norm": 1.6809049533228873, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 34321 + }, + { + "epoch": 0.34322, + "grad_norm": 1.6173632982729587, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34322 + }, + { + "epoch": 0.34323, + "grad_norm": 1.6565792381607922, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 34323 + }, + { + "epoch": 0.34324, + "grad_norm": 1.4900399955721857, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 34324 + }, + { + "epoch": 0.34325, + "grad_norm": 1.292116117166001, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 34325 + }, + { + "epoch": 0.34326, + "grad_norm": 1.643243595450466, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34326 + }, + { + "epoch": 0.34327, + "grad_norm": 1.3579679653250039, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 34327 + }, + { + "epoch": 0.34328, + "grad_norm": 1.618970744756156, + "learning_rate": 0.003, + "loss": 4.036, + "step": 34328 + }, + { + "epoch": 0.34329, + "grad_norm": 1.3184057530695767, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 34329 + }, + { + "epoch": 0.3433, + "grad_norm": 1.8820015477813112, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 34330 + }, + { + "epoch": 0.34331, + "grad_norm": 1.3910912237211994, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 34331 + }, + { + "epoch": 0.34332, + "grad_norm": 1.418548736944478, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 34332 + }, + { + "epoch": 0.34333, + "grad_norm": 1.445392078922259, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34333 + }, + { + "epoch": 0.34334, + "grad_norm": 1.3497710287641667, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 34334 + }, + { + "epoch": 0.34335, + "grad_norm": 1.6557741565157416, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34335 + }, + { + "epoch": 0.34336, + "grad_norm": 1.4231005265862877, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34336 + }, + { + "epoch": 0.34337, + "grad_norm": 1.3167419277537464, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34337 + }, + { + "epoch": 0.34338, + "grad_norm": 1.6605502002395809, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 34338 + }, + { + "epoch": 0.34339, + "grad_norm": 1.3202595354451243, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34339 + }, + { + "epoch": 0.3434, + "grad_norm": 1.9338213080405993, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34340 + }, + { + "epoch": 0.34341, + "grad_norm": 1.4852930017931165, + "learning_rate": 0.003, + "loss": 4.024, + "step": 34341 + }, + { + "epoch": 0.34342, + "grad_norm": 1.4481972317272866, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 34342 + }, + { + "epoch": 0.34343, + "grad_norm": 1.359595595411339, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 34343 + }, + { + "epoch": 0.34344, + "grad_norm": 1.7846265241359058, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 34344 + }, + { + "epoch": 0.34345, + "grad_norm": 1.4102995568552485, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 34345 + }, + { + "epoch": 0.34346, + "grad_norm": 1.3575599026745, + "learning_rate": 0.003, + "loss": 4.028, + "step": 34346 + }, + { + "epoch": 0.34347, + "grad_norm": 1.4126991023946602, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34347 + }, + { + "epoch": 0.34348, + "grad_norm": 1.5491524567229027, + "learning_rate": 0.003, + "loss": 4.016, + "step": 34348 + }, + { + "epoch": 0.34349, + "grad_norm": 1.4674665161235299, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 34349 + }, + { + "epoch": 0.3435, + "grad_norm": 1.5719010053346933, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34350 + }, + { + "epoch": 0.34351, + "grad_norm": 1.6776692160764222, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34351 + }, + { + "epoch": 0.34352, + "grad_norm": 1.1375767664563299, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 34352 + }, + { + "epoch": 0.34353, + "grad_norm": 1.4970453382172852, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 34353 + }, + { + "epoch": 0.34354, + "grad_norm": 1.4449357585844043, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 34354 + }, + { + "epoch": 0.34355, + "grad_norm": 1.351791480744486, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34355 + }, + { + "epoch": 0.34356, + "grad_norm": 1.3716436856350644, + "learning_rate": 0.003, + "loss": 4.038, + "step": 34356 + }, + { + "epoch": 0.34357, + "grad_norm": 1.420011734842814, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 34357 + }, + { + "epoch": 0.34358, + "grad_norm": 1.403972656287465, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 34358 + }, + { + "epoch": 0.34359, + "grad_norm": 1.630968669940623, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 34359 + }, + { + "epoch": 0.3436, + "grad_norm": 1.6285663737897005, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 34360 + }, + { + "epoch": 0.34361, + "grad_norm": 1.9669979234629598, + "learning_rate": 0.003, + "loss": 4.078, + "step": 34361 + }, + { + "epoch": 0.34362, + "grad_norm": 1.4607348430094425, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 34362 + }, + { + "epoch": 0.34363, + "grad_norm": 1.3561352349681086, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 34363 + }, + { + "epoch": 0.34364, + "grad_norm": 1.4806697279602417, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 34364 + }, + { + "epoch": 0.34365, + "grad_norm": 1.6001819351151085, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 34365 + }, + { + "epoch": 0.34366, + "grad_norm": 1.3053240022891137, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 34366 + }, + { + "epoch": 0.34367, + "grad_norm": 1.8823803775085397, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34367 + }, + { + "epoch": 0.34368, + "grad_norm": 1.4281276530326947, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34368 + }, + { + "epoch": 0.34369, + "grad_norm": 1.6660340493531507, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 34369 + }, + { + "epoch": 0.3437, + "grad_norm": 1.3237121809151315, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 34370 + }, + { + "epoch": 0.34371, + "grad_norm": 1.709752433534112, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 34371 + }, + { + "epoch": 0.34372, + "grad_norm": 1.3268429174498484, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34372 + }, + { + "epoch": 0.34373, + "grad_norm": 1.6487348033563867, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 34373 + }, + { + "epoch": 0.34374, + "grad_norm": 1.2590470139499217, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34374 + }, + { + "epoch": 0.34375, + "grad_norm": 1.5830516057979798, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34375 + }, + { + "epoch": 0.34376, + "grad_norm": 1.3471878595800535, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34376 + }, + { + "epoch": 0.34377, + "grad_norm": 1.9529282126767928, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34377 + }, + { + "epoch": 0.34378, + "grad_norm": 1.559499858147031, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 34378 + }, + { + "epoch": 0.34379, + "grad_norm": 1.4357181311307923, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 34379 + }, + { + "epoch": 0.3438, + "grad_norm": 1.5776731259862082, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 34380 + }, + { + "epoch": 0.34381, + "grad_norm": 1.3762933357028646, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34381 + }, + { + "epoch": 0.34382, + "grad_norm": 1.3871188440337292, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 34382 + }, + { + "epoch": 0.34383, + "grad_norm": 1.508772497878746, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34383 + }, + { + "epoch": 0.34384, + "grad_norm": 1.5484526332407318, + "learning_rate": 0.003, + "loss": 4.059, + "step": 34384 + }, + { + "epoch": 0.34385, + "grad_norm": 1.3807377110788062, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34385 + }, + { + "epoch": 0.34386, + "grad_norm": 1.430565141210872, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34386 + }, + { + "epoch": 0.34387, + "grad_norm": 1.4292764820146218, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 34387 + }, + { + "epoch": 0.34388, + "grad_norm": 1.4828279982945591, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 34388 + }, + { + "epoch": 0.34389, + "grad_norm": 1.642901789512018, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34389 + }, + { + "epoch": 0.3439, + "grad_norm": 1.5043292174125638, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34390 + }, + { + "epoch": 0.34391, + "grad_norm": 1.4066339935923446, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 34391 + }, + { + "epoch": 0.34392, + "grad_norm": 1.5923787292062026, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34392 + }, + { + "epoch": 0.34393, + "grad_norm": 1.416713763462943, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 34393 + }, + { + "epoch": 0.34394, + "grad_norm": 1.648389284166871, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34394 + }, + { + "epoch": 0.34395, + "grad_norm": 1.5286178190592925, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 34395 + }, + { + "epoch": 0.34396, + "grad_norm": 1.4859175073748037, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 34396 + }, + { + "epoch": 0.34397, + "grad_norm": 1.5283369638164348, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34397 + }, + { + "epoch": 0.34398, + "grad_norm": 1.361536953968013, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 34398 + }, + { + "epoch": 0.34399, + "grad_norm": 1.6931834785176563, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 34399 + }, + { + "epoch": 0.344, + "grad_norm": 1.5163178717055332, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34400 + }, + { + "epoch": 0.34401, + "grad_norm": 1.4932553540784617, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34401 + }, + { + "epoch": 0.34402, + "grad_norm": 1.7058703851175483, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34402 + }, + { + "epoch": 0.34403, + "grad_norm": 1.5621953129042003, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34403 + }, + { + "epoch": 0.34404, + "grad_norm": 1.6748958411844765, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 34404 + }, + { + "epoch": 0.34405, + "grad_norm": 1.4702196769739873, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 34405 + }, + { + "epoch": 0.34406, + "grad_norm": 1.355807491994499, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 34406 + }, + { + "epoch": 0.34407, + "grad_norm": 1.385126916169077, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 34407 + }, + { + "epoch": 0.34408, + "grad_norm": 1.4378087613804893, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 34408 + }, + { + "epoch": 0.34409, + "grad_norm": 1.4229861704400562, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34409 + }, + { + "epoch": 0.3441, + "grad_norm": 1.3734228814690017, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34410 + }, + { + "epoch": 0.34411, + "grad_norm": 1.7664098966860309, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 34411 + }, + { + "epoch": 0.34412, + "grad_norm": 1.491745347098319, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 34412 + }, + { + "epoch": 0.34413, + "grad_norm": 1.8831138092164748, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34413 + }, + { + "epoch": 0.34414, + "grad_norm": 1.3642374416543408, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 34414 + }, + { + "epoch": 0.34415, + "grad_norm": 1.5111964347286189, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 34415 + }, + { + "epoch": 0.34416, + "grad_norm": 1.4446306160432525, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 34416 + }, + { + "epoch": 0.34417, + "grad_norm": 1.5540557764686123, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 34417 + }, + { + "epoch": 0.34418, + "grad_norm": 1.4178982939786282, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 34418 + }, + { + "epoch": 0.34419, + "grad_norm": 1.4131809457529207, + "learning_rate": 0.003, + "loss": 4.058, + "step": 34419 + }, + { + "epoch": 0.3442, + "grad_norm": 1.5526423962979994, + "learning_rate": 0.003, + "loss": 4.027, + "step": 34420 + }, + { + "epoch": 0.34421, + "grad_norm": 1.973310373356072, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 34421 + }, + { + "epoch": 0.34422, + "grad_norm": 1.0743142066787816, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 34422 + }, + { + "epoch": 0.34423, + "grad_norm": 1.8999400947142213, + "learning_rate": 0.003, + "loss": 4.057, + "step": 34423 + }, + { + "epoch": 0.34424, + "grad_norm": 1.576636751385354, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34424 + }, + { + "epoch": 0.34425, + "grad_norm": 1.7922239843440044, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34425 + }, + { + "epoch": 0.34426, + "grad_norm": 1.4507204241195062, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 34426 + }, + { + "epoch": 0.34427, + "grad_norm": 1.4534422093308637, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 34427 + }, + { + "epoch": 0.34428, + "grad_norm": 1.6619317761942534, + "learning_rate": 0.003, + "loss": 4.055, + "step": 34428 + }, + { + "epoch": 0.34429, + "grad_norm": 1.5192300448763345, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 34429 + }, + { + "epoch": 0.3443, + "grad_norm": 1.4950666164028283, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34430 + }, + { + "epoch": 0.34431, + "grad_norm": 1.147881067465936, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 34431 + }, + { + "epoch": 0.34432, + "grad_norm": 1.4785553657912593, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 34432 + }, + { + "epoch": 0.34433, + "grad_norm": 1.3501294246235565, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 34433 + }, + { + "epoch": 0.34434, + "grad_norm": 1.6940574301167397, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 34434 + }, + { + "epoch": 0.34435, + "grad_norm": 1.1625608105547331, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34435 + }, + { + "epoch": 0.34436, + "grad_norm": 1.5822267548816968, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34436 + }, + { + "epoch": 0.34437, + "grad_norm": 1.509277680020534, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 34437 + }, + { + "epoch": 0.34438, + "grad_norm": 1.5006812804776908, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 34438 + }, + { + "epoch": 0.34439, + "grad_norm": 1.6812592491765714, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 34439 + }, + { + "epoch": 0.3444, + "grad_norm": 1.273112578187822, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 34440 + }, + { + "epoch": 0.34441, + "grad_norm": 1.6774123968241177, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34441 + }, + { + "epoch": 0.34442, + "grad_norm": 1.2944498016440038, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 34442 + }, + { + "epoch": 0.34443, + "grad_norm": 1.5761886841019042, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 34443 + }, + { + "epoch": 0.34444, + "grad_norm": 2.002894434512832, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34444 + }, + { + "epoch": 0.34445, + "grad_norm": 1.4503255807719229, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34445 + }, + { + "epoch": 0.34446, + "grad_norm": 1.6777682619903795, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 34446 + }, + { + "epoch": 0.34447, + "grad_norm": 1.5010407198986984, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34447 + }, + { + "epoch": 0.34448, + "grad_norm": 1.8088807711662114, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34448 + }, + { + "epoch": 0.34449, + "grad_norm": 1.3352551539325792, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 34449 + }, + { + "epoch": 0.3445, + "grad_norm": 1.6759415826548123, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 34450 + }, + { + "epoch": 0.34451, + "grad_norm": 1.4155555652898353, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 34451 + }, + { + "epoch": 0.34452, + "grad_norm": 1.7608457507325348, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 34452 + }, + { + "epoch": 0.34453, + "grad_norm": 1.4789002453972624, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 34453 + }, + { + "epoch": 0.34454, + "grad_norm": 1.4889046826733243, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34454 + }, + { + "epoch": 0.34455, + "grad_norm": 1.3558635347215298, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34455 + }, + { + "epoch": 0.34456, + "grad_norm": 1.5495995559813003, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 34456 + }, + { + "epoch": 0.34457, + "grad_norm": 1.170754624257845, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 34457 + }, + { + "epoch": 0.34458, + "grad_norm": 1.6541165281577292, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 34458 + }, + { + "epoch": 0.34459, + "grad_norm": 1.626862135714322, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34459 + }, + { + "epoch": 0.3446, + "grad_norm": 1.1673691613641937, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 34460 + }, + { + "epoch": 0.34461, + "grad_norm": 1.6183653170516357, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 34461 + }, + { + "epoch": 0.34462, + "grad_norm": 1.4193538951172175, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 34462 + }, + { + "epoch": 0.34463, + "grad_norm": 1.506662108978628, + "learning_rate": 0.003, + "loss": 4.05, + "step": 34463 + }, + { + "epoch": 0.34464, + "grad_norm": 1.3805092338688574, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 34464 + }, + { + "epoch": 0.34465, + "grad_norm": 1.8812638326980813, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34465 + }, + { + "epoch": 0.34466, + "grad_norm": 1.4484945083141436, + "learning_rate": 0.003, + "loss": 4.022, + "step": 34466 + }, + { + "epoch": 0.34467, + "grad_norm": 1.2508992179043705, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34467 + }, + { + "epoch": 0.34468, + "grad_norm": 1.8515661832916728, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34468 + }, + { + "epoch": 0.34469, + "grad_norm": 1.2981742928470827, + "learning_rate": 0.003, + "loss": 4.064, + "step": 34469 + }, + { + "epoch": 0.3447, + "grad_norm": 1.4559040685088351, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34470 + }, + { + "epoch": 0.34471, + "grad_norm": 1.6632305607151008, + "learning_rate": 0.003, + "loss": 4.039, + "step": 34471 + }, + { + "epoch": 0.34472, + "grad_norm": 1.3277576048495376, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 34472 + }, + { + "epoch": 0.34473, + "grad_norm": 1.296993087427553, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 34473 + }, + { + "epoch": 0.34474, + "grad_norm": 1.547447652551516, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 34474 + }, + { + "epoch": 0.34475, + "grad_norm": 1.4746753846832061, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34475 + }, + { + "epoch": 0.34476, + "grad_norm": 1.5981494510870344, + "learning_rate": 0.003, + "loss": 4.018, + "step": 34476 + }, + { + "epoch": 0.34477, + "grad_norm": 1.5724543582998955, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 34477 + }, + { + "epoch": 0.34478, + "grad_norm": 1.2848134279944425, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34478 + }, + { + "epoch": 0.34479, + "grad_norm": 1.4441988293927135, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34479 + }, + { + "epoch": 0.3448, + "grad_norm": 1.5214048150787909, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 34480 + }, + { + "epoch": 0.34481, + "grad_norm": 1.3047498061269929, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 34481 + }, + { + "epoch": 0.34482, + "grad_norm": 1.6981700887222795, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 34482 + }, + { + "epoch": 0.34483, + "grad_norm": 1.3606914668212373, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 34483 + }, + { + "epoch": 0.34484, + "grad_norm": 1.7011657174210904, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34484 + }, + { + "epoch": 0.34485, + "grad_norm": 1.5460759793555734, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34485 + }, + { + "epoch": 0.34486, + "grad_norm": 1.630769252192713, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 34486 + }, + { + "epoch": 0.34487, + "grad_norm": 1.6480796529568205, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34487 + }, + { + "epoch": 0.34488, + "grad_norm": 1.3632109070824399, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 34488 + }, + { + "epoch": 0.34489, + "grad_norm": 1.3948562531593325, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 34489 + }, + { + "epoch": 0.3449, + "grad_norm": 1.2771832277670598, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 34490 + }, + { + "epoch": 0.34491, + "grad_norm": 1.5388467617453148, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34491 + }, + { + "epoch": 0.34492, + "grad_norm": 1.4806365593424065, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 34492 + }, + { + "epoch": 0.34493, + "grad_norm": 1.6681166356793387, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34493 + }, + { + "epoch": 0.34494, + "grad_norm": 1.3564709907535202, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 34494 + }, + { + "epoch": 0.34495, + "grad_norm": 1.661263769326808, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34495 + }, + { + "epoch": 0.34496, + "grad_norm": 1.3081937231831366, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 34496 + }, + { + "epoch": 0.34497, + "grad_norm": 1.6476103757923863, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 34497 + }, + { + "epoch": 0.34498, + "grad_norm": 1.3431157540002878, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34498 + }, + { + "epoch": 0.34499, + "grad_norm": 1.5621653303627043, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 34499 + }, + { + "epoch": 0.345, + "grad_norm": 1.6472340000345906, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 34500 + }, + { + "epoch": 0.34501, + "grad_norm": 1.3871598741068631, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 34501 + }, + { + "epoch": 0.34502, + "grad_norm": 1.514379284298346, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 34502 + }, + { + "epoch": 0.34503, + "grad_norm": 1.6095543211302603, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 34503 + }, + { + "epoch": 0.34504, + "grad_norm": 1.5169265969237984, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 34504 + }, + { + "epoch": 0.34505, + "grad_norm": 1.4095782487536062, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34505 + }, + { + "epoch": 0.34506, + "grad_norm": 1.537011987017943, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 34506 + }, + { + "epoch": 0.34507, + "grad_norm": 1.1446984952631918, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 34507 + }, + { + "epoch": 0.34508, + "grad_norm": 1.409407079993378, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 34508 + }, + { + "epoch": 0.34509, + "grad_norm": 1.4467184038938479, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 34509 + }, + { + "epoch": 0.3451, + "grad_norm": 1.6818078884461196, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 34510 + }, + { + "epoch": 0.34511, + "grad_norm": 1.3931645946432787, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 34511 + }, + { + "epoch": 0.34512, + "grad_norm": 1.4832298444157028, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 34512 + }, + { + "epoch": 0.34513, + "grad_norm": 1.1636680733405633, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 34513 + }, + { + "epoch": 0.34514, + "grad_norm": 1.5597067682753063, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 34514 + }, + { + "epoch": 0.34515, + "grad_norm": 1.6063323036392332, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34515 + }, + { + "epoch": 0.34516, + "grad_norm": 1.963725517799927, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34516 + }, + { + "epoch": 0.34517, + "grad_norm": 1.4716843629713958, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 34517 + }, + { + "epoch": 0.34518, + "grad_norm": 1.7271599080166629, + "learning_rate": 0.003, + "loss": 4.023, + "step": 34518 + }, + { + "epoch": 0.34519, + "grad_norm": 1.4289591932802863, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34519 + }, + { + "epoch": 0.3452, + "grad_norm": 1.6316558102360341, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34520 + }, + { + "epoch": 0.34521, + "grad_norm": 1.2837880890916527, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34521 + }, + { + "epoch": 0.34522, + "grad_norm": 1.5688011827256338, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 34522 + }, + { + "epoch": 0.34523, + "grad_norm": 1.405867454645281, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 34523 + }, + { + "epoch": 0.34524, + "grad_norm": 1.5291613657122927, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 34524 + }, + { + "epoch": 0.34525, + "grad_norm": 1.2422747021280154, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34525 + }, + { + "epoch": 0.34526, + "grad_norm": 1.8844845066478817, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34526 + }, + { + "epoch": 0.34527, + "grad_norm": 1.683635477993362, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 34527 + }, + { + "epoch": 0.34528, + "grad_norm": 1.3102476050682688, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 34528 + }, + { + "epoch": 0.34529, + "grad_norm": 1.8608585601559915, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34529 + }, + { + "epoch": 0.3453, + "grad_norm": 1.3288045544060763, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34530 + }, + { + "epoch": 0.34531, + "grad_norm": 1.296513096718531, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34531 + }, + { + "epoch": 0.34532, + "grad_norm": 1.5880502168903892, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 34532 + }, + { + "epoch": 0.34533, + "grad_norm": 1.2565324408645266, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 34533 + }, + { + "epoch": 0.34534, + "grad_norm": 1.4990599305524372, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 34534 + }, + { + "epoch": 0.34535, + "grad_norm": 1.2086113529374896, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34535 + }, + { + "epoch": 0.34536, + "grad_norm": 1.5248764130072425, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34536 + }, + { + "epoch": 0.34537, + "grad_norm": 1.500400135756852, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 34537 + }, + { + "epoch": 0.34538, + "grad_norm": 1.591184697538816, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 34538 + }, + { + "epoch": 0.34539, + "grad_norm": 1.7160173545738324, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 34539 + }, + { + "epoch": 0.3454, + "grad_norm": 1.6941706398748286, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34540 + }, + { + "epoch": 0.34541, + "grad_norm": 1.4296975039163076, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 34541 + }, + { + "epoch": 0.34542, + "grad_norm": 1.5929511090969408, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 34542 + }, + { + "epoch": 0.34543, + "grad_norm": 1.544825692709214, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 34543 + }, + { + "epoch": 0.34544, + "grad_norm": 1.4252985476505895, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 34544 + }, + { + "epoch": 0.34545, + "grad_norm": 1.4530168013868565, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34545 + }, + { + "epoch": 0.34546, + "grad_norm": 1.6113555880039718, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34546 + }, + { + "epoch": 0.34547, + "grad_norm": 1.4691100038717007, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34547 + }, + { + "epoch": 0.34548, + "grad_norm": 1.3704349836539997, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 34548 + }, + { + "epoch": 0.34549, + "grad_norm": 1.6081932977818265, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 34549 + }, + { + "epoch": 0.3455, + "grad_norm": 1.4036549906498623, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 34550 + }, + { + "epoch": 0.34551, + "grad_norm": 1.5933311423487244, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34551 + }, + { + "epoch": 0.34552, + "grad_norm": 1.5317541264561925, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 34552 + }, + { + "epoch": 0.34553, + "grad_norm": 1.4791518414734819, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34553 + }, + { + "epoch": 0.34554, + "grad_norm": 1.4078828317034888, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 34554 + }, + { + "epoch": 0.34555, + "grad_norm": 1.384473501688924, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34555 + }, + { + "epoch": 0.34556, + "grad_norm": 1.4309053904719555, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 34556 + }, + { + "epoch": 0.34557, + "grad_norm": 1.555971530426214, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 34557 + }, + { + "epoch": 0.34558, + "grad_norm": 1.797981194023014, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34558 + }, + { + "epoch": 0.34559, + "grad_norm": 1.7002274919820868, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34559 + }, + { + "epoch": 0.3456, + "grad_norm": 1.2233155829842297, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34560 + }, + { + "epoch": 0.34561, + "grad_norm": 1.6597483292341093, + "learning_rate": 0.003, + "loss": 4.074, + "step": 34561 + }, + { + "epoch": 0.34562, + "grad_norm": 1.7823621255337339, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 34562 + }, + { + "epoch": 0.34563, + "grad_norm": 1.2811036806549065, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 34563 + }, + { + "epoch": 0.34564, + "grad_norm": 1.7980847015488155, + "learning_rate": 0.003, + "loss": 4.054, + "step": 34564 + }, + { + "epoch": 0.34565, + "grad_norm": 1.2698370350067372, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 34565 + }, + { + "epoch": 0.34566, + "grad_norm": 1.5055623769293174, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 34566 + }, + { + "epoch": 0.34567, + "grad_norm": 1.680911132337614, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 34567 + }, + { + "epoch": 0.34568, + "grad_norm": 1.5742932355428008, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 34568 + }, + { + "epoch": 0.34569, + "grad_norm": 1.4578305486378622, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 34569 + }, + { + "epoch": 0.3457, + "grad_norm": 1.758936837426216, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 34570 + }, + { + "epoch": 0.34571, + "grad_norm": 1.2744020961393918, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 34571 + }, + { + "epoch": 0.34572, + "grad_norm": 1.523420206821775, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 34572 + }, + { + "epoch": 0.34573, + "grad_norm": 1.4830821916137185, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 34573 + }, + { + "epoch": 0.34574, + "grad_norm": 1.5590189874646612, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 34574 + }, + { + "epoch": 0.34575, + "grad_norm": 1.3133278744317753, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 34575 + }, + { + "epoch": 0.34576, + "grad_norm": 1.5689870655144518, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 34576 + }, + { + "epoch": 0.34577, + "grad_norm": 1.3301995415539496, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 34577 + }, + { + "epoch": 0.34578, + "grad_norm": 1.6951431174259926, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34578 + }, + { + "epoch": 0.34579, + "grad_norm": 1.4031531728526654, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34579 + }, + { + "epoch": 0.3458, + "grad_norm": 1.5811193395892156, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 34580 + }, + { + "epoch": 0.34581, + "grad_norm": 1.7779219524668148, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 34581 + }, + { + "epoch": 0.34582, + "grad_norm": 1.4685196759250758, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 34582 + }, + { + "epoch": 0.34583, + "grad_norm": 1.3658047775029456, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34583 + }, + { + "epoch": 0.34584, + "grad_norm": 1.5740463620256142, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 34584 + }, + { + "epoch": 0.34585, + "grad_norm": 1.429926049130142, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34585 + }, + { + "epoch": 0.34586, + "grad_norm": 1.5683272801434507, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 34586 + }, + { + "epoch": 0.34587, + "grad_norm": 1.3142369569688275, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34587 + }, + { + "epoch": 0.34588, + "grad_norm": 1.6597776823968584, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 34588 + }, + { + "epoch": 0.34589, + "grad_norm": 1.2523329458079233, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 34589 + }, + { + "epoch": 0.3459, + "grad_norm": 1.7201760734200096, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34590 + }, + { + "epoch": 0.34591, + "grad_norm": 1.5561234443700251, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 34591 + }, + { + "epoch": 0.34592, + "grad_norm": 1.5284597039633914, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34592 + }, + { + "epoch": 0.34593, + "grad_norm": 1.51400966161712, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 34593 + }, + { + "epoch": 0.34594, + "grad_norm": 1.5034486420533708, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 34594 + }, + { + "epoch": 0.34595, + "grad_norm": 1.6970075276999526, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 34595 + }, + { + "epoch": 0.34596, + "grad_norm": 1.3695699074969914, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 34596 + }, + { + "epoch": 0.34597, + "grad_norm": 1.5438162776311568, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 34597 + }, + { + "epoch": 0.34598, + "grad_norm": 1.2246672941273435, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 34598 + }, + { + "epoch": 0.34599, + "grad_norm": 1.3864266133706367, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 34599 + }, + { + "epoch": 0.346, + "grad_norm": 1.5768950842147913, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34600 + }, + { + "epoch": 0.34601, + "grad_norm": 1.3281999585196385, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 34601 + }, + { + "epoch": 0.34602, + "grad_norm": 1.4314383725172715, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34602 + }, + { + "epoch": 0.34603, + "grad_norm": 1.7441385263720315, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 34603 + }, + { + "epoch": 0.34604, + "grad_norm": 1.9663087534542838, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34604 + }, + { + "epoch": 0.34605, + "grad_norm": 1.1770784538633592, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 34605 + }, + { + "epoch": 0.34606, + "grad_norm": 1.710639282349521, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34606 + }, + { + "epoch": 0.34607, + "grad_norm": 1.8140759909387536, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34607 + }, + { + "epoch": 0.34608, + "grad_norm": 1.5819391025989633, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 34608 + }, + { + "epoch": 0.34609, + "grad_norm": 1.2937778928192394, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34609 + }, + { + "epoch": 0.3461, + "grad_norm": 1.5193281242498977, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34610 + }, + { + "epoch": 0.34611, + "grad_norm": 1.3364028781120185, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 34611 + }, + { + "epoch": 0.34612, + "grad_norm": 1.662506311473199, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 34612 + }, + { + "epoch": 0.34613, + "grad_norm": 1.2625033258479823, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34613 + }, + { + "epoch": 0.34614, + "grad_norm": 1.5795853676705862, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 34614 + }, + { + "epoch": 0.34615, + "grad_norm": 1.4674505527052844, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 34615 + }, + { + "epoch": 0.34616, + "grad_norm": 1.5628602300330658, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 34616 + }, + { + "epoch": 0.34617, + "grad_norm": 1.6492028895673676, + "learning_rate": 0.003, + "loss": 4.043, + "step": 34617 + }, + { + "epoch": 0.34618, + "grad_norm": 1.5750578189933002, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 34618 + }, + { + "epoch": 0.34619, + "grad_norm": 2.0258178451198696, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 34619 + }, + { + "epoch": 0.3462, + "grad_norm": 1.291614649895676, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 34620 + }, + { + "epoch": 0.34621, + "grad_norm": 1.4942420785963209, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34621 + }, + { + "epoch": 0.34622, + "grad_norm": 1.522039546188995, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34622 + }, + { + "epoch": 0.34623, + "grad_norm": 1.377432840706467, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34623 + }, + { + "epoch": 0.34624, + "grad_norm": 1.8249464302645162, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 34624 + }, + { + "epoch": 0.34625, + "grad_norm": 1.4425853717744712, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 34625 + }, + { + "epoch": 0.34626, + "grad_norm": 1.4371859147976909, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 34626 + }, + { + "epoch": 0.34627, + "grad_norm": 1.312016715504202, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 34627 + }, + { + "epoch": 0.34628, + "grad_norm": 1.7377446914501873, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34628 + }, + { + "epoch": 0.34629, + "grad_norm": 1.3021789976892904, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 34629 + }, + { + "epoch": 0.3463, + "grad_norm": 1.4149028312170038, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 34630 + }, + { + "epoch": 0.34631, + "grad_norm": 1.6566946271551635, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 34631 + }, + { + "epoch": 0.34632, + "grad_norm": 1.5057743353222532, + "learning_rate": 0.003, + "loss": 4.056, + "step": 34632 + }, + { + "epoch": 0.34633, + "grad_norm": 1.6842916105971353, + "learning_rate": 0.003, + "loss": 4.06, + "step": 34633 + }, + { + "epoch": 0.34634, + "grad_norm": 1.1946348958109534, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 34634 + }, + { + "epoch": 0.34635, + "grad_norm": 1.9147865256065748, + "learning_rate": 0.003, + "loss": 4.07, + "step": 34635 + }, + { + "epoch": 0.34636, + "grad_norm": 1.3967301039992739, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34636 + }, + { + "epoch": 0.34637, + "grad_norm": 1.5486030087147529, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 34637 + }, + { + "epoch": 0.34638, + "grad_norm": 1.5808868062515018, + "learning_rate": 0.003, + "loss": 4.052, + "step": 34638 + }, + { + "epoch": 0.34639, + "grad_norm": 1.298155204053491, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 34639 + }, + { + "epoch": 0.3464, + "grad_norm": 1.8674334879676187, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 34640 + }, + { + "epoch": 0.34641, + "grad_norm": 1.4659545219724224, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34641 + }, + { + "epoch": 0.34642, + "grad_norm": 1.7199062759393564, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34642 + }, + { + "epoch": 0.34643, + "grad_norm": 1.4087866438555428, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 34643 + }, + { + "epoch": 0.34644, + "grad_norm": 1.5467075988278554, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 34644 + }, + { + "epoch": 0.34645, + "grad_norm": 1.5531998627589392, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 34645 + }, + { + "epoch": 0.34646, + "grad_norm": 1.3704156237381093, + "learning_rate": 0.003, + "loss": 4.047, + "step": 34646 + }, + { + "epoch": 0.34647, + "grad_norm": 1.537372274487844, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34647 + }, + { + "epoch": 0.34648, + "grad_norm": 1.3482396562845043, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 34648 + }, + { + "epoch": 0.34649, + "grad_norm": 1.3145655476748122, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 34649 + }, + { + "epoch": 0.3465, + "grad_norm": 1.594079714986879, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 34650 + }, + { + "epoch": 0.34651, + "grad_norm": 1.5126469551925545, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34651 + }, + { + "epoch": 0.34652, + "grad_norm": 1.3593911450629015, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 34652 + }, + { + "epoch": 0.34653, + "grad_norm": 1.3045236749362805, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 34653 + }, + { + "epoch": 0.34654, + "grad_norm": 1.788846945211801, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34654 + }, + { + "epoch": 0.34655, + "grad_norm": 1.5263837137122769, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 34655 + }, + { + "epoch": 0.34656, + "grad_norm": 1.4104252509570083, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 34656 + }, + { + "epoch": 0.34657, + "grad_norm": 1.2952263856763877, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 34657 + }, + { + "epoch": 0.34658, + "grad_norm": 1.3731110439192724, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 34658 + }, + { + "epoch": 0.34659, + "grad_norm": 1.64314393719272, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 34659 + }, + { + "epoch": 0.3466, + "grad_norm": 1.2308721700920613, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34660 + }, + { + "epoch": 0.34661, + "grad_norm": 1.726958400763375, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34661 + }, + { + "epoch": 0.34662, + "grad_norm": 1.4929798715284557, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 34662 + }, + { + "epoch": 0.34663, + "grad_norm": 1.2792329308364783, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 34663 + }, + { + "epoch": 0.34664, + "grad_norm": 2.01965802436738, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 34664 + }, + { + "epoch": 0.34665, + "grad_norm": 1.8270514116927667, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 34665 + }, + { + "epoch": 0.34666, + "grad_norm": 1.4618496584770175, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 34666 + }, + { + "epoch": 0.34667, + "grad_norm": 1.5468898522512977, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 34667 + }, + { + "epoch": 0.34668, + "grad_norm": 1.2656853640833021, + "learning_rate": 0.003, + "loss": 4.026, + "step": 34668 + }, + { + "epoch": 0.34669, + "grad_norm": 2.171201442269304, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34669 + }, + { + "epoch": 0.3467, + "grad_norm": 1.144210202222561, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 34670 + }, + { + "epoch": 0.34671, + "grad_norm": 1.9822219236332528, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 34671 + }, + { + "epoch": 0.34672, + "grad_norm": 1.5347056523363583, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 34672 + }, + { + "epoch": 0.34673, + "grad_norm": 1.7248909581931922, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 34673 + }, + { + "epoch": 0.34674, + "grad_norm": 1.2413294655787535, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 34674 + }, + { + "epoch": 0.34675, + "grad_norm": 1.519057676901208, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34675 + }, + { + "epoch": 0.34676, + "grad_norm": 1.5856050235489465, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 34676 + }, + { + "epoch": 0.34677, + "grad_norm": 1.5700100911294745, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 34677 + }, + { + "epoch": 0.34678, + "grad_norm": 1.4332193311255517, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 34678 + }, + { + "epoch": 0.34679, + "grad_norm": 1.281861977720368, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 34679 + }, + { + "epoch": 0.3468, + "grad_norm": 1.6269774205303846, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34680 + }, + { + "epoch": 0.34681, + "grad_norm": 1.1897930634434908, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 34681 + }, + { + "epoch": 0.34682, + "grad_norm": 1.5616744034660077, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 34682 + }, + { + "epoch": 0.34683, + "grad_norm": 1.5017945603144907, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 34683 + }, + { + "epoch": 0.34684, + "grad_norm": 1.485945752309125, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34684 + }, + { + "epoch": 0.34685, + "grad_norm": 1.3015834912266182, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 34685 + }, + { + "epoch": 0.34686, + "grad_norm": 1.7353290333514797, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 34686 + }, + { + "epoch": 0.34687, + "grad_norm": 1.5918902899933984, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 34687 + }, + { + "epoch": 0.34688, + "grad_norm": 1.7827119571665373, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 34688 + }, + { + "epoch": 0.34689, + "grad_norm": 1.3740067362407178, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34689 + }, + { + "epoch": 0.3469, + "grad_norm": 1.6118481894027794, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34690 + }, + { + "epoch": 0.34691, + "grad_norm": 1.5139992024123576, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 34691 + }, + { + "epoch": 0.34692, + "grad_norm": 1.2567849496925658, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 34692 + }, + { + "epoch": 0.34693, + "grad_norm": 1.6171790205180545, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 34693 + }, + { + "epoch": 0.34694, + "grad_norm": 1.4015015704174296, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 34694 + }, + { + "epoch": 0.34695, + "grad_norm": 1.455741016669495, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 34695 + }, + { + "epoch": 0.34696, + "grad_norm": 1.2779455886339601, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 34696 + }, + { + "epoch": 0.34697, + "grad_norm": 1.4819864012570492, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 34697 + }, + { + "epoch": 0.34698, + "grad_norm": 1.6145618026838648, + "learning_rate": 0.003, + "loss": 4.058, + "step": 34698 + }, + { + "epoch": 0.34699, + "grad_norm": 1.4858798679951561, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34699 + }, + { + "epoch": 0.347, + "grad_norm": 1.7112894935547525, + "learning_rate": 0.003, + "loss": 4.025, + "step": 34700 + }, + { + "epoch": 0.34701, + "grad_norm": 1.5717920591129415, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 34701 + }, + { + "epoch": 0.34702, + "grad_norm": 1.2952466396957254, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 34702 + }, + { + "epoch": 0.34703, + "grad_norm": 1.7057418365783337, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 34703 + }, + { + "epoch": 0.34704, + "grad_norm": 1.3958101082243097, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 34704 + }, + { + "epoch": 0.34705, + "grad_norm": 1.4444784751350666, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34705 + }, + { + "epoch": 0.34706, + "grad_norm": 1.6383991809579477, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 34706 + }, + { + "epoch": 0.34707, + "grad_norm": 1.4118694691670983, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 34707 + }, + { + "epoch": 0.34708, + "grad_norm": 1.4517288203818302, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34708 + }, + { + "epoch": 0.34709, + "grad_norm": 1.2579919546526623, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34709 + }, + { + "epoch": 0.3471, + "grad_norm": 1.2359626736731166, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 34710 + }, + { + "epoch": 0.34711, + "grad_norm": 1.6222814143309336, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34711 + }, + { + "epoch": 0.34712, + "grad_norm": 1.3072880636990374, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 34712 + }, + { + "epoch": 0.34713, + "grad_norm": 1.4880973380873868, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 34713 + }, + { + "epoch": 0.34714, + "grad_norm": 1.3595853464173717, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34714 + }, + { + "epoch": 0.34715, + "grad_norm": 1.3721049859073886, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34715 + }, + { + "epoch": 0.34716, + "grad_norm": 1.729682549552856, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 34716 + }, + { + "epoch": 0.34717, + "grad_norm": 1.6432599576048061, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 34717 + }, + { + "epoch": 0.34718, + "grad_norm": 1.5909292536708783, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 34718 + }, + { + "epoch": 0.34719, + "grad_norm": 1.3602120468748897, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 34719 + }, + { + "epoch": 0.3472, + "grad_norm": 1.5331766831199816, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 34720 + }, + { + "epoch": 0.34721, + "grad_norm": 1.2585967795041666, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 34721 + }, + { + "epoch": 0.34722, + "grad_norm": 1.5640574539296375, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 34722 + }, + { + "epoch": 0.34723, + "grad_norm": 1.3091448669460701, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 34723 + }, + { + "epoch": 0.34724, + "grad_norm": 1.7060809910954722, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 34724 + }, + { + "epoch": 0.34725, + "grad_norm": 1.3256913185490073, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 34725 + }, + { + "epoch": 0.34726, + "grad_norm": 1.470985307301563, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 34726 + }, + { + "epoch": 0.34727, + "grad_norm": 1.4398819116857724, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34727 + }, + { + "epoch": 0.34728, + "grad_norm": 1.5793038911538329, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 34728 + }, + { + "epoch": 0.34729, + "grad_norm": 1.3762835074215183, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 34729 + }, + { + "epoch": 0.3473, + "grad_norm": 1.4072956088199338, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 34730 + }, + { + "epoch": 0.34731, + "grad_norm": 1.835745298145599, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 34731 + }, + { + "epoch": 0.34732, + "grad_norm": 1.822914553574521, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 34732 + }, + { + "epoch": 0.34733, + "grad_norm": 1.6624717203130794, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34733 + }, + { + "epoch": 0.34734, + "grad_norm": 1.5820602373729247, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 34734 + }, + { + "epoch": 0.34735, + "grad_norm": 1.551674651219849, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 34735 + }, + { + "epoch": 0.34736, + "grad_norm": 1.568573433327253, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 34736 + }, + { + "epoch": 0.34737, + "grad_norm": 1.4584744689043097, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 34737 + }, + { + "epoch": 0.34738, + "grad_norm": 1.4301048031180825, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 34738 + }, + { + "epoch": 0.34739, + "grad_norm": 1.3839790021911038, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 34739 + }, + { + "epoch": 0.3474, + "grad_norm": 1.4123469076690829, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34740 + }, + { + "epoch": 0.34741, + "grad_norm": 1.5233540083262294, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 34741 + }, + { + "epoch": 0.34742, + "grad_norm": 1.4842815142986718, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34742 + }, + { + "epoch": 0.34743, + "grad_norm": 1.3141923879245319, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 34743 + }, + { + "epoch": 0.34744, + "grad_norm": 1.4130198152164242, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 34744 + }, + { + "epoch": 0.34745, + "grad_norm": 1.6176334917686845, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34745 + }, + { + "epoch": 0.34746, + "grad_norm": 1.2289418655715276, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 34746 + }, + { + "epoch": 0.34747, + "grad_norm": 1.9370204142460459, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 34747 + }, + { + "epoch": 0.34748, + "grad_norm": 1.284687181186189, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 34748 + }, + { + "epoch": 0.34749, + "grad_norm": 1.7215844672704983, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 34749 + }, + { + "epoch": 0.3475, + "grad_norm": 1.4609742397652066, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 34750 + }, + { + "epoch": 0.34751, + "grad_norm": 1.3058019712018765, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34751 + }, + { + "epoch": 0.34752, + "grad_norm": 1.689583737322603, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34752 + }, + { + "epoch": 0.34753, + "grad_norm": 1.42723194416402, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 34753 + }, + { + "epoch": 0.34754, + "grad_norm": 1.6150643924252017, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 34754 + }, + { + "epoch": 0.34755, + "grad_norm": 1.3072156016890126, + "learning_rate": 0.003, + "loss": 4.038, + "step": 34755 + }, + { + "epoch": 0.34756, + "grad_norm": 1.688803150439331, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 34756 + }, + { + "epoch": 0.34757, + "grad_norm": 1.437596873759606, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 34757 + }, + { + "epoch": 0.34758, + "grad_norm": 1.4994838185915276, + "learning_rate": 0.003, + "loss": 4.043, + "step": 34758 + }, + { + "epoch": 0.34759, + "grad_norm": 1.7330592916687018, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34759 + }, + { + "epoch": 0.3476, + "grad_norm": 1.3822233343210377, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 34760 + }, + { + "epoch": 0.34761, + "grad_norm": 1.4680066252297785, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34761 + }, + { + "epoch": 0.34762, + "grad_norm": 1.2393453524854379, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 34762 + }, + { + "epoch": 0.34763, + "grad_norm": 1.891645068774812, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34763 + }, + { + "epoch": 0.34764, + "grad_norm": 1.1861960354197418, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 34764 + }, + { + "epoch": 0.34765, + "grad_norm": 1.5459631881488862, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 34765 + }, + { + "epoch": 0.34766, + "grad_norm": 1.463601636153855, + "learning_rate": 0.003, + "loss": 4.025, + "step": 34766 + }, + { + "epoch": 0.34767, + "grad_norm": 1.7618022138075715, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34767 + }, + { + "epoch": 0.34768, + "grad_norm": 1.224351711190075, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 34768 + }, + { + "epoch": 0.34769, + "grad_norm": 1.7917916401227498, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 34769 + }, + { + "epoch": 0.3477, + "grad_norm": 1.32649277230292, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 34770 + }, + { + "epoch": 0.34771, + "grad_norm": 1.470810262141655, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 34771 + }, + { + "epoch": 0.34772, + "grad_norm": 1.2677985288621194, + "learning_rate": 0.003, + "loss": 4.017, + "step": 34772 + }, + { + "epoch": 0.34773, + "grad_norm": 1.8229575341657196, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 34773 + }, + { + "epoch": 0.34774, + "grad_norm": 1.139991426872548, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 34774 + }, + { + "epoch": 0.34775, + "grad_norm": 1.929281748423481, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 34775 + }, + { + "epoch": 0.34776, + "grad_norm": 1.4399248919350638, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 34776 + }, + { + "epoch": 0.34777, + "grad_norm": 1.554680043550741, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 34777 + }, + { + "epoch": 0.34778, + "grad_norm": 1.5013455909446285, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 34778 + }, + { + "epoch": 0.34779, + "grad_norm": 1.7738219590683018, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 34779 + }, + { + "epoch": 0.3478, + "grad_norm": 1.466593808222189, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 34780 + }, + { + "epoch": 0.34781, + "grad_norm": 1.4562328729304306, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 34781 + }, + { + "epoch": 0.34782, + "grad_norm": 1.8076741809638484, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 34782 + }, + { + "epoch": 0.34783, + "grad_norm": 1.421377498006397, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34783 + }, + { + "epoch": 0.34784, + "grad_norm": 1.4247547924659392, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 34784 + }, + { + "epoch": 0.34785, + "grad_norm": 1.329916915285088, + "learning_rate": 0.003, + "loss": 4.041, + "step": 34785 + }, + { + "epoch": 0.34786, + "grad_norm": 1.712777116768574, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 34786 + }, + { + "epoch": 0.34787, + "grad_norm": 1.3940834967814784, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 34787 + }, + { + "epoch": 0.34788, + "grad_norm": 1.509461220105555, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34788 + }, + { + "epoch": 0.34789, + "grad_norm": 1.4305774519498446, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 34789 + }, + { + "epoch": 0.3479, + "grad_norm": 1.7365526797109911, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 34790 + }, + { + "epoch": 0.34791, + "grad_norm": 1.3134583602278422, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 34791 + }, + { + "epoch": 0.34792, + "grad_norm": 1.8672828002198938, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 34792 + }, + { + "epoch": 0.34793, + "grad_norm": 1.407694220135817, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 34793 + }, + { + "epoch": 0.34794, + "grad_norm": 1.5246914422165627, + "learning_rate": 0.003, + "loss": 4.072, + "step": 34794 + }, + { + "epoch": 0.34795, + "grad_norm": 1.7746040542409345, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 34795 + }, + { + "epoch": 0.34796, + "grad_norm": 1.3173554949122317, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 34796 + }, + { + "epoch": 0.34797, + "grad_norm": 1.7334027476911016, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 34797 + }, + { + "epoch": 0.34798, + "grad_norm": 1.4156892617327785, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 34798 + }, + { + "epoch": 0.34799, + "grad_norm": 1.5450025333642377, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 34799 + }, + { + "epoch": 0.348, + "grad_norm": 1.3227672188296777, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 34800 + }, + { + "epoch": 0.34801, + "grad_norm": 1.4350628653982287, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 34801 + }, + { + "epoch": 0.34802, + "grad_norm": 1.351574679407687, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 34802 + }, + { + "epoch": 0.34803, + "grad_norm": 1.590489036645301, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34803 + }, + { + "epoch": 0.34804, + "grad_norm": 1.5268532239271022, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 34804 + }, + { + "epoch": 0.34805, + "grad_norm": 2.0380760925236574, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 34805 + }, + { + "epoch": 0.34806, + "grad_norm": 1.3461033207812139, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 34806 + }, + { + "epoch": 0.34807, + "grad_norm": 1.586166365686095, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 34807 + }, + { + "epoch": 0.34808, + "grad_norm": 1.5854809466991209, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 34808 + }, + { + "epoch": 0.34809, + "grad_norm": 1.1521024883126931, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 34809 + }, + { + "epoch": 0.3481, + "grad_norm": 1.6710680007367353, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 34810 + }, + { + "epoch": 0.34811, + "grad_norm": 1.3573105604404487, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 34811 + }, + { + "epoch": 0.34812, + "grad_norm": 1.6103758550436316, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 34812 + }, + { + "epoch": 0.34813, + "grad_norm": 1.1791986478922152, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34813 + }, + { + "epoch": 0.34814, + "grad_norm": 1.7541984973202087, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34814 + }, + { + "epoch": 0.34815, + "grad_norm": 1.2958864046046936, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34815 + }, + { + "epoch": 0.34816, + "grad_norm": 2.1779953756515127, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34816 + }, + { + "epoch": 0.34817, + "grad_norm": 1.636331030185478, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 34817 + }, + { + "epoch": 0.34818, + "grad_norm": 1.853491705913424, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 34818 + }, + { + "epoch": 0.34819, + "grad_norm": 1.612184963444205, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34819 + }, + { + "epoch": 0.3482, + "grad_norm": 1.5187763066912356, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 34820 + }, + { + "epoch": 0.34821, + "grad_norm": 1.2406667345360392, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34821 + }, + { + "epoch": 0.34822, + "grad_norm": 1.5860913747116432, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 34822 + }, + { + "epoch": 0.34823, + "grad_norm": 1.4384611951528687, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 34823 + }, + { + "epoch": 0.34824, + "grad_norm": 1.4373752263619592, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 34824 + }, + { + "epoch": 0.34825, + "grad_norm": 1.613525950134743, + "learning_rate": 0.003, + "loss": 4.062, + "step": 34825 + }, + { + "epoch": 0.34826, + "grad_norm": 1.5133565045469528, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 34826 + }, + { + "epoch": 0.34827, + "grad_norm": 1.4264231176392144, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 34827 + }, + { + "epoch": 0.34828, + "grad_norm": 1.4725006008296286, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 34828 + }, + { + "epoch": 0.34829, + "grad_norm": 1.3821867478738024, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 34829 + }, + { + "epoch": 0.3483, + "grad_norm": 1.5521445380710854, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 34830 + }, + { + "epoch": 0.34831, + "grad_norm": 1.5232687711549946, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 34831 + }, + { + "epoch": 0.34832, + "grad_norm": 1.5732285035438704, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 34832 + }, + { + "epoch": 0.34833, + "grad_norm": 1.3817429850692486, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 34833 + }, + { + "epoch": 0.34834, + "grad_norm": 1.7777242604145493, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 34834 + }, + { + "epoch": 0.34835, + "grad_norm": 1.8125317215336305, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 34835 + }, + { + "epoch": 0.34836, + "grad_norm": 1.7013280795817196, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 34836 + }, + { + "epoch": 0.34837, + "grad_norm": 1.457797139389556, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 34837 + }, + { + "epoch": 0.34838, + "grad_norm": 1.4925538753816834, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 34838 + }, + { + "epoch": 0.34839, + "grad_norm": 1.6398881281161164, + "learning_rate": 0.003, + "loss": 4.032, + "step": 34839 + }, + { + "epoch": 0.3484, + "grad_norm": 1.20830546106118, + "learning_rate": 0.003, + "loss": 4.045, + "step": 34840 + }, + { + "epoch": 0.34841, + "grad_norm": 1.7490787059128745, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 34841 + }, + { + "epoch": 0.34842, + "grad_norm": 1.4981978194070498, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 34842 + }, + { + "epoch": 0.34843, + "grad_norm": 1.5879828913847558, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 34843 + }, + { + "epoch": 0.34844, + "grad_norm": 1.1829578641149365, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 34844 + }, + { + "epoch": 0.34845, + "grad_norm": 1.4685268013251218, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 34845 + }, + { + "epoch": 0.34846, + "grad_norm": 1.3352460184888955, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 34846 + }, + { + "epoch": 0.34847, + "grad_norm": 1.7682327277029992, + "learning_rate": 0.003, + "loss": 4.03, + "step": 34847 + }, + { + "epoch": 0.34848, + "grad_norm": 1.4416549773007796, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 34848 + }, + { + "epoch": 0.34849, + "grad_norm": 1.2765528052363075, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 34849 + }, + { + "epoch": 0.3485, + "grad_norm": 1.5831214449633764, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 34850 + }, + { + "epoch": 0.34851, + "grad_norm": 1.2470340947217715, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34851 + }, + { + "epoch": 0.34852, + "grad_norm": 1.7975200749657654, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 34852 + }, + { + "epoch": 0.34853, + "grad_norm": 1.1753477008588251, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 34853 + }, + { + "epoch": 0.34854, + "grad_norm": 1.7644010896452307, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 34854 + }, + { + "epoch": 0.34855, + "grad_norm": 1.2589723080910058, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 34855 + }, + { + "epoch": 0.34856, + "grad_norm": 1.5916032832774247, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 34856 + }, + { + "epoch": 0.34857, + "grad_norm": 1.4073473661264146, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 34857 + }, + { + "epoch": 0.34858, + "grad_norm": 1.4976066973720512, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 34858 + }, + { + "epoch": 0.34859, + "grad_norm": 1.567043746288624, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34859 + }, + { + "epoch": 0.3486, + "grad_norm": 2.013120874158658, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 34860 + }, + { + "epoch": 0.34861, + "grad_norm": 1.2300743211857215, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 34861 + }, + { + "epoch": 0.34862, + "grad_norm": 1.6169362891334156, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 34862 + }, + { + "epoch": 0.34863, + "grad_norm": 1.5082447042116338, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34863 + }, + { + "epoch": 0.34864, + "grad_norm": 1.585061579197254, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 34864 + }, + { + "epoch": 0.34865, + "grad_norm": 1.4883557619491057, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 34865 + }, + { + "epoch": 0.34866, + "grad_norm": 1.4994421655163446, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 34866 + }, + { + "epoch": 0.34867, + "grad_norm": 1.464801533222055, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 34867 + }, + { + "epoch": 0.34868, + "grad_norm": 1.4793184870577405, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 34868 + }, + { + "epoch": 0.34869, + "grad_norm": 1.5724360941778814, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 34869 + }, + { + "epoch": 0.3487, + "grad_norm": 1.320982714089201, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 34870 + }, + { + "epoch": 0.34871, + "grad_norm": 1.5461224112958705, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 34871 + }, + { + "epoch": 0.34872, + "grad_norm": 1.8576671710073676, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 34872 + }, + { + "epoch": 0.34873, + "grad_norm": 1.231807900873598, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 34873 + }, + { + "epoch": 0.34874, + "grad_norm": 1.4443691583294513, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 34874 + }, + { + "epoch": 0.34875, + "grad_norm": 1.7596744783302583, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 34875 + }, + { + "epoch": 0.34876, + "grad_norm": 1.3517301749787256, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 34876 + }, + { + "epoch": 0.34877, + "grad_norm": 1.5622284472714323, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 34877 + }, + { + "epoch": 0.34878, + "grad_norm": 1.4925414818129707, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 34878 + }, + { + "epoch": 0.34879, + "grad_norm": 1.5465404811954768, + "learning_rate": 0.003, + "loss": 4.04, + "step": 34879 + }, + { + "epoch": 0.3488, + "grad_norm": 1.5301438682060637, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 34880 + }, + { + "epoch": 0.34881, + "grad_norm": 1.332262190158077, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 34881 + }, + { + "epoch": 0.34882, + "grad_norm": 1.7622025896663516, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 34882 + }, + { + "epoch": 0.34883, + "grad_norm": 1.4083495433327031, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 34883 + }, + { + "epoch": 0.34884, + "grad_norm": 1.5048882341947143, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 34884 + }, + { + "epoch": 0.34885, + "grad_norm": 1.5795882822316647, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 34885 + }, + { + "epoch": 0.34886, + "grad_norm": 1.7545455765517077, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 34886 + }, + { + "epoch": 0.34887, + "grad_norm": 1.3472958996343136, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 34887 + }, + { + "epoch": 0.34888, + "grad_norm": 1.7685143877969305, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 34888 + }, + { + "epoch": 0.34889, + "grad_norm": 1.3860854347229943, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34889 + }, + { + "epoch": 0.3489, + "grad_norm": 1.4396215659379241, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 34890 + }, + { + "epoch": 0.34891, + "grad_norm": 1.3924978785614825, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 34891 + }, + { + "epoch": 0.34892, + "grad_norm": 1.5489123598064616, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 34892 + }, + { + "epoch": 0.34893, + "grad_norm": 1.3101874671956957, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 34893 + }, + { + "epoch": 0.34894, + "grad_norm": 1.4871021304522392, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 34894 + }, + { + "epoch": 0.34895, + "grad_norm": 1.5746180579604936, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 34895 + }, + { + "epoch": 0.34896, + "grad_norm": 1.2481368826399142, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 34896 + }, + { + "epoch": 0.34897, + "grad_norm": 1.5571101356223676, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 34897 + }, + { + "epoch": 0.34898, + "grad_norm": 1.4873997514840198, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 34898 + }, + { + "epoch": 0.34899, + "grad_norm": 1.7605592364840348, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 34899 + }, + { + "epoch": 0.349, + "grad_norm": 1.279160801208172, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 34900 + }, + { + "epoch": 0.34901, + "grad_norm": 1.7878415235667882, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 34901 + }, + { + "epoch": 0.34902, + "grad_norm": 1.6863348675502596, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 34902 + }, + { + "epoch": 0.34903, + "grad_norm": 1.4800470806973527, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 34903 + }, + { + "epoch": 0.34904, + "grad_norm": 1.6499016704317697, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 34904 + }, + { + "epoch": 0.34905, + "grad_norm": 1.4041685341533925, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 34905 + }, + { + "epoch": 0.34906, + "grad_norm": 1.42212548065224, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 34906 + }, + { + "epoch": 0.34907, + "grad_norm": 1.2543141849730277, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 34907 + }, + { + "epoch": 0.34908, + "grad_norm": 1.3541617164276847, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 34908 + }, + { + "epoch": 0.34909, + "grad_norm": 1.5463653892431641, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 34909 + }, + { + "epoch": 0.3491, + "grad_norm": 1.6372257559491885, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 34910 + }, + { + "epoch": 0.34911, + "grad_norm": 1.4567511856478828, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34911 + }, + { + "epoch": 0.34912, + "grad_norm": 1.878787584339788, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 34912 + }, + { + "epoch": 0.34913, + "grad_norm": 1.4166077129182515, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 34913 + }, + { + "epoch": 0.34914, + "grad_norm": 1.3005708570944945, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 34914 + }, + { + "epoch": 0.34915, + "grad_norm": 1.496846018638403, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 34915 + }, + { + "epoch": 0.34916, + "grad_norm": 1.4169840403125074, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 34916 + }, + { + "epoch": 0.34917, + "grad_norm": 1.4400968736816733, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 34917 + }, + { + "epoch": 0.34918, + "grad_norm": 1.2375846557342745, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 34918 + }, + { + "epoch": 0.34919, + "grad_norm": 1.388606409893364, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 34919 + }, + { + "epoch": 0.3492, + "grad_norm": 1.7187788494644636, + "learning_rate": 0.003, + "loss": 4.057, + "step": 34920 + }, + { + "epoch": 0.34921, + "grad_norm": 1.3066158165852981, + "learning_rate": 0.003, + "loss": 4.023, + "step": 34921 + }, + { + "epoch": 0.34922, + "grad_norm": 1.9110442467402653, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 34922 + }, + { + "epoch": 0.34923, + "grad_norm": 1.5507366298128558, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 34923 + }, + { + "epoch": 0.34924, + "grad_norm": 1.636114755135031, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 34924 + }, + { + "epoch": 0.34925, + "grad_norm": 1.6336712231231942, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 34925 + }, + { + "epoch": 0.34926, + "grad_norm": 1.3527273927791759, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 34926 + }, + { + "epoch": 0.34927, + "grad_norm": 1.8317003399423577, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 34927 + }, + { + "epoch": 0.34928, + "grad_norm": 1.4851754269951962, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 34928 + }, + { + "epoch": 0.34929, + "grad_norm": 1.255334997749649, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34929 + }, + { + "epoch": 0.3493, + "grad_norm": 1.6572995625526525, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 34930 + }, + { + "epoch": 0.34931, + "grad_norm": 1.2266219782140393, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 34931 + }, + { + "epoch": 0.34932, + "grad_norm": 1.853986518300175, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34932 + }, + { + "epoch": 0.34933, + "grad_norm": 1.1895607003032822, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 34933 + }, + { + "epoch": 0.34934, + "grad_norm": 1.6002921108489512, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 34934 + }, + { + "epoch": 0.34935, + "grad_norm": 1.6169944088244426, + "learning_rate": 0.003, + "loss": 4.059, + "step": 34935 + }, + { + "epoch": 0.34936, + "grad_norm": 1.5865416056164519, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 34936 + }, + { + "epoch": 0.34937, + "grad_norm": 1.590063524945143, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 34937 + }, + { + "epoch": 0.34938, + "grad_norm": 1.4352860905968672, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 34938 + }, + { + "epoch": 0.34939, + "grad_norm": 1.5457997522509788, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 34939 + }, + { + "epoch": 0.3494, + "grad_norm": 1.3250605562971778, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 34940 + }, + { + "epoch": 0.34941, + "grad_norm": 1.4982777147199453, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 34941 + }, + { + "epoch": 0.34942, + "grad_norm": 1.3657435063841508, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 34942 + }, + { + "epoch": 0.34943, + "grad_norm": 1.8201090339635526, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 34943 + }, + { + "epoch": 0.34944, + "grad_norm": 1.776898992324537, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 34944 + }, + { + "epoch": 0.34945, + "grad_norm": 1.5773580108232486, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34945 + }, + { + "epoch": 0.34946, + "grad_norm": 1.4930511276552971, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34946 + }, + { + "epoch": 0.34947, + "grad_norm": 1.733076213990444, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 34947 + }, + { + "epoch": 0.34948, + "grad_norm": 1.5156519011246605, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 34948 + }, + { + "epoch": 0.34949, + "grad_norm": 1.4061117871350315, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34949 + }, + { + "epoch": 0.3495, + "grad_norm": 1.530760046015532, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 34950 + }, + { + "epoch": 0.34951, + "grad_norm": 1.5669522313510054, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 34951 + }, + { + "epoch": 0.34952, + "grad_norm": 1.40870566404172, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 34952 + }, + { + "epoch": 0.34953, + "grad_norm": 1.2404894623166347, + "learning_rate": 0.003, + "loss": 4.029, + "step": 34953 + }, + { + "epoch": 0.34954, + "grad_norm": 1.8402160562313028, + "learning_rate": 0.003, + "loss": 4.046, + "step": 34954 + }, + { + "epoch": 0.34955, + "grad_norm": 1.2207976970024994, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 34955 + }, + { + "epoch": 0.34956, + "grad_norm": 1.569687451562287, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 34956 + }, + { + "epoch": 0.34957, + "grad_norm": 1.781183227734289, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 34957 + }, + { + "epoch": 0.34958, + "grad_norm": 1.2989516011956013, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 34958 + }, + { + "epoch": 0.34959, + "grad_norm": 1.5850457209037394, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 34959 + }, + { + "epoch": 0.3496, + "grad_norm": 1.3416054832441335, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 34960 + }, + { + "epoch": 0.34961, + "grad_norm": 1.5882407878611964, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 34961 + }, + { + "epoch": 0.34962, + "grad_norm": 1.5332004835951103, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 34962 + }, + { + "epoch": 0.34963, + "grad_norm": 1.3880601560451549, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 34963 + }, + { + "epoch": 0.34964, + "grad_norm": 1.7998958872479678, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 34964 + }, + { + "epoch": 0.34965, + "grad_norm": 1.1648496500212022, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 34965 + }, + { + "epoch": 0.34966, + "grad_norm": 1.689379263320182, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 34966 + }, + { + "epoch": 0.34967, + "grad_norm": 1.4176946664000851, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 34967 + }, + { + "epoch": 0.34968, + "grad_norm": 1.4843640026986937, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 34968 + }, + { + "epoch": 0.34969, + "grad_norm": 1.3964695624129653, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 34969 + }, + { + "epoch": 0.3497, + "grad_norm": 1.512435678994243, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 34970 + }, + { + "epoch": 0.34971, + "grad_norm": 1.429249391930057, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 34971 + }, + { + "epoch": 0.34972, + "grad_norm": 1.3957416114397676, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 34972 + }, + { + "epoch": 0.34973, + "grad_norm": 1.5472474167309533, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 34973 + }, + { + "epoch": 0.34974, + "grad_norm": 1.6504980695330305, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 34974 + }, + { + "epoch": 0.34975, + "grad_norm": 1.7289114278058781, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34975 + }, + { + "epoch": 0.34976, + "grad_norm": 1.3665974277884683, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 34976 + }, + { + "epoch": 0.34977, + "grad_norm": 1.4611998267824542, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 34977 + }, + { + "epoch": 0.34978, + "grad_norm": 1.3718350785032374, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 34978 + }, + { + "epoch": 0.34979, + "grad_norm": 1.5573462670581735, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 34979 + }, + { + "epoch": 0.3498, + "grad_norm": 1.4387510851242173, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 34980 + }, + { + "epoch": 0.34981, + "grad_norm": 1.6565049202869069, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 34981 + }, + { + "epoch": 0.34982, + "grad_norm": 1.4371917261542688, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 34982 + }, + { + "epoch": 0.34983, + "grad_norm": 1.650138189911944, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 34983 + }, + { + "epoch": 0.34984, + "grad_norm": 1.2788595378383014, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 34984 + }, + { + "epoch": 0.34985, + "grad_norm": 1.9674928941281082, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 34985 + }, + { + "epoch": 0.34986, + "grad_norm": 1.2153826191485022, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 34986 + }, + { + "epoch": 0.34987, + "grad_norm": 1.71794079173821, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 34987 + }, + { + "epoch": 0.34988, + "grad_norm": 1.5860904661527848, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 34988 + }, + { + "epoch": 0.34989, + "grad_norm": 1.4832627582173743, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 34989 + }, + { + "epoch": 0.3499, + "grad_norm": 1.3712156154006752, + "learning_rate": 0.003, + "loss": 4.071, + "step": 34990 + }, + { + "epoch": 0.34991, + "grad_norm": 1.4089996891221641, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 34991 + }, + { + "epoch": 0.34992, + "grad_norm": 1.6027485464985902, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 34992 + }, + { + "epoch": 0.34993, + "grad_norm": 1.3846766959224888, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34993 + }, + { + "epoch": 0.34994, + "grad_norm": 1.5505469185081142, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 34994 + }, + { + "epoch": 0.34995, + "grad_norm": 1.2891140136292196, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 34995 + }, + { + "epoch": 0.34996, + "grad_norm": 1.558762105867318, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 34996 + }, + { + "epoch": 0.34997, + "grad_norm": 1.3637667535782982, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 34997 + }, + { + "epoch": 0.34998, + "grad_norm": 1.6381073031674054, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 34998 + }, + { + "epoch": 0.34999, + "grad_norm": 1.5392306703992635, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 34999 + }, + { + "epoch": 0.35, + "grad_norm": 1.4506686554311263, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 35000 + }, + { + "epoch": 0.35001, + "grad_norm": 1.4701307078634847, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 35001 + }, + { + "epoch": 0.35002, + "grad_norm": 1.612544512301308, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 35002 + }, + { + "epoch": 0.35003, + "grad_norm": 1.5800343128437986, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 35003 + }, + { + "epoch": 0.35004, + "grad_norm": 1.6568207421733887, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35004 + }, + { + "epoch": 0.35005, + "grad_norm": 1.4659404109920706, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 35005 + }, + { + "epoch": 0.35006, + "grad_norm": 1.4279411945580414, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35006 + }, + { + "epoch": 0.35007, + "grad_norm": 1.6019524457022365, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35007 + }, + { + "epoch": 0.35008, + "grad_norm": 1.3481009479745787, + "learning_rate": 0.003, + "loss": 4.07, + "step": 35008 + }, + { + "epoch": 0.35009, + "grad_norm": 1.9563312299224507, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 35009 + }, + { + "epoch": 0.3501, + "grad_norm": 1.0869720735607984, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35010 + }, + { + "epoch": 0.35011, + "grad_norm": 1.8188220978467928, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35011 + }, + { + "epoch": 0.35012, + "grad_norm": 1.5202220706066871, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35012 + }, + { + "epoch": 0.35013, + "grad_norm": 1.4777742072292106, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 35013 + }, + { + "epoch": 0.35014, + "grad_norm": 1.680690300481342, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 35014 + }, + { + "epoch": 0.35015, + "grad_norm": 1.437082357756077, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 35015 + }, + { + "epoch": 0.35016, + "grad_norm": 1.579181908517349, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35016 + }, + { + "epoch": 0.35017, + "grad_norm": 1.310327495010364, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35017 + }, + { + "epoch": 0.35018, + "grad_norm": 1.7919026019482258, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 35018 + }, + { + "epoch": 0.35019, + "grad_norm": 1.2906982805201965, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35019 + }, + { + "epoch": 0.3502, + "grad_norm": 1.4877340080185908, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 35020 + }, + { + "epoch": 0.35021, + "grad_norm": 1.7318960803450478, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 35021 + }, + { + "epoch": 0.35022, + "grad_norm": 1.229096661557151, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35022 + }, + { + "epoch": 0.35023, + "grad_norm": 1.6039058906751404, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35023 + }, + { + "epoch": 0.35024, + "grad_norm": 1.3020936922075699, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35024 + }, + { + "epoch": 0.35025, + "grad_norm": 1.4160826211343964, + "learning_rate": 0.003, + "loss": 4.055, + "step": 35025 + }, + { + "epoch": 0.35026, + "grad_norm": 1.5771524242920925, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35026 + }, + { + "epoch": 0.35027, + "grad_norm": 1.4709119086882427, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35027 + }, + { + "epoch": 0.35028, + "grad_norm": 1.5755406999009538, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 35028 + }, + { + "epoch": 0.35029, + "grad_norm": 1.3469965077677692, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 35029 + }, + { + "epoch": 0.3503, + "grad_norm": 1.8277158427081266, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 35030 + }, + { + "epoch": 0.35031, + "grad_norm": 1.2393759904139674, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 35031 + }, + { + "epoch": 0.35032, + "grad_norm": 1.8158865431127273, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 35032 + }, + { + "epoch": 0.35033, + "grad_norm": 1.5354578844777746, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35033 + }, + { + "epoch": 0.35034, + "grad_norm": 1.555156684346834, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 35034 + }, + { + "epoch": 0.35035, + "grad_norm": 1.5516265907760958, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 35035 + }, + { + "epoch": 0.35036, + "grad_norm": 1.6819187700966454, + "learning_rate": 0.003, + "loss": 4.068, + "step": 35036 + }, + { + "epoch": 0.35037, + "grad_norm": 1.3310089605183155, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 35037 + }, + { + "epoch": 0.35038, + "grad_norm": 1.468269036883575, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35038 + }, + { + "epoch": 0.35039, + "grad_norm": 1.6362795679254383, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35039 + }, + { + "epoch": 0.3504, + "grad_norm": 1.2807509138867885, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 35040 + }, + { + "epoch": 0.35041, + "grad_norm": 1.5759918811268168, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 35041 + }, + { + "epoch": 0.35042, + "grad_norm": 1.1661946406133432, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 35042 + }, + { + "epoch": 0.35043, + "grad_norm": 1.8207191577708666, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35043 + }, + { + "epoch": 0.35044, + "grad_norm": 1.131081778163181, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 35044 + }, + { + "epoch": 0.35045, + "grad_norm": 2.139317612693495, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 35045 + }, + { + "epoch": 0.35046, + "grad_norm": 1.4388019364597038, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 35046 + }, + { + "epoch": 0.35047, + "grad_norm": 1.8549363762369986, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 35047 + }, + { + "epoch": 0.35048, + "grad_norm": 1.7705044460526678, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 35048 + }, + { + "epoch": 0.35049, + "grad_norm": 1.1772338541011131, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35049 + }, + { + "epoch": 0.3505, + "grad_norm": 1.3928381410516888, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 35050 + }, + { + "epoch": 0.35051, + "grad_norm": 1.6327427853095655, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 35051 + }, + { + "epoch": 0.35052, + "grad_norm": 1.5663272032703335, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 35052 + }, + { + "epoch": 0.35053, + "grad_norm": 1.7096656265051173, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 35053 + }, + { + "epoch": 0.35054, + "grad_norm": 1.536473941266241, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 35054 + }, + { + "epoch": 0.35055, + "grad_norm": 1.529570850216347, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 35055 + }, + { + "epoch": 0.35056, + "grad_norm": 2.0479520893617815, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 35056 + }, + { + "epoch": 0.35057, + "grad_norm": 1.279595081351948, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 35057 + }, + { + "epoch": 0.35058, + "grad_norm": 1.6864435658279016, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 35058 + }, + { + "epoch": 0.35059, + "grad_norm": 1.7055299188447715, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 35059 + }, + { + "epoch": 0.3506, + "grad_norm": 1.3203935590917864, + "learning_rate": 0.003, + "loss": 4.07, + "step": 35060 + }, + { + "epoch": 0.35061, + "grad_norm": 1.4903358748520406, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 35061 + }, + { + "epoch": 0.35062, + "grad_norm": 1.2310693653558127, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 35062 + }, + { + "epoch": 0.35063, + "grad_norm": 1.9047530097220806, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 35063 + }, + { + "epoch": 0.35064, + "grad_norm": 1.129330042425607, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35064 + }, + { + "epoch": 0.35065, + "grad_norm": 1.6720687352276686, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 35065 + }, + { + "epoch": 0.35066, + "grad_norm": 1.1964285598955127, + "learning_rate": 0.003, + "loss": 4.067, + "step": 35066 + }, + { + "epoch": 0.35067, + "grad_norm": 1.5045383471885994, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35067 + }, + { + "epoch": 0.35068, + "grad_norm": 1.5625769672328491, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 35068 + }, + { + "epoch": 0.35069, + "grad_norm": 1.56597658041091, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 35069 + }, + { + "epoch": 0.3507, + "grad_norm": 1.6584357981808056, + "learning_rate": 0.003, + "loss": 4.057, + "step": 35070 + }, + { + "epoch": 0.35071, + "grad_norm": 1.5855672361486086, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35071 + }, + { + "epoch": 0.35072, + "grad_norm": 1.3391699936591297, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 35072 + }, + { + "epoch": 0.35073, + "grad_norm": 1.8866020877026188, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 35073 + }, + { + "epoch": 0.35074, + "grad_norm": 1.609953506670132, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 35074 + }, + { + "epoch": 0.35075, + "grad_norm": 1.4205384793308544, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 35075 + }, + { + "epoch": 0.35076, + "grad_norm": 1.6032759734978301, + "learning_rate": 0.003, + "loss": 4.04, + "step": 35076 + }, + { + "epoch": 0.35077, + "grad_norm": 1.4670393567830948, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 35077 + }, + { + "epoch": 0.35078, + "grad_norm": 1.4295695334904224, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35078 + }, + { + "epoch": 0.35079, + "grad_norm": 1.4711304635047924, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 35079 + }, + { + "epoch": 0.3508, + "grad_norm": 1.4152950749843336, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35080 + }, + { + "epoch": 0.35081, + "grad_norm": 1.517335087444576, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 35081 + }, + { + "epoch": 0.35082, + "grad_norm": 1.6293146636135474, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 35082 + }, + { + "epoch": 0.35083, + "grad_norm": 1.5654497177479383, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 35083 + }, + { + "epoch": 0.35084, + "grad_norm": 1.9480497595471986, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 35084 + }, + { + "epoch": 0.35085, + "grad_norm": 1.323174888192764, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 35085 + }, + { + "epoch": 0.35086, + "grad_norm": 1.2328865726375453, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35086 + }, + { + "epoch": 0.35087, + "grad_norm": 1.5075240084199106, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 35087 + }, + { + "epoch": 0.35088, + "grad_norm": 1.4165236666549432, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 35088 + }, + { + "epoch": 0.35089, + "grad_norm": 1.339710037516548, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 35089 + }, + { + "epoch": 0.3509, + "grad_norm": 1.67571218031503, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 35090 + }, + { + "epoch": 0.35091, + "grad_norm": 1.429289689404427, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 35091 + }, + { + "epoch": 0.35092, + "grad_norm": 1.627067055704178, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 35092 + }, + { + "epoch": 0.35093, + "grad_norm": 1.4216677316284665, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 35093 + }, + { + "epoch": 0.35094, + "grad_norm": 1.7013883752468908, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35094 + }, + { + "epoch": 0.35095, + "grad_norm": 1.5292736892319327, + "learning_rate": 0.003, + "loss": 4.025, + "step": 35095 + }, + { + "epoch": 0.35096, + "grad_norm": 1.5778796512335687, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35096 + }, + { + "epoch": 0.35097, + "grad_norm": 1.4662571623206933, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 35097 + }, + { + "epoch": 0.35098, + "grad_norm": 1.2806278571173684, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 35098 + }, + { + "epoch": 0.35099, + "grad_norm": 1.5174032689016754, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 35099 + }, + { + "epoch": 0.351, + "grad_norm": 1.6586271505099663, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 35100 + }, + { + "epoch": 0.35101, + "grad_norm": 1.2517765498905344, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 35101 + }, + { + "epoch": 0.35102, + "grad_norm": 1.6834062168579744, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35102 + }, + { + "epoch": 0.35103, + "grad_norm": 1.36864793337455, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35103 + }, + { + "epoch": 0.35104, + "grad_norm": 1.6896539357834668, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 35104 + }, + { + "epoch": 0.35105, + "grad_norm": 1.6823759861207006, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 35105 + }, + { + "epoch": 0.35106, + "grad_norm": 1.6033678873957236, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 35106 + }, + { + "epoch": 0.35107, + "grad_norm": 1.5339304794622044, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 35107 + }, + { + "epoch": 0.35108, + "grad_norm": 1.4152796496498756, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35108 + }, + { + "epoch": 0.35109, + "grad_norm": 1.3320989337185933, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 35109 + }, + { + "epoch": 0.3511, + "grad_norm": 1.462371826262672, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 35110 + }, + { + "epoch": 0.35111, + "grad_norm": 1.4432489476616976, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 35111 + }, + { + "epoch": 0.35112, + "grad_norm": 1.779186927753767, + "learning_rate": 0.003, + "loss": 4.037, + "step": 35112 + }, + { + "epoch": 0.35113, + "grad_norm": 1.405209795387293, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35113 + }, + { + "epoch": 0.35114, + "grad_norm": 1.7111165714462229, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 35114 + }, + { + "epoch": 0.35115, + "grad_norm": 1.4217466283221116, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 35115 + }, + { + "epoch": 0.35116, + "grad_norm": 1.5710962854291306, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 35116 + }, + { + "epoch": 0.35117, + "grad_norm": 1.347421766677916, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 35117 + }, + { + "epoch": 0.35118, + "grad_norm": 1.7613256040500298, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 35118 + }, + { + "epoch": 0.35119, + "grad_norm": 1.2292057261479972, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 35119 + }, + { + "epoch": 0.3512, + "grad_norm": 1.8024745626733085, + "learning_rate": 0.003, + "loss": 4.033, + "step": 35120 + }, + { + "epoch": 0.35121, + "grad_norm": 1.6249805376570083, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 35121 + }, + { + "epoch": 0.35122, + "grad_norm": 1.4062998463352012, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 35122 + }, + { + "epoch": 0.35123, + "grad_norm": 1.536381662814781, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 35123 + }, + { + "epoch": 0.35124, + "grad_norm": 1.4766031158986859, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 35124 + }, + { + "epoch": 0.35125, + "grad_norm": 1.55210648799449, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35125 + }, + { + "epoch": 0.35126, + "grad_norm": 1.4810777463243834, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 35126 + }, + { + "epoch": 0.35127, + "grad_norm": 1.3966911594527969, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 35127 + }, + { + "epoch": 0.35128, + "grad_norm": 1.4055320980458703, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 35128 + }, + { + "epoch": 0.35129, + "grad_norm": 1.590518379148855, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 35129 + }, + { + "epoch": 0.3513, + "grad_norm": 1.2930865603478874, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35130 + }, + { + "epoch": 0.35131, + "grad_norm": 1.7141733336120264, + "learning_rate": 0.003, + "loss": 4.056, + "step": 35131 + }, + { + "epoch": 0.35132, + "grad_norm": 1.5875824423280955, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35132 + }, + { + "epoch": 0.35133, + "grad_norm": 1.874153404692566, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 35133 + }, + { + "epoch": 0.35134, + "grad_norm": 1.2951656959919506, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 35134 + }, + { + "epoch": 0.35135, + "grad_norm": 1.5497852112129193, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35135 + }, + { + "epoch": 0.35136, + "grad_norm": 1.5516029435300913, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35136 + }, + { + "epoch": 0.35137, + "grad_norm": 1.4189245457769355, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35137 + }, + { + "epoch": 0.35138, + "grad_norm": 1.3762850011962335, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 35138 + }, + { + "epoch": 0.35139, + "grad_norm": 1.5696360476341968, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35139 + }, + { + "epoch": 0.3514, + "grad_norm": 1.4021756878893052, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 35140 + }, + { + "epoch": 0.35141, + "grad_norm": 1.3899306735026886, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 35141 + }, + { + "epoch": 0.35142, + "grad_norm": 1.6001097018992179, + "learning_rate": 0.003, + "loss": 4.065, + "step": 35142 + }, + { + "epoch": 0.35143, + "grad_norm": 1.6313068295495994, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35143 + }, + { + "epoch": 0.35144, + "grad_norm": 1.691110351129734, + "learning_rate": 0.003, + "loss": 4.067, + "step": 35144 + }, + { + "epoch": 0.35145, + "grad_norm": 1.6745176752936568, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35145 + }, + { + "epoch": 0.35146, + "grad_norm": 1.4302091308537164, + "learning_rate": 0.003, + "loss": 4.037, + "step": 35146 + }, + { + "epoch": 0.35147, + "grad_norm": 1.5774094799198712, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35147 + }, + { + "epoch": 0.35148, + "grad_norm": 1.5259460661346609, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 35148 + }, + { + "epoch": 0.35149, + "grad_norm": 1.4701900215510064, + "learning_rate": 0.003, + "loss": 4.028, + "step": 35149 + }, + { + "epoch": 0.3515, + "grad_norm": 1.4037127454557148, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35150 + }, + { + "epoch": 0.35151, + "grad_norm": 1.7090161615604373, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 35151 + }, + { + "epoch": 0.35152, + "grad_norm": 1.2417460730941758, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 35152 + }, + { + "epoch": 0.35153, + "grad_norm": 1.4617563254269679, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35153 + }, + { + "epoch": 0.35154, + "grad_norm": 1.4966819337637665, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 35154 + }, + { + "epoch": 0.35155, + "grad_norm": 1.4069187539331385, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 35155 + }, + { + "epoch": 0.35156, + "grad_norm": 1.6820822182414212, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35156 + }, + { + "epoch": 0.35157, + "grad_norm": 1.502911700078775, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 35157 + }, + { + "epoch": 0.35158, + "grad_norm": 1.469803846135169, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 35158 + }, + { + "epoch": 0.35159, + "grad_norm": 1.7918341397103434, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35159 + }, + { + "epoch": 0.3516, + "grad_norm": 1.315449235897319, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 35160 + }, + { + "epoch": 0.35161, + "grad_norm": 1.5403690578638707, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35161 + }, + { + "epoch": 0.35162, + "grad_norm": 1.8529234874711662, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 35162 + }, + { + "epoch": 0.35163, + "grad_norm": 1.127970119161522, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 35163 + }, + { + "epoch": 0.35164, + "grad_norm": 1.8902236823014662, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 35164 + }, + { + "epoch": 0.35165, + "grad_norm": 1.2282384377784732, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 35165 + }, + { + "epoch": 0.35166, + "grad_norm": 1.7106651688998944, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 35166 + }, + { + "epoch": 0.35167, + "grad_norm": 1.5645585714866104, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 35167 + }, + { + "epoch": 0.35168, + "grad_norm": 1.4062815506863307, + "learning_rate": 0.003, + "loss": 4.013, + "step": 35168 + }, + { + "epoch": 0.35169, + "grad_norm": 1.854155095325715, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35169 + }, + { + "epoch": 0.3517, + "grad_norm": 1.2516523389236796, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35170 + }, + { + "epoch": 0.35171, + "grad_norm": 1.7203561911182643, + "learning_rate": 0.003, + "loss": 4.039, + "step": 35171 + }, + { + "epoch": 0.35172, + "grad_norm": 1.375254857961233, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 35172 + }, + { + "epoch": 0.35173, + "grad_norm": 1.470197302940404, + "learning_rate": 0.003, + "loss": 4.052, + "step": 35173 + }, + { + "epoch": 0.35174, + "grad_norm": 1.4499180407343546, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 35174 + }, + { + "epoch": 0.35175, + "grad_norm": 1.5028865174082893, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 35175 + }, + { + "epoch": 0.35176, + "grad_norm": 1.4380995035765776, + "learning_rate": 0.003, + "loss": 4.02, + "step": 35176 + }, + { + "epoch": 0.35177, + "grad_norm": 1.293318129116422, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35177 + }, + { + "epoch": 0.35178, + "grad_norm": 1.5842533675141528, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 35178 + }, + { + "epoch": 0.35179, + "grad_norm": 1.3611351069604283, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 35179 + }, + { + "epoch": 0.3518, + "grad_norm": 1.578844635131463, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 35180 + }, + { + "epoch": 0.35181, + "grad_norm": 1.4375570961890831, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 35181 + }, + { + "epoch": 0.35182, + "grad_norm": 1.535618871567259, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 35182 + }, + { + "epoch": 0.35183, + "grad_norm": 1.8410689984218447, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 35183 + }, + { + "epoch": 0.35184, + "grad_norm": 1.1581896662243762, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 35184 + }, + { + "epoch": 0.35185, + "grad_norm": 1.8413435181356688, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 35185 + }, + { + "epoch": 0.35186, + "grad_norm": 1.2472240153472531, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 35186 + }, + { + "epoch": 0.35187, + "grad_norm": 1.5654275414882437, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 35187 + }, + { + "epoch": 0.35188, + "grad_norm": 1.6813040802716803, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 35188 + }, + { + "epoch": 0.35189, + "grad_norm": 1.4107390316822992, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 35189 + }, + { + "epoch": 0.3519, + "grad_norm": 1.6207879319225023, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 35190 + }, + { + "epoch": 0.35191, + "grad_norm": 1.5028326439661608, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 35191 + }, + { + "epoch": 0.35192, + "grad_norm": 1.7280135324508052, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35192 + }, + { + "epoch": 0.35193, + "grad_norm": 1.1743207390968153, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 35193 + }, + { + "epoch": 0.35194, + "grad_norm": 1.5141392141591656, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35194 + }, + { + "epoch": 0.35195, + "grad_norm": 1.778215619462536, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 35195 + }, + { + "epoch": 0.35196, + "grad_norm": 1.1009986022489147, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 35196 + }, + { + "epoch": 0.35197, + "grad_norm": 1.7036506302356713, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 35197 + }, + { + "epoch": 0.35198, + "grad_norm": 1.3280144936407368, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35198 + }, + { + "epoch": 0.35199, + "grad_norm": 1.5007943390490905, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 35199 + }, + { + "epoch": 0.352, + "grad_norm": 1.390940621754904, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 35200 + }, + { + "epoch": 0.35201, + "grad_norm": 1.8401852219513677, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 35201 + }, + { + "epoch": 0.35202, + "grad_norm": 1.6299734347587433, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 35202 + }, + { + "epoch": 0.35203, + "grad_norm": 1.3808775142199647, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 35203 + }, + { + "epoch": 0.35204, + "grad_norm": 1.7589430603189276, + "learning_rate": 0.003, + "loss": 4.091, + "step": 35204 + }, + { + "epoch": 0.35205, + "grad_norm": 1.3379308405162154, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 35205 + }, + { + "epoch": 0.35206, + "grad_norm": 1.6663674420298211, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 35206 + }, + { + "epoch": 0.35207, + "grad_norm": 1.5589179062610607, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 35207 + }, + { + "epoch": 0.35208, + "grad_norm": 1.5311099797374015, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 35208 + }, + { + "epoch": 0.35209, + "grad_norm": 1.2273129852122238, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35209 + }, + { + "epoch": 0.3521, + "grad_norm": 1.6958740637891572, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 35210 + }, + { + "epoch": 0.35211, + "grad_norm": 1.3223830411079367, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 35211 + }, + { + "epoch": 0.35212, + "grad_norm": 1.6432480892839614, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35212 + }, + { + "epoch": 0.35213, + "grad_norm": 1.350122165410413, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 35213 + }, + { + "epoch": 0.35214, + "grad_norm": 1.8978195098242971, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35214 + }, + { + "epoch": 0.35215, + "grad_norm": 1.213766041562782, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 35215 + }, + { + "epoch": 0.35216, + "grad_norm": 1.5417136366469177, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35216 + }, + { + "epoch": 0.35217, + "grad_norm": 1.9634803002338117, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35217 + }, + { + "epoch": 0.35218, + "grad_norm": 1.3043837204084419, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 35218 + }, + { + "epoch": 0.35219, + "grad_norm": 1.687126646870818, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 35219 + }, + { + "epoch": 0.3522, + "grad_norm": 1.2942592677096534, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 35220 + }, + { + "epoch": 0.35221, + "grad_norm": 1.6413762005164871, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35221 + }, + { + "epoch": 0.35222, + "grad_norm": 1.5210231451156682, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35222 + }, + { + "epoch": 0.35223, + "grad_norm": 1.1996097781051722, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35223 + }, + { + "epoch": 0.35224, + "grad_norm": 1.800629233625781, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35224 + }, + { + "epoch": 0.35225, + "grad_norm": 1.1018099534447148, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35225 + }, + { + "epoch": 0.35226, + "grad_norm": 1.718407649269904, + "learning_rate": 0.003, + "loss": 4.052, + "step": 35226 + }, + { + "epoch": 0.35227, + "grad_norm": 1.689608746670893, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35227 + }, + { + "epoch": 0.35228, + "grad_norm": 1.7428814523193188, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 35228 + }, + { + "epoch": 0.35229, + "grad_norm": 1.5155921085222062, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 35229 + }, + { + "epoch": 0.3523, + "grad_norm": 1.4447721210044262, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 35230 + }, + { + "epoch": 0.35231, + "grad_norm": 1.2660637340730991, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 35231 + }, + { + "epoch": 0.35232, + "grad_norm": 1.6569896943413789, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 35232 + }, + { + "epoch": 0.35233, + "grad_norm": 1.4516037792202823, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 35233 + }, + { + "epoch": 0.35234, + "grad_norm": 1.3965970670959258, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 35234 + }, + { + "epoch": 0.35235, + "grad_norm": 1.5081146036665347, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35235 + }, + { + "epoch": 0.35236, + "grad_norm": 1.247585075047215, + "learning_rate": 0.003, + "loss": 4.022, + "step": 35236 + }, + { + "epoch": 0.35237, + "grad_norm": 1.556044098581186, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 35237 + }, + { + "epoch": 0.35238, + "grad_norm": 1.277705355031927, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 35238 + }, + { + "epoch": 0.35239, + "grad_norm": 1.586802386658393, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35239 + }, + { + "epoch": 0.3524, + "grad_norm": 1.532535479739389, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35240 + }, + { + "epoch": 0.35241, + "grad_norm": 1.76398019379505, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 35241 + }, + { + "epoch": 0.35242, + "grad_norm": 1.274124936957698, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35242 + }, + { + "epoch": 0.35243, + "grad_norm": 1.5083172633721778, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 35243 + }, + { + "epoch": 0.35244, + "grad_norm": 1.599615970378757, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35244 + }, + { + "epoch": 0.35245, + "grad_norm": 1.4373693412926025, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 35245 + }, + { + "epoch": 0.35246, + "grad_norm": 1.483884688052219, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 35246 + }, + { + "epoch": 0.35247, + "grad_norm": 1.6743742068958232, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 35247 + }, + { + "epoch": 0.35248, + "grad_norm": 1.6144001179970875, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 35248 + }, + { + "epoch": 0.35249, + "grad_norm": 1.5029895779518947, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35249 + }, + { + "epoch": 0.3525, + "grad_norm": 1.5120760091345868, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 35250 + }, + { + "epoch": 0.35251, + "grad_norm": 1.505554149510165, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 35251 + }, + { + "epoch": 0.35252, + "grad_norm": 1.3031465272298324, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 35252 + }, + { + "epoch": 0.35253, + "grad_norm": 1.7797198041746274, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 35253 + }, + { + "epoch": 0.35254, + "grad_norm": 1.3994990481258032, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 35254 + }, + { + "epoch": 0.35255, + "grad_norm": 1.6439860037702476, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 35255 + }, + { + "epoch": 0.35256, + "grad_norm": 1.1996816249183369, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 35256 + }, + { + "epoch": 0.35257, + "grad_norm": 1.6498490564563937, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35257 + }, + { + "epoch": 0.35258, + "grad_norm": 1.4622070935022078, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 35258 + }, + { + "epoch": 0.35259, + "grad_norm": 1.448698774978711, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35259 + }, + { + "epoch": 0.3526, + "grad_norm": 1.642593329725574, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 35260 + }, + { + "epoch": 0.35261, + "grad_norm": 1.4658709569868307, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35261 + }, + { + "epoch": 0.35262, + "grad_norm": 1.2570474783547356, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 35262 + }, + { + "epoch": 0.35263, + "grad_norm": 1.5104524855312769, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 35263 + }, + { + "epoch": 0.35264, + "grad_norm": 1.7355524672094218, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35264 + }, + { + "epoch": 0.35265, + "grad_norm": 1.3616899132954603, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35265 + }, + { + "epoch": 0.35266, + "grad_norm": 1.3903204722124842, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 35266 + }, + { + "epoch": 0.35267, + "grad_norm": 1.6842879119747505, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35267 + }, + { + "epoch": 0.35268, + "grad_norm": 1.282192031358656, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35268 + }, + { + "epoch": 0.35269, + "grad_norm": 1.4013044363840441, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 35269 + }, + { + "epoch": 0.3527, + "grad_norm": 1.500390687136034, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 35270 + }, + { + "epoch": 0.35271, + "grad_norm": 1.5476672855290725, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 35271 + }, + { + "epoch": 0.35272, + "grad_norm": 1.632856199020147, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 35272 + }, + { + "epoch": 0.35273, + "grad_norm": 1.4048595719700085, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 35273 + }, + { + "epoch": 0.35274, + "grad_norm": 1.4447718977037687, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35274 + }, + { + "epoch": 0.35275, + "grad_norm": 1.6286692902342654, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35275 + }, + { + "epoch": 0.35276, + "grad_norm": 1.3022523913010131, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 35276 + }, + { + "epoch": 0.35277, + "grad_norm": 1.6205256977603977, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 35277 + }, + { + "epoch": 0.35278, + "grad_norm": 1.5534864372841009, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 35278 + }, + { + "epoch": 0.35279, + "grad_norm": 1.37637166648596, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35279 + }, + { + "epoch": 0.3528, + "grad_norm": 1.4394722365271426, + "learning_rate": 0.003, + "loss": 4.04, + "step": 35280 + }, + { + "epoch": 0.35281, + "grad_norm": 1.410807875522901, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35281 + }, + { + "epoch": 0.35282, + "grad_norm": 1.4831626854280038, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 35282 + }, + { + "epoch": 0.35283, + "grad_norm": 1.5492308402913901, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 35283 + }, + { + "epoch": 0.35284, + "grad_norm": 1.3002390293380264, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 35284 + }, + { + "epoch": 0.35285, + "grad_norm": 1.5140578939985276, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35285 + }, + { + "epoch": 0.35286, + "grad_norm": 1.45440828161312, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35286 + }, + { + "epoch": 0.35287, + "grad_norm": 1.8349727938971216, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 35287 + }, + { + "epoch": 0.35288, + "grad_norm": 1.557480495936823, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 35288 + }, + { + "epoch": 0.35289, + "grad_norm": 2.0480696522627513, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 35289 + }, + { + "epoch": 0.3529, + "grad_norm": 1.1740493670991219, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35290 + }, + { + "epoch": 0.35291, + "grad_norm": 1.4426617006270557, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35291 + }, + { + "epoch": 0.35292, + "grad_norm": 1.3098459708994814, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 35292 + }, + { + "epoch": 0.35293, + "grad_norm": 1.4942763582014367, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 35293 + }, + { + "epoch": 0.35294, + "grad_norm": 1.437625875352967, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 35294 + }, + { + "epoch": 0.35295, + "grad_norm": 1.4078903905011928, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 35295 + }, + { + "epoch": 0.35296, + "grad_norm": 1.3545242848957322, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 35296 + }, + { + "epoch": 0.35297, + "grad_norm": 1.4955098705611087, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35297 + }, + { + "epoch": 0.35298, + "grad_norm": 1.6750406459733966, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35298 + }, + { + "epoch": 0.35299, + "grad_norm": 1.6070596149048513, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35299 + }, + { + "epoch": 0.353, + "grad_norm": 1.6872208868266123, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35300 + }, + { + "epoch": 0.35301, + "grad_norm": 1.4122474840236503, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35301 + }, + { + "epoch": 0.35302, + "grad_norm": 1.4329790513039333, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35302 + }, + { + "epoch": 0.35303, + "grad_norm": 1.6081428451773128, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 35303 + }, + { + "epoch": 0.35304, + "grad_norm": 1.3286518497242548, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 35304 + }, + { + "epoch": 0.35305, + "grad_norm": 1.5968933724611398, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 35305 + }, + { + "epoch": 0.35306, + "grad_norm": 1.524992658950734, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35306 + }, + { + "epoch": 0.35307, + "grad_norm": 1.5150330041199627, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 35307 + }, + { + "epoch": 0.35308, + "grad_norm": 1.2888611666884457, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 35308 + }, + { + "epoch": 0.35309, + "grad_norm": 1.6173267558165076, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 35309 + }, + { + "epoch": 0.3531, + "grad_norm": 1.4953361818052548, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 35310 + }, + { + "epoch": 0.35311, + "grad_norm": 1.504231392163388, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 35311 + }, + { + "epoch": 0.35312, + "grad_norm": 1.5706296245911298, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 35312 + }, + { + "epoch": 0.35313, + "grad_norm": 1.3056378953473986, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 35313 + }, + { + "epoch": 0.35314, + "grad_norm": 2.2514238682262677, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35314 + }, + { + "epoch": 0.35315, + "grad_norm": 1.1569506466934545, + "learning_rate": 0.003, + "loss": 4.02, + "step": 35315 + }, + { + "epoch": 0.35316, + "grad_norm": 1.4746772099441376, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 35316 + }, + { + "epoch": 0.35317, + "grad_norm": 1.4966323254088114, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 35317 + }, + { + "epoch": 0.35318, + "grad_norm": 2.0753069031847797, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35318 + }, + { + "epoch": 0.35319, + "grad_norm": 1.0193091847004823, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35319 + }, + { + "epoch": 0.3532, + "grad_norm": 1.7162093517521917, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 35320 + }, + { + "epoch": 0.35321, + "grad_norm": 1.3656987792637683, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 35321 + }, + { + "epoch": 0.35322, + "grad_norm": 1.7595724632025602, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 35322 + }, + { + "epoch": 0.35323, + "grad_norm": 1.4349687245590934, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 35323 + }, + { + "epoch": 0.35324, + "grad_norm": 1.605374675645456, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 35324 + }, + { + "epoch": 0.35325, + "grad_norm": 1.785156636052491, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 35325 + }, + { + "epoch": 0.35326, + "grad_norm": 1.4580205922858271, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 35326 + }, + { + "epoch": 0.35327, + "grad_norm": 1.5821998949242524, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 35327 + }, + { + "epoch": 0.35328, + "grad_norm": 1.7389704476397656, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 35328 + }, + { + "epoch": 0.35329, + "grad_norm": 1.5164066906478546, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 35329 + }, + { + "epoch": 0.3533, + "grad_norm": 1.496753060137909, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35330 + }, + { + "epoch": 0.35331, + "grad_norm": 1.4669401054077993, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 35331 + }, + { + "epoch": 0.35332, + "grad_norm": 1.3222501224289929, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 35332 + }, + { + "epoch": 0.35333, + "grad_norm": 1.4646123267505426, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 35333 + }, + { + "epoch": 0.35334, + "grad_norm": 1.3960756531583394, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 35334 + }, + { + "epoch": 0.35335, + "grad_norm": 1.7387471335452063, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 35335 + }, + { + "epoch": 0.35336, + "grad_norm": 1.3572674060980445, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 35336 + }, + { + "epoch": 0.35337, + "grad_norm": 1.7296359154799892, + "learning_rate": 0.003, + "loss": 4.019, + "step": 35337 + }, + { + "epoch": 0.35338, + "grad_norm": 1.99315997696877, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 35338 + }, + { + "epoch": 0.35339, + "grad_norm": 1.19188738439977, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 35339 + }, + { + "epoch": 0.3534, + "grad_norm": 1.6347498981088526, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 35340 + }, + { + "epoch": 0.35341, + "grad_norm": 1.5530938256528157, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 35341 + }, + { + "epoch": 0.35342, + "grad_norm": 1.7406141363303174, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 35342 + }, + { + "epoch": 0.35343, + "grad_norm": 1.355994424378076, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35343 + }, + { + "epoch": 0.35344, + "grad_norm": 1.4281965628014524, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 35344 + }, + { + "epoch": 0.35345, + "grad_norm": 1.464908630575462, + "learning_rate": 0.003, + "loss": 4.072, + "step": 35345 + }, + { + "epoch": 0.35346, + "grad_norm": 1.3195867029010597, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 35346 + }, + { + "epoch": 0.35347, + "grad_norm": 1.5385313200188764, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 35347 + }, + { + "epoch": 0.35348, + "grad_norm": 1.4657602240675447, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35348 + }, + { + "epoch": 0.35349, + "grad_norm": 1.512730368578992, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 35349 + }, + { + "epoch": 0.3535, + "grad_norm": 1.6656835905167096, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35350 + }, + { + "epoch": 0.35351, + "grad_norm": 1.3734823698412664, + "learning_rate": 0.003, + "loss": 4.016, + "step": 35351 + }, + { + "epoch": 0.35352, + "grad_norm": 1.7259376426753228, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35352 + }, + { + "epoch": 0.35353, + "grad_norm": 1.5430950606941003, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 35353 + }, + { + "epoch": 0.35354, + "grad_norm": 1.346655275217158, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35354 + }, + { + "epoch": 0.35355, + "grad_norm": 1.599250289722964, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35355 + }, + { + "epoch": 0.35356, + "grad_norm": 1.237825029373226, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35356 + }, + { + "epoch": 0.35357, + "grad_norm": 1.874536937148921, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 35357 + }, + { + "epoch": 0.35358, + "grad_norm": 1.3365214033273134, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 35358 + }, + { + "epoch": 0.35359, + "grad_norm": 1.6098879559574395, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35359 + }, + { + "epoch": 0.3536, + "grad_norm": 1.482847470014683, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35360 + }, + { + "epoch": 0.35361, + "grad_norm": 1.5813500122682507, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 35361 + }, + { + "epoch": 0.35362, + "grad_norm": 1.583843227309563, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35362 + }, + { + "epoch": 0.35363, + "grad_norm": 1.2241877274587079, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 35363 + }, + { + "epoch": 0.35364, + "grad_norm": 1.6333597045763273, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 35364 + }, + { + "epoch": 0.35365, + "grad_norm": 1.4777276674633013, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35365 + }, + { + "epoch": 0.35366, + "grad_norm": 1.2532067282804609, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 35366 + }, + { + "epoch": 0.35367, + "grad_norm": 1.613918346305319, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 35367 + }, + { + "epoch": 0.35368, + "grad_norm": 1.231868411202342, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 35368 + }, + { + "epoch": 0.35369, + "grad_norm": 1.9212967577550246, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 35369 + }, + { + "epoch": 0.3537, + "grad_norm": 1.1122768470227702, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 35370 + }, + { + "epoch": 0.35371, + "grad_norm": 1.8366655374011902, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 35371 + }, + { + "epoch": 0.35372, + "grad_norm": 1.4960938385228089, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 35372 + }, + { + "epoch": 0.35373, + "grad_norm": 2.174654867640534, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 35373 + }, + { + "epoch": 0.35374, + "grad_norm": 1.32668269362226, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 35374 + }, + { + "epoch": 0.35375, + "grad_norm": 1.302943547442381, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 35375 + }, + { + "epoch": 0.35376, + "grad_norm": 1.4053562192096098, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35376 + }, + { + "epoch": 0.35377, + "grad_norm": 1.4655708378208758, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35377 + }, + { + "epoch": 0.35378, + "grad_norm": 1.5367166669783268, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 35378 + }, + { + "epoch": 0.35379, + "grad_norm": 1.9085598094203882, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35379 + }, + { + "epoch": 0.3538, + "grad_norm": 1.3482088273611403, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 35380 + }, + { + "epoch": 0.35381, + "grad_norm": 1.4712566570711159, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35381 + }, + { + "epoch": 0.35382, + "grad_norm": 1.5084168001425322, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 35382 + }, + { + "epoch": 0.35383, + "grad_norm": 1.2865291427264223, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 35383 + }, + { + "epoch": 0.35384, + "grad_norm": 1.7935873548221093, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 35384 + }, + { + "epoch": 0.35385, + "grad_norm": 1.3159664860172753, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 35385 + }, + { + "epoch": 0.35386, + "grad_norm": 1.8404597217099856, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35386 + }, + { + "epoch": 0.35387, + "grad_norm": 1.358654878566607, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35387 + }, + { + "epoch": 0.35388, + "grad_norm": 1.751942817671539, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 35388 + }, + { + "epoch": 0.35389, + "grad_norm": 1.2068605581523495, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 35389 + }, + { + "epoch": 0.3539, + "grad_norm": 1.7319654744880917, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 35390 + }, + { + "epoch": 0.35391, + "grad_norm": 1.489470481712508, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35391 + }, + { + "epoch": 0.35392, + "grad_norm": 1.5987355684301718, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 35392 + }, + { + "epoch": 0.35393, + "grad_norm": 1.4363423981030665, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 35393 + }, + { + "epoch": 0.35394, + "grad_norm": 1.2650741982518023, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 35394 + }, + { + "epoch": 0.35395, + "grad_norm": 1.6623526031805689, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 35395 + }, + { + "epoch": 0.35396, + "grad_norm": 1.9519266041985137, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 35396 + }, + { + "epoch": 0.35397, + "grad_norm": 1.3004101709433822, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 35397 + }, + { + "epoch": 0.35398, + "grad_norm": 1.422858073543274, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 35398 + }, + { + "epoch": 0.35399, + "grad_norm": 1.5547992752004112, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 35399 + }, + { + "epoch": 0.354, + "grad_norm": 1.170500461899887, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35400 + }, + { + "epoch": 0.35401, + "grad_norm": 1.831479983683586, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 35401 + }, + { + "epoch": 0.35402, + "grad_norm": 1.2284624525705945, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35402 + }, + { + "epoch": 0.35403, + "grad_norm": 1.604201165725805, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 35403 + }, + { + "epoch": 0.35404, + "grad_norm": 1.2632157559374344, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35404 + }, + { + "epoch": 0.35405, + "grad_norm": 1.5930635658157013, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35405 + }, + { + "epoch": 0.35406, + "grad_norm": 1.3879974762451923, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 35406 + }, + { + "epoch": 0.35407, + "grad_norm": 1.648443576898452, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 35407 + }, + { + "epoch": 0.35408, + "grad_norm": 1.2949804762624308, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35408 + }, + { + "epoch": 0.35409, + "grad_norm": 2.09287211990343, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 35409 + }, + { + "epoch": 0.3541, + "grad_norm": 1.7721730773662707, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35410 + }, + { + "epoch": 0.35411, + "grad_norm": 1.395525321482202, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 35411 + }, + { + "epoch": 0.35412, + "grad_norm": 1.6025935090735028, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 35412 + }, + { + "epoch": 0.35413, + "grad_norm": 1.6082913238872723, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35413 + }, + { + "epoch": 0.35414, + "grad_norm": 1.360132410595095, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 35414 + }, + { + "epoch": 0.35415, + "grad_norm": 1.5958852553474043, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35415 + }, + { + "epoch": 0.35416, + "grad_norm": 1.2678234541140354, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 35416 + }, + { + "epoch": 0.35417, + "grad_norm": 1.6021192958481794, + "learning_rate": 0.003, + "loss": 4.037, + "step": 35417 + }, + { + "epoch": 0.35418, + "grad_norm": 1.179049141953349, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 35418 + }, + { + "epoch": 0.35419, + "grad_norm": 1.6528847603119612, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 35419 + }, + { + "epoch": 0.3542, + "grad_norm": 1.3959643386666072, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 35420 + }, + { + "epoch": 0.35421, + "grad_norm": 1.7962908536015616, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35421 + }, + { + "epoch": 0.35422, + "grad_norm": 1.45502671059648, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35422 + }, + { + "epoch": 0.35423, + "grad_norm": 1.2764118101128732, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 35423 + }, + { + "epoch": 0.35424, + "grad_norm": 1.4950948190185647, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 35424 + }, + { + "epoch": 0.35425, + "grad_norm": 1.3863477256523737, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 35425 + }, + { + "epoch": 0.35426, + "grad_norm": 1.5339879553276585, + "learning_rate": 0.003, + "loss": 4.073, + "step": 35426 + }, + { + "epoch": 0.35427, + "grad_norm": 1.3502682301883968, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35427 + }, + { + "epoch": 0.35428, + "grad_norm": 1.654770713152518, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 35428 + }, + { + "epoch": 0.35429, + "grad_norm": 1.2746895445982287, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 35429 + }, + { + "epoch": 0.3543, + "grad_norm": 1.6650239056298535, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35430 + }, + { + "epoch": 0.35431, + "grad_norm": 1.287604935871147, + "learning_rate": 0.003, + "loss": 4.06, + "step": 35431 + }, + { + "epoch": 0.35432, + "grad_norm": 1.4801597580646788, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35432 + }, + { + "epoch": 0.35433, + "grad_norm": 1.2396170252335337, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 35433 + }, + { + "epoch": 0.35434, + "grad_norm": 1.9389705807662554, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35434 + }, + { + "epoch": 0.35435, + "grad_norm": 1.2104736088955281, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 35435 + }, + { + "epoch": 0.35436, + "grad_norm": 1.8879211036838122, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 35436 + }, + { + "epoch": 0.35437, + "grad_norm": 1.478087500755164, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35437 + }, + { + "epoch": 0.35438, + "grad_norm": 1.741257869953437, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 35438 + }, + { + "epoch": 0.35439, + "grad_norm": 1.4680196196915818, + "learning_rate": 0.003, + "loss": 4.03, + "step": 35439 + }, + { + "epoch": 0.3544, + "grad_norm": 1.508710963490542, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35440 + }, + { + "epoch": 0.35441, + "grad_norm": 1.4382776485353292, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35441 + }, + { + "epoch": 0.35442, + "grad_norm": 1.5884040888078195, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 35442 + }, + { + "epoch": 0.35443, + "grad_norm": 1.335736129489664, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35443 + }, + { + "epoch": 0.35444, + "grad_norm": 1.6676373772607362, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 35444 + }, + { + "epoch": 0.35445, + "grad_norm": 1.5579648705274298, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 35445 + }, + { + "epoch": 0.35446, + "grad_norm": 1.4475827316065002, + "learning_rate": 0.003, + "loss": 4.01, + "step": 35446 + }, + { + "epoch": 0.35447, + "grad_norm": 1.10632077583456, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 35447 + }, + { + "epoch": 0.35448, + "grad_norm": 1.6596600299688389, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 35448 + }, + { + "epoch": 0.35449, + "grad_norm": 1.4255348861301123, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 35449 + }, + { + "epoch": 0.3545, + "grad_norm": 1.506866091892107, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35450 + }, + { + "epoch": 0.35451, + "grad_norm": 1.5828949237025607, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 35451 + }, + { + "epoch": 0.35452, + "grad_norm": 1.4804529599304428, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 35452 + }, + { + "epoch": 0.35453, + "grad_norm": 1.2979174993805525, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 35453 + }, + { + "epoch": 0.35454, + "grad_norm": 1.5250095252105271, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 35454 + }, + { + "epoch": 0.35455, + "grad_norm": 1.456870004071397, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35455 + }, + { + "epoch": 0.35456, + "grad_norm": 1.9537853400664265, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 35456 + }, + { + "epoch": 0.35457, + "grad_norm": 1.1722467431406314, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35457 + }, + { + "epoch": 0.35458, + "grad_norm": 1.7389143465575883, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 35458 + }, + { + "epoch": 0.35459, + "grad_norm": 1.329352205072214, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35459 + }, + { + "epoch": 0.3546, + "grad_norm": 1.5478514023292569, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35460 + }, + { + "epoch": 0.35461, + "grad_norm": 1.394060170269757, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 35461 + }, + { + "epoch": 0.35462, + "grad_norm": 1.2512976710796488, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35462 + }, + { + "epoch": 0.35463, + "grad_norm": 1.4446442570166889, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 35463 + }, + { + "epoch": 0.35464, + "grad_norm": 1.4309784552494575, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 35464 + }, + { + "epoch": 0.35465, + "grad_norm": 1.9282920908247192, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 35465 + }, + { + "epoch": 0.35466, + "grad_norm": 1.3310105470734581, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 35466 + }, + { + "epoch": 0.35467, + "grad_norm": 1.5613855106587928, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35467 + }, + { + "epoch": 0.35468, + "grad_norm": 1.6567138961484797, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 35468 + }, + { + "epoch": 0.35469, + "grad_norm": 1.3123915616525612, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 35469 + }, + { + "epoch": 0.3547, + "grad_norm": 1.6427304906549736, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 35470 + }, + { + "epoch": 0.35471, + "grad_norm": 1.3289665813066809, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 35471 + }, + { + "epoch": 0.35472, + "grad_norm": 1.3372479978621794, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 35472 + }, + { + "epoch": 0.35473, + "grad_norm": 1.2911592173573423, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 35473 + }, + { + "epoch": 0.35474, + "grad_norm": 1.7497788757218333, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35474 + }, + { + "epoch": 0.35475, + "grad_norm": 1.449212584806385, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 35475 + }, + { + "epoch": 0.35476, + "grad_norm": 1.7600593670101614, + "learning_rate": 0.003, + "loss": 4.054, + "step": 35476 + }, + { + "epoch": 0.35477, + "grad_norm": 1.435099195982482, + "learning_rate": 0.003, + "loss": 4.032, + "step": 35477 + }, + { + "epoch": 0.35478, + "grad_norm": 1.944264927355946, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 35478 + }, + { + "epoch": 0.35479, + "grad_norm": 1.2410104163291438, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 35479 + }, + { + "epoch": 0.3548, + "grad_norm": 1.6135935573205125, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 35480 + }, + { + "epoch": 0.35481, + "grad_norm": 1.5553757356990432, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35481 + }, + { + "epoch": 0.35482, + "grad_norm": 1.4553159270475091, + "learning_rate": 0.003, + "loss": 4.068, + "step": 35482 + }, + { + "epoch": 0.35483, + "grad_norm": 1.5089163531294212, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35483 + }, + { + "epoch": 0.35484, + "grad_norm": 1.3508059672336727, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 35484 + }, + { + "epoch": 0.35485, + "grad_norm": 1.439011152376032, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 35485 + }, + { + "epoch": 0.35486, + "grad_norm": 1.7866223155693126, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 35486 + }, + { + "epoch": 0.35487, + "grad_norm": 1.520776596292522, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35487 + }, + { + "epoch": 0.35488, + "grad_norm": 1.504906289292244, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 35488 + }, + { + "epoch": 0.35489, + "grad_norm": 1.503629348889581, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 35489 + }, + { + "epoch": 0.3549, + "grad_norm": 1.6528224826160491, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 35490 + }, + { + "epoch": 0.35491, + "grad_norm": 1.2738974265417975, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 35491 + }, + { + "epoch": 0.35492, + "grad_norm": 1.524609002444839, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35492 + }, + { + "epoch": 0.35493, + "grad_norm": 1.6025014185248494, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35493 + }, + { + "epoch": 0.35494, + "grad_norm": 1.105968297907291, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 35494 + }, + { + "epoch": 0.35495, + "grad_norm": 1.8757671724557572, + "learning_rate": 0.003, + "loss": 4.065, + "step": 35495 + }, + { + "epoch": 0.35496, + "grad_norm": 1.2696597938163316, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35496 + }, + { + "epoch": 0.35497, + "grad_norm": 1.8651831401781476, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35497 + }, + { + "epoch": 0.35498, + "grad_norm": 1.3749745064407324, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 35498 + }, + { + "epoch": 0.35499, + "grad_norm": 1.240035884939716, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35499 + }, + { + "epoch": 0.355, + "grad_norm": 1.728587940430368, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 35500 + }, + { + "epoch": 0.35501, + "grad_norm": 1.393449037561414, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 35501 + }, + { + "epoch": 0.35502, + "grad_norm": 1.433946106549142, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 35502 + }, + { + "epoch": 0.35503, + "grad_norm": 1.5725854482730535, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 35503 + }, + { + "epoch": 0.35504, + "grad_norm": 1.589753142075018, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35504 + }, + { + "epoch": 0.35505, + "grad_norm": 1.3787110631737218, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 35505 + }, + { + "epoch": 0.35506, + "grad_norm": 1.4119909984053094, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35506 + }, + { + "epoch": 0.35507, + "grad_norm": 1.5447469127263282, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 35507 + }, + { + "epoch": 0.35508, + "grad_norm": 1.3723917411694029, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35508 + }, + { + "epoch": 0.35509, + "grad_norm": 1.4578394612491385, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 35509 + }, + { + "epoch": 0.3551, + "grad_norm": 1.5476067333347134, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 35510 + }, + { + "epoch": 0.35511, + "grad_norm": 1.340607368001277, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 35511 + }, + { + "epoch": 0.35512, + "grad_norm": 1.369781045872829, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35512 + }, + { + "epoch": 0.35513, + "grad_norm": 1.7639080061902916, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 35513 + }, + { + "epoch": 0.35514, + "grad_norm": 1.4292791158390996, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 35514 + }, + { + "epoch": 0.35515, + "grad_norm": 1.7253647109737722, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 35515 + }, + { + "epoch": 0.35516, + "grad_norm": 1.240361233578258, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 35516 + }, + { + "epoch": 0.35517, + "grad_norm": 1.5118154568587643, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 35517 + }, + { + "epoch": 0.35518, + "grad_norm": 1.4731304967084735, + "learning_rate": 0.003, + "loss": 4.05, + "step": 35518 + }, + { + "epoch": 0.35519, + "grad_norm": 1.591021695380248, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 35519 + }, + { + "epoch": 0.3552, + "grad_norm": 1.4918141916857504, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 35520 + }, + { + "epoch": 0.35521, + "grad_norm": 1.474465871199967, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35521 + }, + { + "epoch": 0.35522, + "grad_norm": 1.4258296944235838, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 35522 + }, + { + "epoch": 0.35523, + "grad_norm": 1.4440507150094386, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35523 + }, + { + "epoch": 0.35524, + "grad_norm": 1.3430240616292417, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35524 + }, + { + "epoch": 0.35525, + "grad_norm": 1.4921199921899033, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 35525 + }, + { + "epoch": 0.35526, + "grad_norm": 1.3934893601024427, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 35526 + }, + { + "epoch": 0.35527, + "grad_norm": 1.6349010866471767, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35527 + }, + { + "epoch": 0.35528, + "grad_norm": 1.6044543972494223, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 35528 + }, + { + "epoch": 0.35529, + "grad_norm": 1.471535359716184, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 35529 + }, + { + "epoch": 0.3553, + "grad_norm": 1.4955764954952941, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 35530 + }, + { + "epoch": 0.35531, + "grad_norm": 1.524249001137581, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 35531 + }, + { + "epoch": 0.35532, + "grad_norm": 1.5676840044267524, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 35532 + }, + { + "epoch": 0.35533, + "grad_norm": 1.5387379238194645, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 35533 + }, + { + "epoch": 0.35534, + "grad_norm": 1.6608721588515816, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 35534 + }, + { + "epoch": 0.35535, + "grad_norm": 1.4405641426161861, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35535 + }, + { + "epoch": 0.35536, + "grad_norm": 1.4504668907859741, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 35536 + }, + { + "epoch": 0.35537, + "grad_norm": 1.4606200440924548, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 35537 + }, + { + "epoch": 0.35538, + "grad_norm": 1.4061736661667086, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35538 + }, + { + "epoch": 0.35539, + "grad_norm": 1.5296598249986515, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35539 + }, + { + "epoch": 0.3554, + "grad_norm": 1.3507247863167575, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 35540 + }, + { + "epoch": 0.35541, + "grad_norm": 1.2861007146441452, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 35541 + }, + { + "epoch": 0.35542, + "grad_norm": 1.884613056645582, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 35542 + }, + { + "epoch": 0.35543, + "grad_norm": 1.4694305630967284, + "learning_rate": 0.003, + "loss": 4.063, + "step": 35543 + }, + { + "epoch": 0.35544, + "grad_norm": 1.8424681411909467, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 35544 + }, + { + "epoch": 0.35545, + "grad_norm": 1.3707245290264425, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 35545 + }, + { + "epoch": 0.35546, + "grad_norm": 1.7883685434458358, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35546 + }, + { + "epoch": 0.35547, + "grad_norm": 1.298615875613292, + "learning_rate": 0.003, + "loss": 4.036, + "step": 35547 + }, + { + "epoch": 0.35548, + "grad_norm": 1.5678029057933482, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 35548 + }, + { + "epoch": 0.35549, + "grad_norm": 1.562059408603056, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 35549 + }, + { + "epoch": 0.3555, + "grad_norm": 1.4295712958323517, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 35550 + }, + { + "epoch": 0.35551, + "grad_norm": 1.5790627889501128, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 35551 + }, + { + "epoch": 0.35552, + "grad_norm": 1.359533513704048, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 35552 + }, + { + "epoch": 0.35553, + "grad_norm": 1.4914321667889385, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 35553 + }, + { + "epoch": 0.35554, + "grad_norm": 1.6905565413196861, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 35554 + }, + { + "epoch": 0.35555, + "grad_norm": 1.4279868775798277, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 35555 + }, + { + "epoch": 0.35556, + "grad_norm": 1.4316996725143776, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 35556 + }, + { + "epoch": 0.35557, + "grad_norm": 1.5241267732678583, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 35557 + }, + { + "epoch": 0.35558, + "grad_norm": 1.6029691678487459, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 35558 + }, + { + "epoch": 0.35559, + "grad_norm": 1.3218708820301948, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35559 + }, + { + "epoch": 0.3556, + "grad_norm": 1.7462807318224027, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 35560 + }, + { + "epoch": 0.35561, + "grad_norm": 1.3933586991327511, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 35561 + }, + { + "epoch": 0.35562, + "grad_norm": 1.522400687011281, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 35562 + }, + { + "epoch": 0.35563, + "grad_norm": 1.3448290296904006, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35563 + }, + { + "epoch": 0.35564, + "grad_norm": 1.6974578366159625, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 35564 + }, + { + "epoch": 0.35565, + "grad_norm": 1.5716811394128545, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 35565 + }, + { + "epoch": 0.35566, + "grad_norm": 1.6688319133570564, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 35566 + }, + { + "epoch": 0.35567, + "grad_norm": 1.644429827014162, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 35567 + }, + { + "epoch": 0.35568, + "grad_norm": 1.3366145946734862, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 35568 + }, + { + "epoch": 0.35569, + "grad_norm": 1.4514942138433116, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35569 + }, + { + "epoch": 0.3557, + "grad_norm": 1.7765066015725561, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 35570 + }, + { + "epoch": 0.35571, + "grad_norm": 1.2702208376505975, + "learning_rate": 0.003, + "loss": 4.076, + "step": 35571 + }, + { + "epoch": 0.35572, + "grad_norm": 1.5839778475274335, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 35572 + }, + { + "epoch": 0.35573, + "grad_norm": 1.55237767631114, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 35573 + }, + { + "epoch": 0.35574, + "grad_norm": 1.6902392337468735, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35574 + }, + { + "epoch": 0.35575, + "grad_norm": 1.1370691704921323, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 35575 + }, + { + "epoch": 0.35576, + "grad_norm": 1.6619111536314881, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 35576 + }, + { + "epoch": 0.35577, + "grad_norm": 1.3397175028050736, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 35577 + }, + { + "epoch": 0.35578, + "grad_norm": 1.9704339253888068, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 35578 + }, + { + "epoch": 0.35579, + "grad_norm": 1.5579288449600321, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 35579 + }, + { + "epoch": 0.3558, + "grad_norm": 1.4126830114551967, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 35580 + }, + { + "epoch": 0.35581, + "grad_norm": 1.7568428127828113, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 35581 + }, + { + "epoch": 0.35582, + "grad_norm": 1.3658571536922657, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 35582 + }, + { + "epoch": 0.35583, + "grad_norm": 1.5882434599317543, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 35583 + }, + { + "epoch": 0.35584, + "grad_norm": 1.2435104457461172, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 35584 + }, + { + "epoch": 0.35585, + "grad_norm": 1.5223124267853263, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 35585 + }, + { + "epoch": 0.35586, + "grad_norm": 1.352119662924248, + "learning_rate": 0.003, + "loss": 3.999, + "step": 35586 + }, + { + "epoch": 0.35587, + "grad_norm": 1.9809180855248183, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 35587 + }, + { + "epoch": 0.35588, + "grad_norm": 1.3481721817617804, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 35588 + }, + { + "epoch": 0.35589, + "grad_norm": 1.7362841581745991, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 35589 + }, + { + "epoch": 0.3559, + "grad_norm": 1.3344957304118636, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 35590 + }, + { + "epoch": 0.35591, + "grad_norm": 1.5245700748465818, + "learning_rate": 0.003, + "loss": 4.04, + "step": 35591 + }, + { + "epoch": 0.35592, + "grad_norm": 1.4492690784088396, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35592 + }, + { + "epoch": 0.35593, + "grad_norm": 1.4586058639447215, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 35593 + }, + { + "epoch": 0.35594, + "grad_norm": 1.3084803136603362, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 35594 + }, + { + "epoch": 0.35595, + "grad_norm": 1.4632194165977561, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35595 + }, + { + "epoch": 0.35596, + "grad_norm": 1.5257179023814116, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 35596 + }, + { + "epoch": 0.35597, + "grad_norm": 1.3798384653155764, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 35597 + }, + { + "epoch": 0.35598, + "grad_norm": 1.5755713164854244, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 35598 + }, + { + "epoch": 0.35599, + "grad_norm": 1.4794313022208072, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35599 + }, + { + "epoch": 0.356, + "grad_norm": 1.5741855221167642, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 35600 + }, + { + "epoch": 0.35601, + "grad_norm": 1.26270460805024, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 35601 + }, + { + "epoch": 0.35602, + "grad_norm": 1.6620046811016806, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35602 + }, + { + "epoch": 0.35603, + "grad_norm": 1.3489291959395653, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 35603 + }, + { + "epoch": 0.35604, + "grad_norm": 1.5445831099419083, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 35604 + }, + { + "epoch": 0.35605, + "grad_norm": 1.2551182830270475, + "learning_rate": 0.003, + "loss": 4.027, + "step": 35605 + }, + { + "epoch": 0.35606, + "grad_norm": 1.3368110189132731, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 35606 + }, + { + "epoch": 0.35607, + "grad_norm": 1.9707781961829869, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 35607 + }, + { + "epoch": 0.35608, + "grad_norm": 1.1401121280771687, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 35608 + }, + { + "epoch": 0.35609, + "grad_norm": 1.7746739921203352, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 35609 + }, + { + "epoch": 0.3561, + "grad_norm": 1.4294666332708896, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35610 + }, + { + "epoch": 0.35611, + "grad_norm": 1.668742514172612, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 35611 + }, + { + "epoch": 0.35612, + "grad_norm": 1.5366187005209133, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35612 + }, + { + "epoch": 0.35613, + "grad_norm": 1.5714648567057492, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 35613 + }, + { + "epoch": 0.35614, + "grad_norm": 1.3896333826924676, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35614 + }, + { + "epoch": 0.35615, + "grad_norm": 1.3547574314284252, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 35615 + }, + { + "epoch": 0.35616, + "grad_norm": 1.605087137727629, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 35616 + }, + { + "epoch": 0.35617, + "grad_norm": 1.7113880893366045, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 35617 + }, + { + "epoch": 0.35618, + "grad_norm": 1.2980600759513652, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 35618 + }, + { + "epoch": 0.35619, + "grad_norm": 1.6634564097791837, + "learning_rate": 0.003, + "loss": 4.017, + "step": 35619 + }, + { + "epoch": 0.3562, + "grad_norm": 1.6171205096141177, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 35620 + }, + { + "epoch": 0.35621, + "grad_norm": 1.4476943722952271, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35621 + }, + { + "epoch": 0.35622, + "grad_norm": 1.3669493523457414, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 35622 + }, + { + "epoch": 0.35623, + "grad_norm": 1.627793119665197, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 35623 + }, + { + "epoch": 0.35624, + "grad_norm": 1.3512065958782777, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 35624 + }, + { + "epoch": 0.35625, + "grad_norm": 1.627030790277026, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35625 + }, + { + "epoch": 0.35626, + "grad_norm": 1.6026446515568922, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35626 + }, + { + "epoch": 0.35627, + "grad_norm": 1.2859045819908992, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 35627 + }, + { + "epoch": 0.35628, + "grad_norm": 1.641257997524683, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 35628 + }, + { + "epoch": 0.35629, + "grad_norm": 1.2648642419845437, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 35629 + }, + { + "epoch": 0.3563, + "grad_norm": 1.665515722823435, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 35630 + }, + { + "epoch": 0.35631, + "grad_norm": 1.3217891189121893, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 35631 + }, + { + "epoch": 0.35632, + "grad_norm": 1.7045089671041116, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 35632 + }, + { + "epoch": 0.35633, + "grad_norm": 1.2777082429971052, + "learning_rate": 0.003, + "loss": 4.009, + "step": 35633 + }, + { + "epoch": 0.35634, + "grad_norm": 1.8012199343944286, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35634 + }, + { + "epoch": 0.35635, + "grad_norm": 1.4115844049492892, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 35635 + }, + { + "epoch": 0.35636, + "grad_norm": 1.6180104357837568, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 35636 + }, + { + "epoch": 0.35637, + "grad_norm": 1.7222805743475353, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 35637 + }, + { + "epoch": 0.35638, + "grad_norm": 1.3480687809925274, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 35638 + }, + { + "epoch": 0.35639, + "grad_norm": 1.4069759731646905, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 35639 + }, + { + "epoch": 0.3564, + "grad_norm": 1.2327573244280021, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 35640 + }, + { + "epoch": 0.35641, + "grad_norm": 1.409193708183077, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35641 + }, + { + "epoch": 0.35642, + "grad_norm": 1.3969333083000945, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 35642 + }, + { + "epoch": 0.35643, + "grad_norm": 1.4386362791557004, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 35643 + }, + { + "epoch": 0.35644, + "grad_norm": 1.5398697118708664, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35644 + }, + { + "epoch": 0.35645, + "grad_norm": 1.6792592506776873, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35645 + }, + { + "epoch": 0.35646, + "grad_norm": 1.420365449982015, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 35646 + }, + { + "epoch": 0.35647, + "grad_norm": 1.696241338460024, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 35647 + }, + { + "epoch": 0.35648, + "grad_norm": 1.4865601219731075, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 35648 + }, + { + "epoch": 0.35649, + "grad_norm": 1.7213657936292424, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 35649 + }, + { + "epoch": 0.3565, + "grad_norm": 1.2489506856608352, + "learning_rate": 0.003, + "loss": 4.023, + "step": 35650 + }, + { + "epoch": 0.35651, + "grad_norm": 1.489870516524094, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 35651 + }, + { + "epoch": 0.35652, + "grad_norm": 1.7799661424667266, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 35652 + }, + { + "epoch": 0.35653, + "grad_norm": 1.4493607271942177, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 35653 + }, + { + "epoch": 0.35654, + "grad_norm": 1.496680679059927, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35654 + }, + { + "epoch": 0.35655, + "grad_norm": 1.5118334071497574, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 35655 + }, + { + "epoch": 0.35656, + "grad_norm": 1.4794723155531648, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 35656 + }, + { + "epoch": 0.35657, + "grad_norm": 1.5011536026394052, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35657 + }, + { + "epoch": 0.35658, + "grad_norm": 1.4191182076512332, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 35658 + }, + { + "epoch": 0.35659, + "grad_norm": 1.6859226300436851, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35659 + }, + { + "epoch": 0.3566, + "grad_norm": 1.4780520278399327, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 35660 + }, + { + "epoch": 0.35661, + "grad_norm": 1.2841494087582048, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 35661 + }, + { + "epoch": 0.35662, + "grad_norm": 1.4075945434982053, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 35662 + }, + { + "epoch": 0.35663, + "grad_norm": 1.606196427927515, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35663 + }, + { + "epoch": 0.35664, + "grad_norm": 1.607305205094362, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 35664 + }, + { + "epoch": 0.35665, + "grad_norm": 1.5642360853269968, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 35665 + }, + { + "epoch": 0.35666, + "grad_norm": 1.6107105024464317, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35666 + }, + { + "epoch": 0.35667, + "grad_norm": 1.1847786530845368, + "learning_rate": 0.003, + "loss": 4.02, + "step": 35667 + }, + { + "epoch": 0.35668, + "grad_norm": 1.3607952119504705, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35668 + }, + { + "epoch": 0.35669, + "grad_norm": 1.4727349564524472, + "learning_rate": 0.003, + "loss": 4.045, + "step": 35669 + }, + { + "epoch": 0.3567, + "grad_norm": 1.6715459972855946, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 35670 + }, + { + "epoch": 0.35671, + "grad_norm": 1.734943847522449, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35671 + }, + { + "epoch": 0.35672, + "grad_norm": 1.9084808411963072, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 35672 + }, + { + "epoch": 0.35673, + "grad_norm": 1.4524369546235905, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 35673 + }, + { + "epoch": 0.35674, + "grad_norm": 1.4087916282938322, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 35674 + }, + { + "epoch": 0.35675, + "grad_norm": 1.3095419913268291, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 35675 + }, + { + "epoch": 0.35676, + "grad_norm": 1.6010770355998396, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 35676 + }, + { + "epoch": 0.35677, + "grad_norm": 1.6231417043275238, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 35677 + }, + { + "epoch": 0.35678, + "grad_norm": 1.446903857286104, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 35678 + }, + { + "epoch": 0.35679, + "grad_norm": 1.405456193902152, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 35679 + }, + { + "epoch": 0.3568, + "grad_norm": 1.729781409325779, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 35680 + }, + { + "epoch": 0.35681, + "grad_norm": 1.185849284916496, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 35681 + }, + { + "epoch": 0.35682, + "grad_norm": 1.9041251719226029, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 35682 + }, + { + "epoch": 0.35683, + "grad_norm": 1.2952421032958663, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35683 + }, + { + "epoch": 0.35684, + "grad_norm": 1.4867487238720172, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 35684 + }, + { + "epoch": 0.35685, + "grad_norm": 1.2621370532143146, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 35685 + }, + { + "epoch": 0.35686, + "grad_norm": 1.666876130515192, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 35686 + }, + { + "epoch": 0.35687, + "grad_norm": 1.474140965835461, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 35687 + }, + { + "epoch": 0.35688, + "grad_norm": 1.290426927219482, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 35688 + }, + { + "epoch": 0.35689, + "grad_norm": 1.8685574536273313, + "learning_rate": 0.003, + "loss": 4.028, + "step": 35689 + }, + { + "epoch": 0.3569, + "grad_norm": 1.4407630401315126, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 35690 + }, + { + "epoch": 0.35691, + "grad_norm": 1.3567425597440537, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 35691 + }, + { + "epoch": 0.35692, + "grad_norm": 1.4537405222399726, + "learning_rate": 0.003, + "loss": 4.049, + "step": 35692 + }, + { + "epoch": 0.35693, + "grad_norm": 1.5018143871989462, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 35693 + }, + { + "epoch": 0.35694, + "grad_norm": 1.622908839891153, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 35694 + }, + { + "epoch": 0.35695, + "grad_norm": 1.6340822202064373, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 35695 + }, + { + "epoch": 0.35696, + "grad_norm": 1.4153360002320852, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 35696 + }, + { + "epoch": 0.35697, + "grad_norm": 1.3814122194746874, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 35697 + }, + { + "epoch": 0.35698, + "grad_norm": 1.8602676694793023, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 35698 + }, + { + "epoch": 0.35699, + "grad_norm": 1.2412309684524048, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 35699 + }, + { + "epoch": 0.357, + "grad_norm": 1.9194984342208283, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35700 + }, + { + "epoch": 0.35701, + "grad_norm": 1.2146891400923197, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 35701 + }, + { + "epoch": 0.35702, + "grad_norm": 1.743294795229955, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35702 + }, + { + "epoch": 0.35703, + "grad_norm": 1.3205073040068833, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 35703 + }, + { + "epoch": 0.35704, + "grad_norm": 1.4253570881733129, + "learning_rate": 0.003, + "loss": 4.037, + "step": 35704 + }, + { + "epoch": 0.35705, + "grad_norm": 1.5789312177658652, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 35705 + }, + { + "epoch": 0.35706, + "grad_norm": 1.372718505350424, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 35706 + }, + { + "epoch": 0.35707, + "grad_norm": 1.653265783707475, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 35707 + }, + { + "epoch": 0.35708, + "grad_norm": 1.4069025547211487, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 35708 + }, + { + "epoch": 0.35709, + "grad_norm": 1.6133527161138752, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 35709 + }, + { + "epoch": 0.3571, + "grad_norm": 1.3578909533019385, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 35710 + }, + { + "epoch": 0.35711, + "grad_norm": 1.429450513546705, + "learning_rate": 0.003, + "loss": 4.032, + "step": 35711 + }, + { + "epoch": 0.35712, + "grad_norm": 1.6286113728114293, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 35712 + }, + { + "epoch": 0.35713, + "grad_norm": 1.7252753824962948, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 35713 + }, + { + "epoch": 0.35714, + "grad_norm": 1.1766600589248346, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 35714 + }, + { + "epoch": 0.35715, + "grad_norm": 1.5859659650883058, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35715 + }, + { + "epoch": 0.35716, + "grad_norm": 1.5061096922935646, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 35716 + }, + { + "epoch": 0.35717, + "grad_norm": 1.8206389698280578, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 35717 + }, + { + "epoch": 0.35718, + "grad_norm": 1.4210083171800678, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 35718 + }, + { + "epoch": 0.35719, + "grad_norm": 1.4722866467516595, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 35719 + }, + { + "epoch": 0.3572, + "grad_norm": 1.4877846687369056, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 35720 + }, + { + "epoch": 0.35721, + "grad_norm": 1.4100297997574471, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 35721 + }, + { + "epoch": 0.35722, + "grad_norm": 1.714214248690339, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 35722 + }, + { + "epoch": 0.35723, + "grad_norm": 1.280986016519116, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 35723 + }, + { + "epoch": 0.35724, + "grad_norm": 1.5956529706694322, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 35724 + }, + { + "epoch": 0.35725, + "grad_norm": 1.343713011825708, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35725 + }, + { + "epoch": 0.35726, + "grad_norm": 1.5966682127273017, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 35726 + }, + { + "epoch": 0.35727, + "grad_norm": 1.5731211804881318, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 35727 + }, + { + "epoch": 0.35728, + "grad_norm": 1.5836906795010388, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 35728 + }, + { + "epoch": 0.35729, + "grad_norm": 1.7679620359029267, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35729 + }, + { + "epoch": 0.3573, + "grad_norm": 1.568235312426561, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 35730 + }, + { + "epoch": 0.35731, + "grad_norm": 1.6145971599904867, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 35731 + }, + { + "epoch": 0.35732, + "grad_norm": 1.5099808773691508, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 35732 + }, + { + "epoch": 0.35733, + "grad_norm": 1.2842581211391026, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 35733 + }, + { + "epoch": 0.35734, + "grad_norm": 1.6915014032755886, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 35734 + }, + { + "epoch": 0.35735, + "grad_norm": 1.2479718206076198, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 35735 + }, + { + "epoch": 0.35736, + "grad_norm": 1.5208998931142312, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35736 + }, + { + "epoch": 0.35737, + "grad_norm": 1.350704216463525, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 35737 + }, + { + "epoch": 0.35738, + "grad_norm": 2.0548030591201507, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 35738 + }, + { + "epoch": 0.35739, + "grad_norm": 1.620688736284845, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 35739 + }, + { + "epoch": 0.3574, + "grad_norm": 1.391122374066644, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35740 + }, + { + "epoch": 0.35741, + "grad_norm": 1.4107212673359577, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 35741 + }, + { + "epoch": 0.35742, + "grad_norm": 1.631652065731085, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 35742 + }, + { + "epoch": 0.35743, + "grad_norm": 1.4006780390660873, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 35743 + }, + { + "epoch": 0.35744, + "grad_norm": 1.4552271944034039, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 35744 + }, + { + "epoch": 0.35745, + "grad_norm": 1.4976346006601544, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35745 + }, + { + "epoch": 0.35746, + "grad_norm": 1.4547900559031097, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 35746 + }, + { + "epoch": 0.35747, + "grad_norm": 1.4983810195333573, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 35747 + }, + { + "epoch": 0.35748, + "grad_norm": 1.6305187507390007, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 35748 + }, + { + "epoch": 0.35749, + "grad_norm": 1.320937427476209, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 35749 + }, + { + "epoch": 0.3575, + "grad_norm": 1.6040565921314316, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 35750 + }, + { + "epoch": 0.35751, + "grad_norm": 1.2853115726211173, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 35751 + }, + { + "epoch": 0.35752, + "grad_norm": 1.6771144388544863, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35752 + }, + { + "epoch": 0.35753, + "grad_norm": 1.1410595162986912, + "learning_rate": 0.003, + "loss": 4.059, + "step": 35753 + }, + { + "epoch": 0.35754, + "grad_norm": 1.7850970872481429, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 35754 + }, + { + "epoch": 0.35755, + "grad_norm": 1.7327783453286831, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 35755 + }, + { + "epoch": 0.35756, + "grad_norm": 1.5712563951138097, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 35756 + }, + { + "epoch": 0.35757, + "grad_norm": 1.4406998182325244, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 35757 + }, + { + "epoch": 0.35758, + "grad_norm": 1.4682453116943437, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 35758 + }, + { + "epoch": 0.35759, + "grad_norm": 1.61464740633974, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 35759 + }, + { + "epoch": 0.3576, + "grad_norm": 1.3833115206591566, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 35760 + }, + { + "epoch": 0.35761, + "grad_norm": 1.6543326725714436, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 35761 + }, + { + "epoch": 0.35762, + "grad_norm": 1.556036191316868, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35762 + }, + { + "epoch": 0.35763, + "grad_norm": 1.3189175009970078, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 35763 + }, + { + "epoch": 0.35764, + "grad_norm": 1.710519155579211, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 35764 + }, + { + "epoch": 0.35765, + "grad_norm": 1.3922864505118173, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 35765 + }, + { + "epoch": 0.35766, + "grad_norm": 1.4591565090291787, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 35766 + }, + { + "epoch": 0.35767, + "grad_norm": 1.4208825278815254, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 35767 + }, + { + "epoch": 0.35768, + "grad_norm": 1.6653549061350394, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 35768 + }, + { + "epoch": 0.35769, + "grad_norm": 1.50623597983103, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 35769 + }, + { + "epoch": 0.3577, + "grad_norm": 1.2229967446940444, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 35770 + }, + { + "epoch": 0.35771, + "grad_norm": 1.713560739358256, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 35771 + }, + { + "epoch": 0.35772, + "grad_norm": 1.4221160322594983, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 35772 + }, + { + "epoch": 0.35773, + "grad_norm": 1.8492859410087619, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 35773 + }, + { + "epoch": 0.35774, + "grad_norm": 1.7177996968286022, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 35774 + }, + { + "epoch": 0.35775, + "grad_norm": 1.4991722454684777, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35775 + }, + { + "epoch": 0.35776, + "grad_norm": 1.5657212411044177, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 35776 + }, + { + "epoch": 0.35777, + "grad_norm": 1.651976566662742, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 35777 + }, + { + "epoch": 0.35778, + "grad_norm": 1.2446122141553273, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 35778 + }, + { + "epoch": 0.35779, + "grad_norm": 1.5511945238122866, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35779 + }, + { + "epoch": 0.3578, + "grad_norm": 1.470652861159185, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 35780 + }, + { + "epoch": 0.35781, + "grad_norm": 1.5718775016973106, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 35781 + }, + { + "epoch": 0.35782, + "grad_norm": 1.3469490881091173, + "learning_rate": 0.003, + "loss": 4.036, + "step": 35782 + }, + { + "epoch": 0.35783, + "grad_norm": 1.9299664758589254, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 35783 + }, + { + "epoch": 0.35784, + "grad_norm": 1.1530051472943292, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 35784 + }, + { + "epoch": 0.35785, + "grad_norm": 2.216822076278861, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 35785 + }, + { + "epoch": 0.35786, + "grad_norm": 1.3193684653164655, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 35786 + }, + { + "epoch": 0.35787, + "grad_norm": 1.6240600116908996, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 35787 + }, + { + "epoch": 0.35788, + "grad_norm": 1.5097561877305545, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 35788 + }, + { + "epoch": 0.35789, + "grad_norm": 1.7905314831535615, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 35789 + }, + { + "epoch": 0.3579, + "grad_norm": 1.52772339667006, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 35790 + }, + { + "epoch": 0.35791, + "grad_norm": 1.320885326621326, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35791 + }, + { + "epoch": 0.35792, + "grad_norm": 1.3792524204810186, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 35792 + }, + { + "epoch": 0.35793, + "grad_norm": 1.6312103527246113, + "learning_rate": 0.003, + "loss": 4.058, + "step": 35793 + }, + { + "epoch": 0.35794, + "grad_norm": 1.2566629904364084, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 35794 + }, + { + "epoch": 0.35795, + "grad_norm": 1.6774233848019675, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 35795 + }, + { + "epoch": 0.35796, + "grad_norm": 1.2710034594167334, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35796 + }, + { + "epoch": 0.35797, + "grad_norm": 1.5285840041004513, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 35797 + }, + { + "epoch": 0.35798, + "grad_norm": 1.3188993299805447, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 35798 + }, + { + "epoch": 0.35799, + "grad_norm": 1.80406340503135, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 35799 + }, + { + "epoch": 0.358, + "grad_norm": 1.392381431468969, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35800 + }, + { + "epoch": 0.35801, + "grad_norm": 1.6729366020758951, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 35801 + }, + { + "epoch": 0.35802, + "grad_norm": 1.4913579288693781, + "learning_rate": 0.003, + "loss": 4.055, + "step": 35802 + }, + { + "epoch": 0.35803, + "grad_norm": 1.5363631306349825, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 35803 + }, + { + "epoch": 0.35804, + "grad_norm": 1.308078748638449, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 35804 + }, + { + "epoch": 0.35805, + "grad_norm": 1.5525234311280265, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 35805 + }, + { + "epoch": 0.35806, + "grad_norm": 1.7531717845579626, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 35806 + }, + { + "epoch": 0.35807, + "grad_norm": 1.2762619722738033, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 35807 + }, + { + "epoch": 0.35808, + "grad_norm": 1.470831615767065, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 35808 + }, + { + "epoch": 0.35809, + "grad_norm": 1.352060951881289, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 35809 + }, + { + "epoch": 0.3581, + "grad_norm": 1.5514414560714802, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 35810 + }, + { + "epoch": 0.35811, + "grad_norm": 1.7251326501396775, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 35811 + }, + { + "epoch": 0.35812, + "grad_norm": 1.4660985684180117, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 35812 + }, + { + "epoch": 0.35813, + "grad_norm": 1.8232734834663435, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35813 + }, + { + "epoch": 0.35814, + "grad_norm": 1.475752104101199, + "learning_rate": 0.003, + "loss": 4.029, + "step": 35814 + }, + { + "epoch": 0.35815, + "grad_norm": 1.3446939426821238, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 35815 + }, + { + "epoch": 0.35816, + "grad_norm": 1.4087935702751033, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 35816 + }, + { + "epoch": 0.35817, + "grad_norm": 1.6798137445900267, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 35817 + }, + { + "epoch": 0.35818, + "grad_norm": 1.4393443462484774, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 35818 + }, + { + "epoch": 0.35819, + "grad_norm": 1.4883591823777471, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 35819 + }, + { + "epoch": 0.3582, + "grad_norm": 1.511470667202524, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 35820 + }, + { + "epoch": 0.35821, + "grad_norm": 1.7788783350632682, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35821 + }, + { + "epoch": 0.35822, + "grad_norm": 1.261836570302086, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 35822 + }, + { + "epoch": 0.35823, + "grad_norm": 1.7080162157216987, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 35823 + }, + { + "epoch": 0.35824, + "grad_norm": 1.3870676484598083, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 35824 + }, + { + "epoch": 0.35825, + "grad_norm": 1.3275221865848144, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 35825 + }, + { + "epoch": 0.35826, + "grad_norm": 1.6875966153315096, + "learning_rate": 0.003, + "loss": 4.05, + "step": 35826 + }, + { + "epoch": 0.35827, + "grad_norm": 1.3758465136857103, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 35827 + }, + { + "epoch": 0.35828, + "grad_norm": 1.7663869395841225, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 35828 + }, + { + "epoch": 0.35829, + "grad_norm": 1.1687764505926586, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 35829 + }, + { + "epoch": 0.3583, + "grad_norm": 1.6752023965241964, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 35830 + }, + { + "epoch": 0.35831, + "grad_norm": 1.6429987798494883, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35831 + }, + { + "epoch": 0.35832, + "grad_norm": 1.5480880094767893, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 35832 + }, + { + "epoch": 0.35833, + "grad_norm": 1.3972029635683256, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 35833 + }, + { + "epoch": 0.35834, + "grad_norm": 1.4299622862534573, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35834 + }, + { + "epoch": 0.35835, + "grad_norm": 1.2796690578062615, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 35835 + }, + { + "epoch": 0.35836, + "grad_norm": 1.583598530146338, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35836 + }, + { + "epoch": 0.35837, + "grad_norm": 1.3674177877405667, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 35837 + }, + { + "epoch": 0.35838, + "grad_norm": 1.5095180486274258, + "learning_rate": 0.003, + "loss": 4.025, + "step": 35838 + }, + { + "epoch": 0.35839, + "grad_norm": 1.6349234002899011, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 35839 + }, + { + "epoch": 0.3584, + "grad_norm": 1.6117002195212091, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35840 + }, + { + "epoch": 0.35841, + "grad_norm": 1.4228664252001981, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35841 + }, + { + "epoch": 0.35842, + "grad_norm": 1.6385819795861407, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 35842 + }, + { + "epoch": 0.35843, + "grad_norm": 1.3715111088843233, + "learning_rate": 0.003, + "loss": 4.044, + "step": 35843 + }, + { + "epoch": 0.35844, + "grad_norm": 1.5103874758289804, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35844 + }, + { + "epoch": 0.35845, + "grad_norm": 1.2027506493456488, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35845 + }, + { + "epoch": 0.35846, + "grad_norm": 1.7212929822506042, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 35846 + }, + { + "epoch": 0.35847, + "grad_norm": 1.4001029564624874, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 35847 + }, + { + "epoch": 0.35848, + "grad_norm": 1.499473268020664, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 35848 + }, + { + "epoch": 0.35849, + "grad_norm": 1.256057713772303, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 35849 + }, + { + "epoch": 0.3585, + "grad_norm": 1.612820336752933, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 35850 + }, + { + "epoch": 0.35851, + "grad_norm": 1.3762389751943718, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 35851 + }, + { + "epoch": 0.35852, + "grad_norm": 1.6656273818454863, + "learning_rate": 0.003, + "loss": 4.057, + "step": 35852 + }, + { + "epoch": 0.35853, + "grad_norm": 1.915532215560251, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 35853 + }, + { + "epoch": 0.35854, + "grad_norm": 1.399710950402098, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 35854 + }, + { + "epoch": 0.35855, + "grad_norm": 1.5005539145791544, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 35855 + }, + { + "epoch": 0.35856, + "grad_norm": 1.5142682002181287, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 35856 + }, + { + "epoch": 0.35857, + "grad_norm": 1.655453320254841, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 35857 + }, + { + "epoch": 0.35858, + "grad_norm": 1.438027717930875, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 35858 + }, + { + "epoch": 0.35859, + "grad_norm": 1.4275117522983563, + "learning_rate": 0.003, + "loss": 4.034, + "step": 35859 + }, + { + "epoch": 0.3586, + "grad_norm": 1.6524740848175234, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35860 + }, + { + "epoch": 0.35861, + "grad_norm": 1.497940928047594, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 35861 + }, + { + "epoch": 0.35862, + "grad_norm": 1.4890706469695205, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 35862 + }, + { + "epoch": 0.35863, + "grad_norm": 1.6858780923584025, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 35863 + }, + { + "epoch": 0.35864, + "grad_norm": 1.5155724711906748, + "learning_rate": 0.003, + "loss": 4.009, + "step": 35864 + }, + { + "epoch": 0.35865, + "grad_norm": 1.500955109427621, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 35865 + }, + { + "epoch": 0.35866, + "grad_norm": 1.9465856296081905, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 35866 + }, + { + "epoch": 0.35867, + "grad_norm": 1.2729267049129296, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 35867 + }, + { + "epoch": 0.35868, + "grad_norm": 1.4683531119858377, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35868 + }, + { + "epoch": 0.35869, + "grad_norm": 1.7998947340207414, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 35869 + }, + { + "epoch": 0.3587, + "grad_norm": 1.3199731394741496, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 35870 + }, + { + "epoch": 0.35871, + "grad_norm": 1.8392048014221805, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 35871 + }, + { + "epoch": 0.35872, + "grad_norm": 1.3329226810479164, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 35872 + }, + { + "epoch": 0.35873, + "grad_norm": 1.5396901023565503, + "learning_rate": 0.003, + "loss": 4.039, + "step": 35873 + }, + { + "epoch": 0.35874, + "grad_norm": 1.3416110817101106, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 35874 + }, + { + "epoch": 0.35875, + "grad_norm": 1.8462404948000102, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 35875 + }, + { + "epoch": 0.35876, + "grad_norm": 1.4478009192047405, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 35876 + }, + { + "epoch": 0.35877, + "grad_norm": 1.6293435150463882, + "learning_rate": 0.003, + "loss": 4.046, + "step": 35877 + }, + { + "epoch": 0.35878, + "grad_norm": 1.6102203332064962, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 35878 + }, + { + "epoch": 0.35879, + "grad_norm": 1.4027764809255947, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 35879 + }, + { + "epoch": 0.3588, + "grad_norm": 1.5899012631054845, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 35880 + }, + { + "epoch": 0.35881, + "grad_norm": 1.49328241728702, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 35881 + }, + { + "epoch": 0.35882, + "grad_norm": 1.5059717194149618, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 35882 + }, + { + "epoch": 0.35883, + "grad_norm": 1.4589669654089106, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 35883 + }, + { + "epoch": 0.35884, + "grad_norm": 1.421486985959723, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 35884 + }, + { + "epoch": 0.35885, + "grad_norm": 1.4144641848925366, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 35885 + }, + { + "epoch": 0.35886, + "grad_norm": 1.3989314875821846, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 35886 + }, + { + "epoch": 0.35887, + "grad_norm": 1.4463781548311523, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 35887 + }, + { + "epoch": 0.35888, + "grad_norm": 1.6599896414050666, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 35888 + }, + { + "epoch": 0.35889, + "grad_norm": 1.4330069769090816, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 35889 + }, + { + "epoch": 0.3589, + "grad_norm": 1.4962219431521366, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 35890 + }, + { + "epoch": 0.35891, + "grad_norm": 1.7604893636798153, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 35891 + }, + { + "epoch": 0.35892, + "grad_norm": 1.371792849697054, + "learning_rate": 0.003, + "loss": 4.015, + "step": 35892 + }, + { + "epoch": 0.35893, + "grad_norm": 1.4372273384296423, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 35893 + }, + { + "epoch": 0.35894, + "grad_norm": 1.4899198061966195, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 35894 + }, + { + "epoch": 0.35895, + "grad_norm": 1.641283539892612, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 35895 + }, + { + "epoch": 0.35896, + "grad_norm": 1.2998065929911464, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35896 + }, + { + "epoch": 0.35897, + "grad_norm": 1.673229219701244, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 35897 + }, + { + "epoch": 0.35898, + "grad_norm": 1.5977367553536006, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 35898 + }, + { + "epoch": 0.35899, + "grad_norm": 1.6692761424543778, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35899 + }, + { + "epoch": 0.359, + "grad_norm": 1.2473879550051732, + "learning_rate": 0.003, + "loss": 4.035, + "step": 35900 + }, + { + "epoch": 0.35901, + "grad_norm": 1.5875196409199435, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 35901 + }, + { + "epoch": 0.35902, + "grad_norm": 1.773508685440876, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 35902 + }, + { + "epoch": 0.35903, + "grad_norm": 1.8546014490292493, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 35903 + }, + { + "epoch": 0.35904, + "grad_norm": 1.175286939074762, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 35904 + }, + { + "epoch": 0.35905, + "grad_norm": 1.6060889008262476, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 35905 + }, + { + "epoch": 0.35906, + "grad_norm": 1.7726931443552036, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 35906 + }, + { + "epoch": 0.35907, + "grad_norm": 1.4152265751548077, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 35907 + }, + { + "epoch": 0.35908, + "grad_norm": 1.6010937661138893, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35908 + }, + { + "epoch": 0.35909, + "grad_norm": 1.4853690079881956, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 35909 + }, + { + "epoch": 0.3591, + "grad_norm": 1.4441117172098905, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 35910 + }, + { + "epoch": 0.35911, + "grad_norm": 1.4485673901610592, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35911 + }, + { + "epoch": 0.35912, + "grad_norm": 1.3670057399193798, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 35912 + }, + { + "epoch": 0.35913, + "grad_norm": 1.7502258605200456, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 35913 + }, + { + "epoch": 0.35914, + "grad_norm": 1.368169149778141, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 35914 + }, + { + "epoch": 0.35915, + "grad_norm": 1.697972691317173, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 35915 + }, + { + "epoch": 0.35916, + "grad_norm": 1.2835801189764946, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 35916 + }, + { + "epoch": 0.35917, + "grad_norm": 1.5620931434172254, + "learning_rate": 0.003, + "loss": 4.033, + "step": 35917 + }, + { + "epoch": 0.35918, + "grad_norm": 1.4907521204492888, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 35918 + }, + { + "epoch": 0.35919, + "grad_norm": 1.240785496897055, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 35919 + }, + { + "epoch": 0.3592, + "grad_norm": 1.6038850436362846, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 35920 + }, + { + "epoch": 0.35921, + "grad_norm": 1.5930853180526772, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 35921 + }, + { + "epoch": 0.35922, + "grad_norm": 1.5439366513833805, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 35922 + }, + { + "epoch": 0.35923, + "grad_norm": 1.28483976340618, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 35923 + }, + { + "epoch": 0.35924, + "grad_norm": 2.0048860297901934, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 35924 + }, + { + "epoch": 0.35925, + "grad_norm": 1.070669184662594, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 35925 + }, + { + "epoch": 0.35926, + "grad_norm": 1.5567727062295096, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 35926 + }, + { + "epoch": 0.35927, + "grad_norm": 1.3767757870731885, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 35927 + }, + { + "epoch": 0.35928, + "grad_norm": 1.981069583603281, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 35928 + }, + { + "epoch": 0.35929, + "grad_norm": 1.5307421190304122, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 35929 + }, + { + "epoch": 0.3593, + "grad_norm": 1.7123981157505177, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 35930 + }, + { + "epoch": 0.35931, + "grad_norm": 1.6010462446159044, + "learning_rate": 0.003, + "loss": 4.048, + "step": 35931 + }, + { + "epoch": 0.35932, + "grad_norm": 1.7413286651440458, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 35932 + }, + { + "epoch": 0.35933, + "grad_norm": 1.4015764505616393, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 35933 + }, + { + "epoch": 0.35934, + "grad_norm": 1.7081613340733932, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 35934 + }, + { + "epoch": 0.35935, + "grad_norm": 1.476508172674924, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 35935 + }, + { + "epoch": 0.35936, + "grad_norm": 1.4951266635843974, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 35936 + }, + { + "epoch": 0.35937, + "grad_norm": 1.5738670296885113, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 35937 + }, + { + "epoch": 0.35938, + "grad_norm": 1.3502164895845603, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 35938 + }, + { + "epoch": 0.35939, + "grad_norm": 1.297741642942793, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 35939 + }, + { + "epoch": 0.3594, + "grad_norm": 1.8510856450407753, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 35940 + }, + { + "epoch": 0.35941, + "grad_norm": 1.4056926936973384, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 35941 + }, + { + "epoch": 0.35942, + "grad_norm": 1.7763117243568396, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 35942 + }, + { + "epoch": 0.35943, + "grad_norm": 1.4012081692823624, + "learning_rate": 0.003, + "loss": 4.036, + "step": 35943 + }, + { + "epoch": 0.35944, + "grad_norm": 1.500337088979123, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 35944 + }, + { + "epoch": 0.35945, + "grad_norm": 1.4401511461830923, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 35945 + }, + { + "epoch": 0.35946, + "grad_norm": 1.684868512177644, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 35946 + }, + { + "epoch": 0.35947, + "grad_norm": 1.4633020120851707, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 35947 + }, + { + "epoch": 0.35948, + "grad_norm": 1.3414176058131781, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 35948 + }, + { + "epoch": 0.35949, + "grad_norm": 1.4123832971824766, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 35949 + }, + { + "epoch": 0.3595, + "grad_norm": 1.5375742819220612, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 35950 + }, + { + "epoch": 0.35951, + "grad_norm": 1.4014421356933626, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 35951 + }, + { + "epoch": 0.35952, + "grad_norm": 1.4694230189309794, + "learning_rate": 0.003, + "loss": 4.055, + "step": 35952 + }, + { + "epoch": 0.35953, + "grad_norm": 1.646044964054573, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 35953 + }, + { + "epoch": 0.35954, + "grad_norm": 1.4303504375105147, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 35954 + }, + { + "epoch": 0.35955, + "grad_norm": 1.4630234293579327, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 35955 + }, + { + "epoch": 0.35956, + "grad_norm": 1.5606879211867968, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 35956 + }, + { + "epoch": 0.35957, + "grad_norm": 1.2929871906218704, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 35957 + }, + { + "epoch": 0.35958, + "grad_norm": 1.737875753957061, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 35958 + }, + { + "epoch": 0.35959, + "grad_norm": 1.132313492981009, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 35959 + }, + { + "epoch": 0.3596, + "grad_norm": 1.6946601324010477, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 35960 + }, + { + "epoch": 0.35961, + "grad_norm": 1.2662637839643363, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 35961 + }, + { + "epoch": 0.35962, + "grad_norm": 2.0354937765717187, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 35962 + }, + { + "epoch": 0.35963, + "grad_norm": 1.4050390283829708, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 35963 + }, + { + "epoch": 0.35964, + "grad_norm": 2.2641311979751353, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 35964 + }, + { + "epoch": 0.35965, + "grad_norm": 1.2252349977969843, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 35965 + }, + { + "epoch": 0.35966, + "grad_norm": 1.3319606564975577, + "learning_rate": 0.003, + "loss": 4.05, + "step": 35966 + }, + { + "epoch": 0.35967, + "grad_norm": 1.680073477795038, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 35967 + }, + { + "epoch": 0.35968, + "grad_norm": 1.5012200731811312, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 35968 + }, + { + "epoch": 0.35969, + "grad_norm": 1.499746630333825, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 35969 + }, + { + "epoch": 0.3597, + "grad_norm": 1.285195597585608, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 35970 + }, + { + "epoch": 0.35971, + "grad_norm": 1.9534862546495462, + "learning_rate": 0.003, + "loss": 4.053, + "step": 35971 + }, + { + "epoch": 0.35972, + "grad_norm": 1.4620143540701143, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35972 + }, + { + "epoch": 0.35973, + "grad_norm": 1.701997734945575, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 35973 + }, + { + "epoch": 0.35974, + "grad_norm": 1.5492776990890595, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 35974 + }, + { + "epoch": 0.35975, + "grad_norm": 1.636920619760644, + "learning_rate": 0.003, + "loss": 4.042, + "step": 35975 + }, + { + "epoch": 0.35976, + "grad_norm": 1.6518750238256463, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 35976 + }, + { + "epoch": 0.35977, + "grad_norm": 1.2154411885239822, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 35977 + }, + { + "epoch": 0.35978, + "grad_norm": 1.7305165266218918, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 35978 + }, + { + "epoch": 0.35979, + "grad_norm": 1.4552357652264039, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 35979 + }, + { + "epoch": 0.3598, + "grad_norm": 1.2800431744387564, + "learning_rate": 0.003, + "loss": 4.007, + "step": 35980 + }, + { + "epoch": 0.35981, + "grad_norm": 1.6342212570569734, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 35981 + }, + { + "epoch": 0.35982, + "grad_norm": 1.2319568176689502, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 35982 + }, + { + "epoch": 0.35983, + "grad_norm": 1.4996887837938169, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 35983 + }, + { + "epoch": 0.35984, + "grad_norm": 1.4586621480193507, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 35984 + }, + { + "epoch": 0.35985, + "grad_norm": 1.5568950792157485, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35985 + }, + { + "epoch": 0.35986, + "grad_norm": 1.5417921404753194, + "learning_rate": 0.003, + "loss": 4.043, + "step": 35986 + }, + { + "epoch": 0.35987, + "grad_norm": 1.5250062771814643, + "learning_rate": 0.003, + "loss": 4.047, + "step": 35987 + }, + { + "epoch": 0.35988, + "grad_norm": 1.3604855809572096, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 35988 + }, + { + "epoch": 0.35989, + "grad_norm": 1.8526634448535286, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 35989 + }, + { + "epoch": 0.3599, + "grad_norm": 1.729250350182403, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 35990 + }, + { + "epoch": 0.35991, + "grad_norm": 1.500811271873874, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 35991 + }, + { + "epoch": 0.35992, + "grad_norm": 1.7581901815052208, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 35992 + }, + { + "epoch": 0.35993, + "grad_norm": 1.2718456035031147, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 35993 + }, + { + "epoch": 0.35994, + "grad_norm": 1.474096943504827, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 35994 + }, + { + "epoch": 0.35995, + "grad_norm": 1.6036165219047895, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 35995 + }, + { + "epoch": 0.35996, + "grad_norm": 1.6041232341621592, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 35996 + }, + { + "epoch": 0.35997, + "grad_norm": 1.4987539667038399, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 35997 + }, + { + "epoch": 0.35998, + "grad_norm": 1.5188452267804873, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 35998 + }, + { + "epoch": 0.35999, + "grad_norm": 1.4928486335294038, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 35999 + }, + { + "epoch": 0.36, + "grad_norm": 1.521139044489307, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 36000 + }, + { + "epoch": 0.36001, + "grad_norm": 1.4017996126436723, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36001 + }, + { + "epoch": 0.36002, + "grad_norm": 1.6527484613605146, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 36002 + }, + { + "epoch": 0.36003, + "grad_norm": 1.3589991106474761, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 36003 + }, + { + "epoch": 0.36004, + "grad_norm": 1.998303307557316, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 36004 + }, + { + "epoch": 0.36005, + "grad_norm": 1.3051986617315008, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 36005 + }, + { + "epoch": 0.36006, + "grad_norm": 1.71850823049628, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 36006 + }, + { + "epoch": 0.36007, + "grad_norm": 1.4846498419581928, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 36007 + }, + { + "epoch": 0.36008, + "grad_norm": 1.336038462373874, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36008 + }, + { + "epoch": 0.36009, + "grad_norm": 1.7320140775140238, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36009 + }, + { + "epoch": 0.3601, + "grad_norm": 1.709257158148347, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36010 + }, + { + "epoch": 0.36011, + "grad_norm": 1.3538868861585256, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 36011 + }, + { + "epoch": 0.36012, + "grad_norm": 1.6204546459192655, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36012 + }, + { + "epoch": 0.36013, + "grad_norm": 1.7224657606860827, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 36013 + }, + { + "epoch": 0.36014, + "grad_norm": 1.2098554485048636, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36014 + }, + { + "epoch": 0.36015, + "grad_norm": 1.6440607733428285, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36015 + }, + { + "epoch": 0.36016, + "grad_norm": 1.3960707693735617, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 36016 + }, + { + "epoch": 0.36017, + "grad_norm": 1.8059762826198194, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 36017 + }, + { + "epoch": 0.36018, + "grad_norm": 1.2203332572073236, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 36018 + }, + { + "epoch": 0.36019, + "grad_norm": 1.7026767965546141, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 36019 + }, + { + "epoch": 0.3602, + "grad_norm": 1.387675958836972, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 36020 + }, + { + "epoch": 0.36021, + "grad_norm": 1.5805579556303926, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36021 + }, + { + "epoch": 0.36022, + "grad_norm": 1.3970684738931343, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 36022 + }, + { + "epoch": 0.36023, + "grad_norm": 1.5068496419749862, + "learning_rate": 0.003, + "loss": 4.02, + "step": 36023 + }, + { + "epoch": 0.36024, + "grad_norm": 1.4962258063794436, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 36024 + }, + { + "epoch": 0.36025, + "grad_norm": 1.5374900047102278, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 36025 + }, + { + "epoch": 0.36026, + "grad_norm": 1.6053886190705526, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 36026 + }, + { + "epoch": 0.36027, + "grad_norm": 1.4268285717159208, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 36027 + }, + { + "epoch": 0.36028, + "grad_norm": 1.4537379191274766, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36028 + }, + { + "epoch": 0.36029, + "grad_norm": 1.3757385170043352, + "learning_rate": 0.003, + "loss": 4.032, + "step": 36029 + }, + { + "epoch": 0.3603, + "grad_norm": 1.4724984647144295, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 36030 + }, + { + "epoch": 0.36031, + "grad_norm": 1.800055626907293, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 36031 + }, + { + "epoch": 0.36032, + "grad_norm": 1.5171902100732322, + "learning_rate": 0.003, + "loss": 4.073, + "step": 36032 + }, + { + "epoch": 0.36033, + "grad_norm": 1.9487674345864372, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 36033 + }, + { + "epoch": 0.36034, + "grad_norm": 1.530354515921578, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36034 + }, + { + "epoch": 0.36035, + "grad_norm": 1.386266504967763, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 36035 + }, + { + "epoch": 0.36036, + "grad_norm": 1.6489394323237767, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 36036 + }, + { + "epoch": 0.36037, + "grad_norm": 1.2288834594791715, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 36037 + }, + { + "epoch": 0.36038, + "grad_norm": 1.579685744015952, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 36038 + }, + { + "epoch": 0.36039, + "grad_norm": 1.3420671473678631, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 36039 + }, + { + "epoch": 0.3604, + "grad_norm": 1.5557540045131004, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 36040 + }, + { + "epoch": 0.36041, + "grad_norm": 1.388651850309693, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36041 + }, + { + "epoch": 0.36042, + "grad_norm": 1.5026605197266203, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 36042 + }, + { + "epoch": 0.36043, + "grad_norm": 1.4147520111247744, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 36043 + }, + { + "epoch": 0.36044, + "grad_norm": 1.4361769081729385, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 36044 + }, + { + "epoch": 0.36045, + "grad_norm": 1.543199874044095, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 36045 + }, + { + "epoch": 0.36046, + "grad_norm": 1.4196628816958217, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 36046 + }, + { + "epoch": 0.36047, + "grad_norm": 1.9507055035017966, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 36047 + }, + { + "epoch": 0.36048, + "grad_norm": 1.1511352360107325, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 36048 + }, + { + "epoch": 0.36049, + "grad_norm": 1.639671850449491, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 36049 + }, + { + "epoch": 0.3605, + "grad_norm": 1.489748106253112, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36050 + }, + { + "epoch": 0.36051, + "grad_norm": 1.5592083914828034, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36051 + }, + { + "epoch": 0.36052, + "grad_norm": 1.511143138302487, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36052 + }, + { + "epoch": 0.36053, + "grad_norm": 1.4525020454334474, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 36053 + }, + { + "epoch": 0.36054, + "grad_norm": 1.7468315147196343, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 36054 + }, + { + "epoch": 0.36055, + "grad_norm": 1.2572451191137244, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36055 + }, + { + "epoch": 0.36056, + "grad_norm": 1.613882140455092, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 36056 + }, + { + "epoch": 0.36057, + "grad_norm": 1.2500424933366514, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36057 + }, + { + "epoch": 0.36058, + "grad_norm": 1.6454362729799727, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 36058 + }, + { + "epoch": 0.36059, + "grad_norm": 1.4301854347563132, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 36059 + }, + { + "epoch": 0.3606, + "grad_norm": 1.8901282735966345, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 36060 + }, + { + "epoch": 0.36061, + "grad_norm": 1.726204195990012, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 36061 + }, + { + "epoch": 0.36062, + "grad_norm": 1.6188416677835586, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36062 + }, + { + "epoch": 0.36063, + "grad_norm": 1.3536681324725353, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 36063 + }, + { + "epoch": 0.36064, + "grad_norm": 1.5152379403270433, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36064 + }, + { + "epoch": 0.36065, + "grad_norm": 1.463144964930123, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36065 + }, + { + "epoch": 0.36066, + "grad_norm": 1.579828994654416, + "learning_rate": 0.003, + "loss": 4.017, + "step": 36066 + }, + { + "epoch": 0.36067, + "grad_norm": 1.5917090909497786, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 36067 + }, + { + "epoch": 0.36068, + "grad_norm": 1.5344135543860078, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36068 + }, + { + "epoch": 0.36069, + "grad_norm": 1.5044447757952197, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36069 + }, + { + "epoch": 0.3607, + "grad_norm": 1.2466548384554488, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 36070 + }, + { + "epoch": 0.36071, + "grad_norm": 1.8605423241719103, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36071 + }, + { + "epoch": 0.36072, + "grad_norm": 1.1607953808076559, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 36072 + }, + { + "epoch": 0.36073, + "grad_norm": 1.6183167165875596, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 36073 + }, + { + "epoch": 0.36074, + "grad_norm": 1.345917182616513, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36074 + }, + { + "epoch": 0.36075, + "grad_norm": 1.773745507251321, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 36075 + }, + { + "epoch": 0.36076, + "grad_norm": 1.4832174026901932, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36076 + }, + { + "epoch": 0.36077, + "grad_norm": 1.6314584418492437, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 36077 + }, + { + "epoch": 0.36078, + "grad_norm": 1.4162629865361576, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 36078 + }, + { + "epoch": 0.36079, + "grad_norm": 1.5850651016093422, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 36079 + }, + { + "epoch": 0.3608, + "grad_norm": 1.3068161693047395, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 36080 + }, + { + "epoch": 0.36081, + "grad_norm": 1.7889174685292764, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 36081 + }, + { + "epoch": 0.36082, + "grad_norm": 1.1991210645523631, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 36082 + }, + { + "epoch": 0.36083, + "grad_norm": 1.6131909355719967, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 36083 + }, + { + "epoch": 0.36084, + "grad_norm": 1.5217723553887772, + "learning_rate": 0.003, + "loss": 4.024, + "step": 36084 + }, + { + "epoch": 0.36085, + "grad_norm": 1.6928508525817618, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36085 + }, + { + "epoch": 0.36086, + "grad_norm": 1.5099889538394435, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 36086 + }, + { + "epoch": 0.36087, + "grad_norm": 1.5529075279569713, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36087 + }, + { + "epoch": 0.36088, + "grad_norm": 1.445820072402498, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 36088 + }, + { + "epoch": 0.36089, + "grad_norm": 1.4236892053587107, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 36089 + }, + { + "epoch": 0.3609, + "grad_norm": 1.340949709423032, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 36090 + }, + { + "epoch": 0.36091, + "grad_norm": 1.5277235244616647, + "learning_rate": 0.003, + "loss": 4.035, + "step": 36091 + }, + { + "epoch": 0.36092, + "grad_norm": 1.5830392921811567, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 36092 + }, + { + "epoch": 0.36093, + "grad_norm": 1.7088462232019432, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36093 + }, + { + "epoch": 0.36094, + "grad_norm": 1.5045372152967529, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 36094 + }, + { + "epoch": 0.36095, + "grad_norm": 1.6525701535745336, + "learning_rate": 0.003, + "loss": 4.04, + "step": 36095 + }, + { + "epoch": 0.36096, + "grad_norm": 1.4280199938400173, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36096 + }, + { + "epoch": 0.36097, + "grad_norm": 1.414388960088719, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 36097 + }, + { + "epoch": 0.36098, + "grad_norm": 1.4997672199053973, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 36098 + }, + { + "epoch": 0.36099, + "grad_norm": 1.3253564769451216, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36099 + }, + { + "epoch": 0.361, + "grad_norm": 1.4396963035923365, + "learning_rate": 0.003, + "loss": 4.036, + "step": 36100 + }, + { + "epoch": 0.36101, + "grad_norm": 1.2866524630816203, + "learning_rate": 0.003, + "loss": 4.036, + "step": 36101 + }, + { + "epoch": 0.36102, + "grad_norm": 1.6493040325141803, + "learning_rate": 0.003, + "loss": 4.054, + "step": 36102 + }, + { + "epoch": 0.36103, + "grad_norm": 1.4538233793033677, + "learning_rate": 0.003, + "loss": 4.058, + "step": 36103 + }, + { + "epoch": 0.36104, + "grad_norm": 1.4524066311480794, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36104 + }, + { + "epoch": 0.36105, + "grad_norm": 1.4904032739241038, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36105 + }, + { + "epoch": 0.36106, + "grad_norm": 1.8175849545095057, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 36106 + }, + { + "epoch": 0.36107, + "grad_norm": 1.4871906988256747, + "learning_rate": 0.003, + "loss": 4.016, + "step": 36107 + }, + { + "epoch": 0.36108, + "grad_norm": 1.661078132615046, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 36108 + }, + { + "epoch": 0.36109, + "grad_norm": 1.528140449328241, + "learning_rate": 0.003, + "loss": 4.008, + "step": 36109 + }, + { + "epoch": 0.3611, + "grad_norm": 1.7213386008047564, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 36110 + }, + { + "epoch": 0.36111, + "grad_norm": 1.3862315284275983, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 36111 + }, + { + "epoch": 0.36112, + "grad_norm": 1.4592184925492224, + "learning_rate": 0.003, + "loss": 4.028, + "step": 36112 + }, + { + "epoch": 0.36113, + "grad_norm": 1.3619072769343967, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 36113 + }, + { + "epoch": 0.36114, + "grad_norm": 1.7549909421640502, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36114 + }, + { + "epoch": 0.36115, + "grad_norm": 1.3318941967140578, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 36115 + }, + { + "epoch": 0.36116, + "grad_norm": 1.4860931290530974, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 36116 + }, + { + "epoch": 0.36117, + "grad_norm": 1.7573571167570403, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 36117 + }, + { + "epoch": 0.36118, + "grad_norm": 1.421871030718518, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36118 + }, + { + "epoch": 0.36119, + "grad_norm": 1.3678486770287226, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 36119 + }, + { + "epoch": 0.3612, + "grad_norm": 1.5082578067585333, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 36120 + }, + { + "epoch": 0.36121, + "grad_norm": 1.4412584618591815, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 36121 + }, + { + "epoch": 0.36122, + "grad_norm": 1.376107250053677, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 36122 + }, + { + "epoch": 0.36123, + "grad_norm": 1.7078562105490365, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36123 + }, + { + "epoch": 0.36124, + "grad_norm": 1.4845366751170648, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 36124 + }, + { + "epoch": 0.36125, + "grad_norm": 1.5786351803689882, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 36125 + }, + { + "epoch": 0.36126, + "grad_norm": 1.554785418453854, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 36126 + }, + { + "epoch": 0.36127, + "grad_norm": 1.664258134335546, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36127 + }, + { + "epoch": 0.36128, + "grad_norm": 1.4443505380168729, + "learning_rate": 0.003, + "loss": 4.064, + "step": 36128 + }, + { + "epoch": 0.36129, + "grad_norm": 1.431124524574865, + "learning_rate": 0.003, + "loss": 4.026, + "step": 36129 + }, + { + "epoch": 0.3613, + "grad_norm": 1.8893942713591554, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36130 + }, + { + "epoch": 0.36131, + "grad_norm": 1.4105661107590532, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36131 + }, + { + "epoch": 0.36132, + "grad_norm": 1.7888917449506605, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 36132 + }, + { + "epoch": 0.36133, + "grad_norm": 1.198999732963673, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 36133 + }, + { + "epoch": 0.36134, + "grad_norm": 1.7427555983364076, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 36134 + }, + { + "epoch": 0.36135, + "grad_norm": 1.153547065618868, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 36135 + }, + { + "epoch": 0.36136, + "grad_norm": 2.002413607289644, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36136 + }, + { + "epoch": 0.36137, + "grad_norm": 1.5160948479303717, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 36137 + }, + { + "epoch": 0.36138, + "grad_norm": 1.3392172233483954, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36138 + }, + { + "epoch": 0.36139, + "grad_norm": 1.8132224904862257, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 36139 + }, + { + "epoch": 0.3614, + "grad_norm": 1.327643672560484, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 36140 + }, + { + "epoch": 0.36141, + "grad_norm": 1.4830483179194465, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 36141 + }, + { + "epoch": 0.36142, + "grad_norm": 1.5182104192725248, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 36142 + }, + { + "epoch": 0.36143, + "grad_norm": 1.2675300792365398, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36143 + }, + { + "epoch": 0.36144, + "grad_norm": 1.7052022990545246, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36144 + }, + { + "epoch": 0.36145, + "grad_norm": 1.3576153576518106, + "learning_rate": 0.003, + "loss": 4.034, + "step": 36145 + }, + { + "epoch": 0.36146, + "grad_norm": 1.523262438185443, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 36146 + }, + { + "epoch": 0.36147, + "grad_norm": 1.1532114564520892, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 36147 + }, + { + "epoch": 0.36148, + "grad_norm": 1.5311799564252284, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 36148 + }, + { + "epoch": 0.36149, + "grad_norm": 1.5563899520286477, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 36149 + }, + { + "epoch": 0.3615, + "grad_norm": 1.4011397855295098, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 36150 + }, + { + "epoch": 0.36151, + "grad_norm": 1.5779484204227068, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 36151 + }, + { + "epoch": 0.36152, + "grad_norm": 1.318111287309369, + "learning_rate": 0.003, + "loss": 4.027, + "step": 36152 + }, + { + "epoch": 0.36153, + "grad_norm": 1.5340055547521863, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36153 + }, + { + "epoch": 0.36154, + "grad_norm": 1.2395759992003315, + "learning_rate": 0.003, + "loss": 4.056, + "step": 36154 + }, + { + "epoch": 0.36155, + "grad_norm": 1.7244818203304477, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 36155 + }, + { + "epoch": 0.36156, + "grad_norm": 1.7785536602246859, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 36156 + }, + { + "epoch": 0.36157, + "grad_norm": 1.5427285040869498, + "learning_rate": 0.003, + "loss": 4.081, + "step": 36157 + }, + { + "epoch": 0.36158, + "grad_norm": 2.2021855130315293, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 36158 + }, + { + "epoch": 0.36159, + "grad_norm": 1.370139401635361, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36159 + }, + { + "epoch": 0.3616, + "grad_norm": 1.4487984604855728, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36160 + }, + { + "epoch": 0.36161, + "grad_norm": 1.4988594974285672, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 36161 + }, + { + "epoch": 0.36162, + "grad_norm": 1.6625714000941476, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36162 + }, + { + "epoch": 0.36163, + "grad_norm": 1.6238959287919865, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 36163 + }, + { + "epoch": 0.36164, + "grad_norm": 1.5673907236987117, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 36164 + }, + { + "epoch": 0.36165, + "grad_norm": 1.5046656422692712, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 36165 + }, + { + "epoch": 0.36166, + "grad_norm": 1.3513644830650722, + "learning_rate": 0.003, + "loss": 4.056, + "step": 36166 + }, + { + "epoch": 0.36167, + "grad_norm": 1.5767634493074003, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 36167 + }, + { + "epoch": 0.36168, + "grad_norm": 1.6027490818591665, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 36168 + }, + { + "epoch": 0.36169, + "grad_norm": 1.4557799196399832, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36169 + }, + { + "epoch": 0.3617, + "grad_norm": 1.4657044781166841, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36170 + }, + { + "epoch": 0.36171, + "grad_norm": 1.3736491085028972, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 36171 + }, + { + "epoch": 0.36172, + "grad_norm": 1.504144011971518, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 36172 + }, + { + "epoch": 0.36173, + "grad_norm": 1.5367621659131134, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 36173 + }, + { + "epoch": 0.36174, + "grad_norm": 1.4803892144437065, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 36174 + }, + { + "epoch": 0.36175, + "grad_norm": 1.6490487065080235, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 36175 + }, + { + "epoch": 0.36176, + "grad_norm": 1.3015973771887954, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 36176 + }, + { + "epoch": 0.36177, + "grad_norm": 1.6844386823904465, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 36177 + }, + { + "epoch": 0.36178, + "grad_norm": 1.3988221571292587, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 36178 + }, + { + "epoch": 0.36179, + "grad_norm": 1.5605422321782598, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 36179 + }, + { + "epoch": 0.3618, + "grad_norm": 1.4520192506553122, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 36180 + }, + { + "epoch": 0.36181, + "grad_norm": 1.5344186848129742, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 36181 + }, + { + "epoch": 0.36182, + "grad_norm": 1.312959728529919, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36182 + }, + { + "epoch": 0.36183, + "grad_norm": 1.5402807349682703, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36183 + }, + { + "epoch": 0.36184, + "grad_norm": 1.3664821622134515, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36184 + }, + { + "epoch": 0.36185, + "grad_norm": 1.4673532441372852, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 36185 + }, + { + "epoch": 0.36186, + "grad_norm": 1.5638221252661637, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36186 + }, + { + "epoch": 0.36187, + "grad_norm": 1.036774093489755, + "learning_rate": 0.003, + "loss": 4.063, + "step": 36187 + }, + { + "epoch": 0.36188, + "grad_norm": 1.8512570935805843, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36188 + }, + { + "epoch": 0.36189, + "grad_norm": 1.3494484164589897, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 36189 + }, + { + "epoch": 0.3619, + "grad_norm": 2.064811386990466, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 36190 + }, + { + "epoch": 0.36191, + "grad_norm": 1.630723224684275, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36191 + }, + { + "epoch": 0.36192, + "grad_norm": 1.5340394522275258, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 36192 + }, + { + "epoch": 0.36193, + "grad_norm": 1.6339965407098396, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 36193 + }, + { + "epoch": 0.36194, + "grad_norm": 1.2318289623104852, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 36194 + }, + { + "epoch": 0.36195, + "grad_norm": 1.8087775872322394, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 36195 + }, + { + "epoch": 0.36196, + "grad_norm": 1.2858258081301708, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 36196 + }, + { + "epoch": 0.36197, + "grad_norm": 1.7453493138676042, + "learning_rate": 0.003, + "loss": 4.067, + "step": 36197 + }, + { + "epoch": 0.36198, + "grad_norm": 1.3258170293987768, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 36198 + }, + { + "epoch": 0.36199, + "grad_norm": 1.6135575515970273, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 36199 + }, + { + "epoch": 0.362, + "grad_norm": 1.4467662618402397, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 36200 + }, + { + "epoch": 0.36201, + "grad_norm": 1.5883071013992842, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 36201 + }, + { + "epoch": 0.36202, + "grad_norm": 1.5763458281601086, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36202 + }, + { + "epoch": 0.36203, + "grad_norm": 1.7240598764950434, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 36203 + }, + { + "epoch": 0.36204, + "grad_norm": 1.2142499427088702, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 36204 + }, + { + "epoch": 0.36205, + "grad_norm": 1.5540036476050418, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 36205 + }, + { + "epoch": 0.36206, + "grad_norm": 1.293411674043065, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36206 + }, + { + "epoch": 0.36207, + "grad_norm": 2.139574135157503, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 36207 + }, + { + "epoch": 0.36208, + "grad_norm": 1.1892451299320776, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36208 + }, + { + "epoch": 0.36209, + "grad_norm": 1.7908036331715247, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 36209 + }, + { + "epoch": 0.3621, + "grad_norm": 1.2874748412445434, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36210 + }, + { + "epoch": 0.36211, + "grad_norm": 1.3203032975345894, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 36211 + }, + { + "epoch": 0.36212, + "grad_norm": 1.4272239004400138, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36212 + }, + { + "epoch": 0.36213, + "grad_norm": 1.6221231275818393, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 36213 + }, + { + "epoch": 0.36214, + "grad_norm": 1.5544273949793, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 36214 + }, + { + "epoch": 0.36215, + "grad_norm": 1.2158041980365966, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 36215 + }, + { + "epoch": 0.36216, + "grad_norm": 1.5772542062899726, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 36216 + }, + { + "epoch": 0.36217, + "grad_norm": 1.5812687623378447, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 36217 + }, + { + "epoch": 0.36218, + "grad_norm": 1.797306330513616, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 36218 + }, + { + "epoch": 0.36219, + "grad_norm": 1.693489686087149, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 36219 + }, + { + "epoch": 0.3622, + "grad_norm": 1.6536629821836841, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36220 + }, + { + "epoch": 0.36221, + "grad_norm": 1.502674572571785, + "learning_rate": 0.003, + "loss": 4.078, + "step": 36221 + }, + { + "epoch": 0.36222, + "grad_norm": 1.6485685091248543, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 36222 + }, + { + "epoch": 0.36223, + "grad_norm": 1.52253317840081, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 36223 + }, + { + "epoch": 0.36224, + "grad_norm": 1.5016102890942884, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 36224 + }, + { + "epoch": 0.36225, + "grad_norm": 1.5588009043158835, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 36225 + }, + { + "epoch": 0.36226, + "grad_norm": 1.4968169638365028, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 36226 + }, + { + "epoch": 0.36227, + "grad_norm": 1.298066411203515, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 36227 + }, + { + "epoch": 0.36228, + "grad_norm": 1.6421625478217656, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 36228 + }, + { + "epoch": 0.36229, + "grad_norm": 1.631933311318339, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 36229 + }, + { + "epoch": 0.3623, + "grad_norm": 1.7818327974317907, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 36230 + }, + { + "epoch": 0.36231, + "grad_norm": 1.208471085608682, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36231 + }, + { + "epoch": 0.36232, + "grad_norm": 1.6955108218229866, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 36232 + }, + { + "epoch": 0.36233, + "grad_norm": 1.5269458018112456, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 36233 + }, + { + "epoch": 0.36234, + "grad_norm": 1.558729595492303, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36234 + }, + { + "epoch": 0.36235, + "grad_norm": 1.5464957353240487, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 36235 + }, + { + "epoch": 0.36236, + "grad_norm": 1.2374859868846335, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 36236 + }, + { + "epoch": 0.36237, + "grad_norm": 1.4671451307133507, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36237 + }, + { + "epoch": 0.36238, + "grad_norm": 1.293192079382326, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 36238 + }, + { + "epoch": 0.36239, + "grad_norm": 1.714379553781851, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 36239 + }, + { + "epoch": 0.3624, + "grad_norm": 1.2884958019706576, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36240 + }, + { + "epoch": 0.36241, + "grad_norm": 1.6047434784492352, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 36241 + }, + { + "epoch": 0.36242, + "grad_norm": 1.4068770457921012, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36242 + }, + { + "epoch": 0.36243, + "grad_norm": 1.8108973796468892, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 36243 + }, + { + "epoch": 0.36244, + "grad_norm": 1.528277541233194, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36244 + }, + { + "epoch": 0.36245, + "grad_norm": 1.6960374749344425, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 36245 + }, + { + "epoch": 0.36246, + "grad_norm": 1.3721356105440543, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 36246 + }, + { + "epoch": 0.36247, + "grad_norm": 1.4683229905684827, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36247 + }, + { + "epoch": 0.36248, + "grad_norm": 1.5350445899168568, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 36248 + }, + { + "epoch": 0.36249, + "grad_norm": 1.3761740024576536, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 36249 + }, + { + "epoch": 0.3625, + "grad_norm": 1.7245748001814956, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 36250 + }, + { + "epoch": 0.36251, + "grad_norm": 1.343481146332567, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 36251 + }, + { + "epoch": 0.36252, + "grad_norm": 1.714228441277517, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36252 + }, + { + "epoch": 0.36253, + "grad_norm": 1.2703424180171823, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 36253 + }, + { + "epoch": 0.36254, + "grad_norm": 1.711309940150909, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36254 + }, + { + "epoch": 0.36255, + "grad_norm": 1.4109623758481427, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 36255 + }, + { + "epoch": 0.36256, + "grad_norm": 1.55504230833735, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 36256 + }, + { + "epoch": 0.36257, + "grad_norm": 1.2963283748131191, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 36257 + }, + { + "epoch": 0.36258, + "grad_norm": 1.5734776789205203, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 36258 + }, + { + "epoch": 0.36259, + "grad_norm": 1.380179264943417, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 36259 + }, + { + "epoch": 0.3626, + "grad_norm": 1.4110257431488136, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 36260 + }, + { + "epoch": 0.36261, + "grad_norm": 1.5019525775408535, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36261 + }, + { + "epoch": 0.36262, + "grad_norm": 1.452464372225178, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 36262 + }, + { + "epoch": 0.36263, + "grad_norm": 1.5138812446441288, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36263 + }, + { + "epoch": 0.36264, + "grad_norm": 1.3741757477110124, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 36264 + }, + { + "epoch": 0.36265, + "grad_norm": 1.3756406788457163, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 36265 + }, + { + "epoch": 0.36266, + "grad_norm": 1.5342619048980994, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 36266 + }, + { + "epoch": 0.36267, + "grad_norm": 1.5356563454883523, + "learning_rate": 0.003, + "loss": 4.047, + "step": 36267 + }, + { + "epoch": 0.36268, + "grad_norm": 1.6713719775394407, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36268 + }, + { + "epoch": 0.36269, + "grad_norm": 1.4623241558925004, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 36269 + }, + { + "epoch": 0.3627, + "grad_norm": 2.4037123207974522, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 36270 + }, + { + "epoch": 0.36271, + "grad_norm": 1.094625409980331, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 36271 + }, + { + "epoch": 0.36272, + "grad_norm": 1.8337448223268389, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 36272 + }, + { + "epoch": 0.36273, + "grad_norm": 1.903670294952772, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 36273 + }, + { + "epoch": 0.36274, + "grad_norm": 2.2195598995157946, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 36274 + }, + { + "epoch": 0.36275, + "grad_norm": 1.5082777322169603, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 36275 + }, + { + "epoch": 0.36276, + "grad_norm": 1.43164002836934, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 36276 + }, + { + "epoch": 0.36277, + "grad_norm": 2.5547281501649155, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 36277 + }, + { + "epoch": 0.36278, + "grad_norm": 2.7002523317472082, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 36278 + }, + { + "epoch": 0.36279, + "grad_norm": 1.578035610073124, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 36279 + }, + { + "epoch": 0.3628, + "grad_norm": 1.4041778983802409, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 36280 + }, + { + "epoch": 0.36281, + "grad_norm": 1.574893329944384, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 36281 + }, + { + "epoch": 0.36282, + "grad_norm": 1.6943985576186706, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 36282 + }, + { + "epoch": 0.36283, + "grad_norm": 1.7538955283992532, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 36283 + }, + { + "epoch": 0.36284, + "grad_norm": 1.8465817982081394, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 36284 + }, + { + "epoch": 0.36285, + "grad_norm": 1.945525215298637, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 36285 + }, + { + "epoch": 0.36286, + "grad_norm": 1.6237326659056484, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 36286 + }, + { + "epoch": 0.36287, + "grad_norm": 1.6874066893369766, + "learning_rate": 0.003, + "loss": 4.076, + "step": 36287 + }, + { + "epoch": 0.36288, + "grad_norm": 1.4778149335858561, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 36288 + }, + { + "epoch": 0.36289, + "grad_norm": 1.4037455801635583, + "learning_rate": 0.003, + "loss": 4.066, + "step": 36289 + }, + { + "epoch": 0.3629, + "grad_norm": 1.3300647832640693, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 36290 + }, + { + "epoch": 0.36291, + "grad_norm": 1.6377469107315377, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 36291 + }, + { + "epoch": 0.36292, + "grad_norm": 1.6425704745505179, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 36292 + }, + { + "epoch": 0.36293, + "grad_norm": 1.9776806806882519, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 36293 + }, + { + "epoch": 0.36294, + "grad_norm": 1.4353587162545545, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36294 + }, + { + "epoch": 0.36295, + "grad_norm": 1.8997781790293204, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 36295 + }, + { + "epoch": 0.36296, + "grad_norm": 1.5464782618268966, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 36296 + }, + { + "epoch": 0.36297, + "grad_norm": 1.7303161879893973, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 36297 + }, + { + "epoch": 0.36298, + "grad_norm": 1.517810086351791, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 36298 + }, + { + "epoch": 0.36299, + "grad_norm": 1.3769021551447138, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 36299 + }, + { + "epoch": 0.363, + "grad_norm": 1.507740734625753, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 36300 + }, + { + "epoch": 0.36301, + "grad_norm": 1.4491844912003675, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 36301 + }, + { + "epoch": 0.36302, + "grad_norm": 1.6596452065645417, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 36302 + }, + { + "epoch": 0.36303, + "grad_norm": 1.8203543258964816, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 36303 + }, + { + "epoch": 0.36304, + "grad_norm": 1.2478478244352922, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 36304 + }, + { + "epoch": 0.36305, + "grad_norm": 1.8273432856492813, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 36305 + }, + { + "epoch": 0.36306, + "grad_norm": 1.8647019498858433, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 36306 + }, + { + "epoch": 0.36307, + "grad_norm": 1.202935194653971, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36307 + }, + { + "epoch": 0.36308, + "grad_norm": 1.6353462855150858, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 36308 + }, + { + "epoch": 0.36309, + "grad_norm": 1.1720423956835373, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 36309 + }, + { + "epoch": 0.3631, + "grad_norm": 1.6257517958338357, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 36310 + }, + { + "epoch": 0.36311, + "grad_norm": 1.5875512471596147, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 36311 + }, + { + "epoch": 0.36312, + "grad_norm": 1.7482472002766487, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 36312 + }, + { + "epoch": 0.36313, + "grad_norm": 1.5510355283353303, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 36313 + }, + { + "epoch": 0.36314, + "grad_norm": 1.505366138330431, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36314 + }, + { + "epoch": 0.36315, + "grad_norm": 1.7989321048873712, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 36315 + }, + { + "epoch": 0.36316, + "grad_norm": 1.2414410512084892, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 36316 + }, + { + "epoch": 0.36317, + "grad_norm": 1.5451254257334708, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36317 + }, + { + "epoch": 0.36318, + "grad_norm": 1.370644549520707, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 36318 + }, + { + "epoch": 0.36319, + "grad_norm": 1.3032787264674361, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 36319 + }, + { + "epoch": 0.3632, + "grad_norm": 1.5542491062773633, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 36320 + }, + { + "epoch": 0.36321, + "grad_norm": 1.429916372338992, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 36321 + }, + { + "epoch": 0.36322, + "grad_norm": 1.791164432797486, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 36322 + }, + { + "epoch": 0.36323, + "grad_norm": 1.4260251426423853, + "learning_rate": 0.003, + "loss": 4.042, + "step": 36323 + }, + { + "epoch": 0.36324, + "grad_norm": 2.0131129089267907, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 36324 + }, + { + "epoch": 0.36325, + "grad_norm": 1.3492302910739842, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 36325 + }, + { + "epoch": 0.36326, + "grad_norm": 1.3646442571342763, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 36326 + }, + { + "epoch": 0.36327, + "grad_norm": 1.6423624224483555, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36327 + }, + { + "epoch": 0.36328, + "grad_norm": 1.5118252103449898, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 36328 + }, + { + "epoch": 0.36329, + "grad_norm": 1.477301528742465, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36329 + }, + { + "epoch": 0.3633, + "grad_norm": 1.5874184356176904, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36330 + }, + { + "epoch": 0.36331, + "grad_norm": 1.5105389919401704, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36331 + }, + { + "epoch": 0.36332, + "grad_norm": 1.2566545337225097, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 36332 + }, + { + "epoch": 0.36333, + "grad_norm": 1.5397747461134959, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 36333 + }, + { + "epoch": 0.36334, + "grad_norm": 1.4494500184463794, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 36334 + }, + { + "epoch": 0.36335, + "grad_norm": 1.6481186654314015, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 36335 + }, + { + "epoch": 0.36336, + "grad_norm": 1.3375783992580712, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 36336 + }, + { + "epoch": 0.36337, + "grad_norm": 1.4108842125688672, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 36337 + }, + { + "epoch": 0.36338, + "grad_norm": 1.2975611556555773, + "learning_rate": 0.003, + "loss": 4.069, + "step": 36338 + }, + { + "epoch": 0.36339, + "grad_norm": 1.730703390392309, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36339 + }, + { + "epoch": 0.3634, + "grad_norm": 1.5833950142706528, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36340 + }, + { + "epoch": 0.36341, + "grad_norm": 1.7479154965778674, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 36341 + }, + { + "epoch": 0.36342, + "grad_norm": 1.4998061541911596, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 36342 + }, + { + "epoch": 0.36343, + "grad_norm": 1.271194558692791, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 36343 + }, + { + "epoch": 0.36344, + "grad_norm": 1.3846096368492897, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 36344 + }, + { + "epoch": 0.36345, + "grad_norm": 1.5915211490288361, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36345 + }, + { + "epoch": 0.36346, + "grad_norm": 1.5357597803272947, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 36346 + }, + { + "epoch": 0.36347, + "grad_norm": 1.2476923213371383, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 36347 + }, + { + "epoch": 0.36348, + "grad_norm": 1.4726272588236546, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 36348 + }, + { + "epoch": 0.36349, + "grad_norm": 1.5604937118502173, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 36349 + }, + { + "epoch": 0.3635, + "grad_norm": 1.3390330431119601, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 36350 + }, + { + "epoch": 0.36351, + "grad_norm": 1.6750303529063428, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36351 + }, + { + "epoch": 0.36352, + "grad_norm": 1.362500603825235, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 36352 + }, + { + "epoch": 0.36353, + "grad_norm": 1.8095635805070889, + "learning_rate": 0.003, + "loss": 4.042, + "step": 36353 + }, + { + "epoch": 0.36354, + "grad_norm": 1.7467898130414055, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 36354 + }, + { + "epoch": 0.36355, + "grad_norm": 1.8849587811478892, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 36355 + }, + { + "epoch": 0.36356, + "grad_norm": 1.3924012152264518, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 36356 + }, + { + "epoch": 0.36357, + "grad_norm": 1.731696051403879, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 36357 + }, + { + "epoch": 0.36358, + "grad_norm": 1.2160835405805026, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36358 + }, + { + "epoch": 0.36359, + "grad_norm": 1.7660139086104603, + "learning_rate": 0.003, + "loss": 4.037, + "step": 36359 + }, + { + "epoch": 0.3636, + "grad_norm": 1.3796678652204197, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 36360 + }, + { + "epoch": 0.36361, + "grad_norm": 1.7149331729260744, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 36361 + }, + { + "epoch": 0.36362, + "grad_norm": 1.2407538804228544, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 36362 + }, + { + "epoch": 0.36363, + "grad_norm": 1.9221063043927662, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 36363 + }, + { + "epoch": 0.36364, + "grad_norm": 1.5136304900024942, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 36364 + }, + { + "epoch": 0.36365, + "grad_norm": 1.5003399439524099, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36365 + }, + { + "epoch": 0.36366, + "grad_norm": 1.4742568875908222, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36366 + }, + { + "epoch": 0.36367, + "grad_norm": 1.8037165711219296, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 36367 + }, + { + "epoch": 0.36368, + "grad_norm": 1.4295325199434867, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36368 + }, + { + "epoch": 0.36369, + "grad_norm": 1.4445198509755097, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 36369 + }, + { + "epoch": 0.3637, + "grad_norm": 1.7821691043551118, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36370 + }, + { + "epoch": 0.36371, + "grad_norm": 1.409088821879396, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 36371 + }, + { + "epoch": 0.36372, + "grad_norm": 1.2515138312632335, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 36372 + }, + { + "epoch": 0.36373, + "grad_norm": 1.3634120482561203, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 36373 + }, + { + "epoch": 0.36374, + "grad_norm": 1.6277635746483563, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36374 + }, + { + "epoch": 0.36375, + "grad_norm": 1.2726278591491655, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36375 + }, + { + "epoch": 0.36376, + "grad_norm": 1.5497123402343689, + "learning_rate": 0.003, + "loss": 4.018, + "step": 36376 + }, + { + "epoch": 0.36377, + "grad_norm": 1.6088742206065356, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36377 + }, + { + "epoch": 0.36378, + "grad_norm": 1.4714120104213064, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36378 + }, + { + "epoch": 0.36379, + "grad_norm": 1.750797717357481, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36379 + }, + { + "epoch": 0.3638, + "grad_norm": 1.4168493332125502, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 36380 + }, + { + "epoch": 0.36381, + "grad_norm": 1.6695259885358766, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 36381 + }, + { + "epoch": 0.36382, + "grad_norm": 1.513543890153754, + "learning_rate": 0.003, + "loss": 4.023, + "step": 36382 + }, + { + "epoch": 0.36383, + "grad_norm": 1.4213005573400102, + "learning_rate": 0.003, + "loss": 4.051, + "step": 36383 + }, + { + "epoch": 0.36384, + "grad_norm": 1.5323070317667975, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 36384 + }, + { + "epoch": 0.36385, + "grad_norm": 1.2316251307580313, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 36385 + }, + { + "epoch": 0.36386, + "grad_norm": 1.5030123805387081, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 36386 + }, + { + "epoch": 0.36387, + "grad_norm": 1.2628414007433313, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 36387 + }, + { + "epoch": 0.36388, + "grad_norm": 1.516436801666122, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36388 + }, + { + "epoch": 0.36389, + "grad_norm": 1.2867000458157378, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36389 + }, + { + "epoch": 0.3639, + "grad_norm": 1.7756220177370152, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 36390 + }, + { + "epoch": 0.36391, + "grad_norm": 1.311637719790229, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36391 + }, + { + "epoch": 0.36392, + "grad_norm": 1.8086007006798963, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 36392 + }, + { + "epoch": 0.36393, + "grad_norm": 1.2608271260739228, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 36393 + }, + { + "epoch": 0.36394, + "grad_norm": 1.5587998230603624, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36394 + }, + { + "epoch": 0.36395, + "grad_norm": 1.5177040033536746, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 36395 + }, + { + "epoch": 0.36396, + "grad_norm": 1.7070346687617188, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 36396 + }, + { + "epoch": 0.36397, + "grad_norm": 2.0583795410475694, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 36397 + }, + { + "epoch": 0.36398, + "grad_norm": 1.3335599997615837, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 36398 + }, + { + "epoch": 0.36399, + "grad_norm": 1.586161005069523, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 36399 + }, + { + "epoch": 0.364, + "grad_norm": 1.5610403690763603, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36400 + }, + { + "epoch": 0.36401, + "grad_norm": 1.4050304807250407, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 36401 + }, + { + "epoch": 0.36402, + "grad_norm": 1.4272847330978065, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36402 + }, + { + "epoch": 0.36403, + "grad_norm": 1.5788994528947817, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36403 + }, + { + "epoch": 0.36404, + "grad_norm": 1.444553191966854, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 36404 + }, + { + "epoch": 0.36405, + "grad_norm": 1.7573406751758829, + "learning_rate": 0.003, + "loss": 4.068, + "step": 36405 + }, + { + "epoch": 0.36406, + "grad_norm": 1.2991806465356097, + "learning_rate": 0.003, + "loss": 4.008, + "step": 36406 + }, + { + "epoch": 0.36407, + "grad_norm": 1.4607571458958541, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 36407 + }, + { + "epoch": 0.36408, + "grad_norm": 1.2486279329272965, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 36408 + }, + { + "epoch": 0.36409, + "grad_norm": 1.6931421859351108, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36409 + }, + { + "epoch": 0.3641, + "grad_norm": 1.2921694142119582, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36410 + }, + { + "epoch": 0.36411, + "grad_norm": 1.6246473832947166, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36411 + }, + { + "epoch": 0.36412, + "grad_norm": 1.27219072815467, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 36412 + }, + { + "epoch": 0.36413, + "grad_norm": 1.674130357348134, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 36413 + }, + { + "epoch": 0.36414, + "grad_norm": 1.869932500535831, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36414 + }, + { + "epoch": 0.36415, + "grad_norm": 1.2459067782727595, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 36415 + }, + { + "epoch": 0.36416, + "grad_norm": 1.7318058738540034, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 36416 + }, + { + "epoch": 0.36417, + "grad_norm": 1.3507400834133212, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 36417 + }, + { + "epoch": 0.36418, + "grad_norm": 1.5805972158979027, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36418 + }, + { + "epoch": 0.36419, + "grad_norm": 1.5827265455647626, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 36419 + }, + { + "epoch": 0.3642, + "grad_norm": 1.2719875972152, + "learning_rate": 0.003, + "loss": 4.057, + "step": 36420 + }, + { + "epoch": 0.36421, + "grad_norm": 1.5178926956395897, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 36421 + }, + { + "epoch": 0.36422, + "grad_norm": 1.350141165775416, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 36422 + }, + { + "epoch": 0.36423, + "grad_norm": 1.7399739731081862, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36423 + }, + { + "epoch": 0.36424, + "grad_norm": 1.4770691565230014, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 36424 + }, + { + "epoch": 0.36425, + "grad_norm": 1.5808918595172434, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 36425 + }, + { + "epoch": 0.36426, + "grad_norm": 1.6359790493892903, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 36426 + }, + { + "epoch": 0.36427, + "grad_norm": 1.269966149139685, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36427 + }, + { + "epoch": 0.36428, + "grad_norm": 1.6697753541668638, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36428 + }, + { + "epoch": 0.36429, + "grad_norm": 1.2445738702470999, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 36429 + }, + { + "epoch": 0.3643, + "grad_norm": 1.7684903263452778, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 36430 + }, + { + "epoch": 0.36431, + "grad_norm": 1.9953909048149316, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 36431 + }, + { + "epoch": 0.36432, + "grad_norm": 1.4458693750109743, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36432 + }, + { + "epoch": 0.36433, + "grad_norm": 1.517542880108081, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 36433 + }, + { + "epoch": 0.36434, + "grad_norm": 1.6947334742766322, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 36434 + }, + { + "epoch": 0.36435, + "grad_norm": 1.2434569977703949, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36435 + }, + { + "epoch": 0.36436, + "grad_norm": 1.3996444549169795, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 36436 + }, + { + "epoch": 0.36437, + "grad_norm": 1.3696054782883282, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 36437 + }, + { + "epoch": 0.36438, + "grad_norm": 1.6477946866735609, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 36438 + }, + { + "epoch": 0.36439, + "grad_norm": 1.4354836306362473, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 36439 + }, + { + "epoch": 0.3644, + "grad_norm": 1.6853085231788842, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 36440 + }, + { + "epoch": 0.36441, + "grad_norm": 1.4588491496435085, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 36441 + }, + { + "epoch": 0.36442, + "grad_norm": 1.5426947445991055, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 36442 + }, + { + "epoch": 0.36443, + "grad_norm": 1.2775259012942852, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 36443 + }, + { + "epoch": 0.36444, + "grad_norm": 1.6811014980691554, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 36444 + }, + { + "epoch": 0.36445, + "grad_norm": 1.3943586018126346, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 36445 + }, + { + "epoch": 0.36446, + "grad_norm": 1.570062850786294, + "learning_rate": 0.003, + "loss": 4.057, + "step": 36446 + }, + { + "epoch": 0.36447, + "grad_norm": 1.4011300313662458, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 36447 + }, + { + "epoch": 0.36448, + "grad_norm": 1.3210413697009946, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 36448 + }, + { + "epoch": 0.36449, + "grad_norm": 1.6499093974921508, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 36449 + }, + { + "epoch": 0.3645, + "grad_norm": 1.2477095448042448, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 36450 + }, + { + "epoch": 0.36451, + "grad_norm": 1.6724251716919818, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36451 + }, + { + "epoch": 0.36452, + "grad_norm": 1.4527075936912979, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 36452 + }, + { + "epoch": 0.36453, + "grad_norm": 1.5593145500464007, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 36453 + }, + { + "epoch": 0.36454, + "grad_norm": 1.7212985218027266, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 36454 + }, + { + "epoch": 0.36455, + "grad_norm": 1.3850936753803924, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 36455 + }, + { + "epoch": 0.36456, + "grad_norm": 1.6215503838690362, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36456 + }, + { + "epoch": 0.36457, + "grad_norm": 1.3793643912979534, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36457 + }, + { + "epoch": 0.36458, + "grad_norm": 1.235372958490554, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 36458 + }, + { + "epoch": 0.36459, + "grad_norm": 1.6167540867331218, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 36459 + }, + { + "epoch": 0.3646, + "grad_norm": 1.4280737133712875, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 36460 + }, + { + "epoch": 0.36461, + "grad_norm": 1.861334061798417, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36461 + }, + { + "epoch": 0.36462, + "grad_norm": 1.148880335206981, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 36462 + }, + { + "epoch": 0.36463, + "grad_norm": 1.9339478338829212, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 36463 + }, + { + "epoch": 0.36464, + "grad_norm": 1.2857508479458808, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 36464 + }, + { + "epoch": 0.36465, + "grad_norm": 1.7193163392449151, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 36465 + }, + { + "epoch": 0.36466, + "grad_norm": 1.53681333252708, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36466 + }, + { + "epoch": 0.36467, + "grad_norm": 1.479303325263685, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36467 + }, + { + "epoch": 0.36468, + "grad_norm": 2.0632504751076675, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 36468 + }, + { + "epoch": 0.36469, + "grad_norm": 1.3089384938050195, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36469 + }, + { + "epoch": 0.3647, + "grad_norm": 1.474914235401585, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 36470 + }, + { + "epoch": 0.36471, + "grad_norm": 1.7922869376414885, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 36471 + }, + { + "epoch": 0.36472, + "grad_norm": 1.6582902379070674, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 36472 + }, + { + "epoch": 0.36473, + "grad_norm": 1.378172533756844, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 36473 + }, + { + "epoch": 0.36474, + "grad_norm": 1.4009564875233855, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 36474 + }, + { + "epoch": 0.36475, + "grad_norm": 1.5634202940326178, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36475 + }, + { + "epoch": 0.36476, + "grad_norm": 1.4151750481916767, + "learning_rate": 0.003, + "loss": 4.036, + "step": 36476 + }, + { + "epoch": 0.36477, + "grad_norm": 1.565108247269881, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36477 + }, + { + "epoch": 0.36478, + "grad_norm": 1.3930498612910611, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 36478 + }, + { + "epoch": 0.36479, + "grad_norm": 1.5789364816688418, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36479 + }, + { + "epoch": 0.3648, + "grad_norm": 1.103019332592006, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 36480 + }, + { + "epoch": 0.36481, + "grad_norm": 1.6877065161272626, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 36481 + }, + { + "epoch": 0.36482, + "grad_norm": 1.2448974057609188, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 36482 + }, + { + "epoch": 0.36483, + "grad_norm": 1.7242442604907202, + "learning_rate": 0.003, + "loss": 4.027, + "step": 36483 + }, + { + "epoch": 0.36484, + "grad_norm": 1.4335012275284726, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 36484 + }, + { + "epoch": 0.36485, + "grad_norm": 1.779375113757303, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36485 + }, + { + "epoch": 0.36486, + "grad_norm": 1.1630457623620947, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36486 + }, + { + "epoch": 0.36487, + "grad_norm": 1.6142808483850601, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 36487 + }, + { + "epoch": 0.36488, + "grad_norm": 1.5779835332716137, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 36488 + }, + { + "epoch": 0.36489, + "grad_norm": 1.7167792331968341, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 36489 + }, + { + "epoch": 0.3649, + "grad_norm": 1.2537063399732735, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 36490 + }, + { + "epoch": 0.36491, + "grad_norm": 1.6814686987703067, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 36491 + }, + { + "epoch": 0.36492, + "grad_norm": 1.4425069554154324, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 36492 + }, + { + "epoch": 0.36493, + "grad_norm": 1.2109601395163863, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36493 + }, + { + "epoch": 0.36494, + "grad_norm": 1.7372117134754033, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 36494 + }, + { + "epoch": 0.36495, + "grad_norm": 1.2178120029539552, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 36495 + }, + { + "epoch": 0.36496, + "grad_norm": 1.879130461215432, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 36496 + }, + { + "epoch": 0.36497, + "grad_norm": 1.3887458028721458, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 36497 + }, + { + "epoch": 0.36498, + "grad_norm": 1.8118615541116205, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 36498 + }, + { + "epoch": 0.36499, + "grad_norm": 1.7067162716096118, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 36499 + }, + { + "epoch": 0.365, + "grad_norm": 1.3345560149295106, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36500 + }, + { + "epoch": 0.36501, + "grad_norm": 1.5745345349224247, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 36501 + }, + { + "epoch": 0.36502, + "grad_norm": 1.297872791881372, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 36502 + }, + { + "epoch": 0.36503, + "grad_norm": 1.5720370017789111, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 36503 + }, + { + "epoch": 0.36504, + "grad_norm": 1.594859935532613, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 36504 + }, + { + "epoch": 0.36505, + "grad_norm": 1.4549093128623463, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 36505 + }, + { + "epoch": 0.36506, + "grad_norm": 1.6730515574240914, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 36506 + }, + { + "epoch": 0.36507, + "grad_norm": 1.2256532944572434, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 36507 + }, + { + "epoch": 0.36508, + "grad_norm": 1.7411497047413753, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 36508 + }, + { + "epoch": 0.36509, + "grad_norm": 1.1168988786388072, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 36509 + }, + { + "epoch": 0.3651, + "grad_norm": 1.6404274885802812, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 36510 + }, + { + "epoch": 0.36511, + "grad_norm": 1.3469428442938718, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 36511 + }, + { + "epoch": 0.36512, + "grad_norm": 1.607062890170393, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 36512 + }, + { + "epoch": 0.36513, + "grad_norm": 1.4767408703587224, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36513 + }, + { + "epoch": 0.36514, + "grad_norm": 1.4919954307790784, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36514 + }, + { + "epoch": 0.36515, + "grad_norm": 1.908530796466242, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 36515 + }, + { + "epoch": 0.36516, + "grad_norm": 1.3291536890415432, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36516 + }, + { + "epoch": 0.36517, + "grad_norm": 1.7461564389509208, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36517 + }, + { + "epoch": 0.36518, + "grad_norm": 1.267117910168044, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 36518 + }, + { + "epoch": 0.36519, + "grad_norm": 1.9073432888953292, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 36519 + }, + { + "epoch": 0.3652, + "grad_norm": 1.2676344062829854, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 36520 + }, + { + "epoch": 0.36521, + "grad_norm": 1.64701498114414, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 36521 + }, + { + "epoch": 0.36522, + "grad_norm": 1.4489263331333264, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 36522 + }, + { + "epoch": 0.36523, + "grad_norm": 1.3704697021438546, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 36523 + }, + { + "epoch": 0.36524, + "grad_norm": 1.3642808303440548, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 36524 + }, + { + "epoch": 0.36525, + "grad_norm": 1.8067913693458104, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36525 + }, + { + "epoch": 0.36526, + "grad_norm": 1.687205719447072, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 36526 + }, + { + "epoch": 0.36527, + "grad_norm": 1.4126662624771493, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36527 + }, + { + "epoch": 0.36528, + "grad_norm": 1.6725450226674197, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36528 + }, + { + "epoch": 0.36529, + "grad_norm": 1.442013370165052, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 36529 + }, + { + "epoch": 0.3653, + "grad_norm": 1.7043433873680445, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 36530 + }, + { + "epoch": 0.36531, + "grad_norm": 1.1541350621659798, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 36531 + }, + { + "epoch": 0.36532, + "grad_norm": 1.8272203287381514, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 36532 + }, + { + "epoch": 0.36533, + "grad_norm": 1.3615146180510123, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 36533 + }, + { + "epoch": 0.36534, + "grad_norm": 1.3483540537542782, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 36534 + }, + { + "epoch": 0.36535, + "grad_norm": 1.6715182105723896, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36535 + }, + { + "epoch": 0.36536, + "grad_norm": 1.2813617402703634, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36536 + }, + { + "epoch": 0.36537, + "grad_norm": 1.5496915215333946, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 36537 + }, + { + "epoch": 0.36538, + "grad_norm": 1.31718638958454, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 36538 + }, + { + "epoch": 0.36539, + "grad_norm": 1.7904497310871115, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 36539 + }, + { + "epoch": 0.3654, + "grad_norm": 1.384490944861008, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 36540 + }, + { + "epoch": 0.36541, + "grad_norm": 1.431304554414664, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36541 + }, + { + "epoch": 0.36542, + "grad_norm": 1.3676351134262201, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 36542 + }, + { + "epoch": 0.36543, + "grad_norm": 1.4667540012717328, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 36543 + }, + { + "epoch": 0.36544, + "grad_norm": 1.2909433770086052, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36544 + }, + { + "epoch": 0.36545, + "grad_norm": 1.5430666014109489, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36545 + }, + { + "epoch": 0.36546, + "grad_norm": 1.7182986483323384, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36546 + }, + { + "epoch": 0.36547, + "grad_norm": 1.228066640428083, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36547 + }, + { + "epoch": 0.36548, + "grad_norm": 1.4826305940077287, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 36548 + }, + { + "epoch": 0.36549, + "grad_norm": 1.5322596881927781, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36549 + }, + { + "epoch": 0.3655, + "grad_norm": 1.6884301554412704, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36550 + }, + { + "epoch": 0.36551, + "grad_norm": 1.7556617332618105, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 36551 + }, + { + "epoch": 0.36552, + "grad_norm": 1.5317705182958599, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 36552 + }, + { + "epoch": 0.36553, + "grad_norm": 1.798693218628884, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 36553 + }, + { + "epoch": 0.36554, + "grad_norm": 1.2038225728522443, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 36554 + }, + { + "epoch": 0.36555, + "grad_norm": 1.552332571888773, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36555 + }, + { + "epoch": 0.36556, + "grad_norm": 1.3556790828649379, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 36556 + }, + { + "epoch": 0.36557, + "grad_norm": 1.5481049292739688, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36557 + }, + { + "epoch": 0.36558, + "grad_norm": 1.3327915182211307, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36558 + }, + { + "epoch": 0.36559, + "grad_norm": 1.7633372855993437, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 36559 + }, + { + "epoch": 0.3656, + "grad_norm": 1.29809001208167, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 36560 + }, + { + "epoch": 0.36561, + "grad_norm": 1.8020582461519743, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 36561 + }, + { + "epoch": 0.36562, + "grad_norm": 1.501136852536817, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 36562 + }, + { + "epoch": 0.36563, + "grad_norm": 1.690371615530966, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 36563 + }, + { + "epoch": 0.36564, + "grad_norm": 1.4114774269756087, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 36564 + }, + { + "epoch": 0.36565, + "grad_norm": 1.4421339724317384, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36565 + }, + { + "epoch": 0.36566, + "grad_norm": 1.636537535576763, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 36566 + }, + { + "epoch": 0.36567, + "grad_norm": 1.2998722058438716, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36567 + }, + { + "epoch": 0.36568, + "grad_norm": 1.5439699314150652, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 36568 + }, + { + "epoch": 0.36569, + "grad_norm": 1.3912049853792443, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 36569 + }, + { + "epoch": 0.3657, + "grad_norm": 1.427925472185574, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36570 + }, + { + "epoch": 0.36571, + "grad_norm": 1.5595574856016146, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 36571 + }, + { + "epoch": 0.36572, + "grad_norm": 1.4673937265576409, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 36572 + }, + { + "epoch": 0.36573, + "grad_norm": 1.698910265639058, + "learning_rate": 0.003, + "loss": 4.037, + "step": 36573 + }, + { + "epoch": 0.36574, + "grad_norm": 1.3761840860059573, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 36574 + }, + { + "epoch": 0.36575, + "grad_norm": 1.7404466593722498, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 36575 + }, + { + "epoch": 0.36576, + "grad_norm": 1.6263455344521731, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 36576 + }, + { + "epoch": 0.36577, + "grad_norm": 1.7988188366977893, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 36577 + }, + { + "epoch": 0.36578, + "grad_norm": 1.5745419738005344, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36578 + }, + { + "epoch": 0.36579, + "grad_norm": 1.518452865204436, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 36579 + }, + { + "epoch": 0.3658, + "grad_norm": 1.782624064959908, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36580 + }, + { + "epoch": 0.36581, + "grad_norm": 1.2683232499667403, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 36581 + }, + { + "epoch": 0.36582, + "grad_norm": 1.6726916129805476, + "learning_rate": 0.003, + "loss": 4.026, + "step": 36582 + }, + { + "epoch": 0.36583, + "grad_norm": 1.2758125059768624, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 36583 + }, + { + "epoch": 0.36584, + "grad_norm": 1.995898070087574, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 36584 + }, + { + "epoch": 0.36585, + "grad_norm": 1.3637921046554797, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 36585 + }, + { + "epoch": 0.36586, + "grad_norm": 1.462106119926543, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 36586 + }, + { + "epoch": 0.36587, + "grad_norm": 1.652856250045337, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 36587 + }, + { + "epoch": 0.36588, + "grad_norm": 1.4369156275202195, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 36588 + }, + { + "epoch": 0.36589, + "grad_norm": 1.3114601623723645, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 36589 + }, + { + "epoch": 0.3659, + "grad_norm": 1.5554583205420611, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36590 + }, + { + "epoch": 0.36591, + "grad_norm": 1.332627315633685, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 36591 + }, + { + "epoch": 0.36592, + "grad_norm": 1.2986661995918334, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 36592 + }, + { + "epoch": 0.36593, + "grad_norm": 1.5475056613997968, + "learning_rate": 0.003, + "loss": 4.067, + "step": 36593 + }, + { + "epoch": 0.36594, + "grad_norm": 1.3441420808329791, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36594 + }, + { + "epoch": 0.36595, + "grad_norm": 1.5548130409859175, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 36595 + }, + { + "epoch": 0.36596, + "grad_norm": 1.286957993311821, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36596 + }, + { + "epoch": 0.36597, + "grad_norm": 1.7192598643148047, + "learning_rate": 0.003, + "loss": 4.051, + "step": 36597 + }, + { + "epoch": 0.36598, + "grad_norm": 1.3125006434896402, + "learning_rate": 0.003, + "loss": 4.044, + "step": 36598 + }, + { + "epoch": 0.36599, + "grad_norm": 1.476153311391923, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 36599 + }, + { + "epoch": 0.366, + "grad_norm": 1.5568279422464524, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36600 + }, + { + "epoch": 0.36601, + "grad_norm": 1.5663178354263132, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36601 + }, + { + "epoch": 0.36602, + "grad_norm": 1.6438142647055975, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36602 + }, + { + "epoch": 0.36603, + "grad_norm": 1.5118494347141942, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 36603 + }, + { + "epoch": 0.36604, + "grad_norm": 1.3566751165741102, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36604 + }, + { + "epoch": 0.36605, + "grad_norm": 1.6042811885896429, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 36605 + }, + { + "epoch": 0.36606, + "grad_norm": 1.5688009453980267, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 36606 + }, + { + "epoch": 0.36607, + "grad_norm": 1.651646372583363, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36607 + }, + { + "epoch": 0.36608, + "grad_norm": 1.288094493801379, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 36608 + }, + { + "epoch": 0.36609, + "grad_norm": 1.7120071323486354, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 36609 + }, + { + "epoch": 0.3661, + "grad_norm": 1.1061518152002994, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 36610 + }, + { + "epoch": 0.36611, + "grad_norm": 1.802586847103903, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 36611 + }, + { + "epoch": 0.36612, + "grad_norm": 1.2787098695870849, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 36612 + }, + { + "epoch": 0.36613, + "grad_norm": 1.779940329158256, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36613 + }, + { + "epoch": 0.36614, + "grad_norm": 1.253520043406426, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 36614 + }, + { + "epoch": 0.36615, + "grad_norm": 1.7240669290728718, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36615 + }, + { + "epoch": 0.36616, + "grad_norm": 1.6103953925570802, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 36616 + }, + { + "epoch": 0.36617, + "grad_norm": 1.5743462918166014, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 36617 + }, + { + "epoch": 0.36618, + "grad_norm": 1.4860138478910991, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 36618 + }, + { + "epoch": 0.36619, + "grad_norm": 1.6483728637737165, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 36619 + }, + { + "epoch": 0.3662, + "grad_norm": 1.3852518663279203, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36620 + }, + { + "epoch": 0.36621, + "grad_norm": 1.613737679031107, + "learning_rate": 0.003, + "loss": 4.036, + "step": 36621 + }, + { + "epoch": 0.36622, + "grad_norm": 1.4892808841266636, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 36622 + }, + { + "epoch": 0.36623, + "grad_norm": 1.7397954533284807, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 36623 + }, + { + "epoch": 0.36624, + "grad_norm": 1.1715364681485179, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36624 + }, + { + "epoch": 0.36625, + "grad_norm": 1.9145125694964176, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 36625 + }, + { + "epoch": 0.36626, + "grad_norm": 1.2062633766886006, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 36626 + }, + { + "epoch": 0.36627, + "grad_norm": 1.5767851240536, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 36627 + }, + { + "epoch": 0.36628, + "grad_norm": 1.659536026378073, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 36628 + }, + { + "epoch": 0.36629, + "grad_norm": 1.2809788591902806, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 36629 + }, + { + "epoch": 0.3663, + "grad_norm": 1.6613386765533893, + "learning_rate": 0.003, + "loss": 4.024, + "step": 36630 + }, + { + "epoch": 0.36631, + "grad_norm": 1.301835462801165, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36631 + }, + { + "epoch": 0.36632, + "grad_norm": 1.6090499508347218, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 36632 + }, + { + "epoch": 0.36633, + "grad_norm": 1.5167706899278277, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 36633 + }, + { + "epoch": 0.36634, + "grad_norm": 1.3840377817512732, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 36634 + }, + { + "epoch": 0.36635, + "grad_norm": 1.6007434546147232, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 36635 + }, + { + "epoch": 0.36636, + "grad_norm": 1.4797938177247192, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36636 + }, + { + "epoch": 0.36637, + "grad_norm": 1.468858892530777, + "learning_rate": 0.003, + "loss": 4.039, + "step": 36637 + }, + { + "epoch": 0.36638, + "grad_norm": 1.708214495327177, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36638 + }, + { + "epoch": 0.36639, + "grad_norm": 1.5177249672681032, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 36639 + }, + { + "epoch": 0.3664, + "grad_norm": 1.3412224544209952, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 36640 + }, + { + "epoch": 0.36641, + "grad_norm": 1.452474119237781, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 36641 + }, + { + "epoch": 0.36642, + "grad_norm": 1.5905734419869364, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 36642 + }, + { + "epoch": 0.36643, + "grad_norm": 1.511871296890441, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 36643 + }, + { + "epoch": 0.36644, + "grad_norm": 1.407078132563326, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 36644 + }, + { + "epoch": 0.36645, + "grad_norm": 1.5981594432452295, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 36645 + }, + { + "epoch": 0.36646, + "grad_norm": 1.3462768501746036, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 36646 + }, + { + "epoch": 0.36647, + "grad_norm": 1.6748721090253942, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 36647 + }, + { + "epoch": 0.36648, + "grad_norm": 1.2999325262314578, + "learning_rate": 0.003, + "loss": 4.047, + "step": 36648 + }, + { + "epoch": 0.36649, + "grad_norm": 1.5558938135790539, + "learning_rate": 0.003, + "loss": 4.062, + "step": 36649 + }, + { + "epoch": 0.3665, + "grad_norm": 1.1279198566721882, + "learning_rate": 0.003, + "loss": 4.038, + "step": 36650 + }, + { + "epoch": 0.36651, + "grad_norm": 1.5226203184922396, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 36651 + }, + { + "epoch": 0.36652, + "grad_norm": 1.601576759328702, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 36652 + }, + { + "epoch": 0.36653, + "grad_norm": 1.4726294948413972, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 36653 + }, + { + "epoch": 0.36654, + "grad_norm": 1.6905245403352043, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 36654 + }, + { + "epoch": 0.36655, + "grad_norm": 1.277281751036978, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 36655 + }, + { + "epoch": 0.36656, + "grad_norm": 1.6597688490576654, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 36656 + }, + { + "epoch": 0.36657, + "grad_norm": 1.266306279340477, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 36657 + }, + { + "epoch": 0.36658, + "grad_norm": 1.7643700511032596, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 36658 + }, + { + "epoch": 0.36659, + "grad_norm": 1.4956895663291514, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 36659 + }, + { + "epoch": 0.3666, + "grad_norm": 1.5660340242352102, + "learning_rate": 0.003, + "loss": 4.052, + "step": 36660 + }, + { + "epoch": 0.36661, + "grad_norm": 1.7909247613243693, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36661 + }, + { + "epoch": 0.36662, + "grad_norm": 1.4840845222465986, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36662 + }, + { + "epoch": 0.36663, + "grad_norm": 1.9147950672404948, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 36663 + }, + { + "epoch": 0.36664, + "grad_norm": 1.4475557813791171, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 36664 + }, + { + "epoch": 0.36665, + "grad_norm": 1.67299763399459, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 36665 + }, + { + "epoch": 0.36666, + "grad_norm": 1.617515380843916, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 36666 + }, + { + "epoch": 0.36667, + "grad_norm": 1.72642901739503, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 36667 + }, + { + "epoch": 0.36668, + "grad_norm": 1.4170747920621702, + "learning_rate": 0.003, + "loss": 4.084, + "step": 36668 + }, + { + "epoch": 0.36669, + "grad_norm": 1.6009437318445225, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 36669 + }, + { + "epoch": 0.3667, + "grad_norm": 1.1654126058923422, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 36670 + }, + { + "epoch": 0.36671, + "grad_norm": 1.7079821127104948, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 36671 + }, + { + "epoch": 0.36672, + "grad_norm": 1.285205943767477, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 36672 + }, + { + "epoch": 0.36673, + "grad_norm": 1.7291513084373984, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 36673 + }, + { + "epoch": 0.36674, + "grad_norm": 1.1277442022685453, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36674 + }, + { + "epoch": 0.36675, + "grad_norm": 1.8680516155952678, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 36675 + }, + { + "epoch": 0.36676, + "grad_norm": 1.336285900187545, + "learning_rate": 0.003, + "loss": 4.068, + "step": 36676 + }, + { + "epoch": 0.36677, + "grad_norm": 1.5994555770848435, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 36677 + }, + { + "epoch": 0.36678, + "grad_norm": 1.7475656487986808, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 36678 + }, + { + "epoch": 0.36679, + "grad_norm": 1.7283575256592614, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 36679 + }, + { + "epoch": 0.3668, + "grad_norm": 1.3115441806455288, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36680 + }, + { + "epoch": 0.36681, + "grad_norm": 1.9790579029330195, + "learning_rate": 0.003, + "loss": 4.023, + "step": 36681 + }, + { + "epoch": 0.36682, + "grad_norm": 1.3327334825278148, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 36682 + }, + { + "epoch": 0.36683, + "grad_norm": 1.7695375725603164, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 36683 + }, + { + "epoch": 0.36684, + "grad_norm": 1.5590594312622248, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36684 + }, + { + "epoch": 0.36685, + "grad_norm": 1.4119220484743245, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 36685 + }, + { + "epoch": 0.36686, + "grad_norm": 1.29216554066928, + "learning_rate": 0.003, + "loss": 4.05, + "step": 36686 + }, + { + "epoch": 0.36687, + "grad_norm": 1.6609857825455367, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 36687 + }, + { + "epoch": 0.36688, + "grad_norm": 1.3875949431977994, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 36688 + }, + { + "epoch": 0.36689, + "grad_norm": 1.4373773715341733, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 36689 + }, + { + "epoch": 0.3669, + "grad_norm": 1.5579888026978486, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 36690 + }, + { + "epoch": 0.36691, + "grad_norm": 1.3811682849780595, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36691 + }, + { + "epoch": 0.36692, + "grad_norm": 1.580416327135083, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 36692 + }, + { + "epoch": 0.36693, + "grad_norm": 1.1428805309792995, + "learning_rate": 0.003, + "loss": 4.022, + "step": 36693 + }, + { + "epoch": 0.36694, + "grad_norm": 1.5009641933325129, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 36694 + }, + { + "epoch": 0.36695, + "grad_norm": 1.2688705128253077, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36695 + }, + { + "epoch": 0.36696, + "grad_norm": 1.5550457509540279, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 36696 + }, + { + "epoch": 0.36697, + "grad_norm": 1.5257405040828516, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 36697 + }, + { + "epoch": 0.36698, + "grad_norm": 1.562334995114236, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 36698 + }, + { + "epoch": 0.36699, + "grad_norm": 1.8232471268138828, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36699 + }, + { + "epoch": 0.367, + "grad_norm": 2.2777317674363857, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 36700 + }, + { + "epoch": 0.36701, + "grad_norm": 1.117953746179033, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 36701 + }, + { + "epoch": 0.36702, + "grad_norm": 1.6109163874637034, + "learning_rate": 0.003, + "loss": 4.071, + "step": 36702 + }, + { + "epoch": 0.36703, + "grad_norm": 1.610404281206735, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 36703 + }, + { + "epoch": 0.36704, + "grad_norm": 1.5790930050549845, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 36704 + }, + { + "epoch": 0.36705, + "grad_norm": 1.6304246794102955, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 36705 + }, + { + "epoch": 0.36706, + "grad_norm": 2.027123435822263, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 36706 + }, + { + "epoch": 0.36707, + "grad_norm": 0.9258726887629956, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 36707 + }, + { + "epoch": 0.36708, + "grad_norm": 1.4959761323042962, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 36708 + }, + { + "epoch": 0.36709, + "grad_norm": 1.2223469176732877, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 36709 + }, + { + "epoch": 0.3671, + "grad_norm": 1.5713253295816683, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 36710 + }, + { + "epoch": 0.36711, + "grad_norm": 1.3744256055313666, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 36711 + }, + { + "epoch": 0.36712, + "grad_norm": 1.5654424737314854, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 36712 + }, + { + "epoch": 0.36713, + "grad_norm": 1.6544719984859946, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 36713 + }, + { + "epoch": 0.36714, + "grad_norm": 1.4129952394801568, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 36714 + }, + { + "epoch": 0.36715, + "grad_norm": 1.8217072901342848, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 36715 + }, + { + "epoch": 0.36716, + "grad_norm": 1.533081969525413, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 36716 + }, + { + "epoch": 0.36717, + "grad_norm": 1.5392032597780332, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 36717 + }, + { + "epoch": 0.36718, + "grad_norm": 1.5715585202204791, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 36718 + }, + { + "epoch": 0.36719, + "grad_norm": 1.3253314059594052, + "learning_rate": 0.003, + "loss": 4.057, + "step": 36719 + }, + { + "epoch": 0.3672, + "grad_norm": 1.3424465679386028, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36720 + }, + { + "epoch": 0.36721, + "grad_norm": 1.4513805390855299, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 36721 + }, + { + "epoch": 0.36722, + "grad_norm": 1.2507327761009732, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 36722 + }, + { + "epoch": 0.36723, + "grad_norm": 1.3888347722200005, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 36723 + }, + { + "epoch": 0.36724, + "grad_norm": 1.4075265459671964, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 36724 + }, + { + "epoch": 0.36725, + "grad_norm": 1.6540418553371448, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 36725 + }, + { + "epoch": 0.36726, + "grad_norm": 1.9497639562307585, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 36726 + }, + { + "epoch": 0.36727, + "grad_norm": 1.2415501744040165, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36727 + }, + { + "epoch": 0.36728, + "grad_norm": 1.7761245408278077, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36728 + }, + { + "epoch": 0.36729, + "grad_norm": 1.4981637144872073, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 36729 + }, + { + "epoch": 0.3673, + "grad_norm": 1.6621752533573484, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 36730 + }, + { + "epoch": 0.36731, + "grad_norm": 1.3273835224275752, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 36731 + }, + { + "epoch": 0.36732, + "grad_norm": 1.5286999345905983, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36732 + }, + { + "epoch": 0.36733, + "grad_norm": 1.480896717409359, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36733 + }, + { + "epoch": 0.36734, + "grad_norm": 1.3017957172414547, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 36734 + }, + { + "epoch": 0.36735, + "grad_norm": 1.8493488958367656, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 36735 + }, + { + "epoch": 0.36736, + "grad_norm": 1.6809776440212196, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 36736 + }, + { + "epoch": 0.36737, + "grad_norm": 1.3382604185324973, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 36737 + }, + { + "epoch": 0.36738, + "grad_norm": 1.5691038468904348, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 36738 + }, + { + "epoch": 0.36739, + "grad_norm": 1.393689116474566, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 36739 + }, + { + "epoch": 0.3674, + "grad_norm": 1.4791959119601614, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 36740 + }, + { + "epoch": 0.36741, + "grad_norm": 1.564792895805104, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36741 + }, + { + "epoch": 0.36742, + "grad_norm": 1.3642918190503648, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 36742 + }, + { + "epoch": 0.36743, + "grad_norm": 1.8180022152598445, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36743 + }, + { + "epoch": 0.36744, + "grad_norm": 1.3662490022358005, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 36744 + }, + { + "epoch": 0.36745, + "grad_norm": 1.6952646107756162, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 36745 + }, + { + "epoch": 0.36746, + "grad_norm": 1.5548724741202558, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 36746 + }, + { + "epoch": 0.36747, + "grad_norm": 1.5418288759538323, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 36747 + }, + { + "epoch": 0.36748, + "grad_norm": 1.5059813180285695, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36748 + }, + { + "epoch": 0.36749, + "grad_norm": 1.7455408007349322, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36749 + }, + { + "epoch": 0.3675, + "grad_norm": 1.5040756867562755, + "learning_rate": 0.003, + "loss": 4.034, + "step": 36750 + }, + { + "epoch": 0.36751, + "grad_norm": 1.4682846282536426, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 36751 + }, + { + "epoch": 0.36752, + "grad_norm": 1.582389071272787, + "learning_rate": 0.003, + "loss": 4.053, + "step": 36752 + }, + { + "epoch": 0.36753, + "grad_norm": 1.5089996562319314, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36753 + }, + { + "epoch": 0.36754, + "grad_norm": 1.3959327018811052, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 36754 + }, + { + "epoch": 0.36755, + "grad_norm": 1.3737709603609651, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 36755 + }, + { + "epoch": 0.36756, + "grad_norm": 1.4292486613323876, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 36756 + }, + { + "epoch": 0.36757, + "grad_norm": 1.763911026708902, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36757 + }, + { + "epoch": 0.36758, + "grad_norm": 1.4045584308022798, + "learning_rate": 0.003, + "loss": 4.035, + "step": 36758 + }, + { + "epoch": 0.36759, + "grad_norm": 1.6768330461420013, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36759 + }, + { + "epoch": 0.3676, + "grad_norm": 1.2537525655936212, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36760 + }, + { + "epoch": 0.36761, + "grad_norm": 1.9768891972087772, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 36761 + }, + { + "epoch": 0.36762, + "grad_norm": 1.4150970362061508, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 36762 + }, + { + "epoch": 0.36763, + "grad_norm": 1.7541570249496115, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36763 + }, + { + "epoch": 0.36764, + "grad_norm": 1.4661331881815962, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 36764 + }, + { + "epoch": 0.36765, + "grad_norm": 1.3701709238438595, + "learning_rate": 0.003, + "loss": 4.015, + "step": 36765 + }, + { + "epoch": 0.36766, + "grad_norm": 1.5162971510216174, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 36766 + }, + { + "epoch": 0.36767, + "grad_norm": 1.3970533517843684, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36767 + }, + { + "epoch": 0.36768, + "grad_norm": 1.5625447746150087, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 36768 + }, + { + "epoch": 0.36769, + "grad_norm": 1.4546360058243326, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36769 + }, + { + "epoch": 0.3677, + "grad_norm": 1.3177843587609113, + "learning_rate": 0.003, + "loss": 4.024, + "step": 36770 + }, + { + "epoch": 0.36771, + "grad_norm": 1.5446173816245161, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 36771 + }, + { + "epoch": 0.36772, + "grad_norm": 1.331479672717301, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 36772 + }, + { + "epoch": 0.36773, + "grad_norm": 1.7117272745408223, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 36773 + }, + { + "epoch": 0.36774, + "grad_norm": 1.5060565341237728, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 36774 + }, + { + "epoch": 0.36775, + "grad_norm": 1.5222089998421113, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 36775 + }, + { + "epoch": 0.36776, + "grad_norm": 1.6927738840763236, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36776 + }, + { + "epoch": 0.36777, + "grad_norm": 1.1067354711402715, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 36777 + }, + { + "epoch": 0.36778, + "grad_norm": 1.6902291210328126, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 36778 + }, + { + "epoch": 0.36779, + "grad_norm": 1.4766678507990407, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 36779 + }, + { + "epoch": 0.3678, + "grad_norm": 1.3983176111880669, + "learning_rate": 0.003, + "loss": 4.053, + "step": 36780 + }, + { + "epoch": 0.36781, + "grad_norm": 1.698540034688076, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 36781 + }, + { + "epoch": 0.36782, + "grad_norm": 1.4236695088078233, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 36782 + }, + { + "epoch": 0.36783, + "grad_norm": 1.5721018419604567, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 36783 + }, + { + "epoch": 0.36784, + "grad_norm": 1.441305770220044, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 36784 + }, + { + "epoch": 0.36785, + "grad_norm": 1.3272938725099317, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 36785 + }, + { + "epoch": 0.36786, + "grad_norm": 1.7379131094801938, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 36786 + }, + { + "epoch": 0.36787, + "grad_norm": 1.2081123050141205, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 36787 + }, + { + "epoch": 0.36788, + "grad_norm": 1.812452657899524, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 36788 + }, + { + "epoch": 0.36789, + "grad_norm": 1.6157427472490031, + "learning_rate": 0.003, + "loss": 4.043, + "step": 36789 + }, + { + "epoch": 0.3679, + "grad_norm": 1.6957353003684228, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36790 + }, + { + "epoch": 0.36791, + "grad_norm": 1.5219243839848522, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 36791 + }, + { + "epoch": 0.36792, + "grad_norm": 1.4988877544491535, + "learning_rate": 0.003, + "loss": 4.047, + "step": 36792 + }, + { + "epoch": 0.36793, + "grad_norm": 1.5588149796924375, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 36793 + }, + { + "epoch": 0.36794, + "grad_norm": 1.2403109394839062, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 36794 + }, + { + "epoch": 0.36795, + "grad_norm": 1.6333193169571658, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 36795 + }, + { + "epoch": 0.36796, + "grad_norm": 1.1480694810873386, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 36796 + }, + { + "epoch": 0.36797, + "grad_norm": 1.6164563298070813, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 36797 + }, + { + "epoch": 0.36798, + "grad_norm": 1.3819218431155402, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 36798 + }, + { + "epoch": 0.36799, + "grad_norm": 1.596699609400049, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 36799 + }, + { + "epoch": 0.368, + "grad_norm": 1.226049209488199, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 36800 + }, + { + "epoch": 0.36801, + "grad_norm": 1.6810926478176396, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 36801 + }, + { + "epoch": 0.36802, + "grad_norm": 1.4913387345312112, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36802 + }, + { + "epoch": 0.36803, + "grad_norm": 2.0570064120510296, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 36803 + }, + { + "epoch": 0.36804, + "grad_norm": 1.431975532543942, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 36804 + }, + { + "epoch": 0.36805, + "grad_norm": 1.4644126305876617, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 36805 + }, + { + "epoch": 0.36806, + "grad_norm": 1.4538007470954453, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 36806 + }, + { + "epoch": 0.36807, + "grad_norm": 1.768107405268253, + "learning_rate": 0.003, + "loss": 4.058, + "step": 36807 + }, + { + "epoch": 0.36808, + "grad_norm": 1.2507478745204879, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 36808 + }, + { + "epoch": 0.36809, + "grad_norm": 1.6574165498536746, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 36809 + }, + { + "epoch": 0.3681, + "grad_norm": 1.4622025536352998, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36810 + }, + { + "epoch": 0.36811, + "grad_norm": 1.4589720940804778, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 36811 + }, + { + "epoch": 0.36812, + "grad_norm": 1.432742600336829, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 36812 + }, + { + "epoch": 0.36813, + "grad_norm": 1.427349899052509, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 36813 + }, + { + "epoch": 0.36814, + "grad_norm": 1.5236357744350646, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36814 + }, + { + "epoch": 0.36815, + "grad_norm": 1.5031511053944118, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36815 + }, + { + "epoch": 0.36816, + "grad_norm": 1.4181060190559482, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 36816 + }, + { + "epoch": 0.36817, + "grad_norm": 1.437677727061111, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 36817 + }, + { + "epoch": 0.36818, + "grad_norm": 1.8762830197825597, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 36818 + }, + { + "epoch": 0.36819, + "grad_norm": 1.239827363602783, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 36819 + }, + { + "epoch": 0.3682, + "grad_norm": 2.0685277772119877, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 36820 + }, + { + "epoch": 0.36821, + "grad_norm": 1.4464158573466899, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 36821 + }, + { + "epoch": 0.36822, + "grad_norm": 1.243708850360996, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 36822 + }, + { + "epoch": 0.36823, + "grad_norm": 1.5972485673182895, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 36823 + }, + { + "epoch": 0.36824, + "grad_norm": 1.5538774287821693, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 36824 + }, + { + "epoch": 0.36825, + "grad_norm": 1.361429050165019, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 36825 + }, + { + "epoch": 0.36826, + "grad_norm": 1.4327617549280407, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 36826 + }, + { + "epoch": 0.36827, + "grad_norm": 1.6641604005963986, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 36827 + }, + { + "epoch": 0.36828, + "grad_norm": 1.5006511406954777, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 36828 + }, + { + "epoch": 0.36829, + "grad_norm": 1.5934615311188858, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 36829 + }, + { + "epoch": 0.3683, + "grad_norm": 1.4722525410536846, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 36830 + }, + { + "epoch": 0.36831, + "grad_norm": 1.396095811766188, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 36831 + }, + { + "epoch": 0.36832, + "grad_norm": 1.4777562344124888, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 36832 + }, + { + "epoch": 0.36833, + "grad_norm": 1.5308110752901505, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 36833 + }, + { + "epoch": 0.36834, + "grad_norm": 1.5873011246590778, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 36834 + }, + { + "epoch": 0.36835, + "grad_norm": 1.4653431248829398, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 36835 + }, + { + "epoch": 0.36836, + "grad_norm": 1.585129443682628, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 36836 + }, + { + "epoch": 0.36837, + "grad_norm": 1.3588536619687486, + "learning_rate": 0.003, + "loss": 4.043, + "step": 36837 + }, + { + "epoch": 0.36838, + "grad_norm": 1.720937056058445, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 36838 + }, + { + "epoch": 0.36839, + "grad_norm": 1.2239841951411738, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 36839 + }, + { + "epoch": 0.3684, + "grad_norm": 1.6611969404670115, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 36840 + }, + { + "epoch": 0.36841, + "grad_norm": 1.6220467214582726, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 36841 + }, + { + "epoch": 0.36842, + "grad_norm": 1.6460547521301259, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36842 + }, + { + "epoch": 0.36843, + "grad_norm": 1.3517290916282703, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36843 + }, + { + "epoch": 0.36844, + "grad_norm": 1.4393536848368373, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36844 + }, + { + "epoch": 0.36845, + "grad_norm": 1.5295735159890154, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 36845 + }, + { + "epoch": 0.36846, + "grad_norm": 1.4329146835840583, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 36846 + }, + { + "epoch": 0.36847, + "grad_norm": 1.3424340371484726, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 36847 + }, + { + "epoch": 0.36848, + "grad_norm": 1.3727930927163225, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 36848 + }, + { + "epoch": 0.36849, + "grad_norm": 1.4452484326439599, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 36849 + }, + { + "epoch": 0.3685, + "grad_norm": 1.3666243640944948, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 36850 + }, + { + "epoch": 0.36851, + "grad_norm": 1.7028457799774666, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 36851 + }, + { + "epoch": 0.36852, + "grad_norm": 1.5923970701960124, + "learning_rate": 0.003, + "loss": 4.048, + "step": 36852 + }, + { + "epoch": 0.36853, + "grad_norm": 1.8035867643677226, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 36853 + }, + { + "epoch": 0.36854, + "grad_norm": 1.2592048576078003, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 36854 + }, + { + "epoch": 0.36855, + "grad_norm": 1.5748840697234685, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 36855 + }, + { + "epoch": 0.36856, + "grad_norm": 1.582181299935453, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 36856 + }, + { + "epoch": 0.36857, + "grad_norm": 1.5589168305308743, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 36857 + }, + { + "epoch": 0.36858, + "grad_norm": 1.4086127255716214, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36858 + }, + { + "epoch": 0.36859, + "grad_norm": 2.199336707388514, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36859 + }, + { + "epoch": 0.3686, + "grad_norm": 1.0189747773467757, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 36860 + }, + { + "epoch": 0.36861, + "grad_norm": 1.8983140783686347, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 36861 + }, + { + "epoch": 0.36862, + "grad_norm": 1.7585032313782198, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 36862 + }, + { + "epoch": 0.36863, + "grad_norm": 1.597747834155473, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 36863 + }, + { + "epoch": 0.36864, + "grad_norm": 1.6846747751338327, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 36864 + }, + { + "epoch": 0.36865, + "grad_norm": 1.4842103411985803, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 36865 + }, + { + "epoch": 0.36866, + "grad_norm": 1.4779881166208113, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 36866 + }, + { + "epoch": 0.36867, + "grad_norm": 1.251008711120232, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 36867 + }, + { + "epoch": 0.36868, + "grad_norm": 1.470119581284028, + "learning_rate": 0.003, + "loss": 4.053, + "step": 36868 + }, + { + "epoch": 0.36869, + "grad_norm": 1.7992377959126988, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 36869 + }, + { + "epoch": 0.3687, + "grad_norm": 1.2627245159941807, + "learning_rate": 0.003, + "loss": 4.03, + "step": 36870 + }, + { + "epoch": 0.36871, + "grad_norm": 1.5814422955807164, + "learning_rate": 0.003, + "loss": 4.045, + "step": 36871 + }, + { + "epoch": 0.36872, + "grad_norm": 1.3803047283373433, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 36872 + }, + { + "epoch": 0.36873, + "grad_norm": 1.5092272541577871, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 36873 + }, + { + "epoch": 0.36874, + "grad_norm": 1.4759026416986147, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 36874 + }, + { + "epoch": 0.36875, + "grad_norm": 1.3647517577232646, + "learning_rate": 0.003, + "loss": 4.031, + "step": 36875 + }, + { + "epoch": 0.36876, + "grad_norm": 1.343651895820688, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 36876 + }, + { + "epoch": 0.36877, + "grad_norm": 1.5240113881462338, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 36877 + }, + { + "epoch": 0.36878, + "grad_norm": 1.344300736140805, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 36878 + }, + { + "epoch": 0.36879, + "grad_norm": 1.901086144726974, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 36879 + }, + { + "epoch": 0.3688, + "grad_norm": 1.4614950104666455, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 36880 + }, + { + "epoch": 0.36881, + "grad_norm": 1.42574882823125, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 36881 + }, + { + "epoch": 0.36882, + "grad_norm": 1.9756096692583547, + "learning_rate": 0.003, + "loss": 4.06, + "step": 36882 + }, + { + "epoch": 0.36883, + "grad_norm": 1.2492290535546848, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36883 + }, + { + "epoch": 0.36884, + "grad_norm": 1.9604058144132146, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 36884 + }, + { + "epoch": 0.36885, + "grad_norm": 1.289784281640748, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 36885 + }, + { + "epoch": 0.36886, + "grad_norm": 1.6614429807360847, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 36886 + }, + { + "epoch": 0.36887, + "grad_norm": 1.450180798345175, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 36887 + }, + { + "epoch": 0.36888, + "grad_norm": 1.6031237395495517, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 36888 + }, + { + "epoch": 0.36889, + "grad_norm": 1.3309569102050742, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 36889 + }, + { + "epoch": 0.3689, + "grad_norm": 1.4291350839067771, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 36890 + }, + { + "epoch": 0.36891, + "grad_norm": 1.5132407742119145, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 36891 + }, + { + "epoch": 0.36892, + "grad_norm": 1.5036229369711693, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36892 + }, + { + "epoch": 0.36893, + "grad_norm": 1.8164885116241534, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 36893 + }, + { + "epoch": 0.36894, + "grad_norm": 1.236095627728231, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 36894 + }, + { + "epoch": 0.36895, + "grad_norm": 1.7665933092578123, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 36895 + }, + { + "epoch": 0.36896, + "grad_norm": 1.3882836032572066, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36896 + }, + { + "epoch": 0.36897, + "grad_norm": 1.4503125351305632, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 36897 + }, + { + "epoch": 0.36898, + "grad_norm": 1.587489856129675, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 36898 + }, + { + "epoch": 0.36899, + "grad_norm": 1.3864577658556902, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36899 + }, + { + "epoch": 0.369, + "grad_norm": 1.4124417979921895, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 36900 + }, + { + "epoch": 0.36901, + "grad_norm": 1.4159723205134, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 36901 + }, + { + "epoch": 0.36902, + "grad_norm": 1.447161250135972, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 36902 + }, + { + "epoch": 0.36903, + "grad_norm": 1.3237561747464104, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36903 + }, + { + "epoch": 0.36904, + "grad_norm": 1.3019066585419858, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 36904 + }, + { + "epoch": 0.36905, + "grad_norm": 1.6111607223969262, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 36905 + }, + { + "epoch": 0.36906, + "grad_norm": 1.5109389968538716, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36906 + }, + { + "epoch": 0.36907, + "grad_norm": 1.5905596925589098, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 36907 + }, + { + "epoch": 0.36908, + "grad_norm": 1.8850524956484618, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 36908 + }, + { + "epoch": 0.36909, + "grad_norm": 1.4765556021471786, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 36909 + }, + { + "epoch": 0.3691, + "grad_norm": 1.3888113877250456, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 36910 + }, + { + "epoch": 0.36911, + "grad_norm": 1.6783203738944792, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 36911 + }, + { + "epoch": 0.36912, + "grad_norm": 1.9080802613207153, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 36912 + }, + { + "epoch": 0.36913, + "grad_norm": 1.3254471458849368, + "learning_rate": 0.003, + "loss": 4.061, + "step": 36913 + }, + { + "epoch": 0.36914, + "grad_norm": 1.6576812661909226, + "learning_rate": 0.003, + "loss": 4.033, + "step": 36914 + }, + { + "epoch": 0.36915, + "grad_norm": 1.4713216736125452, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 36915 + }, + { + "epoch": 0.36916, + "grad_norm": 1.3780054097765384, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 36916 + }, + { + "epoch": 0.36917, + "grad_norm": 1.5041485605719498, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36917 + }, + { + "epoch": 0.36918, + "grad_norm": 1.36716630357803, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 36918 + }, + { + "epoch": 0.36919, + "grad_norm": 1.5317213794133353, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 36919 + }, + { + "epoch": 0.3692, + "grad_norm": 1.431842925032379, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 36920 + }, + { + "epoch": 0.36921, + "grad_norm": 1.6088221016041007, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36921 + }, + { + "epoch": 0.36922, + "grad_norm": 1.4100745824872136, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 36922 + }, + { + "epoch": 0.36923, + "grad_norm": 1.5528293425499158, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 36923 + }, + { + "epoch": 0.36924, + "grad_norm": 1.475720605532244, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 36924 + }, + { + "epoch": 0.36925, + "grad_norm": 1.7863717245189834, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 36925 + }, + { + "epoch": 0.36926, + "grad_norm": 1.3969918557845118, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 36926 + }, + { + "epoch": 0.36927, + "grad_norm": 1.617166418434999, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 36927 + }, + { + "epoch": 0.36928, + "grad_norm": 1.399856258103612, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 36928 + }, + { + "epoch": 0.36929, + "grad_norm": 1.5541581633266044, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 36929 + }, + { + "epoch": 0.3693, + "grad_norm": 1.2300914531850986, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 36930 + }, + { + "epoch": 0.36931, + "grad_norm": 1.87612639876198, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 36931 + }, + { + "epoch": 0.36932, + "grad_norm": 1.183287752725886, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 36932 + }, + { + "epoch": 0.36933, + "grad_norm": 1.7288722287127039, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 36933 + }, + { + "epoch": 0.36934, + "grad_norm": 1.0993503871447363, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 36934 + }, + { + "epoch": 0.36935, + "grad_norm": 1.6495806594093945, + "learning_rate": 0.003, + "loss": 4.064, + "step": 36935 + }, + { + "epoch": 0.36936, + "grad_norm": 1.5047561084916308, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 36936 + }, + { + "epoch": 0.36937, + "grad_norm": 1.5665575803435359, + "learning_rate": 0.003, + "loss": 4.025, + "step": 36937 + }, + { + "epoch": 0.36938, + "grad_norm": 1.6551445548525978, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 36938 + }, + { + "epoch": 0.36939, + "grad_norm": 1.7912439305141186, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 36939 + }, + { + "epoch": 0.3694, + "grad_norm": 1.9015275076514095, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 36940 + }, + { + "epoch": 0.36941, + "grad_norm": 1.2053751725395185, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 36941 + }, + { + "epoch": 0.36942, + "grad_norm": 1.4294059705702442, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 36942 + }, + { + "epoch": 0.36943, + "grad_norm": 1.9082071389130497, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 36943 + }, + { + "epoch": 0.36944, + "grad_norm": 1.2960054657354239, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 36944 + }, + { + "epoch": 0.36945, + "grad_norm": 1.6159651300188267, + "learning_rate": 0.003, + "loss": 4.054, + "step": 36945 + }, + { + "epoch": 0.36946, + "grad_norm": 1.4688018121911446, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 36946 + }, + { + "epoch": 0.36947, + "grad_norm": 1.7718327807317926, + "learning_rate": 0.003, + "loss": 4.053, + "step": 36947 + }, + { + "epoch": 0.36948, + "grad_norm": 1.6811971968072956, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 36948 + }, + { + "epoch": 0.36949, + "grad_norm": 1.2892604903658553, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 36949 + }, + { + "epoch": 0.3695, + "grad_norm": 1.5607554297253026, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 36950 + }, + { + "epoch": 0.36951, + "grad_norm": 1.4437301912723939, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 36951 + }, + { + "epoch": 0.36952, + "grad_norm": 1.2761531142211848, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 36952 + }, + { + "epoch": 0.36953, + "grad_norm": 1.420869931188905, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 36953 + }, + { + "epoch": 0.36954, + "grad_norm": 1.335413371165907, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 36954 + }, + { + "epoch": 0.36955, + "grad_norm": 1.441265183742144, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 36955 + }, + { + "epoch": 0.36956, + "grad_norm": 1.7635739344896684, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 36956 + }, + { + "epoch": 0.36957, + "grad_norm": 1.6560630296537662, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 36957 + }, + { + "epoch": 0.36958, + "grad_norm": 2.187286082969296, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 36958 + }, + { + "epoch": 0.36959, + "grad_norm": 1.0314739296267645, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 36959 + }, + { + "epoch": 0.3696, + "grad_norm": 1.4960310872434086, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 36960 + }, + { + "epoch": 0.36961, + "grad_norm": 1.4564287909853875, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36961 + }, + { + "epoch": 0.36962, + "grad_norm": 1.5522751539296404, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 36962 + }, + { + "epoch": 0.36963, + "grad_norm": 1.515262260887205, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36963 + }, + { + "epoch": 0.36964, + "grad_norm": 1.9677384255213086, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 36964 + }, + { + "epoch": 0.36965, + "grad_norm": 1.5474454012062342, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 36965 + }, + { + "epoch": 0.36966, + "grad_norm": 1.5018546669357373, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 36966 + }, + { + "epoch": 0.36967, + "grad_norm": 1.4995448032079928, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36967 + }, + { + "epoch": 0.36968, + "grad_norm": 1.2170278242229744, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 36968 + }, + { + "epoch": 0.36969, + "grad_norm": 1.8133574383455857, + "learning_rate": 0.003, + "loss": 4.046, + "step": 36969 + }, + { + "epoch": 0.3697, + "grad_norm": 1.2250640885027844, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 36970 + }, + { + "epoch": 0.36971, + "grad_norm": 1.4784894993900415, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 36971 + }, + { + "epoch": 0.36972, + "grad_norm": 1.5417374360617055, + "learning_rate": 0.003, + "loss": 4.086, + "step": 36972 + }, + { + "epoch": 0.36973, + "grad_norm": 1.9587333109555538, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 36973 + }, + { + "epoch": 0.36974, + "grad_norm": 1.053589220088777, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 36974 + }, + { + "epoch": 0.36975, + "grad_norm": 1.8126761941152836, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 36975 + }, + { + "epoch": 0.36976, + "grad_norm": 1.7676299882026172, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 36976 + }, + { + "epoch": 0.36977, + "grad_norm": 1.7202754832009763, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 36977 + }, + { + "epoch": 0.36978, + "grad_norm": 1.3251264061917507, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 36978 + }, + { + "epoch": 0.36979, + "grad_norm": 1.6784088268847628, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 36979 + }, + { + "epoch": 0.3698, + "grad_norm": 1.590579824804402, + "learning_rate": 0.003, + "loss": 4.041, + "step": 36980 + }, + { + "epoch": 0.36981, + "grad_norm": 1.3538217049594488, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 36981 + }, + { + "epoch": 0.36982, + "grad_norm": 1.3959711042994858, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 36982 + }, + { + "epoch": 0.36983, + "grad_norm": 1.3814775876772163, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 36983 + }, + { + "epoch": 0.36984, + "grad_norm": 1.4781291694819856, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 36984 + }, + { + "epoch": 0.36985, + "grad_norm": 1.467116488241281, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 36985 + }, + { + "epoch": 0.36986, + "grad_norm": 1.367280588222359, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 36986 + }, + { + "epoch": 0.36987, + "grad_norm": 1.7370179617403159, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 36987 + }, + { + "epoch": 0.36988, + "grad_norm": 1.7229462581396806, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 36988 + }, + { + "epoch": 0.36989, + "grad_norm": 1.3443352367660464, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 36989 + }, + { + "epoch": 0.3699, + "grad_norm": 1.593382079785703, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 36990 + }, + { + "epoch": 0.36991, + "grad_norm": 1.6337864848448853, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 36991 + }, + { + "epoch": 0.36992, + "grad_norm": 1.638721257023527, + "learning_rate": 0.003, + "loss": 4.035, + "step": 36992 + }, + { + "epoch": 0.36993, + "grad_norm": 1.6210461169759232, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 36993 + }, + { + "epoch": 0.36994, + "grad_norm": 1.294770031709204, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 36994 + }, + { + "epoch": 0.36995, + "grad_norm": 1.7404892487361148, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 36995 + }, + { + "epoch": 0.36996, + "grad_norm": 1.130674171849899, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 36996 + }, + { + "epoch": 0.36997, + "grad_norm": 1.7444320180436972, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 36997 + }, + { + "epoch": 0.36998, + "grad_norm": 1.3490305650579044, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 36998 + }, + { + "epoch": 0.36999, + "grad_norm": 1.5772348765428146, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 36999 + }, + { + "epoch": 0.37, + "grad_norm": 1.7572798810641796, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37000 + }, + { + "epoch": 0.37001, + "grad_norm": 1.240205885588833, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 37001 + }, + { + "epoch": 0.37002, + "grad_norm": 1.4553488133893875, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 37002 + }, + { + "epoch": 0.37003, + "grad_norm": 1.5101197418687162, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 37003 + }, + { + "epoch": 0.37004, + "grad_norm": 1.5453224083841892, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 37004 + }, + { + "epoch": 0.37005, + "grad_norm": 1.3256266142151576, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 37005 + }, + { + "epoch": 0.37006, + "grad_norm": 1.575877749472833, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37006 + }, + { + "epoch": 0.37007, + "grad_norm": 1.4195730317964839, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 37007 + }, + { + "epoch": 0.37008, + "grad_norm": 1.5975916288439471, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 37008 + }, + { + "epoch": 0.37009, + "grad_norm": 1.6430594749227654, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 37009 + }, + { + "epoch": 0.3701, + "grad_norm": 1.630628717774746, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37010 + }, + { + "epoch": 0.37011, + "grad_norm": 1.6130515021803522, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 37011 + }, + { + "epoch": 0.37012, + "grad_norm": 1.521470922002301, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 37012 + }, + { + "epoch": 0.37013, + "grad_norm": 1.4506928258801235, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 37013 + }, + { + "epoch": 0.37014, + "grad_norm": 1.6628095041747624, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 37014 + }, + { + "epoch": 0.37015, + "grad_norm": 1.5338871843057877, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 37015 + }, + { + "epoch": 0.37016, + "grad_norm": 1.4217135709957784, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 37016 + }, + { + "epoch": 0.37017, + "grad_norm": 1.6368770079518082, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37017 + }, + { + "epoch": 0.37018, + "grad_norm": 1.3542857578797038, + "learning_rate": 0.003, + "loss": 4.014, + "step": 37018 + }, + { + "epoch": 0.37019, + "grad_norm": 1.338839746391794, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 37019 + }, + { + "epoch": 0.3702, + "grad_norm": 1.4863053775172534, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 37020 + }, + { + "epoch": 0.37021, + "grad_norm": 1.4153656214313248, + "learning_rate": 0.003, + "loss": 4.011, + "step": 37021 + }, + { + "epoch": 0.37022, + "grad_norm": 1.5289417511358818, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 37022 + }, + { + "epoch": 0.37023, + "grad_norm": 1.6872677288233968, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 37023 + }, + { + "epoch": 0.37024, + "grad_norm": 1.3005636224241217, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37024 + }, + { + "epoch": 0.37025, + "grad_norm": 1.452808907731351, + "learning_rate": 0.003, + "loss": 4.041, + "step": 37025 + }, + { + "epoch": 0.37026, + "grad_norm": 1.5416092614909174, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 37026 + }, + { + "epoch": 0.37027, + "grad_norm": 1.2766118524971104, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 37027 + }, + { + "epoch": 0.37028, + "grad_norm": 1.9287390709977024, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 37028 + }, + { + "epoch": 0.37029, + "grad_norm": 1.4348375964740703, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 37029 + }, + { + "epoch": 0.3703, + "grad_norm": 1.7955674729032365, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37030 + }, + { + "epoch": 0.37031, + "grad_norm": 1.4970260882461022, + "learning_rate": 0.003, + "loss": 4.041, + "step": 37031 + }, + { + "epoch": 0.37032, + "grad_norm": 1.5429231610969214, + "learning_rate": 0.003, + "loss": 4.019, + "step": 37032 + }, + { + "epoch": 0.37033, + "grad_norm": 1.309464699493241, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 37033 + }, + { + "epoch": 0.37034, + "grad_norm": 1.67995398214165, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 37034 + }, + { + "epoch": 0.37035, + "grad_norm": 1.1865053914820656, + "learning_rate": 0.003, + "loss": 4.034, + "step": 37035 + }, + { + "epoch": 0.37036, + "grad_norm": 1.5264419390569466, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 37036 + }, + { + "epoch": 0.37037, + "grad_norm": 1.279872306413552, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 37037 + }, + { + "epoch": 0.37038, + "grad_norm": 1.7397039717449772, + "learning_rate": 0.003, + "loss": 4.057, + "step": 37038 + }, + { + "epoch": 0.37039, + "grad_norm": 1.2463341045796572, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37039 + }, + { + "epoch": 0.3704, + "grad_norm": 1.5123220991244537, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 37040 + }, + { + "epoch": 0.37041, + "grad_norm": 1.6361470562570253, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 37041 + }, + { + "epoch": 0.37042, + "grad_norm": 1.4939267177991393, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 37042 + }, + { + "epoch": 0.37043, + "grad_norm": 1.6202092232686116, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 37043 + }, + { + "epoch": 0.37044, + "grad_norm": 1.4225985266045331, + "learning_rate": 0.003, + "loss": 4.047, + "step": 37044 + }, + { + "epoch": 0.37045, + "grad_norm": 1.4251776843385204, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37045 + }, + { + "epoch": 0.37046, + "grad_norm": 1.684242675265506, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37046 + }, + { + "epoch": 0.37047, + "grad_norm": 1.335190038528461, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 37047 + }, + { + "epoch": 0.37048, + "grad_norm": 1.590010392966198, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37048 + }, + { + "epoch": 0.37049, + "grad_norm": 1.7711673644851937, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 37049 + }, + { + "epoch": 0.3705, + "grad_norm": 1.5607996217692366, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 37050 + }, + { + "epoch": 0.37051, + "grad_norm": 1.4381730442354723, + "learning_rate": 0.003, + "loss": 4.03, + "step": 37051 + }, + { + "epoch": 0.37052, + "grad_norm": 1.6901907646506649, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37052 + }, + { + "epoch": 0.37053, + "grad_norm": 1.3342822398561744, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37053 + }, + { + "epoch": 0.37054, + "grad_norm": 1.6218856804594755, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 37054 + }, + { + "epoch": 0.37055, + "grad_norm": 1.4535598791762423, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37055 + }, + { + "epoch": 0.37056, + "grad_norm": 1.2195495026314258, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37056 + }, + { + "epoch": 0.37057, + "grad_norm": 1.4902614454971532, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37057 + }, + { + "epoch": 0.37058, + "grad_norm": 1.339057419938805, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37058 + }, + { + "epoch": 0.37059, + "grad_norm": 1.7331396307995783, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 37059 + }, + { + "epoch": 0.3706, + "grad_norm": 1.3752635251720116, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 37060 + }, + { + "epoch": 0.37061, + "grad_norm": 1.4894402892962304, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 37061 + }, + { + "epoch": 0.37062, + "grad_norm": 1.4873971374236012, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37062 + }, + { + "epoch": 0.37063, + "grad_norm": 1.5297381284783917, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 37063 + }, + { + "epoch": 0.37064, + "grad_norm": 1.5513415458764832, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37064 + }, + { + "epoch": 0.37065, + "grad_norm": 1.8811938369965453, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37065 + }, + { + "epoch": 0.37066, + "grad_norm": 1.3275869403877159, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 37066 + }, + { + "epoch": 0.37067, + "grad_norm": 1.5128880659870083, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37067 + }, + { + "epoch": 0.37068, + "grad_norm": 1.6792677859695404, + "learning_rate": 0.003, + "loss": 4.041, + "step": 37068 + }, + { + "epoch": 0.37069, + "grad_norm": 1.5057251312449564, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 37069 + }, + { + "epoch": 0.3707, + "grad_norm": 1.4474961775372333, + "learning_rate": 0.003, + "loss": 4.046, + "step": 37070 + }, + { + "epoch": 0.37071, + "grad_norm": 1.7791774264953804, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37071 + }, + { + "epoch": 0.37072, + "grad_norm": 1.1428692913764955, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 37072 + }, + { + "epoch": 0.37073, + "grad_norm": 1.5208983070346935, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37073 + }, + { + "epoch": 0.37074, + "grad_norm": 1.3619676224468134, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37074 + }, + { + "epoch": 0.37075, + "grad_norm": 1.6050883332044767, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 37075 + }, + { + "epoch": 0.37076, + "grad_norm": 1.59865125658924, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 37076 + }, + { + "epoch": 0.37077, + "grad_norm": 1.355405405724074, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37077 + }, + { + "epoch": 0.37078, + "grad_norm": 1.5881613192179522, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37078 + }, + { + "epoch": 0.37079, + "grad_norm": 1.3868655028758496, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37079 + }, + { + "epoch": 0.3708, + "grad_norm": 1.5118358445487219, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 37080 + }, + { + "epoch": 0.37081, + "grad_norm": 1.4997671879838481, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37081 + }, + { + "epoch": 0.37082, + "grad_norm": 1.4233698276513533, + "learning_rate": 0.003, + "loss": 4.02, + "step": 37082 + }, + { + "epoch": 0.37083, + "grad_norm": 1.4686470827968032, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37083 + }, + { + "epoch": 0.37084, + "grad_norm": 1.467410804022657, + "learning_rate": 0.003, + "loss": 4.067, + "step": 37084 + }, + { + "epoch": 0.37085, + "grad_norm": 1.7115607564036937, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 37085 + }, + { + "epoch": 0.37086, + "grad_norm": 1.5553184328384677, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 37086 + }, + { + "epoch": 0.37087, + "grad_norm": 1.4710635978634057, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 37087 + }, + { + "epoch": 0.37088, + "grad_norm": 1.5605830563283458, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37088 + }, + { + "epoch": 0.37089, + "grad_norm": 1.255857830020857, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 37089 + }, + { + "epoch": 0.3709, + "grad_norm": 1.8139566635117796, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37090 + }, + { + "epoch": 0.37091, + "grad_norm": 1.7055248768018278, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 37091 + }, + { + "epoch": 0.37092, + "grad_norm": 1.8814415035020184, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 37092 + }, + { + "epoch": 0.37093, + "grad_norm": 1.1884267726084452, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 37093 + }, + { + "epoch": 0.37094, + "grad_norm": 1.454736233665365, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 37094 + }, + { + "epoch": 0.37095, + "grad_norm": 1.5451770532127784, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 37095 + }, + { + "epoch": 0.37096, + "grad_norm": 1.5246794340940086, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 37096 + }, + { + "epoch": 0.37097, + "grad_norm": 1.276145156093401, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37097 + }, + { + "epoch": 0.37098, + "grad_norm": 1.7639413127643244, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 37098 + }, + { + "epoch": 0.37099, + "grad_norm": 1.1272222589517416, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 37099 + }, + { + "epoch": 0.371, + "grad_norm": 2.1216307414451525, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 37100 + }, + { + "epoch": 0.37101, + "grad_norm": 1.392438526540701, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 37101 + }, + { + "epoch": 0.37102, + "grad_norm": 2.067332725720101, + "learning_rate": 0.003, + "loss": 4.06, + "step": 37102 + }, + { + "epoch": 0.37103, + "grad_norm": 1.7260901634141586, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 37103 + }, + { + "epoch": 0.37104, + "grad_norm": 1.492290709580146, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 37104 + }, + { + "epoch": 0.37105, + "grad_norm": 1.450527954412329, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 37105 + }, + { + "epoch": 0.37106, + "grad_norm": 1.509649426835975, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 37106 + }, + { + "epoch": 0.37107, + "grad_norm": 1.4467106685253097, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 37107 + }, + { + "epoch": 0.37108, + "grad_norm": 1.5387051943621437, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37108 + }, + { + "epoch": 0.37109, + "grad_norm": 1.7055455630367609, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37109 + }, + { + "epoch": 0.3711, + "grad_norm": 1.2528747877403281, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 37110 + }, + { + "epoch": 0.37111, + "grad_norm": 1.5499948783846513, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37111 + }, + { + "epoch": 0.37112, + "grad_norm": 1.6654088526545443, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 37112 + }, + { + "epoch": 0.37113, + "grad_norm": 1.3134747282407242, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 37113 + }, + { + "epoch": 0.37114, + "grad_norm": 1.5604023204650297, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 37114 + }, + { + "epoch": 0.37115, + "grad_norm": 1.478210614606876, + "learning_rate": 0.003, + "loss": 4.046, + "step": 37115 + }, + { + "epoch": 0.37116, + "grad_norm": 1.520026015128451, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 37116 + }, + { + "epoch": 0.37117, + "grad_norm": 1.2821650922852232, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 37117 + }, + { + "epoch": 0.37118, + "grad_norm": 1.6366689145796713, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37118 + }, + { + "epoch": 0.37119, + "grad_norm": 1.6941969444734708, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 37119 + }, + { + "epoch": 0.3712, + "grad_norm": 1.9123124148005683, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 37120 + }, + { + "epoch": 0.37121, + "grad_norm": 1.343543344238271, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 37121 + }, + { + "epoch": 0.37122, + "grad_norm": 1.416938763601328, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37122 + }, + { + "epoch": 0.37123, + "grad_norm": 1.8778992886702082, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37123 + }, + { + "epoch": 0.37124, + "grad_norm": 1.2203379951928148, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37124 + }, + { + "epoch": 0.37125, + "grad_norm": 1.630728818444625, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 37125 + }, + { + "epoch": 0.37126, + "grad_norm": 1.185916889142491, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37126 + }, + { + "epoch": 0.37127, + "grad_norm": 1.7202538585973253, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37127 + }, + { + "epoch": 0.37128, + "grad_norm": 1.2991122540414959, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 37128 + }, + { + "epoch": 0.37129, + "grad_norm": 1.732913831068528, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37129 + }, + { + "epoch": 0.3713, + "grad_norm": 1.9115814332259051, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 37130 + }, + { + "epoch": 0.37131, + "grad_norm": 1.662203240977509, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 37131 + }, + { + "epoch": 0.37132, + "grad_norm": 1.6374838546399078, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37132 + }, + { + "epoch": 0.37133, + "grad_norm": 1.5579692802944456, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37133 + }, + { + "epoch": 0.37134, + "grad_norm": 1.185627426442694, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 37134 + }, + { + "epoch": 0.37135, + "grad_norm": 1.8393628484796603, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 37135 + }, + { + "epoch": 0.37136, + "grad_norm": 1.099248632267115, + "learning_rate": 0.003, + "loss": 4.021, + "step": 37136 + }, + { + "epoch": 0.37137, + "grad_norm": 1.599211576017926, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 37137 + }, + { + "epoch": 0.37138, + "grad_norm": 1.481754688618116, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 37138 + }, + { + "epoch": 0.37139, + "grad_norm": 1.5700039680956719, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 37139 + }, + { + "epoch": 0.3714, + "grad_norm": 1.6385775046793976, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 37140 + }, + { + "epoch": 0.37141, + "grad_norm": 1.7488668905637985, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 37141 + }, + { + "epoch": 0.37142, + "grad_norm": 1.87692440127728, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37142 + }, + { + "epoch": 0.37143, + "grad_norm": 1.114674795850613, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 37143 + }, + { + "epoch": 0.37144, + "grad_norm": 1.5221358491167178, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37144 + }, + { + "epoch": 0.37145, + "grad_norm": 1.352114581124833, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37145 + }, + { + "epoch": 0.37146, + "grad_norm": 1.68788744742339, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 37146 + }, + { + "epoch": 0.37147, + "grad_norm": 1.4081367557981934, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 37147 + }, + { + "epoch": 0.37148, + "grad_norm": 1.5348525616641484, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 37148 + }, + { + "epoch": 0.37149, + "grad_norm": 1.4998861135594732, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 37149 + }, + { + "epoch": 0.3715, + "grad_norm": 1.6781337421614306, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37150 + }, + { + "epoch": 0.37151, + "grad_norm": 1.4145092298479482, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 37151 + }, + { + "epoch": 0.37152, + "grad_norm": 1.4284805987744258, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 37152 + }, + { + "epoch": 0.37153, + "grad_norm": 1.4580553144333896, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 37153 + }, + { + "epoch": 0.37154, + "grad_norm": 1.5714190309805793, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37154 + }, + { + "epoch": 0.37155, + "grad_norm": 1.6006465828963072, + "learning_rate": 0.003, + "loss": 4.07, + "step": 37155 + }, + { + "epoch": 0.37156, + "grad_norm": 1.7563317533328053, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 37156 + }, + { + "epoch": 0.37157, + "grad_norm": 1.4326205720536993, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 37157 + }, + { + "epoch": 0.37158, + "grad_norm": 1.6203063943275218, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37158 + }, + { + "epoch": 0.37159, + "grad_norm": 1.5461823507697914, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 37159 + }, + { + "epoch": 0.3716, + "grad_norm": 1.3916754072520812, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37160 + }, + { + "epoch": 0.37161, + "grad_norm": 1.6827513156086011, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37161 + }, + { + "epoch": 0.37162, + "grad_norm": 1.276426955957325, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37162 + }, + { + "epoch": 0.37163, + "grad_norm": 1.6697522852769253, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 37163 + }, + { + "epoch": 0.37164, + "grad_norm": 1.3054124890324588, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37164 + }, + { + "epoch": 0.37165, + "grad_norm": 1.670414055586408, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 37165 + }, + { + "epoch": 0.37166, + "grad_norm": 1.2168748405148673, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37166 + }, + { + "epoch": 0.37167, + "grad_norm": 2.1168101389857776, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 37167 + }, + { + "epoch": 0.37168, + "grad_norm": 1.2485277668871806, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 37168 + }, + { + "epoch": 0.37169, + "grad_norm": 1.7646116738856856, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 37169 + }, + { + "epoch": 0.3717, + "grad_norm": 1.3978248788144283, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 37170 + }, + { + "epoch": 0.37171, + "grad_norm": 1.6312560192875594, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37171 + }, + { + "epoch": 0.37172, + "grad_norm": 1.511158585251026, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 37172 + }, + { + "epoch": 0.37173, + "grad_norm": 1.6314481639719625, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 37173 + }, + { + "epoch": 0.37174, + "grad_norm": 1.5540393738086018, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37174 + }, + { + "epoch": 0.37175, + "grad_norm": 1.5960907701009885, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37175 + }, + { + "epoch": 0.37176, + "grad_norm": 1.420483141616188, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37176 + }, + { + "epoch": 0.37177, + "grad_norm": 1.439716091062099, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 37177 + }, + { + "epoch": 0.37178, + "grad_norm": 1.4233781595251669, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37178 + }, + { + "epoch": 0.37179, + "grad_norm": 1.6276891844302956, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 37179 + }, + { + "epoch": 0.3718, + "grad_norm": 1.2840948530900052, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 37180 + }, + { + "epoch": 0.37181, + "grad_norm": 1.544427915865939, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 37181 + }, + { + "epoch": 0.37182, + "grad_norm": 1.248232852369635, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 37182 + }, + { + "epoch": 0.37183, + "grad_norm": 1.7950707946748152, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 37183 + }, + { + "epoch": 0.37184, + "grad_norm": 1.4211550496382908, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 37184 + }, + { + "epoch": 0.37185, + "grad_norm": 1.4774625681017113, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 37185 + }, + { + "epoch": 0.37186, + "grad_norm": 1.6297664159444767, + "learning_rate": 0.003, + "loss": 4.064, + "step": 37186 + }, + { + "epoch": 0.37187, + "grad_norm": 1.409590219283483, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 37187 + }, + { + "epoch": 0.37188, + "grad_norm": 1.2967919297934793, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37188 + }, + { + "epoch": 0.37189, + "grad_norm": 1.4010470893298523, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 37189 + }, + { + "epoch": 0.3719, + "grad_norm": 1.5645992203841075, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 37190 + }, + { + "epoch": 0.37191, + "grad_norm": 1.6220779424917984, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37191 + }, + { + "epoch": 0.37192, + "grad_norm": 1.418873289342756, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37192 + }, + { + "epoch": 0.37193, + "grad_norm": 1.6279737947420148, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37193 + }, + { + "epoch": 0.37194, + "grad_norm": 1.6305124894300453, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 37194 + }, + { + "epoch": 0.37195, + "grad_norm": 1.682512683890224, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 37195 + }, + { + "epoch": 0.37196, + "grad_norm": 1.524839858057616, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 37196 + }, + { + "epoch": 0.37197, + "grad_norm": 1.5108389253645913, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37197 + }, + { + "epoch": 0.37198, + "grad_norm": 1.3258208052931837, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37198 + }, + { + "epoch": 0.37199, + "grad_norm": 1.5864577075695314, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37199 + }, + { + "epoch": 0.372, + "grad_norm": 1.4506588707565609, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 37200 + }, + { + "epoch": 0.37201, + "grad_norm": 1.509314352610027, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 37201 + }, + { + "epoch": 0.37202, + "grad_norm": 1.5361300067407495, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 37202 + }, + { + "epoch": 0.37203, + "grad_norm": 1.6004316284352633, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 37203 + }, + { + "epoch": 0.37204, + "grad_norm": 1.7938823973688995, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37204 + }, + { + "epoch": 0.37205, + "grad_norm": 1.203018947027216, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 37205 + }, + { + "epoch": 0.37206, + "grad_norm": 1.6561024040473638, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37206 + }, + { + "epoch": 0.37207, + "grad_norm": 1.6206797121438645, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 37207 + }, + { + "epoch": 0.37208, + "grad_norm": 1.684277809564797, + "learning_rate": 0.003, + "loss": 4.066, + "step": 37208 + }, + { + "epoch": 0.37209, + "grad_norm": 1.3286244653514172, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 37209 + }, + { + "epoch": 0.3721, + "grad_norm": 1.335269923728003, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 37210 + }, + { + "epoch": 0.37211, + "grad_norm": 1.6578346023639636, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37211 + }, + { + "epoch": 0.37212, + "grad_norm": 1.2046123766200285, + "learning_rate": 0.003, + "loss": 4.006, + "step": 37212 + }, + { + "epoch": 0.37213, + "grad_norm": 1.6802288404887442, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 37213 + }, + { + "epoch": 0.37214, + "grad_norm": 1.523467511395088, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 37214 + }, + { + "epoch": 0.37215, + "grad_norm": 1.5784509563745492, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 37215 + }, + { + "epoch": 0.37216, + "grad_norm": 1.564819130397969, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 37216 + }, + { + "epoch": 0.37217, + "grad_norm": 1.8463024763659601, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 37217 + }, + { + "epoch": 0.37218, + "grad_norm": 1.5925203163826511, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37218 + }, + { + "epoch": 0.37219, + "grad_norm": 1.4332274713486897, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 37219 + }, + { + "epoch": 0.3722, + "grad_norm": 1.5259919899553873, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 37220 + }, + { + "epoch": 0.37221, + "grad_norm": 1.4448685887131987, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 37221 + }, + { + "epoch": 0.37222, + "grad_norm": 1.4967265898677764, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37222 + }, + { + "epoch": 0.37223, + "grad_norm": 1.5206698757674926, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37223 + }, + { + "epoch": 0.37224, + "grad_norm": 1.1479118924627811, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 37224 + }, + { + "epoch": 0.37225, + "grad_norm": 1.7149519797259865, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 37225 + }, + { + "epoch": 0.37226, + "grad_norm": 1.0880154408264826, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 37226 + }, + { + "epoch": 0.37227, + "grad_norm": 1.8177662976877915, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 37227 + }, + { + "epoch": 0.37228, + "grad_norm": 1.2672412364529388, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 37228 + }, + { + "epoch": 0.37229, + "grad_norm": 1.470697658200143, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37229 + }, + { + "epoch": 0.3723, + "grad_norm": 1.4901665289553976, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 37230 + }, + { + "epoch": 0.37231, + "grad_norm": 1.788087453571012, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 37231 + }, + { + "epoch": 0.37232, + "grad_norm": 1.8098538910870803, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 37232 + }, + { + "epoch": 0.37233, + "grad_norm": 1.440598827227318, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 37233 + }, + { + "epoch": 0.37234, + "grad_norm": 1.338656238264564, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37234 + }, + { + "epoch": 0.37235, + "grad_norm": 1.8848655402680696, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 37235 + }, + { + "epoch": 0.37236, + "grad_norm": 1.546200141518357, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37236 + }, + { + "epoch": 0.37237, + "grad_norm": 1.2363692944268752, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 37237 + }, + { + "epoch": 0.37238, + "grad_norm": 1.6764983503183115, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 37238 + }, + { + "epoch": 0.37239, + "grad_norm": 1.4704712005897815, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 37239 + }, + { + "epoch": 0.3724, + "grad_norm": 1.2728208047088965, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 37240 + }, + { + "epoch": 0.37241, + "grad_norm": 1.7075824764934433, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 37241 + }, + { + "epoch": 0.37242, + "grad_norm": 1.485658088138563, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 37242 + }, + { + "epoch": 0.37243, + "grad_norm": 1.4105150040919332, + "learning_rate": 0.003, + "loss": 4.042, + "step": 37243 + }, + { + "epoch": 0.37244, + "grad_norm": 1.4513510452561695, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37244 + }, + { + "epoch": 0.37245, + "grad_norm": 1.8800392317019474, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37245 + }, + { + "epoch": 0.37246, + "grad_norm": 1.4441803735456769, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37246 + }, + { + "epoch": 0.37247, + "grad_norm": 1.7497600914643563, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 37247 + }, + { + "epoch": 0.37248, + "grad_norm": 1.4450708291704202, + "learning_rate": 0.003, + "loss": 4.043, + "step": 37248 + }, + { + "epoch": 0.37249, + "grad_norm": 1.6181569088461558, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37249 + }, + { + "epoch": 0.3725, + "grad_norm": 1.4282009636943505, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 37250 + }, + { + "epoch": 0.37251, + "grad_norm": 1.4662158591300056, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 37251 + }, + { + "epoch": 0.37252, + "grad_norm": 1.6688680163567637, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 37252 + }, + { + "epoch": 0.37253, + "grad_norm": 1.3999322669244472, + "learning_rate": 0.003, + "loss": 4.021, + "step": 37253 + }, + { + "epoch": 0.37254, + "grad_norm": 1.6139511468955607, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 37254 + }, + { + "epoch": 0.37255, + "grad_norm": 1.346483592974611, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 37255 + }, + { + "epoch": 0.37256, + "grad_norm": 1.7238691027948838, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37256 + }, + { + "epoch": 0.37257, + "grad_norm": 1.2354652256998817, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37257 + }, + { + "epoch": 0.37258, + "grad_norm": 1.583084776708334, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37258 + }, + { + "epoch": 0.37259, + "grad_norm": 1.3179692920247936, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37259 + }, + { + "epoch": 0.3726, + "grad_norm": 1.6717821231869294, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 37260 + }, + { + "epoch": 0.37261, + "grad_norm": 1.2924543064688296, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 37261 + }, + { + "epoch": 0.37262, + "grad_norm": 1.811721569267591, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 37262 + }, + { + "epoch": 0.37263, + "grad_norm": 1.5255666399459438, + "learning_rate": 0.003, + "loss": 4.064, + "step": 37263 + }, + { + "epoch": 0.37264, + "grad_norm": 1.6407321767402405, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37264 + }, + { + "epoch": 0.37265, + "grad_norm": 1.310638506004886, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37265 + }, + { + "epoch": 0.37266, + "grad_norm": 1.6129340613471665, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 37266 + }, + { + "epoch": 0.37267, + "grad_norm": 1.490012647439361, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37267 + }, + { + "epoch": 0.37268, + "grad_norm": 1.3710324897791202, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 37268 + }, + { + "epoch": 0.37269, + "grad_norm": 1.6216973255210843, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 37269 + }, + { + "epoch": 0.3727, + "grad_norm": 1.3506216250434784, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37270 + }, + { + "epoch": 0.37271, + "grad_norm": 1.4880269556494157, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37271 + }, + { + "epoch": 0.37272, + "grad_norm": 1.3109276037287483, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 37272 + }, + { + "epoch": 0.37273, + "grad_norm": 1.478459688771538, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 37273 + }, + { + "epoch": 0.37274, + "grad_norm": 1.41453616177596, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 37274 + }, + { + "epoch": 0.37275, + "grad_norm": 1.879306517068022, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 37275 + }, + { + "epoch": 0.37276, + "grad_norm": 1.24458959705186, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 37276 + }, + { + "epoch": 0.37277, + "grad_norm": 1.7380597662746833, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37277 + }, + { + "epoch": 0.37278, + "grad_norm": 1.5203559771476836, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 37278 + }, + { + "epoch": 0.37279, + "grad_norm": 1.7938252035334714, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37279 + }, + { + "epoch": 0.3728, + "grad_norm": 1.2257508789076057, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 37280 + }, + { + "epoch": 0.37281, + "grad_norm": 1.5254672778944987, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37281 + }, + { + "epoch": 0.37282, + "grad_norm": 1.3982441791473565, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37282 + }, + { + "epoch": 0.37283, + "grad_norm": 1.5341750855048342, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37283 + }, + { + "epoch": 0.37284, + "grad_norm": 1.4918890181827495, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37284 + }, + { + "epoch": 0.37285, + "grad_norm": 1.44585104514623, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37285 + }, + { + "epoch": 0.37286, + "grad_norm": 1.9503594877661359, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 37286 + }, + { + "epoch": 0.37287, + "grad_norm": 1.2518638973620813, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 37287 + }, + { + "epoch": 0.37288, + "grad_norm": 1.4293074003893806, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 37288 + }, + { + "epoch": 0.37289, + "grad_norm": 1.6859364284730867, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 37289 + }, + { + "epoch": 0.3729, + "grad_norm": 1.5884139248114084, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37290 + }, + { + "epoch": 0.37291, + "grad_norm": 1.5592606756169267, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37291 + }, + { + "epoch": 0.37292, + "grad_norm": 1.7098046975214154, + "learning_rate": 0.003, + "loss": 4.052, + "step": 37292 + }, + { + "epoch": 0.37293, + "grad_norm": 1.5509672283020979, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 37293 + }, + { + "epoch": 0.37294, + "grad_norm": 1.715142874230295, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 37294 + }, + { + "epoch": 0.37295, + "grad_norm": 1.2988920584766597, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 37295 + }, + { + "epoch": 0.37296, + "grad_norm": 1.6396020023193913, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 37296 + }, + { + "epoch": 0.37297, + "grad_norm": 1.5316106453092129, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 37297 + }, + { + "epoch": 0.37298, + "grad_norm": 1.3701980176193365, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 37298 + }, + { + "epoch": 0.37299, + "grad_norm": 1.5524776734476724, + "learning_rate": 0.003, + "loss": 4.061, + "step": 37299 + }, + { + "epoch": 0.373, + "grad_norm": 1.719093929329888, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37300 + }, + { + "epoch": 0.37301, + "grad_norm": 1.3692306082331618, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 37301 + }, + { + "epoch": 0.37302, + "grad_norm": 1.7122117101940757, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 37302 + }, + { + "epoch": 0.37303, + "grad_norm": 1.3162403877788849, + "learning_rate": 0.003, + "loss": 3.994, + "step": 37303 + }, + { + "epoch": 0.37304, + "grad_norm": 1.7468257210722329, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 37304 + }, + { + "epoch": 0.37305, + "grad_norm": 1.538204762377728, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 37305 + }, + { + "epoch": 0.37306, + "grad_norm": 1.4481706850685618, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37306 + }, + { + "epoch": 0.37307, + "grad_norm": 1.5626055046560978, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 37307 + }, + { + "epoch": 0.37308, + "grad_norm": 1.4609122956027683, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37308 + }, + { + "epoch": 0.37309, + "grad_norm": 1.4454928121917907, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 37309 + }, + { + "epoch": 0.3731, + "grad_norm": 1.422999752856955, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 37310 + }, + { + "epoch": 0.37311, + "grad_norm": 1.3948836619468987, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 37311 + }, + { + "epoch": 0.37312, + "grad_norm": 1.736113213138571, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 37312 + }, + { + "epoch": 0.37313, + "grad_norm": 1.5904910834110473, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 37313 + }, + { + "epoch": 0.37314, + "grad_norm": 1.7687106176112906, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 37314 + }, + { + "epoch": 0.37315, + "grad_norm": 1.3634653867850337, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37315 + }, + { + "epoch": 0.37316, + "grad_norm": 1.7164035711309578, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37316 + }, + { + "epoch": 0.37317, + "grad_norm": 1.3197866403878689, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 37317 + }, + { + "epoch": 0.37318, + "grad_norm": 1.606331901387925, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 37318 + }, + { + "epoch": 0.37319, + "grad_norm": 1.5383779653229908, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 37319 + }, + { + "epoch": 0.3732, + "grad_norm": 1.5311537791512102, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37320 + }, + { + "epoch": 0.37321, + "grad_norm": 1.5109861048705273, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 37321 + }, + { + "epoch": 0.37322, + "grad_norm": 1.4125746629055116, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37322 + }, + { + "epoch": 0.37323, + "grad_norm": 1.7492958782158063, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 37323 + }, + { + "epoch": 0.37324, + "grad_norm": 1.348466585936138, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 37324 + }, + { + "epoch": 0.37325, + "grad_norm": 1.8588492555231686, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37325 + }, + { + "epoch": 0.37326, + "grad_norm": 1.4025546294227276, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37326 + }, + { + "epoch": 0.37327, + "grad_norm": 1.2922077543074018, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 37327 + }, + { + "epoch": 0.37328, + "grad_norm": 1.745260707844679, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37328 + }, + { + "epoch": 0.37329, + "grad_norm": 1.3819004969593498, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 37329 + }, + { + "epoch": 0.3733, + "grad_norm": 1.484169391214619, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 37330 + }, + { + "epoch": 0.37331, + "grad_norm": 1.5574036181318842, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 37331 + }, + { + "epoch": 0.37332, + "grad_norm": 1.989494767936185, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 37332 + }, + { + "epoch": 0.37333, + "grad_norm": 1.0934562808816037, + "learning_rate": 0.003, + "loss": 4.01, + "step": 37333 + }, + { + "epoch": 0.37334, + "grad_norm": 1.927838474164398, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 37334 + }, + { + "epoch": 0.37335, + "grad_norm": 1.3270676391836682, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37335 + }, + { + "epoch": 0.37336, + "grad_norm": 1.8241920057881174, + "learning_rate": 0.003, + "loss": 4.091, + "step": 37336 + }, + { + "epoch": 0.37337, + "grad_norm": 1.3486911615014763, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 37337 + }, + { + "epoch": 0.37338, + "grad_norm": 1.7531760257099358, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 37338 + }, + { + "epoch": 0.37339, + "grad_norm": 1.2835149125736558, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 37339 + }, + { + "epoch": 0.3734, + "grad_norm": 1.4713690637404846, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 37340 + }, + { + "epoch": 0.37341, + "grad_norm": 1.227511533442222, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37341 + }, + { + "epoch": 0.37342, + "grad_norm": 1.7549140176745934, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37342 + }, + { + "epoch": 0.37343, + "grad_norm": 2.0296767701634733, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 37343 + }, + { + "epoch": 0.37344, + "grad_norm": 1.297721315652186, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 37344 + }, + { + "epoch": 0.37345, + "grad_norm": 1.708132383512617, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 37345 + }, + { + "epoch": 0.37346, + "grad_norm": 1.2630070405246179, + "learning_rate": 0.003, + "loss": 4.047, + "step": 37346 + }, + { + "epoch": 0.37347, + "grad_norm": 1.5574201330829327, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37347 + }, + { + "epoch": 0.37348, + "grad_norm": 1.2527440233536595, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 37348 + }, + { + "epoch": 0.37349, + "grad_norm": 1.420371031604146, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 37349 + }, + { + "epoch": 0.3735, + "grad_norm": 1.509462314650278, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37350 + }, + { + "epoch": 0.37351, + "grad_norm": 1.7581211107441521, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 37351 + }, + { + "epoch": 0.37352, + "grad_norm": 1.4392971068354168, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 37352 + }, + { + "epoch": 0.37353, + "grad_norm": 1.506701293216335, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 37353 + }, + { + "epoch": 0.37354, + "grad_norm": 1.627284733232933, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 37354 + }, + { + "epoch": 0.37355, + "grad_norm": 1.3435122248335052, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 37355 + }, + { + "epoch": 0.37356, + "grad_norm": 1.4472846372645998, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 37356 + }, + { + "epoch": 0.37357, + "grad_norm": 1.7041740131532104, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37357 + }, + { + "epoch": 0.37358, + "grad_norm": 1.4045730697681718, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37358 + }, + { + "epoch": 0.37359, + "grad_norm": 1.4214532420163724, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 37359 + }, + { + "epoch": 0.3736, + "grad_norm": 1.573428091545066, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37360 + }, + { + "epoch": 0.37361, + "grad_norm": 1.2900441567467025, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 37361 + }, + { + "epoch": 0.37362, + "grad_norm": 1.7122677290412054, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 37362 + }, + { + "epoch": 0.37363, + "grad_norm": 1.4167882183420453, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 37363 + }, + { + "epoch": 0.37364, + "grad_norm": 1.4021223666226061, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 37364 + }, + { + "epoch": 0.37365, + "grad_norm": 1.5331178965118892, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 37365 + }, + { + "epoch": 0.37366, + "grad_norm": 1.5865347162203345, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 37366 + }, + { + "epoch": 0.37367, + "grad_norm": 1.497562132389858, + "learning_rate": 0.003, + "loss": 4.045, + "step": 37367 + }, + { + "epoch": 0.37368, + "grad_norm": 1.3506867420882853, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 37368 + }, + { + "epoch": 0.37369, + "grad_norm": 1.4781946533927033, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 37369 + }, + { + "epoch": 0.3737, + "grad_norm": 1.265752873024114, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 37370 + }, + { + "epoch": 0.37371, + "grad_norm": 1.9013003361702343, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 37371 + }, + { + "epoch": 0.37372, + "grad_norm": 1.4273275286810467, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 37372 + }, + { + "epoch": 0.37373, + "grad_norm": 1.9061303898451203, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 37373 + }, + { + "epoch": 0.37374, + "grad_norm": 1.3291728685190838, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 37374 + }, + { + "epoch": 0.37375, + "grad_norm": 1.9173394851735124, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 37375 + }, + { + "epoch": 0.37376, + "grad_norm": 1.3084082497280471, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 37376 + }, + { + "epoch": 0.37377, + "grad_norm": 1.7258049776487456, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37377 + }, + { + "epoch": 0.37378, + "grad_norm": 1.282956604012261, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37378 + }, + { + "epoch": 0.37379, + "grad_norm": 1.5761057733763941, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 37379 + }, + { + "epoch": 0.3738, + "grad_norm": 1.4862185573949103, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 37380 + }, + { + "epoch": 0.37381, + "grad_norm": 1.751677333664468, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 37381 + }, + { + "epoch": 0.37382, + "grad_norm": 1.219830742814978, + "learning_rate": 0.003, + "loss": 4.072, + "step": 37382 + }, + { + "epoch": 0.37383, + "grad_norm": 1.8754617071829656, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 37383 + }, + { + "epoch": 0.37384, + "grad_norm": 1.3467276122163365, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37384 + }, + { + "epoch": 0.37385, + "grad_norm": 1.7924135396554295, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 37385 + }, + { + "epoch": 0.37386, + "grad_norm": 1.4109934407001707, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37386 + }, + { + "epoch": 0.37387, + "grad_norm": 2.0233023217112156, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37387 + }, + { + "epoch": 0.37388, + "grad_norm": 1.3943829844924192, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 37388 + }, + { + "epoch": 0.37389, + "grad_norm": 1.6580304547792144, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37389 + }, + { + "epoch": 0.3739, + "grad_norm": 1.4679987938434274, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 37390 + }, + { + "epoch": 0.37391, + "grad_norm": 1.4956131373754533, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37391 + }, + { + "epoch": 0.37392, + "grad_norm": 1.57734650720773, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37392 + }, + { + "epoch": 0.37393, + "grad_norm": 1.4970773026790318, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 37393 + }, + { + "epoch": 0.37394, + "grad_norm": 1.6304276006654503, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37394 + }, + { + "epoch": 0.37395, + "grad_norm": 1.5441655325181478, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 37395 + }, + { + "epoch": 0.37396, + "grad_norm": 1.4974644838014683, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 37396 + }, + { + "epoch": 0.37397, + "grad_norm": 1.2699826435257828, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37397 + }, + { + "epoch": 0.37398, + "grad_norm": 1.7002682114916323, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 37398 + }, + { + "epoch": 0.37399, + "grad_norm": 1.4623780333859306, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37399 + }, + { + "epoch": 0.374, + "grad_norm": 1.4085937242116484, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37400 + }, + { + "epoch": 0.37401, + "grad_norm": 1.3569840112671963, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 37401 + }, + { + "epoch": 0.37402, + "grad_norm": 1.5201669960342705, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 37402 + }, + { + "epoch": 0.37403, + "grad_norm": 1.3633414140478886, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37403 + }, + { + "epoch": 0.37404, + "grad_norm": 1.8325817289959272, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37404 + }, + { + "epoch": 0.37405, + "grad_norm": 1.5840816399112703, + "learning_rate": 0.003, + "loss": 4.028, + "step": 37405 + }, + { + "epoch": 0.37406, + "grad_norm": 1.3503535665855904, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 37406 + }, + { + "epoch": 0.37407, + "grad_norm": 1.626017998851459, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 37407 + }, + { + "epoch": 0.37408, + "grad_norm": 1.4605882905608083, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 37408 + }, + { + "epoch": 0.37409, + "grad_norm": 1.513856842022384, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 37409 + }, + { + "epoch": 0.3741, + "grad_norm": 1.2695626959492998, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 37410 + }, + { + "epoch": 0.37411, + "grad_norm": 1.527502657497381, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 37411 + }, + { + "epoch": 0.37412, + "grad_norm": 1.3926735978145948, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 37412 + }, + { + "epoch": 0.37413, + "grad_norm": 1.133480363642256, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37413 + }, + { + "epoch": 0.37414, + "grad_norm": 1.9421497907152079, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 37414 + }, + { + "epoch": 0.37415, + "grad_norm": 1.3101366046001093, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 37415 + }, + { + "epoch": 0.37416, + "grad_norm": 1.8715352496417905, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 37416 + }, + { + "epoch": 0.37417, + "grad_norm": 1.4081476411520077, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37417 + }, + { + "epoch": 0.37418, + "grad_norm": 1.7046477183431539, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37418 + }, + { + "epoch": 0.37419, + "grad_norm": 1.6203848980503408, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 37419 + }, + { + "epoch": 0.3742, + "grad_norm": 1.3641793265494107, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 37420 + }, + { + "epoch": 0.37421, + "grad_norm": 1.7042965362106466, + "learning_rate": 0.003, + "loss": 4.07, + "step": 37421 + }, + { + "epoch": 0.37422, + "grad_norm": 1.4600740819243918, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 37422 + }, + { + "epoch": 0.37423, + "grad_norm": 1.5840553572521758, + "learning_rate": 0.003, + "loss": 4.047, + "step": 37423 + }, + { + "epoch": 0.37424, + "grad_norm": 1.5448903085850496, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 37424 + }, + { + "epoch": 0.37425, + "grad_norm": 1.604604145196993, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 37425 + }, + { + "epoch": 0.37426, + "grad_norm": 1.6435751924931365, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 37426 + }, + { + "epoch": 0.37427, + "grad_norm": 1.3479726226626503, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 37427 + }, + { + "epoch": 0.37428, + "grad_norm": 1.456986641660377, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 37428 + }, + { + "epoch": 0.37429, + "grad_norm": 1.3962958949059943, + "learning_rate": 0.003, + "loss": 4.066, + "step": 37429 + }, + { + "epoch": 0.3743, + "grad_norm": 1.5659840847435536, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 37430 + }, + { + "epoch": 0.37431, + "grad_norm": 1.5581541166640553, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 37431 + }, + { + "epoch": 0.37432, + "grad_norm": 1.4723165749296983, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 37432 + }, + { + "epoch": 0.37433, + "grad_norm": 1.2231372064720587, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37433 + }, + { + "epoch": 0.37434, + "grad_norm": 1.5421144226536248, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 37434 + }, + { + "epoch": 0.37435, + "grad_norm": 1.5566782835980748, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 37435 + }, + { + "epoch": 0.37436, + "grad_norm": 1.555434215863295, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 37436 + }, + { + "epoch": 0.37437, + "grad_norm": 1.2704455730568973, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 37437 + }, + { + "epoch": 0.37438, + "grad_norm": 1.7709510050019548, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 37438 + }, + { + "epoch": 0.37439, + "grad_norm": 1.313148176233231, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37439 + }, + { + "epoch": 0.3744, + "grad_norm": 1.656246432162967, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37440 + }, + { + "epoch": 0.37441, + "grad_norm": 1.303735377720755, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 37441 + }, + { + "epoch": 0.37442, + "grad_norm": 1.7283668690986185, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 37442 + }, + { + "epoch": 0.37443, + "grad_norm": 1.5179644674121386, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 37443 + }, + { + "epoch": 0.37444, + "grad_norm": 1.3606259481919762, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 37444 + }, + { + "epoch": 0.37445, + "grad_norm": 1.556232213659483, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37445 + }, + { + "epoch": 0.37446, + "grad_norm": 1.109147707921754, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 37446 + }, + { + "epoch": 0.37447, + "grad_norm": 1.7863069169077124, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 37447 + }, + { + "epoch": 0.37448, + "grad_norm": 1.1198399393074758, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 37448 + }, + { + "epoch": 0.37449, + "grad_norm": 2.007620015656822, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 37449 + }, + { + "epoch": 0.3745, + "grad_norm": 1.4356867002467646, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37450 + }, + { + "epoch": 0.37451, + "grad_norm": 1.340661737378064, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 37451 + }, + { + "epoch": 0.37452, + "grad_norm": 1.7017944938919305, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 37452 + }, + { + "epoch": 0.37453, + "grad_norm": 1.3494872554921469, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 37453 + }, + { + "epoch": 0.37454, + "grad_norm": 1.5765383758600218, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37454 + }, + { + "epoch": 0.37455, + "grad_norm": 1.5185253749241243, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 37455 + }, + { + "epoch": 0.37456, + "grad_norm": 1.607893771808914, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37456 + }, + { + "epoch": 0.37457, + "grad_norm": 1.7017806378404288, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 37457 + }, + { + "epoch": 0.37458, + "grad_norm": 1.4897259647550756, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 37458 + }, + { + "epoch": 0.37459, + "grad_norm": 1.6206081347714185, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37459 + }, + { + "epoch": 0.3746, + "grad_norm": 1.4579809243219122, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 37460 + }, + { + "epoch": 0.37461, + "grad_norm": 1.6674725086638214, + "learning_rate": 0.003, + "loss": 4.024, + "step": 37461 + }, + { + "epoch": 0.37462, + "grad_norm": 1.2909578636046541, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 37462 + }, + { + "epoch": 0.37463, + "grad_norm": 1.9240583577159185, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37463 + }, + { + "epoch": 0.37464, + "grad_norm": 1.1365444942391816, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 37464 + }, + { + "epoch": 0.37465, + "grad_norm": 1.46677713983129, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37465 + }, + { + "epoch": 0.37466, + "grad_norm": 1.6633924844909787, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 37466 + }, + { + "epoch": 0.37467, + "grad_norm": 1.3841850121614372, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37467 + }, + { + "epoch": 0.37468, + "grad_norm": 1.4309504128937423, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37468 + }, + { + "epoch": 0.37469, + "grad_norm": 1.635370158628571, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 37469 + }, + { + "epoch": 0.3747, + "grad_norm": 1.5591015258930532, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 37470 + }, + { + "epoch": 0.37471, + "grad_norm": 1.4159004330766678, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 37471 + }, + { + "epoch": 0.37472, + "grad_norm": 1.4307142660146668, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 37472 + }, + { + "epoch": 0.37473, + "grad_norm": 1.800022823200498, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 37473 + }, + { + "epoch": 0.37474, + "grad_norm": 1.4768266061239357, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37474 + }, + { + "epoch": 0.37475, + "grad_norm": 1.4106232513136605, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 37475 + }, + { + "epoch": 0.37476, + "grad_norm": 1.3593050132872804, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 37476 + }, + { + "epoch": 0.37477, + "grad_norm": 1.4079077196922227, + "learning_rate": 0.003, + "loss": 4.074, + "step": 37477 + }, + { + "epoch": 0.37478, + "grad_norm": 1.6946250014639597, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 37478 + }, + { + "epoch": 0.37479, + "grad_norm": 1.734946958757428, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 37479 + }, + { + "epoch": 0.3748, + "grad_norm": 1.8191057813716511, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37480 + }, + { + "epoch": 0.37481, + "grad_norm": 1.4409739972251134, + "learning_rate": 0.003, + "loss": 4.042, + "step": 37481 + }, + { + "epoch": 0.37482, + "grad_norm": 1.4690053134824341, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 37482 + }, + { + "epoch": 0.37483, + "grad_norm": 1.329260889558525, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 37483 + }, + { + "epoch": 0.37484, + "grad_norm": 1.5912690694283316, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 37484 + }, + { + "epoch": 0.37485, + "grad_norm": 1.485259492364329, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37485 + }, + { + "epoch": 0.37486, + "grad_norm": 1.4249736144586789, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 37486 + }, + { + "epoch": 0.37487, + "grad_norm": 1.3881066041627235, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 37487 + }, + { + "epoch": 0.37488, + "grad_norm": 1.719891630699829, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 37488 + }, + { + "epoch": 0.37489, + "grad_norm": 1.2954991603498955, + "learning_rate": 0.003, + "loss": 4.026, + "step": 37489 + }, + { + "epoch": 0.3749, + "grad_norm": 2.2966900629753626, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 37490 + }, + { + "epoch": 0.37491, + "grad_norm": 1.602821942542837, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 37491 + }, + { + "epoch": 0.37492, + "grad_norm": 1.4215632759005523, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 37492 + }, + { + "epoch": 0.37493, + "grad_norm": 1.6778683607384195, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 37493 + }, + { + "epoch": 0.37494, + "grad_norm": 1.5139702764605052, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 37494 + }, + { + "epoch": 0.37495, + "grad_norm": 1.2848163885242145, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 37495 + }, + { + "epoch": 0.37496, + "grad_norm": 1.727278286510156, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 37496 + }, + { + "epoch": 0.37497, + "grad_norm": 1.268277508599558, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 37497 + }, + { + "epoch": 0.37498, + "grad_norm": 1.5242399400041544, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 37498 + }, + { + "epoch": 0.37499, + "grad_norm": 1.522794590424317, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 37499 + }, + { + "epoch": 0.375, + "grad_norm": 1.5744964064286535, + "learning_rate": 0.003, + "loss": 4.044, + "step": 37500 + }, + { + "epoch": 0.37501, + "grad_norm": 1.3604006371221267, + "learning_rate": 0.003, + "loss": 4.04, + "step": 37501 + }, + { + "epoch": 0.37502, + "grad_norm": 1.587413894262418, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 37502 + }, + { + "epoch": 0.37503, + "grad_norm": 1.3271146870136539, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 37503 + }, + { + "epoch": 0.37504, + "grad_norm": 1.3471416132173073, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 37504 + }, + { + "epoch": 0.37505, + "grad_norm": 1.5903063019515988, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 37505 + }, + { + "epoch": 0.37506, + "grad_norm": 1.2912077425732227, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 37506 + }, + { + "epoch": 0.37507, + "grad_norm": 1.849607724612634, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 37507 + }, + { + "epoch": 0.37508, + "grad_norm": 1.5007532755900892, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 37508 + }, + { + "epoch": 0.37509, + "grad_norm": 1.5218563917985366, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 37509 + }, + { + "epoch": 0.3751, + "grad_norm": 1.5914058186910716, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37510 + }, + { + "epoch": 0.37511, + "grad_norm": 1.3124258643343263, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 37511 + }, + { + "epoch": 0.37512, + "grad_norm": 1.5571847579869933, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 37512 + }, + { + "epoch": 0.37513, + "grad_norm": 1.2400510561070615, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 37513 + }, + { + "epoch": 0.37514, + "grad_norm": 1.4133506792639874, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 37514 + }, + { + "epoch": 0.37515, + "grad_norm": 1.5928598555638698, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37515 + }, + { + "epoch": 0.37516, + "grad_norm": 1.4768215399500995, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37516 + }, + { + "epoch": 0.37517, + "grad_norm": 1.6451315818961534, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 37517 + }, + { + "epoch": 0.37518, + "grad_norm": 2.138455791400785, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 37518 + }, + { + "epoch": 0.37519, + "grad_norm": 1.3486846603140275, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37519 + }, + { + "epoch": 0.3752, + "grad_norm": 1.5353839821138402, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 37520 + }, + { + "epoch": 0.37521, + "grad_norm": 1.3950089332513875, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 37521 + }, + { + "epoch": 0.37522, + "grad_norm": 1.8069046329136875, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37522 + }, + { + "epoch": 0.37523, + "grad_norm": 1.3905789098658585, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 37523 + }, + { + "epoch": 0.37524, + "grad_norm": 1.5571317913210874, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 37524 + }, + { + "epoch": 0.37525, + "grad_norm": 1.5607376939752289, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 37525 + }, + { + "epoch": 0.37526, + "grad_norm": 1.1489393498508813, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 37526 + }, + { + "epoch": 0.37527, + "grad_norm": 1.827673802643873, + "learning_rate": 0.003, + "loss": 4.063, + "step": 37527 + }, + { + "epoch": 0.37528, + "grad_norm": 1.217108030116619, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 37528 + }, + { + "epoch": 0.37529, + "grad_norm": 1.5840175125556841, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37529 + }, + { + "epoch": 0.3753, + "grad_norm": 1.2892054911019641, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 37530 + }, + { + "epoch": 0.37531, + "grad_norm": 1.558667033283634, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37531 + }, + { + "epoch": 0.37532, + "grad_norm": 1.4406368215887257, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 37532 + }, + { + "epoch": 0.37533, + "grad_norm": 1.434038164241871, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 37533 + }, + { + "epoch": 0.37534, + "grad_norm": 1.4563172499437216, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 37534 + }, + { + "epoch": 0.37535, + "grad_norm": 1.8186211012285014, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 37535 + }, + { + "epoch": 0.37536, + "grad_norm": 1.611272329120169, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37536 + }, + { + "epoch": 0.37537, + "grad_norm": 1.5115748524357935, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 37537 + }, + { + "epoch": 0.37538, + "grad_norm": 1.5366986952548822, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37538 + }, + { + "epoch": 0.37539, + "grad_norm": 1.576371898504236, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 37539 + }, + { + "epoch": 0.3754, + "grad_norm": 1.56986544017252, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37540 + }, + { + "epoch": 0.37541, + "grad_norm": 1.288072810229134, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37541 + }, + { + "epoch": 0.37542, + "grad_norm": 1.6260050802474364, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 37542 + }, + { + "epoch": 0.37543, + "grad_norm": 1.4154184969901402, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 37543 + }, + { + "epoch": 0.37544, + "grad_norm": 1.5172228553111387, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 37544 + }, + { + "epoch": 0.37545, + "grad_norm": 1.8570297958008313, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37545 + }, + { + "epoch": 0.37546, + "grad_norm": 1.4055534101809728, + "learning_rate": 0.003, + "loss": 4.018, + "step": 37546 + }, + { + "epoch": 0.37547, + "grad_norm": 1.8346778005931914, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 37547 + }, + { + "epoch": 0.37548, + "grad_norm": 1.3985036375041326, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37548 + }, + { + "epoch": 0.37549, + "grad_norm": 1.7678366647100607, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 37549 + }, + { + "epoch": 0.3755, + "grad_norm": 1.5957537074815333, + "learning_rate": 0.003, + "loss": 4.022, + "step": 37550 + }, + { + "epoch": 0.37551, + "grad_norm": 1.4396867663181114, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37551 + }, + { + "epoch": 0.37552, + "grad_norm": 1.5391838969558227, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 37552 + }, + { + "epoch": 0.37553, + "grad_norm": 1.3138459652234127, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37553 + }, + { + "epoch": 0.37554, + "grad_norm": 1.61855270709907, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37554 + }, + { + "epoch": 0.37555, + "grad_norm": 1.2925651512567662, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 37555 + }, + { + "epoch": 0.37556, + "grad_norm": 1.6249729246196336, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 37556 + }, + { + "epoch": 0.37557, + "grad_norm": 1.345477631215769, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 37557 + }, + { + "epoch": 0.37558, + "grad_norm": 1.3453421505391623, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 37558 + }, + { + "epoch": 0.37559, + "grad_norm": 1.4512777345686485, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37559 + }, + { + "epoch": 0.3756, + "grad_norm": 1.653589250915749, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 37560 + }, + { + "epoch": 0.37561, + "grad_norm": 1.274678944018881, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 37561 + }, + { + "epoch": 0.37562, + "grad_norm": 1.8713318619794206, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 37562 + }, + { + "epoch": 0.37563, + "grad_norm": 1.2822346764370627, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 37563 + }, + { + "epoch": 0.37564, + "grad_norm": 1.532032709942697, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37564 + }, + { + "epoch": 0.37565, + "grad_norm": 1.5368168174945667, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37565 + }, + { + "epoch": 0.37566, + "grad_norm": 2.157787869781742, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37566 + }, + { + "epoch": 0.37567, + "grad_norm": 1.3357892529496351, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 37567 + }, + { + "epoch": 0.37568, + "grad_norm": 1.5698669236085911, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37568 + }, + { + "epoch": 0.37569, + "grad_norm": 1.6030393533243381, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37569 + }, + { + "epoch": 0.3757, + "grad_norm": 1.3159345227521773, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 37570 + }, + { + "epoch": 0.37571, + "grad_norm": 1.6874471023241748, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 37571 + }, + { + "epoch": 0.37572, + "grad_norm": 1.3779971502637296, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37572 + }, + { + "epoch": 0.37573, + "grad_norm": 1.784425618911355, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 37573 + }, + { + "epoch": 0.37574, + "grad_norm": 1.2460485879230625, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 37574 + }, + { + "epoch": 0.37575, + "grad_norm": 1.69692410220807, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37575 + }, + { + "epoch": 0.37576, + "grad_norm": 1.51391185869218, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 37576 + }, + { + "epoch": 0.37577, + "grad_norm": 1.5610864160900957, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37577 + }, + { + "epoch": 0.37578, + "grad_norm": 1.6318862695872738, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 37578 + }, + { + "epoch": 0.37579, + "grad_norm": 1.5355459698961096, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37579 + }, + { + "epoch": 0.3758, + "grad_norm": 1.4413833409444294, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37580 + }, + { + "epoch": 0.37581, + "grad_norm": 1.6871266422895275, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 37581 + }, + { + "epoch": 0.37582, + "grad_norm": 1.475615638895395, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 37582 + }, + { + "epoch": 0.37583, + "grad_norm": 1.5630622891983796, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37583 + }, + { + "epoch": 0.37584, + "grad_norm": 1.4368068955714237, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 37584 + }, + { + "epoch": 0.37585, + "grad_norm": 1.4976876714127962, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37585 + }, + { + "epoch": 0.37586, + "grad_norm": 1.3613142230405473, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 37586 + }, + { + "epoch": 0.37587, + "grad_norm": 1.5380592534368425, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 37587 + }, + { + "epoch": 0.37588, + "grad_norm": 1.4725783774746137, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 37588 + }, + { + "epoch": 0.37589, + "grad_norm": 1.6417879484443518, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 37589 + }, + { + "epoch": 0.3759, + "grad_norm": 1.84097106315218, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 37590 + }, + { + "epoch": 0.37591, + "grad_norm": 1.4802536834271307, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 37591 + }, + { + "epoch": 0.37592, + "grad_norm": 1.5230022246662747, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37592 + }, + { + "epoch": 0.37593, + "grad_norm": 1.5322594376464973, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 37593 + }, + { + "epoch": 0.37594, + "grad_norm": 1.7490816053994835, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 37594 + }, + { + "epoch": 0.37595, + "grad_norm": 1.4404203848781039, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 37595 + }, + { + "epoch": 0.37596, + "grad_norm": 1.750937352937849, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37596 + }, + { + "epoch": 0.37597, + "grad_norm": 1.3232028922714043, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37597 + }, + { + "epoch": 0.37598, + "grad_norm": 1.5011890443782328, + "learning_rate": 0.003, + "loss": 4.025, + "step": 37598 + }, + { + "epoch": 0.37599, + "grad_norm": 1.3299597377239336, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 37599 + }, + { + "epoch": 0.376, + "grad_norm": 1.5063438266273954, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37600 + }, + { + "epoch": 0.37601, + "grad_norm": 1.2399255511809908, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37601 + }, + { + "epoch": 0.37602, + "grad_norm": 1.6865249820516806, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 37602 + }, + { + "epoch": 0.37603, + "grad_norm": 1.4703617871744903, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 37603 + }, + { + "epoch": 0.37604, + "grad_norm": 1.7008845517776305, + "learning_rate": 0.003, + "loss": 4.045, + "step": 37604 + }, + { + "epoch": 0.37605, + "grad_norm": 1.6561888230632664, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 37605 + }, + { + "epoch": 0.37606, + "grad_norm": 1.77769495907868, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 37606 + }, + { + "epoch": 0.37607, + "grad_norm": 1.4071331928974637, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 37607 + }, + { + "epoch": 0.37608, + "grad_norm": 1.3473718045680192, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37608 + }, + { + "epoch": 0.37609, + "grad_norm": 1.8632842660943798, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 37609 + }, + { + "epoch": 0.3761, + "grad_norm": 1.4253197193055833, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37610 + }, + { + "epoch": 0.37611, + "grad_norm": 1.4179164159152584, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37611 + }, + { + "epoch": 0.37612, + "grad_norm": 1.533772130085798, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37612 + }, + { + "epoch": 0.37613, + "grad_norm": 1.3252921162473983, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 37613 + }, + { + "epoch": 0.37614, + "grad_norm": 1.4382994182007687, + "learning_rate": 0.003, + "loss": 4.066, + "step": 37614 + }, + { + "epoch": 0.37615, + "grad_norm": 1.4551003233264952, + "learning_rate": 0.003, + "loss": 4.041, + "step": 37615 + }, + { + "epoch": 0.37616, + "grad_norm": 1.2601240459585368, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37616 + }, + { + "epoch": 0.37617, + "grad_norm": 1.568197855541647, + "learning_rate": 0.003, + "loss": 4.02, + "step": 37617 + }, + { + "epoch": 0.37618, + "grad_norm": 1.5203915111080273, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 37618 + }, + { + "epoch": 0.37619, + "grad_norm": 1.3488166783109459, + "learning_rate": 0.003, + "loss": 4.03, + "step": 37619 + }, + { + "epoch": 0.3762, + "grad_norm": 1.75137441491537, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 37620 + }, + { + "epoch": 0.37621, + "grad_norm": 1.661668893819902, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 37621 + }, + { + "epoch": 0.37622, + "grad_norm": 1.795071433850129, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 37622 + }, + { + "epoch": 0.37623, + "grad_norm": 1.6735351973019859, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37623 + }, + { + "epoch": 0.37624, + "grad_norm": 1.246154970777671, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37624 + }, + { + "epoch": 0.37625, + "grad_norm": 1.5433931645646017, + "learning_rate": 0.003, + "loss": 4.051, + "step": 37625 + }, + { + "epoch": 0.37626, + "grad_norm": 1.5349465176534864, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37626 + }, + { + "epoch": 0.37627, + "grad_norm": 1.6385378540995046, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37627 + }, + { + "epoch": 0.37628, + "grad_norm": 1.6373360957663197, + "learning_rate": 0.003, + "loss": 4.029, + "step": 37628 + }, + { + "epoch": 0.37629, + "grad_norm": 1.4686268370095843, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 37629 + }, + { + "epoch": 0.3763, + "grad_norm": 1.5721871031627006, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37630 + }, + { + "epoch": 0.37631, + "grad_norm": 1.27615211150912, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37631 + }, + { + "epoch": 0.37632, + "grad_norm": 1.4731141745573078, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 37632 + }, + { + "epoch": 0.37633, + "grad_norm": 1.3725697325995845, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37633 + }, + { + "epoch": 0.37634, + "grad_norm": 1.8411588797635265, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 37634 + }, + { + "epoch": 0.37635, + "grad_norm": 1.5084862196682152, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 37635 + }, + { + "epoch": 0.37636, + "grad_norm": 1.9911778535595532, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 37636 + }, + { + "epoch": 0.37637, + "grad_norm": 1.205909792642477, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37637 + }, + { + "epoch": 0.37638, + "grad_norm": 1.396765381758369, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 37638 + }, + { + "epoch": 0.37639, + "grad_norm": 1.8887782988217219, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 37639 + }, + { + "epoch": 0.3764, + "grad_norm": 1.5895522762235448, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 37640 + }, + { + "epoch": 0.37641, + "grad_norm": 1.6907537277020634, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 37641 + }, + { + "epoch": 0.37642, + "grad_norm": 1.2862705059037085, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 37642 + }, + { + "epoch": 0.37643, + "grad_norm": 1.591402662921012, + "learning_rate": 0.003, + "loss": 4.075, + "step": 37643 + }, + { + "epoch": 0.37644, + "grad_norm": 1.3399699407693828, + "learning_rate": 0.003, + "loss": 4.036, + "step": 37644 + }, + { + "epoch": 0.37645, + "grad_norm": 1.5588427391715896, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 37645 + }, + { + "epoch": 0.37646, + "grad_norm": 1.593180977267478, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37646 + }, + { + "epoch": 0.37647, + "grad_norm": 1.2813888432803968, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 37647 + }, + { + "epoch": 0.37648, + "grad_norm": 1.7039973095182792, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 37648 + }, + { + "epoch": 0.37649, + "grad_norm": 1.3842678928993821, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 37649 + }, + { + "epoch": 0.3765, + "grad_norm": 1.7826049698544162, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 37650 + }, + { + "epoch": 0.37651, + "grad_norm": 1.0601561143874016, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37651 + }, + { + "epoch": 0.37652, + "grad_norm": 1.9044369532550576, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 37652 + }, + { + "epoch": 0.37653, + "grad_norm": 1.2108921383864646, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 37653 + }, + { + "epoch": 0.37654, + "grad_norm": 1.8901941351618408, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 37654 + }, + { + "epoch": 0.37655, + "grad_norm": 1.4462751636052948, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 37655 + }, + { + "epoch": 0.37656, + "grad_norm": 1.5672920884111827, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37656 + }, + { + "epoch": 0.37657, + "grad_norm": 1.9036625752941325, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37657 + }, + { + "epoch": 0.37658, + "grad_norm": 1.349701139746263, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 37658 + }, + { + "epoch": 0.37659, + "grad_norm": 1.5910602898077029, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37659 + }, + { + "epoch": 0.3766, + "grad_norm": 1.410139235425275, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 37660 + }, + { + "epoch": 0.37661, + "grad_norm": 1.633466124186963, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 37661 + }, + { + "epoch": 0.37662, + "grad_norm": 1.2778798353884284, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 37662 + }, + { + "epoch": 0.37663, + "grad_norm": 1.6860454379971173, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 37663 + }, + { + "epoch": 0.37664, + "grad_norm": 1.315892511557433, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 37664 + }, + { + "epoch": 0.37665, + "grad_norm": 1.5063120234031318, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 37665 + }, + { + "epoch": 0.37666, + "grad_norm": 1.8221616009314217, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 37666 + }, + { + "epoch": 0.37667, + "grad_norm": 1.7191392617650385, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 37667 + }, + { + "epoch": 0.37668, + "grad_norm": 1.180233184997646, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37668 + }, + { + "epoch": 0.37669, + "grad_norm": 1.6459096895575436, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 37669 + }, + { + "epoch": 0.3767, + "grad_norm": 1.617823896466139, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37670 + }, + { + "epoch": 0.37671, + "grad_norm": 1.4615457560481613, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37671 + }, + { + "epoch": 0.37672, + "grad_norm": 1.2603846741301878, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 37672 + }, + { + "epoch": 0.37673, + "grad_norm": 1.6893427960678487, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 37673 + }, + { + "epoch": 0.37674, + "grad_norm": 1.3408972902648384, + "learning_rate": 0.003, + "loss": 4.084, + "step": 37674 + }, + { + "epoch": 0.37675, + "grad_norm": 1.4382014008976785, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 37675 + }, + { + "epoch": 0.37676, + "grad_norm": 1.5638838556689414, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37676 + }, + { + "epoch": 0.37677, + "grad_norm": 1.5090364532007599, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37677 + }, + { + "epoch": 0.37678, + "grad_norm": 1.6376299327338806, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 37678 + }, + { + "epoch": 0.37679, + "grad_norm": 1.4813984178585404, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 37679 + }, + { + "epoch": 0.3768, + "grad_norm": 1.6795781795799518, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 37680 + }, + { + "epoch": 0.37681, + "grad_norm": 1.407199715759721, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 37681 + }, + { + "epoch": 0.37682, + "grad_norm": 1.3510188411340467, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 37682 + }, + { + "epoch": 0.37683, + "grad_norm": 1.7369212669262193, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 37683 + }, + { + "epoch": 0.37684, + "grad_norm": 1.4195498478323483, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37684 + }, + { + "epoch": 0.37685, + "grad_norm": 1.399831522339098, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 37685 + }, + { + "epoch": 0.37686, + "grad_norm": 1.5556551239418606, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37686 + }, + { + "epoch": 0.37687, + "grad_norm": 1.3667517166689165, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 37687 + }, + { + "epoch": 0.37688, + "grad_norm": 1.5317169755669096, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 37688 + }, + { + "epoch": 0.37689, + "grad_norm": 1.616100678590844, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 37689 + }, + { + "epoch": 0.3769, + "grad_norm": 1.7797765415810889, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 37690 + }, + { + "epoch": 0.37691, + "grad_norm": 1.5442215633764178, + "learning_rate": 0.003, + "loss": 4.048, + "step": 37691 + }, + { + "epoch": 0.37692, + "grad_norm": 1.5913432974380508, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 37692 + }, + { + "epoch": 0.37693, + "grad_norm": 1.7887248464631742, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 37693 + }, + { + "epoch": 0.37694, + "grad_norm": 1.0614225540869573, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 37694 + }, + { + "epoch": 0.37695, + "grad_norm": 1.6075642564961425, + "learning_rate": 0.003, + "loss": 4.066, + "step": 37695 + }, + { + "epoch": 0.37696, + "grad_norm": 1.2894160494200668, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 37696 + }, + { + "epoch": 0.37697, + "grad_norm": 1.8877977355606677, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 37697 + }, + { + "epoch": 0.37698, + "grad_norm": 1.355977086410518, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 37698 + }, + { + "epoch": 0.37699, + "grad_norm": 1.7517886594484375, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 37699 + }, + { + "epoch": 0.377, + "grad_norm": 1.5753979907321116, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 37700 + }, + { + "epoch": 0.37701, + "grad_norm": 1.513814791526171, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 37701 + }, + { + "epoch": 0.37702, + "grad_norm": 1.63013315075503, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 37702 + }, + { + "epoch": 0.37703, + "grad_norm": 1.436819027274203, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 37703 + }, + { + "epoch": 0.37704, + "grad_norm": 1.4964465640763616, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 37704 + }, + { + "epoch": 0.37705, + "grad_norm": 1.5622124460924423, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 37705 + }, + { + "epoch": 0.37706, + "grad_norm": 1.6516591660835003, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 37706 + }, + { + "epoch": 0.37707, + "grad_norm": 1.4755025032065248, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 37707 + }, + { + "epoch": 0.37708, + "grad_norm": 1.6198536024043833, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 37708 + }, + { + "epoch": 0.37709, + "grad_norm": 1.3419477360326293, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37709 + }, + { + "epoch": 0.3771, + "grad_norm": 1.363877020474922, + "learning_rate": 0.003, + "loss": 4.029, + "step": 37710 + }, + { + "epoch": 0.37711, + "grad_norm": 1.4437485519382391, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 37711 + }, + { + "epoch": 0.37712, + "grad_norm": 1.5301634294259827, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 37712 + }, + { + "epoch": 0.37713, + "grad_norm": 1.7356977332143413, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 37713 + }, + { + "epoch": 0.37714, + "grad_norm": 1.3859657872037603, + "learning_rate": 0.003, + "loss": 4.001, + "step": 37714 + }, + { + "epoch": 0.37715, + "grad_norm": 1.518997695142908, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 37715 + }, + { + "epoch": 0.37716, + "grad_norm": 1.426619587171013, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 37716 + }, + { + "epoch": 0.37717, + "grad_norm": 1.2880521202621267, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 37717 + }, + { + "epoch": 0.37718, + "grad_norm": 1.4157649875136697, + "learning_rate": 0.003, + "loss": 4.055, + "step": 37718 + }, + { + "epoch": 0.37719, + "grad_norm": 1.4896912180042052, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 37719 + }, + { + "epoch": 0.3772, + "grad_norm": 1.3792122838981835, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 37720 + }, + { + "epoch": 0.37721, + "grad_norm": 1.3339314694112128, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 37721 + }, + { + "epoch": 0.37722, + "grad_norm": 1.7613784845186728, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37722 + }, + { + "epoch": 0.37723, + "grad_norm": 1.5390197579153075, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 37723 + }, + { + "epoch": 0.37724, + "grad_norm": 1.5339703807420304, + "learning_rate": 0.003, + "loss": 4.044, + "step": 37724 + }, + { + "epoch": 0.37725, + "grad_norm": 1.2304769812597265, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 37725 + }, + { + "epoch": 0.37726, + "grad_norm": 1.8977571145378627, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37726 + }, + { + "epoch": 0.37727, + "grad_norm": 1.1683112083179115, + "learning_rate": 0.003, + "loss": 4.028, + "step": 37727 + }, + { + "epoch": 0.37728, + "grad_norm": 1.8487838836180053, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 37728 + }, + { + "epoch": 0.37729, + "grad_norm": 1.2434902072629, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 37729 + }, + { + "epoch": 0.3773, + "grad_norm": 1.752867253537864, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 37730 + }, + { + "epoch": 0.37731, + "grad_norm": 1.3113544729004627, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 37731 + }, + { + "epoch": 0.37732, + "grad_norm": 1.7163247422592058, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 37732 + }, + { + "epoch": 0.37733, + "grad_norm": 1.3070010969972201, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 37733 + }, + { + "epoch": 0.37734, + "grad_norm": 1.6711107972701291, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 37734 + }, + { + "epoch": 0.37735, + "grad_norm": 1.3015809446524869, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 37735 + }, + { + "epoch": 0.37736, + "grad_norm": 1.8394861790971702, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 37736 + }, + { + "epoch": 0.37737, + "grad_norm": 1.517099955875972, + "learning_rate": 0.003, + "loss": 4.051, + "step": 37737 + }, + { + "epoch": 0.37738, + "grad_norm": 1.821521407951394, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 37738 + }, + { + "epoch": 0.37739, + "grad_norm": 1.5428325445813578, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 37739 + }, + { + "epoch": 0.3774, + "grad_norm": 1.488070965731017, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 37740 + }, + { + "epoch": 0.37741, + "grad_norm": 1.5104564267474565, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37741 + }, + { + "epoch": 0.37742, + "grad_norm": 1.7137616341877073, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 37742 + }, + { + "epoch": 0.37743, + "grad_norm": 1.4995188344268364, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 37743 + }, + { + "epoch": 0.37744, + "grad_norm": 1.6103728972345432, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37744 + }, + { + "epoch": 0.37745, + "grad_norm": 1.7794120239841422, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 37745 + }, + { + "epoch": 0.37746, + "grad_norm": 1.158073631939096, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 37746 + }, + { + "epoch": 0.37747, + "grad_norm": 1.6686846256577903, + "learning_rate": 0.003, + "loss": 4.037, + "step": 37747 + }, + { + "epoch": 0.37748, + "grad_norm": 1.450656059501426, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 37748 + }, + { + "epoch": 0.37749, + "grad_norm": 1.6202987863070593, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 37749 + }, + { + "epoch": 0.3775, + "grad_norm": 1.3536500527328408, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 37750 + }, + { + "epoch": 0.37751, + "grad_norm": 1.530658236813148, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37751 + }, + { + "epoch": 0.37752, + "grad_norm": 1.7184254459912147, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 37752 + }, + { + "epoch": 0.37753, + "grad_norm": 1.3200901511399743, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 37753 + }, + { + "epoch": 0.37754, + "grad_norm": 1.4389972053139086, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 37754 + }, + { + "epoch": 0.37755, + "grad_norm": 1.620331281611047, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 37755 + }, + { + "epoch": 0.37756, + "grad_norm": 1.3652431806655698, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 37756 + }, + { + "epoch": 0.37757, + "grad_norm": 1.7971057531512946, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 37757 + }, + { + "epoch": 0.37758, + "grad_norm": 1.2352639555160203, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 37758 + }, + { + "epoch": 0.37759, + "grad_norm": 1.7134589904745343, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 37759 + }, + { + "epoch": 0.3776, + "grad_norm": 1.3525609750959906, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 37760 + }, + { + "epoch": 0.37761, + "grad_norm": 1.558409033882503, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 37761 + }, + { + "epoch": 0.37762, + "grad_norm": 1.3833777939995322, + "learning_rate": 0.003, + "loss": 4.064, + "step": 37762 + }, + { + "epoch": 0.37763, + "grad_norm": 1.5541145801445067, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 37763 + }, + { + "epoch": 0.37764, + "grad_norm": 1.4755523333687406, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 37764 + }, + { + "epoch": 0.37765, + "grad_norm": 1.6864367855580915, + "learning_rate": 0.003, + "loss": 4.035, + "step": 37765 + }, + { + "epoch": 0.37766, + "grad_norm": 1.6145324286354676, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 37766 + }, + { + "epoch": 0.37767, + "grad_norm": 1.613306420164985, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 37767 + }, + { + "epoch": 0.37768, + "grad_norm": 1.3515452989192671, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 37768 + }, + { + "epoch": 0.37769, + "grad_norm": 1.6036397869547647, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 37769 + }, + { + "epoch": 0.3777, + "grad_norm": 1.4199053229499623, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 37770 + }, + { + "epoch": 0.37771, + "grad_norm": 1.6473993986484603, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 37771 + }, + { + "epoch": 0.37772, + "grad_norm": 1.216367911802994, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 37772 + }, + { + "epoch": 0.37773, + "grad_norm": 1.7649021642527112, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 37773 + }, + { + "epoch": 0.37774, + "grad_norm": 1.4685461126771158, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 37774 + }, + { + "epoch": 0.37775, + "grad_norm": 1.959393818197071, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 37775 + }, + { + "epoch": 0.37776, + "grad_norm": 1.2836620438238249, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 37776 + }, + { + "epoch": 0.37777, + "grad_norm": 1.8068277776269894, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 37777 + }, + { + "epoch": 0.37778, + "grad_norm": 1.3928488662212362, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 37778 + }, + { + "epoch": 0.37779, + "grad_norm": 1.5416458112318068, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 37779 + }, + { + "epoch": 0.3778, + "grad_norm": 1.7253521678640749, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 37780 + }, + { + "epoch": 0.37781, + "grad_norm": 1.7416511403055013, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 37781 + }, + { + "epoch": 0.37782, + "grad_norm": 1.5494895966721427, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 37782 + }, + { + "epoch": 0.37783, + "grad_norm": 1.4691560499253644, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 37783 + }, + { + "epoch": 0.37784, + "grad_norm": 1.6124739560473822, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 37784 + }, + { + "epoch": 0.37785, + "grad_norm": 1.191764466251145, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 37785 + }, + { + "epoch": 0.37786, + "grad_norm": 1.7152180551388405, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 37786 + }, + { + "epoch": 0.37787, + "grad_norm": 1.2927806108046351, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37787 + }, + { + "epoch": 0.37788, + "grad_norm": 1.6662325555590847, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 37788 + }, + { + "epoch": 0.37789, + "grad_norm": 1.5311356654531332, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 37789 + }, + { + "epoch": 0.3779, + "grad_norm": 1.4076896531762824, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 37790 + }, + { + "epoch": 0.37791, + "grad_norm": 1.6516842506094604, + "learning_rate": 0.003, + "loss": 4.053, + "step": 37791 + }, + { + "epoch": 0.37792, + "grad_norm": 1.374918207641917, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 37792 + }, + { + "epoch": 0.37793, + "grad_norm": 1.4480624600118204, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37793 + }, + { + "epoch": 0.37794, + "grad_norm": 1.5819810531613, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 37794 + }, + { + "epoch": 0.37795, + "grad_norm": 1.2415180929631702, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 37795 + }, + { + "epoch": 0.37796, + "grad_norm": 1.6523320545224252, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 37796 + }, + { + "epoch": 0.37797, + "grad_norm": 1.2961839830734592, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 37797 + }, + { + "epoch": 0.37798, + "grad_norm": 1.5494839034253405, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 37798 + }, + { + "epoch": 0.37799, + "grad_norm": 1.2045208509455274, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 37799 + }, + { + "epoch": 0.378, + "grad_norm": 1.847727371552368, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37800 + }, + { + "epoch": 0.37801, + "grad_norm": 1.2576066202421574, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 37801 + }, + { + "epoch": 0.37802, + "grad_norm": 2.023043703745257, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 37802 + }, + { + "epoch": 0.37803, + "grad_norm": 1.600634433680554, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 37803 + }, + { + "epoch": 0.37804, + "grad_norm": 1.481365882294003, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37804 + }, + { + "epoch": 0.37805, + "grad_norm": 1.4453898849559097, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 37805 + }, + { + "epoch": 0.37806, + "grad_norm": 1.7358758374893262, + "learning_rate": 0.003, + "loss": 4.058, + "step": 37806 + }, + { + "epoch": 0.37807, + "grad_norm": 1.2911767511234657, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 37807 + }, + { + "epoch": 0.37808, + "grad_norm": 1.6357381775045106, + "learning_rate": 0.003, + "loss": 4.053, + "step": 37808 + }, + { + "epoch": 0.37809, + "grad_norm": 1.5243193685590786, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37809 + }, + { + "epoch": 0.3781, + "grad_norm": 1.2589776210360175, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 37810 + }, + { + "epoch": 0.37811, + "grad_norm": 1.4260711479101378, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37811 + }, + { + "epoch": 0.37812, + "grad_norm": 1.5267738739924246, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 37812 + }, + { + "epoch": 0.37813, + "grad_norm": 1.901412140834396, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 37813 + }, + { + "epoch": 0.37814, + "grad_norm": 1.1243116100592505, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37814 + }, + { + "epoch": 0.37815, + "grad_norm": 1.6773369411166221, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 37815 + }, + { + "epoch": 0.37816, + "grad_norm": 1.2905877306453184, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 37816 + }, + { + "epoch": 0.37817, + "grad_norm": 1.551844134570513, + "learning_rate": 0.003, + "loss": 4.071, + "step": 37817 + }, + { + "epoch": 0.37818, + "grad_norm": 1.6274727451045343, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37818 + }, + { + "epoch": 0.37819, + "grad_norm": 1.8296712327107674, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 37819 + }, + { + "epoch": 0.3782, + "grad_norm": 1.990881250248969, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 37820 + }, + { + "epoch": 0.37821, + "grad_norm": 1.269402709618264, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 37821 + }, + { + "epoch": 0.37822, + "grad_norm": 1.5278538615238555, + "learning_rate": 0.003, + "loss": 4.069, + "step": 37822 + }, + { + "epoch": 0.37823, + "grad_norm": 1.6090606292410503, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 37823 + }, + { + "epoch": 0.37824, + "grad_norm": 1.5837994943923326, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 37824 + }, + { + "epoch": 0.37825, + "grad_norm": 1.5267752159207715, + "learning_rate": 0.003, + "loss": 4.021, + "step": 37825 + }, + { + "epoch": 0.37826, + "grad_norm": 1.707477062302361, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 37826 + }, + { + "epoch": 0.37827, + "grad_norm": 1.4478586536501905, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37827 + }, + { + "epoch": 0.37828, + "grad_norm": 1.3556394501480649, + "learning_rate": 0.003, + "loss": 4.006, + "step": 37828 + }, + { + "epoch": 0.37829, + "grad_norm": 1.6041232691810696, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37829 + }, + { + "epoch": 0.3783, + "grad_norm": 1.604981186911732, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 37830 + }, + { + "epoch": 0.37831, + "grad_norm": 1.411342911535102, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 37831 + }, + { + "epoch": 0.37832, + "grad_norm": 1.533878657181683, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37832 + }, + { + "epoch": 0.37833, + "grad_norm": 1.6194188007685837, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 37833 + }, + { + "epoch": 0.37834, + "grad_norm": 1.4950198859795683, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 37834 + }, + { + "epoch": 0.37835, + "grad_norm": 1.5623308060949288, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 37835 + }, + { + "epoch": 0.37836, + "grad_norm": 1.4827740435675218, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 37836 + }, + { + "epoch": 0.37837, + "grad_norm": 1.2827577023359278, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 37837 + }, + { + "epoch": 0.37838, + "grad_norm": 1.481120795879637, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 37838 + }, + { + "epoch": 0.37839, + "grad_norm": 1.4380155715267713, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 37839 + }, + { + "epoch": 0.3784, + "grad_norm": 1.29802391954271, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 37840 + }, + { + "epoch": 0.37841, + "grad_norm": 1.9236926472937759, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 37841 + }, + { + "epoch": 0.37842, + "grad_norm": 1.306988499793709, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 37842 + }, + { + "epoch": 0.37843, + "grad_norm": 2.0711795780291506, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 37843 + }, + { + "epoch": 0.37844, + "grad_norm": 1.4451787254101478, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 37844 + }, + { + "epoch": 0.37845, + "grad_norm": 1.6571379702992386, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 37845 + }, + { + "epoch": 0.37846, + "grad_norm": 1.4132156497524988, + "learning_rate": 0.003, + "loss": 4.042, + "step": 37846 + }, + { + "epoch": 0.37847, + "grad_norm": 1.5275881427623224, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 37847 + }, + { + "epoch": 0.37848, + "grad_norm": 1.492154834813886, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 37848 + }, + { + "epoch": 0.37849, + "grad_norm": 1.6311365971657668, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 37849 + }, + { + "epoch": 0.3785, + "grad_norm": 1.479433283908546, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37850 + }, + { + "epoch": 0.37851, + "grad_norm": 1.661331606207619, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 37851 + }, + { + "epoch": 0.37852, + "grad_norm": 1.4151841008310853, + "learning_rate": 0.003, + "loss": 4.075, + "step": 37852 + }, + { + "epoch": 0.37853, + "grad_norm": 1.3928966600824906, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 37853 + }, + { + "epoch": 0.37854, + "grad_norm": 1.4522619419480562, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 37854 + }, + { + "epoch": 0.37855, + "grad_norm": 1.6981016948005059, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 37855 + }, + { + "epoch": 0.37856, + "grad_norm": 1.2971937786708676, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 37856 + }, + { + "epoch": 0.37857, + "grad_norm": 2.0861146902234684, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 37857 + }, + { + "epoch": 0.37858, + "grad_norm": 1.3441632264428474, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 37858 + }, + { + "epoch": 0.37859, + "grad_norm": 1.5786700522664256, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37859 + }, + { + "epoch": 0.3786, + "grad_norm": 1.4071992020425266, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 37860 + }, + { + "epoch": 0.37861, + "grad_norm": 1.8263179501551257, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 37861 + }, + { + "epoch": 0.37862, + "grad_norm": 1.5877344637956818, + "learning_rate": 0.003, + "loss": 4.033, + "step": 37862 + }, + { + "epoch": 0.37863, + "grad_norm": 1.292493089315165, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 37863 + }, + { + "epoch": 0.37864, + "grad_norm": 1.5877851078139875, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 37864 + }, + { + "epoch": 0.37865, + "grad_norm": 1.5139431512167145, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37865 + }, + { + "epoch": 0.37866, + "grad_norm": 1.470120217200821, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 37866 + }, + { + "epoch": 0.37867, + "grad_norm": 1.6551087236592923, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 37867 + }, + { + "epoch": 0.37868, + "grad_norm": 1.5882995704195955, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 37868 + }, + { + "epoch": 0.37869, + "grad_norm": 1.353104925010971, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 37869 + }, + { + "epoch": 0.3787, + "grad_norm": 1.4719642014199639, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 37870 + }, + { + "epoch": 0.37871, + "grad_norm": 1.5854540937789023, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 37871 + }, + { + "epoch": 0.37872, + "grad_norm": 1.401164201814763, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 37872 + }, + { + "epoch": 0.37873, + "grad_norm": 1.8884424448475337, + "learning_rate": 0.003, + "loss": 4.039, + "step": 37873 + }, + { + "epoch": 0.37874, + "grad_norm": 1.3962232459540949, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 37874 + }, + { + "epoch": 0.37875, + "grad_norm": 1.7676402729158516, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 37875 + }, + { + "epoch": 0.37876, + "grad_norm": 1.5127515508129141, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 37876 + }, + { + "epoch": 0.37877, + "grad_norm": 1.4251733784358172, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 37877 + }, + { + "epoch": 0.37878, + "grad_norm": 1.3996381933707331, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 37878 + }, + { + "epoch": 0.37879, + "grad_norm": 1.4510317115835045, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 37879 + }, + { + "epoch": 0.3788, + "grad_norm": 1.5733446425892925, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 37880 + }, + { + "epoch": 0.37881, + "grad_norm": 1.569181746951923, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37881 + }, + { + "epoch": 0.37882, + "grad_norm": 1.6671293551969442, + "learning_rate": 0.003, + "loss": 4.034, + "step": 37882 + }, + { + "epoch": 0.37883, + "grad_norm": 1.2225713360339079, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 37883 + }, + { + "epoch": 0.37884, + "grad_norm": 1.682770261507447, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 37884 + }, + { + "epoch": 0.37885, + "grad_norm": 1.542937725815346, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 37885 + }, + { + "epoch": 0.37886, + "grad_norm": 1.5867521485833616, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 37886 + }, + { + "epoch": 0.37887, + "grad_norm": 1.4236295242882409, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37887 + }, + { + "epoch": 0.37888, + "grad_norm": 1.4513713467111362, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 37888 + }, + { + "epoch": 0.37889, + "grad_norm": 1.6203230107228292, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 37889 + }, + { + "epoch": 0.3789, + "grad_norm": 1.3762807583996668, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 37890 + }, + { + "epoch": 0.37891, + "grad_norm": 1.464075422619567, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 37891 + }, + { + "epoch": 0.37892, + "grad_norm": 1.472296810750596, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 37892 + }, + { + "epoch": 0.37893, + "grad_norm": 1.606173308042842, + "learning_rate": 0.003, + "loss": 4.031, + "step": 37893 + }, + { + "epoch": 0.37894, + "grad_norm": 1.4467199557023014, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 37894 + }, + { + "epoch": 0.37895, + "grad_norm": 1.4573778018038017, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 37895 + }, + { + "epoch": 0.37896, + "grad_norm": 1.4471508339069612, + "learning_rate": 0.003, + "loss": 4.03, + "step": 37896 + }, + { + "epoch": 0.37897, + "grad_norm": 1.5655148672453112, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 37897 + }, + { + "epoch": 0.37898, + "grad_norm": 1.5603931481607205, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 37898 + }, + { + "epoch": 0.37899, + "grad_norm": 1.6525165520301304, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 37899 + }, + { + "epoch": 0.379, + "grad_norm": 1.5068295222957369, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 37900 + }, + { + "epoch": 0.37901, + "grad_norm": 1.5942539393816346, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 37901 + }, + { + "epoch": 0.37902, + "grad_norm": 1.3425677765974302, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 37902 + }, + { + "epoch": 0.37903, + "grad_norm": 1.83913239284698, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 37903 + }, + { + "epoch": 0.37904, + "grad_norm": 1.358360610150749, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 37904 + }, + { + "epoch": 0.37905, + "grad_norm": 1.4935197508990328, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 37905 + }, + { + "epoch": 0.37906, + "grad_norm": 1.294843030870738, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 37906 + }, + { + "epoch": 0.37907, + "grad_norm": 1.5643969132274729, + "learning_rate": 0.003, + "loss": 4.058, + "step": 37907 + }, + { + "epoch": 0.37908, + "grad_norm": 1.323028302826402, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 37908 + }, + { + "epoch": 0.37909, + "grad_norm": 1.5933969848502894, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 37909 + }, + { + "epoch": 0.3791, + "grad_norm": 1.747217590223197, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 37910 + }, + { + "epoch": 0.37911, + "grad_norm": 1.498648962685537, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 37911 + }, + { + "epoch": 0.37912, + "grad_norm": 1.8500835466923724, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37912 + }, + { + "epoch": 0.37913, + "grad_norm": 1.2739449191391587, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 37913 + }, + { + "epoch": 0.37914, + "grad_norm": 1.4661388911651978, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 37914 + }, + { + "epoch": 0.37915, + "grad_norm": 1.5871321556032636, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 37915 + }, + { + "epoch": 0.37916, + "grad_norm": 1.5397855587815663, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 37916 + }, + { + "epoch": 0.37917, + "grad_norm": 1.4629641786621317, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 37917 + }, + { + "epoch": 0.37918, + "grad_norm": 1.5464556909409686, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 37918 + }, + { + "epoch": 0.37919, + "grad_norm": 1.6192957513764878, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 37919 + }, + { + "epoch": 0.3792, + "grad_norm": 1.6746438614499026, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 37920 + }, + { + "epoch": 0.37921, + "grad_norm": 1.7795055113992624, + "learning_rate": 0.003, + "loss": 4.009, + "step": 37921 + }, + { + "epoch": 0.37922, + "grad_norm": 1.599291097816906, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 37922 + }, + { + "epoch": 0.37923, + "grad_norm": 1.4494472815791921, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 37923 + }, + { + "epoch": 0.37924, + "grad_norm": 1.4767547487975228, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 37924 + }, + { + "epoch": 0.37925, + "grad_norm": 1.4596799235853015, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 37925 + }, + { + "epoch": 0.37926, + "grad_norm": 1.4014811968155354, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 37926 + }, + { + "epoch": 0.37927, + "grad_norm": 1.3914365647761082, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 37927 + }, + { + "epoch": 0.37928, + "grad_norm": 1.54614947388452, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37928 + }, + { + "epoch": 0.37929, + "grad_norm": 1.6201098805848173, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 37929 + }, + { + "epoch": 0.3793, + "grad_norm": 1.5678895921964504, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 37930 + }, + { + "epoch": 0.37931, + "grad_norm": 1.6377760856629076, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 37931 + }, + { + "epoch": 0.37932, + "grad_norm": 1.6416868602752386, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 37932 + }, + { + "epoch": 0.37933, + "grad_norm": 1.3311734261832382, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 37933 + }, + { + "epoch": 0.37934, + "grad_norm": 1.4716565353603772, + "learning_rate": 0.003, + "loss": 4.02, + "step": 37934 + }, + { + "epoch": 0.37935, + "grad_norm": 1.5144709915014603, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 37935 + }, + { + "epoch": 0.37936, + "grad_norm": 1.4256473465624768, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37936 + }, + { + "epoch": 0.37937, + "grad_norm": 1.613316513958432, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 37937 + }, + { + "epoch": 0.37938, + "grad_norm": 1.6281405436762142, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 37938 + }, + { + "epoch": 0.37939, + "grad_norm": 1.449838728331943, + "learning_rate": 0.003, + "loss": 4.02, + "step": 37939 + }, + { + "epoch": 0.3794, + "grad_norm": 1.513130424739428, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 37940 + }, + { + "epoch": 0.37941, + "grad_norm": 1.7021236346568667, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37941 + }, + { + "epoch": 0.37942, + "grad_norm": 1.1261121335718924, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 37942 + }, + { + "epoch": 0.37943, + "grad_norm": 1.7073814125264293, + "learning_rate": 0.003, + "loss": 4.041, + "step": 37943 + }, + { + "epoch": 0.37944, + "grad_norm": 1.4022077752786035, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 37944 + }, + { + "epoch": 0.37945, + "grad_norm": 1.6650465117320499, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 37945 + }, + { + "epoch": 0.37946, + "grad_norm": 1.6108108598893371, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 37946 + }, + { + "epoch": 0.37947, + "grad_norm": 1.2803940062454677, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 37947 + }, + { + "epoch": 0.37948, + "grad_norm": 1.7014140733702283, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 37948 + }, + { + "epoch": 0.37949, + "grad_norm": 1.4929971329803484, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 37949 + }, + { + "epoch": 0.3795, + "grad_norm": 1.7541591187132353, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37950 + }, + { + "epoch": 0.37951, + "grad_norm": 1.4930281553415186, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 37951 + }, + { + "epoch": 0.37952, + "grad_norm": 1.4100749557888692, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 37952 + }, + { + "epoch": 0.37953, + "grad_norm": 1.5762715864268182, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 37953 + }, + { + "epoch": 0.37954, + "grad_norm": 1.351945404967754, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 37954 + }, + { + "epoch": 0.37955, + "grad_norm": 1.4598651857998872, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 37955 + }, + { + "epoch": 0.37956, + "grad_norm": 1.6043683620683424, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37956 + }, + { + "epoch": 0.37957, + "grad_norm": 1.505058731321315, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 37957 + }, + { + "epoch": 0.37958, + "grad_norm": 1.4639144154157975, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37958 + }, + { + "epoch": 0.37959, + "grad_norm": 1.4218394413937534, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 37959 + }, + { + "epoch": 0.3796, + "grad_norm": 1.9372742843871433, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 37960 + }, + { + "epoch": 0.37961, + "grad_norm": 1.1994393360463114, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 37961 + }, + { + "epoch": 0.37962, + "grad_norm": 1.7424061491485163, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 37962 + }, + { + "epoch": 0.37963, + "grad_norm": 1.390148144366316, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 37963 + }, + { + "epoch": 0.37964, + "grad_norm": 1.6973916936259463, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 37964 + }, + { + "epoch": 0.37965, + "grad_norm": 1.7992220439390507, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 37965 + }, + { + "epoch": 0.37966, + "grad_norm": 1.261396588152627, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 37966 + }, + { + "epoch": 0.37967, + "grad_norm": 1.7904975844403257, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 37967 + }, + { + "epoch": 0.37968, + "grad_norm": 1.362765962848028, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 37968 + }, + { + "epoch": 0.37969, + "grad_norm": 1.4920117969473008, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 37969 + }, + { + "epoch": 0.3797, + "grad_norm": 1.5257249124749215, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 37970 + }, + { + "epoch": 0.37971, + "grad_norm": 1.6364167377708807, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 37971 + }, + { + "epoch": 0.37972, + "grad_norm": 1.4063215143236205, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 37972 + }, + { + "epoch": 0.37973, + "grad_norm": 1.7960439877710896, + "learning_rate": 0.003, + "loss": 4.056, + "step": 37973 + }, + { + "epoch": 0.37974, + "grad_norm": 1.3284156401810387, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 37974 + }, + { + "epoch": 0.37975, + "grad_norm": 1.4466176171556442, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 37975 + }, + { + "epoch": 0.37976, + "grad_norm": 1.4942987671098757, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 37976 + }, + { + "epoch": 0.37977, + "grad_norm": 1.716651027231441, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 37977 + }, + { + "epoch": 0.37978, + "grad_norm": 1.3226436962908752, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 37978 + }, + { + "epoch": 0.37979, + "grad_norm": 1.4223871308209728, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 37979 + }, + { + "epoch": 0.3798, + "grad_norm": 1.558861328267449, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 37980 + }, + { + "epoch": 0.37981, + "grad_norm": 1.6890180948595939, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 37981 + }, + { + "epoch": 0.37982, + "grad_norm": 1.1989753031725672, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 37982 + }, + { + "epoch": 0.37983, + "grad_norm": 1.4377139607132081, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 37983 + }, + { + "epoch": 0.37984, + "grad_norm": 1.4841839437622666, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 37984 + }, + { + "epoch": 0.37985, + "grad_norm": 1.4992850676207639, + "learning_rate": 0.003, + "loss": 4.06, + "step": 37985 + }, + { + "epoch": 0.37986, + "grad_norm": 1.7299346028044968, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 37986 + }, + { + "epoch": 0.37987, + "grad_norm": 1.5817087617493981, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 37987 + }, + { + "epoch": 0.37988, + "grad_norm": 1.698980081427973, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 37988 + }, + { + "epoch": 0.37989, + "grad_norm": 1.1432551309601837, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37989 + }, + { + "epoch": 0.3799, + "grad_norm": 1.7179355379826495, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 37990 + }, + { + "epoch": 0.37991, + "grad_norm": 1.3753599434521742, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 37991 + }, + { + "epoch": 0.37992, + "grad_norm": 1.9027768419406117, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37992 + }, + { + "epoch": 0.37993, + "grad_norm": 1.48248519854308, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 37993 + }, + { + "epoch": 0.37994, + "grad_norm": 1.5432608936856873, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 37994 + }, + { + "epoch": 0.37995, + "grad_norm": 1.6862985202433924, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 37995 + }, + { + "epoch": 0.37996, + "grad_norm": 1.4690188853821764, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 37996 + }, + { + "epoch": 0.37997, + "grad_norm": 1.5132867007583093, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 37997 + }, + { + "epoch": 0.37998, + "grad_norm": 1.8497922471509454, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37998 + }, + { + "epoch": 0.37999, + "grad_norm": 1.2326865694750202, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 37999 + }, + { + "epoch": 0.38, + "grad_norm": 1.458730651389204, + "learning_rate": 0.003, + "loss": 4.06, + "step": 38000 + }, + { + "epoch": 0.38001, + "grad_norm": 1.574436477919589, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38001 + }, + { + "epoch": 0.38002, + "grad_norm": 1.6293480768155686, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 38002 + }, + { + "epoch": 0.38003, + "grad_norm": 1.4645636477863888, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 38003 + }, + { + "epoch": 0.38004, + "grad_norm": 1.4367447239346787, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 38004 + }, + { + "epoch": 0.38005, + "grad_norm": 1.3951428604431457, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 38005 + }, + { + "epoch": 0.38006, + "grad_norm": 1.5097558399938775, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38006 + }, + { + "epoch": 0.38007, + "grad_norm": 1.3987004318007958, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38007 + }, + { + "epoch": 0.38008, + "grad_norm": 1.8273456532533754, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38008 + }, + { + "epoch": 0.38009, + "grad_norm": 1.715805225803674, + "learning_rate": 0.003, + "loss": 4.067, + "step": 38009 + }, + { + "epoch": 0.3801, + "grad_norm": 1.5630179719467634, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 38010 + }, + { + "epoch": 0.38011, + "grad_norm": 1.6133858361838134, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 38011 + }, + { + "epoch": 0.38012, + "grad_norm": 1.8311959651866256, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 38012 + }, + { + "epoch": 0.38013, + "grad_norm": 1.3840067660849795, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38013 + }, + { + "epoch": 0.38014, + "grad_norm": 1.4943971071384852, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 38014 + }, + { + "epoch": 0.38015, + "grad_norm": 1.6277769876374075, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38015 + }, + { + "epoch": 0.38016, + "grad_norm": 1.48232024714728, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38016 + }, + { + "epoch": 0.38017, + "grad_norm": 1.5400711920047225, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38017 + }, + { + "epoch": 0.38018, + "grad_norm": 1.3490485125576572, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 38018 + }, + { + "epoch": 0.38019, + "grad_norm": 1.7631294384000589, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 38019 + }, + { + "epoch": 0.3802, + "grad_norm": 1.2800143071914352, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 38020 + }, + { + "epoch": 0.38021, + "grad_norm": 1.7559136152676196, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 38021 + }, + { + "epoch": 0.38022, + "grad_norm": 1.15223297054616, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 38022 + }, + { + "epoch": 0.38023, + "grad_norm": 2.493738120377482, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 38023 + }, + { + "epoch": 0.38024, + "grad_norm": 1.2333673990320773, + "learning_rate": 0.003, + "loss": 4.064, + "step": 38024 + }, + { + "epoch": 0.38025, + "grad_norm": 1.609873851282699, + "learning_rate": 0.003, + "loss": 4.064, + "step": 38025 + }, + { + "epoch": 0.38026, + "grad_norm": 1.4745053503812715, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 38026 + }, + { + "epoch": 0.38027, + "grad_norm": 1.5710497925287927, + "learning_rate": 0.003, + "loss": 4.047, + "step": 38027 + }, + { + "epoch": 0.38028, + "grad_norm": 2.0371734294506685, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38028 + }, + { + "epoch": 0.38029, + "grad_norm": 1.4040186429431658, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 38029 + }, + { + "epoch": 0.3803, + "grad_norm": 1.529901085430038, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38030 + }, + { + "epoch": 0.38031, + "grad_norm": 1.5379217012508808, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 38031 + }, + { + "epoch": 0.38032, + "grad_norm": 1.4424921020210397, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38032 + }, + { + "epoch": 0.38033, + "grad_norm": 1.570568478199965, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 38033 + }, + { + "epoch": 0.38034, + "grad_norm": 1.42383386783539, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 38034 + }, + { + "epoch": 0.38035, + "grad_norm": 1.53903399964715, + "learning_rate": 0.003, + "loss": 4.059, + "step": 38035 + }, + { + "epoch": 0.38036, + "grad_norm": 1.628865426180663, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 38036 + }, + { + "epoch": 0.38037, + "grad_norm": 1.6436788259803958, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 38037 + }, + { + "epoch": 0.38038, + "grad_norm": 1.8895404880390392, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38038 + }, + { + "epoch": 0.38039, + "grad_norm": 1.4320258281928884, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 38039 + }, + { + "epoch": 0.3804, + "grad_norm": 1.6616387646124204, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 38040 + }, + { + "epoch": 0.38041, + "grad_norm": 1.6593959381759382, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 38041 + }, + { + "epoch": 0.38042, + "grad_norm": 1.4219757005492881, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 38042 + }, + { + "epoch": 0.38043, + "grad_norm": 1.5025652952759092, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 38043 + }, + { + "epoch": 0.38044, + "grad_norm": 1.3391096598299221, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 38044 + }, + { + "epoch": 0.38045, + "grad_norm": 1.4380171611069597, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38045 + }, + { + "epoch": 0.38046, + "grad_norm": 1.3238918697080204, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 38046 + }, + { + "epoch": 0.38047, + "grad_norm": 1.6087919656273033, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38047 + }, + { + "epoch": 0.38048, + "grad_norm": 1.4554876355390727, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 38048 + }, + { + "epoch": 0.38049, + "grad_norm": 1.6098944735782976, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 38049 + }, + { + "epoch": 0.3805, + "grad_norm": 1.34452267140266, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 38050 + }, + { + "epoch": 0.38051, + "grad_norm": 1.9007764079454583, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38051 + }, + { + "epoch": 0.38052, + "grad_norm": 1.192887730764323, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38052 + }, + { + "epoch": 0.38053, + "grad_norm": 1.640120762965744, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38053 + }, + { + "epoch": 0.38054, + "grad_norm": 1.3639961802804175, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38054 + }, + { + "epoch": 0.38055, + "grad_norm": 1.8883916992465735, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 38055 + }, + { + "epoch": 0.38056, + "grad_norm": 1.4473603997668336, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38056 + }, + { + "epoch": 0.38057, + "grad_norm": 1.3673277263489154, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38057 + }, + { + "epoch": 0.38058, + "grad_norm": 1.5215222493580827, + "learning_rate": 0.003, + "loss": 4.06, + "step": 38058 + }, + { + "epoch": 0.38059, + "grad_norm": 1.4177788482176532, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 38059 + }, + { + "epoch": 0.3806, + "grad_norm": 1.4450012820785387, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 38060 + }, + { + "epoch": 0.38061, + "grad_norm": 1.5140942457460014, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 38061 + }, + { + "epoch": 0.38062, + "grad_norm": 1.9862920329760658, + "learning_rate": 0.003, + "loss": 4.056, + "step": 38062 + }, + { + "epoch": 0.38063, + "grad_norm": 1.129678799943479, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 38063 + }, + { + "epoch": 0.38064, + "grad_norm": 1.8132435690650153, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 38064 + }, + { + "epoch": 0.38065, + "grad_norm": 1.6033668128899603, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 38065 + }, + { + "epoch": 0.38066, + "grad_norm": 1.698128872041192, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38066 + }, + { + "epoch": 0.38067, + "grad_norm": 1.500559636558877, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38067 + }, + { + "epoch": 0.38068, + "grad_norm": 1.414301039656164, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 38068 + }, + { + "epoch": 0.38069, + "grad_norm": 1.6842303799036349, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38069 + }, + { + "epoch": 0.3807, + "grad_norm": 1.193110303544231, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 38070 + }, + { + "epoch": 0.38071, + "grad_norm": 1.6031625013886794, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38071 + }, + { + "epoch": 0.38072, + "grad_norm": 1.347308090101412, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 38072 + }, + { + "epoch": 0.38073, + "grad_norm": 1.740700722772463, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 38073 + }, + { + "epoch": 0.38074, + "grad_norm": 1.2404915180171094, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 38074 + }, + { + "epoch": 0.38075, + "grad_norm": 1.9008891367556222, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 38075 + }, + { + "epoch": 0.38076, + "grad_norm": 1.2249628540524609, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38076 + }, + { + "epoch": 0.38077, + "grad_norm": 1.521561508334356, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 38077 + }, + { + "epoch": 0.38078, + "grad_norm": 1.3453402536484436, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 38078 + }, + { + "epoch": 0.38079, + "grad_norm": 1.612685634518459, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38079 + }, + { + "epoch": 0.3808, + "grad_norm": 1.5857646433601347, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 38080 + }, + { + "epoch": 0.38081, + "grad_norm": 1.7138406707192713, + "learning_rate": 0.003, + "loss": 4.063, + "step": 38081 + }, + { + "epoch": 0.38082, + "grad_norm": 1.5189656501406743, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 38082 + }, + { + "epoch": 0.38083, + "grad_norm": 1.6766662435065016, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38083 + }, + { + "epoch": 0.38084, + "grad_norm": 1.3705548206654299, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 38084 + }, + { + "epoch": 0.38085, + "grad_norm": 1.493370754714028, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 38085 + }, + { + "epoch": 0.38086, + "grad_norm": 1.3619339266926718, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38086 + }, + { + "epoch": 0.38087, + "grad_norm": 1.6069013050229943, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 38087 + }, + { + "epoch": 0.38088, + "grad_norm": 1.6468234374702695, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 38088 + }, + { + "epoch": 0.38089, + "grad_norm": 1.466086940629366, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38089 + }, + { + "epoch": 0.3809, + "grad_norm": 1.4963557368075855, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 38090 + }, + { + "epoch": 0.38091, + "grad_norm": 1.4136284127872643, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38091 + }, + { + "epoch": 0.38092, + "grad_norm": 1.5852755199659525, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 38092 + }, + { + "epoch": 0.38093, + "grad_norm": 1.5585814161450704, + "learning_rate": 0.003, + "loss": 4.023, + "step": 38093 + }, + { + "epoch": 0.38094, + "grad_norm": 1.567291999264119, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 38094 + }, + { + "epoch": 0.38095, + "grad_norm": 1.3491744294189114, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38095 + }, + { + "epoch": 0.38096, + "grad_norm": 1.3336944829096535, + "learning_rate": 0.003, + "loss": 4.022, + "step": 38096 + }, + { + "epoch": 0.38097, + "grad_norm": 1.396947506351399, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 38097 + }, + { + "epoch": 0.38098, + "grad_norm": 1.36099479080887, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38098 + }, + { + "epoch": 0.38099, + "grad_norm": 1.7113352617269042, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38099 + }, + { + "epoch": 0.381, + "grad_norm": 1.429677286670589, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 38100 + }, + { + "epoch": 0.38101, + "grad_norm": 1.857651517502119, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38101 + }, + { + "epoch": 0.38102, + "grad_norm": 1.717887979860227, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 38102 + }, + { + "epoch": 0.38103, + "grad_norm": 1.4002696828256471, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 38103 + }, + { + "epoch": 0.38104, + "grad_norm": 1.4771624136142265, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 38104 + }, + { + "epoch": 0.38105, + "grad_norm": 1.7075318854591288, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 38105 + }, + { + "epoch": 0.38106, + "grad_norm": 1.3199845441477496, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38106 + }, + { + "epoch": 0.38107, + "grad_norm": 1.6456459005748083, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38107 + }, + { + "epoch": 0.38108, + "grad_norm": 1.2697699705125693, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 38108 + }, + { + "epoch": 0.38109, + "grad_norm": 1.486794111588314, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 38109 + }, + { + "epoch": 0.3811, + "grad_norm": 1.3831204156323718, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 38110 + }, + { + "epoch": 0.38111, + "grad_norm": 1.3795902851903574, + "learning_rate": 0.003, + "loss": 4.054, + "step": 38111 + }, + { + "epoch": 0.38112, + "grad_norm": 1.7522628784112733, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 38112 + }, + { + "epoch": 0.38113, + "grad_norm": 2.0383797588091817, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 38113 + }, + { + "epoch": 0.38114, + "grad_norm": 1.3812288020376993, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 38114 + }, + { + "epoch": 0.38115, + "grad_norm": 1.3496306307289567, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 38115 + }, + { + "epoch": 0.38116, + "grad_norm": 1.5497874531372848, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 38116 + }, + { + "epoch": 0.38117, + "grad_norm": 1.516280746895107, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 38117 + }, + { + "epoch": 0.38118, + "grad_norm": 1.4673588979841243, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 38118 + }, + { + "epoch": 0.38119, + "grad_norm": 1.4518186931830015, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 38119 + }, + { + "epoch": 0.3812, + "grad_norm": 1.6773297197026924, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 38120 + }, + { + "epoch": 0.38121, + "grad_norm": 1.6158670217315665, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 38121 + }, + { + "epoch": 0.38122, + "grad_norm": 1.4744649991130476, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 38122 + }, + { + "epoch": 0.38123, + "grad_norm": 1.6080544371972816, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 38123 + }, + { + "epoch": 0.38124, + "grad_norm": 1.5114921119786555, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 38124 + }, + { + "epoch": 0.38125, + "grad_norm": 1.3802751373937836, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 38125 + }, + { + "epoch": 0.38126, + "grad_norm": 1.6929236370823615, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38126 + }, + { + "epoch": 0.38127, + "grad_norm": 1.1776913616831752, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 38127 + }, + { + "epoch": 0.38128, + "grad_norm": 1.673599525074305, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 38128 + }, + { + "epoch": 0.38129, + "grad_norm": 1.607514594730208, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 38129 + }, + { + "epoch": 0.3813, + "grad_norm": 1.6419335542426494, + "learning_rate": 0.003, + "loss": 4.058, + "step": 38130 + }, + { + "epoch": 0.38131, + "grad_norm": 1.7463228434211504, + "learning_rate": 0.003, + "loss": 4.067, + "step": 38131 + }, + { + "epoch": 0.38132, + "grad_norm": 1.4379352505698575, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 38132 + }, + { + "epoch": 0.38133, + "grad_norm": 1.448840609208713, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 38133 + }, + { + "epoch": 0.38134, + "grad_norm": 1.3737107167849696, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38134 + }, + { + "epoch": 0.38135, + "grad_norm": 1.9819594196876114, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 38135 + }, + { + "epoch": 0.38136, + "grad_norm": 1.0233678663274144, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 38136 + }, + { + "epoch": 0.38137, + "grad_norm": 1.8090854561664802, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38137 + }, + { + "epoch": 0.38138, + "grad_norm": 1.5712612553554386, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38138 + }, + { + "epoch": 0.38139, + "grad_norm": 1.6994310387358598, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 38139 + }, + { + "epoch": 0.3814, + "grad_norm": 1.393968907766906, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38140 + }, + { + "epoch": 0.38141, + "grad_norm": 1.6216701596370695, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38141 + }, + { + "epoch": 0.38142, + "grad_norm": 1.329480403774831, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 38142 + }, + { + "epoch": 0.38143, + "grad_norm": 1.6612217292464404, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 38143 + }, + { + "epoch": 0.38144, + "grad_norm": 1.5013700592251358, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 38144 + }, + { + "epoch": 0.38145, + "grad_norm": 1.4731731261620622, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 38145 + }, + { + "epoch": 0.38146, + "grad_norm": 1.4316844087666418, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 38146 + }, + { + "epoch": 0.38147, + "grad_norm": 1.6447956627862839, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 38147 + }, + { + "epoch": 0.38148, + "grad_norm": 1.5104321967141061, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38148 + }, + { + "epoch": 0.38149, + "grad_norm": 1.6485464359398383, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 38149 + }, + { + "epoch": 0.3815, + "grad_norm": 1.3897048473885474, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 38150 + }, + { + "epoch": 0.38151, + "grad_norm": 1.444304124626784, + "learning_rate": 0.003, + "loss": 4.059, + "step": 38151 + }, + { + "epoch": 0.38152, + "grad_norm": 1.4514847574045202, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 38152 + }, + { + "epoch": 0.38153, + "grad_norm": 1.3309283844745192, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38153 + }, + { + "epoch": 0.38154, + "grad_norm": 1.408854732760175, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 38154 + }, + { + "epoch": 0.38155, + "grad_norm": 1.3701683080455458, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 38155 + }, + { + "epoch": 0.38156, + "grad_norm": 1.7125759632619182, + "learning_rate": 0.003, + "loss": 4.034, + "step": 38156 + }, + { + "epoch": 0.38157, + "grad_norm": 1.398812847536106, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 38157 + }, + { + "epoch": 0.38158, + "grad_norm": 1.6775532857331077, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 38158 + }, + { + "epoch": 0.38159, + "grad_norm": 1.520491483262833, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38159 + }, + { + "epoch": 0.3816, + "grad_norm": 1.4043940043485175, + "learning_rate": 0.003, + "loss": 4.031, + "step": 38160 + }, + { + "epoch": 0.38161, + "grad_norm": 1.756152906963884, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38161 + }, + { + "epoch": 0.38162, + "grad_norm": 1.883408807488077, + "learning_rate": 0.003, + "loss": 4.038, + "step": 38162 + }, + { + "epoch": 0.38163, + "grad_norm": 1.1883177317628133, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38163 + }, + { + "epoch": 0.38164, + "grad_norm": 1.786330970131783, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 38164 + }, + { + "epoch": 0.38165, + "grad_norm": 1.3924312425732082, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38165 + }, + { + "epoch": 0.38166, + "grad_norm": 1.6072153879517732, + "learning_rate": 0.003, + "loss": 4.084, + "step": 38166 + }, + { + "epoch": 0.38167, + "grad_norm": 1.3805584157205872, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 38167 + }, + { + "epoch": 0.38168, + "grad_norm": 1.4419399075349033, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38168 + }, + { + "epoch": 0.38169, + "grad_norm": 1.435437708412951, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38169 + }, + { + "epoch": 0.3817, + "grad_norm": 1.558620623803038, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 38170 + }, + { + "epoch": 0.38171, + "grad_norm": 1.4526485327727174, + "learning_rate": 0.003, + "loss": 4.053, + "step": 38171 + }, + { + "epoch": 0.38172, + "grad_norm": 1.685354067679757, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 38172 + }, + { + "epoch": 0.38173, + "grad_norm": 1.944033632622374, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 38173 + }, + { + "epoch": 0.38174, + "grad_norm": 1.3284858768649397, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 38174 + }, + { + "epoch": 0.38175, + "grad_norm": 2.0242121459329376, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 38175 + }, + { + "epoch": 0.38176, + "grad_norm": 1.1840507731170649, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38176 + }, + { + "epoch": 0.38177, + "grad_norm": 1.8015362648800977, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38177 + }, + { + "epoch": 0.38178, + "grad_norm": 1.3930079412454017, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 38178 + }, + { + "epoch": 0.38179, + "grad_norm": 1.5732729110280272, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 38179 + }, + { + "epoch": 0.3818, + "grad_norm": 1.4966171086850908, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 38180 + }, + { + "epoch": 0.38181, + "grad_norm": 1.224764901872778, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 38181 + }, + { + "epoch": 0.38182, + "grad_norm": 1.6743777228326642, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 38182 + }, + { + "epoch": 0.38183, + "grad_norm": 1.4664563139518727, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38183 + }, + { + "epoch": 0.38184, + "grad_norm": 1.713556159867431, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 38184 + }, + { + "epoch": 0.38185, + "grad_norm": 1.4227716828280814, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 38185 + }, + { + "epoch": 0.38186, + "grad_norm": 1.6411978260858577, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 38186 + }, + { + "epoch": 0.38187, + "grad_norm": 1.6662874393846723, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 38187 + }, + { + "epoch": 0.38188, + "grad_norm": 1.4060358513473483, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 38188 + }, + { + "epoch": 0.38189, + "grad_norm": 1.5832887187883937, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 38189 + }, + { + "epoch": 0.3819, + "grad_norm": 1.4873916524874682, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 38190 + }, + { + "epoch": 0.38191, + "grad_norm": 1.6664966975933515, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38191 + }, + { + "epoch": 0.38192, + "grad_norm": 1.6546389370364312, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 38192 + }, + { + "epoch": 0.38193, + "grad_norm": 1.4068477294956818, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 38193 + }, + { + "epoch": 0.38194, + "grad_norm": 1.41790042983329, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 38194 + }, + { + "epoch": 0.38195, + "grad_norm": 1.4549081340350376, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 38195 + }, + { + "epoch": 0.38196, + "grad_norm": 1.3800588630537087, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38196 + }, + { + "epoch": 0.38197, + "grad_norm": 1.5673802575117204, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 38197 + }, + { + "epoch": 0.38198, + "grad_norm": 1.7621498836116818, + "learning_rate": 0.003, + "loss": 4.056, + "step": 38198 + }, + { + "epoch": 0.38199, + "grad_norm": 1.352019488535621, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38199 + }, + { + "epoch": 0.382, + "grad_norm": 1.681214876883562, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38200 + }, + { + "epoch": 0.38201, + "grad_norm": 1.4681537786285794, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 38201 + }, + { + "epoch": 0.38202, + "grad_norm": 1.5649908158204908, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 38202 + }, + { + "epoch": 0.38203, + "grad_norm": 1.4296028784900712, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 38203 + }, + { + "epoch": 0.38204, + "grad_norm": 1.424823193560532, + "learning_rate": 0.003, + "loss": 4.035, + "step": 38204 + }, + { + "epoch": 0.38205, + "grad_norm": 1.6757933423933602, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38205 + }, + { + "epoch": 0.38206, + "grad_norm": 1.1970661793097788, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 38206 + }, + { + "epoch": 0.38207, + "grad_norm": 1.620706916756378, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 38207 + }, + { + "epoch": 0.38208, + "grad_norm": 1.6003748790617929, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38208 + }, + { + "epoch": 0.38209, + "grad_norm": 1.8474252119197, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 38209 + }, + { + "epoch": 0.3821, + "grad_norm": 1.496283701294245, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 38210 + }, + { + "epoch": 0.38211, + "grad_norm": 1.4259362570601186, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 38211 + }, + { + "epoch": 0.38212, + "grad_norm": 1.3982327865858555, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 38212 + }, + { + "epoch": 0.38213, + "grad_norm": 1.4502086740463582, + "learning_rate": 0.003, + "loss": 4.011, + "step": 38213 + }, + { + "epoch": 0.38214, + "grad_norm": 1.8618922315904038, + "learning_rate": 0.003, + "loss": 4.065, + "step": 38214 + }, + { + "epoch": 0.38215, + "grad_norm": 1.4941450904086255, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38215 + }, + { + "epoch": 0.38216, + "grad_norm": 1.3823201869901092, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 38216 + }, + { + "epoch": 0.38217, + "grad_norm": 1.6172477364021631, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38217 + }, + { + "epoch": 0.38218, + "grad_norm": 1.2576208038859087, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 38218 + }, + { + "epoch": 0.38219, + "grad_norm": 1.6598171536566821, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38219 + }, + { + "epoch": 0.3822, + "grad_norm": 1.3264407414088986, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 38220 + }, + { + "epoch": 0.38221, + "grad_norm": 2.0876645065911386, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 38221 + }, + { + "epoch": 0.38222, + "grad_norm": 1.7616189552242247, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 38222 + }, + { + "epoch": 0.38223, + "grad_norm": 1.5248676020205267, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 38223 + }, + { + "epoch": 0.38224, + "grad_norm": 1.702156923225873, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 38224 + }, + { + "epoch": 0.38225, + "grad_norm": 1.7954471517933361, + "learning_rate": 0.003, + "loss": 4.03, + "step": 38225 + }, + { + "epoch": 0.38226, + "grad_norm": 1.1432220880274977, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 38226 + }, + { + "epoch": 0.38227, + "grad_norm": 1.603009624874351, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 38227 + }, + { + "epoch": 0.38228, + "grad_norm": 1.3046527337911815, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 38228 + }, + { + "epoch": 0.38229, + "grad_norm": 1.4972527178316302, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 38229 + }, + { + "epoch": 0.3823, + "grad_norm": 1.7482815985965945, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38230 + }, + { + "epoch": 0.38231, + "grad_norm": 1.3161437935984373, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 38231 + }, + { + "epoch": 0.38232, + "grad_norm": 1.637133719783138, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 38232 + }, + { + "epoch": 0.38233, + "grad_norm": 1.3743222187927104, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38233 + }, + { + "epoch": 0.38234, + "grad_norm": 1.7239412122440283, + "learning_rate": 0.003, + "loss": 4.075, + "step": 38234 + }, + { + "epoch": 0.38235, + "grad_norm": 1.1720004957357344, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 38235 + }, + { + "epoch": 0.38236, + "grad_norm": 1.5955675330055188, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 38236 + }, + { + "epoch": 0.38237, + "grad_norm": 1.5073066773539825, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 38237 + }, + { + "epoch": 0.38238, + "grad_norm": 1.5701559896986763, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 38238 + }, + { + "epoch": 0.38239, + "grad_norm": 1.496965430828016, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 38239 + }, + { + "epoch": 0.3824, + "grad_norm": 1.59354399154771, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38240 + }, + { + "epoch": 0.38241, + "grad_norm": 1.6973151026683455, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 38241 + }, + { + "epoch": 0.38242, + "grad_norm": 1.5707705841690711, + "learning_rate": 0.003, + "loss": 4.021, + "step": 38242 + }, + { + "epoch": 0.38243, + "grad_norm": 1.4083328582423702, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 38243 + }, + { + "epoch": 0.38244, + "grad_norm": 1.465374651094211, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 38244 + }, + { + "epoch": 0.38245, + "grad_norm": 1.4152591058231712, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38245 + }, + { + "epoch": 0.38246, + "grad_norm": 1.5423749683029222, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 38246 + }, + { + "epoch": 0.38247, + "grad_norm": 1.3765097737956036, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 38247 + }, + { + "epoch": 0.38248, + "grad_norm": 1.6212642907847992, + "learning_rate": 0.003, + "loss": 4.025, + "step": 38248 + }, + { + "epoch": 0.38249, + "grad_norm": 1.2824695337001397, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 38249 + }, + { + "epoch": 0.3825, + "grad_norm": 1.727310559941878, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38250 + }, + { + "epoch": 0.38251, + "grad_norm": 1.2809879570538458, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 38251 + }, + { + "epoch": 0.38252, + "grad_norm": 1.8056193564657599, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 38252 + }, + { + "epoch": 0.38253, + "grad_norm": 1.4494707136601226, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 38253 + }, + { + "epoch": 0.38254, + "grad_norm": 1.5374799378513018, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 38254 + }, + { + "epoch": 0.38255, + "grad_norm": 1.7378750219630217, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38255 + }, + { + "epoch": 0.38256, + "grad_norm": 1.1231889441770855, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38256 + }, + { + "epoch": 0.38257, + "grad_norm": 1.8253539422845841, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38257 + }, + { + "epoch": 0.38258, + "grad_norm": 1.3297410527111788, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38258 + }, + { + "epoch": 0.38259, + "grad_norm": 1.8320675073057064, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 38259 + }, + { + "epoch": 0.3826, + "grad_norm": 1.8460441515420398, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38260 + }, + { + "epoch": 0.38261, + "grad_norm": 1.6030851512008881, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 38261 + }, + { + "epoch": 0.38262, + "grad_norm": 1.3092816882092404, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 38262 + }, + { + "epoch": 0.38263, + "grad_norm": 1.774461950119579, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 38263 + }, + { + "epoch": 0.38264, + "grad_norm": 1.3571984696698092, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38264 + }, + { + "epoch": 0.38265, + "grad_norm": 1.7443895837781698, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 38265 + }, + { + "epoch": 0.38266, + "grad_norm": 1.4726739798401198, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 38266 + }, + { + "epoch": 0.38267, + "grad_norm": 1.546455299485055, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 38267 + }, + { + "epoch": 0.38268, + "grad_norm": 1.2960845260829843, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38268 + }, + { + "epoch": 0.38269, + "grad_norm": 1.5281119336829627, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38269 + }, + { + "epoch": 0.3827, + "grad_norm": 1.55982132167772, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 38270 + }, + { + "epoch": 0.38271, + "grad_norm": 1.3712206256058608, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 38271 + }, + { + "epoch": 0.38272, + "grad_norm": 1.4967819121184007, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38272 + }, + { + "epoch": 0.38273, + "grad_norm": 1.5481948626984197, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 38273 + }, + { + "epoch": 0.38274, + "grad_norm": 1.5767125133329556, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38274 + }, + { + "epoch": 0.38275, + "grad_norm": 1.2881049437496594, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 38275 + }, + { + "epoch": 0.38276, + "grad_norm": 1.54951512996983, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 38276 + }, + { + "epoch": 0.38277, + "grad_norm": 1.6224684424055855, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 38277 + }, + { + "epoch": 0.38278, + "grad_norm": 1.5463535285193881, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38278 + }, + { + "epoch": 0.38279, + "grad_norm": 1.2512750750629933, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38279 + }, + { + "epoch": 0.3828, + "grad_norm": 1.6907348447823376, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 38280 + }, + { + "epoch": 0.38281, + "grad_norm": 1.4264675697185798, + "learning_rate": 0.003, + "loss": 4.038, + "step": 38281 + }, + { + "epoch": 0.38282, + "grad_norm": 1.3860282749688058, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 38282 + }, + { + "epoch": 0.38283, + "grad_norm": 1.6333814658210806, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 38283 + }, + { + "epoch": 0.38284, + "grad_norm": 1.4826054380425635, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 38284 + }, + { + "epoch": 0.38285, + "grad_norm": 1.5317068332809833, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 38285 + }, + { + "epoch": 0.38286, + "grad_norm": 1.3851555964753652, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 38286 + }, + { + "epoch": 0.38287, + "grad_norm": 1.802536423764432, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38287 + }, + { + "epoch": 0.38288, + "grad_norm": 1.2585200859338794, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 38288 + }, + { + "epoch": 0.38289, + "grad_norm": 1.7494708808314086, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38289 + }, + { + "epoch": 0.3829, + "grad_norm": 1.7969793604315276, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38290 + }, + { + "epoch": 0.38291, + "grad_norm": 1.4459008887974578, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 38291 + }, + { + "epoch": 0.38292, + "grad_norm": 1.606204095750402, + "learning_rate": 0.003, + "loss": 4.019, + "step": 38292 + }, + { + "epoch": 0.38293, + "grad_norm": 1.399420023128102, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 38293 + }, + { + "epoch": 0.38294, + "grad_norm": 1.3916890028699684, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 38294 + }, + { + "epoch": 0.38295, + "grad_norm": 1.5010048778674123, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 38295 + }, + { + "epoch": 0.38296, + "grad_norm": 1.4150848695861273, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 38296 + }, + { + "epoch": 0.38297, + "grad_norm": 1.361251330581607, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 38297 + }, + { + "epoch": 0.38298, + "grad_norm": 1.6647535619416962, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 38298 + }, + { + "epoch": 0.38299, + "grad_norm": 1.4358302187834695, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 38299 + }, + { + "epoch": 0.383, + "grad_norm": 1.4456472349239426, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 38300 + }, + { + "epoch": 0.38301, + "grad_norm": 1.6272170638529528, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 38301 + }, + { + "epoch": 0.38302, + "grad_norm": 1.7579368159752236, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38302 + }, + { + "epoch": 0.38303, + "grad_norm": 2.108536568800147, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 38303 + }, + { + "epoch": 0.38304, + "grad_norm": 1.3316629123686083, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 38304 + }, + { + "epoch": 0.38305, + "grad_norm": 1.8749014314914643, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38305 + }, + { + "epoch": 0.38306, + "grad_norm": 1.280353725010765, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 38306 + }, + { + "epoch": 0.38307, + "grad_norm": 1.7379510518828023, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 38307 + }, + { + "epoch": 0.38308, + "grad_norm": 1.3888818769898152, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38308 + }, + { + "epoch": 0.38309, + "grad_norm": 1.6335537204434132, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 38309 + }, + { + "epoch": 0.3831, + "grad_norm": 1.578820643890547, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 38310 + }, + { + "epoch": 0.38311, + "grad_norm": 1.6273261885005528, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 38311 + }, + { + "epoch": 0.38312, + "grad_norm": 1.717045722326628, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38312 + }, + { + "epoch": 0.38313, + "grad_norm": 1.1656768975014413, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 38313 + }, + { + "epoch": 0.38314, + "grad_norm": 1.4902429329914384, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 38314 + }, + { + "epoch": 0.38315, + "grad_norm": 1.4070415428836567, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 38315 + }, + { + "epoch": 0.38316, + "grad_norm": 1.754595547574413, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 38316 + }, + { + "epoch": 0.38317, + "grad_norm": 1.442689592487331, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 38317 + }, + { + "epoch": 0.38318, + "grad_norm": 1.5017730088086767, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38318 + }, + { + "epoch": 0.38319, + "grad_norm": 1.4881260227315967, + "learning_rate": 0.003, + "loss": 4.052, + "step": 38319 + }, + { + "epoch": 0.3832, + "grad_norm": 1.6624671303217495, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 38320 + }, + { + "epoch": 0.38321, + "grad_norm": 1.2241441179119057, + "learning_rate": 0.003, + "loss": 4.024, + "step": 38321 + }, + { + "epoch": 0.38322, + "grad_norm": 1.692341987753431, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38322 + }, + { + "epoch": 0.38323, + "grad_norm": 1.329567753863701, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 38323 + }, + { + "epoch": 0.38324, + "grad_norm": 1.5254427073773573, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38324 + }, + { + "epoch": 0.38325, + "grad_norm": 1.1899039101400304, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 38325 + }, + { + "epoch": 0.38326, + "grad_norm": 1.8286113377233204, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 38326 + }, + { + "epoch": 0.38327, + "grad_norm": 1.5192886370063872, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 38327 + }, + { + "epoch": 0.38328, + "grad_norm": 1.8939115753375741, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 38328 + }, + { + "epoch": 0.38329, + "grad_norm": 1.669992517336523, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 38329 + }, + { + "epoch": 0.3833, + "grad_norm": 1.4790446576460103, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 38330 + }, + { + "epoch": 0.38331, + "grad_norm": 1.9036513559189998, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 38331 + }, + { + "epoch": 0.38332, + "grad_norm": 1.429513706061288, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 38332 + }, + { + "epoch": 0.38333, + "grad_norm": 1.3572645203475162, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38333 + }, + { + "epoch": 0.38334, + "grad_norm": 1.5663341962309072, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 38334 + }, + { + "epoch": 0.38335, + "grad_norm": 1.4333350177968114, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 38335 + }, + { + "epoch": 0.38336, + "grad_norm": 1.5562300031320166, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 38336 + }, + { + "epoch": 0.38337, + "grad_norm": 1.763630546597831, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38337 + }, + { + "epoch": 0.38338, + "grad_norm": 1.3949130721260967, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38338 + }, + { + "epoch": 0.38339, + "grad_norm": 1.2887368585363503, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38339 + }, + { + "epoch": 0.3834, + "grad_norm": 1.6068923696266177, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 38340 + }, + { + "epoch": 0.38341, + "grad_norm": 1.504772823714911, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38341 + }, + { + "epoch": 0.38342, + "grad_norm": 1.8325892528701353, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 38342 + }, + { + "epoch": 0.38343, + "grad_norm": 1.2928030600716691, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 38343 + }, + { + "epoch": 0.38344, + "grad_norm": 1.6248089637819838, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 38344 + }, + { + "epoch": 0.38345, + "grad_norm": 1.6924072060753221, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38345 + }, + { + "epoch": 0.38346, + "grad_norm": 1.322707270047044, + "learning_rate": 0.003, + "loss": 4.018, + "step": 38346 + }, + { + "epoch": 0.38347, + "grad_norm": 1.7321909095016972, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 38347 + }, + { + "epoch": 0.38348, + "grad_norm": 1.235104184636563, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 38348 + }, + { + "epoch": 0.38349, + "grad_norm": 1.7201139565080892, + "learning_rate": 0.003, + "loss": 4.062, + "step": 38349 + }, + { + "epoch": 0.3835, + "grad_norm": 1.4488362084277642, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38350 + }, + { + "epoch": 0.38351, + "grad_norm": 1.5172108224083816, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 38351 + }, + { + "epoch": 0.38352, + "grad_norm": 1.4870892050868905, + "learning_rate": 0.003, + "loss": 4.021, + "step": 38352 + }, + { + "epoch": 0.38353, + "grad_norm": 1.5518438265693135, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 38353 + }, + { + "epoch": 0.38354, + "grad_norm": 1.4924928750598376, + "learning_rate": 0.003, + "loss": 4.038, + "step": 38354 + }, + { + "epoch": 0.38355, + "grad_norm": 1.756086518494792, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38355 + }, + { + "epoch": 0.38356, + "grad_norm": 1.3061581137755607, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 38356 + }, + { + "epoch": 0.38357, + "grad_norm": 1.9899319137973295, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 38357 + }, + { + "epoch": 0.38358, + "grad_norm": 1.6108758026399645, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38358 + }, + { + "epoch": 0.38359, + "grad_norm": 1.4787981187304238, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38359 + }, + { + "epoch": 0.3836, + "grad_norm": 1.853189614102945, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38360 + }, + { + "epoch": 0.38361, + "grad_norm": 1.4889394434190948, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38361 + }, + { + "epoch": 0.38362, + "grad_norm": 1.5366884627961135, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 38362 + }, + { + "epoch": 0.38363, + "grad_norm": 1.7972866876959903, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38363 + }, + { + "epoch": 0.38364, + "grad_norm": 1.5383802043885524, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 38364 + }, + { + "epoch": 0.38365, + "grad_norm": 1.3785996539337115, + "learning_rate": 0.003, + "loss": 4.012, + "step": 38365 + }, + { + "epoch": 0.38366, + "grad_norm": 1.5487085385993378, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38366 + }, + { + "epoch": 0.38367, + "grad_norm": 1.4658618891582298, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 38367 + }, + { + "epoch": 0.38368, + "grad_norm": 1.6855445772415363, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 38368 + }, + { + "epoch": 0.38369, + "grad_norm": 1.5507515877120759, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 38369 + }, + { + "epoch": 0.3837, + "grad_norm": 1.5599815755512325, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 38370 + }, + { + "epoch": 0.38371, + "grad_norm": 1.2843727436506087, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 38371 + }, + { + "epoch": 0.38372, + "grad_norm": 1.6210097059186848, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 38372 + }, + { + "epoch": 0.38373, + "grad_norm": 1.332149995473835, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 38373 + }, + { + "epoch": 0.38374, + "grad_norm": 1.5495348182500321, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38374 + }, + { + "epoch": 0.38375, + "grad_norm": 1.4981101976691433, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 38375 + }, + { + "epoch": 0.38376, + "grad_norm": 1.4824888932464229, + "learning_rate": 0.003, + "loss": 4.043, + "step": 38376 + }, + { + "epoch": 0.38377, + "grad_norm": 1.7852590943843794, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 38377 + }, + { + "epoch": 0.38378, + "grad_norm": 1.558622329218316, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 38378 + }, + { + "epoch": 0.38379, + "grad_norm": 1.5999036045449402, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 38379 + }, + { + "epoch": 0.3838, + "grad_norm": 1.6878854969783494, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 38380 + }, + { + "epoch": 0.38381, + "grad_norm": 1.560969467623565, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 38381 + }, + { + "epoch": 0.38382, + "grad_norm": 1.4652882727582919, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 38382 + }, + { + "epoch": 0.38383, + "grad_norm": 1.7504372291341632, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38383 + }, + { + "epoch": 0.38384, + "grad_norm": 1.3771937021163363, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 38384 + }, + { + "epoch": 0.38385, + "grad_norm": 1.4115201927696739, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 38385 + }, + { + "epoch": 0.38386, + "grad_norm": 1.275900020266946, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 38386 + }, + { + "epoch": 0.38387, + "grad_norm": 1.6413053616219546, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 38387 + }, + { + "epoch": 0.38388, + "grad_norm": 1.4404303808221173, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38388 + }, + { + "epoch": 0.38389, + "grad_norm": 1.5063998061026056, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 38389 + }, + { + "epoch": 0.3839, + "grad_norm": 1.6099600531245646, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 38390 + }, + { + "epoch": 0.38391, + "grad_norm": 1.7440486907220143, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38391 + }, + { + "epoch": 0.38392, + "grad_norm": 1.5508608826722472, + "learning_rate": 0.003, + "loss": 4.028, + "step": 38392 + }, + { + "epoch": 0.38393, + "grad_norm": 1.399953788142759, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 38393 + }, + { + "epoch": 0.38394, + "grad_norm": 1.5513483969123625, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38394 + }, + { + "epoch": 0.38395, + "grad_norm": 1.5900083849078535, + "learning_rate": 0.003, + "loss": 4.064, + "step": 38395 + }, + { + "epoch": 0.38396, + "grad_norm": 1.3473084977824306, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 38396 + }, + { + "epoch": 0.38397, + "grad_norm": 1.5844457238659384, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 38397 + }, + { + "epoch": 0.38398, + "grad_norm": 1.539440557408991, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 38398 + }, + { + "epoch": 0.38399, + "grad_norm": 1.3258385323943993, + "learning_rate": 0.003, + "loss": 4.053, + "step": 38399 + }, + { + "epoch": 0.384, + "grad_norm": 1.9006424967851867, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 38400 + }, + { + "epoch": 0.38401, + "grad_norm": 1.5527642964567563, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38401 + }, + { + "epoch": 0.38402, + "grad_norm": 1.7372559762157984, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38402 + }, + { + "epoch": 0.38403, + "grad_norm": 1.381879651234559, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38403 + }, + { + "epoch": 0.38404, + "grad_norm": 1.6652764790957013, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 38404 + }, + { + "epoch": 0.38405, + "grad_norm": 1.640747513206889, + "learning_rate": 0.003, + "loss": 4.029, + "step": 38405 + }, + { + "epoch": 0.38406, + "grad_norm": 1.403051441099557, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 38406 + }, + { + "epoch": 0.38407, + "grad_norm": 1.588127085852381, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38407 + }, + { + "epoch": 0.38408, + "grad_norm": 1.3094072043855265, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38408 + }, + { + "epoch": 0.38409, + "grad_norm": 1.714098021282935, + "learning_rate": 0.003, + "loss": 3.997, + "step": 38409 + }, + { + "epoch": 0.3841, + "grad_norm": 1.3980300915015498, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 38410 + }, + { + "epoch": 0.38411, + "grad_norm": 1.4350650524714668, + "learning_rate": 0.003, + "loss": 4.052, + "step": 38411 + }, + { + "epoch": 0.38412, + "grad_norm": 1.5054781280593121, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38412 + }, + { + "epoch": 0.38413, + "grad_norm": 1.5171029048525868, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 38413 + }, + { + "epoch": 0.38414, + "grad_norm": 1.6627064549861252, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38414 + }, + { + "epoch": 0.38415, + "grad_norm": 1.3968392722705745, + "learning_rate": 0.003, + "loss": 4.01, + "step": 38415 + }, + { + "epoch": 0.38416, + "grad_norm": 1.718854538878358, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38416 + }, + { + "epoch": 0.38417, + "grad_norm": 1.4277493483531247, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 38417 + }, + { + "epoch": 0.38418, + "grad_norm": 1.5389211971608987, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38418 + }, + { + "epoch": 0.38419, + "grad_norm": 1.5930920459903057, + "learning_rate": 0.003, + "loss": 4.038, + "step": 38419 + }, + { + "epoch": 0.3842, + "grad_norm": 1.2780603039097664, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 38420 + }, + { + "epoch": 0.38421, + "grad_norm": 1.783134751617627, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 38421 + }, + { + "epoch": 0.38422, + "grad_norm": 1.3050200846277218, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 38422 + }, + { + "epoch": 0.38423, + "grad_norm": 1.6295113513293236, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 38423 + }, + { + "epoch": 0.38424, + "grad_norm": 1.4015670962466678, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 38424 + }, + { + "epoch": 0.38425, + "grad_norm": 1.717049025696753, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 38425 + }, + { + "epoch": 0.38426, + "grad_norm": 1.938931301372068, + "learning_rate": 0.003, + "loss": 4.077, + "step": 38426 + }, + { + "epoch": 0.38427, + "grad_norm": 1.4444407470303404, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 38427 + }, + { + "epoch": 0.38428, + "grad_norm": 1.7439846187050725, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 38428 + }, + { + "epoch": 0.38429, + "grad_norm": 1.4683569933291518, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 38429 + }, + { + "epoch": 0.3843, + "grad_norm": 1.6975980504483674, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 38430 + }, + { + "epoch": 0.38431, + "grad_norm": 1.3336844417792433, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 38431 + }, + { + "epoch": 0.38432, + "grad_norm": 1.7543455590971841, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38432 + }, + { + "epoch": 0.38433, + "grad_norm": 1.1690187857908583, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38433 + }, + { + "epoch": 0.38434, + "grad_norm": 1.6068410451005803, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38434 + }, + { + "epoch": 0.38435, + "grad_norm": 1.355277736641451, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 38435 + }, + { + "epoch": 0.38436, + "grad_norm": 1.9753271790849967, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 38436 + }, + { + "epoch": 0.38437, + "grad_norm": 1.3940595042469486, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38437 + }, + { + "epoch": 0.38438, + "grad_norm": 1.6002312508725747, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 38438 + }, + { + "epoch": 0.38439, + "grad_norm": 1.4015550070378737, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 38439 + }, + { + "epoch": 0.3844, + "grad_norm": 1.5137730132007756, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38440 + }, + { + "epoch": 0.38441, + "grad_norm": 1.393594311001204, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 38441 + }, + { + "epoch": 0.38442, + "grad_norm": 1.4567050869356133, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 38442 + }, + { + "epoch": 0.38443, + "grad_norm": 1.5545816792417906, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 38443 + }, + { + "epoch": 0.38444, + "grad_norm": 1.3622684027631298, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 38444 + }, + { + "epoch": 0.38445, + "grad_norm": 1.5149865103668367, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 38445 + }, + { + "epoch": 0.38446, + "grad_norm": 1.7364748904303755, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 38446 + }, + { + "epoch": 0.38447, + "grad_norm": 1.5484744775999035, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 38447 + }, + { + "epoch": 0.38448, + "grad_norm": 1.296134455458775, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 38448 + }, + { + "epoch": 0.38449, + "grad_norm": 1.4947634418474074, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 38449 + }, + { + "epoch": 0.3845, + "grad_norm": 1.5213433687920364, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 38450 + }, + { + "epoch": 0.38451, + "grad_norm": 1.5964220346344251, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38451 + }, + { + "epoch": 0.38452, + "grad_norm": 1.722947780426428, + "learning_rate": 0.003, + "loss": 4.032, + "step": 38452 + }, + { + "epoch": 0.38453, + "grad_norm": 1.7331041992846934, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 38453 + }, + { + "epoch": 0.38454, + "grad_norm": 1.2650297867806388, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38454 + }, + { + "epoch": 0.38455, + "grad_norm": 1.491842148263649, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 38455 + }, + { + "epoch": 0.38456, + "grad_norm": 1.8757974586623916, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 38456 + }, + { + "epoch": 0.38457, + "grad_norm": 1.4826938205233922, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 38457 + }, + { + "epoch": 0.38458, + "grad_norm": 1.57917303291955, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 38458 + }, + { + "epoch": 0.38459, + "grad_norm": 1.4160723263357187, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38459 + }, + { + "epoch": 0.3846, + "grad_norm": 1.6411760727131517, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 38460 + }, + { + "epoch": 0.38461, + "grad_norm": 1.4632718380747822, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 38461 + }, + { + "epoch": 0.38462, + "grad_norm": 1.4985924797090977, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 38462 + }, + { + "epoch": 0.38463, + "grad_norm": 1.513389436581235, + "learning_rate": 0.003, + "loss": 4.021, + "step": 38463 + }, + { + "epoch": 0.38464, + "grad_norm": 1.9743167221337494, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38464 + }, + { + "epoch": 0.38465, + "grad_norm": 1.2847623347320924, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 38465 + }, + { + "epoch": 0.38466, + "grad_norm": 1.543122914200918, + "learning_rate": 0.003, + "loss": 4.024, + "step": 38466 + }, + { + "epoch": 0.38467, + "grad_norm": 1.4266358886527288, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 38467 + }, + { + "epoch": 0.38468, + "grad_norm": 1.6932046420881224, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 38468 + }, + { + "epoch": 0.38469, + "grad_norm": 1.4040591175589172, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38469 + }, + { + "epoch": 0.3847, + "grad_norm": 1.5108841346827888, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 38470 + }, + { + "epoch": 0.38471, + "grad_norm": 1.3974658339155943, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38471 + }, + { + "epoch": 0.38472, + "grad_norm": 1.3688484172713116, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38472 + }, + { + "epoch": 0.38473, + "grad_norm": 1.4045768094063502, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38473 + }, + { + "epoch": 0.38474, + "grad_norm": 1.7587738536492972, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38474 + }, + { + "epoch": 0.38475, + "grad_norm": 1.2687213728706819, + "learning_rate": 0.003, + "loss": 4.041, + "step": 38475 + }, + { + "epoch": 0.38476, + "grad_norm": 1.6988814346161916, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 38476 + }, + { + "epoch": 0.38477, + "grad_norm": 1.171999330990128, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38477 + }, + { + "epoch": 0.38478, + "grad_norm": 1.842180792620365, + "learning_rate": 0.003, + "loss": 4.047, + "step": 38478 + }, + { + "epoch": 0.38479, + "grad_norm": 1.3766896829642536, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38479 + }, + { + "epoch": 0.3848, + "grad_norm": 1.6992459223583816, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 38480 + }, + { + "epoch": 0.38481, + "grad_norm": 2.0336765333440154, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 38481 + }, + { + "epoch": 0.38482, + "grad_norm": 1.5449217596358178, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 38482 + }, + { + "epoch": 0.38483, + "grad_norm": 1.5103735135049032, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 38483 + }, + { + "epoch": 0.38484, + "grad_norm": 1.637168084392902, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 38484 + }, + { + "epoch": 0.38485, + "grad_norm": 1.6404531066023917, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 38485 + }, + { + "epoch": 0.38486, + "grad_norm": 1.47394023449808, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38486 + }, + { + "epoch": 0.38487, + "grad_norm": 1.5841791400399006, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38487 + }, + { + "epoch": 0.38488, + "grad_norm": 1.3754183296111742, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38488 + }, + { + "epoch": 0.38489, + "grad_norm": 1.7154307776718156, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 38489 + }, + { + "epoch": 0.3849, + "grad_norm": 1.2257128428627067, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 38490 + }, + { + "epoch": 0.38491, + "grad_norm": 1.668315860203252, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 38491 + }, + { + "epoch": 0.38492, + "grad_norm": 1.5161631275572462, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38492 + }, + { + "epoch": 0.38493, + "grad_norm": 1.5148844276254034, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 38493 + }, + { + "epoch": 0.38494, + "grad_norm": 1.4124139374108264, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38494 + }, + { + "epoch": 0.38495, + "grad_norm": 1.7464807112054481, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 38495 + }, + { + "epoch": 0.38496, + "grad_norm": 1.570091632098605, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 38496 + }, + { + "epoch": 0.38497, + "grad_norm": 1.5372341879073572, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38497 + }, + { + "epoch": 0.38498, + "grad_norm": 1.6878655238571698, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 38498 + }, + { + "epoch": 0.38499, + "grad_norm": 1.5478097889168692, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 38499 + }, + { + "epoch": 0.385, + "grad_norm": 1.252442051365372, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 38500 + }, + { + "epoch": 0.38501, + "grad_norm": 1.5662115338759204, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 38501 + }, + { + "epoch": 0.38502, + "grad_norm": 1.5606927273918214, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 38502 + }, + { + "epoch": 0.38503, + "grad_norm": 1.5770592177724196, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 38503 + }, + { + "epoch": 0.38504, + "grad_norm": 1.4265637725717384, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 38504 + }, + { + "epoch": 0.38505, + "grad_norm": 1.5187688503296948, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 38505 + }, + { + "epoch": 0.38506, + "grad_norm": 1.485356473482925, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38506 + }, + { + "epoch": 0.38507, + "grad_norm": 1.488395832845041, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 38507 + }, + { + "epoch": 0.38508, + "grad_norm": 1.4972786057061096, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 38508 + }, + { + "epoch": 0.38509, + "grad_norm": 1.2218036238131242, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 38509 + }, + { + "epoch": 0.3851, + "grad_norm": 1.2435079606388046, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38510 + }, + { + "epoch": 0.38511, + "grad_norm": 1.746322481227093, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38511 + }, + { + "epoch": 0.38512, + "grad_norm": 1.241805849850967, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38512 + }, + { + "epoch": 0.38513, + "grad_norm": 2.0002696646656406, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 38513 + }, + { + "epoch": 0.38514, + "grad_norm": 1.80170026962112, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38514 + }, + { + "epoch": 0.38515, + "grad_norm": 1.7289644277599934, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 38515 + }, + { + "epoch": 0.38516, + "grad_norm": 1.6648286875509652, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38516 + }, + { + "epoch": 0.38517, + "grad_norm": 1.45912208947607, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 38517 + }, + { + "epoch": 0.38518, + "grad_norm": 1.3867228980280388, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 38518 + }, + { + "epoch": 0.38519, + "grad_norm": 1.791662578297118, + "learning_rate": 0.003, + "loss": 4.046, + "step": 38519 + }, + { + "epoch": 0.3852, + "grad_norm": 1.321102198132431, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 38520 + }, + { + "epoch": 0.38521, + "grad_norm": 1.9680993196615175, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 38521 + }, + { + "epoch": 0.38522, + "grad_norm": 1.356310289854958, + "learning_rate": 0.003, + "loss": 4.042, + "step": 38522 + }, + { + "epoch": 0.38523, + "grad_norm": 1.5998628770356125, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 38523 + }, + { + "epoch": 0.38524, + "grad_norm": 1.4047055916637061, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 38524 + }, + { + "epoch": 0.38525, + "grad_norm": 1.6665251710711189, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 38525 + }, + { + "epoch": 0.38526, + "grad_norm": 1.301694347544213, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 38526 + }, + { + "epoch": 0.38527, + "grad_norm": 1.456645341696021, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 38527 + }, + { + "epoch": 0.38528, + "grad_norm": 1.2349006697912892, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 38528 + }, + { + "epoch": 0.38529, + "grad_norm": 1.6581122300761209, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38529 + }, + { + "epoch": 0.3853, + "grad_norm": 1.4039888557090747, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 38530 + }, + { + "epoch": 0.38531, + "grad_norm": 1.601767774505015, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 38531 + }, + { + "epoch": 0.38532, + "grad_norm": 1.468851113815337, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 38532 + }, + { + "epoch": 0.38533, + "grad_norm": 1.6864072592706285, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38533 + }, + { + "epoch": 0.38534, + "grad_norm": 1.4859756593765885, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 38534 + }, + { + "epoch": 0.38535, + "grad_norm": 1.8786592805197997, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38535 + }, + { + "epoch": 0.38536, + "grad_norm": 1.338467976037726, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 38536 + }, + { + "epoch": 0.38537, + "grad_norm": 1.6953653697429256, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 38537 + }, + { + "epoch": 0.38538, + "grad_norm": 1.628859130861397, + "learning_rate": 0.003, + "loss": 4.048, + "step": 38538 + }, + { + "epoch": 0.38539, + "grad_norm": 1.285978412917146, + "learning_rate": 0.003, + "loss": 4.041, + "step": 38539 + }, + { + "epoch": 0.3854, + "grad_norm": 1.7603030531175983, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 38540 + }, + { + "epoch": 0.38541, + "grad_norm": 1.543537955832053, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38541 + }, + { + "epoch": 0.38542, + "grad_norm": 1.487362739168186, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 38542 + }, + { + "epoch": 0.38543, + "grad_norm": 1.3193793257180981, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 38543 + }, + { + "epoch": 0.38544, + "grad_norm": 1.5110299028868446, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 38544 + }, + { + "epoch": 0.38545, + "grad_norm": 1.4395124005426432, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 38545 + }, + { + "epoch": 0.38546, + "grad_norm": 1.4422141841157812, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 38546 + }, + { + "epoch": 0.38547, + "grad_norm": 1.4582818346894055, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 38547 + }, + { + "epoch": 0.38548, + "grad_norm": 1.4711050024708237, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 38548 + }, + { + "epoch": 0.38549, + "grad_norm": 1.380165306970563, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38549 + }, + { + "epoch": 0.3855, + "grad_norm": 1.708033945977362, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38550 + }, + { + "epoch": 0.38551, + "grad_norm": 1.2537981219619736, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 38551 + }, + { + "epoch": 0.38552, + "grad_norm": 1.9203953057386625, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38552 + }, + { + "epoch": 0.38553, + "grad_norm": 1.347169549913741, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 38553 + }, + { + "epoch": 0.38554, + "grad_norm": 1.8825003083846885, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 38554 + }, + { + "epoch": 0.38555, + "grad_norm": 1.4987057864684172, + "learning_rate": 0.003, + "loss": 4.035, + "step": 38555 + }, + { + "epoch": 0.38556, + "grad_norm": 1.743007048919216, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 38556 + }, + { + "epoch": 0.38557, + "grad_norm": 1.3176924941598809, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 38557 + }, + { + "epoch": 0.38558, + "grad_norm": 1.4527827029288545, + "learning_rate": 0.003, + "loss": 4.017, + "step": 38558 + }, + { + "epoch": 0.38559, + "grad_norm": 1.5215192747798085, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 38559 + }, + { + "epoch": 0.3856, + "grad_norm": 1.4702937888733525, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38560 + }, + { + "epoch": 0.38561, + "grad_norm": 1.6336652025372413, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 38561 + }, + { + "epoch": 0.38562, + "grad_norm": 1.4278679255303843, + "learning_rate": 0.003, + "loss": 4.048, + "step": 38562 + }, + { + "epoch": 0.38563, + "grad_norm": 1.5151901805967185, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 38563 + }, + { + "epoch": 0.38564, + "grad_norm": 1.4808631493431572, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 38564 + }, + { + "epoch": 0.38565, + "grad_norm": 1.4948395579953258, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 38565 + }, + { + "epoch": 0.38566, + "grad_norm": 1.37513643989554, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 38566 + }, + { + "epoch": 0.38567, + "grad_norm": 1.6251365741954622, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38567 + }, + { + "epoch": 0.38568, + "grad_norm": 1.8182540066618496, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 38568 + }, + { + "epoch": 0.38569, + "grad_norm": 1.3332696834860505, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 38569 + }, + { + "epoch": 0.3857, + "grad_norm": 1.533484892645711, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38570 + }, + { + "epoch": 0.38571, + "grad_norm": 1.476082152091461, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 38571 + }, + { + "epoch": 0.38572, + "grad_norm": 1.9050901221949461, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 38572 + }, + { + "epoch": 0.38573, + "grad_norm": 1.275363913516145, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 38573 + }, + { + "epoch": 0.38574, + "grad_norm": 1.5457750620933493, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38574 + }, + { + "epoch": 0.38575, + "grad_norm": 1.5188568204065998, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 38575 + }, + { + "epoch": 0.38576, + "grad_norm": 1.4363859277771027, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38576 + }, + { + "epoch": 0.38577, + "grad_norm": 1.6059433544835022, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38577 + }, + { + "epoch": 0.38578, + "grad_norm": 1.432853462403582, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38578 + }, + { + "epoch": 0.38579, + "grad_norm": 1.4717592272444038, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38579 + }, + { + "epoch": 0.3858, + "grad_norm": 1.8880561771640196, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 38580 + }, + { + "epoch": 0.38581, + "grad_norm": 1.4960100843070527, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 38581 + }, + { + "epoch": 0.38582, + "grad_norm": 1.4779503678982024, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 38582 + }, + { + "epoch": 0.38583, + "grad_norm": 1.4901158355500175, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 38583 + }, + { + "epoch": 0.38584, + "grad_norm": 1.3455672501971827, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 38584 + }, + { + "epoch": 0.38585, + "grad_norm": 1.5150471424148464, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 38585 + }, + { + "epoch": 0.38586, + "grad_norm": 1.4752842016698913, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 38586 + }, + { + "epoch": 0.38587, + "grad_norm": 1.9255970785982803, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 38587 + }, + { + "epoch": 0.38588, + "grad_norm": 1.394970650296837, + "learning_rate": 0.003, + "loss": 4.075, + "step": 38588 + }, + { + "epoch": 0.38589, + "grad_norm": 1.5834354210316925, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38589 + }, + { + "epoch": 0.3859, + "grad_norm": 1.2173927441969938, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 38590 + }, + { + "epoch": 0.38591, + "grad_norm": 1.629232460204511, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 38591 + }, + { + "epoch": 0.38592, + "grad_norm": 1.278987752396149, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 38592 + }, + { + "epoch": 0.38593, + "grad_norm": 1.482054573880177, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 38593 + }, + { + "epoch": 0.38594, + "grad_norm": 1.468457178093815, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 38594 + }, + { + "epoch": 0.38595, + "grad_norm": 1.3941620405113795, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 38595 + }, + { + "epoch": 0.38596, + "grad_norm": 1.5656242219394085, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 38596 + }, + { + "epoch": 0.38597, + "grad_norm": 1.4913801736603247, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 38597 + }, + { + "epoch": 0.38598, + "grad_norm": 1.7087813767456435, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 38598 + }, + { + "epoch": 0.38599, + "grad_norm": 1.576862557797441, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38599 + }, + { + "epoch": 0.386, + "grad_norm": 1.5413271658052776, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 38600 + }, + { + "epoch": 0.38601, + "grad_norm": 1.4574631586725204, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 38601 + }, + { + "epoch": 0.38602, + "grad_norm": 1.6492284182235852, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 38602 + }, + { + "epoch": 0.38603, + "grad_norm": 1.2840181876195798, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 38603 + }, + { + "epoch": 0.38604, + "grad_norm": 1.8268357607262165, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 38604 + }, + { + "epoch": 0.38605, + "grad_norm": 1.2896661959292721, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 38605 + }, + { + "epoch": 0.38606, + "grad_norm": 2.163843869870334, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 38606 + }, + { + "epoch": 0.38607, + "grad_norm": 1.0591846999698526, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 38607 + }, + { + "epoch": 0.38608, + "grad_norm": 1.7267993287164722, + "learning_rate": 0.003, + "loss": 4.06, + "step": 38608 + }, + { + "epoch": 0.38609, + "grad_norm": 1.5730301174555845, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 38609 + }, + { + "epoch": 0.3861, + "grad_norm": 1.3751624409155379, + "learning_rate": 0.003, + "loss": 4.038, + "step": 38610 + }, + { + "epoch": 0.38611, + "grad_norm": 1.5689623665945283, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 38611 + }, + { + "epoch": 0.38612, + "grad_norm": 1.6175618152375593, + "learning_rate": 0.003, + "loss": 4.06, + "step": 38612 + }, + { + "epoch": 0.38613, + "grad_norm": 1.4732834422852494, + "learning_rate": 0.003, + "loss": 4.023, + "step": 38613 + }, + { + "epoch": 0.38614, + "grad_norm": 1.8075406109997876, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38614 + }, + { + "epoch": 0.38615, + "grad_norm": 1.3470690564382173, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 38615 + }, + { + "epoch": 0.38616, + "grad_norm": 1.6812341413128828, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 38616 + }, + { + "epoch": 0.38617, + "grad_norm": 1.5267145678777192, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 38617 + }, + { + "epoch": 0.38618, + "grad_norm": 1.6122071648775496, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38618 + }, + { + "epoch": 0.38619, + "grad_norm": 1.5715381584393555, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 38619 + }, + { + "epoch": 0.3862, + "grad_norm": 1.371387596319893, + "learning_rate": 0.003, + "loss": 4.06, + "step": 38620 + }, + { + "epoch": 0.38621, + "grad_norm": 1.6368529091748196, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 38621 + }, + { + "epoch": 0.38622, + "grad_norm": 1.350774562665224, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 38622 + }, + { + "epoch": 0.38623, + "grad_norm": 1.4853587485147246, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 38623 + }, + { + "epoch": 0.38624, + "grad_norm": 1.4514045331813932, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 38624 + }, + { + "epoch": 0.38625, + "grad_norm": 1.4469402789082009, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38625 + }, + { + "epoch": 0.38626, + "grad_norm": 1.4910170021940659, + "learning_rate": 0.003, + "loss": 4.022, + "step": 38626 + }, + { + "epoch": 0.38627, + "grad_norm": 1.8447255471178456, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38627 + }, + { + "epoch": 0.38628, + "grad_norm": 1.2262581555451457, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 38628 + }, + { + "epoch": 0.38629, + "grad_norm": 1.6178815108454812, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 38629 + }, + { + "epoch": 0.3863, + "grad_norm": 1.9427176144318807, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 38630 + }, + { + "epoch": 0.38631, + "grad_norm": 1.524905226042078, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 38631 + }, + { + "epoch": 0.38632, + "grad_norm": 1.5491722997961712, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38632 + }, + { + "epoch": 0.38633, + "grad_norm": 1.3710204254425802, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 38633 + }, + { + "epoch": 0.38634, + "grad_norm": 1.4319956235291353, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 38634 + }, + { + "epoch": 0.38635, + "grad_norm": 1.6419802919829123, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38635 + }, + { + "epoch": 0.38636, + "grad_norm": 1.3510034743388846, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38636 + }, + { + "epoch": 0.38637, + "grad_norm": 1.7641649434054445, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 38637 + }, + { + "epoch": 0.38638, + "grad_norm": 1.4234172848652198, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38638 + }, + { + "epoch": 0.38639, + "grad_norm": 1.6562998799800384, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 38639 + }, + { + "epoch": 0.3864, + "grad_norm": 1.5116372949789587, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 38640 + }, + { + "epoch": 0.38641, + "grad_norm": 1.7331099485370816, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38641 + }, + { + "epoch": 0.38642, + "grad_norm": 1.291447207980926, + "learning_rate": 0.003, + "loss": 4.029, + "step": 38642 + }, + { + "epoch": 0.38643, + "grad_norm": 1.6930001346487387, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 38643 + }, + { + "epoch": 0.38644, + "grad_norm": 1.3079570528568887, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 38644 + }, + { + "epoch": 0.38645, + "grad_norm": 1.7851936524655023, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 38645 + }, + { + "epoch": 0.38646, + "grad_norm": 1.288562652326159, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 38646 + }, + { + "epoch": 0.38647, + "grad_norm": 1.6550916510185796, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 38647 + }, + { + "epoch": 0.38648, + "grad_norm": 1.8346458883479366, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38648 + }, + { + "epoch": 0.38649, + "grad_norm": 1.4074674162294247, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38649 + }, + { + "epoch": 0.3865, + "grad_norm": 1.5407525409431229, + "learning_rate": 0.003, + "loss": 4.021, + "step": 38650 + }, + { + "epoch": 0.38651, + "grad_norm": 1.6523367034100978, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 38651 + }, + { + "epoch": 0.38652, + "grad_norm": 1.3986796963893877, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 38652 + }, + { + "epoch": 0.38653, + "grad_norm": 1.5254897769954445, + "learning_rate": 0.003, + "loss": 4.047, + "step": 38653 + }, + { + "epoch": 0.38654, + "grad_norm": 1.3380737546555523, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 38654 + }, + { + "epoch": 0.38655, + "grad_norm": 1.5096934375367033, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 38655 + }, + { + "epoch": 0.38656, + "grad_norm": 1.6709331533815361, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 38656 + }, + { + "epoch": 0.38657, + "grad_norm": 1.3041066303610247, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 38657 + }, + { + "epoch": 0.38658, + "grad_norm": 1.5929367266521068, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 38658 + }, + { + "epoch": 0.38659, + "grad_norm": 1.5038035387138216, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38659 + }, + { + "epoch": 0.3866, + "grad_norm": 1.8422606609319676, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 38660 + }, + { + "epoch": 0.38661, + "grad_norm": 1.198007454517008, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 38661 + }, + { + "epoch": 0.38662, + "grad_norm": 1.7540121838397562, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 38662 + }, + { + "epoch": 0.38663, + "grad_norm": 1.5453192950873718, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 38663 + }, + { + "epoch": 0.38664, + "grad_norm": 1.477558149307919, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 38664 + }, + { + "epoch": 0.38665, + "grad_norm": 1.5041808848567815, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 38665 + }, + { + "epoch": 0.38666, + "grad_norm": 1.5920941465239653, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 38666 + }, + { + "epoch": 0.38667, + "grad_norm": 1.2150642919029173, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 38667 + }, + { + "epoch": 0.38668, + "grad_norm": 1.8606752509971507, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 38668 + }, + { + "epoch": 0.38669, + "grad_norm": 1.321431930873787, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 38669 + }, + { + "epoch": 0.3867, + "grad_norm": 1.8763534595316063, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 38670 + }, + { + "epoch": 0.38671, + "grad_norm": 1.3537495145198064, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 38671 + }, + { + "epoch": 0.38672, + "grad_norm": 1.7355004669392704, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 38672 + }, + { + "epoch": 0.38673, + "grad_norm": 1.5757923123908464, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 38673 + }, + { + "epoch": 0.38674, + "grad_norm": 1.6019644450569863, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 38674 + }, + { + "epoch": 0.38675, + "grad_norm": 1.6228946693413246, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 38675 + }, + { + "epoch": 0.38676, + "grad_norm": 1.4972891833882593, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 38676 + }, + { + "epoch": 0.38677, + "grad_norm": 1.1473968397525323, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 38677 + }, + { + "epoch": 0.38678, + "grad_norm": 1.5219861411093827, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 38678 + }, + { + "epoch": 0.38679, + "grad_norm": 1.5136191792880982, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38679 + }, + { + "epoch": 0.3868, + "grad_norm": 1.6769013605711593, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 38680 + }, + { + "epoch": 0.38681, + "grad_norm": 1.6275939808810727, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38681 + }, + { + "epoch": 0.38682, + "grad_norm": 1.4261384236613561, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 38682 + }, + { + "epoch": 0.38683, + "grad_norm": 1.4045907130380417, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38683 + }, + { + "epoch": 0.38684, + "grad_norm": 1.7676196345647857, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 38684 + }, + { + "epoch": 0.38685, + "grad_norm": 1.4303942362921116, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 38685 + }, + { + "epoch": 0.38686, + "grad_norm": 1.5859772999403974, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 38686 + }, + { + "epoch": 0.38687, + "grad_norm": 1.3305762115590227, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38687 + }, + { + "epoch": 0.38688, + "grad_norm": 1.5272344796100648, + "learning_rate": 0.003, + "loss": 4.059, + "step": 38688 + }, + { + "epoch": 0.38689, + "grad_norm": 1.3864690910154829, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38689 + }, + { + "epoch": 0.3869, + "grad_norm": 1.6118112612242819, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 38690 + }, + { + "epoch": 0.38691, + "grad_norm": 1.5182661228151875, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38691 + }, + { + "epoch": 0.38692, + "grad_norm": 1.3837006746214313, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38692 + }, + { + "epoch": 0.38693, + "grad_norm": 1.8384194699191674, + "learning_rate": 0.003, + "loss": 4.049, + "step": 38693 + }, + { + "epoch": 0.38694, + "grad_norm": 1.885192082237664, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 38694 + }, + { + "epoch": 0.38695, + "grad_norm": 1.4284855704135797, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38695 + }, + { + "epoch": 0.38696, + "grad_norm": 1.7728743534929086, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 38696 + }, + { + "epoch": 0.38697, + "grad_norm": 1.2475401796732783, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 38697 + }, + { + "epoch": 0.38698, + "grad_norm": 1.8838245996385239, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 38698 + }, + { + "epoch": 0.38699, + "grad_norm": 1.1509342619883018, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 38699 + }, + { + "epoch": 0.387, + "grad_norm": 1.92624476160236, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 38700 + }, + { + "epoch": 0.38701, + "grad_norm": 1.6654815797683238, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 38701 + }, + { + "epoch": 0.38702, + "grad_norm": 1.4945233809516816, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 38702 + }, + { + "epoch": 0.38703, + "grad_norm": 1.682370970862992, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38703 + }, + { + "epoch": 0.38704, + "grad_norm": 1.7041363427951188, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 38704 + }, + { + "epoch": 0.38705, + "grad_norm": 1.2103204816623434, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 38705 + }, + { + "epoch": 0.38706, + "grad_norm": 1.6800853654202998, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 38706 + }, + { + "epoch": 0.38707, + "grad_norm": 1.2880937165864392, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 38707 + }, + { + "epoch": 0.38708, + "grad_norm": 1.9495510840042092, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38708 + }, + { + "epoch": 0.38709, + "grad_norm": 1.19522261219684, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 38709 + }, + { + "epoch": 0.3871, + "grad_norm": 1.5902902310567508, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38710 + }, + { + "epoch": 0.38711, + "grad_norm": 1.6087253340838328, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 38711 + }, + { + "epoch": 0.38712, + "grad_norm": 1.4542477923515007, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 38712 + }, + { + "epoch": 0.38713, + "grad_norm": 1.4283434492980462, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 38713 + }, + { + "epoch": 0.38714, + "grad_norm": 1.4204263426989128, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 38714 + }, + { + "epoch": 0.38715, + "grad_norm": 1.3995881017948855, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38715 + }, + { + "epoch": 0.38716, + "grad_norm": 1.4894502523127346, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38716 + }, + { + "epoch": 0.38717, + "grad_norm": 1.444462220027978, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 38717 + }, + { + "epoch": 0.38718, + "grad_norm": 1.635360632221773, + "learning_rate": 0.003, + "loss": 4.035, + "step": 38718 + }, + { + "epoch": 0.38719, + "grad_norm": 1.5056905889929755, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 38719 + }, + { + "epoch": 0.3872, + "grad_norm": 1.642471908450372, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 38720 + }, + { + "epoch": 0.38721, + "grad_norm": 1.5418054485698582, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 38721 + }, + { + "epoch": 0.38722, + "grad_norm": 1.4297514691206752, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 38722 + }, + { + "epoch": 0.38723, + "grad_norm": 1.4676491101082796, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 38723 + }, + { + "epoch": 0.38724, + "grad_norm": 1.527408600451223, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 38724 + }, + { + "epoch": 0.38725, + "grad_norm": 1.4894640377369104, + "learning_rate": 0.003, + "loss": 4.056, + "step": 38725 + }, + { + "epoch": 0.38726, + "grad_norm": 1.5542268988041152, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 38726 + }, + { + "epoch": 0.38727, + "grad_norm": 1.635684927397225, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 38727 + }, + { + "epoch": 0.38728, + "grad_norm": 1.5173482280125126, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38728 + }, + { + "epoch": 0.38729, + "grad_norm": 1.3616079115380784, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 38729 + }, + { + "epoch": 0.3873, + "grad_norm": 1.5578999507889837, + "learning_rate": 0.003, + "loss": 4.058, + "step": 38730 + }, + { + "epoch": 0.38731, + "grad_norm": 1.462997304611008, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 38731 + }, + { + "epoch": 0.38732, + "grad_norm": 1.5902627223171122, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38732 + }, + { + "epoch": 0.38733, + "grad_norm": 1.5079471003428293, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 38733 + }, + { + "epoch": 0.38734, + "grad_norm": 1.5331037680158721, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 38734 + }, + { + "epoch": 0.38735, + "grad_norm": 1.4869064827970404, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 38735 + }, + { + "epoch": 0.38736, + "grad_norm": 1.8116847658984494, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 38736 + }, + { + "epoch": 0.38737, + "grad_norm": 1.503269628524926, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 38737 + }, + { + "epoch": 0.38738, + "grad_norm": 1.9481290139415925, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 38738 + }, + { + "epoch": 0.38739, + "grad_norm": 1.5329467251238533, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 38739 + }, + { + "epoch": 0.3874, + "grad_norm": 1.5053257670411944, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 38740 + }, + { + "epoch": 0.38741, + "grad_norm": 1.679623349802803, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 38741 + }, + { + "epoch": 0.38742, + "grad_norm": 1.7634717293934599, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 38742 + }, + { + "epoch": 0.38743, + "grad_norm": 1.3359248027521022, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 38743 + }, + { + "epoch": 0.38744, + "grad_norm": 1.4036680323076536, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38744 + }, + { + "epoch": 0.38745, + "grad_norm": 1.5141615366521095, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 38745 + }, + { + "epoch": 0.38746, + "grad_norm": 1.3727444484106728, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38746 + }, + { + "epoch": 0.38747, + "grad_norm": 1.7279720805309526, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 38747 + }, + { + "epoch": 0.38748, + "grad_norm": 1.6560556670252402, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 38748 + }, + { + "epoch": 0.38749, + "grad_norm": 1.6581266584211516, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 38749 + }, + { + "epoch": 0.3875, + "grad_norm": 1.4954984652669692, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 38750 + }, + { + "epoch": 0.38751, + "grad_norm": 1.4711744851908188, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 38751 + }, + { + "epoch": 0.38752, + "grad_norm": 1.6573073630463948, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 38752 + }, + { + "epoch": 0.38753, + "grad_norm": 1.4522278008486773, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 38753 + }, + { + "epoch": 0.38754, + "grad_norm": 1.4382623176106208, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 38754 + }, + { + "epoch": 0.38755, + "grad_norm": 1.4922500063109898, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 38755 + }, + { + "epoch": 0.38756, + "grad_norm": 1.3703627917478491, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 38756 + }, + { + "epoch": 0.38757, + "grad_norm": 1.5641707403648184, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 38757 + }, + { + "epoch": 0.38758, + "grad_norm": 1.6132267840528116, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 38758 + }, + { + "epoch": 0.38759, + "grad_norm": 1.58982061547507, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 38759 + }, + { + "epoch": 0.3876, + "grad_norm": 1.5452344305136805, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 38760 + }, + { + "epoch": 0.38761, + "grad_norm": 1.7093747355638227, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 38761 + }, + { + "epoch": 0.38762, + "grad_norm": 1.3752368773246446, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 38762 + }, + { + "epoch": 0.38763, + "grad_norm": 1.5654901852042147, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 38763 + }, + { + "epoch": 0.38764, + "grad_norm": 1.3272734569053435, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 38764 + }, + { + "epoch": 0.38765, + "grad_norm": 1.5856874774054972, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 38765 + }, + { + "epoch": 0.38766, + "grad_norm": 1.864917663627813, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 38766 + }, + { + "epoch": 0.38767, + "grad_norm": 1.370617873826358, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 38767 + }, + { + "epoch": 0.38768, + "grad_norm": 1.8092087403539376, + "learning_rate": 0.003, + "loss": 4.044, + "step": 38768 + }, + { + "epoch": 0.38769, + "grad_norm": 1.21492912566961, + "learning_rate": 0.003, + "loss": 4.025, + "step": 38769 + }, + { + "epoch": 0.3877, + "grad_norm": 1.6159163347993866, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38770 + }, + { + "epoch": 0.38771, + "grad_norm": 1.395673139125319, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 38771 + }, + { + "epoch": 0.38772, + "grad_norm": 1.7996461324844037, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 38772 + }, + { + "epoch": 0.38773, + "grad_norm": 1.3804000687462614, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38773 + }, + { + "epoch": 0.38774, + "grad_norm": 1.3688825377355613, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38774 + }, + { + "epoch": 0.38775, + "grad_norm": 1.610307173920792, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 38775 + }, + { + "epoch": 0.38776, + "grad_norm": 1.3163477928795768, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38776 + }, + { + "epoch": 0.38777, + "grad_norm": 1.7332992772140865, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 38777 + }, + { + "epoch": 0.38778, + "grad_norm": 1.4581834738869537, + "learning_rate": 0.003, + "loss": 4.039, + "step": 38778 + }, + { + "epoch": 0.38779, + "grad_norm": 2.258398139686029, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 38779 + }, + { + "epoch": 0.3878, + "grad_norm": 1.429510823988078, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 38780 + }, + { + "epoch": 0.38781, + "grad_norm": 1.5503922353074404, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 38781 + }, + { + "epoch": 0.38782, + "grad_norm": 1.6154051043452462, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 38782 + }, + { + "epoch": 0.38783, + "grad_norm": 1.5948871914498426, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 38783 + }, + { + "epoch": 0.38784, + "grad_norm": 1.4528462866719467, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38784 + }, + { + "epoch": 0.38785, + "grad_norm": 1.5267941270017746, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 38785 + }, + { + "epoch": 0.38786, + "grad_norm": 1.6379219922545092, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 38786 + }, + { + "epoch": 0.38787, + "grad_norm": 1.2991452403856545, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 38787 + }, + { + "epoch": 0.38788, + "grad_norm": 1.7732115052371475, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 38788 + }, + { + "epoch": 0.38789, + "grad_norm": 1.3767564516531081, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38789 + }, + { + "epoch": 0.3879, + "grad_norm": 1.5694204902094082, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 38790 + }, + { + "epoch": 0.38791, + "grad_norm": 1.2656226023002843, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 38791 + }, + { + "epoch": 0.38792, + "grad_norm": 1.4725276201695574, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 38792 + }, + { + "epoch": 0.38793, + "grad_norm": 1.4201308265201058, + "learning_rate": 0.003, + "loss": 4.043, + "step": 38793 + }, + { + "epoch": 0.38794, + "grad_norm": 1.2737981902578708, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38794 + }, + { + "epoch": 0.38795, + "grad_norm": 1.5926118698831797, + "learning_rate": 0.003, + "loss": 4.065, + "step": 38795 + }, + { + "epoch": 0.38796, + "grad_norm": 1.5424208085751903, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 38796 + }, + { + "epoch": 0.38797, + "grad_norm": 2.1490504652797795, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 38797 + }, + { + "epoch": 0.38798, + "grad_norm": 1.2739363809290527, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 38798 + }, + { + "epoch": 0.38799, + "grad_norm": 2.1013269764482314, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38799 + }, + { + "epoch": 0.388, + "grad_norm": 1.6659996795775183, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 38800 + }, + { + "epoch": 0.38801, + "grad_norm": 1.4692961192956528, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 38801 + }, + { + "epoch": 0.38802, + "grad_norm": 1.4627493504618814, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 38802 + }, + { + "epoch": 0.38803, + "grad_norm": 1.6851637283848462, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 38803 + }, + { + "epoch": 0.38804, + "grad_norm": 1.1099675829073432, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 38804 + }, + { + "epoch": 0.38805, + "grad_norm": 1.4498148005105393, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 38805 + }, + { + "epoch": 0.38806, + "grad_norm": 1.4092998923781823, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 38806 + }, + { + "epoch": 0.38807, + "grad_norm": 1.6834834684217315, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38807 + }, + { + "epoch": 0.38808, + "grad_norm": 1.2806868946291723, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 38808 + }, + { + "epoch": 0.38809, + "grad_norm": 1.8207063215573396, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 38809 + }, + { + "epoch": 0.3881, + "grad_norm": 2.125875513170036, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 38810 + }, + { + "epoch": 0.38811, + "grad_norm": 1.860896665831221, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 38811 + }, + { + "epoch": 0.38812, + "grad_norm": 1.364837868446146, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 38812 + }, + { + "epoch": 0.38813, + "grad_norm": 1.8633364878871255, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 38813 + }, + { + "epoch": 0.38814, + "grad_norm": 1.273620900991391, + "learning_rate": 0.003, + "loss": 4.045, + "step": 38814 + }, + { + "epoch": 0.38815, + "grad_norm": 1.7156882830509406, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 38815 + }, + { + "epoch": 0.38816, + "grad_norm": 1.432943355372162, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 38816 + }, + { + "epoch": 0.38817, + "grad_norm": 1.4973297989198946, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 38817 + }, + { + "epoch": 0.38818, + "grad_norm": 1.3947037377857348, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 38818 + }, + { + "epoch": 0.38819, + "grad_norm": 1.533390973084299, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 38819 + }, + { + "epoch": 0.3882, + "grad_norm": 1.760909393306886, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 38820 + }, + { + "epoch": 0.38821, + "grad_norm": 1.3126508728375181, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 38821 + }, + { + "epoch": 0.38822, + "grad_norm": 1.6617051863934615, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 38822 + }, + { + "epoch": 0.38823, + "grad_norm": 1.5226565945130712, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 38823 + }, + { + "epoch": 0.38824, + "grad_norm": 1.9249139882710644, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 38824 + }, + { + "epoch": 0.38825, + "grad_norm": 1.4221681626986937, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 38825 + }, + { + "epoch": 0.38826, + "grad_norm": 1.5411474996635286, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 38826 + }, + { + "epoch": 0.38827, + "grad_norm": 1.570351670389688, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 38827 + }, + { + "epoch": 0.38828, + "grad_norm": 1.379233435444864, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 38828 + }, + { + "epoch": 0.38829, + "grad_norm": 1.6486634236156006, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38829 + }, + { + "epoch": 0.3883, + "grad_norm": 1.4007118477216092, + "learning_rate": 0.003, + "loss": 4.036, + "step": 38830 + }, + { + "epoch": 0.38831, + "grad_norm": 1.8002141076821754, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38831 + }, + { + "epoch": 0.38832, + "grad_norm": 1.4852282808600163, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38832 + }, + { + "epoch": 0.38833, + "grad_norm": 2.026392380240002, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 38833 + }, + { + "epoch": 0.38834, + "grad_norm": 1.4368723955840532, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 38834 + }, + { + "epoch": 0.38835, + "grad_norm": 1.5327218240971785, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 38835 + }, + { + "epoch": 0.38836, + "grad_norm": 1.3693905454070547, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38836 + }, + { + "epoch": 0.38837, + "grad_norm": 1.6572242780567241, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 38837 + }, + { + "epoch": 0.38838, + "grad_norm": 1.2931441877380339, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 38838 + }, + { + "epoch": 0.38839, + "grad_norm": 1.6719270983826025, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 38839 + }, + { + "epoch": 0.3884, + "grad_norm": 1.4582855874037668, + "learning_rate": 0.003, + "loss": 4.041, + "step": 38840 + }, + { + "epoch": 0.38841, + "grad_norm": 1.4482989939531725, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 38841 + }, + { + "epoch": 0.38842, + "grad_norm": 1.3565622744881687, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 38842 + }, + { + "epoch": 0.38843, + "grad_norm": 1.7928871164885418, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 38843 + }, + { + "epoch": 0.38844, + "grad_norm": 1.5888707451898667, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 38844 + }, + { + "epoch": 0.38845, + "grad_norm": 1.4252305210108358, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 38845 + }, + { + "epoch": 0.38846, + "grad_norm": 1.5339242536842936, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 38846 + }, + { + "epoch": 0.38847, + "grad_norm": 1.9884350914987488, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 38847 + }, + { + "epoch": 0.38848, + "grad_norm": 1.146849556406207, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 38848 + }, + { + "epoch": 0.38849, + "grad_norm": 1.553835370425014, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 38849 + }, + { + "epoch": 0.3885, + "grad_norm": 1.499291045939731, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38850 + }, + { + "epoch": 0.38851, + "grad_norm": 1.512272055277199, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 38851 + }, + { + "epoch": 0.38852, + "grad_norm": 1.5869326970255382, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 38852 + }, + { + "epoch": 0.38853, + "grad_norm": 1.4598898219094338, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 38853 + }, + { + "epoch": 0.38854, + "grad_norm": 1.6091742230372093, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 38854 + }, + { + "epoch": 0.38855, + "grad_norm": 1.508163903704725, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 38855 + }, + { + "epoch": 0.38856, + "grad_norm": 1.7967882947765799, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 38856 + }, + { + "epoch": 0.38857, + "grad_norm": 1.6076551414297982, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38857 + }, + { + "epoch": 0.38858, + "grad_norm": 1.5932878760234463, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 38858 + }, + { + "epoch": 0.38859, + "grad_norm": 1.6122478433419378, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 38859 + }, + { + "epoch": 0.3886, + "grad_norm": 1.2850911300549535, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 38860 + }, + { + "epoch": 0.38861, + "grad_norm": 1.6123521811320327, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 38861 + }, + { + "epoch": 0.38862, + "grad_norm": 1.5130755921274437, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 38862 + }, + { + "epoch": 0.38863, + "grad_norm": 1.7250873516368084, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 38863 + }, + { + "epoch": 0.38864, + "grad_norm": 1.6741224392296887, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 38864 + }, + { + "epoch": 0.38865, + "grad_norm": 1.6621346998419282, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 38865 + }, + { + "epoch": 0.38866, + "grad_norm": 1.2930281573276186, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 38866 + }, + { + "epoch": 0.38867, + "grad_norm": 1.663483923368135, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 38867 + }, + { + "epoch": 0.38868, + "grad_norm": 1.6269769561191711, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 38868 + }, + { + "epoch": 0.38869, + "grad_norm": 1.5453484955716954, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 38869 + }, + { + "epoch": 0.3887, + "grad_norm": 1.5409302759687746, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 38870 + }, + { + "epoch": 0.38871, + "grad_norm": 1.5966514107753034, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 38871 + }, + { + "epoch": 0.38872, + "grad_norm": 1.5268429226662616, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 38872 + }, + { + "epoch": 0.38873, + "grad_norm": 1.4358649354672846, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 38873 + }, + { + "epoch": 0.38874, + "grad_norm": 1.6079382660922337, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 38874 + }, + { + "epoch": 0.38875, + "grad_norm": 1.4080294575181045, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 38875 + }, + { + "epoch": 0.38876, + "grad_norm": 1.4762677400688642, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 38876 + }, + { + "epoch": 0.38877, + "grad_norm": 1.80820589463122, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38877 + }, + { + "epoch": 0.38878, + "grad_norm": 1.5867639375889786, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 38878 + }, + { + "epoch": 0.38879, + "grad_norm": 1.7019828225097202, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 38879 + }, + { + "epoch": 0.3888, + "grad_norm": 1.2626217535850963, + "learning_rate": 0.003, + "loss": 4.055, + "step": 38880 + }, + { + "epoch": 0.38881, + "grad_norm": 1.6838696672356264, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38881 + }, + { + "epoch": 0.38882, + "grad_norm": 1.4294353037816858, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38882 + }, + { + "epoch": 0.38883, + "grad_norm": 1.30457711606019, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38883 + }, + { + "epoch": 0.38884, + "grad_norm": 1.4823450262107778, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 38884 + }, + { + "epoch": 0.38885, + "grad_norm": 1.5695577930333817, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 38885 + }, + { + "epoch": 0.38886, + "grad_norm": 1.4203706539285736, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 38886 + }, + { + "epoch": 0.38887, + "grad_norm": 1.6233638669968578, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 38887 + }, + { + "epoch": 0.38888, + "grad_norm": 1.358189522758427, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 38888 + }, + { + "epoch": 0.38889, + "grad_norm": 1.7766743644640408, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 38889 + }, + { + "epoch": 0.3889, + "grad_norm": 1.3245641901445009, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38890 + }, + { + "epoch": 0.38891, + "grad_norm": 1.6755395328766012, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 38891 + }, + { + "epoch": 0.38892, + "grad_norm": 1.6318364097032052, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 38892 + }, + { + "epoch": 0.38893, + "grad_norm": 1.3917644491530041, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 38893 + }, + { + "epoch": 0.38894, + "grad_norm": 1.7098402284770835, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 38894 + }, + { + "epoch": 0.38895, + "grad_norm": 1.35266848570429, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 38895 + }, + { + "epoch": 0.38896, + "grad_norm": 1.8019913016301654, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 38896 + }, + { + "epoch": 0.38897, + "grad_norm": 1.3510264761601918, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 38897 + }, + { + "epoch": 0.38898, + "grad_norm": 1.7827309856869762, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 38898 + }, + { + "epoch": 0.38899, + "grad_norm": 1.6643244731543436, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 38899 + }, + { + "epoch": 0.389, + "grad_norm": 1.6246297884123426, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 38900 + }, + { + "epoch": 0.38901, + "grad_norm": 1.4436600116986669, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 38901 + }, + { + "epoch": 0.38902, + "grad_norm": 1.3814046685705947, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 38902 + }, + { + "epoch": 0.38903, + "grad_norm": 1.6848503878335135, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 38903 + }, + { + "epoch": 0.38904, + "grad_norm": 1.5823386945798619, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 38904 + }, + { + "epoch": 0.38905, + "grad_norm": 1.472164493986636, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 38905 + }, + { + "epoch": 0.38906, + "grad_norm": 1.50570177624655, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 38906 + }, + { + "epoch": 0.38907, + "grad_norm": 1.997146478633414, + "learning_rate": 0.003, + "loss": 4.051, + "step": 38907 + }, + { + "epoch": 0.38908, + "grad_norm": 1.4275081634689037, + "learning_rate": 0.003, + "loss": 4.011, + "step": 38908 + }, + { + "epoch": 0.38909, + "grad_norm": 1.6612010573734342, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 38909 + }, + { + "epoch": 0.3891, + "grad_norm": 1.483107511706225, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 38910 + }, + { + "epoch": 0.38911, + "grad_norm": 1.4782497111551987, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 38911 + }, + { + "epoch": 0.38912, + "grad_norm": 1.5004174508083832, + "learning_rate": 0.003, + "loss": 4.04, + "step": 38912 + }, + { + "epoch": 0.38913, + "grad_norm": 1.8777557208612115, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 38913 + }, + { + "epoch": 0.38914, + "grad_norm": 1.3618514562214612, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 38914 + }, + { + "epoch": 0.38915, + "grad_norm": 1.5545637373379129, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 38915 + }, + { + "epoch": 0.38916, + "grad_norm": 1.6339395267741228, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 38916 + }, + { + "epoch": 0.38917, + "grad_norm": 1.5088870215880015, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38917 + }, + { + "epoch": 0.38918, + "grad_norm": 1.5696368661231224, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 38918 + }, + { + "epoch": 0.38919, + "grad_norm": 1.4417625067128208, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 38919 + }, + { + "epoch": 0.3892, + "grad_norm": 1.671941940127066, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 38920 + }, + { + "epoch": 0.38921, + "grad_norm": 1.4022058932329229, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 38921 + }, + { + "epoch": 0.38922, + "grad_norm": 1.3880508516743204, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 38922 + }, + { + "epoch": 0.38923, + "grad_norm": 1.2793283181588087, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 38923 + }, + { + "epoch": 0.38924, + "grad_norm": 1.5048512902600863, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 38924 + }, + { + "epoch": 0.38925, + "grad_norm": 1.4722818445557073, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 38925 + }, + { + "epoch": 0.38926, + "grad_norm": 1.675859988530999, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 38926 + }, + { + "epoch": 0.38927, + "grad_norm": 1.43692804722252, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 38927 + }, + { + "epoch": 0.38928, + "grad_norm": 1.7969961537067183, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 38928 + }, + { + "epoch": 0.38929, + "grad_norm": 1.455144446025814, + "learning_rate": 0.003, + "loss": 4.051, + "step": 38929 + }, + { + "epoch": 0.3893, + "grad_norm": 1.8041981618732101, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 38930 + }, + { + "epoch": 0.38931, + "grad_norm": 1.4768678924440932, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 38931 + }, + { + "epoch": 0.38932, + "grad_norm": 1.6401175941218884, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 38932 + }, + { + "epoch": 0.38933, + "grad_norm": 1.5045941423543536, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 38933 + }, + { + "epoch": 0.38934, + "grad_norm": 1.5585855457589775, + "learning_rate": 0.003, + "loss": 4.054, + "step": 38934 + }, + { + "epoch": 0.38935, + "grad_norm": 1.4845426347948403, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 38935 + }, + { + "epoch": 0.38936, + "grad_norm": 1.451350291917564, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 38936 + }, + { + "epoch": 0.38937, + "grad_norm": 1.4040781463704155, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 38937 + }, + { + "epoch": 0.38938, + "grad_norm": 1.7301428279915567, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 38938 + }, + { + "epoch": 0.38939, + "grad_norm": 1.2129292863790349, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 38939 + }, + { + "epoch": 0.3894, + "grad_norm": 1.756205913016322, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 38940 + }, + { + "epoch": 0.38941, + "grad_norm": 1.3040960091849239, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 38941 + }, + { + "epoch": 0.38942, + "grad_norm": 1.8159322006634298, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 38942 + }, + { + "epoch": 0.38943, + "grad_norm": 1.2293878806914607, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 38943 + }, + { + "epoch": 0.38944, + "grad_norm": 2.2315932948128623, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 38944 + }, + { + "epoch": 0.38945, + "grad_norm": 1.8805998762879292, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 38945 + }, + { + "epoch": 0.38946, + "grad_norm": 1.292998738960854, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 38946 + }, + { + "epoch": 0.38947, + "grad_norm": 1.6471678249465351, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38947 + }, + { + "epoch": 0.38948, + "grad_norm": 1.3303449527455389, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 38948 + }, + { + "epoch": 0.38949, + "grad_norm": 1.4416264637079639, + "learning_rate": 0.003, + "loss": 4.066, + "step": 38949 + }, + { + "epoch": 0.3895, + "grad_norm": 1.6101522298878737, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 38950 + }, + { + "epoch": 0.38951, + "grad_norm": 1.3437658107443966, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 38951 + }, + { + "epoch": 0.38952, + "grad_norm": 1.5238999234801611, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 38952 + }, + { + "epoch": 0.38953, + "grad_norm": 1.5194612029081158, + "learning_rate": 0.003, + "loss": 4.013, + "step": 38953 + }, + { + "epoch": 0.38954, + "grad_norm": 1.6690282884352232, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 38954 + }, + { + "epoch": 0.38955, + "grad_norm": 1.309059055110225, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 38955 + }, + { + "epoch": 0.38956, + "grad_norm": 1.9087993107538481, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 38956 + }, + { + "epoch": 0.38957, + "grad_norm": 1.5879512152039725, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 38957 + }, + { + "epoch": 0.38958, + "grad_norm": 1.7676592233513104, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 38958 + }, + { + "epoch": 0.38959, + "grad_norm": 1.4687338477942997, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38959 + }, + { + "epoch": 0.3896, + "grad_norm": 1.2711648490821341, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38960 + }, + { + "epoch": 0.38961, + "grad_norm": 1.4543048378724681, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 38961 + }, + { + "epoch": 0.38962, + "grad_norm": 1.5419516483092333, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 38962 + }, + { + "epoch": 0.38963, + "grad_norm": 1.1523831686990382, + "learning_rate": 0.003, + "loss": 4.024, + "step": 38963 + }, + { + "epoch": 0.38964, + "grad_norm": 1.813186225616357, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 38964 + }, + { + "epoch": 0.38965, + "grad_norm": 1.2954024692220396, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 38965 + }, + { + "epoch": 0.38966, + "grad_norm": 1.6841093009801238, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 38966 + }, + { + "epoch": 0.38967, + "grad_norm": 1.4495904605198011, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38967 + }, + { + "epoch": 0.38968, + "grad_norm": 1.5985606451755368, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 38968 + }, + { + "epoch": 0.38969, + "grad_norm": 1.6053969874522742, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 38969 + }, + { + "epoch": 0.3897, + "grad_norm": 1.45035473524328, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 38970 + }, + { + "epoch": 0.38971, + "grad_norm": 1.6717726265533372, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 38971 + }, + { + "epoch": 0.38972, + "grad_norm": 1.5337064465068022, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 38972 + }, + { + "epoch": 0.38973, + "grad_norm": 1.740717907351314, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 38973 + }, + { + "epoch": 0.38974, + "grad_norm": 1.6681058137976637, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 38974 + }, + { + "epoch": 0.38975, + "grad_norm": 1.4672845008544104, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 38975 + }, + { + "epoch": 0.38976, + "grad_norm": 1.3521793950878607, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 38976 + }, + { + "epoch": 0.38977, + "grad_norm": 1.418589407687543, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 38977 + }, + { + "epoch": 0.38978, + "grad_norm": 1.4838166417994982, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 38978 + }, + { + "epoch": 0.38979, + "grad_norm": 1.6766769986893877, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 38979 + }, + { + "epoch": 0.3898, + "grad_norm": 1.3012400189831548, + "learning_rate": 0.003, + "loss": 4.04, + "step": 38980 + }, + { + "epoch": 0.38981, + "grad_norm": 1.6321203442980716, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 38981 + }, + { + "epoch": 0.38982, + "grad_norm": 1.232080531895431, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 38982 + }, + { + "epoch": 0.38983, + "grad_norm": 1.5906193658109786, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 38983 + }, + { + "epoch": 0.38984, + "grad_norm": 1.4724067288530824, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 38984 + }, + { + "epoch": 0.38985, + "grad_norm": 1.3992692983413988, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 38985 + }, + { + "epoch": 0.38986, + "grad_norm": 1.7394703802898306, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 38986 + }, + { + "epoch": 0.38987, + "grad_norm": 1.4418093428912389, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 38987 + }, + { + "epoch": 0.38988, + "grad_norm": 1.9918866198795706, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38988 + }, + { + "epoch": 0.38989, + "grad_norm": 1.445630052274399, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 38989 + }, + { + "epoch": 0.3899, + "grad_norm": 1.8839786395642264, + "learning_rate": 0.003, + "loss": 3.998, + "step": 38990 + }, + { + "epoch": 0.38991, + "grad_norm": 1.0985232481160276, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 38991 + }, + { + "epoch": 0.38992, + "grad_norm": 2.0340421561428306, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 38992 + }, + { + "epoch": 0.38993, + "grad_norm": 1.4408477517260743, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 38993 + }, + { + "epoch": 0.38994, + "grad_norm": 1.550887974990108, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 38994 + }, + { + "epoch": 0.38995, + "grad_norm": 1.5658500760631933, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 38995 + }, + { + "epoch": 0.38996, + "grad_norm": 1.5293712835482156, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 38996 + }, + { + "epoch": 0.38997, + "grad_norm": 1.5929291344033665, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 38997 + }, + { + "epoch": 0.38998, + "grad_norm": 1.4315591602916447, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 38998 + }, + { + "epoch": 0.38999, + "grad_norm": 1.5240129461943386, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 38999 + }, + { + "epoch": 0.39, + "grad_norm": 1.4131441829880198, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 39000 + }, + { + "epoch": 0.39001, + "grad_norm": 1.5496932625582895, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 39001 + }, + { + "epoch": 0.39002, + "grad_norm": 1.326160931058954, + "learning_rate": 0.003, + "loss": 4.006, + "step": 39002 + }, + { + "epoch": 0.39003, + "grad_norm": 1.8922399864447597, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 39003 + }, + { + "epoch": 0.39004, + "grad_norm": 1.516539942893102, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39004 + }, + { + "epoch": 0.39005, + "grad_norm": 1.6513374749858383, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39005 + }, + { + "epoch": 0.39006, + "grad_norm": 1.820026974964863, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 39006 + }, + { + "epoch": 0.39007, + "grad_norm": 1.4592938732215328, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39007 + }, + { + "epoch": 0.39008, + "grad_norm": 1.488762322397853, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39008 + }, + { + "epoch": 0.39009, + "grad_norm": 1.4685129222416475, + "learning_rate": 0.003, + "loss": 4.031, + "step": 39009 + }, + { + "epoch": 0.3901, + "grad_norm": 1.4417391568154552, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39010 + }, + { + "epoch": 0.39011, + "grad_norm": 1.4005916642327423, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 39011 + }, + { + "epoch": 0.39012, + "grad_norm": 1.4369389933581402, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 39012 + }, + { + "epoch": 0.39013, + "grad_norm": 1.46363049466004, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 39013 + }, + { + "epoch": 0.39014, + "grad_norm": 1.2738655057070885, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39014 + }, + { + "epoch": 0.39015, + "grad_norm": 1.936308955371726, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 39015 + }, + { + "epoch": 0.39016, + "grad_norm": 1.2204703326599766, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 39016 + }, + { + "epoch": 0.39017, + "grad_norm": 1.7010195758025295, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 39017 + }, + { + "epoch": 0.39018, + "grad_norm": 1.3380117397972906, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 39018 + }, + { + "epoch": 0.39019, + "grad_norm": 1.5284138989549247, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39019 + }, + { + "epoch": 0.3902, + "grad_norm": 1.660123433369315, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39020 + }, + { + "epoch": 0.39021, + "grad_norm": 1.6480648262267505, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 39021 + }, + { + "epoch": 0.39022, + "grad_norm": 1.8810124472488141, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 39022 + }, + { + "epoch": 0.39023, + "grad_norm": 1.5065883231457553, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 39023 + }, + { + "epoch": 0.39024, + "grad_norm": 1.2909412037814032, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 39024 + }, + { + "epoch": 0.39025, + "grad_norm": 1.5692316046673702, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39025 + }, + { + "epoch": 0.39026, + "grad_norm": 1.5100197840245824, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 39026 + }, + { + "epoch": 0.39027, + "grad_norm": 1.229511445836706, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 39027 + }, + { + "epoch": 0.39028, + "grad_norm": 1.6428159328451715, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 39028 + }, + { + "epoch": 0.39029, + "grad_norm": 1.550471373848979, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 39029 + }, + { + "epoch": 0.3903, + "grad_norm": 1.8002593305062002, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 39030 + }, + { + "epoch": 0.39031, + "grad_norm": 1.4908026595888055, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39031 + }, + { + "epoch": 0.39032, + "grad_norm": 1.4381751775995455, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 39032 + }, + { + "epoch": 0.39033, + "grad_norm": 1.718461693500414, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39033 + }, + { + "epoch": 0.39034, + "grad_norm": 1.220964115398869, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39034 + }, + { + "epoch": 0.39035, + "grad_norm": 1.7104026917980173, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 39035 + }, + { + "epoch": 0.39036, + "grad_norm": 1.2464887318695013, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39036 + }, + { + "epoch": 0.39037, + "grad_norm": 1.7305776384956084, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 39037 + }, + { + "epoch": 0.39038, + "grad_norm": 1.5314716044275922, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 39038 + }, + { + "epoch": 0.39039, + "grad_norm": 1.6043618829898438, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 39039 + }, + { + "epoch": 0.3904, + "grad_norm": 1.811620945921955, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 39040 + }, + { + "epoch": 0.39041, + "grad_norm": 1.302836689439044, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 39041 + }, + { + "epoch": 0.39042, + "grad_norm": 1.7804379901746572, + "learning_rate": 0.003, + "loss": 4.053, + "step": 39042 + }, + { + "epoch": 0.39043, + "grad_norm": 1.7734641792500268, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 39043 + }, + { + "epoch": 0.39044, + "grad_norm": 1.238008796641479, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 39044 + }, + { + "epoch": 0.39045, + "grad_norm": 1.6293259780036708, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 39045 + }, + { + "epoch": 0.39046, + "grad_norm": 1.367872319245828, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 39046 + }, + { + "epoch": 0.39047, + "grad_norm": 1.6446238888337372, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 39047 + }, + { + "epoch": 0.39048, + "grad_norm": 1.1826379463751056, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 39048 + }, + { + "epoch": 0.39049, + "grad_norm": 1.6352340925399547, + "learning_rate": 0.003, + "loss": 4.031, + "step": 39049 + }, + { + "epoch": 0.3905, + "grad_norm": 1.8262695911690814, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39050 + }, + { + "epoch": 0.39051, + "grad_norm": 1.4819174155668313, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 39051 + }, + { + "epoch": 0.39052, + "grad_norm": 1.54029070944047, + "learning_rate": 0.003, + "loss": 4.057, + "step": 39052 + }, + { + "epoch": 0.39053, + "grad_norm": 1.6636046763381704, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 39053 + }, + { + "epoch": 0.39054, + "grad_norm": 1.4248798428210314, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39054 + }, + { + "epoch": 0.39055, + "grad_norm": 1.5549242647751549, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 39055 + }, + { + "epoch": 0.39056, + "grad_norm": 1.4236092711378188, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 39056 + }, + { + "epoch": 0.39057, + "grad_norm": 1.5044360651087236, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 39057 + }, + { + "epoch": 0.39058, + "grad_norm": 1.377155053828152, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 39058 + }, + { + "epoch": 0.39059, + "grad_norm": 1.428501970656722, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 39059 + }, + { + "epoch": 0.3906, + "grad_norm": 1.5969644083557548, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 39060 + }, + { + "epoch": 0.39061, + "grad_norm": 1.458236966635893, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 39061 + }, + { + "epoch": 0.39062, + "grad_norm": 1.4471666509779477, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 39062 + }, + { + "epoch": 0.39063, + "grad_norm": 1.335364135376702, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 39063 + }, + { + "epoch": 0.39064, + "grad_norm": 1.4526386337582367, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 39064 + }, + { + "epoch": 0.39065, + "grad_norm": 1.5333241975238285, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 39065 + }, + { + "epoch": 0.39066, + "grad_norm": 1.940003690834732, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 39066 + }, + { + "epoch": 0.39067, + "grad_norm": 1.84488398579094, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 39067 + }, + { + "epoch": 0.39068, + "grad_norm": 1.3072454609451916, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 39068 + }, + { + "epoch": 0.39069, + "grad_norm": 1.7571853023282926, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 39069 + }, + { + "epoch": 0.3907, + "grad_norm": 1.3447767222349174, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 39070 + }, + { + "epoch": 0.39071, + "grad_norm": 1.745306172204631, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 39071 + }, + { + "epoch": 0.39072, + "grad_norm": 1.3824237527696726, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 39072 + }, + { + "epoch": 0.39073, + "grad_norm": 1.5852936318405637, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39073 + }, + { + "epoch": 0.39074, + "grad_norm": 1.4389809246329541, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39074 + }, + { + "epoch": 0.39075, + "grad_norm": 1.3221773890487392, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 39075 + }, + { + "epoch": 0.39076, + "grad_norm": 1.4761422471541092, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 39076 + }, + { + "epoch": 0.39077, + "grad_norm": 1.533868850138156, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 39077 + }, + { + "epoch": 0.39078, + "grad_norm": 1.441773603927473, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 39078 + }, + { + "epoch": 0.39079, + "grad_norm": 1.528706001342156, + "learning_rate": 0.003, + "loss": 4.033, + "step": 39079 + }, + { + "epoch": 0.3908, + "grad_norm": 1.484168428283128, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39080 + }, + { + "epoch": 0.39081, + "grad_norm": 1.6507958406955558, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 39081 + }, + { + "epoch": 0.39082, + "grad_norm": 1.7170545512209576, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 39082 + }, + { + "epoch": 0.39083, + "grad_norm": 1.8490528021571322, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 39083 + }, + { + "epoch": 0.39084, + "grad_norm": 1.3213654589950992, + "learning_rate": 0.003, + "loss": 4.057, + "step": 39084 + }, + { + "epoch": 0.39085, + "grad_norm": 1.7315857143494862, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 39085 + }, + { + "epoch": 0.39086, + "grad_norm": 1.6447343183133258, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 39086 + }, + { + "epoch": 0.39087, + "grad_norm": 1.3054166468165518, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 39087 + }, + { + "epoch": 0.39088, + "grad_norm": 1.8245531980521899, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 39088 + }, + { + "epoch": 0.39089, + "grad_norm": 1.353870819648898, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 39089 + }, + { + "epoch": 0.3909, + "grad_norm": 1.6483728334942218, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39090 + }, + { + "epoch": 0.39091, + "grad_norm": 1.5335706369051736, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 39091 + }, + { + "epoch": 0.39092, + "grad_norm": 1.324861869656761, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 39092 + }, + { + "epoch": 0.39093, + "grad_norm": 1.3338204970713294, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39093 + }, + { + "epoch": 0.39094, + "grad_norm": 1.6267270704868655, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 39094 + }, + { + "epoch": 0.39095, + "grad_norm": 1.5369842965424827, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 39095 + }, + { + "epoch": 0.39096, + "grad_norm": 1.563440648466432, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 39096 + }, + { + "epoch": 0.39097, + "grad_norm": 1.359348243993864, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 39097 + }, + { + "epoch": 0.39098, + "grad_norm": 1.569244260061089, + "learning_rate": 0.003, + "loss": 4.061, + "step": 39098 + }, + { + "epoch": 0.39099, + "grad_norm": 1.4117349357221824, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 39099 + }, + { + "epoch": 0.391, + "grad_norm": 1.6314172880297233, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 39100 + }, + { + "epoch": 0.39101, + "grad_norm": 1.7208749428557373, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 39101 + }, + { + "epoch": 0.39102, + "grad_norm": 1.5296524267147797, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 39102 + }, + { + "epoch": 0.39103, + "grad_norm": 1.8082379542225004, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 39103 + }, + { + "epoch": 0.39104, + "grad_norm": 1.6248558030892593, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39104 + }, + { + "epoch": 0.39105, + "grad_norm": 1.4391684686331032, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 39105 + }, + { + "epoch": 0.39106, + "grad_norm": 1.6935318470647114, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 39106 + }, + { + "epoch": 0.39107, + "grad_norm": 1.1881925022426196, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 39107 + }, + { + "epoch": 0.39108, + "grad_norm": 1.5524438472442343, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 39108 + }, + { + "epoch": 0.39109, + "grad_norm": 1.7149098892768002, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 39109 + }, + { + "epoch": 0.3911, + "grad_norm": 1.5667403129174573, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 39110 + }, + { + "epoch": 0.39111, + "grad_norm": 1.3914364779988526, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 39111 + }, + { + "epoch": 0.39112, + "grad_norm": 2.0554875812067617, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39112 + }, + { + "epoch": 0.39113, + "grad_norm": 1.1256107884956028, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39113 + }, + { + "epoch": 0.39114, + "grad_norm": 1.669333993276301, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 39114 + }, + { + "epoch": 0.39115, + "grad_norm": 1.5990611908076304, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 39115 + }, + { + "epoch": 0.39116, + "grad_norm": 1.4887439035349284, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 39116 + }, + { + "epoch": 0.39117, + "grad_norm": 1.5320242347469581, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 39117 + }, + { + "epoch": 0.39118, + "grad_norm": 1.714388576494124, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 39118 + }, + { + "epoch": 0.39119, + "grad_norm": 1.4936637150446512, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 39119 + }, + { + "epoch": 0.3912, + "grad_norm": 1.41856992491173, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39120 + }, + { + "epoch": 0.39121, + "grad_norm": 1.4326871999245498, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39121 + }, + { + "epoch": 0.39122, + "grad_norm": 1.4785436053328116, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39122 + }, + { + "epoch": 0.39123, + "grad_norm": 1.2852115265087272, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39123 + }, + { + "epoch": 0.39124, + "grad_norm": 1.7155385559298175, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 39124 + }, + { + "epoch": 0.39125, + "grad_norm": 1.5308242645368728, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 39125 + }, + { + "epoch": 0.39126, + "grad_norm": 1.41273266611867, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 39126 + }, + { + "epoch": 0.39127, + "grad_norm": 1.791857953753863, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 39127 + }, + { + "epoch": 0.39128, + "grad_norm": 1.374664179012995, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 39128 + }, + { + "epoch": 0.39129, + "grad_norm": 1.9196944466535268, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 39129 + }, + { + "epoch": 0.3913, + "grad_norm": 1.3303357332166592, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 39130 + }, + { + "epoch": 0.39131, + "grad_norm": 1.3463361880989793, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 39131 + }, + { + "epoch": 0.39132, + "grad_norm": 1.6260473630764622, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39132 + }, + { + "epoch": 0.39133, + "grad_norm": 1.4033268556208194, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 39133 + }, + { + "epoch": 0.39134, + "grad_norm": 1.5316067951513719, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 39134 + }, + { + "epoch": 0.39135, + "grad_norm": 1.2343075044613114, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 39135 + }, + { + "epoch": 0.39136, + "grad_norm": 1.9389287230503511, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 39136 + }, + { + "epoch": 0.39137, + "grad_norm": 1.6724242388712198, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39137 + }, + { + "epoch": 0.39138, + "grad_norm": 1.5182333861278516, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39138 + }, + { + "epoch": 0.39139, + "grad_norm": 1.3735819664900082, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 39139 + }, + { + "epoch": 0.3914, + "grad_norm": 1.5671301107944238, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 39140 + }, + { + "epoch": 0.39141, + "grad_norm": 1.5298428014860392, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 39141 + }, + { + "epoch": 0.39142, + "grad_norm": 1.245025641785305, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 39142 + }, + { + "epoch": 0.39143, + "grad_norm": 1.5649899565563077, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 39143 + }, + { + "epoch": 0.39144, + "grad_norm": 1.373570701728577, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 39144 + }, + { + "epoch": 0.39145, + "grad_norm": 1.994724941735637, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 39145 + }, + { + "epoch": 0.39146, + "grad_norm": 1.2326333509882506, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 39146 + }, + { + "epoch": 0.39147, + "grad_norm": 1.700384709072263, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 39147 + }, + { + "epoch": 0.39148, + "grad_norm": 1.8182937782215134, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 39148 + }, + { + "epoch": 0.39149, + "grad_norm": 1.2285160001388673, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39149 + }, + { + "epoch": 0.3915, + "grad_norm": 1.6859664271805177, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 39150 + }, + { + "epoch": 0.39151, + "grad_norm": 1.6364865884397433, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 39151 + }, + { + "epoch": 0.39152, + "grad_norm": 1.772343933304195, + "learning_rate": 0.003, + "loss": 4.054, + "step": 39152 + }, + { + "epoch": 0.39153, + "grad_norm": 1.3833321873178284, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 39153 + }, + { + "epoch": 0.39154, + "grad_norm": 1.5385623463630635, + "learning_rate": 0.003, + "loss": 4.077, + "step": 39154 + }, + { + "epoch": 0.39155, + "grad_norm": 1.4500547531695562, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 39155 + }, + { + "epoch": 0.39156, + "grad_norm": 1.5408413000221988, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 39156 + }, + { + "epoch": 0.39157, + "grad_norm": 1.2369803192601452, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39157 + }, + { + "epoch": 0.39158, + "grad_norm": 1.5502725635520649, + "learning_rate": 0.003, + "loss": 4.045, + "step": 39158 + }, + { + "epoch": 0.39159, + "grad_norm": 1.5267829508999895, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 39159 + }, + { + "epoch": 0.3916, + "grad_norm": 1.548781969066665, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39160 + }, + { + "epoch": 0.39161, + "grad_norm": 1.8038684734564197, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39161 + }, + { + "epoch": 0.39162, + "grad_norm": 1.5084302570946173, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 39162 + }, + { + "epoch": 0.39163, + "grad_norm": 1.7721084927450095, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39163 + }, + { + "epoch": 0.39164, + "grad_norm": 1.7233206646404304, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 39164 + }, + { + "epoch": 0.39165, + "grad_norm": 1.5785072062624717, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 39165 + }, + { + "epoch": 0.39166, + "grad_norm": 1.4309171376749612, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 39166 + }, + { + "epoch": 0.39167, + "grad_norm": 1.564582357575792, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 39167 + }, + { + "epoch": 0.39168, + "grad_norm": 1.4614532763368138, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 39168 + }, + { + "epoch": 0.39169, + "grad_norm": 1.4785117110179449, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 39169 + }, + { + "epoch": 0.3917, + "grad_norm": 1.3158767352486171, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39170 + }, + { + "epoch": 0.39171, + "grad_norm": 1.6931979718149242, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 39171 + }, + { + "epoch": 0.39172, + "grad_norm": 1.3661679975414556, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 39172 + }, + { + "epoch": 0.39173, + "grad_norm": 1.5940517935146077, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39173 + }, + { + "epoch": 0.39174, + "grad_norm": 1.5336865101119117, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39174 + }, + { + "epoch": 0.39175, + "grad_norm": 1.6782358551741257, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 39175 + }, + { + "epoch": 0.39176, + "grad_norm": 1.4533731276564332, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39176 + }, + { + "epoch": 0.39177, + "grad_norm": 1.6904579269324769, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 39177 + }, + { + "epoch": 0.39178, + "grad_norm": 1.4251023044907944, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 39178 + }, + { + "epoch": 0.39179, + "grad_norm": 1.7070068093954975, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 39179 + }, + { + "epoch": 0.3918, + "grad_norm": 1.3850167840196341, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39180 + }, + { + "epoch": 0.39181, + "grad_norm": 1.679056629970491, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 39181 + }, + { + "epoch": 0.39182, + "grad_norm": 1.232062483877146, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 39182 + }, + { + "epoch": 0.39183, + "grad_norm": 1.9178506616074522, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39183 + }, + { + "epoch": 0.39184, + "grad_norm": 1.2484781877290954, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39184 + }, + { + "epoch": 0.39185, + "grad_norm": 1.6932771396521764, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 39185 + }, + { + "epoch": 0.39186, + "grad_norm": 1.338127027346343, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39186 + }, + { + "epoch": 0.39187, + "grad_norm": 1.8028960687834952, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39187 + }, + { + "epoch": 0.39188, + "grad_norm": 1.6779997715326902, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 39188 + }, + { + "epoch": 0.39189, + "grad_norm": 1.8031731771665334, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39189 + }, + { + "epoch": 0.3919, + "grad_norm": 1.4750093487065017, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 39190 + }, + { + "epoch": 0.39191, + "grad_norm": 1.5457557255951626, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 39191 + }, + { + "epoch": 0.39192, + "grad_norm": 1.3723810416613322, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39192 + }, + { + "epoch": 0.39193, + "grad_norm": 1.765910825521709, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 39193 + }, + { + "epoch": 0.39194, + "grad_norm": 1.306455994436169, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 39194 + }, + { + "epoch": 0.39195, + "grad_norm": 1.874698803381305, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 39195 + }, + { + "epoch": 0.39196, + "grad_norm": 1.618136881790597, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39196 + }, + { + "epoch": 0.39197, + "grad_norm": 1.2432586646731891, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39197 + }, + { + "epoch": 0.39198, + "grad_norm": 1.3126907368383742, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39198 + }, + { + "epoch": 0.39199, + "grad_norm": 1.8709450186235834, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 39199 + }, + { + "epoch": 0.392, + "grad_norm": 1.3307211436955193, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39200 + }, + { + "epoch": 0.39201, + "grad_norm": 1.6345604419867918, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39201 + }, + { + "epoch": 0.39202, + "grad_norm": 1.5659188740057512, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 39202 + }, + { + "epoch": 0.39203, + "grad_norm": 1.3014962640773529, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39203 + }, + { + "epoch": 0.39204, + "grad_norm": 1.5245255875775725, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 39204 + }, + { + "epoch": 0.39205, + "grad_norm": 1.641203375278123, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 39205 + }, + { + "epoch": 0.39206, + "grad_norm": 1.0850335117569516, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 39206 + }, + { + "epoch": 0.39207, + "grad_norm": 1.947502286864688, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 39207 + }, + { + "epoch": 0.39208, + "grad_norm": 1.2614032684413778, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39208 + }, + { + "epoch": 0.39209, + "grad_norm": 1.6629034802476523, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 39209 + }, + { + "epoch": 0.3921, + "grad_norm": 1.6116604981419347, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39210 + }, + { + "epoch": 0.39211, + "grad_norm": 1.7371238803820017, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 39211 + }, + { + "epoch": 0.39212, + "grad_norm": 1.6866757431946973, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 39212 + }, + { + "epoch": 0.39213, + "grad_norm": 1.3482992357915338, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 39213 + }, + { + "epoch": 0.39214, + "grad_norm": 1.4716849251062236, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 39214 + }, + { + "epoch": 0.39215, + "grad_norm": 1.6162519804592124, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 39215 + }, + { + "epoch": 0.39216, + "grad_norm": 1.3442836090757173, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39216 + }, + { + "epoch": 0.39217, + "grad_norm": 1.4609113864678374, + "learning_rate": 0.003, + "loss": 4.025, + "step": 39217 + }, + { + "epoch": 0.39218, + "grad_norm": 1.6014037735814362, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 39218 + }, + { + "epoch": 0.39219, + "grad_norm": 1.6328060396981452, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 39219 + }, + { + "epoch": 0.3922, + "grad_norm": 1.6038009268135769, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 39220 + }, + { + "epoch": 0.39221, + "grad_norm": 1.277362780683777, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 39221 + }, + { + "epoch": 0.39222, + "grad_norm": 1.5904976644760558, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 39222 + }, + { + "epoch": 0.39223, + "grad_norm": 1.313722373241692, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 39223 + }, + { + "epoch": 0.39224, + "grad_norm": 1.6580438854601622, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 39224 + }, + { + "epoch": 0.39225, + "grad_norm": 1.380233076269285, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 39225 + }, + { + "epoch": 0.39226, + "grad_norm": 1.669333924331191, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 39226 + }, + { + "epoch": 0.39227, + "grad_norm": 1.5957098987921443, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 39227 + }, + { + "epoch": 0.39228, + "grad_norm": 1.4015514299550962, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 39228 + }, + { + "epoch": 0.39229, + "grad_norm": 1.3939461728287268, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 39229 + }, + { + "epoch": 0.3923, + "grad_norm": 1.5061438546674002, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39230 + }, + { + "epoch": 0.39231, + "grad_norm": 1.3425004227569022, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39231 + }, + { + "epoch": 0.39232, + "grad_norm": 1.7141526079507468, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 39232 + }, + { + "epoch": 0.39233, + "grad_norm": 1.4166035796350616, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 39233 + }, + { + "epoch": 0.39234, + "grad_norm": 1.6807004114713597, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 39234 + }, + { + "epoch": 0.39235, + "grad_norm": 1.4418174258977943, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39235 + }, + { + "epoch": 0.39236, + "grad_norm": 1.6019419165474016, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 39236 + }, + { + "epoch": 0.39237, + "grad_norm": 1.4739123143031672, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 39237 + }, + { + "epoch": 0.39238, + "grad_norm": 1.6185625408442776, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39238 + }, + { + "epoch": 0.39239, + "grad_norm": 1.4695326492493446, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 39239 + }, + { + "epoch": 0.3924, + "grad_norm": 1.4428145877608112, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 39240 + }, + { + "epoch": 0.39241, + "grad_norm": 1.6658189725062846, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 39241 + }, + { + "epoch": 0.39242, + "grad_norm": 1.5509969416624387, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 39242 + }, + { + "epoch": 0.39243, + "grad_norm": 1.6049781931543132, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 39243 + }, + { + "epoch": 0.39244, + "grad_norm": 1.512271617289721, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 39244 + }, + { + "epoch": 0.39245, + "grad_norm": 1.4745481123728252, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 39245 + }, + { + "epoch": 0.39246, + "grad_norm": 1.5381300981380082, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39246 + }, + { + "epoch": 0.39247, + "grad_norm": 1.401069284105515, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 39247 + }, + { + "epoch": 0.39248, + "grad_norm": 1.8406462930361958, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 39248 + }, + { + "epoch": 0.39249, + "grad_norm": 1.391714197182955, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39249 + }, + { + "epoch": 0.3925, + "grad_norm": 1.7630025626866357, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39250 + }, + { + "epoch": 0.39251, + "grad_norm": 1.4288002295322892, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 39251 + }, + { + "epoch": 0.39252, + "grad_norm": 1.441488487961739, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 39252 + }, + { + "epoch": 0.39253, + "grad_norm": 1.3635659887304847, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 39253 + }, + { + "epoch": 0.39254, + "grad_norm": 1.7997856248490833, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 39254 + }, + { + "epoch": 0.39255, + "grad_norm": 1.3588371946856632, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39255 + }, + { + "epoch": 0.39256, + "grad_norm": 1.5648799707332837, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 39256 + }, + { + "epoch": 0.39257, + "grad_norm": 1.572529773733751, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 39257 + }, + { + "epoch": 0.39258, + "grad_norm": 1.6372005050762355, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 39258 + }, + { + "epoch": 0.39259, + "grad_norm": 1.2853198720995604, + "learning_rate": 0.003, + "loss": 4.056, + "step": 39259 + }, + { + "epoch": 0.3926, + "grad_norm": 1.609771286065536, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 39260 + }, + { + "epoch": 0.39261, + "grad_norm": 1.629195598417752, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 39261 + }, + { + "epoch": 0.39262, + "grad_norm": 1.413007792826098, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39262 + }, + { + "epoch": 0.39263, + "grad_norm": 1.5589084130658022, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 39263 + }, + { + "epoch": 0.39264, + "grad_norm": 1.6823701432929559, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 39264 + }, + { + "epoch": 0.39265, + "grad_norm": 1.5701961541982705, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 39265 + }, + { + "epoch": 0.39266, + "grad_norm": 1.3641641499735357, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 39266 + }, + { + "epoch": 0.39267, + "grad_norm": 1.5209973308906428, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 39267 + }, + { + "epoch": 0.39268, + "grad_norm": 1.7736069570635753, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 39268 + }, + { + "epoch": 0.39269, + "grad_norm": 1.3838845107334832, + "learning_rate": 0.003, + "loss": 4.025, + "step": 39269 + }, + { + "epoch": 0.3927, + "grad_norm": 1.8249445540443943, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 39270 + }, + { + "epoch": 0.39271, + "grad_norm": 1.364334332497821, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39271 + }, + { + "epoch": 0.39272, + "grad_norm": 1.756745446431008, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 39272 + }, + { + "epoch": 0.39273, + "grad_norm": 1.4153417116517633, + "learning_rate": 0.003, + "loss": 4.059, + "step": 39273 + }, + { + "epoch": 0.39274, + "grad_norm": 1.6030424329024997, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39274 + }, + { + "epoch": 0.39275, + "grad_norm": 1.4341345051028738, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 39275 + }, + { + "epoch": 0.39276, + "grad_norm": 1.5753318168758859, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 39276 + }, + { + "epoch": 0.39277, + "grad_norm": 1.8871738969472263, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39277 + }, + { + "epoch": 0.39278, + "grad_norm": 1.3084249726889896, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 39278 + }, + { + "epoch": 0.39279, + "grad_norm": 1.6206016020396565, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 39279 + }, + { + "epoch": 0.3928, + "grad_norm": 1.653983523337921, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 39280 + }, + { + "epoch": 0.39281, + "grad_norm": 1.5858729124375457, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 39281 + }, + { + "epoch": 0.39282, + "grad_norm": 1.4429328512351163, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39282 + }, + { + "epoch": 0.39283, + "grad_norm": 1.5669620624145468, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39283 + }, + { + "epoch": 0.39284, + "grad_norm": 1.9305611984142284, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 39284 + }, + { + "epoch": 0.39285, + "grad_norm": 1.4513055812697615, + "learning_rate": 0.003, + "loss": 4.021, + "step": 39285 + }, + { + "epoch": 0.39286, + "grad_norm": 1.4681898025056295, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 39286 + }, + { + "epoch": 0.39287, + "grad_norm": 1.6038460769961271, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 39287 + }, + { + "epoch": 0.39288, + "grad_norm": 1.3176872288206323, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 39288 + }, + { + "epoch": 0.39289, + "grad_norm": 1.59222883719982, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 39289 + }, + { + "epoch": 0.3929, + "grad_norm": 1.5022299026025219, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39290 + }, + { + "epoch": 0.39291, + "grad_norm": 1.6970596345476203, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 39291 + }, + { + "epoch": 0.39292, + "grad_norm": 1.112716992568877, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 39292 + }, + { + "epoch": 0.39293, + "grad_norm": 1.5486834571593218, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 39293 + }, + { + "epoch": 0.39294, + "grad_norm": 1.3933347890687424, + "learning_rate": 0.003, + "loss": 4.053, + "step": 39294 + }, + { + "epoch": 0.39295, + "grad_norm": 1.6670708349758532, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39295 + }, + { + "epoch": 0.39296, + "grad_norm": 1.3410467478639283, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 39296 + }, + { + "epoch": 0.39297, + "grad_norm": 1.812966870475585, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39297 + }, + { + "epoch": 0.39298, + "grad_norm": 1.3334630869594721, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 39298 + }, + { + "epoch": 0.39299, + "grad_norm": 1.412863710986945, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 39299 + }, + { + "epoch": 0.393, + "grad_norm": 1.5157882896269739, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 39300 + }, + { + "epoch": 0.39301, + "grad_norm": 1.4823275871522836, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39301 + }, + { + "epoch": 0.39302, + "grad_norm": 1.5119739795232592, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39302 + }, + { + "epoch": 0.39303, + "grad_norm": 1.5831905155628032, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 39303 + }, + { + "epoch": 0.39304, + "grad_norm": 1.3482505479738083, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 39304 + }, + { + "epoch": 0.39305, + "grad_norm": 1.7904879073462552, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 39305 + }, + { + "epoch": 0.39306, + "grad_norm": 1.7388218975283563, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 39306 + }, + { + "epoch": 0.39307, + "grad_norm": 1.4387645713538204, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 39307 + }, + { + "epoch": 0.39308, + "grad_norm": 1.682654717194914, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 39308 + }, + { + "epoch": 0.39309, + "grad_norm": 1.6659504101892786, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 39309 + }, + { + "epoch": 0.3931, + "grad_norm": 1.5362728071349812, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 39310 + }, + { + "epoch": 0.39311, + "grad_norm": 1.6262864298509248, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39311 + }, + { + "epoch": 0.39312, + "grad_norm": 1.3594128051916428, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 39312 + }, + { + "epoch": 0.39313, + "grad_norm": 1.5109750680156915, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 39313 + }, + { + "epoch": 0.39314, + "grad_norm": 1.2754646762471933, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 39314 + }, + { + "epoch": 0.39315, + "grad_norm": 1.9435772158194025, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 39315 + }, + { + "epoch": 0.39316, + "grad_norm": 1.2619301626578168, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39316 + }, + { + "epoch": 0.39317, + "grad_norm": 1.6910805006524166, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 39317 + }, + { + "epoch": 0.39318, + "grad_norm": 1.605391739400307, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39318 + }, + { + "epoch": 0.39319, + "grad_norm": 1.5513323517701747, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39319 + }, + { + "epoch": 0.3932, + "grad_norm": 1.6247035975172366, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 39320 + }, + { + "epoch": 0.39321, + "grad_norm": 1.6758866697002681, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 39321 + }, + { + "epoch": 0.39322, + "grad_norm": 1.3395296092647102, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 39322 + }, + { + "epoch": 0.39323, + "grad_norm": 1.8427642804681552, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 39323 + }, + { + "epoch": 0.39324, + "grad_norm": 1.2169052795900572, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 39324 + }, + { + "epoch": 0.39325, + "grad_norm": 1.700696977917184, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 39325 + }, + { + "epoch": 0.39326, + "grad_norm": 1.2495660154789074, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 39326 + }, + { + "epoch": 0.39327, + "grad_norm": 1.5728871335131218, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 39327 + }, + { + "epoch": 0.39328, + "grad_norm": 1.558234511018839, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 39328 + }, + { + "epoch": 0.39329, + "grad_norm": 1.7006533023754207, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 39329 + }, + { + "epoch": 0.3933, + "grad_norm": 1.475825201354357, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 39330 + }, + { + "epoch": 0.39331, + "grad_norm": 1.7477545738584783, + "learning_rate": 0.003, + "loss": 4.04, + "step": 39331 + }, + { + "epoch": 0.39332, + "grad_norm": 1.5090394774021918, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 39332 + }, + { + "epoch": 0.39333, + "grad_norm": 1.4043138737748235, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 39333 + }, + { + "epoch": 0.39334, + "grad_norm": 1.6887679534320668, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 39334 + }, + { + "epoch": 0.39335, + "grad_norm": 1.407544048958776, + "learning_rate": 0.003, + "loss": 4.055, + "step": 39335 + }, + { + "epoch": 0.39336, + "grad_norm": 1.6219034905534304, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39336 + }, + { + "epoch": 0.39337, + "grad_norm": 1.4396657825489865, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39337 + }, + { + "epoch": 0.39338, + "grad_norm": 1.2577493870674137, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 39338 + }, + { + "epoch": 0.39339, + "grad_norm": 1.6137177168472088, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 39339 + }, + { + "epoch": 0.3934, + "grad_norm": 1.4819457368985989, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39340 + }, + { + "epoch": 0.39341, + "grad_norm": 1.4889512601062653, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 39341 + }, + { + "epoch": 0.39342, + "grad_norm": 1.500985132803061, + "learning_rate": 0.003, + "loss": 4.016, + "step": 39342 + }, + { + "epoch": 0.39343, + "grad_norm": 1.6223310109371796, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 39343 + }, + { + "epoch": 0.39344, + "grad_norm": 1.2812938113590773, + "learning_rate": 0.003, + "loss": 4.048, + "step": 39344 + }, + { + "epoch": 0.39345, + "grad_norm": 1.6873037398302784, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39345 + }, + { + "epoch": 0.39346, + "grad_norm": 1.3168897149503374, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 39346 + }, + { + "epoch": 0.39347, + "grad_norm": 1.9030203471920075, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 39347 + }, + { + "epoch": 0.39348, + "grad_norm": 1.122508672674065, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 39348 + }, + { + "epoch": 0.39349, + "grad_norm": 1.8839461465426561, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39349 + }, + { + "epoch": 0.3935, + "grad_norm": 1.4883762014741613, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 39350 + }, + { + "epoch": 0.39351, + "grad_norm": 2.1756300392763617, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 39351 + }, + { + "epoch": 0.39352, + "grad_norm": 1.2769857986343418, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 39352 + }, + { + "epoch": 0.39353, + "grad_norm": 1.8611672726869757, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 39353 + }, + { + "epoch": 0.39354, + "grad_norm": 1.6785783888141608, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39354 + }, + { + "epoch": 0.39355, + "grad_norm": 1.6300505504642724, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 39355 + }, + { + "epoch": 0.39356, + "grad_norm": 1.754388753655807, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 39356 + }, + { + "epoch": 0.39357, + "grad_norm": 1.3104776833493825, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39357 + }, + { + "epoch": 0.39358, + "grad_norm": 1.464753200658128, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 39358 + }, + { + "epoch": 0.39359, + "grad_norm": 1.6444616520400592, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 39359 + }, + { + "epoch": 0.3936, + "grad_norm": 1.2409419741757393, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39360 + }, + { + "epoch": 0.39361, + "grad_norm": 1.5461126482624403, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 39361 + }, + { + "epoch": 0.39362, + "grad_norm": 1.3650516623711872, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 39362 + }, + { + "epoch": 0.39363, + "grad_norm": 1.5657673716918794, + "learning_rate": 0.003, + "loss": 4.031, + "step": 39363 + }, + { + "epoch": 0.39364, + "grad_norm": 1.4503127156768516, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 39364 + }, + { + "epoch": 0.39365, + "grad_norm": 1.7561743561007037, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 39365 + }, + { + "epoch": 0.39366, + "grad_norm": 1.1810285138972187, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39366 + }, + { + "epoch": 0.39367, + "grad_norm": 1.7445556453128994, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 39367 + }, + { + "epoch": 0.39368, + "grad_norm": 1.3561553334723182, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 39368 + }, + { + "epoch": 0.39369, + "grad_norm": 1.83557715480671, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 39369 + }, + { + "epoch": 0.3937, + "grad_norm": 1.3418062724355448, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 39370 + }, + { + "epoch": 0.39371, + "grad_norm": 1.5674399940319617, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 39371 + }, + { + "epoch": 0.39372, + "grad_norm": 1.3928203421963838, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 39372 + }, + { + "epoch": 0.39373, + "grad_norm": 1.5126048271534753, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39373 + }, + { + "epoch": 0.39374, + "grad_norm": 1.4453867422941673, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 39374 + }, + { + "epoch": 0.39375, + "grad_norm": 1.2823624202294823, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 39375 + }, + { + "epoch": 0.39376, + "grad_norm": 1.7721611713873948, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 39376 + }, + { + "epoch": 0.39377, + "grad_norm": 1.6943581324756805, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 39377 + }, + { + "epoch": 0.39378, + "grad_norm": 1.9473711327337178, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 39378 + }, + { + "epoch": 0.39379, + "grad_norm": 1.3836483667015333, + "learning_rate": 0.003, + "loss": 4.02, + "step": 39379 + }, + { + "epoch": 0.3938, + "grad_norm": 1.8593473828665164, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39380 + }, + { + "epoch": 0.39381, + "grad_norm": 1.0407054862758875, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 39381 + }, + { + "epoch": 0.39382, + "grad_norm": 1.7261782460968098, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39382 + }, + { + "epoch": 0.39383, + "grad_norm": 1.3690864621537902, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 39383 + }, + { + "epoch": 0.39384, + "grad_norm": 1.7921867873525492, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 39384 + }, + { + "epoch": 0.39385, + "grad_norm": 1.4777774385156788, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39385 + }, + { + "epoch": 0.39386, + "grad_norm": 1.8162606686947724, + "learning_rate": 0.003, + "loss": 4.063, + "step": 39386 + }, + { + "epoch": 0.39387, + "grad_norm": 1.2989694006167407, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 39387 + }, + { + "epoch": 0.39388, + "grad_norm": 1.5873844213406887, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 39388 + }, + { + "epoch": 0.39389, + "grad_norm": 1.1454514080963143, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 39389 + }, + { + "epoch": 0.3939, + "grad_norm": 1.8548034995536595, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39390 + }, + { + "epoch": 0.39391, + "grad_norm": 1.5163162954714944, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 39391 + }, + { + "epoch": 0.39392, + "grad_norm": 1.5029109909743976, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39392 + }, + { + "epoch": 0.39393, + "grad_norm": 1.4672739115289217, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 39393 + }, + { + "epoch": 0.39394, + "grad_norm": 1.4757650727057745, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 39394 + }, + { + "epoch": 0.39395, + "grad_norm": 1.3036982223907447, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 39395 + }, + { + "epoch": 0.39396, + "grad_norm": 1.4716980375704733, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39396 + }, + { + "epoch": 0.39397, + "grad_norm": 1.466230548926502, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 39397 + }, + { + "epoch": 0.39398, + "grad_norm": 1.5322884130401666, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 39398 + }, + { + "epoch": 0.39399, + "grad_norm": 1.4590621422463177, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 39399 + }, + { + "epoch": 0.394, + "grad_norm": 1.5215295238809916, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39400 + }, + { + "epoch": 0.39401, + "grad_norm": 1.539761771088104, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 39401 + }, + { + "epoch": 0.39402, + "grad_norm": 1.5567522399113738, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39402 + }, + { + "epoch": 0.39403, + "grad_norm": 1.6009203492019366, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 39403 + }, + { + "epoch": 0.39404, + "grad_norm": 1.7052187194285808, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 39404 + }, + { + "epoch": 0.39405, + "grad_norm": 1.5116254672811855, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 39405 + }, + { + "epoch": 0.39406, + "grad_norm": 1.6539101315160363, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39406 + }, + { + "epoch": 0.39407, + "grad_norm": 1.317601689274933, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39407 + }, + { + "epoch": 0.39408, + "grad_norm": 1.504722094263392, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 39408 + }, + { + "epoch": 0.39409, + "grad_norm": 1.2462312322079752, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 39409 + }, + { + "epoch": 0.3941, + "grad_norm": 2.1348325745610253, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 39410 + }, + { + "epoch": 0.39411, + "grad_norm": 1.2524024729224077, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 39411 + }, + { + "epoch": 0.39412, + "grad_norm": 1.6636022336464775, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39412 + }, + { + "epoch": 0.39413, + "grad_norm": 1.5348380033722224, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 39413 + }, + { + "epoch": 0.39414, + "grad_norm": 1.37065731831991, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 39414 + }, + { + "epoch": 0.39415, + "grad_norm": 1.4692323685644895, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 39415 + }, + { + "epoch": 0.39416, + "grad_norm": 1.2942405288270302, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 39416 + }, + { + "epoch": 0.39417, + "grad_norm": 1.5631467464229878, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39417 + }, + { + "epoch": 0.39418, + "grad_norm": 1.2413269850220507, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 39418 + }, + { + "epoch": 0.39419, + "grad_norm": 1.6188097715586574, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 39419 + }, + { + "epoch": 0.3942, + "grad_norm": 1.520151290086518, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 39420 + }, + { + "epoch": 0.39421, + "grad_norm": 1.314943618062636, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 39421 + }, + { + "epoch": 0.39422, + "grad_norm": 1.426445649082407, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 39422 + }, + { + "epoch": 0.39423, + "grad_norm": 1.432770397504007, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 39423 + }, + { + "epoch": 0.39424, + "grad_norm": 1.5930694962950478, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39424 + }, + { + "epoch": 0.39425, + "grad_norm": 1.7784282838534506, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 39425 + }, + { + "epoch": 0.39426, + "grad_norm": 1.2694453444170792, + "learning_rate": 0.003, + "loss": 4.01, + "step": 39426 + }, + { + "epoch": 0.39427, + "grad_norm": 1.7578838012287668, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 39427 + }, + { + "epoch": 0.39428, + "grad_norm": 1.2819340835747184, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 39428 + }, + { + "epoch": 0.39429, + "grad_norm": 1.7018482071638288, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 39429 + }, + { + "epoch": 0.3943, + "grad_norm": 1.8749158183606018, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 39430 + }, + { + "epoch": 0.39431, + "grad_norm": 1.515549397467005, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 39431 + }, + { + "epoch": 0.39432, + "grad_norm": 1.4018972625466788, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 39432 + }, + { + "epoch": 0.39433, + "grad_norm": 1.5852163708008462, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39433 + }, + { + "epoch": 0.39434, + "grad_norm": 1.862350227441563, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 39434 + }, + { + "epoch": 0.39435, + "grad_norm": 1.3954274680996857, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 39435 + }, + { + "epoch": 0.39436, + "grad_norm": 1.5227897919291657, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 39436 + }, + { + "epoch": 0.39437, + "grad_norm": 1.6664244849459524, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 39437 + }, + { + "epoch": 0.39438, + "grad_norm": 1.7973881859835825, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 39438 + }, + { + "epoch": 0.39439, + "grad_norm": 1.2868280852221337, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 39439 + }, + { + "epoch": 0.3944, + "grad_norm": 1.670619517199469, + "learning_rate": 0.003, + "loss": 4.018, + "step": 39440 + }, + { + "epoch": 0.39441, + "grad_norm": 1.3004634435743188, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 39441 + }, + { + "epoch": 0.39442, + "grad_norm": 1.862952344822722, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39442 + }, + { + "epoch": 0.39443, + "grad_norm": 1.4034714770528494, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39443 + }, + { + "epoch": 0.39444, + "grad_norm": 1.6809075180033923, + "learning_rate": 0.003, + "loss": 4.073, + "step": 39444 + }, + { + "epoch": 0.39445, + "grad_norm": 1.435370883756479, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 39445 + }, + { + "epoch": 0.39446, + "grad_norm": 1.6252685151555317, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39446 + }, + { + "epoch": 0.39447, + "grad_norm": 1.6103253696521687, + "learning_rate": 0.003, + "loss": 4.032, + "step": 39447 + }, + { + "epoch": 0.39448, + "grad_norm": 1.700072547130205, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 39448 + }, + { + "epoch": 0.39449, + "grad_norm": 1.1984106267361092, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 39449 + }, + { + "epoch": 0.3945, + "grad_norm": 1.6810093016568821, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 39450 + }, + { + "epoch": 0.39451, + "grad_norm": 1.5891712312541368, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 39451 + }, + { + "epoch": 0.39452, + "grad_norm": 1.329599202669022, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 39452 + }, + { + "epoch": 0.39453, + "grad_norm": 1.7214672050750106, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 39453 + }, + { + "epoch": 0.39454, + "grad_norm": 1.381755265266865, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 39454 + }, + { + "epoch": 0.39455, + "grad_norm": 1.872769856941442, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 39455 + }, + { + "epoch": 0.39456, + "grad_norm": 1.2802310556590841, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 39456 + }, + { + "epoch": 0.39457, + "grad_norm": 1.5607774601920654, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 39457 + }, + { + "epoch": 0.39458, + "grad_norm": 1.2622407172824295, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 39458 + }, + { + "epoch": 0.39459, + "grad_norm": 1.668165291559792, + "learning_rate": 0.003, + "loss": 4.027, + "step": 39459 + }, + { + "epoch": 0.3946, + "grad_norm": 1.5245493702451933, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 39460 + }, + { + "epoch": 0.39461, + "grad_norm": 1.4347656053884574, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 39461 + }, + { + "epoch": 0.39462, + "grad_norm": 1.4507364543292887, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 39462 + }, + { + "epoch": 0.39463, + "grad_norm": 1.514591696948452, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 39463 + }, + { + "epoch": 0.39464, + "grad_norm": 1.763073425006857, + "learning_rate": 0.003, + "loss": 4.014, + "step": 39464 + }, + { + "epoch": 0.39465, + "grad_norm": 1.3366342599364713, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39465 + }, + { + "epoch": 0.39466, + "grad_norm": 1.5368189599633837, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 39466 + }, + { + "epoch": 0.39467, + "grad_norm": 1.6817903077252656, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 39467 + }, + { + "epoch": 0.39468, + "grad_norm": 1.5559525836387962, + "learning_rate": 0.003, + "loss": 4.01, + "step": 39468 + }, + { + "epoch": 0.39469, + "grad_norm": 1.4525984919645898, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 39469 + }, + { + "epoch": 0.3947, + "grad_norm": 1.2868253304835937, + "learning_rate": 0.003, + "loss": 4.054, + "step": 39470 + }, + { + "epoch": 0.39471, + "grad_norm": 1.6227804633319423, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 39471 + }, + { + "epoch": 0.39472, + "grad_norm": 1.1629651871333833, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 39472 + }, + { + "epoch": 0.39473, + "grad_norm": 1.5774165895015255, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 39473 + }, + { + "epoch": 0.39474, + "grad_norm": 1.309359755618836, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 39474 + }, + { + "epoch": 0.39475, + "grad_norm": 1.682704359903517, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39475 + }, + { + "epoch": 0.39476, + "grad_norm": 1.6877049061749456, + "learning_rate": 0.003, + "loss": 4.048, + "step": 39476 + }, + { + "epoch": 0.39477, + "grad_norm": 1.6107524859594005, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39477 + }, + { + "epoch": 0.39478, + "grad_norm": 1.7647369853209836, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39478 + }, + { + "epoch": 0.39479, + "grad_norm": 1.5150590573774774, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 39479 + }, + { + "epoch": 0.3948, + "grad_norm": 1.6658866996167683, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 39480 + }, + { + "epoch": 0.39481, + "grad_norm": 1.5346589198457647, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 39481 + }, + { + "epoch": 0.39482, + "grad_norm": 1.2086303965702847, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 39482 + }, + { + "epoch": 0.39483, + "grad_norm": 1.6786310475430701, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39483 + }, + { + "epoch": 0.39484, + "grad_norm": 1.2729399049385197, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 39484 + }, + { + "epoch": 0.39485, + "grad_norm": 1.5296316021643128, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 39485 + }, + { + "epoch": 0.39486, + "grad_norm": 1.561904738560953, + "learning_rate": 0.003, + "loss": 4.017, + "step": 39486 + }, + { + "epoch": 0.39487, + "grad_norm": 1.717254962045785, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 39487 + }, + { + "epoch": 0.39488, + "grad_norm": 1.677860396510333, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 39488 + }, + { + "epoch": 0.39489, + "grad_norm": 1.5884874684412789, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 39489 + }, + { + "epoch": 0.3949, + "grad_norm": 1.6482861993696243, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 39490 + }, + { + "epoch": 0.39491, + "grad_norm": 1.6702198165660054, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 39491 + }, + { + "epoch": 0.39492, + "grad_norm": 1.6006435076861703, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 39492 + }, + { + "epoch": 0.39493, + "grad_norm": 1.5333519654790457, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39493 + }, + { + "epoch": 0.39494, + "grad_norm": 1.4385481699078062, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39494 + }, + { + "epoch": 0.39495, + "grad_norm": 1.4642231923286606, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 39495 + }, + { + "epoch": 0.39496, + "grad_norm": 1.5469879744602848, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 39496 + }, + { + "epoch": 0.39497, + "grad_norm": 1.425246894429167, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39497 + }, + { + "epoch": 0.39498, + "grad_norm": 1.3843903834500368, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39498 + }, + { + "epoch": 0.39499, + "grad_norm": 1.4705349823950458, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39499 + }, + { + "epoch": 0.395, + "grad_norm": 1.6346069695263032, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 39500 + }, + { + "epoch": 0.39501, + "grad_norm": 1.3584465569496484, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 39501 + }, + { + "epoch": 0.39502, + "grad_norm": 1.5373142116535035, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 39502 + }, + { + "epoch": 0.39503, + "grad_norm": 1.5680447915210256, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 39503 + }, + { + "epoch": 0.39504, + "grad_norm": 1.4842892180929521, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39504 + }, + { + "epoch": 0.39505, + "grad_norm": 1.6229912722328996, + "learning_rate": 0.003, + "loss": 4.058, + "step": 39505 + }, + { + "epoch": 0.39506, + "grad_norm": 1.5784472898727933, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 39506 + }, + { + "epoch": 0.39507, + "grad_norm": 1.5376520404136111, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 39507 + }, + { + "epoch": 0.39508, + "grad_norm": 1.5440550168621772, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 39508 + }, + { + "epoch": 0.39509, + "grad_norm": 1.4832938950152312, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 39509 + }, + { + "epoch": 0.3951, + "grad_norm": 1.4200493560289442, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39510 + }, + { + "epoch": 0.39511, + "grad_norm": 1.4796700349924246, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39511 + }, + { + "epoch": 0.39512, + "grad_norm": 1.3111263262800656, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 39512 + }, + { + "epoch": 0.39513, + "grad_norm": 1.4980738049656976, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 39513 + }, + { + "epoch": 0.39514, + "grad_norm": 1.485768484515352, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39514 + }, + { + "epoch": 0.39515, + "grad_norm": 1.8183175872063175, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 39515 + }, + { + "epoch": 0.39516, + "grad_norm": 1.640146398204085, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 39516 + }, + { + "epoch": 0.39517, + "grad_norm": 1.417009653851308, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 39517 + }, + { + "epoch": 0.39518, + "grad_norm": 1.6137214068636014, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 39518 + }, + { + "epoch": 0.39519, + "grad_norm": 1.4119933107876845, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 39519 + }, + { + "epoch": 0.3952, + "grad_norm": 1.6520923095275448, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 39520 + }, + { + "epoch": 0.39521, + "grad_norm": 1.3559766583227906, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 39521 + }, + { + "epoch": 0.39522, + "grad_norm": 1.9273753904469841, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 39522 + }, + { + "epoch": 0.39523, + "grad_norm": 1.5290509826175016, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 39523 + }, + { + "epoch": 0.39524, + "grad_norm": 1.3749165118230657, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 39524 + }, + { + "epoch": 0.39525, + "grad_norm": 1.725541268076751, + "learning_rate": 0.003, + "loss": 4.053, + "step": 39525 + }, + { + "epoch": 0.39526, + "grad_norm": 1.3589982013884991, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39526 + }, + { + "epoch": 0.39527, + "grad_norm": 1.6545004935610597, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 39527 + }, + { + "epoch": 0.39528, + "grad_norm": 1.5205004987240112, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 39528 + }, + { + "epoch": 0.39529, + "grad_norm": 1.6192011902345962, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 39529 + }, + { + "epoch": 0.3953, + "grad_norm": 1.7642355772459792, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 39530 + }, + { + "epoch": 0.39531, + "grad_norm": 1.3932653764646308, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39531 + }, + { + "epoch": 0.39532, + "grad_norm": 1.445450410548313, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 39532 + }, + { + "epoch": 0.39533, + "grad_norm": 1.4453099141323276, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 39533 + }, + { + "epoch": 0.39534, + "grad_norm": 1.5136908548555177, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 39534 + }, + { + "epoch": 0.39535, + "grad_norm": 1.4951186785321144, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39535 + }, + { + "epoch": 0.39536, + "grad_norm": 1.7489214631449772, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 39536 + }, + { + "epoch": 0.39537, + "grad_norm": 1.273049888721383, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 39537 + }, + { + "epoch": 0.39538, + "grad_norm": 1.5186921400284816, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39538 + }, + { + "epoch": 0.39539, + "grad_norm": 1.533263235725368, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39539 + }, + { + "epoch": 0.3954, + "grad_norm": 1.6412053212826105, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 39540 + }, + { + "epoch": 0.39541, + "grad_norm": 1.1489131008234534, + "learning_rate": 0.003, + "loss": 4.028, + "step": 39541 + }, + { + "epoch": 0.39542, + "grad_norm": 1.908597488249271, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39542 + }, + { + "epoch": 0.39543, + "grad_norm": 1.4646314006929413, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39543 + }, + { + "epoch": 0.39544, + "grad_norm": 1.4080462616292697, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 39544 + }, + { + "epoch": 0.39545, + "grad_norm": 1.6708634559845674, + "learning_rate": 0.003, + "loss": 4.055, + "step": 39545 + }, + { + "epoch": 0.39546, + "grad_norm": 1.4214182688361623, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 39546 + }, + { + "epoch": 0.39547, + "grad_norm": 1.4648059589141438, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 39547 + }, + { + "epoch": 0.39548, + "grad_norm": 1.6833382115343303, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 39548 + }, + { + "epoch": 0.39549, + "grad_norm": 1.4049345547607235, + "learning_rate": 0.003, + "loss": 4.01, + "step": 39549 + }, + { + "epoch": 0.3955, + "grad_norm": 1.418332097781672, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 39550 + }, + { + "epoch": 0.39551, + "grad_norm": 1.5190077847311347, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 39551 + }, + { + "epoch": 0.39552, + "grad_norm": 1.7504438756449163, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 39552 + }, + { + "epoch": 0.39553, + "grad_norm": 1.994704702309346, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 39553 + }, + { + "epoch": 0.39554, + "grad_norm": 1.9918060195136191, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 39554 + }, + { + "epoch": 0.39555, + "grad_norm": 1.1948761343716643, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 39555 + }, + { + "epoch": 0.39556, + "grad_norm": 1.684131816075993, + "learning_rate": 0.003, + "loss": 4.07, + "step": 39556 + }, + { + "epoch": 0.39557, + "grad_norm": 1.9769476345240364, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39557 + }, + { + "epoch": 0.39558, + "grad_norm": 1.411275503609052, + "learning_rate": 0.003, + "loss": 4.029, + "step": 39558 + }, + { + "epoch": 0.39559, + "grad_norm": 1.6063109922641772, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 39559 + }, + { + "epoch": 0.3956, + "grad_norm": 1.369585276329642, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 39560 + }, + { + "epoch": 0.39561, + "grad_norm": 1.9639404162697818, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 39561 + }, + { + "epoch": 0.39562, + "grad_norm": 1.2727517567784001, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 39562 + }, + { + "epoch": 0.39563, + "grad_norm": 1.4770482646524805, + "learning_rate": 0.003, + "loss": 4.047, + "step": 39563 + }, + { + "epoch": 0.39564, + "grad_norm": 1.5983356258465462, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39564 + }, + { + "epoch": 0.39565, + "grad_norm": 1.3798035638379003, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 39565 + }, + { + "epoch": 0.39566, + "grad_norm": 1.6240731906713721, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39566 + }, + { + "epoch": 0.39567, + "grad_norm": 1.6017822046431789, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39567 + }, + { + "epoch": 0.39568, + "grad_norm": 1.804752852739869, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 39568 + }, + { + "epoch": 0.39569, + "grad_norm": 1.2266859731605433, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 39569 + }, + { + "epoch": 0.3957, + "grad_norm": 1.6821411996984483, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39570 + }, + { + "epoch": 0.39571, + "grad_norm": 1.431126171229569, + "learning_rate": 0.003, + "loss": 4.034, + "step": 39571 + }, + { + "epoch": 0.39572, + "grad_norm": 1.4678266416629195, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39572 + }, + { + "epoch": 0.39573, + "grad_norm": 1.493463173922419, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 39573 + }, + { + "epoch": 0.39574, + "grad_norm": 1.4173891134246528, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 39574 + }, + { + "epoch": 0.39575, + "grad_norm": 1.4739100801541996, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 39575 + }, + { + "epoch": 0.39576, + "grad_norm": 1.6559009846136337, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 39576 + }, + { + "epoch": 0.39577, + "grad_norm": 1.5491766764756147, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 39577 + }, + { + "epoch": 0.39578, + "grad_norm": 1.2483228384854874, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 39578 + }, + { + "epoch": 0.39579, + "grad_norm": 1.8154939824928777, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 39579 + }, + { + "epoch": 0.3958, + "grad_norm": 1.037306989002405, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 39580 + }, + { + "epoch": 0.39581, + "grad_norm": 2.2718236142253487, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 39581 + }, + { + "epoch": 0.39582, + "grad_norm": 1.313829449006874, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 39582 + }, + { + "epoch": 0.39583, + "grad_norm": 1.5231590049504997, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 39583 + }, + { + "epoch": 0.39584, + "grad_norm": 1.4506787302363506, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 39584 + }, + { + "epoch": 0.39585, + "grad_norm": 1.777131204666326, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 39585 + }, + { + "epoch": 0.39586, + "grad_norm": 1.6050906759292107, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 39586 + }, + { + "epoch": 0.39587, + "grad_norm": 1.5408937614798959, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 39587 + }, + { + "epoch": 0.39588, + "grad_norm": 1.3242741552645547, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 39588 + }, + { + "epoch": 0.39589, + "grad_norm": 1.5380686426777124, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39589 + }, + { + "epoch": 0.3959, + "grad_norm": 1.8408322198594864, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 39590 + }, + { + "epoch": 0.39591, + "grad_norm": 1.1890920452551248, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 39591 + }, + { + "epoch": 0.39592, + "grad_norm": 1.790628714088605, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 39592 + }, + { + "epoch": 0.39593, + "grad_norm": 1.370649606466643, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 39593 + }, + { + "epoch": 0.39594, + "grad_norm": 1.8654207892050239, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39594 + }, + { + "epoch": 0.39595, + "grad_norm": 1.5502045266926292, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 39595 + }, + { + "epoch": 0.39596, + "grad_norm": 1.4999338463734244, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39596 + }, + { + "epoch": 0.39597, + "grad_norm": 1.4321496235597329, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 39597 + }, + { + "epoch": 0.39598, + "grad_norm": 1.591368954506369, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 39598 + }, + { + "epoch": 0.39599, + "grad_norm": 1.7503627261607835, + "learning_rate": 0.003, + "loss": 4.091, + "step": 39599 + }, + { + "epoch": 0.396, + "grad_norm": 1.4287132480751266, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 39600 + }, + { + "epoch": 0.39601, + "grad_norm": 1.6440254657021742, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 39601 + }, + { + "epoch": 0.39602, + "grad_norm": 1.6010294826998628, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 39602 + }, + { + "epoch": 0.39603, + "grad_norm": 1.4126881484379392, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 39603 + }, + { + "epoch": 0.39604, + "grad_norm": 1.6685446904853976, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 39604 + }, + { + "epoch": 0.39605, + "grad_norm": 1.080439335424512, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 39605 + }, + { + "epoch": 0.39606, + "grad_norm": 1.7696419342212588, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 39606 + }, + { + "epoch": 0.39607, + "grad_norm": 1.0062254329527902, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 39607 + }, + { + "epoch": 0.39608, + "grad_norm": 1.8850494140724467, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 39608 + }, + { + "epoch": 0.39609, + "grad_norm": 1.4648587054688254, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39609 + }, + { + "epoch": 0.3961, + "grad_norm": 1.6944817438488873, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 39610 + }, + { + "epoch": 0.39611, + "grad_norm": 1.3658178062297928, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39611 + }, + { + "epoch": 0.39612, + "grad_norm": 1.8178652013945056, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 39612 + }, + { + "epoch": 0.39613, + "grad_norm": 1.4861999436448692, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 39613 + }, + { + "epoch": 0.39614, + "grad_norm": 1.3587458379333055, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 39614 + }, + { + "epoch": 0.39615, + "grad_norm": 1.354513006909419, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 39615 + }, + { + "epoch": 0.39616, + "grad_norm": 1.6279186841903739, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 39616 + }, + { + "epoch": 0.39617, + "grad_norm": 1.615613920651119, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 39617 + }, + { + "epoch": 0.39618, + "grad_norm": 1.2775625875875498, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 39618 + }, + { + "epoch": 0.39619, + "grad_norm": 1.7198116250998767, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 39619 + }, + { + "epoch": 0.3962, + "grad_norm": 1.8855636362480601, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 39620 + }, + { + "epoch": 0.39621, + "grad_norm": 1.615253866358725, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 39621 + }, + { + "epoch": 0.39622, + "grad_norm": 1.5891173390025264, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 39622 + }, + { + "epoch": 0.39623, + "grad_norm": 1.5810001204155448, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 39623 + }, + { + "epoch": 0.39624, + "grad_norm": 1.4646935397138576, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39624 + }, + { + "epoch": 0.39625, + "grad_norm": 1.3654242911334933, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 39625 + }, + { + "epoch": 0.39626, + "grad_norm": 1.7038766935777212, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39626 + }, + { + "epoch": 0.39627, + "grad_norm": 1.3349709657929414, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 39627 + }, + { + "epoch": 0.39628, + "grad_norm": 1.6043034225067834, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 39628 + }, + { + "epoch": 0.39629, + "grad_norm": 1.4629932594730433, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 39629 + }, + { + "epoch": 0.3963, + "grad_norm": 1.5388395873783287, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 39630 + }, + { + "epoch": 0.39631, + "grad_norm": 1.446157520691181, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 39631 + }, + { + "epoch": 0.39632, + "grad_norm": 1.4808298093789813, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 39632 + }, + { + "epoch": 0.39633, + "grad_norm": 1.4089441208064433, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 39633 + }, + { + "epoch": 0.39634, + "grad_norm": 1.5681978923784985, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 39634 + }, + { + "epoch": 0.39635, + "grad_norm": 1.31683024273901, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 39635 + }, + { + "epoch": 0.39636, + "grad_norm": 1.7930989958316348, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 39636 + }, + { + "epoch": 0.39637, + "grad_norm": 1.0801528351521401, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 39637 + }, + { + "epoch": 0.39638, + "grad_norm": 2.2029655931667795, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 39638 + }, + { + "epoch": 0.39639, + "grad_norm": 1.5050592099200162, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 39639 + }, + { + "epoch": 0.3964, + "grad_norm": 1.3570400363668849, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 39640 + }, + { + "epoch": 0.39641, + "grad_norm": 1.506640695724285, + "learning_rate": 0.003, + "loss": 4.028, + "step": 39641 + }, + { + "epoch": 0.39642, + "grad_norm": 1.577793153141463, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 39642 + }, + { + "epoch": 0.39643, + "grad_norm": 1.5196384077419869, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 39643 + }, + { + "epoch": 0.39644, + "grad_norm": 1.3514647659100942, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 39644 + }, + { + "epoch": 0.39645, + "grad_norm": 1.4511136972693635, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 39645 + }, + { + "epoch": 0.39646, + "grad_norm": 1.4475947140119108, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 39646 + }, + { + "epoch": 0.39647, + "grad_norm": 1.1960355369760407, + "learning_rate": 0.003, + "loss": 4.017, + "step": 39647 + }, + { + "epoch": 0.39648, + "grad_norm": 1.6316912896179865, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 39648 + }, + { + "epoch": 0.39649, + "grad_norm": 1.3534783306679585, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 39649 + }, + { + "epoch": 0.3965, + "grad_norm": 1.4785989111987494, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39650 + }, + { + "epoch": 0.39651, + "grad_norm": 1.6164892262908972, + "learning_rate": 0.003, + "loss": 4.073, + "step": 39651 + }, + { + "epoch": 0.39652, + "grad_norm": 1.6068457534634502, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 39652 + }, + { + "epoch": 0.39653, + "grad_norm": 1.7539139419129106, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39653 + }, + { + "epoch": 0.39654, + "grad_norm": 1.45503495321086, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 39654 + }, + { + "epoch": 0.39655, + "grad_norm": 1.6233940491611991, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 39655 + }, + { + "epoch": 0.39656, + "grad_norm": 1.696744822317181, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 39656 + }, + { + "epoch": 0.39657, + "grad_norm": 1.4323822973135871, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 39657 + }, + { + "epoch": 0.39658, + "grad_norm": 1.4295592204158265, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 39658 + }, + { + "epoch": 0.39659, + "grad_norm": 1.4170524724262095, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39659 + }, + { + "epoch": 0.3966, + "grad_norm": 1.3329711449342259, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 39660 + }, + { + "epoch": 0.39661, + "grad_norm": 1.8530033590865995, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 39661 + }, + { + "epoch": 0.39662, + "grad_norm": 1.1556804719511296, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 39662 + }, + { + "epoch": 0.39663, + "grad_norm": 1.952578759833693, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 39663 + }, + { + "epoch": 0.39664, + "grad_norm": 1.2780576042641305, + "learning_rate": 0.003, + "loss": 4.051, + "step": 39664 + }, + { + "epoch": 0.39665, + "grad_norm": 1.7265435216337657, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 39665 + }, + { + "epoch": 0.39666, + "grad_norm": 1.435574461346075, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39666 + }, + { + "epoch": 0.39667, + "grad_norm": 1.9211214142211377, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 39667 + }, + { + "epoch": 0.39668, + "grad_norm": 1.6233956676449088, + "learning_rate": 0.003, + "loss": 4.047, + "step": 39668 + }, + { + "epoch": 0.39669, + "grad_norm": 1.3743939246038193, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 39669 + }, + { + "epoch": 0.3967, + "grad_norm": 1.4100026545574174, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 39670 + }, + { + "epoch": 0.39671, + "grad_norm": 1.6580435491459444, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 39671 + }, + { + "epoch": 0.39672, + "grad_norm": 1.5680860316715908, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 39672 + }, + { + "epoch": 0.39673, + "grad_norm": 1.2926263166149585, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39673 + }, + { + "epoch": 0.39674, + "grad_norm": 1.7295217936157368, + "learning_rate": 0.003, + "loss": 4.048, + "step": 39674 + }, + { + "epoch": 0.39675, + "grad_norm": 1.2123891876373074, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39675 + }, + { + "epoch": 0.39676, + "grad_norm": 1.7609121511013794, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 39676 + }, + { + "epoch": 0.39677, + "grad_norm": 1.412046061840759, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39677 + }, + { + "epoch": 0.39678, + "grad_norm": 1.702288542704342, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 39678 + }, + { + "epoch": 0.39679, + "grad_norm": 1.2893277172666033, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 39679 + }, + { + "epoch": 0.3968, + "grad_norm": 1.6419800766077317, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 39680 + }, + { + "epoch": 0.39681, + "grad_norm": 1.6690396038060153, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 39681 + }, + { + "epoch": 0.39682, + "grad_norm": 1.824437704347926, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 39682 + }, + { + "epoch": 0.39683, + "grad_norm": 1.3260340581575183, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 39683 + }, + { + "epoch": 0.39684, + "grad_norm": 1.369027234563807, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 39684 + }, + { + "epoch": 0.39685, + "grad_norm": 1.7374709666874326, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 39685 + }, + { + "epoch": 0.39686, + "grad_norm": 1.3664357627291484, + "learning_rate": 0.003, + "loss": 4.023, + "step": 39686 + }, + { + "epoch": 0.39687, + "grad_norm": 1.8948881278673109, + "learning_rate": 0.003, + "loss": 4.039, + "step": 39687 + }, + { + "epoch": 0.39688, + "grad_norm": 1.2413053293194078, + "learning_rate": 0.003, + "loss": 4.073, + "step": 39688 + }, + { + "epoch": 0.39689, + "grad_norm": 1.8837140421517855, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 39689 + }, + { + "epoch": 0.3969, + "grad_norm": 1.2671890752864352, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 39690 + }, + { + "epoch": 0.39691, + "grad_norm": 1.9556110125266692, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 39691 + }, + { + "epoch": 0.39692, + "grad_norm": 1.3102653701751705, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 39692 + }, + { + "epoch": 0.39693, + "grad_norm": 1.7647800723304634, + "learning_rate": 0.003, + "loss": 3.992, + "step": 39693 + }, + { + "epoch": 0.39694, + "grad_norm": 1.4335365833961025, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 39694 + }, + { + "epoch": 0.39695, + "grad_norm": 1.3425722518768357, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 39695 + }, + { + "epoch": 0.39696, + "grad_norm": 1.56761030064612, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 39696 + }, + { + "epoch": 0.39697, + "grad_norm": 1.3927564338481393, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 39697 + }, + { + "epoch": 0.39698, + "grad_norm": 1.3349129819875658, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 39698 + }, + { + "epoch": 0.39699, + "grad_norm": 1.6407716502459329, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 39699 + }, + { + "epoch": 0.397, + "grad_norm": 1.8085727888173402, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 39700 + }, + { + "epoch": 0.39701, + "grad_norm": 1.5510136492625564, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 39701 + }, + { + "epoch": 0.39702, + "grad_norm": 1.7931901154044037, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 39702 + }, + { + "epoch": 0.39703, + "grad_norm": 1.3450858949822875, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 39703 + }, + { + "epoch": 0.39704, + "grad_norm": 1.716292295384698, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 39704 + }, + { + "epoch": 0.39705, + "grad_norm": 1.6459749456592547, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 39705 + }, + { + "epoch": 0.39706, + "grad_norm": 1.3215707899642295, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 39706 + }, + { + "epoch": 0.39707, + "grad_norm": 1.471702185126159, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 39707 + }, + { + "epoch": 0.39708, + "grad_norm": 1.7186724754060965, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 39708 + }, + { + "epoch": 0.39709, + "grad_norm": 1.2301822037429235, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 39709 + }, + { + "epoch": 0.3971, + "grad_norm": 1.6356626303313768, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 39710 + }, + { + "epoch": 0.39711, + "grad_norm": 1.431006028944067, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 39711 + }, + { + "epoch": 0.39712, + "grad_norm": 1.5538566323421699, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 39712 + }, + { + "epoch": 0.39713, + "grad_norm": 1.2045241628634165, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 39713 + }, + { + "epoch": 0.39714, + "grad_norm": 1.7165183044644305, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 39714 + }, + { + "epoch": 0.39715, + "grad_norm": 1.1629581007313414, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 39715 + }, + { + "epoch": 0.39716, + "grad_norm": 1.784149391106247, + "learning_rate": 0.003, + "loss": 4.034, + "step": 39716 + }, + { + "epoch": 0.39717, + "grad_norm": 1.3892407213747395, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 39717 + }, + { + "epoch": 0.39718, + "grad_norm": 1.727926395206028, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 39718 + }, + { + "epoch": 0.39719, + "grad_norm": 1.487665818506082, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 39719 + }, + { + "epoch": 0.3972, + "grad_norm": 1.4874179149552382, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 39720 + }, + { + "epoch": 0.39721, + "grad_norm": 1.6813210541710528, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 39721 + }, + { + "epoch": 0.39722, + "grad_norm": 1.2217528249953815, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 39722 + }, + { + "epoch": 0.39723, + "grad_norm": 1.555636685589338, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 39723 + }, + { + "epoch": 0.39724, + "grad_norm": 1.439886874666372, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 39724 + }, + { + "epoch": 0.39725, + "grad_norm": 1.618394829341988, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 39725 + }, + { + "epoch": 0.39726, + "grad_norm": 1.5985986630050386, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39726 + }, + { + "epoch": 0.39727, + "grad_norm": 1.4549380044883702, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39727 + }, + { + "epoch": 0.39728, + "grad_norm": 1.5229355316890651, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 39728 + }, + { + "epoch": 0.39729, + "grad_norm": 1.8059581490241177, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 39729 + }, + { + "epoch": 0.3973, + "grad_norm": 1.2670916576023201, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 39730 + }, + { + "epoch": 0.39731, + "grad_norm": 1.596054683567535, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 39731 + }, + { + "epoch": 0.39732, + "grad_norm": 1.7025997742550485, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 39732 + }, + { + "epoch": 0.39733, + "grad_norm": 1.6279326382360462, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 39733 + }, + { + "epoch": 0.39734, + "grad_norm": 1.605387401157188, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 39734 + }, + { + "epoch": 0.39735, + "grad_norm": 1.5043246517047515, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 39735 + }, + { + "epoch": 0.39736, + "grad_norm": 1.434315553157176, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 39736 + }, + { + "epoch": 0.39737, + "grad_norm": 1.6041974557918723, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39737 + }, + { + "epoch": 0.39738, + "grad_norm": 1.1869797480539428, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 39738 + }, + { + "epoch": 0.39739, + "grad_norm": 1.9596748502907868, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 39739 + }, + { + "epoch": 0.3974, + "grad_norm": 1.3091216584503615, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 39740 + }, + { + "epoch": 0.39741, + "grad_norm": 1.557764086321222, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 39741 + }, + { + "epoch": 0.39742, + "grad_norm": 1.286248828727619, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 39742 + }, + { + "epoch": 0.39743, + "grad_norm": 1.6375502864915363, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 39743 + }, + { + "epoch": 0.39744, + "grad_norm": 1.486934439561368, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 39744 + }, + { + "epoch": 0.39745, + "grad_norm": 1.5576770889010805, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 39745 + }, + { + "epoch": 0.39746, + "grad_norm": 1.448114917215483, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 39746 + }, + { + "epoch": 0.39747, + "grad_norm": 1.4033681006513645, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39747 + }, + { + "epoch": 0.39748, + "grad_norm": 1.3680947394088419, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 39748 + }, + { + "epoch": 0.39749, + "grad_norm": 1.7772539368293638, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39749 + }, + { + "epoch": 0.3975, + "grad_norm": 1.5109549918832776, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 39750 + }, + { + "epoch": 0.39751, + "grad_norm": 1.9985311422731835, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 39751 + }, + { + "epoch": 0.39752, + "grad_norm": 1.4406936086907056, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 39752 + }, + { + "epoch": 0.39753, + "grad_norm": 1.2598926970282214, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39753 + }, + { + "epoch": 0.39754, + "grad_norm": 1.704707157700129, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 39754 + }, + { + "epoch": 0.39755, + "grad_norm": 1.4717787006948384, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 39755 + }, + { + "epoch": 0.39756, + "grad_norm": 1.6161965185330407, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39756 + }, + { + "epoch": 0.39757, + "grad_norm": 1.6828461703813546, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 39757 + }, + { + "epoch": 0.39758, + "grad_norm": 1.4322210795744927, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 39758 + }, + { + "epoch": 0.39759, + "grad_norm": 1.6650748742345505, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 39759 + }, + { + "epoch": 0.3976, + "grad_norm": 1.508213726086184, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 39760 + }, + { + "epoch": 0.39761, + "grad_norm": 1.4959847231275465, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 39761 + }, + { + "epoch": 0.39762, + "grad_norm": 1.623956696907353, + "learning_rate": 0.003, + "loss": 4.063, + "step": 39762 + }, + { + "epoch": 0.39763, + "grad_norm": 1.387686044332844, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 39763 + }, + { + "epoch": 0.39764, + "grad_norm": 1.7013383634114325, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 39764 + }, + { + "epoch": 0.39765, + "grad_norm": 1.0817265616273766, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 39765 + }, + { + "epoch": 0.39766, + "grad_norm": 1.8822923548535544, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 39766 + }, + { + "epoch": 0.39767, + "grad_norm": 1.4350336870835343, + "learning_rate": 0.003, + "loss": 4.043, + "step": 39767 + }, + { + "epoch": 0.39768, + "grad_norm": 1.5946765982376663, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39768 + }, + { + "epoch": 0.39769, + "grad_norm": 1.4320882035269265, + "learning_rate": 0.003, + "loss": 4.04, + "step": 39769 + }, + { + "epoch": 0.3977, + "grad_norm": 2.1601726714033216, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 39770 + }, + { + "epoch": 0.39771, + "grad_norm": 1.3333186954125744, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 39771 + }, + { + "epoch": 0.39772, + "grad_norm": 1.6582436965730833, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 39772 + }, + { + "epoch": 0.39773, + "grad_norm": 1.6680243025416484, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 39773 + }, + { + "epoch": 0.39774, + "grad_norm": 1.1976297146699473, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 39774 + }, + { + "epoch": 0.39775, + "grad_norm": 2.3290821019525203, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 39775 + }, + { + "epoch": 0.39776, + "grad_norm": 1.3121065557566072, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 39776 + }, + { + "epoch": 0.39777, + "grad_norm": 1.4458093245868717, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 39777 + }, + { + "epoch": 0.39778, + "grad_norm": 1.6054881510601122, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39778 + }, + { + "epoch": 0.39779, + "grad_norm": 1.6643555844208093, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 39779 + }, + { + "epoch": 0.3978, + "grad_norm": 1.4528230936732758, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 39780 + }, + { + "epoch": 0.39781, + "grad_norm": 1.4077037765210576, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 39781 + }, + { + "epoch": 0.39782, + "grad_norm": 1.2696179559082865, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 39782 + }, + { + "epoch": 0.39783, + "grad_norm": 1.7329508788161618, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 39783 + }, + { + "epoch": 0.39784, + "grad_norm": 1.3835916987222527, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 39784 + }, + { + "epoch": 0.39785, + "grad_norm": 1.5722557449129726, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 39785 + }, + { + "epoch": 0.39786, + "grad_norm": 1.4608984227527473, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39786 + }, + { + "epoch": 0.39787, + "grad_norm": 1.4608394994405214, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 39787 + }, + { + "epoch": 0.39788, + "grad_norm": 1.8685347558048082, + "learning_rate": 0.003, + "loss": 4.055, + "step": 39788 + }, + { + "epoch": 0.39789, + "grad_norm": 1.292046731791235, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 39789 + }, + { + "epoch": 0.3979, + "grad_norm": 1.596713295801229, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 39790 + }, + { + "epoch": 0.39791, + "grad_norm": 1.3881400972234916, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 39791 + }, + { + "epoch": 0.39792, + "grad_norm": 1.7698132520182834, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 39792 + }, + { + "epoch": 0.39793, + "grad_norm": 1.230519459208385, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39793 + }, + { + "epoch": 0.39794, + "grad_norm": 1.6396701022696216, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 39794 + }, + { + "epoch": 0.39795, + "grad_norm": 1.6247053620586367, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39795 + }, + { + "epoch": 0.39796, + "grad_norm": 1.554544428441785, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39796 + }, + { + "epoch": 0.39797, + "grad_norm": 1.303648101167199, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 39797 + }, + { + "epoch": 0.39798, + "grad_norm": 1.5438967741196243, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 39798 + }, + { + "epoch": 0.39799, + "grad_norm": 1.7553686970636875, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39799 + }, + { + "epoch": 0.398, + "grad_norm": 1.2433619962420215, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 39800 + }, + { + "epoch": 0.39801, + "grad_norm": 1.9352653146876915, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 39801 + }, + { + "epoch": 0.39802, + "grad_norm": 1.1730980134048439, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 39802 + }, + { + "epoch": 0.39803, + "grad_norm": 1.7000159638528978, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 39803 + }, + { + "epoch": 0.39804, + "grad_norm": 1.4654687351907236, + "learning_rate": 0.003, + "loss": 4.082, + "step": 39804 + }, + { + "epoch": 0.39805, + "grad_norm": 1.445255884560971, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39805 + }, + { + "epoch": 0.39806, + "grad_norm": 1.5308944978556762, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 39806 + }, + { + "epoch": 0.39807, + "grad_norm": 1.743916199673013, + "learning_rate": 0.003, + "loss": 4.09, + "step": 39807 + }, + { + "epoch": 0.39808, + "grad_norm": 1.1986753374306685, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 39808 + }, + { + "epoch": 0.39809, + "grad_norm": 1.7858292522657313, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 39809 + }, + { + "epoch": 0.3981, + "grad_norm": 1.1375199130159095, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 39810 + }, + { + "epoch": 0.39811, + "grad_norm": 1.6220852207007783, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 39811 + }, + { + "epoch": 0.39812, + "grad_norm": 1.6860770537772634, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 39812 + }, + { + "epoch": 0.39813, + "grad_norm": 1.5748893546708511, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 39813 + }, + { + "epoch": 0.39814, + "grad_norm": 1.8932911167243405, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 39814 + }, + { + "epoch": 0.39815, + "grad_norm": 1.1654465329476877, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 39815 + }, + { + "epoch": 0.39816, + "grad_norm": 1.7225634529763265, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 39816 + }, + { + "epoch": 0.39817, + "grad_norm": 1.5033481327728906, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 39817 + }, + { + "epoch": 0.39818, + "grad_norm": 1.6635243046948738, + "learning_rate": 0.003, + "loss": 4.049, + "step": 39818 + }, + { + "epoch": 0.39819, + "grad_norm": 1.7886402329164328, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 39819 + }, + { + "epoch": 0.3982, + "grad_norm": 1.3625679750397541, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39820 + }, + { + "epoch": 0.39821, + "grad_norm": 1.7420432286814729, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 39821 + }, + { + "epoch": 0.39822, + "grad_norm": 1.4076562400641526, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 39822 + }, + { + "epoch": 0.39823, + "grad_norm": 1.374198129047758, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 39823 + }, + { + "epoch": 0.39824, + "grad_norm": 1.5004689663465836, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 39824 + }, + { + "epoch": 0.39825, + "grad_norm": 1.7416197074858235, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 39825 + }, + { + "epoch": 0.39826, + "grad_norm": 1.2157465147950013, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 39826 + }, + { + "epoch": 0.39827, + "grad_norm": 1.6414141767660895, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 39827 + }, + { + "epoch": 0.39828, + "grad_norm": 1.4969231592352952, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 39828 + }, + { + "epoch": 0.39829, + "grad_norm": 1.6987912083075916, + "learning_rate": 0.003, + "loss": 4.037, + "step": 39829 + }, + { + "epoch": 0.3983, + "grad_norm": 1.2640659986567238, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 39830 + }, + { + "epoch": 0.39831, + "grad_norm": 1.4275636603747686, + "learning_rate": 0.003, + "loss": 4.063, + "step": 39831 + }, + { + "epoch": 0.39832, + "grad_norm": 1.8149049126512162, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39832 + }, + { + "epoch": 0.39833, + "grad_norm": 1.2723327601690433, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 39833 + }, + { + "epoch": 0.39834, + "grad_norm": 1.8146216679103726, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 39834 + }, + { + "epoch": 0.39835, + "grad_norm": 1.3780855770421916, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 39835 + }, + { + "epoch": 0.39836, + "grad_norm": 1.8360555397765743, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 39836 + }, + { + "epoch": 0.39837, + "grad_norm": 1.3897966955115184, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 39837 + }, + { + "epoch": 0.39838, + "grad_norm": 1.569033802216872, + "learning_rate": 0.003, + "loss": 4.042, + "step": 39838 + }, + { + "epoch": 0.39839, + "grad_norm": 1.3951369805896685, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 39839 + }, + { + "epoch": 0.3984, + "grad_norm": 1.5026113673111359, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 39840 + }, + { + "epoch": 0.39841, + "grad_norm": 1.4579940803480718, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 39841 + }, + { + "epoch": 0.39842, + "grad_norm": 1.491761100144407, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 39842 + }, + { + "epoch": 0.39843, + "grad_norm": 1.4265460195265236, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 39843 + }, + { + "epoch": 0.39844, + "grad_norm": 1.5866448252236491, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 39844 + }, + { + "epoch": 0.39845, + "grad_norm": 1.487296165407626, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 39845 + }, + { + "epoch": 0.39846, + "grad_norm": 1.5288669238068613, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 39846 + }, + { + "epoch": 0.39847, + "grad_norm": 1.4607362447240126, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 39847 + }, + { + "epoch": 0.39848, + "grad_norm": 1.4827613514555527, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 39848 + }, + { + "epoch": 0.39849, + "grad_norm": 1.4948632350854814, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39849 + }, + { + "epoch": 0.3985, + "grad_norm": 1.5255127967323228, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 39850 + }, + { + "epoch": 0.39851, + "grad_norm": 1.2736090680897585, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 39851 + }, + { + "epoch": 0.39852, + "grad_norm": 1.6326442937728256, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 39852 + }, + { + "epoch": 0.39853, + "grad_norm": 1.273512165305973, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 39853 + }, + { + "epoch": 0.39854, + "grad_norm": 1.608126101636675, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 39854 + }, + { + "epoch": 0.39855, + "grad_norm": 1.4227260663137267, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39855 + }, + { + "epoch": 0.39856, + "grad_norm": 1.7341809299896571, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39856 + }, + { + "epoch": 0.39857, + "grad_norm": 1.361255730927451, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 39857 + }, + { + "epoch": 0.39858, + "grad_norm": 1.7644212162012618, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 39858 + }, + { + "epoch": 0.39859, + "grad_norm": 1.458233422286207, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 39859 + }, + { + "epoch": 0.3986, + "grad_norm": 1.714909859124863, + "learning_rate": 0.003, + "loss": 4.027, + "step": 39860 + }, + { + "epoch": 0.39861, + "grad_norm": 1.492265597106838, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39861 + }, + { + "epoch": 0.39862, + "grad_norm": 1.6111501088080238, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 39862 + }, + { + "epoch": 0.39863, + "grad_norm": 1.3957436791052902, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 39863 + }, + { + "epoch": 0.39864, + "grad_norm": 1.7607124031766026, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39864 + }, + { + "epoch": 0.39865, + "grad_norm": 1.3572197106699158, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 39865 + }, + { + "epoch": 0.39866, + "grad_norm": 1.4517884204118037, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 39866 + }, + { + "epoch": 0.39867, + "grad_norm": 1.659077828831311, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 39867 + }, + { + "epoch": 0.39868, + "grad_norm": 1.3654414713455238, + "learning_rate": 0.003, + "loss": 4.03, + "step": 39868 + }, + { + "epoch": 0.39869, + "grad_norm": 1.6398613869686165, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39869 + }, + { + "epoch": 0.3987, + "grad_norm": 1.3997523177769429, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 39870 + }, + { + "epoch": 0.39871, + "grad_norm": 1.5043528969056972, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 39871 + }, + { + "epoch": 0.39872, + "grad_norm": 1.448204151725989, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 39872 + }, + { + "epoch": 0.39873, + "grad_norm": 1.6356429806295714, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 39873 + }, + { + "epoch": 0.39874, + "grad_norm": 1.4313551678176657, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 39874 + }, + { + "epoch": 0.39875, + "grad_norm": 1.7739801077700084, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 39875 + }, + { + "epoch": 0.39876, + "grad_norm": 1.3249454089641952, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 39876 + }, + { + "epoch": 0.39877, + "grad_norm": 1.4698197817262442, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 39877 + }, + { + "epoch": 0.39878, + "grad_norm": 1.4432770897089133, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 39878 + }, + { + "epoch": 0.39879, + "grad_norm": 1.4947267434639948, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 39879 + }, + { + "epoch": 0.3988, + "grad_norm": 1.5833638610358955, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 39880 + }, + { + "epoch": 0.39881, + "grad_norm": 1.1886534047922352, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 39881 + }, + { + "epoch": 0.39882, + "grad_norm": 1.8368026866346414, + "learning_rate": 0.003, + "loss": 4.064, + "step": 39882 + }, + { + "epoch": 0.39883, + "grad_norm": 1.4173462016283263, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 39883 + }, + { + "epoch": 0.39884, + "grad_norm": 1.7677489971015858, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 39884 + }, + { + "epoch": 0.39885, + "grad_norm": 1.3842455376766174, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 39885 + }, + { + "epoch": 0.39886, + "grad_norm": 1.9130582365256843, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 39886 + }, + { + "epoch": 0.39887, + "grad_norm": 1.4943391856832253, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39887 + }, + { + "epoch": 0.39888, + "grad_norm": 1.313538906043359, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 39888 + }, + { + "epoch": 0.39889, + "grad_norm": 1.7651062952762013, + "learning_rate": 0.003, + "loss": 4.041, + "step": 39889 + }, + { + "epoch": 0.3989, + "grad_norm": 1.6030767027849846, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 39890 + }, + { + "epoch": 0.39891, + "grad_norm": 1.4832641753236027, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 39891 + }, + { + "epoch": 0.39892, + "grad_norm": 1.4021083346959258, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 39892 + }, + { + "epoch": 0.39893, + "grad_norm": 1.4342879259464596, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 39893 + }, + { + "epoch": 0.39894, + "grad_norm": 1.5290477374763962, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 39894 + }, + { + "epoch": 0.39895, + "grad_norm": 1.464539220808174, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 39895 + }, + { + "epoch": 0.39896, + "grad_norm": 1.4276830514955463, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 39896 + }, + { + "epoch": 0.39897, + "grad_norm": 1.6956719501204203, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 39897 + }, + { + "epoch": 0.39898, + "grad_norm": 1.5476988471435413, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 39898 + }, + { + "epoch": 0.39899, + "grad_norm": 1.8191941781654952, + "learning_rate": 0.003, + "loss": 4.069, + "step": 39899 + }, + { + "epoch": 0.399, + "grad_norm": 1.508437749789631, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 39900 + }, + { + "epoch": 0.39901, + "grad_norm": 1.5483485691084955, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 39901 + }, + { + "epoch": 0.39902, + "grad_norm": 1.3191930687778082, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 39902 + }, + { + "epoch": 0.39903, + "grad_norm": 1.7174751700130029, + "learning_rate": 0.003, + "loss": 4.048, + "step": 39903 + }, + { + "epoch": 0.39904, + "grad_norm": 1.145834666155481, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 39904 + }, + { + "epoch": 0.39905, + "grad_norm": 1.5625353287312802, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 39905 + }, + { + "epoch": 0.39906, + "grad_norm": 1.8534049783548368, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 39906 + }, + { + "epoch": 0.39907, + "grad_norm": 1.4041007654215147, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 39907 + }, + { + "epoch": 0.39908, + "grad_norm": 1.5705471374339415, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 39908 + }, + { + "epoch": 0.39909, + "grad_norm": 1.6707766227481264, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 39909 + }, + { + "epoch": 0.3991, + "grad_norm": 1.5375352166161627, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 39910 + }, + { + "epoch": 0.39911, + "grad_norm": 1.465120606781287, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 39911 + }, + { + "epoch": 0.39912, + "grad_norm": 1.4376226996078743, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 39912 + }, + { + "epoch": 0.39913, + "grad_norm": 1.546497579121612, + "learning_rate": 0.003, + "loss": 4.04, + "step": 39913 + }, + { + "epoch": 0.39914, + "grad_norm": 1.2396092360330315, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 39914 + }, + { + "epoch": 0.39915, + "grad_norm": 1.7122796786640209, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 39915 + }, + { + "epoch": 0.39916, + "grad_norm": 1.1366235423881417, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 39916 + }, + { + "epoch": 0.39917, + "grad_norm": 1.7782564786226682, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 39917 + }, + { + "epoch": 0.39918, + "grad_norm": 1.6763077071143335, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 39918 + }, + { + "epoch": 0.39919, + "grad_norm": 1.7168439192674967, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 39919 + }, + { + "epoch": 0.3992, + "grad_norm": 1.1312454339589035, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 39920 + }, + { + "epoch": 0.39921, + "grad_norm": 1.8554896185350516, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 39921 + }, + { + "epoch": 0.39922, + "grad_norm": 1.5495107563673516, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 39922 + }, + { + "epoch": 0.39923, + "grad_norm": 1.6604739508614401, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39923 + }, + { + "epoch": 0.39924, + "grad_norm": 1.6737703372513903, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 39924 + }, + { + "epoch": 0.39925, + "grad_norm": 1.4262983267903682, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39925 + }, + { + "epoch": 0.39926, + "grad_norm": 1.4648522869356575, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 39926 + }, + { + "epoch": 0.39927, + "grad_norm": 1.5483731465325858, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 39927 + }, + { + "epoch": 0.39928, + "grad_norm": 1.5503158771291246, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 39928 + }, + { + "epoch": 0.39929, + "grad_norm": 1.6391187472218007, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39929 + }, + { + "epoch": 0.3993, + "grad_norm": 1.6944440710387358, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 39930 + }, + { + "epoch": 0.39931, + "grad_norm": 1.2935656737067198, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39931 + }, + { + "epoch": 0.39932, + "grad_norm": 1.5435071888118677, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 39932 + }, + { + "epoch": 0.39933, + "grad_norm": 1.4845584925242246, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 39933 + }, + { + "epoch": 0.39934, + "grad_norm": 1.611478959096227, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 39934 + }, + { + "epoch": 0.39935, + "grad_norm": 1.5065475956260526, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 39935 + }, + { + "epoch": 0.39936, + "grad_norm": 1.6890795855781213, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 39936 + }, + { + "epoch": 0.39937, + "grad_norm": 1.505094288345835, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 39937 + }, + { + "epoch": 0.39938, + "grad_norm": 1.733917906946783, + "learning_rate": 0.003, + "loss": 4.046, + "step": 39938 + }, + { + "epoch": 0.39939, + "grad_norm": 1.7352781932050119, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 39939 + }, + { + "epoch": 0.3994, + "grad_norm": 1.4016802024570394, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 39940 + }, + { + "epoch": 0.39941, + "grad_norm": 1.4242761795014973, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 39941 + }, + { + "epoch": 0.39942, + "grad_norm": 1.5973721818019513, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 39942 + }, + { + "epoch": 0.39943, + "grad_norm": 1.4018917777032691, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 39943 + }, + { + "epoch": 0.39944, + "grad_norm": 1.3888211877774155, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 39944 + }, + { + "epoch": 0.39945, + "grad_norm": 1.691368722231022, + "learning_rate": 0.003, + "loss": 4.027, + "step": 39945 + }, + { + "epoch": 0.39946, + "grad_norm": 1.415429391165691, + "learning_rate": 0.003, + "loss": 4.036, + "step": 39946 + }, + { + "epoch": 0.39947, + "grad_norm": 1.5091397475738009, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 39947 + }, + { + "epoch": 0.39948, + "grad_norm": 1.1899172674453034, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 39948 + }, + { + "epoch": 0.39949, + "grad_norm": 1.8079195226550109, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 39949 + }, + { + "epoch": 0.3995, + "grad_norm": 1.3800649855596747, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 39950 + }, + { + "epoch": 0.39951, + "grad_norm": 1.6126892562941928, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 39951 + }, + { + "epoch": 0.39952, + "grad_norm": 1.4870024832027489, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39952 + }, + { + "epoch": 0.39953, + "grad_norm": 1.564822940228039, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39953 + }, + { + "epoch": 0.39954, + "grad_norm": 1.533900095722206, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 39954 + }, + { + "epoch": 0.39955, + "grad_norm": 1.7394995786468228, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39955 + }, + { + "epoch": 0.39956, + "grad_norm": 1.233682608426138, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 39956 + }, + { + "epoch": 0.39957, + "grad_norm": 1.7036446056856702, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 39957 + }, + { + "epoch": 0.39958, + "grad_norm": 1.6739888741488818, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 39958 + }, + { + "epoch": 0.39959, + "grad_norm": 1.5598528543861168, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 39959 + }, + { + "epoch": 0.3996, + "grad_norm": 1.4666032368887196, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 39960 + }, + { + "epoch": 0.39961, + "grad_norm": 1.62236493538615, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 39961 + }, + { + "epoch": 0.39962, + "grad_norm": 1.478237722747433, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 39962 + }, + { + "epoch": 0.39963, + "grad_norm": 1.3996989932707757, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 39963 + }, + { + "epoch": 0.39964, + "grad_norm": 1.555932050924877, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 39964 + }, + { + "epoch": 0.39965, + "grad_norm": 1.5400345545053518, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 39965 + }, + { + "epoch": 0.39966, + "grad_norm": 1.6753920992745581, + "learning_rate": 0.003, + "loss": 4.033, + "step": 39966 + }, + { + "epoch": 0.39967, + "grad_norm": 1.5915389834088245, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 39967 + }, + { + "epoch": 0.39968, + "grad_norm": 1.4410491174412505, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 39968 + }, + { + "epoch": 0.39969, + "grad_norm": 1.4135016978502783, + "learning_rate": 0.003, + "loss": 4.038, + "step": 39969 + }, + { + "epoch": 0.3997, + "grad_norm": 1.6264280471872972, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 39970 + }, + { + "epoch": 0.39971, + "grad_norm": 1.346114349690583, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 39971 + }, + { + "epoch": 0.39972, + "grad_norm": 1.9705667195669636, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39972 + }, + { + "epoch": 0.39973, + "grad_norm": 1.3476490929011347, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 39973 + }, + { + "epoch": 0.39974, + "grad_norm": 1.8283769519092608, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 39974 + }, + { + "epoch": 0.39975, + "grad_norm": 1.3890974754556584, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 39975 + }, + { + "epoch": 0.39976, + "grad_norm": 1.6344857163161342, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 39976 + }, + { + "epoch": 0.39977, + "grad_norm": 1.5487705963234952, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 39977 + }, + { + "epoch": 0.39978, + "grad_norm": 1.4895551301181134, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 39978 + }, + { + "epoch": 0.39979, + "grad_norm": 1.5661012090767874, + "learning_rate": 0.003, + "loss": 4.044, + "step": 39979 + }, + { + "epoch": 0.3998, + "grad_norm": 1.3888905504103808, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 39980 + }, + { + "epoch": 0.39981, + "grad_norm": 1.5910410702767972, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 39981 + }, + { + "epoch": 0.39982, + "grad_norm": 1.5476054778206147, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 39982 + }, + { + "epoch": 0.39983, + "grad_norm": 1.3954110223085596, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 39983 + }, + { + "epoch": 0.39984, + "grad_norm": 1.6946482720685199, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 39984 + }, + { + "epoch": 0.39985, + "grad_norm": 1.4077919982294045, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 39985 + }, + { + "epoch": 0.39986, + "grad_norm": 1.443844333437085, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 39986 + }, + { + "epoch": 0.39987, + "grad_norm": 1.4140901113489104, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 39987 + }, + { + "epoch": 0.39988, + "grad_norm": 1.5768009427986913, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 39988 + }, + { + "epoch": 0.39989, + "grad_norm": 1.669616125778035, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 39989 + }, + { + "epoch": 0.3999, + "grad_norm": 1.5286561992038108, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 39990 + }, + { + "epoch": 0.39991, + "grad_norm": 1.6378504195701056, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 39991 + }, + { + "epoch": 0.39992, + "grad_norm": 1.5259248959672356, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 39992 + }, + { + "epoch": 0.39993, + "grad_norm": 1.6777614021669425, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 39993 + }, + { + "epoch": 0.39994, + "grad_norm": 1.6150009783701877, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 39994 + }, + { + "epoch": 0.39995, + "grad_norm": 1.4028821038434574, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 39995 + }, + { + "epoch": 0.39996, + "grad_norm": 1.6715024636722164, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 39996 + }, + { + "epoch": 0.39997, + "grad_norm": 1.778592406324181, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 39997 + }, + { + "epoch": 0.39998, + "grad_norm": 1.3129874927900669, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 39998 + }, + { + "epoch": 0.39999, + "grad_norm": 1.8463101173987457, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 39999 + }, + { + "epoch": 0.4, + "grad_norm": 1.23937433573568, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 40000 + }, + { + "epoch": 0.40001, + "grad_norm": 1.8016146244777353, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 40001 + }, + { + "epoch": 0.40002, + "grad_norm": 1.3634119538505767, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 40002 + }, + { + "epoch": 0.40003, + "grad_norm": 1.830687704007794, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 40003 + }, + { + "epoch": 0.40004, + "grad_norm": 1.4831244080784103, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40004 + }, + { + "epoch": 0.40005, + "grad_norm": 1.6353233268502645, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 40005 + }, + { + "epoch": 0.40006, + "grad_norm": 1.441529254704694, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 40006 + }, + { + "epoch": 0.40007, + "grad_norm": 1.7184142657947996, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 40007 + }, + { + "epoch": 0.40008, + "grad_norm": 1.3752260517698611, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40008 + }, + { + "epoch": 0.40009, + "grad_norm": 1.2173714148271841, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 40009 + }, + { + "epoch": 0.4001, + "grad_norm": 1.8543767659867616, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 40010 + }, + { + "epoch": 0.40011, + "grad_norm": 1.1104616725713912, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 40011 + }, + { + "epoch": 0.40012, + "grad_norm": 1.8526995314466403, + "learning_rate": 0.003, + "loss": 4.034, + "step": 40012 + }, + { + "epoch": 0.40013, + "grad_norm": 1.5204222671288514, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 40013 + }, + { + "epoch": 0.40014, + "grad_norm": 1.7507612146759892, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 40014 + }, + { + "epoch": 0.40015, + "grad_norm": 1.544743521859538, + "learning_rate": 0.003, + "loss": 4.023, + "step": 40015 + }, + { + "epoch": 0.40016, + "grad_norm": 1.5027815901257424, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40016 + }, + { + "epoch": 0.40017, + "grad_norm": 1.4881752667666532, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 40017 + }, + { + "epoch": 0.40018, + "grad_norm": 1.3757641149649977, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40018 + }, + { + "epoch": 0.40019, + "grad_norm": 1.3748298099434646, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40019 + }, + { + "epoch": 0.4002, + "grad_norm": 1.5010559362513494, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40020 + }, + { + "epoch": 0.40021, + "grad_norm": 1.6232049961925235, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 40021 + }, + { + "epoch": 0.40022, + "grad_norm": 1.5105311282330578, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 40022 + }, + { + "epoch": 0.40023, + "grad_norm": 1.4515336882861982, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 40023 + }, + { + "epoch": 0.40024, + "grad_norm": 1.5871012067406887, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 40024 + }, + { + "epoch": 0.40025, + "grad_norm": 1.702430076575302, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 40025 + }, + { + "epoch": 0.40026, + "grad_norm": 1.3357898021968588, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 40026 + }, + { + "epoch": 0.40027, + "grad_norm": 1.738777755020147, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 40027 + }, + { + "epoch": 0.40028, + "grad_norm": 1.504396837683168, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40028 + }, + { + "epoch": 0.40029, + "grad_norm": 1.4048176266889374, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 40029 + }, + { + "epoch": 0.4003, + "grad_norm": 1.7947082153186449, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 40030 + }, + { + "epoch": 0.40031, + "grad_norm": 1.44325522739487, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 40031 + }, + { + "epoch": 0.40032, + "grad_norm": 1.4500748243182853, + "learning_rate": 0.003, + "loss": 4.055, + "step": 40032 + }, + { + "epoch": 0.40033, + "grad_norm": 1.3345221402942382, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 40033 + }, + { + "epoch": 0.40034, + "grad_norm": 1.4386683064163426, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40034 + }, + { + "epoch": 0.40035, + "grad_norm": 1.5547936613006053, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 40035 + }, + { + "epoch": 0.40036, + "grad_norm": 1.6355308566578397, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 40036 + }, + { + "epoch": 0.40037, + "grad_norm": 1.6621347640853885, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 40037 + }, + { + "epoch": 0.40038, + "grad_norm": 1.5992003396582557, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 40038 + }, + { + "epoch": 0.40039, + "grad_norm": 1.3063578673407694, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 40039 + }, + { + "epoch": 0.4004, + "grad_norm": 1.7287412599754655, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 40040 + }, + { + "epoch": 0.40041, + "grad_norm": 1.2598849091126654, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40041 + }, + { + "epoch": 0.40042, + "grad_norm": 1.67009605973195, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 40042 + }, + { + "epoch": 0.40043, + "grad_norm": 1.4896097674586741, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40043 + }, + { + "epoch": 0.40044, + "grad_norm": 1.806878263929135, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40044 + }, + { + "epoch": 0.40045, + "grad_norm": 1.2974001578813648, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 40045 + }, + { + "epoch": 0.40046, + "grad_norm": 1.5849674991811584, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40046 + }, + { + "epoch": 0.40047, + "grad_norm": 1.7194439869016536, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 40047 + }, + { + "epoch": 0.40048, + "grad_norm": 1.333112091736035, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40048 + }, + { + "epoch": 0.40049, + "grad_norm": 1.8356241438621712, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40049 + }, + { + "epoch": 0.4005, + "grad_norm": 1.624700261749295, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40050 + }, + { + "epoch": 0.40051, + "grad_norm": 1.3733403033084894, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 40051 + }, + { + "epoch": 0.40052, + "grad_norm": 1.5925598312287836, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 40052 + }, + { + "epoch": 0.40053, + "grad_norm": 1.4485965266532552, + "learning_rate": 0.003, + "loss": 4.018, + "step": 40053 + }, + { + "epoch": 0.40054, + "grad_norm": 1.4981343081645746, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 40054 + }, + { + "epoch": 0.40055, + "grad_norm": 1.6343076606375107, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40055 + }, + { + "epoch": 0.40056, + "grad_norm": 1.4703164894356602, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 40056 + }, + { + "epoch": 0.40057, + "grad_norm": 1.378526877518949, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40057 + }, + { + "epoch": 0.40058, + "grad_norm": 1.398142696823653, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 40058 + }, + { + "epoch": 0.40059, + "grad_norm": 1.2723956901602216, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40059 + }, + { + "epoch": 0.4006, + "grad_norm": 1.8442423609184473, + "learning_rate": 0.003, + "loss": 4.044, + "step": 40060 + }, + { + "epoch": 0.40061, + "grad_norm": 1.3539605529382521, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 40061 + }, + { + "epoch": 0.40062, + "grad_norm": 1.9256868049988904, + "learning_rate": 0.003, + "loss": 4.036, + "step": 40062 + }, + { + "epoch": 0.40063, + "grad_norm": 1.508816834044207, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 40063 + }, + { + "epoch": 0.40064, + "grad_norm": 1.6162744528091835, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40064 + }, + { + "epoch": 0.40065, + "grad_norm": 1.4708318342053808, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40065 + }, + { + "epoch": 0.40066, + "grad_norm": 1.4302299916612475, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 40066 + }, + { + "epoch": 0.40067, + "grad_norm": 1.9623464690365278, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40067 + }, + { + "epoch": 0.40068, + "grad_norm": 1.6835682412716801, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 40068 + }, + { + "epoch": 0.40069, + "grad_norm": 1.264580391119604, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 40069 + }, + { + "epoch": 0.4007, + "grad_norm": 1.6873593865634633, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 40070 + }, + { + "epoch": 0.40071, + "grad_norm": 1.3888764213722917, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40071 + }, + { + "epoch": 0.40072, + "grad_norm": 1.4988198761867486, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 40072 + }, + { + "epoch": 0.40073, + "grad_norm": 1.5556945458165063, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 40073 + }, + { + "epoch": 0.40074, + "grad_norm": 1.8594900892038442, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40074 + }, + { + "epoch": 0.40075, + "grad_norm": 1.392995383828197, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40075 + }, + { + "epoch": 0.40076, + "grad_norm": 1.9508009104912782, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 40076 + }, + { + "epoch": 0.40077, + "grad_norm": 1.2180682484597547, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 40077 + }, + { + "epoch": 0.40078, + "grad_norm": 1.6445943057434795, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 40078 + }, + { + "epoch": 0.40079, + "grad_norm": 1.7637552302978572, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 40079 + }, + { + "epoch": 0.4008, + "grad_norm": 1.8658337127624873, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 40080 + }, + { + "epoch": 0.40081, + "grad_norm": 1.4561529425369768, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 40081 + }, + { + "epoch": 0.40082, + "grad_norm": 1.5412586876637528, + "learning_rate": 0.003, + "loss": 4.03, + "step": 40082 + }, + { + "epoch": 0.40083, + "grad_norm": 1.7375508449182466, + "learning_rate": 0.003, + "loss": 4.065, + "step": 40083 + }, + { + "epoch": 0.40084, + "grad_norm": 1.2194607142790055, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40084 + }, + { + "epoch": 0.40085, + "grad_norm": 1.4921723969220557, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 40085 + }, + { + "epoch": 0.40086, + "grad_norm": 1.3604108340362655, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 40086 + }, + { + "epoch": 0.40087, + "grad_norm": 1.6023705033579518, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 40087 + }, + { + "epoch": 0.40088, + "grad_norm": 1.5058272574418479, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 40088 + }, + { + "epoch": 0.40089, + "grad_norm": 1.4554585422134958, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40089 + }, + { + "epoch": 0.4009, + "grad_norm": 1.558277981602594, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 40090 + }, + { + "epoch": 0.40091, + "grad_norm": 1.3422160197095705, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 40091 + }, + { + "epoch": 0.40092, + "grad_norm": 1.4982726957435357, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 40092 + }, + { + "epoch": 0.40093, + "grad_norm": 1.4735970729210495, + "learning_rate": 0.003, + "loss": 4.049, + "step": 40093 + }, + { + "epoch": 0.40094, + "grad_norm": 1.8909814083934542, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 40094 + }, + { + "epoch": 0.40095, + "grad_norm": 1.1828110189803496, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 40095 + }, + { + "epoch": 0.40096, + "grad_norm": 1.730677611227787, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 40096 + }, + { + "epoch": 0.40097, + "grad_norm": 1.6901491928810448, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40097 + }, + { + "epoch": 0.40098, + "grad_norm": 1.680161747056, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 40098 + }, + { + "epoch": 0.40099, + "grad_norm": 1.4118979876856368, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40099 + }, + { + "epoch": 0.401, + "grad_norm": 1.5309707562512629, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 40100 + }, + { + "epoch": 0.40101, + "grad_norm": 1.6043550843316756, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40101 + }, + { + "epoch": 0.40102, + "grad_norm": 1.5846110376677607, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 40102 + }, + { + "epoch": 0.40103, + "grad_norm": 1.4850945151438566, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 40103 + }, + { + "epoch": 0.40104, + "grad_norm": 1.5029516523464295, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40104 + }, + { + "epoch": 0.40105, + "grad_norm": 1.7732545818997056, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40105 + }, + { + "epoch": 0.40106, + "grad_norm": 1.4694308386669577, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 40106 + }, + { + "epoch": 0.40107, + "grad_norm": 1.4116045074986618, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 40107 + }, + { + "epoch": 0.40108, + "grad_norm": 1.6589988347510658, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 40108 + }, + { + "epoch": 0.40109, + "grad_norm": 1.3363791026035738, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 40109 + }, + { + "epoch": 0.4011, + "grad_norm": 1.7192665617685305, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 40110 + }, + { + "epoch": 0.40111, + "grad_norm": 1.4376092194264067, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40111 + }, + { + "epoch": 0.40112, + "grad_norm": 1.4738752982202776, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 40112 + }, + { + "epoch": 0.40113, + "grad_norm": 1.4280293361429937, + "learning_rate": 0.003, + "loss": 4.022, + "step": 40113 + }, + { + "epoch": 0.40114, + "grad_norm": 1.4735022880884208, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 40114 + }, + { + "epoch": 0.40115, + "grad_norm": 1.4705057261482697, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 40115 + }, + { + "epoch": 0.40116, + "grad_norm": 1.475847226016636, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 40116 + }, + { + "epoch": 0.40117, + "grad_norm": 1.529302812805843, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40117 + }, + { + "epoch": 0.40118, + "grad_norm": 1.5404866659010619, + "learning_rate": 0.003, + "loss": 4.059, + "step": 40118 + }, + { + "epoch": 0.40119, + "grad_norm": 1.503796248676906, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 40119 + }, + { + "epoch": 0.4012, + "grad_norm": 1.904711350186929, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 40120 + }, + { + "epoch": 0.40121, + "grad_norm": 1.3975267874922026, + "learning_rate": 0.003, + "loss": 4.043, + "step": 40121 + }, + { + "epoch": 0.40122, + "grad_norm": 1.535793503100225, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 40122 + }, + { + "epoch": 0.40123, + "grad_norm": 1.5508762303958297, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40123 + }, + { + "epoch": 0.40124, + "grad_norm": 1.8640463220949552, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40124 + }, + { + "epoch": 0.40125, + "grad_norm": 1.3625834049790235, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 40125 + }, + { + "epoch": 0.40126, + "grad_norm": 1.3811424922627256, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 40126 + }, + { + "epoch": 0.40127, + "grad_norm": 1.7139504610787506, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 40127 + }, + { + "epoch": 0.40128, + "grad_norm": 1.4436827565085795, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 40128 + }, + { + "epoch": 0.40129, + "grad_norm": 1.6280276848494224, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 40129 + }, + { + "epoch": 0.4013, + "grad_norm": 1.46719334214457, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 40130 + }, + { + "epoch": 0.40131, + "grad_norm": 1.5268733548878424, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 40131 + }, + { + "epoch": 0.40132, + "grad_norm": 1.448409070380694, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 40132 + }, + { + "epoch": 0.40133, + "grad_norm": 2.046668551454872, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 40133 + }, + { + "epoch": 0.40134, + "grad_norm": 1.4368021585775455, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 40134 + }, + { + "epoch": 0.40135, + "grad_norm": 1.7965421148917238, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 40135 + }, + { + "epoch": 0.40136, + "grad_norm": 1.3326687547911242, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40136 + }, + { + "epoch": 0.40137, + "grad_norm": 1.3863780704243025, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 40137 + }, + { + "epoch": 0.40138, + "grad_norm": 1.463249917744524, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40138 + }, + { + "epoch": 0.40139, + "grad_norm": 1.5973967268247606, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40139 + }, + { + "epoch": 0.4014, + "grad_norm": 1.289181807886211, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 40140 + }, + { + "epoch": 0.40141, + "grad_norm": 1.8034895908191386, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 40141 + }, + { + "epoch": 0.40142, + "grad_norm": 1.4264659882844648, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 40142 + }, + { + "epoch": 0.40143, + "grad_norm": 1.3097833746221854, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40143 + }, + { + "epoch": 0.40144, + "grad_norm": 1.7365122753446642, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 40144 + }, + { + "epoch": 0.40145, + "grad_norm": 1.3817012638576058, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 40145 + }, + { + "epoch": 0.40146, + "grad_norm": 1.8691686669715086, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40146 + }, + { + "epoch": 0.40147, + "grad_norm": 1.5156940692266845, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40147 + }, + { + "epoch": 0.40148, + "grad_norm": 1.52172803850808, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40148 + }, + { + "epoch": 0.40149, + "grad_norm": 1.5265903926376592, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 40149 + }, + { + "epoch": 0.4015, + "grad_norm": 1.4213676223887586, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 40150 + }, + { + "epoch": 0.40151, + "grad_norm": 1.53677454084804, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 40151 + }, + { + "epoch": 0.40152, + "grad_norm": 1.3797369846373513, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40152 + }, + { + "epoch": 0.40153, + "grad_norm": 1.8876811087886507, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 40153 + }, + { + "epoch": 0.40154, + "grad_norm": 1.495155246151637, + "learning_rate": 0.003, + "loss": 4.026, + "step": 40154 + }, + { + "epoch": 0.40155, + "grad_norm": 1.4597173088724946, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 40155 + }, + { + "epoch": 0.40156, + "grad_norm": 1.536241972932325, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 40156 + }, + { + "epoch": 0.40157, + "grad_norm": 1.7050940956674592, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40157 + }, + { + "epoch": 0.40158, + "grad_norm": 1.526994020807725, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 40158 + }, + { + "epoch": 0.40159, + "grad_norm": 1.43564457393086, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 40159 + }, + { + "epoch": 0.4016, + "grad_norm": 1.3784455082205957, + "learning_rate": 0.003, + "loss": 4.007, + "step": 40160 + }, + { + "epoch": 0.40161, + "grad_norm": 1.3507494570789162, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40161 + }, + { + "epoch": 0.40162, + "grad_norm": 1.617310063726163, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 40162 + }, + { + "epoch": 0.40163, + "grad_norm": 1.4031632925932602, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40163 + }, + { + "epoch": 0.40164, + "grad_norm": 1.6138344870079153, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 40164 + }, + { + "epoch": 0.40165, + "grad_norm": 1.2291109183974205, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40165 + }, + { + "epoch": 0.40166, + "grad_norm": 1.9977075315177764, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 40166 + }, + { + "epoch": 0.40167, + "grad_norm": 1.0580062755250377, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 40167 + }, + { + "epoch": 0.40168, + "grad_norm": 1.7401292237292096, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 40168 + }, + { + "epoch": 0.40169, + "grad_norm": 2.0369348649893184, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 40169 + }, + { + "epoch": 0.4017, + "grad_norm": 1.330630498982732, + "learning_rate": 0.003, + "loss": 4.03, + "step": 40170 + }, + { + "epoch": 0.40171, + "grad_norm": 1.5881431955695076, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40171 + }, + { + "epoch": 0.40172, + "grad_norm": 1.7152448080434644, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 40172 + }, + { + "epoch": 0.40173, + "grad_norm": 1.1991917994008323, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 40173 + }, + { + "epoch": 0.40174, + "grad_norm": 1.7805971893657988, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 40174 + }, + { + "epoch": 0.40175, + "grad_norm": 1.2017811638026847, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 40175 + }, + { + "epoch": 0.40176, + "grad_norm": 1.9532273839143606, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 40176 + }, + { + "epoch": 0.40177, + "grad_norm": 1.2937119261087815, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 40177 + }, + { + "epoch": 0.40178, + "grad_norm": 1.9014246970347046, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 40178 + }, + { + "epoch": 0.40179, + "grad_norm": 1.3572042561356323, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 40179 + }, + { + "epoch": 0.4018, + "grad_norm": 2.05482430483293, + "learning_rate": 0.003, + "loss": 4.061, + "step": 40180 + }, + { + "epoch": 0.40181, + "grad_norm": 1.5192113217927123, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40181 + }, + { + "epoch": 0.40182, + "grad_norm": 1.532480715763363, + "learning_rate": 0.003, + "loss": 4.0971, + "step": 40182 + }, + { + "epoch": 0.40183, + "grad_norm": 1.629055591120936, + "learning_rate": 0.003, + "loss": 4.096, + "step": 40183 + }, + { + "epoch": 0.40184, + "grad_norm": 1.5587469635062545, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 40184 + }, + { + "epoch": 0.40185, + "grad_norm": 1.253731030929095, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 40185 + }, + { + "epoch": 0.40186, + "grad_norm": 1.5119886023800646, + "learning_rate": 0.003, + "loss": 4.038, + "step": 40186 + }, + { + "epoch": 0.40187, + "grad_norm": 1.5992422605611512, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 40187 + }, + { + "epoch": 0.40188, + "grad_norm": 1.752490588739094, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 40188 + }, + { + "epoch": 0.40189, + "grad_norm": 1.4392567357547688, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 40189 + }, + { + "epoch": 0.4019, + "grad_norm": 1.729683472983918, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 40190 + }, + { + "epoch": 0.40191, + "grad_norm": 1.3300647216758923, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 40191 + }, + { + "epoch": 0.40192, + "grad_norm": 1.7167910634306256, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 40192 + }, + { + "epoch": 0.40193, + "grad_norm": 1.5147146475248863, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 40193 + }, + { + "epoch": 0.40194, + "grad_norm": 1.420178223790627, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40194 + }, + { + "epoch": 0.40195, + "grad_norm": 1.7393904146429453, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 40195 + }, + { + "epoch": 0.40196, + "grad_norm": 1.1866696427828738, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40196 + }, + { + "epoch": 0.40197, + "grad_norm": 1.731971180396561, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 40197 + }, + { + "epoch": 0.40198, + "grad_norm": 1.2480205826022965, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 40198 + }, + { + "epoch": 0.40199, + "grad_norm": 1.8042821216435623, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40199 + }, + { + "epoch": 0.402, + "grad_norm": 1.3932951202420867, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 40200 + }, + { + "epoch": 0.40201, + "grad_norm": 2.1437081587208273, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 40201 + }, + { + "epoch": 0.40202, + "grad_norm": 1.0620021224447418, + "learning_rate": 0.003, + "loss": 4.005, + "step": 40202 + }, + { + "epoch": 0.40203, + "grad_norm": 1.6562545233760468, + "learning_rate": 0.003, + "loss": 4.047, + "step": 40203 + }, + { + "epoch": 0.40204, + "grad_norm": 1.5896686217634715, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 40204 + }, + { + "epoch": 0.40205, + "grad_norm": 1.8518952335686127, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 40205 + }, + { + "epoch": 0.40206, + "grad_norm": 1.0745648886610757, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40206 + }, + { + "epoch": 0.40207, + "grad_norm": 1.764133024014595, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 40207 + }, + { + "epoch": 0.40208, + "grad_norm": 1.4867634278576676, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 40208 + }, + { + "epoch": 0.40209, + "grad_norm": 1.1412940007679344, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40209 + }, + { + "epoch": 0.4021, + "grad_norm": 1.8277852787957845, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 40210 + }, + { + "epoch": 0.40211, + "grad_norm": 1.3578851960663925, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40211 + }, + { + "epoch": 0.40212, + "grad_norm": 1.846225086381238, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 40212 + }, + { + "epoch": 0.40213, + "grad_norm": 1.6569512768556238, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 40213 + }, + { + "epoch": 0.40214, + "grad_norm": 1.9994420702830593, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 40214 + }, + { + "epoch": 0.40215, + "grad_norm": 1.094916758825337, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 40215 + }, + { + "epoch": 0.40216, + "grad_norm": 1.681426100898334, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 40216 + }, + { + "epoch": 0.40217, + "grad_norm": 1.9064119959484291, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 40217 + }, + { + "epoch": 0.40218, + "grad_norm": 1.670107553428128, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 40218 + }, + { + "epoch": 0.40219, + "grad_norm": 1.4847632306614627, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 40219 + }, + { + "epoch": 0.4022, + "grad_norm": 1.529391770734477, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 40220 + }, + { + "epoch": 0.40221, + "grad_norm": 1.6275245124310322, + "learning_rate": 0.003, + "loss": 4.049, + "step": 40221 + }, + { + "epoch": 0.40222, + "grad_norm": 1.341797476485267, + "learning_rate": 0.003, + "loss": 4.026, + "step": 40222 + }, + { + "epoch": 0.40223, + "grad_norm": 1.5362756780289772, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 40223 + }, + { + "epoch": 0.40224, + "grad_norm": 1.7472452506860983, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40224 + }, + { + "epoch": 0.40225, + "grad_norm": 1.2894166026685436, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 40225 + }, + { + "epoch": 0.40226, + "grad_norm": 1.5193416758420237, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 40226 + }, + { + "epoch": 0.40227, + "grad_norm": 1.5823128500453172, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 40227 + }, + { + "epoch": 0.40228, + "grad_norm": 1.3069855186414532, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 40228 + }, + { + "epoch": 0.40229, + "grad_norm": 1.6043131091947518, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 40229 + }, + { + "epoch": 0.4023, + "grad_norm": 1.2341145550950463, + "learning_rate": 0.003, + "loss": 4.025, + "step": 40230 + }, + { + "epoch": 0.40231, + "grad_norm": 1.8975844200540484, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 40231 + }, + { + "epoch": 0.40232, + "grad_norm": 1.6941772310559011, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 40232 + }, + { + "epoch": 0.40233, + "grad_norm": 1.1781936975893992, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 40233 + }, + { + "epoch": 0.40234, + "grad_norm": 1.6362192655849417, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 40234 + }, + { + "epoch": 0.40235, + "grad_norm": 1.5649201385114055, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40235 + }, + { + "epoch": 0.40236, + "grad_norm": 1.2370299039199513, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40236 + }, + { + "epoch": 0.40237, + "grad_norm": 1.6608742704239405, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 40237 + }, + { + "epoch": 0.40238, + "grad_norm": 1.4840763996562156, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 40238 + }, + { + "epoch": 0.40239, + "grad_norm": 1.5418164405650578, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40239 + }, + { + "epoch": 0.4024, + "grad_norm": 1.5275305116643385, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40240 + }, + { + "epoch": 0.40241, + "grad_norm": 1.722216069266768, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 40241 + }, + { + "epoch": 0.40242, + "grad_norm": 1.3011256067933104, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 40242 + }, + { + "epoch": 0.40243, + "grad_norm": 1.5850967664037992, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40243 + }, + { + "epoch": 0.40244, + "grad_norm": 1.302043981153893, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 40244 + }, + { + "epoch": 0.40245, + "grad_norm": 1.4485568727169222, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 40245 + }, + { + "epoch": 0.40246, + "grad_norm": 1.5107556529506412, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40246 + }, + { + "epoch": 0.40247, + "grad_norm": 1.8625738970752188, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40247 + }, + { + "epoch": 0.40248, + "grad_norm": 1.4237303023585348, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 40248 + }, + { + "epoch": 0.40249, + "grad_norm": 1.6589739951656723, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 40249 + }, + { + "epoch": 0.4025, + "grad_norm": 1.5720369790734583, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 40250 + }, + { + "epoch": 0.40251, + "grad_norm": 1.3183456418623771, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40251 + }, + { + "epoch": 0.40252, + "grad_norm": 1.6442034525530178, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 40252 + }, + { + "epoch": 0.40253, + "grad_norm": 1.635258723508028, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40253 + }, + { + "epoch": 0.40254, + "grad_norm": 1.5567406619491067, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 40254 + }, + { + "epoch": 0.40255, + "grad_norm": 1.5567001756355618, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40255 + }, + { + "epoch": 0.40256, + "grad_norm": 1.4054065450576363, + "learning_rate": 0.003, + "loss": 4.023, + "step": 40256 + }, + { + "epoch": 0.40257, + "grad_norm": 1.6285973712112192, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40257 + }, + { + "epoch": 0.40258, + "grad_norm": 1.3240444708108203, + "learning_rate": 0.003, + "loss": 4.065, + "step": 40258 + }, + { + "epoch": 0.40259, + "grad_norm": 1.884658701000172, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40259 + }, + { + "epoch": 0.4026, + "grad_norm": 1.3602474566975453, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40260 + }, + { + "epoch": 0.40261, + "grad_norm": 1.397003970119708, + "learning_rate": 0.003, + "loss": 4.033, + "step": 40261 + }, + { + "epoch": 0.40262, + "grad_norm": 1.523948153721968, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 40262 + }, + { + "epoch": 0.40263, + "grad_norm": 1.3596441181816517, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 40263 + }, + { + "epoch": 0.40264, + "grad_norm": 2.0588327751502247, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 40264 + }, + { + "epoch": 0.40265, + "grad_norm": 1.313963532245039, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 40265 + }, + { + "epoch": 0.40266, + "grad_norm": 1.4111842258582585, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 40266 + }, + { + "epoch": 0.40267, + "grad_norm": 1.7054809307879617, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 40267 + }, + { + "epoch": 0.40268, + "grad_norm": 1.9683137096396954, + "learning_rate": 0.003, + "loss": 4.043, + "step": 40268 + }, + { + "epoch": 0.40269, + "grad_norm": 1.2246219643088245, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 40269 + }, + { + "epoch": 0.4027, + "grad_norm": 1.5608126262084634, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 40270 + }, + { + "epoch": 0.40271, + "grad_norm": 1.8620153006921096, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 40271 + }, + { + "epoch": 0.40272, + "grad_norm": 1.2571014038965411, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40272 + }, + { + "epoch": 0.40273, + "grad_norm": 1.6341144172199615, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 40273 + }, + { + "epoch": 0.40274, + "grad_norm": 1.4210646952793087, + "learning_rate": 0.003, + "loss": 4.022, + "step": 40274 + }, + { + "epoch": 0.40275, + "grad_norm": 1.859905258952808, + "learning_rate": 0.003, + "loss": 4.073, + "step": 40275 + }, + { + "epoch": 0.40276, + "grad_norm": 1.3434867725612658, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 40276 + }, + { + "epoch": 0.40277, + "grad_norm": 1.752951464537829, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 40277 + }, + { + "epoch": 0.40278, + "grad_norm": 1.7531458166287095, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 40278 + }, + { + "epoch": 0.40279, + "grad_norm": 1.3951464011679353, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 40279 + }, + { + "epoch": 0.4028, + "grad_norm": 1.3931020858418872, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 40280 + }, + { + "epoch": 0.40281, + "grad_norm": 1.76701631137413, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40281 + }, + { + "epoch": 0.40282, + "grad_norm": 1.4032566870852548, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 40282 + }, + { + "epoch": 0.40283, + "grad_norm": 1.4165792610585803, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 40283 + }, + { + "epoch": 0.40284, + "grad_norm": 1.44574522227696, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 40284 + }, + { + "epoch": 0.40285, + "grad_norm": 1.4013828097220413, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 40285 + }, + { + "epoch": 0.40286, + "grad_norm": 1.4859663146302855, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 40286 + }, + { + "epoch": 0.40287, + "grad_norm": 1.5157902945124626, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 40287 + }, + { + "epoch": 0.40288, + "grad_norm": 1.730732341855698, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 40288 + }, + { + "epoch": 0.40289, + "grad_norm": 1.508160588508863, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 40289 + }, + { + "epoch": 0.4029, + "grad_norm": 1.48305655821918, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 40290 + }, + { + "epoch": 0.40291, + "grad_norm": 1.603158122899763, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 40291 + }, + { + "epoch": 0.40292, + "grad_norm": 1.511403025444609, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40292 + }, + { + "epoch": 0.40293, + "grad_norm": 1.5952510049516637, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 40293 + }, + { + "epoch": 0.40294, + "grad_norm": 2.057364760143569, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 40294 + }, + { + "epoch": 0.40295, + "grad_norm": 1.4618830394013875, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 40295 + }, + { + "epoch": 0.40296, + "grad_norm": 1.5889689937761498, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 40296 + }, + { + "epoch": 0.40297, + "grad_norm": 1.6432094017686014, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40297 + }, + { + "epoch": 0.40298, + "grad_norm": 1.6187312119610109, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 40298 + }, + { + "epoch": 0.40299, + "grad_norm": 1.230627040679279, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 40299 + }, + { + "epoch": 0.403, + "grad_norm": 1.8851014199216622, + "learning_rate": 0.003, + "loss": 4.033, + "step": 40300 + }, + { + "epoch": 0.40301, + "grad_norm": 1.2221590971383542, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 40301 + }, + { + "epoch": 0.40302, + "grad_norm": 1.5107542596668573, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 40302 + }, + { + "epoch": 0.40303, + "grad_norm": 1.6133740974864816, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40303 + }, + { + "epoch": 0.40304, + "grad_norm": 1.5266689106998137, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 40304 + }, + { + "epoch": 0.40305, + "grad_norm": 2.1376546130255805, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40305 + }, + { + "epoch": 0.40306, + "grad_norm": 1.0087451167610784, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 40306 + }, + { + "epoch": 0.40307, + "grad_norm": 1.6625555615145793, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 40307 + }, + { + "epoch": 0.40308, + "grad_norm": 1.6550008953278148, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 40308 + }, + { + "epoch": 0.40309, + "grad_norm": 1.7270457683228226, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 40309 + }, + { + "epoch": 0.4031, + "grad_norm": 1.5173056531487372, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 40310 + }, + { + "epoch": 0.40311, + "grad_norm": 1.5280264568102682, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40311 + }, + { + "epoch": 0.40312, + "grad_norm": 1.4723607938330319, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 40312 + }, + { + "epoch": 0.40313, + "grad_norm": 1.3272461862453164, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40313 + }, + { + "epoch": 0.40314, + "grad_norm": 1.4324025992364062, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 40314 + }, + { + "epoch": 0.40315, + "grad_norm": 1.4017520665831544, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 40315 + }, + { + "epoch": 0.40316, + "grad_norm": 1.6417748395958947, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 40316 + }, + { + "epoch": 0.40317, + "grad_norm": 1.6960012446904646, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 40317 + }, + { + "epoch": 0.40318, + "grad_norm": 1.6978888384523312, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40318 + }, + { + "epoch": 0.40319, + "grad_norm": 1.4208533941006642, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 40319 + }, + { + "epoch": 0.4032, + "grad_norm": 1.3914688715711365, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 40320 + }, + { + "epoch": 0.40321, + "grad_norm": 1.6765334286251707, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 40321 + }, + { + "epoch": 0.40322, + "grad_norm": 1.285332671948275, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 40322 + }, + { + "epoch": 0.40323, + "grad_norm": 1.5758344558007906, + "learning_rate": 0.003, + "loss": 4.065, + "step": 40323 + }, + { + "epoch": 0.40324, + "grad_norm": 1.4188286610452994, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40324 + }, + { + "epoch": 0.40325, + "grad_norm": 1.5627552086006402, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40325 + }, + { + "epoch": 0.40326, + "grad_norm": 1.3962803541297608, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 40326 + }, + { + "epoch": 0.40327, + "grad_norm": 1.332972694714629, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 40327 + }, + { + "epoch": 0.40328, + "grad_norm": 1.5418402614423448, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 40328 + }, + { + "epoch": 0.40329, + "grad_norm": 1.8652678552248603, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 40329 + }, + { + "epoch": 0.4033, + "grad_norm": 1.9205861116750635, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 40330 + }, + { + "epoch": 0.40331, + "grad_norm": 1.5960028992065278, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40331 + }, + { + "epoch": 0.40332, + "grad_norm": 1.8319367129240134, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 40332 + }, + { + "epoch": 0.40333, + "grad_norm": 1.6272691854015413, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 40333 + }, + { + "epoch": 0.40334, + "grad_norm": 1.4380696996153561, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 40334 + }, + { + "epoch": 0.40335, + "grad_norm": 1.4520085690740065, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40335 + }, + { + "epoch": 0.40336, + "grad_norm": 1.5968711840155236, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40336 + }, + { + "epoch": 0.40337, + "grad_norm": 1.2822322734771963, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 40337 + }, + { + "epoch": 0.40338, + "grad_norm": 1.4222391140636939, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 40338 + }, + { + "epoch": 0.40339, + "grad_norm": 1.634091903469164, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 40339 + }, + { + "epoch": 0.4034, + "grad_norm": 1.560457943227519, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 40340 + }, + { + "epoch": 0.40341, + "grad_norm": 1.411852304590035, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 40341 + }, + { + "epoch": 0.40342, + "grad_norm": 1.5605228699066525, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40342 + }, + { + "epoch": 0.40343, + "grad_norm": 1.255081794780736, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 40343 + }, + { + "epoch": 0.40344, + "grad_norm": 1.5694313610115929, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40344 + }, + { + "epoch": 0.40345, + "grad_norm": 1.528941020923876, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 40345 + }, + { + "epoch": 0.40346, + "grad_norm": 1.4654658941542231, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 40346 + }, + { + "epoch": 0.40347, + "grad_norm": 1.6427511672670834, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 40347 + }, + { + "epoch": 0.40348, + "grad_norm": 1.341934276733711, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 40348 + }, + { + "epoch": 0.40349, + "grad_norm": 1.7358495717135856, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40349 + }, + { + "epoch": 0.4035, + "grad_norm": 1.4078673699586701, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40350 + }, + { + "epoch": 0.40351, + "grad_norm": 1.8840204725004208, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 40351 + }, + { + "epoch": 0.40352, + "grad_norm": 1.4969176304013776, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 40352 + }, + { + "epoch": 0.40353, + "grad_norm": 1.8822481234376858, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 40353 + }, + { + "epoch": 0.40354, + "grad_norm": 1.3367136155996882, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40354 + }, + { + "epoch": 0.40355, + "grad_norm": 1.5352229322241482, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 40355 + }, + { + "epoch": 0.40356, + "grad_norm": 1.3676740441359891, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40356 + }, + { + "epoch": 0.40357, + "grad_norm": 1.7085656370012614, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 40357 + }, + { + "epoch": 0.40358, + "grad_norm": 1.3559733356871613, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40358 + }, + { + "epoch": 0.40359, + "grad_norm": 1.7367080799156924, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 40359 + }, + { + "epoch": 0.4036, + "grad_norm": 1.5114862075828255, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 40360 + }, + { + "epoch": 0.40361, + "grad_norm": 1.4962240481822722, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 40361 + }, + { + "epoch": 0.40362, + "grad_norm": 1.3198859770339113, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40362 + }, + { + "epoch": 0.40363, + "grad_norm": 1.3859298448762283, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 40363 + }, + { + "epoch": 0.40364, + "grad_norm": 1.7194928160802598, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40364 + }, + { + "epoch": 0.40365, + "grad_norm": 1.882103023935956, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 40365 + }, + { + "epoch": 0.40366, + "grad_norm": 1.239708068523308, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 40366 + }, + { + "epoch": 0.40367, + "grad_norm": 1.4004908880675593, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 40367 + }, + { + "epoch": 0.40368, + "grad_norm": 1.7829592471604194, + "learning_rate": 0.003, + "loss": 4.052, + "step": 40368 + }, + { + "epoch": 0.40369, + "grad_norm": 1.6380748565685048, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40369 + }, + { + "epoch": 0.4037, + "grad_norm": 1.4251476020366058, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 40370 + }, + { + "epoch": 0.40371, + "grad_norm": 1.448934693925133, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40371 + }, + { + "epoch": 0.40372, + "grad_norm": 1.4114216454751047, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40372 + }, + { + "epoch": 0.40373, + "grad_norm": 1.386910109449574, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 40373 + }, + { + "epoch": 0.40374, + "grad_norm": 1.5631401680938133, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40374 + }, + { + "epoch": 0.40375, + "grad_norm": 1.5228342920178632, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 40375 + }, + { + "epoch": 0.40376, + "grad_norm": 1.4723462084101102, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 40376 + }, + { + "epoch": 0.40377, + "grad_norm": 1.7236797971021993, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 40377 + }, + { + "epoch": 0.40378, + "grad_norm": 1.149356565547808, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 40378 + }, + { + "epoch": 0.40379, + "grad_norm": 1.62226260040403, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 40379 + }, + { + "epoch": 0.4038, + "grad_norm": 1.4275551719873485, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 40380 + }, + { + "epoch": 0.40381, + "grad_norm": 1.70236860294372, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 40381 + }, + { + "epoch": 0.40382, + "grad_norm": 1.2963298538492936, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40382 + }, + { + "epoch": 0.40383, + "grad_norm": 1.5868507795555105, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 40383 + }, + { + "epoch": 0.40384, + "grad_norm": 1.4204657940623049, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40384 + }, + { + "epoch": 0.40385, + "grad_norm": 1.5211067018232314, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 40385 + }, + { + "epoch": 0.40386, + "grad_norm": 1.5647591213294412, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 40386 + }, + { + "epoch": 0.40387, + "grad_norm": 1.7280704781915504, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 40387 + }, + { + "epoch": 0.40388, + "grad_norm": 1.5496567033074513, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40388 + }, + { + "epoch": 0.40389, + "grad_norm": 1.836174545308055, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40389 + }, + { + "epoch": 0.4039, + "grad_norm": 1.544733137049979, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 40390 + }, + { + "epoch": 0.40391, + "grad_norm": 1.3153644663434307, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 40391 + }, + { + "epoch": 0.40392, + "grad_norm": 2.2017898495079087, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 40392 + }, + { + "epoch": 0.40393, + "grad_norm": 1.2224182787726854, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40393 + }, + { + "epoch": 0.40394, + "grad_norm": 1.8316669336636995, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40394 + }, + { + "epoch": 0.40395, + "grad_norm": 1.2432474684958215, + "learning_rate": 0.003, + "loss": 4.037, + "step": 40395 + }, + { + "epoch": 0.40396, + "grad_norm": 1.7644144794639691, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 40396 + }, + { + "epoch": 0.40397, + "grad_norm": 1.325069193258963, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40397 + }, + { + "epoch": 0.40398, + "grad_norm": 1.4089630173030117, + "learning_rate": 0.003, + "loss": 4.039, + "step": 40398 + }, + { + "epoch": 0.40399, + "grad_norm": 1.4859113818654568, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 40399 + }, + { + "epoch": 0.404, + "grad_norm": 1.6574259408960008, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 40400 + }, + { + "epoch": 0.40401, + "grad_norm": 1.4550575860861974, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40401 + }, + { + "epoch": 0.40402, + "grad_norm": 1.8372648396538134, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 40402 + }, + { + "epoch": 0.40403, + "grad_norm": 1.4045355857063389, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 40403 + }, + { + "epoch": 0.40404, + "grad_norm": 1.4801564925476185, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 40404 + }, + { + "epoch": 0.40405, + "grad_norm": 1.4278775090144797, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 40405 + }, + { + "epoch": 0.40406, + "grad_norm": 1.7134173843519758, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 40406 + }, + { + "epoch": 0.40407, + "grad_norm": 1.6313585708009346, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 40407 + }, + { + "epoch": 0.40408, + "grad_norm": 1.2925856040688344, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40408 + }, + { + "epoch": 0.40409, + "grad_norm": 1.6048521761541557, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40409 + }, + { + "epoch": 0.4041, + "grad_norm": 1.5357283294798594, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40410 + }, + { + "epoch": 0.40411, + "grad_norm": 1.4146318931791546, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40411 + }, + { + "epoch": 0.40412, + "grad_norm": 2.298512701610411, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 40412 + }, + { + "epoch": 0.40413, + "grad_norm": 1.8927849825643714, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 40413 + }, + { + "epoch": 0.40414, + "grad_norm": 1.478828975904898, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 40414 + }, + { + "epoch": 0.40415, + "grad_norm": 1.6368387891828582, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 40415 + }, + { + "epoch": 0.40416, + "grad_norm": 1.375759797833826, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 40416 + }, + { + "epoch": 0.40417, + "grad_norm": 2.5085222775797655, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 40417 + }, + { + "epoch": 0.40418, + "grad_norm": 2.547019321517509, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40418 + }, + { + "epoch": 0.40419, + "grad_norm": 2.240298422463635, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 40419 + }, + { + "epoch": 0.4042, + "grad_norm": 1.7453535980404822, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 40420 + }, + { + "epoch": 0.40421, + "grad_norm": 1.4116109219990043, + "learning_rate": 0.003, + "loss": 4.071, + "step": 40421 + }, + { + "epoch": 0.40422, + "grad_norm": 2.049175226301433, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 40422 + }, + { + "epoch": 0.40423, + "grad_norm": 1.5774047441221004, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 40423 + }, + { + "epoch": 0.40424, + "grad_norm": 1.8222734808232979, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 40424 + }, + { + "epoch": 0.40425, + "grad_norm": 1.5041343884373244, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 40425 + }, + { + "epoch": 0.40426, + "grad_norm": 1.5053124635914379, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 40426 + }, + { + "epoch": 0.40427, + "grad_norm": 1.6584048868495358, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 40427 + }, + { + "epoch": 0.40428, + "grad_norm": 1.8352910456482296, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 40428 + }, + { + "epoch": 0.40429, + "grad_norm": 1.910366860520186, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 40429 + }, + { + "epoch": 0.4043, + "grad_norm": 2.0974946928226164, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 40430 + }, + { + "epoch": 0.40431, + "grad_norm": 1.7275990995188693, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 40431 + }, + { + "epoch": 0.40432, + "grad_norm": 1.5665060083981435, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 40432 + }, + { + "epoch": 0.40433, + "grad_norm": 1.5041973455328137, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 40433 + }, + { + "epoch": 0.40434, + "grad_norm": 1.7183442067798569, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40434 + }, + { + "epoch": 0.40435, + "grad_norm": 1.463037332324173, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 40435 + }, + { + "epoch": 0.40436, + "grad_norm": 2.2084978165394147, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 40436 + }, + { + "epoch": 0.40437, + "grad_norm": 1.524611507949652, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 40437 + }, + { + "epoch": 0.40438, + "grad_norm": 1.6600011588463819, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 40438 + }, + { + "epoch": 0.40439, + "grad_norm": 1.5222161814695678, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 40439 + }, + { + "epoch": 0.4044, + "grad_norm": 1.6365891203880008, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 40440 + }, + { + "epoch": 0.40441, + "grad_norm": 1.5993679367713514, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40441 + }, + { + "epoch": 0.40442, + "grad_norm": 1.3971453226902681, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40442 + }, + { + "epoch": 0.40443, + "grad_norm": 1.8327417098236805, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40443 + }, + { + "epoch": 0.40444, + "grad_norm": 1.2655288147463968, + "learning_rate": 0.003, + "loss": 4.056, + "step": 40444 + }, + { + "epoch": 0.40445, + "grad_norm": 1.6753735126104565, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 40445 + }, + { + "epoch": 0.40446, + "grad_norm": 1.5410555561328587, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 40446 + }, + { + "epoch": 0.40447, + "grad_norm": 1.6127571272600634, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 40447 + }, + { + "epoch": 0.40448, + "grad_norm": 1.7787853128020565, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 40448 + }, + { + "epoch": 0.40449, + "grad_norm": 1.668380745429667, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 40449 + }, + { + "epoch": 0.4045, + "grad_norm": 1.8934500270164825, + "learning_rate": 0.003, + "loss": 4.087, + "step": 40450 + }, + { + "epoch": 0.40451, + "grad_norm": 1.4630518719241696, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 40451 + }, + { + "epoch": 0.40452, + "grad_norm": 1.6114840472548955, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 40452 + }, + { + "epoch": 0.40453, + "grad_norm": 1.3021428225062188, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 40453 + }, + { + "epoch": 0.40454, + "grad_norm": 1.59865877145538, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 40454 + }, + { + "epoch": 0.40455, + "grad_norm": 1.5126823525563127, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40455 + }, + { + "epoch": 0.40456, + "grad_norm": 1.7511515619970313, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40456 + }, + { + "epoch": 0.40457, + "grad_norm": 1.5987299924672826, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 40457 + }, + { + "epoch": 0.40458, + "grad_norm": 1.2802625167374526, + "learning_rate": 0.003, + "loss": 4.029, + "step": 40458 + }, + { + "epoch": 0.40459, + "grad_norm": 1.5493355913822242, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 40459 + }, + { + "epoch": 0.4046, + "grad_norm": 1.4843868493367582, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40460 + }, + { + "epoch": 0.40461, + "grad_norm": 1.6237745000174546, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 40461 + }, + { + "epoch": 0.40462, + "grad_norm": 1.4058170669531795, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 40462 + }, + { + "epoch": 0.40463, + "grad_norm": 1.4895891731794382, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40463 + }, + { + "epoch": 0.40464, + "grad_norm": 1.3942275592009925, + "learning_rate": 0.003, + "loss": 4.04, + "step": 40464 + }, + { + "epoch": 0.40465, + "grad_norm": 1.7169200839221777, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 40465 + }, + { + "epoch": 0.40466, + "grad_norm": 1.5770007489553202, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 40466 + }, + { + "epoch": 0.40467, + "grad_norm": 1.4550579591720452, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 40467 + }, + { + "epoch": 0.40468, + "grad_norm": 1.5180530242357684, + "learning_rate": 0.003, + "loss": 4.058, + "step": 40468 + }, + { + "epoch": 0.40469, + "grad_norm": 1.3572110750590745, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 40469 + }, + { + "epoch": 0.4047, + "grad_norm": 1.7482156217657592, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 40470 + }, + { + "epoch": 0.40471, + "grad_norm": 1.4021950763371258, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 40471 + }, + { + "epoch": 0.40472, + "grad_norm": 1.531782619632787, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 40472 + }, + { + "epoch": 0.40473, + "grad_norm": 1.3899605186307014, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40473 + }, + { + "epoch": 0.40474, + "grad_norm": 1.8208583143328572, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 40474 + }, + { + "epoch": 0.40475, + "grad_norm": 1.141384376952557, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 40475 + }, + { + "epoch": 0.40476, + "grad_norm": 1.7837273626177519, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 40476 + }, + { + "epoch": 0.40477, + "grad_norm": 1.5799960939560787, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 40477 + }, + { + "epoch": 0.40478, + "grad_norm": 1.591721697820582, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 40478 + }, + { + "epoch": 0.40479, + "grad_norm": 1.4994647427053838, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 40479 + }, + { + "epoch": 0.4048, + "grad_norm": 1.3720938870955883, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 40480 + }, + { + "epoch": 0.40481, + "grad_norm": 1.51457792125528, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 40481 + }, + { + "epoch": 0.40482, + "grad_norm": 1.5407780889945664, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 40482 + }, + { + "epoch": 0.40483, + "grad_norm": 1.4850895660141572, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40483 + }, + { + "epoch": 0.40484, + "grad_norm": 1.47524348610295, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 40484 + }, + { + "epoch": 0.40485, + "grad_norm": 1.607406331448807, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 40485 + }, + { + "epoch": 0.40486, + "grad_norm": 1.4760693985381617, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 40486 + }, + { + "epoch": 0.40487, + "grad_norm": 1.4576428330948321, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 40487 + }, + { + "epoch": 0.40488, + "grad_norm": 1.553317392691861, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40488 + }, + { + "epoch": 0.40489, + "grad_norm": 1.2243522377573894, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 40489 + }, + { + "epoch": 0.4049, + "grad_norm": 1.6780165031822183, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 40490 + }, + { + "epoch": 0.40491, + "grad_norm": 1.3770173933891456, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 40491 + }, + { + "epoch": 0.40492, + "grad_norm": 1.5294802272662342, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 40492 + }, + { + "epoch": 0.40493, + "grad_norm": 1.9918598063741362, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 40493 + }, + { + "epoch": 0.40494, + "grad_norm": 1.6719115292958817, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40494 + }, + { + "epoch": 0.40495, + "grad_norm": 1.389023282399445, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 40495 + }, + { + "epoch": 0.40496, + "grad_norm": 1.5637690124837242, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 40496 + }, + { + "epoch": 0.40497, + "grad_norm": 1.7452455854320155, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40497 + }, + { + "epoch": 0.40498, + "grad_norm": 1.5081926514449366, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40498 + }, + { + "epoch": 0.40499, + "grad_norm": 1.5388175226651495, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40499 + }, + { + "epoch": 0.405, + "grad_norm": 1.3251855805921493, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 40500 + }, + { + "epoch": 0.40501, + "grad_norm": 1.8380410520765171, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40501 + }, + { + "epoch": 0.40502, + "grad_norm": 1.2692904028088865, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 40502 + }, + { + "epoch": 0.40503, + "grad_norm": 1.5077543626614542, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40503 + }, + { + "epoch": 0.40504, + "grad_norm": 1.282387774203741, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40504 + }, + { + "epoch": 0.40505, + "grad_norm": 1.7805907805477228, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 40505 + }, + { + "epoch": 0.40506, + "grad_norm": 1.1113112983232094, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 40506 + }, + { + "epoch": 0.40507, + "grad_norm": 1.6976113250980795, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 40507 + }, + { + "epoch": 0.40508, + "grad_norm": 1.7659709623037014, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40508 + }, + { + "epoch": 0.40509, + "grad_norm": 1.371504940629617, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 40509 + }, + { + "epoch": 0.4051, + "grad_norm": 1.393515256370435, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 40510 + }, + { + "epoch": 0.40511, + "grad_norm": 1.4513108888774837, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 40511 + }, + { + "epoch": 0.40512, + "grad_norm": 1.784368564551658, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 40512 + }, + { + "epoch": 0.40513, + "grad_norm": 1.4097654144149387, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40513 + }, + { + "epoch": 0.40514, + "grad_norm": 1.3871036441017865, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40514 + }, + { + "epoch": 0.40515, + "grad_norm": 1.515683176490838, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 40515 + }, + { + "epoch": 0.40516, + "grad_norm": 1.393064705412407, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 40516 + }, + { + "epoch": 0.40517, + "grad_norm": 1.7398485499222893, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 40517 + }, + { + "epoch": 0.40518, + "grad_norm": 1.285626835033707, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 40518 + }, + { + "epoch": 0.40519, + "grad_norm": 1.9162141141190188, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 40519 + }, + { + "epoch": 0.4052, + "grad_norm": 1.2129847328692593, + "learning_rate": 0.003, + "loss": 4.058, + "step": 40520 + }, + { + "epoch": 0.40521, + "grad_norm": 1.56620722682814, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40521 + }, + { + "epoch": 0.40522, + "grad_norm": 1.202546291874468, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40522 + }, + { + "epoch": 0.40523, + "grad_norm": 1.504450977023936, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 40523 + }, + { + "epoch": 0.40524, + "grad_norm": 1.4989349002223997, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 40524 + }, + { + "epoch": 0.40525, + "grad_norm": 1.553637630286617, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 40525 + }, + { + "epoch": 0.40526, + "grad_norm": 1.5936628870037477, + "learning_rate": 0.003, + "loss": 4.071, + "step": 40526 + }, + { + "epoch": 0.40527, + "grad_norm": 1.5481693438046984, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 40527 + }, + { + "epoch": 0.40528, + "grad_norm": 1.362520120245211, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40528 + }, + { + "epoch": 0.40529, + "grad_norm": 1.7424308993906568, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40529 + }, + { + "epoch": 0.4053, + "grad_norm": 1.4472896036099865, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 40530 + }, + { + "epoch": 0.40531, + "grad_norm": 2.129978143993505, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 40531 + }, + { + "epoch": 0.40532, + "grad_norm": 1.5505148358267085, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 40532 + }, + { + "epoch": 0.40533, + "grad_norm": 1.4833208413825438, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 40533 + }, + { + "epoch": 0.40534, + "grad_norm": 1.279277055555515, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 40534 + }, + { + "epoch": 0.40535, + "grad_norm": 1.7125845901113, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40535 + }, + { + "epoch": 0.40536, + "grad_norm": 1.3638410202641267, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 40536 + }, + { + "epoch": 0.40537, + "grad_norm": 1.9044983239570061, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 40537 + }, + { + "epoch": 0.40538, + "grad_norm": 1.506214719857247, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 40538 + }, + { + "epoch": 0.40539, + "grad_norm": 1.7297937769977445, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 40539 + }, + { + "epoch": 0.4054, + "grad_norm": 1.2779618824919865, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 40540 + }, + { + "epoch": 0.40541, + "grad_norm": 1.7702460624427383, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 40541 + }, + { + "epoch": 0.40542, + "grad_norm": 1.3909820685760093, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40542 + }, + { + "epoch": 0.40543, + "grad_norm": 1.5871800831825897, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 40543 + }, + { + "epoch": 0.40544, + "grad_norm": 1.6664330067415343, + "learning_rate": 0.003, + "loss": 4.054, + "step": 40544 + }, + { + "epoch": 0.40545, + "grad_norm": 1.2275363322712673, + "learning_rate": 0.003, + "loss": 4.021, + "step": 40545 + }, + { + "epoch": 0.40546, + "grad_norm": 1.4897203334034181, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 40546 + }, + { + "epoch": 0.40547, + "grad_norm": 1.4401734268217667, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 40547 + }, + { + "epoch": 0.40548, + "grad_norm": 1.4068563767126323, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 40548 + }, + { + "epoch": 0.40549, + "grad_norm": 1.4454499811572898, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 40549 + }, + { + "epoch": 0.4055, + "grad_norm": 1.7022826452480042, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 40550 + }, + { + "epoch": 0.40551, + "grad_norm": 1.7301444074897059, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 40551 + }, + { + "epoch": 0.40552, + "grad_norm": 1.777249137994327, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 40552 + }, + { + "epoch": 0.40553, + "grad_norm": 1.1964648259311872, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 40553 + }, + { + "epoch": 0.40554, + "grad_norm": 1.5191630799160223, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 40554 + }, + { + "epoch": 0.40555, + "grad_norm": 1.7509845037304068, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 40555 + }, + { + "epoch": 0.40556, + "grad_norm": 1.7605842585488813, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 40556 + }, + { + "epoch": 0.40557, + "grad_norm": 1.2460235768208003, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 40557 + }, + { + "epoch": 0.40558, + "grad_norm": 1.8079038739930648, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 40558 + }, + { + "epoch": 0.40559, + "grad_norm": 1.3701139752510174, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40559 + }, + { + "epoch": 0.4056, + "grad_norm": 1.301214716026571, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 40560 + }, + { + "epoch": 0.40561, + "grad_norm": 1.831951838611972, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 40561 + }, + { + "epoch": 0.40562, + "grad_norm": 1.1965114172833498, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 40562 + }, + { + "epoch": 0.40563, + "grad_norm": 1.8984048966401377, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 40563 + }, + { + "epoch": 0.40564, + "grad_norm": 1.3644082571884797, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 40564 + }, + { + "epoch": 0.40565, + "grad_norm": 1.3770469723114023, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 40565 + }, + { + "epoch": 0.40566, + "grad_norm": 1.729807242664579, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 40566 + }, + { + "epoch": 0.40567, + "grad_norm": 1.3623410024534268, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 40567 + }, + { + "epoch": 0.40568, + "grad_norm": 1.3768089736580582, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 40568 + }, + { + "epoch": 0.40569, + "grad_norm": 1.5457229723174928, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 40569 + }, + { + "epoch": 0.4057, + "grad_norm": 1.6651875673088166, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 40570 + }, + { + "epoch": 0.40571, + "grad_norm": 1.729791173143301, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 40571 + }, + { + "epoch": 0.40572, + "grad_norm": 1.460134459869056, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40572 + }, + { + "epoch": 0.40573, + "grad_norm": 1.493675357406816, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 40573 + }, + { + "epoch": 0.40574, + "grad_norm": 1.4076012400819897, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 40574 + }, + { + "epoch": 0.40575, + "grad_norm": 1.4305032505581643, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 40575 + }, + { + "epoch": 0.40576, + "grad_norm": 1.5244783041288719, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40576 + }, + { + "epoch": 0.40577, + "grad_norm": 1.6013466181937563, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40577 + }, + { + "epoch": 0.40578, + "grad_norm": 1.6518046610626367, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 40578 + }, + { + "epoch": 0.40579, + "grad_norm": 1.5961161490921623, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40579 + }, + { + "epoch": 0.4058, + "grad_norm": 1.472719462696911, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 40580 + }, + { + "epoch": 0.40581, + "grad_norm": 1.540621741078016, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 40581 + }, + { + "epoch": 0.40582, + "grad_norm": 1.5654405124028306, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 40582 + }, + { + "epoch": 0.40583, + "grad_norm": 1.5830831553248457, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 40583 + }, + { + "epoch": 0.40584, + "grad_norm": 1.4429982434418054, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40584 + }, + { + "epoch": 0.40585, + "grad_norm": 1.6518396649991145, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 40585 + }, + { + "epoch": 0.40586, + "grad_norm": 1.5290170200039428, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 40586 + }, + { + "epoch": 0.40587, + "grad_norm": 1.2791205439026407, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40587 + }, + { + "epoch": 0.40588, + "grad_norm": 1.7258659214857404, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 40588 + }, + { + "epoch": 0.40589, + "grad_norm": 1.3794836051085557, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40589 + }, + { + "epoch": 0.4059, + "grad_norm": 1.4794122406426677, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40590 + }, + { + "epoch": 0.40591, + "grad_norm": 1.408321111602885, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40591 + }, + { + "epoch": 0.40592, + "grad_norm": 1.5908923057832856, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 40592 + }, + { + "epoch": 0.40593, + "grad_norm": 1.3438228169106263, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 40593 + }, + { + "epoch": 0.40594, + "grad_norm": 1.6379295328201404, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 40594 + }, + { + "epoch": 0.40595, + "grad_norm": 1.2566074228205069, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 40595 + }, + { + "epoch": 0.40596, + "grad_norm": 1.7634214830150128, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 40596 + }, + { + "epoch": 0.40597, + "grad_norm": 1.1720516412512545, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 40597 + }, + { + "epoch": 0.40598, + "grad_norm": 1.9917716133709489, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 40598 + }, + { + "epoch": 0.40599, + "grad_norm": 1.15396974977853, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 40599 + }, + { + "epoch": 0.406, + "grad_norm": 1.695383000194787, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 40600 + }, + { + "epoch": 0.40601, + "grad_norm": 1.537823519064769, + "learning_rate": 0.003, + "loss": 4.064, + "step": 40601 + }, + { + "epoch": 0.40602, + "grad_norm": 1.673754469576651, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 40602 + }, + { + "epoch": 0.40603, + "grad_norm": 1.8118513630953699, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 40603 + }, + { + "epoch": 0.40604, + "grad_norm": 1.6700444276822584, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 40604 + }, + { + "epoch": 0.40605, + "grad_norm": 1.4675261827774897, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 40605 + }, + { + "epoch": 0.40606, + "grad_norm": 1.4442457586742266, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 40606 + }, + { + "epoch": 0.40607, + "grad_norm": 1.4382513708328106, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 40607 + }, + { + "epoch": 0.40608, + "grad_norm": 1.4040673387224487, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 40608 + }, + { + "epoch": 0.40609, + "grad_norm": 1.4737966881937306, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 40609 + }, + { + "epoch": 0.4061, + "grad_norm": 1.5909591968840533, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 40610 + }, + { + "epoch": 0.40611, + "grad_norm": 1.5843295658648422, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 40611 + }, + { + "epoch": 0.40612, + "grad_norm": 1.2819005010956874, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 40612 + }, + { + "epoch": 0.40613, + "grad_norm": 1.6766266618286307, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 40613 + }, + { + "epoch": 0.40614, + "grad_norm": 1.4039238058106442, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40614 + }, + { + "epoch": 0.40615, + "grad_norm": 1.2925177759025843, + "learning_rate": 0.003, + "loss": 4.058, + "step": 40615 + }, + { + "epoch": 0.40616, + "grad_norm": 1.6281934692250148, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 40616 + }, + { + "epoch": 0.40617, + "grad_norm": 1.3406556270309686, + "learning_rate": 0.003, + "loss": 4.049, + "step": 40617 + }, + { + "epoch": 0.40618, + "grad_norm": 1.5168557428785225, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 40618 + }, + { + "epoch": 0.40619, + "grad_norm": 1.3834701168570982, + "learning_rate": 0.003, + "loss": 4.028, + "step": 40619 + }, + { + "epoch": 0.4062, + "grad_norm": 1.5013398243605145, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 40620 + }, + { + "epoch": 0.40621, + "grad_norm": 1.2720649236634587, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 40621 + }, + { + "epoch": 0.40622, + "grad_norm": 1.7412283386279823, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 40622 + }, + { + "epoch": 0.40623, + "grad_norm": 2.145929482309571, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 40623 + }, + { + "epoch": 0.40624, + "grad_norm": 1.5162931559687298, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 40624 + }, + { + "epoch": 0.40625, + "grad_norm": 1.326774837839541, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40625 + }, + { + "epoch": 0.40626, + "grad_norm": 1.5103893418817367, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 40626 + }, + { + "epoch": 0.40627, + "grad_norm": 2.163788289001314, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 40627 + }, + { + "epoch": 0.40628, + "grad_norm": 0.8711457358396604, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 40628 + }, + { + "epoch": 0.40629, + "grad_norm": 1.5758034592041412, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 40629 + }, + { + "epoch": 0.4063, + "grad_norm": 1.4186699455197704, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 40630 + }, + { + "epoch": 0.40631, + "grad_norm": 1.5832070031734624, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 40631 + }, + { + "epoch": 0.40632, + "grad_norm": 1.5854354260301127, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40632 + }, + { + "epoch": 0.40633, + "grad_norm": 1.73958258508371, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40633 + }, + { + "epoch": 0.40634, + "grad_norm": 1.6527943048643237, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 40634 + }, + { + "epoch": 0.40635, + "grad_norm": 1.6451242552108483, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 40635 + }, + { + "epoch": 0.40636, + "grad_norm": 1.3416072510204582, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 40636 + }, + { + "epoch": 0.40637, + "grad_norm": 1.6885955662856376, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 40637 + }, + { + "epoch": 0.40638, + "grad_norm": 1.31399417236966, + "learning_rate": 0.003, + "loss": 4.043, + "step": 40638 + }, + { + "epoch": 0.40639, + "grad_norm": 1.6232372770778463, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 40639 + }, + { + "epoch": 0.4064, + "grad_norm": 1.3030539854487189, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 40640 + }, + { + "epoch": 0.40641, + "grad_norm": 1.6913157109567454, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 40641 + }, + { + "epoch": 0.40642, + "grad_norm": 1.2276904517717768, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 40642 + }, + { + "epoch": 0.40643, + "grad_norm": 1.6443266005798527, + "learning_rate": 0.003, + "loss": 4.046, + "step": 40643 + }, + { + "epoch": 0.40644, + "grad_norm": 1.249395985605376, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 40644 + }, + { + "epoch": 0.40645, + "grad_norm": 1.8017116644318487, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 40645 + }, + { + "epoch": 0.40646, + "grad_norm": 1.492391735453744, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 40646 + }, + { + "epoch": 0.40647, + "grad_norm": 1.86362779333387, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 40647 + }, + { + "epoch": 0.40648, + "grad_norm": 1.5216806797275289, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 40648 + }, + { + "epoch": 0.40649, + "grad_norm": 1.4291262620487553, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 40649 + }, + { + "epoch": 0.4065, + "grad_norm": 1.4407563501016292, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 40650 + }, + { + "epoch": 0.40651, + "grad_norm": 1.5262428507915555, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 40651 + }, + { + "epoch": 0.40652, + "grad_norm": 1.5210250419107216, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40652 + }, + { + "epoch": 0.40653, + "grad_norm": 1.3987404114725666, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 40653 + }, + { + "epoch": 0.40654, + "grad_norm": 1.5094115329063733, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 40654 + }, + { + "epoch": 0.40655, + "grad_norm": 1.410113379756153, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 40655 + }, + { + "epoch": 0.40656, + "grad_norm": 1.4261515588446307, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 40656 + }, + { + "epoch": 0.40657, + "grad_norm": 1.53696840172399, + "learning_rate": 0.003, + "loss": 4.021, + "step": 40657 + }, + { + "epoch": 0.40658, + "grad_norm": 1.325580511366003, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 40658 + }, + { + "epoch": 0.40659, + "grad_norm": 1.6934118883241176, + "learning_rate": 0.003, + "loss": 4.035, + "step": 40659 + }, + { + "epoch": 0.4066, + "grad_norm": 1.6339009162283675, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 40660 + }, + { + "epoch": 0.40661, + "grad_norm": 1.5559537736326923, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 40661 + }, + { + "epoch": 0.40662, + "grad_norm": 1.6087186646590383, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40662 + }, + { + "epoch": 0.40663, + "grad_norm": 1.6631719177786561, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 40663 + }, + { + "epoch": 0.40664, + "grad_norm": 1.3251650538777897, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 40664 + }, + { + "epoch": 0.40665, + "grad_norm": 1.6252092060230805, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 40665 + }, + { + "epoch": 0.40666, + "grad_norm": 1.5472047310441792, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 40666 + }, + { + "epoch": 0.40667, + "grad_norm": 1.4159790263620144, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40667 + }, + { + "epoch": 0.40668, + "grad_norm": 1.8130557030063492, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40668 + }, + { + "epoch": 0.40669, + "grad_norm": 1.3440906031726247, + "learning_rate": 0.003, + "loss": 4.063, + "step": 40669 + }, + { + "epoch": 0.4067, + "grad_norm": 1.8181408651666846, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 40670 + }, + { + "epoch": 0.40671, + "grad_norm": 1.3779949582337607, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 40671 + }, + { + "epoch": 0.40672, + "grad_norm": 1.972367393117122, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 40672 + }, + { + "epoch": 0.40673, + "grad_norm": 1.1058730890680246, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 40673 + }, + { + "epoch": 0.40674, + "grad_norm": 1.6831631222001728, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 40674 + }, + { + "epoch": 0.40675, + "grad_norm": 1.8838824382218988, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40675 + }, + { + "epoch": 0.40676, + "grad_norm": 1.4212870228445749, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 40676 + }, + { + "epoch": 0.40677, + "grad_norm": 1.6388671127386645, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 40677 + }, + { + "epoch": 0.40678, + "grad_norm": 1.148048025623082, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 40678 + }, + { + "epoch": 0.40679, + "grad_norm": 1.6567207459854025, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 40679 + }, + { + "epoch": 0.4068, + "grad_norm": 1.4009494632304162, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 40680 + }, + { + "epoch": 0.40681, + "grad_norm": 1.5697206675827446, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 40681 + }, + { + "epoch": 0.40682, + "grad_norm": 1.5585937420728275, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 40682 + }, + { + "epoch": 0.40683, + "grad_norm": 1.7419726225284697, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40683 + }, + { + "epoch": 0.40684, + "grad_norm": 1.1196959776018687, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 40684 + }, + { + "epoch": 0.40685, + "grad_norm": 1.7668827063096426, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 40685 + }, + { + "epoch": 0.40686, + "grad_norm": 1.52521445637228, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 40686 + }, + { + "epoch": 0.40687, + "grad_norm": 1.675075279709223, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40687 + }, + { + "epoch": 0.40688, + "grad_norm": 1.5347981021297332, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 40688 + }, + { + "epoch": 0.40689, + "grad_norm": 1.4792814348371506, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40689 + }, + { + "epoch": 0.4069, + "grad_norm": 1.377546569060363, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 40690 + }, + { + "epoch": 0.40691, + "grad_norm": 1.585066635740695, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 40691 + }, + { + "epoch": 0.40692, + "grad_norm": 1.2231112999242, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 40692 + }, + { + "epoch": 0.40693, + "grad_norm": 1.8394841773338517, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 40693 + }, + { + "epoch": 0.40694, + "grad_norm": 1.338626918823969, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 40694 + }, + { + "epoch": 0.40695, + "grad_norm": 1.609117482993662, + "learning_rate": 0.003, + "loss": 4.036, + "step": 40695 + }, + { + "epoch": 0.40696, + "grad_norm": 1.3327798969763316, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 40696 + }, + { + "epoch": 0.40697, + "grad_norm": 1.7977481204048735, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 40697 + }, + { + "epoch": 0.40698, + "grad_norm": 1.542834647977603, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 40698 + }, + { + "epoch": 0.40699, + "grad_norm": 1.4413726822701975, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 40699 + }, + { + "epoch": 0.407, + "grad_norm": 1.5962108260344432, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 40700 + }, + { + "epoch": 0.40701, + "grad_norm": 1.6218220743145177, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 40701 + }, + { + "epoch": 0.40702, + "grad_norm": 1.296040689454445, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 40702 + }, + { + "epoch": 0.40703, + "grad_norm": 1.7293336436179636, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 40703 + }, + { + "epoch": 0.40704, + "grad_norm": 1.6823480525429566, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40704 + }, + { + "epoch": 0.40705, + "grad_norm": 1.3765918705367914, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 40705 + }, + { + "epoch": 0.40706, + "grad_norm": 1.5410522060511944, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40706 + }, + { + "epoch": 0.40707, + "grad_norm": 1.5508098732661653, + "learning_rate": 0.003, + "loss": 4.006, + "step": 40707 + }, + { + "epoch": 0.40708, + "grad_norm": 1.224146427488511, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 40708 + }, + { + "epoch": 0.40709, + "grad_norm": 1.4775578649847592, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40709 + }, + { + "epoch": 0.4071, + "grad_norm": 1.4926003689364506, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40710 + }, + { + "epoch": 0.40711, + "grad_norm": 1.4189803734231232, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 40711 + }, + { + "epoch": 0.40712, + "grad_norm": 1.433899782550048, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 40712 + }, + { + "epoch": 0.40713, + "grad_norm": 1.340545268712381, + "learning_rate": 0.003, + "loss": 4.069, + "step": 40713 + }, + { + "epoch": 0.40714, + "grad_norm": 1.579855797614487, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 40714 + }, + { + "epoch": 0.40715, + "grad_norm": 1.561437968618151, + "learning_rate": 0.003, + "loss": 4.038, + "step": 40715 + }, + { + "epoch": 0.40716, + "grad_norm": 1.5086075017537284, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 40716 + }, + { + "epoch": 0.40717, + "grad_norm": 2.0895742277485003, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 40717 + }, + { + "epoch": 0.40718, + "grad_norm": 1.1435204175325244, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 40718 + }, + { + "epoch": 0.40719, + "grad_norm": 1.8536827252746444, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 40719 + }, + { + "epoch": 0.4072, + "grad_norm": 1.4724979882782911, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 40720 + }, + { + "epoch": 0.40721, + "grad_norm": 1.7991344819442268, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 40721 + }, + { + "epoch": 0.40722, + "grad_norm": 1.274931508184339, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 40722 + }, + { + "epoch": 0.40723, + "grad_norm": 1.6872730260315023, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 40723 + }, + { + "epoch": 0.40724, + "grad_norm": 1.5494613823783043, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 40724 + }, + { + "epoch": 0.40725, + "grad_norm": 1.3734780410644154, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 40725 + }, + { + "epoch": 0.40726, + "grad_norm": 1.4478801271183213, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 40726 + }, + { + "epoch": 0.40727, + "grad_norm": 1.390571874748023, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 40727 + }, + { + "epoch": 0.40728, + "grad_norm": 1.8156178054319005, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 40728 + }, + { + "epoch": 0.40729, + "grad_norm": 1.1880853680519055, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 40729 + }, + { + "epoch": 0.4073, + "grad_norm": 1.5567380349885025, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 40730 + }, + { + "epoch": 0.40731, + "grad_norm": 1.4757765879222493, + "learning_rate": 0.003, + "loss": 4.017, + "step": 40731 + }, + { + "epoch": 0.40732, + "grad_norm": 1.4780464277045704, + "learning_rate": 0.003, + "loss": 4.037, + "step": 40732 + }, + { + "epoch": 0.40733, + "grad_norm": 1.434524629982804, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 40733 + }, + { + "epoch": 0.40734, + "grad_norm": 1.4386702063468448, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40734 + }, + { + "epoch": 0.40735, + "grad_norm": 1.6262092635801242, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 40735 + }, + { + "epoch": 0.40736, + "grad_norm": 1.460155216339771, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 40736 + }, + { + "epoch": 0.40737, + "grad_norm": 1.6228545202398117, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 40737 + }, + { + "epoch": 0.40738, + "grad_norm": 1.341782366060886, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 40738 + }, + { + "epoch": 0.40739, + "grad_norm": 1.7064008577001633, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 40739 + }, + { + "epoch": 0.4074, + "grad_norm": 1.958892152933086, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 40740 + }, + { + "epoch": 0.40741, + "grad_norm": 1.3329709515525547, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 40741 + }, + { + "epoch": 0.40742, + "grad_norm": 1.7723023798154214, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40742 + }, + { + "epoch": 0.40743, + "grad_norm": 1.77543266814757, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 40743 + }, + { + "epoch": 0.40744, + "grad_norm": 1.3048007670011985, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 40744 + }, + { + "epoch": 0.40745, + "grad_norm": 1.6056902822629093, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40745 + }, + { + "epoch": 0.40746, + "grad_norm": 1.2086064222399187, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 40746 + }, + { + "epoch": 0.40747, + "grad_norm": 1.5540932044588787, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 40747 + }, + { + "epoch": 0.40748, + "grad_norm": 1.3878076706225793, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 40748 + }, + { + "epoch": 0.40749, + "grad_norm": 1.5353161044488788, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 40749 + }, + { + "epoch": 0.4075, + "grad_norm": 1.5754779063142865, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 40750 + }, + { + "epoch": 0.40751, + "grad_norm": 1.648002214797735, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 40751 + }, + { + "epoch": 0.40752, + "grad_norm": 1.7413286632552007, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 40752 + }, + { + "epoch": 0.40753, + "grad_norm": 1.3663479990759133, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 40753 + }, + { + "epoch": 0.40754, + "grad_norm": 1.3023117130457613, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 40754 + }, + { + "epoch": 0.40755, + "grad_norm": 1.8620629365462582, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 40755 + }, + { + "epoch": 0.40756, + "grad_norm": 1.3852119227871944, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 40756 + }, + { + "epoch": 0.40757, + "grad_norm": 1.4771354733844195, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 40757 + }, + { + "epoch": 0.40758, + "grad_norm": 1.3807533476032234, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 40758 + }, + { + "epoch": 0.40759, + "grad_norm": 1.5943158167746432, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 40759 + }, + { + "epoch": 0.4076, + "grad_norm": 1.2047031200514509, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 40760 + }, + { + "epoch": 0.40761, + "grad_norm": 1.798863510006458, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 40761 + }, + { + "epoch": 0.40762, + "grad_norm": 1.5613224315969847, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40762 + }, + { + "epoch": 0.40763, + "grad_norm": 1.475010338471291, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 40763 + }, + { + "epoch": 0.40764, + "grad_norm": 1.8302611259660104, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40764 + }, + { + "epoch": 0.40765, + "grad_norm": 1.2622244260752236, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 40765 + }, + { + "epoch": 0.40766, + "grad_norm": 1.7120775770522212, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 40766 + }, + { + "epoch": 0.40767, + "grad_norm": 1.4914995261468484, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 40767 + }, + { + "epoch": 0.40768, + "grad_norm": 1.517263644947462, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40768 + }, + { + "epoch": 0.40769, + "grad_norm": 1.534314113799963, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40769 + }, + { + "epoch": 0.4077, + "grad_norm": 1.5737192381186993, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 40770 + }, + { + "epoch": 0.40771, + "grad_norm": 1.7400638465998903, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 40771 + }, + { + "epoch": 0.40772, + "grad_norm": 1.3130908112236739, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40772 + }, + { + "epoch": 0.40773, + "grad_norm": 1.7191750112514859, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40773 + }, + { + "epoch": 0.40774, + "grad_norm": 1.3084971876621019, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40774 + }, + { + "epoch": 0.40775, + "grad_norm": 1.6695871842960448, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40775 + }, + { + "epoch": 0.40776, + "grad_norm": 1.7111984103868634, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 40776 + }, + { + "epoch": 0.40777, + "grad_norm": 1.5409430961111013, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 40777 + }, + { + "epoch": 0.40778, + "grad_norm": 1.5169096339509744, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 40778 + }, + { + "epoch": 0.40779, + "grad_norm": 1.418006409802813, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 40779 + }, + { + "epoch": 0.4078, + "grad_norm": 1.5543954734260457, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40780 + }, + { + "epoch": 0.40781, + "grad_norm": 1.4418934014665536, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 40781 + }, + { + "epoch": 0.40782, + "grad_norm": 1.626325192480784, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 40782 + }, + { + "epoch": 0.40783, + "grad_norm": 1.575860359631109, + "learning_rate": 0.003, + "loss": 4.048, + "step": 40783 + }, + { + "epoch": 0.40784, + "grad_norm": 1.3129292346501846, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 40784 + }, + { + "epoch": 0.40785, + "grad_norm": 1.6861436448894969, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 40785 + }, + { + "epoch": 0.40786, + "grad_norm": 1.2307377638213857, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 40786 + }, + { + "epoch": 0.40787, + "grad_norm": 1.7576907657710543, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40787 + }, + { + "epoch": 0.40788, + "grad_norm": 1.4248616990359302, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 40788 + }, + { + "epoch": 0.40789, + "grad_norm": 1.7117091657553447, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 40789 + }, + { + "epoch": 0.4079, + "grad_norm": 1.2114107934374658, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 40790 + }, + { + "epoch": 0.40791, + "grad_norm": 1.5394127845936416, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 40791 + }, + { + "epoch": 0.40792, + "grad_norm": 1.329233349598259, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 40792 + }, + { + "epoch": 0.40793, + "grad_norm": 1.7345398621988177, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 40793 + }, + { + "epoch": 0.40794, + "grad_norm": 1.2983184729505042, + "learning_rate": 0.003, + "loss": 4.063, + "step": 40794 + }, + { + "epoch": 0.40795, + "grad_norm": 1.7427668238958145, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 40795 + }, + { + "epoch": 0.40796, + "grad_norm": 1.3833101742586944, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40796 + }, + { + "epoch": 0.40797, + "grad_norm": 1.6518874825033663, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 40797 + }, + { + "epoch": 0.40798, + "grad_norm": 1.495569715057804, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 40798 + }, + { + "epoch": 0.40799, + "grad_norm": 1.5609287811044172, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 40799 + }, + { + "epoch": 0.408, + "grad_norm": 1.5304509517628464, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 40800 + }, + { + "epoch": 0.40801, + "grad_norm": 1.4160632938592481, + "learning_rate": 0.003, + "loss": 4.026, + "step": 40801 + }, + { + "epoch": 0.40802, + "grad_norm": 1.436689875416726, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 40802 + }, + { + "epoch": 0.40803, + "grad_norm": 1.474722978347943, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 40803 + }, + { + "epoch": 0.40804, + "grad_norm": 1.545381118140685, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 40804 + }, + { + "epoch": 0.40805, + "grad_norm": 1.563090107346286, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 40805 + }, + { + "epoch": 0.40806, + "grad_norm": 1.636024009720487, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 40806 + }, + { + "epoch": 0.40807, + "grad_norm": 1.8918171726030306, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40807 + }, + { + "epoch": 0.40808, + "grad_norm": 1.5317209512709715, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 40808 + }, + { + "epoch": 0.40809, + "grad_norm": 1.5363868365572777, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 40809 + }, + { + "epoch": 0.4081, + "grad_norm": 1.7016487403830052, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 40810 + }, + { + "epoch": 0.40811, + "grad_norm": 1.16563273219837, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 40811 + }, + { + "epoch": 0.40812, + "grad_norm": 1.5519736812754588, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 40812 + }, + { + "epoch": 0.40813, + "grad_norm": 1.433411450491914, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 40813 + }, + { + "epoch": 0.40814, + "grad_norm": 1.4616442672838195, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 40814 + }, + { + "epoch": 0.40815, + "grad_norm": 1.520745994867205, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 40815 + }, + { + "epoch": 0.40816, + "grad_norm": 1.4469605198097433, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 40816 + }, + { + "epoch": 0.40817, + "grad_norm": 1.8914782607795784, + "learning_rate": 0.003, + "loss": 4.028, + "step": 40817 + }, + { + "epoch": 0.40818, + "grad_norm": 1.554813645533305, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 40818 + }, + { + "epoch": 0.40819, + "grad_norm": 1.4658046798588373, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 40819 + }, + { + "epoch": 0.4082, + "grad_norm": 1.6818499553919628, + "learning_rate": 0.003, + "loss": 4.05, + "step": 40820 + }, + { + "epoch": 0.40821, + "grad_norm": 1.5501397155359964, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 40821 + }, + { + "epoch": 0.40822, + "grad_norm": 1.6276578485395985, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 40822 + }, + { + "epoch": 0.40823, + "grad_norm": 1.55609507396572, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 40823 + }, + { + "epoch": 0.40824, + "grad_norm": 1.6766329384350764, + "learning_rate": 0.003, + "loss": 4.011, + "step": 40824 + }, + { + "epoch": 0.40825, + "grad_norm": 1.2335028793484093, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40825 + }, + { + "epoch": 0.40826, + "grad_norm": 1.8162606899974427, + "learning_rate": 0.003, + "loss": 4.055, + "step": 40826 + }, + { + "epoch": 0.40827, + "grad_norm": 1.2870078493299808, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40827 + }, + { + "epoch": 0.40828, + "grad_norm": 1.754583693141419, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 40828 + }, + { + "epoch": 0.40829, + "grad_norm": 1.3049545612111608, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 40829 + }, + { + "epoch": 0.4083, + "grad_norm": 2.261385061854426, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 40830 + }, + { + "epoch": 0.40831, + "grad_norm": 1.5736942001775416, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 40831 + }, + { + "epoch": 0.40832, + "grad_norm": 1.478448618550995, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 40832 + }, + { + "epoch": 0.40833, + "grad_norm": 1.6596309060402084, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 40833 + }, + { + "epoch": 0.40834, + "grad_norm": 1.359740782597847, + "learning_rate": 0.003, + "loss": 4.03, + "step": 40834 + }, + { + "epoch": 0.40835, + "grad_norm": 1.4738213087295589, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 40835 + }, + { + "epoch": 0.40836, + "grad_norm": 1.5496771648385457, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 40836 + }, + { + "epoch": 0.40837, + "grad_norm": 1.2855740707645704, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 40837 + }, + { + "epoch": 0.40838, + "grad_norm": 1.767022738675035, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 40838 + }, + { + "epoch": 0.40839, + "grad_norm": 1.5820641879403967, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 40839 + }, + { + "epoch": 0.4084, + "grad_norm": 1.9709765403176829, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 40840 + }, + { + "epoch": 0.40841, + "grad_norm": 1.617996315148943, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 40841 + }, + { + "epoch": 0.40842, + "grad_norm": 1.3696919748821685, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 40842 + }, + { + "epoch": 0.40843, + "grad_norm": 1.616237166470913, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 40843 + }, + { + "epoch": 0.40844, + "grad_norm": 1.7908927606839558, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 40844 + }, + { + "epoch": 0.40845, + "grad_norm": 1.1771316995971868, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 40845 + }, + { + "epoch": 0.40846, + "grad_norm": 1.591847831469772, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 40846 + }, + { + "epoch": 0.40847, + "grad_norm": 1.387047351309697, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 40847 + }, + { + "epoch": 0.40848, + "grad_norm": 1.8588428813984212, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 40848 + }, + { + "epoch": 0.40849, + "grad_norm": 1.4958412590716008, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 40849 + }, + { + "epoch": 0.4085, + "grad_norm": 1.4327341497952384, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40850 + }, + { + "epoch": 0.40851, + "grad_norm": 1.8035903229999972, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 40851 + }, + { + "epoch": 0.40852, + "grad_norm": 1.1813295989670036, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 40852 + }, + { + "epoch": 0.40853, + "grad_norm": 1.6579238184838583, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40853 + }, + { + "epoch": 0.40854, + "grad_norm": 1.4681637778226966, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 40854 + }, + { + "epoch": 0.40855, + "grad_norm": 1.9249909402126588, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40855 + }, + { + "epoch": 0.40856, + "grad_norm": 1.3329843480672117, + "learning_rate": 0.003, + "loss": 4.045, + "step": 40856 + }, + { + "epoch": 0.40857, + "grad_norm": 1.5487734348055582, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 40857 + }, + { + "epoch": 0.40858, + "grad_norm": 1.6962528624097242, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 40858 + }, + { + "epoch": 0.40859, + "grad_norm": 1.4941354160989737, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 40859 + }, + { + "epoch": 0.4086, + "grad_norm": 1.4895144421179718, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 40860 + }, + { + "epoch": 0.40861, + "grad_norm": 1.3428265721466761, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 40861 + }, + { + "epoch": 0.40862, + "grad_norm": 1.4333691356235696, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 40862 + }, + { + "epoch": 0.40863, + "grad_norm": 1.3447645365871057, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 40863 + }, + { + "epoch": 0.40864, + "grad_norm": 1.6601869016967963, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 40864 + }, + { + "epoch": 0.40865, + "grad_norm": 1.4674349603412675, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 40865 + }, + { + "epoch": 0.40866, + "grad_norm": 1.6284634041551667, + "learning_rate": 0.003, + "loss": 4.031, + "step": 40866 + }, + { + "epoch": 0.40867, + "grad_norm": 1.3237340276228888, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 40867 + }, + { + "epoch": 0.40868, + "grad_norm": 1.7800021803112582, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 40868 + }, + { + "epoch": 0.40869, + "grad_norm": 1.3786686783567759, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 40869 + }, + { + "epoch": 0.4087, + "grad_norm": 1.7634561125876838, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 40870 + }, + { + "epoch": 0.40871, + "grad_norm": 1.289956437908221, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 40871 + }, + { + "epoch": 0.40872, + "grad_norm": 1.568299214421058, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 40872 + }, + { + "epoch": 0.40873, + "grad_norm": 1.6450485844624747, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 40873 + }, + { + "epoch": 0.40874, + "grad_norm": 1.5367182717905428, + "learning_rate": 0.003, + "loss": 4.051, + "step": 40874 + }, + { + "epoch": 0.40875, + "grad_norm": 1.3138815676208817, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 40875 + }, + { + "epoch": 0.40876, + "grad_norm": 1.771851712768085, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 40876 + }, + { + "epoch": 0.40877, + "grad_norm": 1.8524203438739244, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 40877 + }, + { + "epoch": 0.40878, + "grad_norm": 1.327048621806493, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 40878 + }, + { + "epoch": 0.40879, + "grad_norm": 1.7143742995184317, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 40879 + }, + { + "epoch": 0.4088, + "grad_norm": 1.5944739599562847, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 40880 + }, + { + "epoch": 0.40881, + "grad_norm": 1.5979718834447212, + "learning_rate": 0.003, + "loss": 4.025, + "step": 40881 + }, + { + "epoch": 0.40882, + "grad_norm": 1.3300895228638254, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 40882 + }, + { + "epoch": 0.40883, + "grad_norm": 1.4449618039652234, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 40883 + }, + { + "epoch": 0.40884, + "grad_norm": 1.3317964937107447, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 40884 + }, + { + "epoch": 0.40885, + "grad_norm": 1.7470769057433466, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 40885 + }, + { + "epoch": 0.40886, + "grad_norm": 1.3331025881691434, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 40886 + }, + { + "epoch": 0.40887, + "grad_norm": 1.530183763112304, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 40887 + }, + { + "epoch": 0.40888, + "grad_norm": 1.5088557542803291, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 40888 + }, + { + "epoch": 0.40889, + "grad_norm": 1.5327182240868271, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 40889 + }, + { + "epoch": 0.4089, + "grad_norm": 1.556523198137825, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 40890 + }, + { + "epoch": 0.40891, + "grad_norm": 1.8267460185585627, + "learning_rate": 0.003, + "loss": 4.041, + "step": 40891 + }, + { + "epoch": 0.40892, + "grad_norm": 1.7022609363447163, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40892 + }, + { + "epoch": 0.40893, + "grad_norm": 1.4766294930742738, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 40893 + }, + { + "epoch": 0.40894, + "grad_norm": 1.4159961045861595, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 40894 + }, + { + "epoch": 0.40895, + "grad_norm": 1.5965807279598772, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 40895 + }, + { + "epoch": 0.40896, + "grad_norm": 1.2002961594041248, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 40896 + }, + { + "epoch": 0.40897, + "grad_norm": 1.7180386630001476, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 40897 + }, + { + "epoch": 0.40898, + "grad_norm": 1.3728764620679828, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40898 + }, + { + "epoch": 0.40899, + "grad_norm": 1.5188955607913028, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40899 + }, + { + "epoch": 0.409, + "grad_norm": 1.285700410908939, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 40900 + }, + { + "epoch": 0.40901, + "grad_norm": 1.7683259600689132, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40901 + }, + { + "epoch": 0.40902, + "grad_norm": 1.2263104661006077, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 40902 + }, + { + "epoch": 0.40903, + "grad_norm": 1.7960642783508411, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 40903 + }, + { + "epoch": 0.40904, + "grad_norm": 1.5384585846191245, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 40904 + }, + { + "epoch": 0.40905, + "grad_norm": 1.7569868688987025, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 40905 + }, + { + "epoch": 0.40906, + "grad_norm": 1.2846904899840257, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40906 + }, + { + "epoch": 0.40907, + "grad_norm": 1.6457572154264026, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 40907 + }, + { + "epoch": 0.40908, + "grad_norm": 1.3370682277988184, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 40908 + }, + { + "epoch": 0.40909, + "grad_norm": 1.5730446686661024, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 40909 + }, + { + "epoch": 0.4091, + "grad_norm": 1.5599961142609857, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 40910 + }, + { + "epoch": 0.40911, + "grad_norm": 1.3827771011259085, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 40911 + }, + { + "epoch": 0.40912, + "grad_norm": 1.8387063175148741, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 40912 + }, + { + "epoch": 0.40913, + "grad_norm": 1.2165064857794465, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 40913 + }, + { + "epoch": 0.40914, + "grad_norm": 1.6408517349144605, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 40914 + }, + { + "epoch": 0.40915, + "grad_norm": 1.616084502210247, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 40915 + }, + { + "epoch": 0.40916, + "grad_norm": 1.4838767628980525, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 40916 + }, + { + "epoch": 0.40917, + "grad_norm": 1.777075886905828, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40917 + }, + { + "epoch": 0.40918, + "grad_norm": 1.6439794015370297, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 40918 + }, + { + "epoch": 0.40919, + "grad_norm": 1.39130860328586, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 40919 + }, + { + "epoch": 0.4092, + "grad_norm": 1.5803041286718584, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 40920 + }, + { + "epoch": 0.40921, + "grad_norm": 1.4514249075203858, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 40921 + }, + { + "epoch": 0.40922, + "grad_norm": 1.2855861819398446, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 40922 + }, + { + "epoch": 0.40923, + "grad_norm": 1.5497966874185962, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 40923 + }, + { + "epoch": 0.40924, + "grad_norm": 1.3791719887097276, + "learning_rate": 0.003, + "loss": 4.032, + "step": 40924 + }, + { + "epoch": 0.40925, + "grad_norm": 1.5951625498727229, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 40925 + }, + { + "epoch": 0.40926, + "grad_norm": 1.4131251041026645, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 40926 + }, + { + "epoch": 0.40927, + "grad_norm": 1.729729681031251, + "learning_rate": 0.003, + "loss": 4.047, + "step": 40927 + }, + { + "epoch": 0.40928, + "grad_norm": 1.3296648781467935, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 40928 + }, + { + "epoch": 0.40929, + "grad_norm": 1.5444246311767622, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 40929 + }, + { + "epoch": 0.4093, + "grad_norm": 1.2792737385464164, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 40930 + }, + { + "epoch": 0.40931, + "grad_norm": 1.78869568918114, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40931 + }, + { + "epoch": 0.40932, + "grad_norm": 1.5240698739053382, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 40932 + }, + { + "epoch": 0.40933, + "grad_norm": 1.868801712815792, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 40933 + }, + { + "epoch": 0.40934, + "grad_norm": 1.55711705665473, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 40934 + }, + { + "epoch": 0.40935, + "grad_norm": 1.546924665830106, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 40935 + }, + { + "epoch": 0.40936, + "grad_norm": 1.566478258124144, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 40936 + }, + { + "epoch": 0.40937, + "grad_norm": 1.5326580761702597, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 40937 + }, + { + "epoch": 0.40938, + "grad_norm": 1.6657334374500035, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 40938 + }, + { + "epoch": 0.40939, + "grad_norm": 1.3187351346733793, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 40939 + }, + { + "epoch": 0.4094, + "grad_norm": 2.2409190103392476, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 40940 + }, + { + "epoch": 0.40941, + "grad_norm": 1.0400195379592372, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 40941 + }, + { + "epoch": 0.40942, + "grad_norm": 1.9367485060600413, + "learning_rate": 0.003, + "loss": 4.057, + "step": 40942 + }, + { + "epoch": 0.40943, + "grad_norm": 1.244630955995404, + "learning_rate": 0.003, + "loss": 4.067, + "step": 40943 + }, + { + "epoch": 0.40944, + "grad_norm": 1.8268327369219142, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 40944 + }, + { + "epoch": 0.40945, + "grad_norm": 1.4888227665297253, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 40945 + }, + { + "epoch": 0.40946, + "grad_norm": 1.7451592543193242, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 40946 + }, + { + "epoch": 0.40947, + "grad_norm": 1.6142312333994728, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40947 + }, + { + "epoch": 0.40948, + "grad_norm": 1.4112065241080358, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 40948 + }, + { + "epoch": 0.40949, + "grad_norm": 1.3890584831606705, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 40949 + }, + { + "epoch": 0.4095, + "grad_norm": 1.4079493544726782, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 40950 + }, + { + "epoch": 0.40951, + "grad_norm": 1.7424707712167475, + "learning_rate": 0.003, + "loss": 4.038, + "step": 40951 + }, + { + "epoch": 0.40952, + "grad_norm": 1.1407661046504942, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 40952 + }, + { + "epoch": 0.40953, + "grad_norm": 1.9935705164892643, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 40953 + }, + { + "epoch": 0.40954, + "grad_norm": 0.9517725362025433, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 40954 + }, + { + "epoch": 0.40955, + "grad_norm": 1.6213137824881911, + "learning_rate": 0.003, + "loss": 4.08, + "step": 40955 + }, + { + "epoch": 0.40956, + "grad_norm": 1.6230078869672837, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 40956 + }, + { + "epoch": 0.40957, + "grad_norm": 1.3869474335548637, + "learning_rate": 0.003, + "loss": 4.06, + "step": 40957 + }, + { + "epoch": 0.40958, + "grad_norm": 1.4980322475863903, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 40958 + }, + { + "epoch": 0.40959, + "grad_norm": 1.3402404620705743, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 40959 + }, + { + "epoch": 0.4096, + "grad_norm": 1.6027772939682319, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 40960 + }, + { + "epoch": 0.40961, + "grad_norm": 1.639641436363637, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 40961 + }, + { + "epoch": 0.40962, + "grad_norm": 1.872047783386089, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 40962 + }, + { + "epoch": 0.40963, + "grad_norm": 1.500049575798862, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 40963 + }, + { + "epoch": 0.40964, + "grad_norm": 1.448999332105586, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 40964 + }, + { + "epoch": 0.40965, + "grad_norm": 1.6183658005544161, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 40965 + }, + { + "epoch": 0.40966, + "grad_norm": 1.4947954307477078, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 40966 + }, + { + "epoch": 0.40967, + "grad_norm": 1.4768038707553997, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 40967 + }, + { + "epoch": 0.40968, + "grad_norm": 1.2743792755897947, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 40968 + }, + { + "epoch": 0.40969, + "grad_norm": 1.501364868711847, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 40969 + }, + { + "epoch": 0.4097, + "grad_norm": 1.7302143063427018, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 40970 + }, + { + "epoch": 0.40971, + "grad_norm": 1.5939475074990919, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 40971 + }, + { + "epoch": 0.40972, + "grad_norm": 1.2364572489457497, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 40972 + }, + { + "epoch": 0.40973, + "grad_norm": 1.6596541381521794, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 40973 + }, + { + "epoch": 0.40974, + "grad_norm": 1.47152830498144, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 40974 + }, + { + "epoch": 0.40975, + "grad_norm": 1.7566235928740503, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 40975 + }, + { + "epoch": 0.40976, + "grad_norm": 1.5734973469226499, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 40976 + }, + { + "epoch": 0.40977, + "grad_norm": 1.202755836961281, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 40977 + }, + { + "epoch": 0.40978, + "grad_norm": 1.5641187857461931, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 40978 + }, + { + "epoch": 0.40979, + "grad_norm": 1.3567371547030673, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 40979 + }, + { + "epoch": 0.4098, + "grad_norm": 1.4450039002603792, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 40980 + }, + { + "epoch": 0.40981, + "grad_norm": 1.3628141315881641, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40981 + }, + { + "epoch": 0.40982, + "grad_norm": 1.5330035558794446, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 40982 + }, + { + "epoch": 0.40983, + "grad_norm": 1.5175123238949735, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 40983 + }, + { + "epoch": 0.40984, + "grad_norm": 1.61396240173662, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 40984 + }, + { + "epoch": 0.40985, + "grad_norm": 1.750445035447545, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 40985 + }, + { + "epoch": 0.40986, + "grad_norm": 2.103587854678312, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 40986 + }, + { + "epoch": 0.40987, + "grad_norm": 1.2603021432205979, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 40987 + }, + { + "epoch": 0.40988, + "grad_norm": 1.50959084265189, + "learning_rate": 0.003, + "loss": 4.065, + "step": 40988 + }, + { + "epoch": 0.40989, + "grad_norm": 1.5031702567932348, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 40989 + }, + { + "epoch": 0.4099, + "grad_norm": 1.7038801811426574, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 40990 + }, + { + "epoch": 0.40991, + "grad_norm": 1.4496146032547184, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 40991 + }, + { + "epoch": 0.40992, + "grad_norm": 1.6879981152041001, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 40992 + }, + { + "epoch": 0.40993, + "grad_norm": 1.4616141594319059, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 40993 + }, + { + "epoch": 0.40994, + "grad_norm": 1.29202158041084, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 40994 + }, + { + "epoch": 0.40995, + "grad_norm": 1.5889534380595451, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 40995 + }, + { + "epoch": 0.40996, + "grad_norm": 1.2916131424519497, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 40996 + }, + { + "epoch": 0.40997, + "grad_norm": 1.651414903336473, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 40997 + }, + { + "epoch": 0.40998, + "grad_norm": 1.5165603119568412, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 40998 + }, + { + "epoch": 0.40999, + "grad_norm": 1.4079268511289404, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 40999 + }, + { + "epoch": 0.41, + "grad_norm": 1.672683058166464, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41000 + }, + { + "epoch": 0.41001, + "grad_norm": 1.1313266683956165, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41001 + }, + { + "epoch": 0.41002, + "grad_norm": 1.9736161080944798, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 41002 + }, + { + "epoch": 0.41003, + "grad_norm": 1.1125858823957562, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41003 + }, + { + "epoch": 0.41004, + "grad_norm": 2.052812842530022, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 41004 + }, + { + "epoch": 0.41005, + "grad_norm": 1.3662704606611231, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 41005 + }, + { + "epoch": 0.41006, + "grad_norm": 1.6571455082392015, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41006 + }, + { + "epoch": 0.41007, + "grad_norm": 1.547685240032256, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 41007 + }, + { + "epoch": 0.41008, + "grad_norm": 1.813114590032506, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 41008 + }, + { + "epoch": 0.41009, + "grad_norm": 1.4022915169387316, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 41009 + }, + { + "epoch": 0.4101, + "grad_norm": 1.6845547920228672, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 41010 + }, + { + "epoch": 0.41011, + "grad_norm": 1.7211622198226142, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41011 + }, + { + "epoch": 0.41012, + "grad_norm": 1.2098799984982418, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41012 + }, + { + "epoch": 0.41013, + "grad_norm": 1.5169621244462805, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41013 + }, + { + "epoch": 0.41014, + "grad_norm": 1.611480974195635, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41014 + }, + { + "epoch": 0.41015, + "grad_norm": 1.8115262499985634, + "learning_rate": 0.003, + "loss": 4.08, + "step": 41015 + }, + { + "epoch": 0.41016, + "grad_norm": 1.2031453308529914, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 41016 + }, + { + "epoch": 0.41017, + "grad_norm": 1.6843529528377448, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 41017 + }, + { + "epoch": 0.41018, + "grad_norm": 1.2908855831180173, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41018 + }, + { + "epoch": 0.41019, + "grad_norm": 1.3099822632119973, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 41019 + }, + { + "epoch": 0.4102, + "grad_norm": 1.4467346438798578, + "learning_rate": 0.003, + "loss": 4.054, + "step": 41020 + }, + { + "epoch": 0.41021, + "grad_norm": 1.3859305846809227, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 41021 + }, + { + "epoch": 0.41022, + "grad_norm": 1.6437496012301067, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41022 + }, + { + "epoch": 0.41023, + "grad_norm": 1.5287672112138493, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 41023 + }, + { + "epoch": 0.41024, + "grad_norm": 1.4897644387084907, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 41024 + }, + { + "epoch": 0.41025, + "grad_norm": 1.621771344949379, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41025 + }, + { + "epoch": 0.41026, + "grad_norm": 1.5713375270890393, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41026 + }, + { + "epoch": 0.41027, + "grad_norm": 1.3468258115093183, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41027 + }, + { + "epoch": 0.41028, + "grad_norm": 1.5156070231108283, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 41028 + }, + { + "epoch": 0.41029, + "grad_norm": 1.6090922861287247, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41029 + }, + { + "epoch": 0.4103, + "grad_norm": 1.4791831560700899, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 41030 + }, + { + "epoch": 0.41031, + "grad_norm": 1.5231244636311774, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 41031 + }, + { + "epoch": 0.41032, + "grad_norm": 1.323603613952192, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 41032 + }, + { + "epoch": 0.41033, + "grad_norm": 1.5510954993788857, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 41033 + }, + { + "epoch": 0.41034, + "grad_norm": 1.277005763345458, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 41034 + }, + { + "epoch": 0.41035, + "grad_norm": 1.7533720882994086, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 41035 + }, + { + "epoch": 0.41036, + "grad_norm": 1.1412004328141523, + "learning_rate": 0.003, + "loss": 4.014, + "step": 41036 + }, + { + "epoch": 0.41037, + "grad_norm": 1.6298840170840252, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41037 + }, + { + "epoch": 0.41038, + "grad_norm": 1.5833935220394784, + "learning_rate": 0.003, + "loss": 4.054, + "step": 41038 + }, + { + "epoch": 0.41039, + "grad_norm": 1.624111674488335, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 41039 + }, + { + "epoch": 0.4104, + "grad_norm": 1.9574743383700897, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 41040 + }, + { + "epoch": 0.41041, + "grad_norm": 1.1109561010988014, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 41041 + }, + { + "epoch": 0.41042, + "grad_norm": 1.6292859806747204, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41042 + }, + { + "epoch": 0.41043, + "grad_norm": 1.5881034830932004, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 41043 + }, + { + "epoch": 0.41044, + "grad_norm": 1.6980164468924672, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 41044 + }, + { + "epoch": 0.41045, + "grad_norm": 1.331866123528162, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41045 + }, + { + "epoch": 0.41046, + "grad_norm": 1.7823381186764182, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 41046 + }, + { + "epoch": 0.41047, + "grad_norm": 1.2814242085635419, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41047 + }, + { + "epoch": 0.41048, + "grad_norm": 1.5006135009917516, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 41048 + }, + { + "epoch": 0.41049, + "grad_norm": 1.524511708717759, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 41049 + }, + { + "epoch": 0.4105, + "grad_norm": 1.7256547879859212, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 41050 + }, + { + "epoch": 0.41051, + "grad_norm": 1.5662635747895925, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 41051 + }, + { + "epoch": 0.41052, + "grad_norm": 1.730655082145875, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 41052 + }, + { + "epoch": 0.41053, + "grad_norm": 1.6375198393187116, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 41053 + }, + { + "epoch": 0.41054, + "grad_norm": 1.5272476266384978, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 41054 + }, + { + "epoch": 0.41055, + "grad_norm": 1.395516308470113, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 41055 + }, + { + "epoch": 0.41056, + "grad_norm": 1.5287851050653587, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 41056 + }, + { + "epoch": 0.41057, + "grad_norm": 1.6484898806781587, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41057 + }, + { + "epoch": 0.41058, + "grad_norm": 1.2995090853075133, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 41058 + }, + { + "epoch": 0.41059, + "grad_norm": 1.4124228840750335, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 41059 + }, + { + "epoch": 0.4106, + "grad_norm": 1.6314875653568395, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 41060 + }, + { + "epoch": 0.41061, + "grad_norm": 1.5157039317579424, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 41061 + }, + { + "epoch": 0.41062, + "grad_norm": 1.5538152578245465, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 41062 + }, + { + "epoch": 0.41063, + "grad_norm": 1.5048227394631704, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 41063 + }, + { + "epoch": 0.41064, + "grad_norm": 1.7145961810709915, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41064 + }, + { + "epoch": 0.41065, + "grad_norm": 1.7572567277605173, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 41065 + }, + { + "epoch": 0.41066, + "grad_norm": 1.3430691455954253, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 41066 + }, + { + "epoch": 0.41067, + "grad_norm": 1.869240678741608, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 41067 + }, + { + "epoch": 0.41068, + "grad_norm": 1.1764119412424774, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41068 + }, + { + "epoch": 0.41069, + "grad_norm": 1.6577942572206743, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 41069 + }, + { + "epoch": 0.4107, + "grad_norm": 1.3743195383992841, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 41070 + }, + { + "epoch": 0.41071, + "grad_norm": 1.8544125975356665, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 41071 + }, + { + "epoch": 0.41072, + "grad_norm": 1.454202849873172, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 41072 + }, + { + "epoch": 0.41073, + "grad_norm": 1.5138167609893283, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 41073 + }, + { + "epoch": 0.41074, + "grad_norm": 1.7407506727922386, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 41074 + }, + { + "epoch": 0.41075, + "grad_norm": 1.575535930010088, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41075 + }, + { + "epoch": 0.41076, + "grad_norm": 1.6461275740121366, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 41076 + }, + { + "epoch": 0.41077, + "grad_norm": 1.2392849115072813, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 41077 + }, + { + "epoch": 0.41078, + "grad_norm": 1.5761679083205746, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 41078 + }, + { + "epoch": 0.41079, + "grad_norm": 1.407377369909002, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 41079 + }, + { + "epoch": 0.4108, + "grad_norm": 1.6421628494870404, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 41080 + }, + { + "epoch": 0.41081, + "grad_norm": 1.3258441031614876, + "learning_rate": 0.003, + "loss": 4.016, + "step": 41081 + }, + { + "epoch": 0.41082, + "grad_norm": 1.641627738060025, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 41082 + }, + { + "epoch": 0.41083, + "grad_norm": 1.256130099405426, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 41083 + }, + { + "epoch": 0.41084, + "grad_norm": 1.5897170976902653, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 41084 + }, + { + "epoch": 0.41085, + "grad_norm": 1.4468626009144272, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41085 + }, + { + "epoch": 0.41086, + "grad_norm": 1.8067294381920838, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41086 + }, + { + "epoch": 0.41087, + "grad_norm": 1.952619162360702, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 41087 + }, + { + "epoch": 0.41088, + "grad_norm": 1.3460084306152797, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 41088 + }, + { + "epoch": 0.41089, + "grad_norm": 1.5210666988455632, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 41089 + }, + { + "epoch": 0.4109, + "grad_norm": 1.6407191182044947, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 41090 + }, + { + "epoch": 0.41091, + "grad_norm": 1.555179876803219, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41091 + }, + { + "epoch": 0.41092, + "grad_norm": 1.3814119501857036, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 41092 + }, + { + "epoch": 0.41093, + "grad_norm": 1.7715276563635363, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 41093 + }, + { + "epoch": 0.41094, + "grad_norm": 1.4728435307046814, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 41094 + }, + { + "epoch": 0.41095, + "grad_norm": 1.354435045287238, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 41095 + }, + { + "epoch": 0.41096, + "grad_norm": 1.5186053569156133, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 41096 + }, + { + "epoch": 0.41097, + "grad_norm": 1.646550189096874, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 41097 + }, + { + "epoch": 0.41098, + "grad_norm": 1.2966043145848762, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 41098 + }, + { + "epoch": 0.41099, + "grad_norm": 1.670839213681658, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 41099 + }, + { + "epoch": 0.411, + "grad_norm": 1.437650847453722, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 41100 + }, + { + "epoch": 0.41101, + "grad_norm": 1.694272118137213, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 41101 + }, + { + "epoch": 0.41102, + "grad_norm": 1.1380912134752244, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 41102 + }, + { + "epoch": 0.41103, + "grad_norm": 1.8580121145857014, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 41103 + }, + { + "epoch": 0.41104, + "grad_norm": 1.6689834599824893, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 41104 + }, + { + "epoch": 0.41105, + "grad_norm": 1.6413191884056402, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 41105 + }, + { + "epoch": 0.41106, + "grad_norm": 1.5851041072031853, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 41106 + }, + { + "epoch": 0.41107, + "grad_norm": 1.5521572884919512, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 41107 + }, + { + "epoch": 0.41108, + "grad_norm": 1.7179622103305974, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41108 + }, + { + "epoch": 0.41109, + "grad_norm": 1.2953368227433544, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 41109 + }, + { + "epoch": 0.4111, + "grad_norm": 1.734806998880313, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 41110 + }, + { + "epoch": 0.41111, + "grad_norm": 1.4052750255097148, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 41111 + }, + { + "epoch": 0.41112, + "grad_norm": 1.6323131625566374, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 41112 + }, + { + "epoch": 0.41113, + "grad_norm": 1.199410316948321, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41113 + }, + { + "epoch": 0.41114, + "grad_norm": 1.5991274741451416, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 41114 + }, + { + "epoch": 0.41115, + "grad_norm": 1.178349127540678, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 41115 + }, + { + "epoch": 0.41116, + "grad_norm": 1.7571812621753378, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 41116 + }, + { + "epoch": 0.41117, + "grad_norm": 1.4548719055201473, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 41117 + }, + { + "epoch": 0.41118, + "grad_norm": 1.7797704309869224, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 41118 + }, + { + "epoch": 0.41119, + "grad_norm": 1.2817409560075077, + "learning_rate": 0.003, + "loss": 4.053, + "step": 41119 + }, + { + "epoch": 0.4112, + "grad_norm": 1.6678391568798159, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 41120 + }, + { + "epoch": 0.41121, + "grad_norm": 1.571965748116856, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41121 + }, + { + "epoch": 0.41122, + "grad_norm": 1.737757804434229, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 41122 + }, + { + "epoch": 0.41123, + "grad_norm": 1.197311476637755, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 41123 + }, + { + "epoch": 0.41124, + "grad_norm": 1.543521031933837, + "learning_rate": 0.003, + "loss": 4.036, + "step": 41124 + }, + { + "epoch": 0.41125, + "grad_norm": 1.416860095453767, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 41125 + }, + { + "epoch": 0.41126, + "grad_norm": 1.4267995778657687, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 41126 + }, + { + "epoch": 0.41127, + "grad_norm": 1.4340796355810679, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 41127 + }, + { + "epoch": 0.41128, + "grad_norm": 1.8183254850867623, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41128 + }, + { + "epoch": 0.41129, + "grad_norm": 1.3727693904658724, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 41129 + }, + { + "epoch": 0.4113, + "grad_norm": 1.8920917932478751, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 41130 + }, + { + "epoch": 0.41131, + "grad_norm": 1.5649476886506932, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 41131 + }, + { + "epoch": 0.41132, + "grad_norm": 1.2189853342788293, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41132 + }, + { + "epoch": 0.41133, + "grad_norm": 1.6235926219482388, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 41133 + }, + { + "epoch": 0.41134, + "grad_norm": 1.3985855537302485, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 41134 + }, + { + "epoch": 0.41135, + "grad_norm": 1.6851715221313128, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 41135 + }, + { + "epoch": 0.41136, + "grad_norm": 1.450976166613318, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41136 + }, + { + "epoch": 0.41137, + "grad_norm": 1.682584783868243, + "learning_rate": 0.003, + "loss": 4.016, + "step": 41137 + }, + { + "epoch": 0.41138, + "grad_norm": 1.410648298662724, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 41138 + }, + { + "epoch": 0.41139, + "grad_norm": 1.7035524643212339, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 41139 + }, + { + "epoch": 0.4114, + "grad_norm": 1.5337212929131563, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 41140 + }, + { + "epoch": 0.41141, + "grad_norm": 1.4166168188916353, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41141 + }, + { + "epoch": 0.41142, + "grad_norm": 1.4788282322697663, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 41142 + }, + { + "epoch": 0.41143, + "grad_norm": 1.2808123661141677, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 41143 + }, + { + "epoch": 0.41144, + "grad_norm": 1.6325760569987466, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41144 + }, + { + "epoch": 0.41145, + "grad_norm": 1.2745984992697825, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 41145 + }, + { + "epoch": 0.41146, + "grad_norm": 1.671019594887441, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41146 + }, + { + "epoch": 0.41147, + "grad_norm": 1.2743957084837692, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 41147 + }, + { + "epoch": 0.41148, + "grad_norm": 2.173743633886764, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 41148 + }, + { + "epoch": 0.41149, + "grad_norm": 1.5123917894924122, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 41149 + }, + { + "epoch": 0.4115, + "grad_norm": 1.3521757854085912, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41150 + }, + { + "epoch": 0.41151, + "grad_norm": 1.5818296141503616, + "learning_rate": 0.003, + "loss": 4.048, + "step": 41151 + }, + { + "epoch": 0.41152, + "grad_norm": 1.6881894198551848, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 41152 + }, + { + "epoch": 0.41153, + "grad_norm": 1.4986683259749385, + "learning_rate": 0.003, + "loss": 4.041, + "step": 41153 + }, + { + "epoch": 0.41154, + "grad_norm": 1.397512301661659, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 41154 + }, + { + "epoch": 0.41155, + "grad_norm": 1.3828188942505482, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41155 + }, + { + "epoch": 0.41156, + "grad_norm": 1.5462801652635205, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 41156 + }, + { + "epoch": 0.41157, + "grad_norm": 1.5517720055634343, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 41157 + }, + { + "epoch": 0.41158, + "grad_norm": 1.645535232442291, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 41158 + }, + { + "epoch": 0.41159, + "grad_norm": 1.651079091377288, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 41159 + }, + { + "epoch": 0.4116, + "grad_norm": 1.3953062502054905, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 41160 + }, + { + "epoch": 0.41161, + "grad_norm": 1.6677138610119528, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 41161 + }, + { + "epoch": 0.41162, + "grad_norm": 1.5355029089995624, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 41162 + }, + { + "epoch": 0.41163, + "grad_norm": 1.2704154746189185, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41163 + }, + { + "epoch": 0.41164, + "grad_norm": 1.5751910384602037, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41164 + }, + { + "epoch": 0.41165, + "grad_norm": 1.3193592762378634, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 41165 + }, + { + "epoch": 0.41166, + "grad_norm": 1.765417995221808, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 41166 + }, + { + "epoch": 0.41167, + "grad_norm": 1.495761175413887, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 41167 + }, + { + "epoch": 0.41168, + "grad_norm": 1.949913451027607, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 41168 + }, + { + "epoch": 0.41169, + "grad_norm": 1.1739173910043383, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41169 + }, + { + "epoch": 0.4117, + "grad_norm": 1.925184086526412, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 41170 + }, + { + "epoch": 0.41171, + "grad_norm": 1.7104469157942732, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 41171 + }, + { + "epoch": 0.41172, + "grad_norm": 1.354558101261012, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 41172 + }, + { + "epoch": 0.41173, + "grad_norm": 1.4494667617884525, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41173 + }, + { + "epoch": 0.41174, + "grad_norm": 1.7277456679439032, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 41174 + }, + { + "epoch": 0.41175, + "grad_norm": 1.4252489946444202, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 41175 + }, + { + "epoch": 0.41176, + "grad_norm": 1.6034238089465185, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41176 + }, + { + "epoch": 0.41177, + "grad_norm": 1.3179667451143202, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 41177 + }, + { + "epoch": 0.41178, + "grad_norm": 1.6815915067496106, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 41178 + }, + { + "epoch": 0.41179, + "grad_norm": 1.4124767379426455, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 41179 + }, + { + "epoch": 0.4118, + "grad_norm": 1.650007354365556, + "learning_rate": 0.003, + "loss": 4.038, + "step": 41180 + }, + { + "epoch": 0.41181, + "grad_norm": 1.3899780524851806, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 41181 + }, + { + "epoch": 0.41182, + "grad_norm": 1.7942246634051897, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 41182 + }, + { + "epoch": 0.41183, + "grad_norm": 1.4642150841639219, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41183 + }, + { + "epoch": 0.41184, + "grad_norm": 1.6390002779485637, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 41184 + }, + { + "epoch": 0.41185, + "grad_norm": 1.4844749511915072, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 41185 + }, + { + "epoch": 0.41186, + "grad_norm": 1.4278045567452555, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 41186 + }, + { + "epoch": 0.41187, + "grad_norm": 1.4673068544537167, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41187 + }, + { + "epoch": 0.41188, + "grad_norm": 1.5337409568401825, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 41188 + }, + { + "epoch": 0.41189, + "grad_norm": 1.3605646114990257, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 41189 + }, + { + "epoch": 0.4119, + "grad_norm": 1.4444926647905232, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 41190 + }, + { + "epoch": 0.41191, + "grad_norm": 1.24553650993116, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41191 + }, + { + "epoch": 0.41192, + "grad_norm": 1.7390872863564704, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 41192 + }, + { + "epoch": 0.41193, + "grad_norm": 1.5856146901260644, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 41193 + }, + { + "epoch": 0.41194, + "grad_norm": 1.7920571525544524, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 41194 + }, + { + "epoch": 0.41195, + "grad_norm": 1.3687468660244162, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41195 + }, + { + "epoch": 0.41196, + "grad_norm": 1.7170820887189135, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41196 + }, + { + "epoch": 0.41197, + "grad_norm": 1.6297523839530812, + "learning_rate": 0.003, + "loss": 4.051, + "step": 41197 + }, + { + "epoch": 0.41198, + "grad_norm": 1.244422525599676, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41198 + }, + { + "epoch": 0.41199, + "grad_norm": 1.8074666478790713, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 41199 + }, + { + "epoch": 0.412, + "grad_norm": 1.2535127763405325, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 41200 + }, + { + "epoch": 0.41201, + "grad_norm": 1.8343271682265248, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 41201 + }, + { + "epoch": 0.41202, + "grad_norm": 1.2662057363118, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41202 + }, + { + "epoch": 0.41203, + "grad_norm": 1.5276119244117452, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 41203 + }, + { + "epoch": 0.41204, + "grad_norm": 1.493388778595177, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 41204 + }, + { + "epoch": 0.41205, + "grad_norm": 1.6305768243042351, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 41205 + }, + { + "epoch": 0.41206, + "grad_norm": 1.6123720164528732, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41206 + }, + { + "epoch": 0.41207, + "grad_norm": 1.5906707964429474, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 41207 + }, + { + "epoch": 0.41208, + "grad_norm": 1.762397714980506, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 41208 + }, + { + "epoch": 0.41209, + "grad_norm": 1.4832738288885856, + "learning_rate": 0.003, + "loss": 4.038, + "step": 41209 + }, + { + "epoch": 0.4121, + "grad_norm": 1.4141732837606729, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41210 + }, + { + "epoch": 0.41211, + "grad_norm": 1.5682365482349616, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41211 + }, + { + "epoch": 0.41212, + "grad_norm": 1.3583466995108668, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 41212 + }, + { + "epoch": 0.41213, + "grad_norm": 1.8786444686088009, + "learning_rate": 0.003, + "loss": 4.023, + "step": 41213 + }, + { + "epoch": 0.41214, + "grad_norm": 1.1870780822133316, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41214 + }, + { + "epoch": 0.41215, + "grad_norm": 2.1669162370633153, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 41215 + }, + { + "epoch": 0.41216, + "grad_norm": 1.3215519513283642, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 41216 + }, + { + "epoch": 0.41217, + "grad_norm": 1.5168237189208806, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 41217 + }, + { + "epoch": 0.41218, + "grad_norm": 1.5558243941175487, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41218 + }, + { + "epoch": 0.41219, + "grad_norm": 1.6627724993748965, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 41219 + }, + { + "epoch": 0.4122, + "grad_norm": 1.4892023076340286, + "learning_rate": 0.003, + "loss": 4.037, + "step": 41220 + }, + { + "epoch": 0.41221, + "grad_norm": 1.5170495507595154, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 41221 + }, + { + "epoch": 0.41222, + "grad_norm": 1.5127129530344685, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 41222 + }, + { + "epoch": 0.41223, + "grad_norm": 1.4274182958355988, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 41223 + }, + { + "epoch": 0.41224, + "grad_norm": 1.537145628360535, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41224 + }, + { + "epoch": 0.41225, + "grad_norm": 1.476550217231673, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41225 + }, + { + "epoch": 0.41226, + "grad_norm": 1.365610508957827, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41226 + }, + { + "epoch": 0.41227, + "grad_norm": 1.4246914144997491, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 41227 + }, + { + "epoch": 0.41228, + "grad_norm": 1.6302832800684905, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41228 + }, + { + "epoch": 0.41229, + "grad_norm": 1.3139521443289182, + "learning_rate": 0.003, + "loss": 4.05, + "step": 41229 + }, + { + "epoch": 0.4123, + "grad_norm": 1.4223252916824998, + "learning_rate": 0.003, + "loss": 4.012, + "step": 41230 + }, + { + "epoch": 0.41231, + "grad_norm": 1.3465366423626484, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41231 + }, + { + "epoch": 0.41232, + "grad_norm": 1.7026922457825007, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 41232 + }, + { + "epoch": 0.41233, + "grad_norm": 1.5084932406819351, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 41233 + }, + { + "epoch": 0.41234, + "grad_norm": 1.359097360450121, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41234 + }, + { + "epoch": 0.41235, + "grad_norm": 1.5867501536067716, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41235 + }, + { + "epoch": 0.41236, + "grad_norm": 1.2115924332267227, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41236 + }, + { + "epoch": 0.41237, + "grad_norm": 1.9245869198314642, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41237 + }, + { + "epoch": 0.41238, + "grad_norm": 1.0895737232866531, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 41238 + }, + { + "epoch": 0.41239, + "grad_norm": 2.1106031957001856, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41239 + }, + { + "epoch": 0.4124, + "grad_norm": 1.4078772946066453, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 41240 + }, + { + "epoch": 0.41241, + "grad_norm": 1.5395884661438073, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 41241 + }, + { + "epoch": 0.41242, + "grad_norm": 1.4769754067442833, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 41242 + }, + { + "epoch": 0.41243, + "grad_norm": 1.9265336721365145, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 41243 + }, + { + "epoch": 0.41244, + "grad_norm": 1.6482778546392438, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41244 + }, + { + "epoch": 0.41245, + "grad_norm": 1.5201260346316114, + "learning_rate": 0.003, + "loss": 4.041, + "step": 41245 + }, + { + "epoch": 0.41246, + "grad_norm": 1.6016164710589729, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41246 + }, + { + "epoch": 0.41247, + "grad_norm": 1.3697511372222222, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41247 + }, + { + "epoch": 0.41248, + "grad_norm": 1.6153241583046998, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 41248 + }, + { + "epoch": 0.41249, + "grad_norm": 1.2627222398641043, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 41249 + }, + { + "epoch": 0.4125, + "grad_norm": 1.5525623956359516, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41250 + }, + { + "epoch": 0.41251, + "grad_norm": 1.302354439686703, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 41251 + }, + { + "epoch": 0.41252, + "grad_norm": 1.6110184470016793, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 41252 + }, + { + "epoch": 0.41253, + "grad_norm": 1.5392705781241915, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 41253 + }, + { + "epoch": 0.41254, + "grad_norm": 1.585413255676591, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41254 + }, + { + "epoch": 0.41255, + "grad_norm": 1.3341534259536552, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41255 + }, + { + "epoch": 0.41256, + "grad_norm": 1.4057773825281112, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 41256 + }, + { + "epoch": 0.41257, + "grad_norm": 1.9529591770632884, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 41257 + }, + { + "epoch": 0.41258, + "grad_norm": 1.4516034338602037, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 41258 + }, + { + "epoch": 0.41259, + "grad_norm": 1.358146982350784, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 41259 + }, + { + "epoch": 0.4126, + "grad_norm": 1.5059462263370118, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 41260 + }, + { + "epoch": 0.41261, + "grad_norm": 1.4129316704572354, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 41261 + }, + { + "epoch": 0.41262, + "grad_norm": 1.5031779565426282, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41262 + }, + { + "epoch": 0.41263, + "grad_norm": 1.4502640093554735, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41263 + }, + { + "epoch": 0.41264, + "grad_norm": 1.5482883273981938, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41264 + }, + { + "epoch": 0.41265, + "grad_norm": 1.459222848816595, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 41265 + }, + { + "epoch": 0.41266, + "grad_norm": 1.3250112069306585, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 41266 + }, + { + "epoch": 0.41267, + "grad_norm": 1.6222373873142928, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41267 + }, + { + "epoch": 0.41268, + "grad_norm": 1.5154035716520613, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41268 + }, + { + "epoch": 0.41269, + "grad_norm": 1.7448272398093656, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 41269 + }, + { + "epoch": 0.4127, + "grad_norm": 1.2731411111620556, + "learning_rate": 0.003, + "loss": 4.04, + "step": 41270 + }, + { + "epoch": 0.41271, + "grad_norm": 1.902859434188703, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41271 + }, + { + "epoch": 0.41272, + "grad_norm": 1.4869826157666202, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 41272 + }, + { + "epoch": 0.41273, + "grad_norm": 1.5471234077472764, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 41273 + }, + { + "epoch": 0.41274, + "grad_norm": 1.4281989701742386, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 41274 + }, + { + "epoch": 0.41275, + "grad_norm": 1.536219376599827, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 41275 + }, + { + "epoch": 0.41276, + "grad_norm": 1.3650601746674875, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 41276 + }, + { + "epoch": 0.41277, + "grad_norm": 1.5409745615183208, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 41277 + }, + { + "epoch": 0.41278, + "grad_norm": 1.4451104771328322, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 41278 + }, + { + "epoch": 0.41279, + "grad_norm": 1.6478078633515916, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 41279 + }, + { + "epoch": 0.4128, + "grad_norm": 1.5205948688131958, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41280 + }, + { + "epoch": 0.41281, + "grad_norm": 1.8339456964583654, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41281 + }, + { + "epoch": 0.41282, + "grad_norm": 2.1015212780482253, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41282 + }, + { + "epoch": 0.41283, + "grad_norm": 1.4342161391937318, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 41283 + }, + { + "epoch": 0.41284, + "grad_norm": 1.6900752100532301, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 41284 + }, + { + "epoch": 0.41285, + "grad_norm": 1.4604173534125737, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41285 + }, + { + "epoch": 0.41286, + "grad_norm": 1.4876955072154645, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41286 + }, + { + "epoch": 0.41287, + "grad_norm": 1.5403036183859118, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 41287 + }, + { + "epoch": 0.41288, + "grad_norm": 1.4377708881951068, + "learning_rate": 0.003, + "loss": 4.067, + "step": 41288 + }, + { + "epoch": 0.41289, + "grad_norm": 1.5609902728515406, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 41289 + }, + { + "epoch": 0.4129, + "grad_norm": 1.3701053099140441, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 41290 + }, + { + "epoch": 0.41291, + "grad_norm": 1.4338282190089866, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 41291 + }, + { + "epoch": 0.41292, + "grad_norm": 1.488631982434633, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 41292 + }, + { + "epoch": 0.41293, + "grad_norm": 1.4775669462893528, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 41293 + }, + { + "epoch": 0.41294, + "grad_norm": 1.4554905331629984, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 41294 + }, + { + "epoch": 0.41295, + "grad_norm": 1.3937350064149054, + "learning_rate": 0.003, + "loss": 4.037, + "step": 41295 + }, + { + "epoch": 0.41296, + "grad_norm": 1.3589287123617089, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 41296 + }, + { + "epoch": 0.41297, + "grad_norm": 1.6522704321054407, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 41297 + }, + { + "epoch": 0.41298, + "grad_norm": 1.4839790656545329, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 41298 + }, + { + "epoch": 0.41299, + "grad_norm": 1.23272848824502, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 41299 + }, + { + "epoch": 0.413, + "grad_norm": 1.6279624226245712, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 41300 + }, + { + "epoch": 0.41301, + "grad_norm": 1.4574828199542127, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 41301 + }, + { + "epoch": 0.41302, + "grad_norm": 1.7467501455398409, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 41302 + }, + { + "epoch": 0.41303, + "grad_norm": 1.9996815995978494, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 41303 + }, + { + "epoch": 0.41304, + "grad_norm": 1.1554536860754214, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41304 + }, + { + "epoch": 0.41305, + "grad_norm": 1.642349228401469, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 41305 + }, + { + "epoch": 0.41306, + "grad_norm": 1.510381786102597, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 41306 + }, + { + "epoch": 0.41307, + "grad_norm": 1.9511902583292784, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 41307 + }, + { + "epoch": 0.41308, + "grad_norm": 1.4936159261883761, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 41308 + }, + { + "epoch": 0.41309, + "grad_norm": 1.7389044113802594, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 41309 + }, + { + "epoch": 0.4131, + "grad_norm": 1.215254335042484, + "learning_rate": 0.003, + "loss": 4.05, + "step": 41310 + }, + { + "epoch": 0.41311, + "grad_norm": 1.7335652130191583, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41311 + }, + { + "epoch": 0.41312, + "grad_norm": 1.468265208190123, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 41312 + }, + { + "epoch": 0.41313, + "grad_norm": 1.527864380308709, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41313 + }, + { + "epoch": 0.41314, + "grad_norm": 1.572083691468712, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 41314 + }, + { + "epoch": 0.41315, + "grad_norm": 1.4655317301860022, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 41315 + }, + { + "epoch": 0.41316, + "grad_norm": 1.4530639517218802, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 41316 + }, + { + "epoch": 0.41317, + "grad_norm": 1.3825261822505788, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 41317 + }, + { + "epoch": 0.41318, + "grad_norm": 1.456844174425141, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 41318 + }, + { + "epoch": 0.41319, + "grad_norm": 1.706241482709638, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 41319 + }, + { + "epoch": 0.4132, + "grad_norm": 1.7316719331463344, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41320 + }, + { + "epoch": 0.41321, + "grad_norm": 1.73653184845402, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 41321 + }, + { + "epoch": 0.41322, + "grad_norm": 1.5951244138573422, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41322 + }, + { + "epoch": 0.41323, + "grad_norm": 1.2477813356892207, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41323 + }, + { + "epoch": 0.41324, + "grad_norm": 1.7587139398874743, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41324 + }, + { + "epoch": 0.41325, + "grad_norm": 1.4786544335046288, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 41325 + }, + { + "epoch": 0.41326, + "grad_norm": 1.5470822742911763, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 41326 + }, + { + "epoch": 0.41327, + "grad_norm": 1.6193083256758174, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41327 + }, + { + "epoch": 0.41328, + "grad_norm": 1.591016097390931, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 41328 + }, + { + "epoch": 0.41329, + "grad_norm": 1.375599911237885, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 41329 + }, + { + "epoch": 0.4133, + "grad_norm": 1.7043730887207968, + "learning_rate": 0.003, + "loss": 4.047, + "step": 41330 + }, + { + "epoch": 0.41331, + "grad_norm": 1.544600493803851, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41331 + }, + { + "epoch": 0.41332, + "grad_norm": 1.4099180309416932, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41332 + }, + { + "epoch": 0.41333, + "grad_norm": 1.5927905435099512, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41333 + }, + { + "epoch": 0.41334, + "grad_norm": 1.38665743568607, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 41334 + }, + { + "epoch": 0.41335, + "grad_norm": 1.5906669263390267, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41335 + }, + { + "epoch": 0.41336, + "grad_norm": 1.4346530263674335, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 41336 + }, + { + "epoch": 0.41337, + "grad_norm": 1.498081141991213, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 41337 + }, + { + "epoch": 0.41338, + "grad_norm": 1.4851335203874847, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 41338 + }, + { + "epoch": 0.41339, + "grad_norm": 1.5466962783575076, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 41339 + }, + { + "epoch": 0.4134, + "grad_norm": 1.4354456937822602, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 41340 + }, + { + "epoch": 0.41341, + "grad_norm": 1.7557658856952034, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 41341 + }, + { + "epoch": 0.41342, + "grad_norm": 1.5900001203303764, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 41342 + }, + { + "epoch": 0.41343, + "grad_norm": 1.8024738849441087, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41343 + }, + { + "epoch": 0.41344, + "grad_norm": 1.4273168666162825, + "learning_rate": 0.003, + "loss": 4.029, + "step": 41344 + }, + { + "epoch": 0.41345, + "grad_norm": 1.4174931732721106, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 41345 + }, + { + "epoch": 0.41346, + "grad_norm": 1.5496553823269552, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 41346 + }, + { + "epoch": 0.41347, + "grad_norm": 1.3802212562442528, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 41347 + }, + { + "epoch": 0.41348, + "grad_norm": 1.8063794778296822, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 41348 + }, + { + "epoch": 0.41349, + "grad_norm": 1.4319698168959167, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 41349 + }, + { + "epoch": 0.4135, + "grad_norm": 1.8034337651354981, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41350 + }, + { + "epoch": 0.41351, + "grad_norm": 1.4857100275895445, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41351 + }, + { + "epoch": 0.41352, + "grad_norm": 1.4160911831334457, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 41352 + }, + { + "epoch": 0.41353, + "grad_norm": 1.5135203576492644, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 41353 + }, + { + "epoch": 0.41354, + "grad_norm": 1.3730009693556016, + "learning_rate": 0.003, + "loss": 4.048, + "step": 41354 + }, + { + "epoch": 0.41355, + "grad_norm": 1.4570311262835847, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41355 + }, + { + "epoch": 0.41356, + "grad_norm": 1.5552601822287622, + "learning_rate": 0.003, + "loss": 4.048, + "step": 41356 + }, + { + "epoch": 0.41357, + "grad_norm": 1.4671393299360933, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41357 + }, + { + "epoch": 0.41358, + "grad_norm": 1.4873984604648811, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41358 + }, + { + "epoch": 0.41359, + "grad_norm": 1.423554583123364, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41359 + }, + { + "epoch": 0.4136, + "grad_norm": 1.7586432796553073, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 41360 + }, + { + "epoch": 0.41361, + "grad_norm": 1.4236948733764136, + "learning_rate": 0.003, + "loss": 4.011, + "step": 41361 + }, + { + "epoch": 0.41362, + "grad_norm": 1.7845065689999011, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 41362 + }, + { + "epoch": 0.41363, + "grad_norm": 1.3923126459858985, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 41363 + }, + { + "epoch": 0.41364, + "grad_norm": 1.544621692160637, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 41364 + }, + { + "epoch": 0.41365, + "grad_norm": 1.5922076877054572, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 41365 + }, + { + "epoch": 0.41366, + "grad_norm": 1.3217467201313566, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41366 + }, + { + "epoch": 0.41367, + "grad_norm": 1.4616047468249833, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 41367 + }, + { + "epoch": 0.41368, + "grad_norm": 1.6653741081020201, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41368 + }, + { + "epoch": 0.41369, + "grad_norm": 1.5533003598293933, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 41369 + }, + { + "epoch": 0.4137, + "grad_norm": 1.394977343133052, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 41370 + }, + { + "epoch": 0.41371, + "grad_norm": 1.5176530913535133, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 41371 + }, + { + "epoch": 0.41372, + "grad_norm": 1.5756053168196262, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 41372 + }, + { + "epoch": 0.41373, + "grad_norm": 1.2974770534091997, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41373 + }, + { + "epoch": 0.41374, + "grad_norm": 1.665362514774941, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 41374 + }, + { + "epoch": 0.41375, + "grad_norm": 1.3869393673318111, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 41375 + }, + { + "epoch": 0.41376, + "grad_norm": 1.6254376918657372, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41376 + }, + { + "epoch": 0.41377, + "grad_norm": 1.3096235319796743, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 41377 + }, + { + "epoch": 0.41378, + "grad_norm": 1.6730008207463625, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 41378 + }, + { + "epoch": 0.41379, + "grad_norm": 1.463128346277146, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 41379 + }, + { + "epoch": 0.4138, + "grad_norm": 1.8342183122006117, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 41380 + }, + { + "epoch": 0.41381, + "grad_norm": 1.4013593911484041, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 41381 + }, + { + "epoch": 0.41382, + "grad_norm": 1.5793171829069361, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41382 + }, + { + "epoch": 0.41383, + "grad_norm": 1.362171621908038, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 41383 + }, + { + "epoch": 0.41384, + "grad_norm": 1.5506436053367545, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 41384 + }, + { + "epoch": 0.41385, + "grad_norm": 1.4073893208336297, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 41385 + }, + { + "epoch": 0.41386, + "grad_norm": 1.5766162033309794, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 41386 + }, + { + "epoch": 0.41387, + "grad_norm": 1.4325781606594263, + "learning_rate": 0.003, + "loss": 4.049, + "step": 41387 + }, + { + "epoch": 0.41388, + "grad_norm": 1.7329533485773048, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 41388 + }, + { + "epoch": 0.41389, + "grad_norm": 1.3753147819118403, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 41389 + }, + { + "epoch": 0.4139, + "grad_norm": 1.6882606007366519, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 41390 + }, + { + "epoch": 0.41391, + "grad_norm": 1.4100627372593597, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 41391 + }, + { + "epoch": 0.41392, + "grad_norm": 1.6051053987608317, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 41392 + }, + { + "epoch": 0.41393, + "grad_norm": 1.5619031410311834, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 41393 + }, + { + "epoch": 0.41394, + "grad_norm": 1.2873312563747594, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 41394 + }, + { + "epoch": 0.41395, + "grad_norm": 1.7374842033291242, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 41395 + }, + { + "epoch": 0.41396, + "grad_norm": 1.7026595890984653, + "learning_rate": 0.003, + "loss": 4.061, + "step": 41396 + }, + { + "epoch": 0.41397, + "grad_norm": 1.6683920684355884, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 41397 + }, + { + "epoch": 0.41398, + "grad_norm": 1.875671475175397, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 41398 + }, + { + "epoch": 0.41399, + "grad_norm": 1.6572413265164923, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 41399 + }, + { + "epoch": 0.414, + "grad_norm": 1.586703834486052, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41400 + }, + { + "epoch": 0.41401, + "grad_norm": 1.5371750777846604, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41401 + }, + { + "epoch": 0.41402, + "grad_norm": 1.4410188068407455, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41402 + }, + { + "epoch": 0.41403, + "grad_norm": 1.485860482612688, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 41403 + }, + { + "epoch": 0.41404, + "grad_norm": 1.698444844406903, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 41404 + }, + { + "epoch": 0.41405, + "grad_norm": 1.3894771301387938, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41405 + }, + { + "epoch": 0.41406, + "grad_norm": 1.532091386520595, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 41406 + }, + { + "epoch": 0.41407, + "grad_norm": 1.7873737395758211, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 41407 + }, + { + "epoch": 0.41408, + "grad_norm": 1.12433663377292, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 41408 + }, + { + "epoch": 0.41409, + "grad_norm": 1.578482263681388, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 41409 + }, + { + "epoch": 0.4141, + "grad_norm": 1.5266175020184802, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 41410 + }, + { + "epoch": 0.41411, + "grad_norm": 1.7799960535932604, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 41411 + }, + { + "epoch": 0.41412, + "grad_norm": 1.3940835702757028, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 41412 + }, + { + "epoch": 0.41413, + "grad_norm": 1.9101002197404777, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 41413 + }, + { + "epoch": 0.41414, + "grad_norm": 1.2893455286490954, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41414 + }, + { + "epoch": 0.41415, + "grad_norm": 1.5732423503043078, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41415 + }, + { + "epoch": 0.41416, + "grad_norm": 1.6318288734352833, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 41416 + }, + { + "epoch": 0.41417, + "grad_norm": 1.572208956204967, + "learning_rate": 0.003, + "loss": 3.997, + "step": 41417 + }, + { + "epoch": 0.41418, + "grad_norm": 1.2717337696805509, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 41418 + }, + { + "epoch": 0.41419, + "grad_norm": 1.7777638054805975, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 41419 + }, + { + "epoch": 0.4142, + "grad_norm": 1.346063065427264, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 41420 + }, + { + "epoch": 0.41421, + "grad_norm": 1.6055311029920556, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 41421 + }, + { + "epoch": 0.41422, + "grad_norm": 1.2864783216771927, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 41422 + }, + { + "epoch": 0.41423, + "grad_norm": 1.6997332706168296, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 41423 + }, + { + "epoch": 0.41424, + "grad_norm": 1.6669169414624927, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 41424 + }, + { + "epoch": 0.41425, + "grad_norm": 1.408633661081413, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 41425 + }, + { + "epoch": 0.41426, + "grad_norm": 1.6193388918780574, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 41426 + }, + { + "epoch": 0.41427, + "grad_norm": 1.6031415431298688, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 41427 + }, + { + "epoch": 0.41428, + "grad_norm": 1.447058943879896, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 41428 + }, + { + "epoch": 0.41429, + "grad_norm": 1.233402746679035, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 41429 + }, + { + "epoch": 0.4143, + "grad_norm": 1.8214702475639113, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 41430 + }, + { + "epoch": 0.41431, + "grad_norm": 1.3348912860155593, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 41431 + }, + { + "epoch": 0.41432, + "grad_norm": 1.9349487304045911, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 41432 + }, + { + "epoch": 0.41433, + "grad_norm": 1.6309494487973601, + "learning_rate": 0.003, + "loss": 4.079, + "step": 41433 + }, + { + "epoch": 0.41434, + "grad_norm": 1.560684530941979, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 41434 + }, + { + "epoch": 0.41435, + "grad_norm": 1.5332507313479602, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41435 + }, + { + "epoch": 0.41436, + "grad_norm": 1.666708605635523, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 41436 + }, + { + "epoch": 0.41437, + "grad_norm": 1.3311273818213147, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 41437 + }, + { + "epoch": 0.41438, + "grad_norm": 1.4074503605401796, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 41438 + }, + { + "epoch": 0.41439, + "grad_norm": 1.6871236655991557, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41439 + }, + { + "epoch": 0.4144, + "grad_norm": 1.3590256840543802, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41440 + }, + { + "epoch": 0.41441, + "grad_norm": 1.855718249153128, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 41441 + }, + { + "epoch": 0.41442, + "grad_norm": 1.5839801369271536, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41442 + }, + { + "epoch": 0.41443, + "grad_norm": 1.6525243854340679, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 41443 + }, + { + "epoch": 0.41444, + "grad_norm": 1.7502404944584462, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 41444 + }, + { + "epoch": 0.41445, + "grad_norm": 1.476004111353999, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 41445 + }, + { + "epoch": 0.41446, + "grad_norm": 1.5692132448645022, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41446 + }, + { + "epoch": 0.41447, + "grad_norm": 1.3680858353706344, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 41447 + }, + { + "epoch": 0.41448, + "grad_norm": 1.4801852772395525, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 41448 + }, + { + "epoch": 0.41449, + "grad_norm": 1.368942363076511, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 41449 + }, + { + "epoch": 0.4145, + "grad_norm": 1.971739796331447, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 41450 + }, + { + "epoch": 0.41451, + "grad_norm": 1.1389942053506983, + "learning_rate": 0.003, + "loss": 4.015, + "step": 41451 + }, + { + "epoch": 0.41452, + "grad_norm": 1.6293749557272066, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 41452 + }, + { + "epoch": 0.41453, + "grad_norm": 1.2972181960044071, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41453 + }, + { + "epoch": 0.41454, + "grad_norm": 1.7334744845464818, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 41454 + }, + { + "epoch": 0.41455, + "grad_norm": 1.4181941545562526, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 41455 + }, + { + "epoch": 0.41456, + "grad_norm": 1.6397766469054353, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 41456 + }, + { + "epoch": 0.41457, + "grad_norm": 1.3480541228383331, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 41457 + }, + { + "epoch": 0.41458, + "grad_norm": 1.7159354201470436, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 41458 + }, + { + "epoch": 0.41459, + "grad_norm": 1.3161230712898557, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 41459 + }, + { + "epoch": 0.4146, + "grad_norm": 1.7200102182083603, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 41460 + }, + { + "epoch": 0.41461, + "grad_norm": 1.4492453258376392, + "learning_rate": 0.003, + "loss": 4.072, + "step": 41461 + }, + { + "epoch": 0.41462, + "grad_norm": 1.4889943913435177, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 41462 + }, + { + "epoch": 0.41463, + "grad_norm": 1.6821470502638411, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41463 + }, + { + "epoch": 0.41464, + "grad_norm": 1.7085614518815582, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 41464 + }, + { + "epoch": 0.41465, + "grad_norm": 1.7159489856401466, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 41465 + }, + { + "epoch": 0.41466, + "grad_norm": 1.4772020779640256, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 41466 + }, + { + "epoch": 0.41467, + "grad_norm": 1.470808765265056, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 41467 + }, + { + "epoch": 0.41468, + "grad_norm": 1.592006966927438, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 41468 + }, + { + "epoch": 0.41469, + "grad_norm": 1.4696811323831327, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 41469 + }, + { + "epoch": 0.4147, + "grad_norm": 1.5558757842284316, + "learning_rate": 0.003, + "loss": 4.021, + "step": 41470 + }, + { + "epoch": 0.41471, + "grad_norm": 1.464275464999203, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 41471 + }, + { + "epoch": 0.41472, + "grad_norm": 1.5130853803739766, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 41472 + }, + { + "epoch": 0.41473, + "grad_norm": 1.720834178658874, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 41473 + }, + { + "epoch": 0.41474, + "grad_norm": 1.5133455176832415, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41474 + }, + { + "epoch": 0.41475, + "grad_norm": 1.3900506140165474, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 41475 + }, + { + "epoch": 0.41476, + "grad_norm": 1.476698437212612, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41476 + }, + { + "epoch": 0.41477, + "grad_norm": 1.5035595573145941, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 41477 + }, + { + "epoch": 0.41478, + "grad_norm": 1.3478818755556055, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 41478 + }, + { + "epoch": 0.41479, + "grad_norm": 1.4614441039584287, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 41479 + }, + { + "epoch": 0.4148, + "grad_norm": 1.6381961828447886, + "learning_rate": 0.003, + "loss": 4.049, + "step": 41480 + }, + { + "epoch": 0.41481, + "grad_norm": 1.492256340579706, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 41481 + }, + { + "epoch": 0.41482, + "grad_norm": 1.4538033192345066, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 41482 + }, + { + "epoch": 0.41483, + "grad_norm": 1.4662430446156756, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 41483 + }, + { + "epoch": 0.41484, + "grad_norm": 1.6361630309635007, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41484 + }, + { + "epoch": 0.41485, + "grad_norm": 1.6439443941748253, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41485 + }, + { + "epoch": 0.41486, + "grad_norm": 1.2877225200805773, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41486 + }, + { + "epoch": 0.41487, + "grad_norm": 1.6249808884317223, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 41487 + }, + { + "epoch": 0.41488, + "grad_norm": 1.884859005724997, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 41488 + }, + { + "epoch": 0.41489, + "grad_norm": 1.4765965141064499, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41489 + }, + { + "epoch": 0.4149, + "grad_norm": 1.698045847411333, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 41490 + }, + { + "epoch": 0.41491, + "grad_norm": 1.446720844416518, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 41491 + }, + { + "epoch": 0.41492, + "grad_norm": 1.9019672829920062, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 41492 + }, + { + "epoch": 0.41493, + "grad_norm": 1.313592855799461, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 41493 + }, + { + "epoch": 0.41494, + "grad_norm": 1.6592412901493554, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 41494 + }, + { + "epoch": 0.41495, + "grad_norm": 1.3938867945667002, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41495 + }, + { + "epoch": 0.41496, + "grad_norm": 1.441989567995528, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 41496 + }, + { + "epoch": 0.41497, + "grad_norm": 1.717442438983677, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41497 + }, + { + "epoch": 0.41498, + "grad_norm": 1.4631653151662434, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 41498 + }, + { + "epoch": 0.41499, + "grad_norm": 1.4304926036334285, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 41499 + }, + { + "epoch": 0.415, + "grad_norm": 1.6784055212972346, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41500 + }, + { + "epoch": 0.41501, + "grad_norm": 1.4302532292123569, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41501 + }, + { + "epoch": 0.41502, + "grad_norm": 1.7749382605725998, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41502 + }, + { + "epoch": 0.41503, + "grad_norm": 1.4912618277414766, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 41503 + }, + { + "epoch": 0.41504, + "grad_norm": 1.8259391736846908, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 41504 + }, + { + "epoch": 0.41505, + "grad_norm": 1.2461292656244531, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 41505 + }, + { + "epoch": 0.41506, + "grad_norm": 1.8458001461583773, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 41506 + }, + { + "epoch": 0.41507, + "grad_norm": 1.468661928706883, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 41507 + }, + { + "epoch": 0.41508, + "grad_norm": 1.4799607434738766, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41508 + }, + { + "epoch": 0.41509, + "grad_norm": 1.4658687511959887, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41509 + }, + { + "epoch": 0.4151, + "grad_norm": 1.4711269719233835, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 41510 + }, + { + "epoch": 0.41511, + "grad_norm": 1.4983036993783412, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41511 + }, + { + "epoch": 0.41512, + "grad_norm": 1.511436280616719, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 41512 + }, + { + "epoch": 0.41513, + "grad_norm": 1.5973947650568838, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 41513 + }, + { + "epoch": 0.41514, + "grad_norm": 1.5246322612409968, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41514 + }, + { + "epoch": 0.41515, + "grad_norm": 1.4427703020002638, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 41515 + }, + { + "epoch": 0.41516, + "grad_norm": 1.6793492381363906, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 41516 + }, + { + "epoch": 0.41517, + "grad_norm": 1.4179705070802493, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 41517 + }, + { + "epoch": 0.41518, + "grad_norm": 1.5656221455794508, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 41518 + }, + { + "epoch": 0.41519, + "grad_norm": 1.741845931017258, + "learning_rate": 0.003, + "loss": 4.037, + "step": 41519 + }, + { + "epoch": 0.4152, + "grad_norm": 1.5945121468218668, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41520 + }, + { + "epoch": 0.41521, + "grad_norm": 1.4822617724567484, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 41521 + }, + { + "epoch": 0.41522, + "grad_norm": 1.2823444661551897, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 41522 + }, + { + "epoch": 0.41523, + "grad_norm": 1.7190234659297172, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41523 + }, + { + "epoch": 0.41524, + "grad_norm": 1.3168103396045208, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 41524 + }, + { + "epoch": 0.41525, + "grad_norm": 1.74787607637339, + "learning_rate": 0.003, + "loss": 4.047, + "step": 41525 + }, + { + "epoch": 0.41526, + "grad_norm": 1.2776980864748952, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 41526 + }, + { + "epoch": 0.41527, + "grad_norm": 1.9723979492250991, + "learning_rate": 0.003, + "loss": 4.077, + "step": 41527 + }, + { + "epoch": 0.41528, + "grad_norm": 1.3674735289700726, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 41528 + }, + { + "epoch": 0.41529, + "grad_norm": 1.4666650013740348, + "learning_rate": 0.003, + "loss": 4.003, + "step": 41529 + }, + { + "epoch": 0.4153, + "grad_norm": 1.7926392060273197, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41530 + }, + { + "epoch": 0.41531, + "grad_norm": 1.9869661171170865, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 41531 + }, + { + "epoch": 0.41532, + "grad_norm": 1.1780222260785016, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41532 + }, + { + "epoch": 0.41533, + "grad_norm": 1.5498294809465523, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 41533 + }, + { + "epoch": 0.41534, + "grad_norm": 1.4926929344269186, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 41534 + }, + { + "epoch": 0.41535, + "grad_norm": 1.881620163634552, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 41535 + }, + { + "epoch": 0.41536, + "grad_norm": 1.3479560327570397, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 41536 + }, + { + "epoch": 0.41537, + "grad_norm": 1.5065095713955905, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 41537 + }, + { + "epoch": 0.41538, + "grad_norm": 1.716910607599555, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 41538 + }, + { + "epoch": 0.41539, + "grad_norm": 1.1409338358185306, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 41539 + }, + { + "epoch": 0.4154, + "grad_norm": 1.8785232645401586, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 41540 + }, + { + "epoch": 0.41541, + "grad_norm": 1.4013708250868193, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 41541 + }, + { + "epoch": 0.41542, + "grad_norm": 1.7811923044936306, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 41542 + }, + { + "epoch": 0.41543, + "grad_norm": 1.5702163610828068, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 41543 + }, + { + "epoch": 0.41544, + "grad_norm": 1.6735474176818792, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 41544 + }, + { + "epoch": 0.41545, + "grad_norm": 1.7133217169996984, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 41545 + }, + { + "epoch": 0.41546, + "grad_norm": 1.6080543832131173, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 41546 + }, + { + "epoch": 0.41547, + "grad_norm": 1.3594137145534302, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 41547 + }, + { + "epoch": 0.41548, + "grad_norm": 1.9378259179541306, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 41548 + }, + { + "epoch": 0.41549, + "grad_norm": 1.2807987484497292, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41549 + }, + { + "epoch": 0.4155, + "grad_norm": 1.7186718529522247, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41550 + }, + { + "epoch": 0.41551, + "grad_norm": 1.4717958250115828, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41551 + }, + { + "epoch": 0.41552, + "grad_norm": 1.4641447646202537, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 41552 + }, + { + "epoch": 0.41553, + "grad_norm": 1.3591139166037298, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 41553 + }, + { + "epoch": 0.41554, + "grad_norm": 1.380602175889241, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 41554 + }, + { + "epoch": 0.41555, + "grad_norm": 1.606160700357714, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 41555 + }, + { + "epoch": 0.41556, + "grad_norm": 1.3800307443822548, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41556 + }, + { + "epoch": 0.41557, + "grad_norm": 1.6084405552698686, + "learning_rate": 0.003, + "loss": 4.048, + "step": 41557 + }, + { + "epoch": 0.41558, + "grad_norm": 1.6231249109641834, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 41558 + }, + { + "epoch": 0.41559, + "grad_norm": 1.9280876455434892, + "learning_rate": 0.003, + "loss": 4.038, + "step": 41559 + }, + { + "epoch": 0.4156, + "grad_norm": 1.541061084457553, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 41560 + }, + { + "epoch": 0.41561, + "grad_norm": 1.4032387940911086, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 41561 + }, + { + "epoch": 0.41562, + "grad_norm": 1.7393583071513659, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 41562 + }, + { + "epoch": 0.41563, + "grad_norm": 1.5650576196645951, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 41563 + }, + { + "epoch": 0.41564, + "grad_norm": 1.371570344329584, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 41564 + }, + { + "epoch": 0.41565, + "grad_norm": 1.5835677494356937, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 41565 + }, + { + "epoch": 0.41566, + "grad_norm": 1.254080099830128, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 41566 + }, + { + "epoch": 0.41567, + "grad_norm": 1.3611172807031655, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 41567 + }, + { + "epoch": 0.41568, + "grad_norm": 1.447424221759769, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 41568 + }, + { + "epoch": 0.41569, + "grad_norm": 1.4379494369162886, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 41569 + }, + { + "epoch": 0.4157, + "grad_norm": 1.8692594904910835, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41570 + }, + { + "epoch": 0.41571, + "grad_norm": 1.3139264480710549, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 41571 + }, + { + "epoch": 0.41572, + "grad_norm": 1.6102591062281681, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41572 + }, + { + "epoch": 0.41573, + "grad_norm": 1.6015826651032181, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 41573 + }, + { + "epoch": 0.41574, + "grad_norm": 1.8453253065540862, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 41574 + }, + { + "epoch": 0.41575, + "grad_norm": 1.5695169391969892, + "learning_rate": 0.003, + "loss": 4.047, + "step": 41575 + }, + { + "epoch": 0.41576, + "grad_norm": 1.724662918233309, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41576 + }, + { + "epoch": 0.41577, + "grad_norm": 1.2464810789805125, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 41577 + }, + { + "epoch": 0.41578, + "grad_norm": 1.841822591404267, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41578 + }, + { + "epoch": 0.41579, + "grad_norm": 1.34250966045372, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41579 + }, + { + "epoch": 0.4158, + "grad_norm": 1.6139828221919652, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41580 + }, + { + "epoch": 0.41581, + "grad_norm": 1.3427013028633934, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 41581 + }, + { + "epoch": 0.41582, + "grad_norm": 1.7579905753232736, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41582 + }, + { + "epoch": 0.41583, + "grad_norm": 1.2985663995643821, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 41583 + }, + { + "epoch": 0.41584, + "grad_norm": 1.5084316963466788, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 41584 + }, + { + "epoch": 0.41585, + "grad_norm": 1.800037059043795, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 41585 + }, + { + "epoch": 0.41586, + "grad_norm": 1.4647486608379146, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 41586 + }, + { + "epoch": 0.41587, + "grad_norm": 1.544455196813831, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 41587 + }, + { + "epoch": 0.41588, + "grad_norm": 1.6737433652347484, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 41588 + }, + { + "epoch": 0.41589, + "grad_norm": 1.3298588636921413, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 41589 + }, + { + "epoch": 0.4159, + "grad_norm": 1.5952783839850007, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 41590 + }, + { + "epoch": 0.41591, + "grad_norm": 1.3632061109464377, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 41591 + }, + { + "epoch": 0.41592, + "grad_norm": 1.5337255371293654, + "learning_rate": 0.003, + "loss": 4.021, + "step": 41592 + }, + { + "epoch": 0.41593, + "grad_norm": 1.3742916782598151, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 41593 + }, + { + "epoch": 0.41594, + "grad_norm": 1.4777525995034912, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 41594 + }, + { + "epoch": 0.41595, + "grad_norm": 1.692439423603803, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41595 + }, + { + "epoch": 0.41596, + "grad_norm": 1.4643995414731255, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 41596 + }, + { + "epoch": 0.41597, + "grad_norm": 1.528724004713093, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 41597 + }, + { + "epoch": 0.41598, + "grad_norm": 1.8854283850907427, + "learning_rate": 0.003, + "loss": 4.044, + "step": 41598 + }, + { + "epoch": 0.41599, + "grad_norm": 1.5594025864474947, + "learning_rate": 0.003, + "loss": 4.028, + "step": 41599 + }, + { + "epoch": 0.416, + "grad_norm": 1.8194683965532326, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 41600 + }, + { + "epoch": 0.41601, + "grad_norm": 1.3693298042231394, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 41601 + }, + { + "epoch": 0.41602, + "grad_norm": 1.6717722484083468, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 41602 + }, + { + "epoch": 0.41603, + "grad_norm": 1.3563861305994458, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 41603 + }, + { + "epoch": 0.41604, + "grad_norm": 1.554911592228804, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 41604 + }, + { + "epoch": 0.41605, + "grad_norm": 1.3776311153080139, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 41605 + }, + { + "epoch": 0.41606, + "grad_norm": 1.6649471517634316, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 41606 + }, + { + "epoch": 0.41607, + "grad_norm": 1.5322654520030456, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 41607 + }, + { + "epoch": 0.41608, + "grad_norm": 1.6853576362767984, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 41608 + }, + { + "epoch": 0.41609, + "grad_norm": 1.6368395888006042, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41609 + }, + { + "epoch": 0.4161, + "grad_norm": 1.7094879002180021, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41610 + }, + { + "epoch": 0.41611, + "grad_norm": 1.2513286681961626, + "learning_rate": 0.003, + "loss": 4.038, + "step": 41611 + }, + { + "epoch": 0.41612, + "grad_norm": 1.6612437588707452, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41612 + }, + { + "epoch": 0.41613, + "grad_norm": 1.6062438511002246, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 41613 + }, + { + "epoch": 0.41614, + "grad_norm": 1.4928849207760393, + "learning_rate": 0.003, + "loss": 4.046, + "step": 41614 + }, + { + "epoch": 0.41615, + "grad_norm": 1.7036952371790437, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 41615 + }, + { + "epoch": 0.41616, + "grad_norm": 1.4523724694179647, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 41616 + }, + { + "epoch": 0.41617, + "grad_norm": 1.1913459782974631, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41617 + }, + { + "epoch": 0.41618, + "grad_norm": 1.8525859482533396, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41618 + }, + { + "epoch": 0.41619, + "grad_norm": 1.186058932878574, + "learning_rate": 0.003, + "loss": 4.035, + "step": 41619 + }, + { + "epoch": 0.4162, + "grad_norm": 1.6752292456391806, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 41620 + }, + { + "epoch": 0.41621, + "grad_norm": 1.284870265834211, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 41621 + }, + { + "epoch": 0.41622, + "grad_norm": 1.8516532370667753, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 41622 + }, + { + "epoch": 0.41623, + "grad_norm": 1.5742310459776434, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 41623 + }, + { + "epoch": 0.41624, + "grad_norm": 1.378817632597197, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 41624 + }, + { + "epoch": 0.41625, + "grad_norm": 1.3500907305627452, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 41625 + }, + { + "epoch": 0.41626, + "grad_norm": 1.8819744974700556, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41626 + }, + { + "epoch": 0.41627, + "grad_norm": 1.1846862653285835, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41627 + }, + { + "epoch": 0.41628, + "grad_norm": 1.6412963056773786, + "learning_rate": 0.003, + "loss": 4.057, + "step": 41628 + }, + { + "epoch": 0.41629, + "grad_norm": 1.4475733598099523, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 41629 + }, + { + "epoch": 0.4163, + "grad_norm": 1.7028769874211565, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 41630 + }, + { + "epoch": 0.41631, + "grad_norm": 1.7560915466709726, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 41631 + }, + { + "epoch": 0.41632, + "grad_norm": 1.2926632416949058, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 41632 + }, + { + "epoch": 0.41633, + "grad_norm": 1.62628430237712, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 41633 + }, + { + "epoch": 0.41634, + "grad_norm": 1.6428576305617075, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 41634 + }, + { + "epoch": 0.41635, + "grad_norm": 1.546157515070754, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 41635 + }, + { + "epoch": 0.41636, + "grad_norm": 1.5940663868473048, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 41636 + }, + { + "epoch": 0.41637, + "grad_norm": 1.5113541019673193, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 41637 + }, + { + "epoch": 0.41638, + "grad_norm": 1.5421547555312891, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 41638 + }, + { + "epoch": 0.41639, + "grad_norm": 1.3358021371942206, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 41639 + }, + { + "epoch": 0.4164, + "grad_norm": 1.4804018434551571, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 41640 + }, + { + "epoch": 0.41641, + "grad_norm": 1.3555548600058311, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 41641 + }, + { + "epoch": 0.41642, + "grad_norm": 1.7363361135263555, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 41642 + }, + { + "epoch": 0.41643, + "grad_norm": 1.5112942145081998, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 41643 + }, + { + "epoch": 0.41644, + "grad_norm": 1.6696407264542583, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 41644 + }, + { + "epoch": 0.41645, + "grad_norm": 1.321653014960955, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 41645 + }, + { + "epoch": 0.41646, + "grad_norm": 1.5980943498215479, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 41646 + }, + { + "epoch": 0.41647, + "grad_norm": 1.0903031921147994, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 41647 + }, + { + "epoch": 0.41648, + "grad_norm": 1.913964657918247, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 41648 + }, + { + "epoch": 0.41649, + "grad_norm": 1.4362666797238295, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 41649 + }, + { + "epoch": 0.4165, + "grad_norm": 1.277357548377839, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 41650 + }, + { + "epoch": 0.41651, + "grad_norm": 1.6452601060975762, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 41651 + }, + { + "epoch": 0.41652, + "grad_norm": 1.414009846905698, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 41652 + }, + { + "epoch": 0.41653, + "grad_norm": 1.7012552708185926, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 41653 + }, + { + "epoch": 0.41654, + "grad_norm": 1.8178178465409207, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 41654 + }, + { + "epoch": 0.41655, + "grad_norm": 1.8099354151410778, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 41655 + }, + { + "epoch": 0.41656, + "grad_norm": 1.4916120823244043, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 41656 + }, + { + "epoch": 0.41657, + "grad_norm": 1.8014375985916768, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 41657 + }, + { + "epoch": 0.41658, + "grad_norm": 1.3315599065852706, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 41658 + }, + { + "epoch": 0.41659, + "grad_norm": 1.894120648678036, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 41659 + }, + { + "epoch": 0.4166, + "grad_norm": 1.2036997028905276, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 41660 + }, + { + "epoch": 0.41661, + "grad_norm": 1.6145831864118867, + "learning_rate": 0.003, + "loss": 3.984, + "step": 41661 + }, + { + "epoch": 0.41662, + "grad_norm": 1.2692131762940286, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 41662 + }, + { + "epoch": 0.41663, + "grad_norm": 1.671660448718446, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 41663 + }, + { + "epoch": 0.41664, + "grad_norm": 1.494219755717791, + "learning_rate": 0.003, + "loss": 4.034, + "step": 41664 + }, + { + "epoch": 0.41665, + "grad_norm": 1.605484174227734, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 41665 + }, + { + "epoch": 0.41666, + "grad_norm": 1.5274458515658325, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41666 + }, + { + "epoch": 0.41667, + "grad_norm": 1.6420519136744027, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 41667 + }, + { + "epoch": 0.41668, + "grad_norm": 1.2920067400890047, + "learning_rate": 0.003, + "loss": 4.051, + "step": 41668 + }, + { + "epoch": 0.41669, + "grad_norm": 1.5539805140317104, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 41669 + }, + { + "epoch": 0.4167, + "grad_norm": 1.4388654612038096, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41670 + }, + { + "epoch": 0.41671, + "grad_norm": 1.2704368393934713, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 41671 + }, + { + "epoch": 0.41672, + "grad_norm": 1.732531602252248, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 41672 + }, + { + "epoch": 0.41673, + "grad_norm": 1.5701531122011307, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 41673 + }, + { + "epoch": 0.41674, + "grad_norm": 1.5121874645903572, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 41674 + }, + { + "epoch": 0.41675, + "grad_norm": 1.4235336500530904, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41675 + }, + { + "epoch": 0.41676, + "grad_norm": 1.4686441446102723, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 41676 + }, + { + "epoch": 0.41677, + "grad_norm": 1.4574253179384635, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 41677 + }, + { + "epoch": 0.41678, + "grad_norm": 1.648480726331975, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 41678 + }, + { + "epoch": 0.41679, + "grad_norm": 1.7209970194350759, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 41679 + }, + { + "epoch": 0.4168, + "grad_norm": 1.4813198505644085, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 41680 + }, + { + "epoch": 0.41681, + "grad_norm": 1.7239440326293336, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 41681 + }, + { + "epoch": 0.41682, + "grad_norm": 1.1796331199246943, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 41682 + }, + { + "epoch": 0.41683, + "grad_norm": 1.7868285718903176, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 41683 + }, + { + "epoch": 0.41684, + "grad_norm": 1.356354485225881, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 41684 + }, + { + "epoch": 0.41685, + "grad_norm": 1.6096390946032038, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41685 + }, + { + "epoch": 0.41686, + "grad_norm": 1.3827461737570335, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 41686 + }, + { + "epoch": 0.41687, + "grad_norm": 1.8145119128814209, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 41687 + }, + { + "epoch": 0.41688, + "grad_norm": 1.589568628077744, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 41688 + }, + { + "epoch": 0.41689, + "grad_norm": 1.3464531867328744, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41689 + }, + { + "epoch": 0.4169, + "grad_norm": 1.5836611176174418, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 41690 + }, + { + "epoch": 0.41691, + "grad_norm": 1.3681977960703815, + "learning_rate": 0.003, + "loss": 4.068, + "step": 41691 + }, + { + "epoch": 0.41692, + "grad_norm": 1.4486996969080495, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 41692 + }, + { + "epoch": 0.41693, + "grad_norm": 1.7249236706775564, + "learning_rate": 0.003, + "loss": 4.025, + "step": 41693 + }, + { + "epoch": 0.41694, + "grad_norm": 1.3639026605919522, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 41694 + }, + { + "epoch": 0.41695, + "grad_norm": 1.6422154495022774, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 41695 + }, + { + "epoch": 0.41696, + "grad_norm": 1.4433065639565745, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 41696 + }, + { + "epoch": 0.41697, + "grad_norm": 1.7108118042975773, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 41697 + }, + { + "epoch": 0.41698, + "grad_norm": 1.3220021792840908, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 41698 + }, + { + "epoch": 0.41699, + "grad_norm": 1.5454758628686076, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 41699 + }, + { + "epoch": 0.417, + "grad_norm": 1.5290289633910978, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 41700 + }, + { + "epoch": 0.41701, + "grad_norm": 1.3866590845953086, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 41701 + }, + { + "epoch": 0.41702, + "grad_norm": 1.7742513053307842, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 41702 + }, + { + "epoch": 0.41703, + "grad_norm": 1.2852163756206143, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 41703 + }, + { + "epoch": 0.41704, + "grad_norm": 1.7027092371326593, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41704 + }, + { + "epoch": 0.41705, + "grad_norm": 1.1564776605675835, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 41705 + }, + { + "epoch": 0.41706, + "grad_norm": 1.9975323657668014, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41706 + }, + { + "epoch": 0.41707, + "grad_norm": 1.4897921397473677, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 41707 + }, + { + "epoch": 0.41708, + "grad_norm": 1.6750455473218173, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 41708 + }, + { + "epoch": 0.41709, + "grad_norm": 1.6243245313482733, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 41709 + }, + { + "epoch": 0.4171, + "grad_norm": 1.571918901260577, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 41710 + }, + { + "epoch": 0.41711, + "grad_norm": 1.4321260240615599, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 41711 + }, + { + "epoch": 0.41712, + "grad_norm": 1.8860401586265059, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 41712 + }, + { + "epoch": 0.41713, + "grad_norm": 1.1303530559035295, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41713 + }, + { + "epoch": 0.41714, + "grad_norm": 1.9051978513537733, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 41714 + }, + { + "epoch": 0.41715, + "grad_norm": 1.4280063718978377, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 41715 + }, + { + "epoch": 0.41716, + "grad_norm": 1.5422857853112542, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 41716 + }, + { + "epoch": 0.41717, + "grad_norm": 1.2936077568982542, + "learning_rate": 0.003, + "loss": 4.041, + "step": 41717 + }, + { + "epoch": 0.41718, + "grad_norm": 1.6429718250878482, + "learning_rate": 0.003, + "loss": 4.06, + "step": 41718 + }, + { + "epoch": 0.41719, + "grad_norm": 1.5045572689363058, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 41719 + }, + { + "epoch": 0.4172, + "grad_norm": 1.4175320353325453, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41720 + }, + { + "epoch": 0.41721, + "grad_norm": 1.70073471549553, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 41721 + }, + { + "epoch": 0.41722, + "grad_norm": 1.7090780935926269, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 41722 + }, + { + "epoch": 0.41723, + "grad_norm": 1.66694215665512, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 41723 + }, + { + "epoch": 0.41724, + "grad_norm": 1.5924699392406396, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41724 + }, + { + "epoch": 0.41725, + "grad_norm": 1.8901341090213881, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 41725 + }, + { + "epoch": 0.41726, + "grad_norm": 1.4454599634425462, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 41726 + }, + { + "epoch": 0.41727, + "grad_norm": 1.400888120667451, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 41727 + }, + { + "epoch": 0.41728, + "grad_norm": 1.5524525438726917, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 41728 + }, + { + "epoch": 0.41729, + "grad_norm": 1.3435591638241307, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 41729 + }, + { + "epoch": 0.4173, + "grad_norm": 1.3771875071672355, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 41730 + }, + { + "epoch": 0.41731, + "grad_norm": 1.7282586239038011, + "learning_rate": 0.003, + "loss": 4.039, + "step": 41731 + }, + { + "epoch": 0.41732, + "grad_norm": 1.2810449650186118, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 41732 + }, + { + "epoch": 0.41733, + "grad_norm": 1.6946263318691306, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 41733 + }, + { + "epoch": 0.41734, + "grad_norm": 1.7601829180801318, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 41734 + }, + { + "epoch": 0.41735, + "grad_norm": 1.4272174895180902, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 41735 + }, + { + "epoch": 0.41736, + "grad_norm": 1.5265754690527191, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 41736 + }, + { + "epoch": 0.41737, + "grad_norm": 1.459275262461954, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 41737 + }, + { + "epoch": 0.41738, + "grad_norm": 1.417365122452981, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 41738 + }, + { + "epoch": 0.41739, + "grad_norm": 1.5474189468790085, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41739 + }, + { + "epoch": 0.4174, + "grad_norm": 1.7059616977281933, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 41740 + }, + { + "epoch": 0.41741, + "grad_norm": 1.3160587480264472, + "learning_rate": 0.003, + "loss": 4.058, + "step": 41741 + }, + { + "epoch": 0.41742, + "grad_norm": 1.5517757027475798, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 41742 + }, + { + "epoch": 0.41743, + "grad_norm": 1.630347730168504, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 41743 + }, + { + "epoch": 0.41744, + "grad_norm": 1.6850844526106286, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 41744 + }, + { + "epoch": 0.41745, + "grad_norm": 1.3828909904505209, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 41745 + }, + { + "epoch": 0.41746, + "grad_norm": 1.368786769264596, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 41746 + }, + { + "epoch": 0.41747, + "grad_norm": 1.5530591908126388, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41747 + }, + { + "epoch": 0.41748, + "grad_norm": 1.509689842605487, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 41748 + }, + { + "epoch": 0.41749, + "grad_norm": 1.702996784319653, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 41749 + }, + { + "epoch": 0.4175, + "grad_norm": 1.7265367300292824, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 41750 + }, + { + "epoch": 0.41751, + "grad_norm": 1.69050248503725, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41751 + }, + { + "epoch": 0.41752, + "grad_norm": 1.6020308091773299, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 41752 + }, + { + "epoch": 0.41753, + "grad_norm": 1.4804267660358592, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41753 + }, + { + "epoch": 0.41754, + "grad_norm": 1.4288574353672647, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 41754 + }, + { + "epoch": 0.41755, + "grad_norm": 1.4623289459949824, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 41755 + }, + { + "epoch": 0.41756, + "grad_norm": 1.575505256745096, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 41756 + }, + { + "epoch": 0.41757, + "grad_norm": 1.627056548650218, + "learning_rate": 0.003, + "loss": 4.053, + "step": 41757 + }, + { + "epoch": 0.41758, + "grad_norm": 1.4533130629043476, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41758 + }, + { + "epoch": 0.41759, + "grad_norm": 1.4615620087784864, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 41759 + }, + { + "epoch": 0.4176, + "grad_norm": 1.3980421770884321, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 41760 + }, + { + "epoch": 0.41761, + "grad_norm": 1.8345009077430812, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 41761 + }, + { + "epoch": 0.41762, + "grad_norm": 1.5611844323906563, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 41762 + }, + { + "epoch": 0.41763, + "grad_norm": 1.844073216806608, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 41763 + }, + { + "epoch": 0.41764, + "grad_norm": 1.8383003642363045, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41764 + }, + { + "epoch": 0.41765, + "grad_norm": 1.9097579703143113, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 41765 + }, + { + "epoch": 0.41766, + "grad_norm": 1.595067199217669, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 41766 + }, + { + "epoch": 0.41767, + "grad_norm": 1.9124940413313007, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 41767 + }, + { + "epoch": 0.41768, + "grad_norm": 1.7958703110419667, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 41768 + }, + { + "epoch": 0.41769, + "grad_norm": 1.552103124930985, + "learning_rate": 0.003, + "loss": 4.073, + "step": 41769 + }, + { + "epoch": 0.4177, + "grad_norm": 1.5560890062583612, + "learning_rate": 0.003, + "loss": 4.011, + "step": 41770 + }, + { + "epoch": 0.41771, + "grad_norm": 1.5489373949974685, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 41771 + }, + { + "epoch": 0.41772, + "grad_norm": 1.2287079625260127, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41772 + }, + { + "epoch": 0.41773, + "grad_norm": 1.6967110898785176, + "learning_rate": 0.003, + "loss": 4.052, + "step": 41773 + }, + { + "epoch": 0.41774, + "grad_norm": 1.3414119497109385, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 41774 + }, + { + "epoch": 0.41775, + "grad_norm": 1.551202716792147, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41775 + }, + { + "epoch": 0.41776, + "grad_norm": 1.602056376456247, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 41776 + }, + { + "epoch": 0.41777, + "grad_norm": 1.6199369855012875, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 41777 + }, + { + "epoch": 0.41778, + "grad_norm": 1.2587902540277536, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 41778 + }, + { + "epoch": 0.41779, + "grad_norm": 1.8308594280096429, + "learning_rate": 0.003, + "loss": 4.051, + "step": 41779 + }, + { + "epoch": 0.4178, + "grad_norm": 1.0955285572980458, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 41780 + }, + { + "epoch": 0.41781, + "grad_norm": 1.8862540451967196, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 41781 + }, + { + "epoch": 0.41782, + "grad_norm": 1.3757364729478183, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 41782 + }, + { + "epoch": 0.41783, + "grad_norm": 1.761571265903107, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 41783 + }, + { + "epoch": 0.41784, + "grad_norm": 1.5652855638345542, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 41784 + }, + { + "epoch": 0.41785, + "grad_norm": 1.6041763094294956, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 41785 + }, + { + "epoch": 0.41786, + "grad_norm": 1.3943443503824204, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 41786 + }, + { + "epoch": 0.41787, + "grad_norm": 1.86007285579664, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 41787 + }, + { + "epoch": 0.41788, + "grad_norm": 1.5987269552115781, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 41788 + }, + { + "epoch": 0.41789, + "grad_norm": 1.3554337155723257, + "learning_rate": 0.003, + "loss": 4.035, + "step": 41789 + }, + { + "epoch": 0.4179, + "grad_norm": 1.4984561740651137, + "learning_rate": 0.003, + "loss": 4.064, + "step": 41790 + }, + { + "epoch": 0.41791, + "grad_norm": 1.4394720610402068, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 41791 + }, + { + "epoch": 0.41792, + "grad_norm": 1.583672789685416, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 41792 + }, + { + "epoch": 0.41793, + "grad_norm": 1.2349698743755264, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 41793 + }, + { + "epoch": 0.41794, + "grad_norm": 1.872617834473143, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 41794 + }, + { + "epoch": 0.41795, + "grad_norm": 1.4044047974138738, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 41795 + }, + { + "epoch": 0.41796, + "grad_norm": 1.969561137177534, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 41796 + }, + { + "epoch": 0.41797, + "grad_norm": 1.7954039103794126, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 41797 + }, + { + "epoch": 0.41798, + "grad_norm": 1.465363725702129, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41798 + }, + { + "epoch": 0.41799, + "grad_norm": 1.5461138071828975, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 41799 + }, + { + "epoch": 0.418, + "grad_norm": 1.384531932557745, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41800 + }, + { + "epoch": 0.41801, + "grad_norm": 1.7502258231894419, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 41801 + }, + { + "epoch": 0.41802, + "grad_norm": 1.2903182929897223, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 41802 + }, + { + "epoch": 0.41803, + "grad_norm": 1.6461049816887907, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 41803 + }, + { + "epoch": 0.41804, + "grad_norm": 1.239687576823504, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 41804 + }, + { + "epoch": 0.41805, + "grad_norm": 1.6516751498596556, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 41805 + }, + { + "epoch": 0.41806, + "grad_norm": 1.6129257173661598, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 41806 + }, + { + "epoch": 0.41807, + "grad_norm": 1.5331723877249033, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41807 + }, + { + "epoch": 0.41808, + "grad_norm": 1.4006826042821634, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41808 + }, + { + "epoch": 0.41809, + "grad_norm": 1.390461121353893, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 41809 + }, + { + "epoch": 0.4181, + "grad_norm": 1.5019196718394456, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 41810 + }, + { + "epoch": 0.41811, + "grad_norm": 1.118420008855586, + "learning_rate": 0.003, + "loss": 4.026, + "step": 41811 + }, + { + "epoch": 0.41812, + "grad_norm": 1.6670467682288643, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 41812 + }, + { + "epoch": 0.41813, + "grad_norm": 1.466355200727519, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 41813 + }, + { + "epoch": 0.41814, + "grad_norm": 1.5817462761835965, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 41814 + }, + { + "epoch": 0.41815, + "grad_norm": 1.4816030337896482, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 41815 + }, + { + "epoch": 0.41816, + "grad_norm": 1.8670580532416303, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 41816 + }, + { + "epoch": 0.41817, + "grad_norm": 1.3646939814819155, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 41817 + }, + { + "epoch": 0.41818, + "grad_norm": 1.4480134840238952, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 41818 + }, + { + "epoch": 0.41819, + "grad_norm": 1.539454529220043, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 41819 + }, + { + "epoch": 0.4182, + "grad_norm": 1.586144335819331, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 41820 + }, + { + "epoch": 0.41821, + "grad_norm": 1.3382660072240036, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41821 + }, + { + "epoch": 0.41822, + "grad_norm": 1.4451652410056974, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 41822 + }, + { + "epoch": 0.41823, + "grad_norm": 1.6241418149845332, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 41823 + }, + { + "epoch": 0.41824, + "grad_norm": 1.7705975999956074, + "learning_rate": 0.003, + "loss": 4.036, + "step": 41824 + }, + { + "epoch": 0.41825, + "grad_norm": 1.5340052774807367, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 41825 + }, + { + "epoch": 0.41826, + "grad_norm": 1.437518201005437, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 41826 + }, + { + "epoch": 0.41827, + "grad_norm": 1.4741104730667518, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 41827 + }, + { + "epoch": 0.41828, + "grad_norm": 1.5009701358098138, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 41828 + }, + { + "epoch": 0.41829, + "grad_norm": 1.2030784566361097, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41829 + }, + { + "epoch": 0.4183, + "grad_norm": 1.597995169919141, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 41830 + }, + { + "epoch": 0.41831, + "grad_norm": 1.4484454933670445, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 41831 + }, + { + "epoch": 0.41832, + "grad_norm": 1.9207963439696683, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 41832 + }, + { + "epoch": 0.41833, + "grad_norm": 1.1532469941016747, + "learning_rate": 0.003, + "loss": 4.026, + "step": 41833 + }, + { + "epoch": 0.41834, + "grad_norm": 2.1753097994357686, + "learning_rate": 0.003, + "loss": 4.057, + "step": 41834 + }, + { + "epoch": 0.41835, + "grad_norm": 1.1664353079998966, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 41835 + }, + { + "epoch": 0.41836, + "grad_norm": 1.6208730139037437, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 41836 + }, + { + "epoch": 0.41837, + "grad_norm": 1.5453174327651615, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 41837 + }, + { + "epoch": 0.41838, + "grad_norm": 1.4974742073939469, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 41838 + }, + { + "epoch": 0.41839, + "grad_norm": 1.6795238881663772, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 41839 + }, + { + "epoch": 0.4184, + "grad_norm": 1.5904199489021977, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 41840 + }, + { + "epoch": 0.41841, + "grad_norm": 1.3549822604711625, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41841 + }, + { + "epoch": 0.41842, + "grad_norm": 1.7551833555728034, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 41842 + }, + { + "epoch": 0.41843, + "grad_norm": 1.3686546801137696, + "learning_rate": 0.003, + "loss": 4.059, + "step": 41843 + }, + { + "epoch": 0.41844, + "grad_norm": 1.7918435443054117, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 41844 + }, + { + "epoch": 0.41845, + "grad_norm": 1.637332863884353, + "learning_rate": 0.003, + "loss": 4.068, + "step": 41845 + }, + { + "epoch": 0.41846, + "grad_norm": 1.3992363485708437, + "learning_rate": 0.003, + "loss": 4.06, + "step": 41846 + }, + { + "epoch": 0.41847, + "grad_norm": 1.4151933905187521, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 41847 + }, + { + "epoch": 0.41848, + "grad_norm": 1.665488474919565, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 41848 + }, + { + "epoch": 0.41849, + "grad_norm": 1.5357704529882716, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 41849 + }, + { + "epoch": 0.4185, + "grad_norm": 1.2139991603004725, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 41850 + }, + { + "epoch": 0.41851, + "grad_norm": 1.605458712616258, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 41851 + }, + { + "epoch": 0.41852, + "grad_norm": 1.3400628311538096, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41852 + }, + { + "epoch": 0.41853, + "grad_norm": 1.8242345862459697, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 41853 + }, + { + "epoch": 0.41854, + "grad_norm": 1.6706129368512146, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 41854 + }, + { + "epoch": 0.41855, + "grad_norm": 1.4179636569939522, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 41855 + }, + { + "epoch": 0.41856, + "grad_norm": 1.7658410347700686, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 41856 + }, + { + "epoch": 0.41857, + "grad_norm": 1.6250729145920872, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 41857 + }, + { + "epoch": 0.41858, + "grad_norm": 1.2682041459828564, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 41858 + }, + { + "epoch": 0.41859, + "grad_norm": 1.72398285106969, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 41859 + }, + { + "epoch": 0.4186, + "grad_norm": 1.217777033483654, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 41860 + }, + { + "epoch": 0.41861, + "grad_norm": 1.7853468514751742, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 41861 + }, + { + "epoch": 0.41862, + "grad_norm": 1.4201450264712938, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 41862 + }, + { + "epoch": 0.41863, + "grad_norm": 1.541193656198914, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 41863 + }, + { + "epoch": 0.41864, + "grad_norm": 1.4920129872099528, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 41864 + }, + { + "epoch": 0.41865, + "grad_norm": 1.4553341912249265, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 41865 + }, + { + "epoch": 0.41866, + "grad_norm": 1.6776220132300461, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 41866 + }, + { + "epoch": 0.41867, + "grad_norm": 1.1284393310493501, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 41867 + }, + { + "epoch": 0.41868, + "grad_norm": 1.6383988511101848, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 41868 + }, + { + "epoch": 0.41869, + "grad_norm": 1.388570795600801, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 41869 + }, + { + "epoch": 0.4187, + "grad_norm": 1.7607888349099627, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 41870 + }, + { + "epoch": 0.41871, + "grad_norm": 1.411951424395042, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41871 + }, + { + "epoch": 0.41872, + "grad_norm": 1.7617636074256942, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 41872 + }, + { + "epoch": 0.41873, + "grad_norm": 1.475953480193187, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 41873 + }, + { + "epoch": 0.41874, + "grad_norm": 1.6210031944384262, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 41874 + }, + { + "epoch": 0.41875, + "grad_norm": 1.6212556101088134, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 41875 + }, + { + "epoch": 0.41876, + "grad_norm": 1.634253519876384, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 41876 + }, + { + "epoch": 0.41877, + "grad_norm": 1.334068246651076, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 41877 + }, + { + "epoch": 0.41878, + "grad_norm": 1.5990011003337092, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 41878 + }, + { + "epoch": 0.41879, + "grad_norm": 1.3357858992346667, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 41879 + }, + { + "epoch": 0.4188, + "grad_norm": 1.6506141685378153, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 41880 + }, + { + "epoch": 0.41881, + "grad_norm": 1.3981023815021905, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 41881 + }, + { + "epoch": 0.41882, + "grad_norm": 1.870851750847253, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 41882 + }, + { + "epoch": 0.41883, + "grad_norm": 1.391224368687125, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 41883 + }, + { + "epoch": 0.41884, + "grad_norm": 1.4754288720547573, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 41884 + }, + { + "epoch": 0.41885, + "grad_norm": 1.5650729546958129, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41885 + }, + { + "epoch": 0.41886, + "grad_norm": 1.5593307729113735, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 41886 + }, + { + "epoch": 0.41887, + "grad_norm": 1.4717578345584341, + "learning_rate": 0.003, + "loss": 4.025, + "step": 41887 + }, + { + "epoch": 0.41888, + "grad_norm": 1.6601248941062365, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 41888 + }, + { + "epoch": 0.41889, + "grad_norm": 1.3831557633747853, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 41889 + }, + { + "epoch": 0.4189, + "grad_norm": 1.6060032737869423, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 41890 + }, + { + "epoch": 0.41891, + "grad_norm": 1.3057955223609552, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 41891 + }, + { + "epoch": 0.41892, + "grad_norm": 1.7339097061845015, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 41892 + }, + { + "epoch": 0.41893, + "grad_norm": 1.4380150436841423, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 41893 + }, + { + "epoch": 0.41894, + "grad_norm": 1.4039146760515215, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 41894 + }, + { + "epoch": 0.41895, + "grad_norm": 1.5976330641267402, + "learning_rate": 0.003, + "loss": 4.008, + "step": 41895 + }, + { + "epoch": 0.41896, + "grad_norm": 1.6792013882683299, + "learning_rate": 0.003, + "loss": 4.007, + "step": 41896 + }, + { + "epoch": 0.41897, + "grad_norm": 1.567986331756742, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 41897 + }, + { + "epoch": 0.41898, + "grad_norm": 1.5567574364636503, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 41898 + }, + { + "epoch": 0.41899, + "grad_norm": 1.4363088211240556, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 41899 + }, + { + "epoch": 0.419, + "grad_norm": 1.6746047827895783, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 41900 + }, + { + "epoch": 0.41901, + "grad_norm": 1.1621844476385597, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 41901 + }, + { + "epoch": 0.41902, + "grad_norm": 1.7564235305222862, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 41902 + }, + { + "epoch": 0.41903, + "grad_norm": 1.375415264822158, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41903 + }, + { + "epoch": 0.41904, + "grad_norm": 1.5638833616498555, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 41904 + }, + { + "epoch": 0.41905, + "grad_norm": 1.7272370419557843, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 41905 + }, + { + "epoch": 0.41906, + "grad_norm": 1.592458217580157, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 41906 + }, + { + "epoch": 0.41907, + "grad_norm": 1.6620289194189817, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 41907 + }, + { + "epoch": 0.41908, + "grad_norm": 1.3057711732196877, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41908 + }, + { + "epoch": 0.41909, + "grad_norm": 1.6493896577365057, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41909 + }, + { + "epoch": 0.4191, + "grad_norm": 1.2179123904526135, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 41910 + }, + { + "epoch": 0.41911, + "grad_norm": 1.6455298113778123, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 41911 + }, + { + "epoch": 0.41912, + "grad_norm": 1.450186376358922, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 41912 + }, + { + "epoch": 0.41913, + "grad_norm": 1.5647289851456658, + "learning_rate": 0.003, + "loss": 4.024, + "step": 41913 + }, + { + "epoch": 0.41914, + "grad_norm": 1.4492087605884387, + "learning_rate": 0.003, + "loss": 4.045, + "step": 41914 + }, + { + "epoch": 0.41915, + "grad_norm": 1.591311074050783, + "learning_rate": 0.003, + "loss": 4.004, + "step": 41915 + }, + { + "epoch": 0.41916, + "grad_norm": 1.417697159113556, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 41916 + }, + { + "epoch": 0.41917, + "grad_norm": 1.495958622235838, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 41917 + }, + { + "epoch": 0.41918, + "grad_norm": 1.4825173512339755, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 41918 + }, + { + "epoch": 0.41919, + "grad_norm": 1.4357491462590957, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 41919 + }, + { + "epoch": 0.4192, + "grad_norm": 1.4602801389642788, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 41920 + }, + { + "epoch": 0.41921, + "grad_norm": 1.3328310563926462, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 41921 + }, + { + "epoch": 0.41922, + "grad_norm": 1.5300692156910982, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 41922 + }, + { + "epoch": 0.41923, + "grad_norm": 1.4466024276015381, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 41923 + }, + { + "epoch": 0.41924, + "grad_norm": 1.8073140500463096, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 41924 + }, + { + "epoch": 0.41925, + "grad_norm": 1.3922506795957652, + "learning_rate": 0.003, + "loss": 4.05, + "step": 41925 + }, + { + "epoch": 0.41926, + "grad_norm": 1.4323467319020209, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 41926 + }, + { + "epoch": 0.41927, + "grad_norm": 1.6523412590766604, + "learning_rate": 0.003, + "loss": 4.027, + "step": 41927 + }, + { + "epoch": 0.41928, + "grad_norm": 2.0704061764911037, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 41928 + }, + { + "epoch": 0.41929, + "grad_norm": 1.776190153683518, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41929 + }, + { + "epoch": 0.4193, + "grad_norm": 1.4098605453567599, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 41930 + }, + { + "epoch": 0.41931, + "grad_norm": 1.4774424831003277, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 41931 + }, + { + "epoch": 0.41932, + "grad_norm": 1.63178605267538, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 41932 + }, + { + "epoch": 0.41933, + "grad_norm": 1.6880745575202962, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 41933 + }, + { + "epoch": 0.41934, + "grad_norm": 1.4884134768092172, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 41934 + }, + { + "epoch": 0.41935, + "grad_norm": 1.3782214306852225, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 41935 + }, + { + "epoch": 0.41936, + "grad_norm": 1.5859143056093723, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 41936 + }, + { + "epoch": 0.41937, + "grad_norm": 1.1551556738379294, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 41937 + }, + { + "epoch": 0.41938, + "grad_norm": 1.8698215114798984, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 41938 + }, + { + "epoch": 0.41939, + "grad_norm": 1.3262001959435918, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 41939 + }, + { + "epoch": 0.4194, + "grad_norm": 1.7248616126409209, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 41940 + }, + { + "epoch": 0.41941, + "grad_norm": 1.3408569502465417, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 41941 + }, + { + "epoch": 0.41942, + "grad_norm": 1.8322417856979714, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 41942 + }, + { + "epoch": 0.41943, + "grad_norm": 1.4343751425663374, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41943 + }, + { + "epoch": 0.41944, + "grad_norm": 1.5256269236099507, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 41944 + }, + { + "epoch": 0.41945, + "grad_norm": 1.5915346115925915, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 41945 + }, + { + "epoch": 0.41946, + "grad_norm": 1.6159677798560235, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 41946 + }, + { + "epoch": 0.41947, + "grad_norm": 1.5715998974534646, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 41947 + }, + { + "epoch": 0.41948, + "grad_norm": 1.3585943304263715, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41948 + }, + { + "epoch": 0.41949, + "grad_norm": 1.349481193206886, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 41949 + }, + { + "epoch": 0.4195, + "grad_norm": 1.734662819960851, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 41950 + }, + { + "epoch": 0.41951, + "grad_norm": 1.6638970121534284, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 41951 + }, + { + "epoch": 0.41952, + "grad_norm": 1.8467388517446963, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 41952 + }, + { + "epoch": 0.41953, + "grad_norm": 1.3111220191188804, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 41953 + }, + { + "epoch": 0.41954, + "grad_norm": 1.527147501986217, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 41954 + }, + { + "epoch": 0.41955, + "grad_norm": 1.2037977527635084, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 41955 + }, + { + "epoch": 0.41956, + "grad_norm": 1.8289398107889334, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 41956 + }, + { + "epoch": 0.41957, + "grad_norm": 1.2863002331240063, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 41957 + }, + { + "epoch": 0.41958, + "grad_norm": 1.5674094073191975, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 41958 + }, + { + "epoch": 0.41959, + "grad_norm": 1.4693422659481712, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 41959 + }, + { + "epoch": 0.4196, + "grad_norm": 1.4225215345528672, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 41960 + }, + { + "epoch": 0.41961, + "grad_norm": 1.6961870902817588, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 41961 + }, + { + "epoch": 0.41962, + "grad_norm": 1.4242574172272022, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 41962 + }, + { + "epoch": 0.41963, + "grad_norm": 1.3550038429309939, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 41963 + }, + { + "epoch": 0.41964, + "grad_norm": 1.5852627421232652, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 41964 + }, + { + "epoch": 0.41965, + "grad_norm": 1.8153587121730872, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 41965 + }, + { + "epoch": 0.41966, + "grad_norm": 2.239219394138386, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 41966 + }, + { + "epoch": 0.41967, + "grad_norm": 1.071865291965343, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 41967 + }, + { + "epoch": 0.41968, + "grad_norm": 1.7749278218533657, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 41968 + }, + { + "epoch": 0.41969, + "grad_norm": 1.6742909172073277, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 41969 + }, + { + "epoch": 0.4197, + "grad_norm": 1.4952562397296727, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 41970 + }, + { + "epoch": 0.41971, + "grad_norm": 1.7008430058307724, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 41971 + }, + { + "epoch": 0.41972, + "grad_norm": 1.446317102127311, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 41972 + }, + { + "epoch": 0.41973, + "grad_norm": 1.6050102670703525, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 41973 + }, + { + "epoch": 0.41974, + "grad_norm": 1.558309876336897, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 41974 + }, + { + "epoch": 0.41975, + "grad_norm": 1.3383301457352645, + "learning_rate": 0.003, + "loss": 4.011, + "step": 41975 + }, + { + "epoch": 0.41976, + "grad_norm": 1.6633253536901123, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 41976 + }, + { + "epoch": 0.41977, + "grad_norm": 1.3827582857112628, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 41977 + }, + { + "epoch": 0.41978, + "grad_norm": 1.919056074190659, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 41978 + }, + { + "epoch": 0.41979, + "grad_norm": 1.2251296687444253, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 41979 + }, + { + "epoch": 0.4198, + "grad_norm": 1.662129119439661, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 41980 + }, + { + "epoch": 0.41981, + "grad_norm": 1.611151698313816, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 41981 + }, + { + "epoch": 0.41982, + "grad_norm": 1.397012034993571, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 41982 + }, + { + "epoch": 0.41983, + "grad_norm": 1.4799498538657643, + "learning_rate": 0.003, + "loss": 4.022, + "step": 41983 + }, + { + "epoch": 0.41984, + "grad_norm": 1.5603601279113042, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 41984 + }, + { + "epoch": 0.41985, + "grad_norm": 1.8711077212841927, + "learning_rate": 0.003, + "loss": 4.051, + "step": 41985 + }, + { + "epoch": 0.41986, + "grad_norm": 1.6455702132561156, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 41986 + }, + { + "epoch": 0.41987, + "grad_norm": 1.472514935683033, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 41987 + }, + { + "epoch": 0.41988, + "grad_norm": 1.6063548328368544, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 41988 + }, + { + "epoch": 0.41989, + "grad_norm": 1.6098779534786025, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 41989 + }, + { + "epoch": 0.4199, + "grad_norm": 1.4053958986662323, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 41990 + }, + { + "epoch": 0.41991, + "grad_norm": 1.3936230248494894, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 41991 + }, + { + "epoch": 0.41992, + "grad_norm": 1.6107909516474455, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 41992 + }, + { + "epoch": 0.41993, + "grad_norm": 1.4121048226196828, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 41993 + }, + { + "epoch": 0.41994, + "grad_norm": 1.55763596240821, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 41994 + }, + { + "epoch": 0.41995, + "grad_norm": 1.4588395556692684, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 41995 + }, + { + "epoch": 0.41996, + "grad_norm": 1.4598123457611374, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41996 + }, + { + "epoch": 0.41997, + "grad_norm": 1.5451271426664743, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 41997 + }, + { + "epoch": 0.41998, + "grad_norm": 1.5229809494344482, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 41998 + }, + { + "epoch": 0.41999, + "grad_norm": 1.5363576264016459, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 41999 + }, + { + "epoch": 0.42, + "grad_norm": 1.4071720511369687, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 42000 + }, + { + "epoch": 0.42001, + "grad_norm": 2.069137887247133, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 42001 + }, + { + "epoch": 0.42002, + "grad_norm": 1.5585425796694967, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 42002 + }, + { + "epoch": 0.42003, + "grad_norm": 1.5851805641888252, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 42003 + }, + { + "epoch": 0.42004, + "grad_norm": 1.461620447903966, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 42004 + }, + { + "epoch": 0.42005, + "grad_norm": 1.7565689819617254, + "learning_rate": 0.003, + "loss": 4.006, + "step": 42005 + }, + { + "epoch": 0.42006, + "grad_norm": 1.3358619147723774, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 42006 + }, + { + "epoch": 0.42007, + "grad_norm": 1.5806975546492907, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42007 + }, + { + "epoch": 0.42008, + "grad_norm": 1.5236493238690234, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42008 + }, + { + "epoch": 0.42009, + "grad_norm": 1.312809689685695, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42009 + }, + { + "epoch": 0.4201, + "grad_norm": 1.4996044128699464, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 42010 + }, + { + "epoch": 0.42011, + "grad_norm": 1.2832463489947268, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42011 + }, + { + "epoch": 0.42012, + "grad_norm": 1.710543885876342, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 42012 + }, + { + "epoch": 0.42013, + "grad_norm": 1.3245834369647245, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42013 + }, + { + "epoch": 0.42014, + "grad_norm": 1.651934705727464, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42014 + }, + { + "epoch": 0.42015, + "grad_norm": 1.6132744635323437, + "learning_rate": 0.003, + "loss": 4.058, + "step": 42015 + }, + { + "epoch": 0.42016, + "grad_norm": 1.6037002941842688, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42016 + }, + { + "epoch": 0.42017, + "grad_norm": 1.3726246174107004, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42017 + }, + { + "epoch": 0.42018, + "grad_norm": 2.1129572348575203, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 42018 + }, + { + "epoch": 0.42019, + "grad_norm": 1.2865443670765513, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 42019 + }, + { + "epoch": 0.4202, + "grad_norm": 1.8102534540172281, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 42020 + }, + { + "epoch": 0.42021, + "grad_norm": 1.7349475517518478, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42021 + }, + { + "epoch": 0.42022, + "grad_norm": 1.2424188431560885, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 42022 + }, + { + "epoch": 0.42023, + "grad_norm": 1.7210227662442925, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42023 + }, + { + "epoch": 0.42024, + "grad_norm": 1.2714197392089666, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 42024 + }, + { + "epoch": 0.42025, + "grad_norm": 1.536290487365259, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42025 + }, + { + "epoch": 0.42026, + "grad_norm": 1.5784705804669226, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 42026 + }, + { + "epoch": 0.42027, + "grad_norm": 1.903826931076763, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42027 + }, + { + "epoch": 0.42028, + "grad_norm": 1.3534052114515631, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 42028 + }, + { + "epoch": 0.42029, + "grad_norm": 1.2820543756662115, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42029 + }, + { + "epoch": 0.4203, + "grad_norm": 1.7157920659937744, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42030 + }, + { + "epoch": 0.42031, + "grad_norm": 1.4453219810307163, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 42031 + }, + { + "epoch": 0.42032, + "grad_norm": 1.5109947634957428, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 42032 + }, + { + "epoch": 0.42033, + "grad_norm": 1.6185545421601337, + "learning_rate": 0.003, + "loss": 4.066, + "step": 42033 + }, + { + "epoch": 0.42034, + "grad_norm": 1.5474984534350502, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 42034 + }, + { + "epoch": 0.42035, + "grad_norm": 1.5859072772455947, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 42035 + }, + { + "epoch": 0.42036, + "grad_norm": 1.5717054041200784, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 42036 + }, + { + "epoch": 0.42037, + "grad_norm": 1.7951578229924872, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42037 + }, + { + "epoch": 0.42038, + "grad_norm": 1.3176974533387418, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 42038 + }, + { + "epoch": 0.42039, + "grad_norm": 1.8825435595763036, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 42039 + }, + { + "epoch": 0.4204, + "grad_norm": 1.4200715812634135, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 42040 + }, + { + "epoch": 0.42041, + "grad_norm": 1.4471484618418544, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 42041 + }, + { + "epoch": 0.42042, + "grad_norm": 1.5900013757113203, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 42042 + }, + { + "epoch": 0.42043, + "grad_norm": 1.571062421601607, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 42043 + }, + { + "epoch": 0.42044, + "grad_norm": 1.7732475221939155, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42044 + }, + { + "epoch": 0.42045, + "grad_norm": 1.5460722468287023, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 42045 + }, + { + "epoch": 0.42046, + "grad_norm": 1.5792068863618627, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42046 + }, + { + "epoch": 0.42047, + "grad_norm": 1.4123046205462597, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42047 + }, + { + "epoch": 0.42048, + "grad_norm": 1.3949181964464192, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 42048 + }, + { + "epoch": 0.42049, + "grad_norm": 1.5294116437365584, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 42049 + }, + { + "epoch": 0.4205, + "grad_norm": 1.674792917035468, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42050 + }, + { + "epoch": 0.42051, + "grad_norm": 1.4812351118560614, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42051 + }, + { + "epoch": 0.42052, + "grad_norm": 1.3546796285061786, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 42052 + }, + { + "epoch": 0.42053, + "grad_norm": 1.605002529137344, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 42053 + }, + { + "epoch": 0.42054, + "grad_norm": 1.600721834105462, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 42054 + }, + { + "epoch": 0.42055, + "grad_norm": 1.6352922884864907, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 42055 + }, + { + "epoch": 0.42056, + "grad_norm": 1.3248716401242728, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 42056 + }, + { + "epoch": 0.42057, + "grad_norm": 1.5568974392272927, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 42057 + }, + { + "epoch": 0.42058, + "grad_norm": 1.6177791302009155, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 42058 + }, + { + "epoch": 0.42059, + "grad_norm": 1.600642986126611, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 42059 + }, + { + "epoch": 0.4206, + "grad_norm": 1.5112281651882256, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42060 + }, + { + "epoch": 0.42061, + "grad_norm": 1.992784157369177, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42061 + }, + { + "epoch": 0.42062, + "grad_norm": 1.0716764957803542, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 42062 + }, + { + "epoch": 0.42063, + "grad_norm": 1.6451637387591325, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 42063 + }, + { + "epoch": 0.42064, + "grad_norm": 1.4963644826423714, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 42064 + }, + { + "epoch": 0.42065, + "grad_norm": 1.9787809851649505, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 42065 + }, + { + "epoch": 0.42066, + "grad_norm": 1.398825370577943, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 42066 + }, + { + "epoch": 0.42067, + "grad_norm": 1.5623790954639634, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 42067 + }, + { + "epoch": 0.42068, + "grad_norm": 1.538375622369539, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42068 + }, + { + "epoch": 0.42069, + "grad_norm": 1.7425048512332528, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 42069 + }, + { + "epoch": 0.4207, + "grad_norm": 1.3404509546853949, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42070 + }, + { + "epoch": 0.42071, + "grad_norm": 1.647300573331925, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 42071 + }, + { + "epoch": 0.42072, + "grad_norm": 1.4045568564547646, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 42072 + }, + { + "epoch": 0.42073, + "grad_norm": 1.6470564543131085, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 42073 + }, + { + "epoch": 0.42074, + "grad_norm": 1.4474738281472461, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42074 + }, + { + "epoch": 0.42075, + "grad_norm": 1.767718814579465, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 42075 + }, + { + "epoch": 0.42076, + "grad_norm": 1.6051936264778326, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 42076 + }, + { + "epoch": 0.42077, + "grad_norm": 1.4858643206995186, + "learning_rate": 0.003, + "loss": 4.053, + "step": 42077 + }, + { + "epoch": 0.42078, + "grad_norm": 1.6277976127746228, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 42078 + }, + { + "epoch": 0.42079, + "grad_norm": 1.315779193370727, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 42079 + }, + { + "epoch": 0.4208, + "grad_norm": 1.8268433430609308, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42080 + }, + { + "epoch": 0.42081, + "grad_norm": 1.4033427875135964, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 42081 + }, + { + "epoch": 0.42082, + "grad_norm": 1.5387290895283092, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42082 + }, + { + "epoch": 0.42083, + "grad_norm": 1.549646032940175, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 42083 + }, + { + "epoch": 0.42084, + "grad_norm": 1.4442634281699263, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42084 + }, + { + "epoch": 0.42085, + "grad_norm": 1.2680539140829623, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42085 + }, + { + "epoch": 0.42086, + "grad_norm": 1.6648583110872748, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 42086 + }, + { + "epoch": 0.42087, + "grad_norm": 1.6538857867849923, + "learning_rate": 0.003, + "loss": 4.054, + "step": 42087 + }, + { + "epoch": 0.42088, + "grad_norm": 1.3159355511595143, + "learning_rate": 0.003, + "loss": 4.062, + "step": 42088 + }, + { + "epoch": 0.42089, + "grad_norm": 1.3927076870040929, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42089 + }, + { + "epoch": 0.4209, + "grad_norm": 1.584223506400029, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 42090 + }, + { + "epoch": 0.42091, + "grad_norm": 1.3913535299661743, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 42091 + }, + { + "epoch": 0.42092, + "grad_norm": 1.67208972745565, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 42092 + }, + { + "epoch": 0.42093, + "grad_norm": 1.3598003670740442, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 42093 + }, + { + "epoch": 0.42094, + "grad_norm": 1.6700930241228218, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42094 + }, + { + "epoch": 0.42095, + "grad_norm": 1.4568920640115341, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 42095 + }, + { + "epoch": 0.42096, + "grad_norm": 1.9797049303630347, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 42096 + }, + { + "epoch": 0.42097, + "grad_norm": 1.3997199759397019, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42097 + }, + { + "epoch": 0.42098, + "grad_norm": 1.5170056408734736, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 42098 + }, + { + "epoch": 0.42099, + "grad_norm": 1.5234329581625816, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42099 + }, + { + "epoch": 0.421, + "grad_norm": 1.8194872093172074, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 42100 + }, + { + "epoch": 0.42101, + "grad_norm": 1.4297193412647873, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42101 + }, + { + "epoch": 0.42102, + "grad_norm": 1.7738339963146221, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 42102 + }, + { + "epoch": 0.42103, + "grad_norm": 1.2198739165062764, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 42103 + }, + { + "epoch": 0.42104, + "grad_norm": 1.6927847574539991, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 42104 + }, + { + "epoch": 0.42105, + "grad_norm": 1.2727744237475203, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42105 + }, + { + "epoch": 0.42106, + "grad_norm": 1.8151459758659914, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42106 + }, + { + "epoch": 0.42107, + "grad_norm": 1.4509833216043924, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 42107 + }, + { + "epoch": 0.42108, + "grad_norm": 1.6164021962391957, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 42108 + }, + { + "epoch": 0.42109, + "grad_norm": 1.4899615880509554, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 42109 + }, + { + "epoch": 0.4211, + "grad_norm": 1.3395638831520817, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 42110 + }, + { + "epoch": 0.42111, + "grad_norm": 1.6487285547375623, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 42111 + }, + { + "epoch": 0.42112, + "grad_norm": 1.2705415065549244, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 42112 + }, + { + "epoch": 0.42113, + "grad_norm": 1.5277676490174872, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 42113 + }, + { + "epoch": 0.42114, + "grad_norm": 1.6411293490136887, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 42114 + }, + { + "epoch": 0.42115, + "grad_norm": 1.476354079126554, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 42115 + }, + { + "epoch": 0.42116, + "grad_norm": 1.580183483811, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 42116 + }, + { + "epoch": 0.42117, + "grad_norm": 1.7347705218034637, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42117 + }, + { + "epoch": 0.42118, + "grad_norm": 1.611324509711604, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 42118 + }, + { + "epoch": 0.42119, + "grad_norm": 1.5742772039052704, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42119 + }, + { + "epoch": 0.4212, + "grad_norm": 1.5279147297797595, + "learning_rate": 0.003, + "loss": 4.061, + "step": 42120 + }, + { + "epoch": 0.42121, + "grad_norm": 1.325027202227869, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 42121 + }, + { + "epoch": 0.42122, + "grad_norm": 1.5698641368103188, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 42122 + }, + { + "epoch": 0.42123, + "grad_norm": 1.323722764196603, + "learning_rate": 0.003, + "loss": 4.045, + "step": 42123 + }, + { + "epoch": 0.42124, + "grad_norm": 1.4735589535380773, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42124 + }, + { + "epoch": 0.42125, + "grad_norm": 1.3755159762955753, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 42125 + }, + { + "epoch": 0.42126, + "grad_norm": 2.237045162354008, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 42126 + }, + { + "epoch": 0.42127, + "grad_norm": 1.274907978298295, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 42127 + }, + { + "epoch": 0.42128, + "grad_norm": 1.7262263600338348, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 42128 + }, + { + "epoch": 0.42129, + "grad_norm": 1.5324044679549131, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 42129 + }, + { + "epoch": 0.4213, + "grad_norm": 1.4433615510672388, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42130 + }, + { + "epoch": 0.42131, + "grad_norm": 1.9360349381619069, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 42131 + }, + { + "epoch": 0.42132, + "grad_norm": 1.088558180926403, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 42132 + }, + { + "epoch": 0.42133, + "grad_norm": 1.5685294840374828, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 42133 + }, + { + "epoch": 0.42134, + "grad_norm": 1.7200903546223036, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 42134 + }, + { + "epoch": 0.42135, + "grad_norm": 1.6324429955251514, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 42135 + }, + { + "epoch": 0.42136, + "grad_norm": 1.41076929040418, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42136 + }, + { + "epoch": 0.42137, + "grad_norm": 1.8604446899825573, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 42137 + }, + { + "epoch": 0.42138, + "grad_norm": 1.1476982049403361, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42138 + }, + { + "epoch": 0.42139, + "grad_norm": 1.5387933098920685, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 42139 + }, + { + "epoch": 0.4214, + "grad_norm": 1.737026841585308, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 42140 + }, + { + "epoch": 0.42141, + "grad_norm": 1.5334079421656464, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 42141 + }, + { + "epoch": 0.42142, + "grad_norm": 1.4979741022959323, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 42142 + }, + { + "epoch": 0.42143, + "grad_norm": 1.4013830452663065, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42143 + }, + { + "epoch": 0.42144, + "grad_norm": 1.4618220016449335, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 42144 + }, + { + "epoch": 0.42145, + "grad_norm": 1.3560184157224793, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 42145 + }, + { + "epoch": 0.42146, + "grad_norm": 1.6612070525998757, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 42146 + }, + { + "epoch": 0.42147, + "grad_norm": 1.455469451994088, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 42147 + }, + { + "epoch": 0.42148, + "grad_norm": 1.5515040090792565, + "learning_rate": 0.003, + "loss": 4.022, + "step": 42148 + }, + { + "epoch": 0.42149, + "grad_norm": 1.2751087327143105, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 42149 + }, + { + "epoch": 0.4215, + "grad_norm": 1.579402203300351, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42150 + }, + { + "epoch": 0.42151, + "grad_norm": 1.161139266287962, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42151 + }, + { + "epoch": 0.42152, + "grad_norm": 1.6481540857924195, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 42152 + }, + { + "epoch": 0.42153, + "grad_norm": 1.3555642820269589, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 42153 + }, + { + "epoch": 0.42154, + "grad_norm": 1.6648434456613244, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 42154 + }, + { + "epoch": 0.42155, + "grad_norm": 1.6078582559005645, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 42155 + }, + { + "epoch": 0.42156, + "grad_norm": 1.8695492925868877, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 42156 + }, + { + "epoch": 0.42157, + "grad_norm": 1.3746017170284015, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 42157 + }, + { + "epoch": 0.42158, + "grad_norm": 1.6137746038415026, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42158 + }, + { + "epoch": 0.42159, + "grad_norm": 1.6230447608905663, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 42159 + }, + { + "epoch": 0.4216, + "grad_norm": 1.506817437786154, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 42160 + }, + { + "epoch": 0.42161, + "grad_norm": 1.271902813861193, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 42161 + }, + { + "epoch": 0.42162, + "grad_norm": 1.614868477561775, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 42162 + }, + { + "epoch": 0.42163, + "grad_norm": 1.6532454051469823, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42163 + }, + { + "epoch": 0.42164, + "grad_norm": 1.402060336052378, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 42164 + }, + { + "epoch": 0.42165, + "grad_norm": 1.5163009322360161, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42165 + }, + { + "epoch": 0.42166, + "grad_norm": 1.4020865123639508, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 42166 + }, + { + "epoch": 0.42167, + "grad_norm": 1.6512768062365313, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 42167 + }, + { + "epoch": 0.42168, + "grad_norm": 1.5671854619946806, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 42168 + }, + { + "epoch": 0.42169, + "grad_norm": 2.0176976123706214, + "learning_rate": 0.003, + "loss": 4.021, + "step": 42169 + }, + { + "epoch": 0.4217, + "grad_norm": 1.4443331799158403, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 42170 + }, + { + "epoch": 0.42171, + "grad_norm": 1.519377087878334, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 42171 + }, + { + "epoch": 0.42172, + "grad_norm": 1.5436634507429705, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 42172 + }, + { + "epoch": 0.42173, + "grad_norm": 1.3229437434818672, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 42173 + }, + { + "epoch": 0.42174, + "grad_norm": 1.6098169665274105, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42174 + }, + { + "epoch": 0.42175, + "grad_norm": 1.5476358367829055, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 42175 + }, + { + "epoch": 0.42176, + "grad_norm": 1.5759081691336185, + "learning_rate": 0.003, + "loss": 4.05, + "step": 42176 + }, + { + "epoch": 0.42177, + "grad_norm": 1.533100518875386, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 42177 + }, + { + "epoch": 0.42178, + "grad_norm": 1.3307600280991292, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 42178 + }, + { + "epoch": 0.42179, + "grad_norm": 1.7404631338120953, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42179 + }, + { + "epoch": 0.4218, + "grad_norm": 1.5261457556555584, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 42180 + }, + { + "epoch": 0.42181, + "grad_norm": 1.6928686089810518, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42181 + }, + { + "epoch": 0.42182, + "grad_norm": 1.539724461208031, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42182 + }, + { + "epoch": 0.42183, + "grad_norm": 1.3111924438185578, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42183 + }, + { + "epoch": 0.42184, + "grad_norm": 1.6769721191237448, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42184 + }, + { + "epoch": 0.42185, + "grad_norm": 1.561905057648116, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 42185 + }, + { + "epoch": 0.42186, + "grad_norm": 1.7327755206735675, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 42186 + }, + { + "epoch": 0.42187, + "grad_norm": 1.2799759896763467, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 42187 + }, + { + "epoch": 0.42188, + "grad_norm": 1.5427397770731028, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 42188 + }, + { + "epoch": 0.42189, + "grad_norm": 1.417303438862292, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 42189 + }, + { + "epoch": 0.4219, + "grad_norm": 1.5704983115453444, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 42190 + }, + { + "epoch": 0.42191, + "grad_norm": 1.7637398853720088, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42191 + }, + { + "epoch": 0.42192, + "grad_norm": 1.452196111528625, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 42192 + }, + { + "epoch": 0.42193, + "grad_norm": 1.6626843399548505, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42193 + }, + { + "epoch": 0.42194, + "grad_norm": 1.3997100156998488, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 42194 + }, + { + "epoch": 0.42195, + "grad_norm": 1.7196448441649401, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 42195 + }, + { + "epoch": 0.42196, + "grad_norm": 1.6118759578179789, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 42196 + }, + { + "epoch": 0.42197, + "grad_norm": 1.2818956944962823, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42197 + }, + { + "epoch": 0.42198, + "grad_norm": 1.598495932747698, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 42198 + }, + { + "epoch": 0.42199, + "grad_norm": 1.4421755077407146, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42199 + }, + { + "epoch": 0.422, + "grad_norm": 1.427479414267211, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 42200 + }, + { + "epoch": 0.42201, + "grad_norm": 1.437496136204997, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 42201 + }, + { + "epoch": 0.42202, + "grad_norm": 1.8197768950887638, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 42202 + }, + { + "epoch": 0.42203, + "grad_norm": 1.1747502792720612, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 42203 + }, + { + "epoch": 0.42204, + "grad_norm": 2.0492284227745667, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 42204 + }, + { + "epoch": 0.42205, + "grad_norm": 1.4130880411973303, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 42205 + }, + { + "epoch": 0.42206, + "grad_norm": 1.6018750514835634, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 42206 + }, + { + "epoch": 0.42207, + "grad_norm": 1.5483604614593762, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 42207 + }, + { + "epoch": 0.42208, + "grad_norm": 1.7003098263726277, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 42208 + }, + { + "epoch": 0.42209, + "grad_norm": 1.4441062550211816, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 42209 + }, + { + "epoch": 0.4221, + "grad_norm": 1.5684953616797812, + "learning_rate": 0.003, + "loss": 4.052, + "step": 42210 + }, + { + "epoch": 0.42211, + "grad_norm": 1.6270467235975303, + "learning_rate": 0.003, + "loss": 4.034, + "step": 42211 + }, + { + "epoch": 0.42212, + "grad_norm": 1.4761810982468526, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42212 + }, + { + "epoch": 0.42213, + "grad_norm": 1.5376127828237705, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 42213 + }, + { + "epoch": 0.42214, + "grad_norm": 1.4217247613814163, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 42214 + }, + { + "epoch": 0.42215, + "grad_norm": 1.3785993381180197, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 42215 + }, + { + "epoch": 0.42216, + "grad_norm": 1.3894133855264263, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42216 + }, + { + "epoch": 0.42217, + "grad_norm": 1.7329119980928862, + "learning_rate": 0.003, + "loss": 4.046, + "step": 42217 + }, + { + "epoch": 0.42218, + "grad_norm": 1.4009272204576193, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 42218 + }, + { + "epoch": 0.42219, + "grad_norm": 2.0679758887566484, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 42219 + }, + { + "epoch": 0.4222, + "grad_norm": 1.4370008559663883, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42220 + }, + { + "epoch": 0.42221, + "grad_norm": 1.6273043922453578, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 42221 + }, + { + "epoch": 0.42222, + "grad_norm": 1.4407304226082578, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 42222 + }, + { + "epoch": 0.42223, + "grad_norm": 1.554541797193351, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 42223 + }, + { + "epoch": 0.42224, + "grad_norm": 1.6048286368991218, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42224 + }, + { + "epoch": 0.42225, + "grad_norm": 1.2691642034397195, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42225 + }, + { + "epoch": 0.42226, + "grad_norm": 1.961726595275884, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 42226 + }, + { + "epoch": 0.42227, + "grad_norm": 1.3373214868351213, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 42227 + }, + { + "epoch": 0.42228, + "grad_norm": 1.4430406211286566, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 42228 + }, + { + "epoch": 0.42229, + "grad_norm": 1.7365831453933118, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 42229 + }, + { + "epoch": 0.4223, + "grad_norm": 1.15158632582919, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42230 + }, + { + "epoch": 0.42231, + "grad_norm": 1.8284169536894683, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 42231 + }, + { + "epoch": 0.42232, + "grad_norm": 1.4473029254903222, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 42232 + }, + { + "epoch": 0.42233, + "grad_norm": 1.4737183970284953, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42233 + }, + { + "epoch": 0.42234, + "grad_norm": 1.4936902180770788, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 42234 + }, + { + "epoch": 0.42235, + "grad_norm": 1.579856285435662, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 42235 + }, + { + "epoch": 0.42236, + "grad_norm": 1.6072187732799939, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 42236 + }, + { + "epoch": 0.42237, + "grad_norm": 1.6544767362075858, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 42237 + }, + { + "epoch": 0.42238, + "grad_norm": 1.2484632569768992, + "learning_rate": 0.003, + "loss": 4.025, + "step": 42238 + }, + { + "epoch": 0.42239, + "grad_norm": 1.654257280810677, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 42239 + }, + { + "epoch": 0.4224, + "grad_norm": 1.1363206002153836, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 42240 + }, + { + "epoch": 0.42241, + "grad_norm": 1.9080025014817543, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 42241 + }, + { + "epoch": 0.42242, + "grad_norm": 1.323924870855244, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42242 + }, + { + "epoch": 0.42243, + "grad_norm": 1.969311604540027, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 42243 + }, + { + "epoch": 0.42244, + "grad_norm": 1.5073155734319874, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 42244 + }, + { + "epoch": 0.42245, + "grad_norm": 1.6910735332472306, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 42245 + }, + { + "epoch": 0.42246, + "grad_norm": 1.574775012438874, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42246 + }, + { + "epoch": 0.42247, + "grad_norm": 1.7569260594449436, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 42247 + }, + { + "epoch": 0.42248, + "grad_norm": 1.3715225520244303, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 42248 + }, + { + "epoch": 0.42249, + "grad_norm": 1.6828402750171452, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 42249 + }, + { + "epoch": 0.4225, + "grad_norm": 1.4192753007426122, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 42250 + }, + { + "epoch": 0.42251, + "grad_norm": 1.5306080480062718, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 42251 + }, + { + "epoch": 0.42252, + "grad_norm": 1.328860962711536, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 42252 + }, + { + "epoch": 0.42253, + "grad_norm": 1.5693524755172021, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42253 + }, + { + "epoch": 0.42254, + "grad_norm": 1.7054055459722695, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42254 + }, + { + "epoch": 0.42255, + "grad_norm": 1.3675533240316726, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42255 + }, + { + "epoch": 0.42256, + "grad_norm": 1.7595715490544817, + "learning_rate": 0.003, + "loss": 4.016, + "step": 42256 + }, + { + "epoch": 0.42257, + "grad_norm": 1.3889558212613309, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 42257 + }, + { + "epoch": 0.42258, + "grad_norm": 1.4206260009045586, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 42258 + }, + { + "epoch": 0.42259, + "grad_norm": 1.6284586251675721, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 42259 + }, + { + "epoch": 0.4226, + "grad_norm": 1.2752356995935685, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 42260 + }, + { + "epoch": 0.42261, + "grad_norm": 1.6204825902449493, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 42261 + }, + { + "epoch": 0.42262, + "grad_norm": 1.2935261191608725, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 42262 + }, + { + "epoch": 0.42263, + "grad_norm": 1.5002296909990624, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 42263 + }, + { + "epoch": 0.42264, + "grad_norm": 1.5251936179583, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 42264 + }, + { + "epoch": 0.42265, + "grad_norm": 1.5164825329895615, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 42265 + }, + { + "epoch": 0.42266, + "grad_norm": 1.5504092251142296, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 42266 + }, + { + "epoch": 0.42267, + "grad_norm": 1.5235240752255392, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 42267 + }, + { + "epoch": 0.42268, + "grad_norm": 1.5154180650822713, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 42268 + }, + { + "epoch": 0.42269, + "grad_norm": 1.7758152572175148, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42269 + }, + { + "epoch": 0.4227, + "grad_norm": 1.6592217674198277, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 42270 + }, + { + "epoch": 0.42271, + "grad_norm": 1.4332301786279624, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 42271 + }, + { + "epoch": 0.42272, + "grad_norm": 1.4411359685048943, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 42272 + }, + { + "epoch": 0.42273, + "grad_norm": 1.8961221322760133, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42273 + }, + { + "epoch": 0.42274, + "grad_norm": 1.2656197011892392, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42274 + }, + { + "epoch": 0.42275, + "grad_norm": 1.6996990181025728, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 42275 + }, + { + "epoch": 0.42276, + "grad_norm": 1.3197987728334277, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42276 + }, + { + "epoch": 0.42277, + "grad_norm": 1.7209242233431623, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 42277 + }, + { + "epoch": 0.42278, + "grad_norm": 1.4082910339795112, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 42278 + }, + { + "epoch": 0.42279, + "grad_norm": 1.4320976263883574, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 42279 + }, + { + "epoch": 0.4228, + "grad_norm": 1.6408877256363967, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 42280 + }, + { + "epoch": 0.42281, + "grad_norm": 1.3830903372956393, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42281 + }, + { + "epoch": 0.42282, + "grad_norm": 1.3596740416296655, + "learning_rate": 0.003, + "loss": 4.037, + "step": 42282 + }, + { + "epoch": 0.42283, + "grad_norm": 1.7423921007693, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 42283 + }, + { + "epoch": 0.42284, + "grad_norm": 1.3037869038846277, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 42284 + }, + { + "epoch": 0.42285, + "grad_norm": 2.0046074678870625, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42285 + }, + { + "epoch": 0.42286, + "grad_norm": 1.6814680995925446, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 42286 + }, + { + "epoch": 0.42287, + "grad_norm": 1.2599821633873651, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42287 + }, + { + "epoch": 0.42288, + "grad_norm": 1.741300342816574, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 42288 + }, + { + "epoch": 0.42289, + "grad_norm": 1.4289694818628402, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 42289 + }, + { + "epoch": 0.4229, + "grad_norm": 1.265317619352791, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 42290 + }, + { + "epoch": 0.42291, + "grad_norm": 1.8448357736529368, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 42291 + }, + { + "epoch": 0.42292, + "grad_norm": 1.4628441709091426, + "learning_rate": 0.003, + "loss": 4.042, + "step": 42292 + }, + { + "epoch": 0.42293, + "grad_norm": 1.5638290805161916, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 42293 + }, + { + "epoch": 0.42294, + "grad_norm": 1.4942202284072599, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 42294 + }, + { + "epoch": 0.42295, + "grad_norm": 1.7435696341059335, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 42295 + }, + { + "epoch": 0.42296, + "grad_norm": 1.6720568045832371, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42296 + }, + { + "epoch": 0.42297, + "grad_norm": 1.2366977603022555, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 42297 + }, + { + "epoch": 0.42298, + "grad_norm": 1.7479795250262051, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 42298 + }, + { + "epoch": 0.42299, + "grad_norm": 1.2059890698203566, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 42299 + }, + { + "epoch": 0.423, + "grad_norm": 1.5058465316743592, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42300 + }, + { + "epoch": 0.42301, + "grad_norm": 1.721547716629741, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 42301 + }, + { + "epoch": 0.42302, + "grad_norm": 1.3688169435941882, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42302 + }, + { + "epoch": 0.42303, + "grad_norm": 1.7396480393865015, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 42303 + }, + { + "epoch": 0.42304, + "grad_norm": 1.4413543879193043, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42304 + }, + { + "epoch": 0.42305, + "grad_norm": 1.6330977807108427, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 42305 + }, + { + "epoch": 0.42306, + "grad_norm": 1.515008317698458, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 42306 + }, + { + "epoch": 0.42307, + "grad_norm": 1.4369612607072715, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 42307 + }, + { + "epoch": 0.42308, + "grad_norm": 1.692071835316365, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 42308 + }, + { + "epoch": 0.42309, + "grad_norm": 1.4292076848229611, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 42309 + }, + { + "epoch": 0.4231, + "grad_norm": 1.3214796092755798, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 42310 + }, + { + "epoch": 0.42311, + "grad_norm": 1.4556656480295371, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 42311 + }, + { + "epoch": 0.42312, + "grad_norm": 1.1627837643128554, + "learning_rate": 0.003, + "loss": 4.051, + "step": 42312 + }, + { + "epoch": 0.42313, + "grad_norm": 1.6873785411293727, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 42313 + }, + { + "epoch": 0.42314, + "grad_norm": 1.3198336639196242, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 42314 + }, + { + "epoch": 0.42315, + "grad_norm": 1.7339970680561716, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 42315 + }, + { + "epoch": 0.42316, + "grad_norm": 1.4359950293424162, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 42316 + }, + { + "epoch": 0.42317, + "grad_norm": 1.8337777570337725, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 42317 + }, + { + "epoch": 0.42318, + "grad_norm": 1.3230028297689547, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 42318 + }, + { + "epoch": 0.42319, + "grad_norm": 1.4691965024422984, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42319 + }, + { + "epoch": 0.4232, + "grad_norm": 1.980277795709122, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 42320 + }, + { + "epoch": 0.42321, + "grad_norm": 1.116559348095621, + "learning_rate": 0.003, + "loss": 4.042, + "step": 42321 + }, + { + "epoch": 0.42322, + "grad_norm": 1.6634757399098579, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 42322 + }, + { + "epoch": 0.42323, + "grad_norm": 1.4580447169173538, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42323 + }, + { + "epoch": 0.42324, + "grad_norm": 1.7143798879641976, + "learning_rate": 0.003, + "loss": 4.055, + "step": 42324 + }, + { + "epoch": 0.42325, + "grad_norm": 1.335932293137067, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 42325 + }, + { + "epoch": 0.42326, + "grad_norm": 1.74493936194905, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 42326 + }, + { + "epoch": 0.42327, + "grad_norm": 1.292682278284292, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42327 + }, + { + "epoch": 0.42328, + "grad_norm": 2.1203728649772184, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 42328 + }, + { + "epoch": 0.42329, + "grad_norm": 1.4604556860733644, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 42329 + }, + { + "epoch": 0.4233, + "grad_norm": 1.5582410151549575, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 42330 + }, + { + "epoch": 0.42331, + "grad_norm": 1.5101724674901835, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42331 + }, + { + "epoch": 0.42332, + "grad_norm": 1.7114412799973064, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 42332 + }, + { + "epoch": 0.42333, + "grad_norm": 1.2406126283852434, + "learning_rate": 0.003, + "loss": 4.073, + "step": 42333 + }, + { + "epoch": 0.42334, + "grad_norm": 1.911468564804136, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 42334 + }, + { + "epoch": 0.42335, + "grad_norm": 1.127641103797356, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 42335 + }, + { + "epoch": 0.42336, + "grad_norm": 2.0496127816868133, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 42336 + }, + { + "epoch": 0.42337, + "grad_norm": 1.4364538171239063, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42337 + }, + { + "epoch": 0.42338, + "grad_norm": 1.4818285218031817, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 42338 + }, + { + "epoch": 0.42339, + "grad_norm": 1.3999228408395472, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 42339 + }, + { + "epoch": 0.4234, + "grad_norm": 1.7318248839231012, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 42340 + }, + { + "epoch": 0.42341, + "grad_norm": 1.6838064554379613, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 42341 + }, + { + "epoch": 0.42342, + "grad_norm": 1.6102413665307465, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 42342 + }, + { + "epoch": 0.42343, + "grad_norm": 1.34535470439001, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 42343 + }, + { + "epoch": 0.42344, + "grad_norm": 1.5121929067979885, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42344 + }, + { + "epoch": 0.42345, + "grad_norm": 1.3282782063141183, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42345 + }, + { + "epoch": 0.42346, + "grad_norm": 1.5842894809738208, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42346 + }, + { + "epoch": 0.42347, + "grad_norm": 1.4305046707477773, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 42347 + }, + { + "epoch": 0.42348, + "grad_norm": 1.546228700668396, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 42348 + }, + { + "epoch": 0.42349, + "grad_norm": 1.1239477351824825, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 42349 + }, + { + "epoch": 0.4235, + "grad_norm": 1.946729744345668, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 42350 + }, + { + "epoch": 0.42351, + "grad_norm": 1.1690147799661763, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 42351 + }, + { + "epoch": 0.42352, + "grad_norm": 1.6556542057096335, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 42352 + }, + { + "epoch": 0.42353, + "grad_norm": 1.4668715548493583, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 42353 + }, + { + "epoch": 0.42354, + "grad_norm": 1.64681162881559, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42354 + }, + { + "epoch": 0.42355, + "grad_norm": 1.5689428271314165, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42355 + }, + { + "epoch": 0.42356, + "grad_norm": 1.6455953316617935, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 42356 + }, + { + "epoch": 0.42357, + "grad_norm": 1.4274042794204087, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 42357 + }, + { + "epoch": 0.42358, + "grad_norm": 1.598400586742122, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 42358 + }, + { + "epoch": 0.42359, + "grad_norm": 1.326098560907333, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 42359 + }, + { + "epoch": 0.4236, + "grad_norm": 1.8006927824225671, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 42360 + }, + { + "epoch": 0.42361, + "grad_norm": 1.7999995420587125, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 42361 + }, + { + "epoch": 0.42362, + "grad_norm": 1.3417928711315612, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 42362 + }, + { + "epoch": 0.42363, + "grad_norm": 1.461719704571237, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 42363 + }, + { + "epoch": 0.42364, + "grad_norm": 1.6160935780742787, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 42364 + }, + { + "epoch": 0.42365, + "grad_norm": 1.521728192216615, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 42365 + }, + { + "epoch": 0.42366, + "grad_norm": 1.3494040582902112, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 42366 + }, + { + "epoch": 0.42367, + "grad_norm": 1.9191820089780016, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 42367 + }, + { + "epoch": 0.42368, + "grad_norm": 1.2534524475007651, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 42368 + }, + { + "epoch": 0.42369, + "grad_norm": 1.5580588649337954, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 42369 + }, + { + "epoch": 0.4237, + "grad_norm": 1.5482187179935407, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 42370 + }, + { + "epoch": 0.42371, + "grad_norm": 1.4978047515035986, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 42371 + }, + { + "epoch": 0.42372, + "grad_norm": 1.2633430144761721, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 42372 + }, + { + "epoch": 0.42373, + "grad_norm": 1.6074013238477385, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 42373 + }, + { + "epoch": 0.42374, + "grad_norm": 1.433985652319794, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42374 + }, + { + "epoch": 0.42375, + "grad_norm": 1.3688377031575172, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42375 + }, + { + "epoch": 0.42376, + "grad_norm": 1.8403159624703762, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 42376 + }, + { + "epoch": 0.42377, + "grad_norm": 1.2209895911160897, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 42377 + }, + { + "epoch": 0.42378, + "grad_norm": 2.014003012439158, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42378 + }, + { + "epoch": 0.42379, + "grad_norm": 1.5775515909463385, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 42379 + }, + { + "epoch": 0.4238, + "grad_norm": 1.8291232422814374, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 42380 + }, + { + "epoch": 0.42381, + "grad_norm": 1.3639871678073985, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 42381 + }, + { + "epoch": 0.42382, + "grad_norm": 1.5590754098801636, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 42382 + }, + { + "epoch": 0.42383, + "grad_norm": 1.5104399348540019, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42383 + }, + { + "epoch": 0.42384, + "grad_norm": 1.3083975071506098, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 42384 + }, + { + "epoch": 0.42385, + "grad_norm": 1.5568692222460985, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 42385 + }, + { + "epoch": 0.42386, + "grad_norm": 1.5909722638627086, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 42386 + }, + { + "epoch": 0.42387, + "grad_norm": 1.8215784568425721, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 42387 + }, + { + "epoch": 0.42388, + "grad_norm": 1.0928412090271034, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42388 + }, + { + "epoch": 0.42389, + "grad_norm": 1.642558159214056, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 42389 + }, + { + "epoch": 0.4239, + "grad_norm": 1.2103712088133831, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 42390 + }, + { + "epoch": 0.42391, + "grad_norm": 1.7082520119972706, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 42391 + }, + { + "epoch": 0.42392, + "grad_norm": 1.3404250219667555, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 42392 + }, + { + "epoch": 0.42393, + "grad_norm": 1.5583634207243053, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 42393 + }, + { + "epoch": 0.42394, + "grad_norm": 1.4564117472377494, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 42394 + }, + { + "epoch": 0.42395, + "grad_norm": 1.577947250929734, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 42395 + }, + { + "epoch": 0.42396, + "grad_norm": 1.9159025483095553, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 42396 + }, + { + "epoch": 0.42397, + "grad_norm": 1.6907830873534166, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42397 + }, + { + "epoch": 0.42398, + "grad_norm": 1.4561718443835379, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 42398 + }, + { + "epoch": 0.42399, + "grad_norm": 1.666619592428573, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 42399 + }, + { + "epoch": 0.424, + "grad_norm": 1.2524417051063048, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 42400 + }, + { + "epoch": 0.42401, + "grad_norm": 1.6845459401319813, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 42401 + }, + { + "epoch": 0.42402, + "grad_norm": 1.3296038265030687, + "learning_rate": 0.003, + "loss": 4.031, + "step": 42402 + }, + { + "epoch": 0.42403, + "grad_norm": 1.366573055797645, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 42403 + }, + { + "epoch": 0.42404, + "grad_norm": 2.3214906280421035, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 42404 + }, + { + "epoch": 0.42405, + "grad_norm": 1.0282395148625905, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42405 + }, + { + "epoch": 0.42406, + "grad_norm": 1.8110036290197733, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 42406 + }, + { + "epoch": 0.42407, + "grad_norm": 1.3269952354317334, + "learning_rate": 0.003, + "loss": 4.052, + "step": 42407 + }, + { + "epoch": 0.42408, + "grad_norm": 1.443036831354477, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 42408 + }, + { + "epoch": 0.42409, + "grad_norm": 1.45760661240061, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 42409 + }, + { + "epoch": 0.4241, + "grad_norm": 1.3164645834287856, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 42410 + }, + { + "epoch": 0.42411, + "grad_norm": 1.629928642923893, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42411 + }, + { + "epoch": 0.42412, + "grad_norm": 1.9026585152869029, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 42412 + }, + { + "epoch": 0.42413, + "grad_norm": 1.3983231240624125, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 42413 + }, + { + "epoch": 0.42414, + "grad_norm": 1.8877051007568704, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 42414 + }, + { + "epoch": 0.42415, + "grad_norm": 1.5912458046075808, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42415 + }, + { + "epoch": 0.42416, + "grad_norm": 1.567457175506356, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 42416 + }, + { + "epoch": 0.42417, + "grad_norm": 1.6939115086448002, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 42417 + }, + { + "epoch": 0.42418, + "grad_norm": 1.3896617519354746, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 42418 + }, + { + "epoch": 0.42419, + "grad_norm": 1.6053713901837126, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 42419 + }, + { + "epoch": 0.4242, + "grad_norm": 1.603912790802956, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42420 + }, + { + "epoch": 0.42421, + "grad_norm": 1.3312773321178641, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 42421 + }, + { + "epoch": 0.42422, + "grad_norm": 1.7603006365080285, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 42422 + }, + { + "epoch": 0.42423, + "grad_norm": 1.317750024756286, + "learning_rate": 0.003, + "loss": 4.07, + "step": 42423 + }, + { + "epoch": 0.42424, + "grad_norm": 1.8181471043093897, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 42424 + }, + { + "epoch": 0.42425, + "grad_norm": 1.5049634325152437, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 42425 + }, + { + "epoch": 0.42426, + "grad_norm": 1.4368122059563184, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 42426 + }, + { + "epoch": 0.42427, + "grad_norm": 1.5282854017653462, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 42427 + }, + { + "epoch": 0.42428, + "grad_norm": 1.6447885428448412, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42428 + }, + { + "epoch": 0.42429, + "grad_norm": 1.3336750389049072, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 42429 + }, + { + "epoch": 0.4243, + "grad_norm": 1.6662732959731354, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 42430 + }, + { + "epoch": 0.42431, + "grad_norm": 1.34623496503156, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 42431 + }, + { + "epoch": 0.42432, + "grad_norm": 1.5001111480867066, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 42432 + }, + { + "epoch": 0.42433, + "grad_norm": 1.730381966992078, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 42433 + }, + { + "epoch": 0.42434, + "grad_norm": 1.5421060422421655, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 42434 + }, + { + "epoch": 0.42435, + "grad_norm": 2.085905365648638, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42435 + }, + { + "epoch": 0.42436, + "grad_norm": 1.647100076449646, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42436 + }, + { + "epoch": 0.42437, + "grad_norm": 1.4482533446341035, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42437 + }, + { + "epoch": 0.42438, + "grad_norm": 1.6804736203717807, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 42438 + }, + { + "epoch": 0.42439, + "grad_norm": 1.3352048784604822, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 42439 + }, + { + "epoch": 0.4244, + "grad_norm": 1.730681476718155, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42440 + }, + { + "epoch": 0.42441, + "grad_norm": 1.4475677165426946, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 42441 + }, + { + "epoch": 0.42442, + "grad_norm": 1.685892696741247, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 42442 + }, + { + "epoch": 0.42443, + "grad_norm": 1.4120939858749824, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 42443 + }, + { + "epoch": 0.42444, + "grad_norm": 1.464379782319354, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 42444 + }, + { + "epoch": 0.42445, + "grad_norm": 1.3437645563917011, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 42445 + }, + { + "epoch": 0.42446, + "grad_norm": 1.696185171787527, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42446 + }, + { + "epoch": 0.42447, + "grad_norm": 1.7288231393091276, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 42447 + }, + { + "epoch": 0.42448, + "grad_norm": 1.2997761823708047, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42448 + }, + { + "epoch": 0.42449, + "grad_norm": 1.7396285249863987, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 42449 + }, + { + "epoch": 0.4245, + "grad_norm": 1.4535174700182243, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 42450 + }, + { + "epoch": 0.42451, + "grad_norm": 1.5298664616867903, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 42451 + }, + { + "epoch": 0.42452, + "grad_norm": 1.6629012387621724, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 42452 + }, + { + "epoch": 0.42453, + "grad_norm": 1.4058916269816168, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 42453 + }, + { + "epoch": 0.42454, + "grad_norm": 1.5890944515047578, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 42454 + }, + { + "epoch": 0.42455, + "grad_norm": 1.3549368832836464, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 42455 + }, + { + "epoch": 0.42456, + "grad_norm": 1.768789095094692, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42456 + }, + { + "epoch": 0.42457, + "grad_norm": 1.3454192621984205, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 42457 + }, + { + "epoch": 0.42458, + "grad_norm": 1.4014904693294374, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 42458 + }, + { + "epoch": 0.42459, + "grad_norm": 1.511837950413729, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 42459 + }, + { + "epoch": 0.4246, + "grad_norm": 1.5205541594756746, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 42460 + }, + { + "epoch": 0.42461, + "grad_norm": 1.3025749149048966, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42461 + }, + { + "epoch": 0.42462, + "grad_norm": 1.4769210251985132, + "learning_rate": 0.003, + "loss": 4.013, + "step": 42462 + }, + { + "epoch": 0.42463, + "grad_norm": 1.578494668327848, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 42463 + }, + { + "epoch": 0.42464, + "grad_norm": 1.9011380280869294, + "learning_rate": 0.003, + "loss": 4.043, + "step": 42464 + }, + { + "epoch": 0.42465, + "grad_norm": 1.3953796815049893, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 42465 + }, + { + "epoch": 0.42466, + "grad_norm": 1.7322194268239475, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 42466 + }, + { + "epoch": 0.42467, + "grad_norm": 1.3494592058418546, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 42467 + }, + { + "epoch": 0.42468, + "grad_norm": 1.8566581487953417, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 42468 + }, + { + "epoch": 0.42469, + "grad_norm": 1.4540816601206168, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 42469 + }, + { + "epoch": 0.4247, + "grad_norm": 1.8074465945577842, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 42470 + }, + { + "epoch": 0.42471, + "grad_norm": 1.3191962147683634, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 42471 + }, + { + "epoch": 0.42472, + "grad_norm": 1.7166767480858456, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42472 + }, + { + "epoch": 0.42473, + "grad_norm": 1.3965313680468328, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42473 + }, + { + "epoch": 0.42474, + "grad_norm": 1.861809784044772, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 42474 + }, + { + "epoch": 0.42475, + "grad_norm": 1.504924338988313, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 42475 + }, + { + "epoch": 0.42476, + "grad_norm": 1.653675014026, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 42476 + }, + { + "epoch": 0.42477, + "grad_norm": 1.0769305680940284, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 42477 + }, + { + "epoch": 0.42478, + "grad_norm": 1.9120951728386684, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 42478 + }, + { + "epoch": 0.42479, + "grad_norm": 1.4000455600551085, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 42479 + }, + { + "epoch": 0.4248, + "grad_norm": 1.6127702120141287, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42480 + }, + { + "epoch": 0.42481, + "grad_norm": 1.519225719833365, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 42481 + }, + { + "epoch": 0.42482, + "grad_norm": 1.5093739214588864, + "learning_rate": 0.003, + "loss": 4.002, + "step": 42482 + }, + { + "epoch": 0.42483, + "grad_norm": 1.5442564505646605, + "learning_rate": 0.003, + "loss": 4.048, + "step": 42483 + }, + { + "epoch": 0.42484, + "grad_norm": 1.5627069386169588, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42484 + }, + { + "epoch": 0.42485, + "grad_norm": 1.6426505812949594, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42485 + }, + { + "epoch": 0.42486, + "grad_norm": 1.6453954477876904, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42486 + }, + { + "epoch": 0.42487, + "grad_norm": 1.6079840138410049, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 42487 + }, + { + "epoch": 0.42488, + "grad_norm": 1.3241466135940057, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 42488 + }, + { + "epoch": 0.42489, + "grad_norm": 1.7286195290925934, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42489 + }, + { + "epoch": 0.4249, + "grad_norm": 1.3542575608934069, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 42490 + }, + { + "epoch": 0.42491, + "grad_norm": 1.6875730738460069, + "learning_rate": 0.003, + "loss": 4.025, + "step": 42491 + }, + { + "epoch": 0.42492, + "grad_norm": 1.4803866606289626, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 42492 + }, + { + "epoch": 0.42493, + "grad_norm": 1.509070075158149, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42493 + }, + { + "epoch": 0.42494, + "grad_norm": 1.3688200301792066, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42494 + }, + { + "epoch": 0.42495, + "grad_norm": 1.5749003633048013, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42495 + }, + { + "epoch": 0.42496, + "grad_norm": 1.4706828538043721, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 42496 + }, + { + "epoch": 0.42497, + "grad_norm": 1.4981800330254027, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42497 + }, + { + "epoch": 0.42498, + "grad_norm": 1.603472451203607, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 42498 + }, + { + "epoch": 0.42499, + "grad_norm": 1.6283886870935904, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 42499 + }, + { + "epoch": 0.425, + "grad_norm": 1.6469295964317623, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 42500 + }, + { + "epoch": 0.42501, + "grad_norm": 1.575540511247959, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 42501 + }, + { + "epoch": 0.42502, + "grad_norm": 1.59733857601778, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42502 + }, + { + "epoch": 0.42503, + "grad_norm": 1.2952609551150074, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 42503 + }, + { + "epoch": 0.42504, + "grad_norm": 1.4733929650303794, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 42504 + }, + { + "epoch": 0.42505, + "grad_norm": 1.4648060161097356, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42505 + }, + { + "epoch": 0.42506, + "grad_norm": 1.476318626431549, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42506 + }, + { + "epoch": 0.42507, + "grad_norm": 1.7381935827256272, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 42507 + }, + { + "epoch": 0.42508, + "grad_norm": 1.235515092512556, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 42508 + }, + { + "epoch": 0.42509, + "grad_norm": 1.9018387159053665, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 42509 + }, + { + "epoch": 0.4251, + "grad_norm": 1.3043024319749659, + "learning_rate": 0.003, + "loss": 4.056, + "step": 42510 + }, + { + "epoch": 0.42511, + "grad_norm": 1.7097971050349539, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 42511 + }, + { + "epoch": 0.42512, + "grad_norm": 1.351019949079863, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 42512 + }, + { + "epoch": 0.42513, + "grad_norm": 1.6527153195574773, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 42513 + }, + { + "epoch": 0.42514, + "grad_norm": 1.5987844946454213, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 42514 + }, + { + "epoch": 0.42515, + "grad_norm": 1.763144865669518, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 42515 + }, + { + "epoch": 0.42516, + "grad_norm": 1.2685115077483824, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 42516 + }, + { + "epoch": 0.42517, + "grad_norm": 1.6679593296388546, + "learning_rate": 0.003, + "loss": 4.036, + "step": 42517 + }, + { + "epoch": 0.42518, + "grad_norm": 1.6001829397393965, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 42518 + }, + { + "epoch": 0.42519, + "grad_norm": 1.5345200214879562, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 42519 + }, + { + "epoch": 0.4252, + "grad_norm": 1.661897544562786, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42520 + }, + { + "epoch": 0.42521, + "grad_norm": 1.4471478485632512, + "learning_rate": 0.003, + "loss": 4.05, + "step": 42521 + }, + { + "epoch": 0.42522, + "grad_norm": 1.6486335435174013, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 42522 + }, + { + "epoch": 0.42523, + "grad_norm": 1.0941973932422737, + "learning_rate": 0.003, + "loss": 4.018, + "step": 42523 + }, + { + "epoch": 0.42524, + "grad_norm": 1.7625141881751432, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 42524 + }, + { + "epoch": 0.42525, + "grad_norm": 1.3860894332181721, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 42525 + }, + { + "epoch": 0.42526, + "grad_norm": 1.425093384046108, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42526 + }, + { + "epoch": 0.42527, + "grad_norm": 1.8023105930063834, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 42527 + }, + { + "epoch": 0.42528, + "grad_norm": 1.248066170715957, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 42528 + }, + { + "epoch": 0.42529, + "grad_norm": 1.8614799220749572, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 42529 + }, + { + "epoch": 0.4253, + "grad_norm": 1.4536723769226083, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 42530 + }, + { + "epoch": 0.42531, + "grad_norm": 1.9067508337540497, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42531 + }, + { + "epoch": 0.42532, + "grad_norm": 1.522571701027982, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42532 + }, + { + "epoch": 0.42533, + "grad_norm": 1.4297111316700322, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 42533 + }, + { + "epoch": 0.42534, + "grad_norm": 1.474025984287538, + "learning_rate": 0.003, + "loss": 4.055, + "step": 42534 + }, + { + "epoch": 0.42535, + "grad_norm": 1.4608969208273528, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 42535 + }, + { + "epoch": 0.42536, + "grad_norm": 1.4895618588084762, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 42536 + }, + { + "epoch": 0.42537, + "grad_norm": 1.4348794714698543, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 42537 + }, + { + "epoch": 0.42538, + "grad_norm": 1.803758899521196, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 42538 + }, + { + "epoch": 0.42539, + "grad_norm": 1.4418480584875863, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 42539 + }, + { + "epoch": 0.4254, + "grad_norm": 1.7394741690101212, + "learning_rate": 0.003, + "loss": 4.034, + "step": 42540 + }, + { + "epoch": 0.42541, + "grad_norm": 1.2777494509962488, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 42541 + }, + { + "epoch": 0.42542, + "grad_norm": 1.7661362702191055, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42542 + }, + { + "epoch": 0.42543, + "grad_norm": 1.3900129427097818, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 42543 + }, + { + "epoch": 0.42544, + "grad_norm": 1.5424467617162132, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 42544 + }, + { + "epoch": 0.42545, + "grad_norm": 1.5681719508886716, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 42545 + }, + { + "epoch": 0.42546, + "grad_norm": 1.516154832134409, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42546 + }, + { + "epoch": 0.42547, + "grad_norm": 1.6966365834564878, + "learning_rate": 0.003, + "loss": 4.07, + "step": 42547 + }, + { + "epoch": 0.42548, + "grad_norm": 1.5632393690950919, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 42548 + }, + { + "epoch": 0.42549, + "grad_norm": 1.3790230756138253, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 42549 + }, + { + "epoch": 0.4255, + "grad_norm": 1.7179559015690316, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 42550 + }, + { + "epoch": 0.42551, + "grad_norm": 1.3803704517049653, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42551 + }, + { + "epoch": 0.42552, + "grad_norm": 1.5181852106668405, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42552 + }, + { + "epoch": 0.42553, + "grad_norm": 1.7359472753146046, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 42553 + }, + { + "epoch": 0.42554, + "grad_norm": 1.279413462878253, + "learning_rate": 0.003, + "loss": 4.025, + "step": 42554 + }, + { + "epoch": 0.42555, + "grad_norm": 1.821393787761916, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 42555 + }, + { + "epoch": 0.42556, + "grad_norm": 1.3707411233241245, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 42556 + }, + { + "epoch": 0.42557, + "grad_norm": 1.7055519537616028, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 42557 + }, + { + "epoch": 0.42558, + "grad_norm": 1.698699506462451, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 42558 + }, + { + "epoch": 0.42559, + "grad_norm": 1.3774798623213118, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42559 + }, + { + "epoch": 0.4256, + "grad_norm": 1.5674253578873727, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 42560 + }, + { + "epoch": 0.42561, + "grad_norm": 1.3308804795109508, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42561 + }, + { + "epoch": 0.42562, + "grad_norm": 1.463781549675307, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42562 + }, + { + "epoch": 0.42563, + "grad_norm": 1.442977263906723, + "learning_rate": 0.003, + "loss": 4.024, + "step": 42563 + }, + { + "epoch": 0.42564, + "grad_norm": 1.6826544599092252, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 42564 + }, + { + "epoch": 0.42565, + "grad_norm": 1.7166276231091184, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 42565 + }, + { + "epoch": 0.42566, + "grad_norm": 1.123508069144283, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 42566 + }, + { + "epoch": 0.42567, + "grad_norm": 1.9239230408088182, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 42567 + }, + { + "epoch": 0.42568, + "grad_norm": 1.663855535737067, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 42568 + }, + { + "epoch": 0.42569, + "grad_norm": 1.4126419140862592, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 42569 + }, + { + "epoch": 0.4257, + "grad_norm": 1.7599609615064116, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 42570 + }, + { + "epoch": 0.42571, + "grad_norm": 1.381339331560802, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 42571 + }, + { + "epoch": 0.42572, + "grad_norm": 1.4431094799525073, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 42572 + }, + { + "epoch": 0.42573, + "grad_norm": 1.6628308739908024, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42573 + }, + { + "epoch": 0.42574, + "grad_norm": 1.3756437956322753, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 42574 + }, + { + "epoch": 0.42575, + "grad_norm": 1.4882358095085726, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 42575 + }, + { + "epoch": 0.42576, + "grad_norm": 1.8420807806480302, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 42576 + }, + { + "epoch": 0.42577, + "grad_norm": 1.6540703390783265, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 42577 + }, + { + "epoch": 0.42578, + "grad_norm": 1.3658429583941458, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 42578 + }, + { + "epoch": 0.42579, + "grad_norm": 1.5350824961349225, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 42579 + }, + { + "epoch": 0.4258, + "grad_norm": 1.356073873460834, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 42580 + }, + { + "epoch": 0.42581, + "grad_norm": 1.7303515883017309, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42581 + }, + { + "epoch": 0.42582, + "grad_norm": 1.5051615914784697, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 42582 + }, + { + "epoch": 0.42583, + "grad_norm": 1.541016700994415, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 42583 + }, + { + "epoch": 0.42584, + "grad_norm": 1.5274905248259125, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 42584 + }, + { + "epoch": 0.42585, + "grad_norm": 1.5396329730460154, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 42585 + }, + { + "epoch": 0.42586, + "grad_norm": 1.3712893110579998, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 42586 + }, + { + "epoch": 0.42587, + "grad_norm": 1.5066906602465464, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 42587 + }, + { + "epoch": 0.42588, + "grad_norm": 1.2985677758732257, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42588 + }, + { + "epoch": 0.42589, + "grad_norm": 1.6398428058076762, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 42589 + }, + { + "epoch": 0.4259, + "grad_norm": 1.360537497288641, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 42590 + }, + { + "epoch": 0.42591, + "grad_norm": 1.4102894709334732, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42591 + }, + { + "epoch": 0.42592, + "grad_norm": 1.5774672263717229, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 42592 + }, + { + "epoch": 0.42593, + "grad_norm": 1.6179834642430049, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 42593 + }, + { + "epoch": 0.42594, + "grad_norm": 1.5828660344311758, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42594 + }, + { + "epoch": 0.42595, + "grad_norm": 1.5311897736600568, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42595 + }, + { + "epoch": 0.42596, + "grad_norm": 1.504973442405231, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 42596 + }, + { + "epoch": 0.42597, + "grad_norm": 1.7014077740724942, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42597 + }, + { + "epoch": 0.42598, + "grad_norm": 1.4317176251828712, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42598 + }, + { + "epoch": 0.42599, + "grad_norm": 1.7567091561504704, + "learning_rate": 0.003, + "loss": 4.035, + "step": 42599 + }, + { + "epoch": 0.426, + "grad_norm": 1.2954774728273741, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 42600 + }, + { + "epoch": 0.42601, + "grad_norm": 1.675105332181385, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 42601 + }, + { + "epoch": 0.42602, + "grad_norm": 1.351133590272317, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 42602 + }, + { + "epoch": 0.42603, + "grad_norm": 2.088827722881763, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 42603 + }, + { + "epoch": 0.42604, + "grad_norm": 1.1338716199232228, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 42604 + }, + { + "epoch": 0.42605, + "grad_norm": 1.8349078363472424, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 42605 + }, + { + "epoch": 0.42606, + "grad_norm": 1.5253538538132452, + "learning_rate": 0.003, + "loss": 4.034, + "step": 42606 + }, + { + "epoch": 0.42607, + "grad_norm": 1.5153566151412121, + "learning_rate": 0.003, + "loss": 4.023, + "step": 42607 + }, + { + "epoch": 0.42608, + "grad_norm": 1.5660938385311713, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 42608 + }, + { + "epoch": 0.42609, + "grad_norm": 1.2628909567324562, + "learning_rate": 0.003, + "loss": 4.017, + "step": 42609 + }, + { + "epoch": 0.4261, + "grad_norm": 1.7216647392862228, + "learning_rate": 0.003, + "loss": 4.046, + "step": 42610 + }, + { + "epoch": 0.42611, + "grad_norm": 1.3691102959091412, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 42611 + }, + { + "epoch": 0.42612, + "grad_norm": 1.719193262434752, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 42612 + }, + { + "epoch": 0.42613, + "grad_norm": 1.3411842150348772, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 42613 + }, + { + "epoch": 0.42614, + "grad_norm": 1.6933790073128898, + "learning_rate": 0.003, + "loss": 4.067, + "step": 42614 + }, + { + "epoch": 0.42615, + "grad_norm": 1.3161798563356324, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 42615 + }, + { + "epoch": 0.42616, + "grad_norm": 1.5784679806658473, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 42616 + }, + { + "epoch": 0.42617, + "grad_norm": 1.5958022739963518, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 42617 + }, + { + "epoch": 0.42618, + "grad_norm": 1.3955393626385502, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 42618 + }, + { + "epoch": 0.42619, + "grad_norm": 1.4961098030189075, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 42619 + }, + { + "epoch": 0.4262, + "grad_norm": 1.357874062566721, + "learning_rate": 0.003, + "loss": 4.038, + "step": 42620 + }, + { + "epoch": 0.42621, + "grad_norm": 1.637352486059841, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 42621 + }, + { + "epoch": 0.42622, + "grad_norm": 1.5125134817225971, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 42622 + }, + { + "epoch": 0.42623, + "grad_norm": 2.037778573080359, + "learning_rate": 0.003, + "loss": 4.045, + "step": 42623 + }, + { + "epoch": 0.42624, + "grad_norm": 1.4447051408068825, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 42624 + }, + { + "epoch": 0.42625, + "grad_norm": 1.614239087704616, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42625 + }, + { + "epoch": 0.42626, + "grad_norm": 1.5986232860502805, + "learning_rate": 0.003, + "loss": 4.062, + "step": 42626 + }, + { + "epoch": 0.42627, + "grad_norm": 1.83748044895175, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42627 + }, + { + "epoch": 0.42628, + "grad_norm": 1.3901197761388309, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 42628 + }, + { + "epoch": 0.42629, + "grad_norm": 1.5552320399577535, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 42629 + }, + { + "epoch": 0.4263, + "grad_norm": 1.687052931781383, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 42630 + }, + { + "epoch": 0.42631, + "grad_norm": 1.240240870461935, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 42631 + }, + { + "epoch": 0.42632, + "grad_norm": 1.6930627820341249, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 42632 + }, + { + "epoch": 0.42633, + "grad_norm": 1.3117816603350678, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42633 + }, + { + "epoch": 0.42634, + "grad_norm": 1.5045064136017283, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 42634 + }, + { + "epoch": 0.42635, + "grad_norm": 1.4001372796455531, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 42635 + }, + { + "epoch": 0.42636, + "grad_norm": 1.548677438022398, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 42636 + }, + { + "epoch": 0.42637, + "grad_norm": 1.7850710746710778, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 42637 + }, + { + "epoch": 0.42638, + "grad_norm": 1.8197931977181885, + "learning_rate": 0.003, + "loss": 4.036, + "step": 42638 + }, + { + "epoch": 0.42639, + "grad_norm": 1.3995845372809492, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 42639 + }, + { + "epoch": 0.4264, + "grad_norm": 1.7018941959784475, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 42640 + }, + { + "epoch": 0.42641, + "grad_norm": 1.6161031433617292, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 42641 + }, + { + "epoch": 0.42642, + "grad_norm": 1.511455327529901, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 42642 + }, + { + "epoch": 0.42643, + "grad_norm": 1.5424568846115896, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 42643 + }, + { + "epoch": 0.42644, + "grad_norm": 1.2046762413872418, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 42644 + }, + { + "epoch": 0.42645, + "grad_norm": 1.6878351636588256, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 42645 + }, + { + "epoch": 0.42646, + "grad_norm": 1.3384219945785694, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 42646 + }, + { + "epoch": 0.42647, + "grad_norm": 1.5182056575000715, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 42647 + }, + { + "epoch": 0.42648, + "grad_norm": 1.3429962096148529, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 42648 + }, + { + "epoch": 0.42649, + "grad_norm": 1.9161139190399694, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42649 + }, + { + "epoch": 0.4265, + "grad_norm": 1.1190525752199434, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 42650 + }, + { + "epoch": 0.42651, + "grad_norm": 1.7275066559070287, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 42651 + }, + { + "epoch": 0.42652, + "grad_norm": 1.4936016251905566, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42652 + }, + { + "epoch": 0.42653, + "grad_norm": 1.7996562542205436, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42653 + }, + { + "epoch": 0.42654, + "grad_norm": 1.2140512490285258, + "learning_rate": 0.003, + "loss": 4.06, + "step": 42654 + }, + { + "epoch": 0.42655, + "grad_norm": 1.790940057901232, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 42655 + }, + { + "epoch": 0.42656, + "grad_norm": 1.074933803708933, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 42656 + }, + { + "epoch": 0.42657, + "grad_norm": 1.713988870820434, + "learning_rate": 0.003, + "loss": 4.042, + "step": 42657 + }, + { + "epoch": 0.42658, + "grad_norm": 1.7698875987049878, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 42658 + }, + { + "epoch": 0.42659, + "grad_norm": 1.6078449927993375, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 42659 + }, + { + "epoch": 0.4266, + "grad_norm": 1.3054034149283567, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 42660 + }, + { + "epoch": 0.42661, + "grad_norm": 1.790919697209442, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 42661 + }, + { + "epoch": 0.42662, + "grad_norm": 1.5774880365534147, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 42662 + }, + { + "epoch": 0.42663, + "grad_norm": 1.5026542297592043, + "learning_rate": 0.003, + "loss": 4.026, + "step": 42663 + }, + { + "epoch": 0.42664, + "grad_norm": 1.5538907245233102, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 42664 + }, + { + "epoch": 0.42665, + "grad_norm": 1.497729299129083, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 42665 + }, + { + "epoch": 0.42666, + "grad_norm": 1.4476556995491212, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 42666 + }, + { + "epoch": 0.42667, + "grad_norm": 1.50714793770638, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 42667 + }, + { + "epoch": 0.42668, + "grad_norm": 1.326205455554897, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42668 + }, + { + "epoch": 0.42669, + "grad_norm": 1.8599298007626621, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42669 + }, + { + "epoch": 0.4267, + "grad_norm": 1.238545380786275, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42670 + }, + { + "epoch": 0.42671, + "grad_norm": 1.827681908670954, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 42671 + }, + { + "epoch": 0.42672, + "grad_norm": 1.390100358766677, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 42672 + }, + { + "epoch": 0.42673, + "grad_norm": 1.5828092926348207, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 42673 + }, + { + "epoch": 0.42674, + "grad_norm": 1.6893737113242648, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 42674 + }, + { + "epoch": 0.42675, + "grad_norm": 1.35972136687345, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 42675 + }, + { + "epoch": 0.42676, + "grad_norm": 1.6045446917320345, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 42676 + }, + { + "epoch": 0.42677, + "grad_norm": 1.5773175083363358, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 42677 + }, + { + "epoch": 0.42678, + "grad_norm": 1.5993912708333573, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 42678 + }, + { + "epoch": 0.42679, + "grad_norm": 1.3998037966791925, + "learning_rate": 0.003, + "loss": 4.037, + "step": 42679 + }, + { + "epoch": 0.4268, + "grad_norm": 1.4361630843182214, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42680 + }, + { + "epoch": 0.42681, + "grad_norm": 1.4224195086224034, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42681 + }, + { + "epoch": 0.42682, + "grad_norm": 1.6312240829304265, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 42682 + }, + { + "epoch": 0.42683, + "grad_norm": 1.453399284271669, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 42683 + }, + { + "epoch": 0.42684, + "grad_norm": 1.2990142894513579, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 42684 + }, + { + "epoch": 0.42685, + "grad_norm": 1.7137532619897158, + "learning_rate": 0.003, + "loss": 4.066, + "step": 42685 + }, + { + "epoch": 0.42686, + "grad_norm": 1.5029964177110744, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 42686 + }, + { + "epoch": 0.42687, + "grad_norm": 1.509515608041772, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 42687 + }, + { + "epoch": 0.42688, + "grad_norm": 1.6834259709396788, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42688 + }, + { + "epoch": 0.42689, + "grad_norm": 1.6504459615063953, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 42689 + }, + { + "epoch": 0.4269, + "grad_norm": 1.6113273379850803, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 42690 + }, + { + "epoch": 0.42691, + "grad_norm": 1.403458381055645, + "learning_rate": 0.003, + "loss": 4.017, + "step": 42691 + }, + { + "epoch": 0.42692, + "grad_norm": 1.4393639545760932, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 42692 + }, + { + "epoch": 0.42693, + "grad_norm": 1.3856154924619755, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42693 + }, + { + "epoch": 0.42694, + "grad_norm": 1.6740080290247457, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42694 + }, + { + "epoch": 0.42695, + "grad_norm": 1.4108999635389383, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 42695 + }, + { + "epoch": 0.42696, + "grad_norm": 1.8275108901363215, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 42696 + }, + { + "epoch": 0.42697, + "grad_norm": 1.275367175498295, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 42697 + }, + { + "epoch": 0.42698, + "grad_norm": 1.6233765711143784, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 42698 + }, + { + "epoch": 0.42699, + "grad_norm": 1.570693390546288, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 42699 + }, + { + "epoch": 0.427, + "grad_norm": 1.4290219795362469, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 42700 + }, + { + "epoch": 0.42701, + "grad_norm": 1.776144196923217, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 42701 + }, + { + "epoch": 0.42702, + "grad_norm": 1.8497167056720343, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 42702 + }, + { + "epoch": 0.42703, + "grad_norm": 1.429660065734541, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 42703 + }, + { + "epoch": 0.42704, + "grad_norm": 1.3418335696333872, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 42704 + }, + { + "epoch": 0.42705, + "grad_norm": 2.095249280515914, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 42705 + }, + { + "epoch": 0.42706, + "grad_norm": 1.3178509747930653, + "learning_rate": 0.003, + "loss": 4.01, + "step": 42706 + }, + { + "epoch": 0.42707, + "grad_norm": 1.9207321669314794, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 42707 + }, + { + "epoch": 0.42708, + "grad_norm": 1.326903027152168, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42708 + }, + { + "epoch": 0.42709, + "grad_norm": 1.4877413275913518, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 42709 + }, + { + "epoch": 0.4271, + "grad_norm": 1.482336335409253, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 42710 + }, + { + "epoch": 0.42711, + "grad_norm": 1.4786883933216577, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 42711 + }, + { + "epoch": 0.42712, + "grad_norm": 1.4291268114309152, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 42712 + }, + { + "epoch": 0.42713, + "grad_norm": 1.5264653656941367, + "learning_rate": 0.003, + "loss": 4.068, + "step": 42713 + }, + { + "epoch": 0.42714, + "grad_norm": 1.5419649001095517, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42714 + }, + { + "epoch": 0.42715, + "grad_norm": 1.8136345912367522, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 42715 + }, + { + "epoch": 0.42716, + "grad_norm": 1.5381730830929174, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 42716 + }, + { + "epoch": 0.42717, + "grad_norm": 1.5834026451606449, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 42717 + }, + { + "epoch": 0.42718, + "grad_norm": 1.6286677437364783, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 42718 + }, + { + "epoch": 0.42719, + "grad_norm": 1.4381753289095232, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 42719 + }, + { + "epoch": 0.4272, + "grad_norm": 1.503682478020474, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 42720 + }, + { + "epoch": 0.42721, + "grad_norm": 1.6606737426755256, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 42721 + }, + { + "epoch": 0.42722, + "grad_norm": 1.2524824797840186, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 42722 + }, + { + "epoch": 0.42723, + "grad_norm": 1.615254162355723, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 42723 + }, + { + "epoch": 0.42724, + "grad_norm": 1.2558811616512742, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 42724 + }, + { + "epoch": 0.42725, + "grad_norm": 1.5912454040744386, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 42725 + }, + { + "epoch": 0.42726, + "grad_norm": 1.622705594715044, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42726 + }, + { + "epoch": 0.42727, + "grad_norm": 1.5226785645955079, + "learning_rate": 0.003, + "loss": 4.036, + "step": 42727 + }, + { + "epoch": 0.42728, + "grad_norm": 1.6916323602949968, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42728 + }, + { + "epoch": 0.42729, + "grad_norm": 1.698084108559628, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 42729 + }, + { + "epoch": 0.4273, + "grad_norm": 1.289264862990976, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 42730 + }, + { + "epoch": 0.42731, + "grad_norm": 1.840025822462225, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 42731 + }, + { + "epoch": 0.42732, + "grad_norm": 1.4792446873066072, + "learning_rate": 0.003, + "loss": 4.065, + "step": 42732 + }, + { + "epoch": 0.42733, + "grad_norm": 1.4662320688577946, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 42733 + }, + { + "epoch": 0.42734, + "grad_norm": 1.6131841855858142, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 42734 + }, + { + "epoch": 0.42735, + "grad_norm": 1.270378950932129, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42735 + }, + { + "epoch": 0.42736, + "grad_norm": 1.7785423750196656, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 42736 + }, + { + "epoch": 0.42737, + "grad_norm": 1.3654585292466244, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 42737 + }, + { + "epoch": 0.42738, + "grad_norm": 1.9555710778849713, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 42738 + }, + { + "epoch": 0.42739, + "grad_norm": 1.6501464715909389, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42739 + }, + { + "epoch": 0.4274, + "grad_norm": 1.5665481500498912, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 42740 + }, + { + "epoch": 0.42741, + "grad_norm": 1.4534368978589203, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 42741 + }, + { + "epoch": 0.42742, + "grad_norm": 1.5410771808720134, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 42742 + }, + { + "epoch": 0.42743, + "grad_norm": 1.259923871253057, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 42743 + }, + { + "epoch": 0.42744, + "grad_norm": 1.8213907052347724, + "learning_rate": 0.003, + "loss": 4.068, + "step": 42744 + }, + { + "epoch": 0.42745, + "grad_norm": 1.4909172106479067, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 42745 + }, + { + "epoch": 0.42746, + "grad_norm": 1.7195272861052864, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 42746 + }, + { + "epoch": 0.42747, + "grad_norm": 1.3167185235004735, + "learning_rate": 0.003, + "loss": 4.041, + "step": 42747 + }, + { + "epoch": 0.42748, + "grad_norm": 1.8362454780546846, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 42748 + }, + { + "epoch": 0.42749, + "grad_norm": 1.2310434120646907, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 42749 + }, + { + "epoch": 0.4275, + "grad_norm": 1.772891068072832, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42750 + }, + { + "epoch": 0.42751, + "grad_norm": 1.3618381480073574, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 42751 + }, + { + "epoch": 0.42752, + "grad_norm": 1.6570406999537406, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 42752 + }, + { + "epoch": 0.42753, + "grad_norm": 1.5806917953177566, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 42753 + }, + { + "epoch": 0.42754, + "grad_norm": 1.3595230758910042, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42754 + }, + { + "epoch": 0.42755, + "grad_norm": 1.8860253826636852, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42755 + }, + { + "epoch": 0.42756, + "grad_norm": 1.3760112122257204, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 42756 + }, + { + "epoch": 0.42757, + "grad_norm": 1.4393570558043893, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 42757 + }, + { + "epoch": 0.42758, + "grad_norm": 1.6901821029012747, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 42758 + }, + { + "epoch": 0.42759, + "grad_norm": 1.709334033405679, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 42759 + }, + { + "epoch": 0.4276, + "grad_norm": 1.4165470222086871, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 42760 + }, + { + "epoch": 0.42761, + "grad_norm": 1.6155646184054042, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 42761 + }, + { + "epoch": 0.42762, + "grad_norm": 1.4414629026177572, + "learning_rate": 0.003, + "loss": 4.03, + "step": 42762 + }, + { + "epoch": 0.42763, + "grad_norm": 1.3990643044821884, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 42763 + }, + { + "epoch": 0.42764, + "grad_norm": 1.3617343359922554, + "learning_rate": 0.003, + "loss": 4.027, + "step": 42764 + }, + { + "epoch": 0.42765, + "grad_norm": 1.4711541501764056, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 42765 + }, + { + "epoch": 0.42766, + "grad_norm": 1.4113662277803105, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 42766 + }, + { + "epoch": 0.42767, + "grad_norm": 1.5967404258376894, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 42767 + }, + { + "epoch": 0.42768, + "grad_norm": 1.5170463186607255, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 42768 + }, + { + "epoch": 0.42769, + "grad_norm": 2.015331929897486, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42769 + }, + { + "epoch": 0.4277, + "grad_norm": 1.573339464854256, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42770 + }, + { + "epoch": 0.42771, + "grad_norm": 1.8488927764640493, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 42771 + }, + { + "epoch": 0.42772, + "grad_norm": 1.1523371885351508, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 42772 + }, + { + "epoch": 0.42773, + "grad_norm": 1.7002750864937612, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 42773 + }, + { + "epoch": 0.42774, + "grad_norm": 1.5126232877771357, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 42774 + }, + { + "epoch": 0.42775, + "grad_norm": 1.538236798609422, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 42775 + }, + { + "epoch": 0.42776, + "grad_norm": 1.694568521906015, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 42776 + }, + { + "epoch": 0.42777, + "grad_norm": 1.4226788368231562, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 42777 + }, + { + "epoch": 0.42778, + "grad_norm": 1.5314318219431895, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 42778 + }, + { + "epoch": 0.42779, + "grad_norm": 1.4818910198811692, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42779 + }, + { + "epoch": 0.4278, + "grad_norm": 1.543035915624846, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 42780 + }, + { + "epoch": 0.42781, + "grad_norm": 1.1969199180458525, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 42781 + }, + { + "epoch": 0.42782, + "grad_norm": 1.8529196707593862, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 42782 + }, + { + "epoch": 0.42783, + "grad_norm": 1.1876914892839665, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 42783 + }, + { + "epoch": 0.42784, + "grad_norm": 1.903693404766572, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 42784 + }, + { + "epoch": 0.42785, + "grad_norm": 1.2771560536540552, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 42785 + }, + { + "epoch": 0.42786, + "grad_norm": 1.7211310416477619, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 42786 + }, + { + "epoch": 0.42787, + "grad_norm": 1.2548595832349219, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 42787 + }, + { + "epoch": 0.42788, + "grad_norm": 1.5908473505376868, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 42788 + }, + { + "epoch": 0.42789, + "grad_norm": 1.6013782025330099, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 42789 + }, + { + "epoch": 0.4279, + "grad_norm": 1.3326225524106765, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42790 + }, + { + "epoch": 0.42791, + "grad_norm": 1.548021502098192, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 42791 + }, + { + "epoch": 0.42792, + "grad_norm": 1.5944915192015443, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 42792 + }, + { + "epoch": 0.42793, + "grad_norm": 1.6887343106324721, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 42793 + }, + { + "epoch": 0.42794, + "grad_norm": 1.4398745353495312, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 42794 + }, + { + "epoch": 0.42795, + "grad_norm": 1.5564292015586656, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 42795 + }, + { + "epoch": 0.42796, + "grad_norm": 1.553281817345925, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 42796 + }, + { + "epoch": 0.42797, + "grad_norm": 1.626730946037817, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 42797 + }, + { + "epoch": 0.42798, + "grad_norm": 1.2997014208219626, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 42798 + }, + { + "epoch": 0.42799, + "grad_norm": 1.3912797244350246, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 42799 + }, + { + "epoch": 0.428, + "grad_norm": 1.6276853147556272, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42800 + }, + { + "epoch": 0.42801, + "grad_norm": 1.6327542717562433, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42801 + }, + { + "epoch": 0.42802, + "grad_norm": 1.523431553997604, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 42802 + }, + { + "epoch": 0.42803, + "grad_norm": 1.554550208265149, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 42803 + }, + { + "epoch": 0.42804, + "grad_norm": 1.805000440152064, + "learning_rate": 0.003, + "loss": 4.044, + "step": 42804 + }, + { + "epoch": 0.42805, + "grad_norm": 1.4055877012338693, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 42805 + }, + { + "epoch": 0.42806, + "grad_norm": 1.5588780946691216, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 42806 + }, + { + "epoch": 0.42807, + "grad_norm": 1.911479598478813, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 42807 + }, + { + "epoch": 0.42808, + "grad_norm": 1.1886562034563737, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 42808 + }, + { + "epoch": 0.42809, + "grad_norm": 1.8272045330052717, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 42809 + }, + { + "epoch": 0.4281, + "grad_norm": 1.1800562452324002, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42810 + }, + { + "epoch": 0.42811, + "grad_norm": 1.8710842524549884, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 42811 + }, + { + "epoch": 0.42812, + "grad_norm": 1.382884073205837, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 42812 + }, + { + "epoch": 0.42813, + "grad_norm": 1.684526440332989, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 42813 + }, + { + "epoch": 0.42814, + "grad_norm": 1.813328884485701, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 42814 + }, + { + "epoch": 0.42815, + "grad_norm": 1.2345295998916221, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 42815 + }, + { + "epoch": 0.42816, + "grad_norm": 1.2469502848439022, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 42816 + }, + { + "epoch": 0.42817, + "grad_norm": 1.6331325502081153, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 42817 + }, + { + "epoch": 0.42818, + "grad_norm": 1.6364844677250916, + "learning_rate": 0.003, + "loss": 4.028, + "step": 42818 + }, + { + "epoch": 0.42819, + "grad_norm": 1.3736151186573526, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 42819 + }, + { + "epoch": 0.4282, + "grad_norm": 1.925736698186627, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 42820 + }, + { + "epoch": 0.42821, + "grad_norm": 1.1535703368556616, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42821 + }, + { + "epoch": 0.42822, + "grad_norm": 1.721666674956854, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 42822 + }, + { + "epoch": 0.42823, + "grad_norm": 1.22683836651265, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42823 + }, + { + "epoch": 0.42824, + "grad_norm": 1.8319438355417532, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 42824 + }, + { + "epoch": 0.42825, + "grad_norm": 1.70004108852703, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42825 + }, + { + "epoch": 0.42826, + "grad_norm": 1.2848623155389785, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 42826 + }, + { + "epoch": 0.42827, + "grad_norm": 1.6683827596478287, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 42827 + }, + { + "epoch": 0.42828, + "grad_norm": 1.424688253321067, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 42828 + }, + { + "epoch": 0.42829, + "grad_norm": 1.60059510168323, + "learning_rate": 0.003, + "loss": 4.054, + "step": 42829 + }, + { + "epoch": 0.4283, + "grad_norm": 1.4434057026894305, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 42830 + }, + { + "epoch": 0.42831, + "grad_norm": 1.497068909384767, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42831 + }, + { + "epoch": 0.42832, + "grad_norm": 1.4332941722413293, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42832 + }, + { + "epoch": 0.42833, + "grad_norm": 1.57330692534269, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 42833 + }, + { + "epoch": 0.42834, + "grad_norm": 1.5545655814777632, + "learning_rate": 0.003, + "loss": 4.063, + "step": 42834 + }, + { + "epoch": 0.42835, + "grad_norm": 1.505117120043619, + "learning_rate": 0.003, + "loss": 4.051, + "step": 42835 + }, + { + "epoch": 0.42836, + "grad_norm": 1.3239826912579609, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 42836 + }, + { + "epoch": 0.42837, + "grad_norm": 1.464803413974138, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 42837 + }, + { + "epoch": 0.42838, + "grad_norm": 1.3325263069897828, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 42838 + }, + { + "epoch": 0.42839, + "grad_norm": 1.5601885561788607, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42839 + }, + { + "epoch": 0.4284, + "grad_norm": 1.326351678631618, + "learning_rate": 0.003, + "loss": 4.013, + "step": 42840 + }, + { + "epoch": 0.42841, + "grad_norm": 1.5874802886365835, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 42841 + }, + { + "epoch": 0.42842, + "grad_norm": 1.6345187590547903, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 42842 + }, + { + "epoch": 0.42843, + "grad_norm": 1.440055993835501, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 42843 + }, + { + "epoch": 0.42844, + "grad_norm": 1.7422397162578678, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 42844 + }, + { + "epoch": 0.42845, + "grad_norm": 1.4643898196986145, + "learning_rate": 0.003, + "loss": 4.022, + "step": 42845 + }, + { + "epoch": 0.42846, + "grad_norm": 1.650851800765047, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 42846 + }, + { + "epoch": 0.42847, + "grad_norm": 1.4208831951888536, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 42847 + }, + { + "epoch": 0.42848, + "grad_norm": 1.5082380397510127, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 42848 + }, + { + "epoch": 0.42849, + "grad_norm": 1.34071311951666, + "learning_rate": 0.003, + "loss": 4.021, + "step": 42849 + }, + { + "epoch": 0.4285, + "grad_norm": 1.601530927404843, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 42850 + }, + { + "epoch": 0.42851, + "grad_norm": 1.3403211075607522, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 42851 + }, + { + "epoch": 0.42852, + "grad_norm": 1.7755194551109308, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 42852 + }, + { + "epoch": 0.42853, + "grad_norm": 1.5499345779885303, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 42853 + }, + { + "epoch": 0.42854, + "grad_norm": 2.0234956094972847, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 42854 + }, + { + "epoch": 0.42855, + "grad_norm": 1.2836128786948855, + "learning_rate": 0.003, + "loss": 4.047, + "step": 42855 + }, + { + "epoch": 0.42856, + "grad_norm": 1.424167712547683, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 42856 + }, + { + "epoch": 0.42857, + "grad_norm": 1.5545657415511498, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 42857 + }, + { + "epoch": 0.42858, + "grad_norm": 1.5864518978560707, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 42858 + }, + { + "epoch": 0.42859, + "grad_norm": 1.9265715682086115, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 42859 + }, + { + "epoch": 0.4286, + "grad_norm": 1.3635069804465492, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 42860 + }, + { + "epoch": 0.42861, + "grad_norm": 1.7945213358197125, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 42861 + }, + { + "epoch": 0.42862, + "grad_norm": 1.1915002084012523, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 42862 + }, + { + "epoch": 0.42863, + "grad_norm": 1.6714256426534693, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 42863 + }, + { + "epoch": 0.42864, + "grad_norm": 1.509474649571617, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 42864 + }, + { + "epoch": 0.42865, + "grad_norm": 1.7245140304676143, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 42865 + }, + { + "epoch": 0.42866, + "grad_norm": 1.506771525337719, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42866 + }, + { + "epoch": 0.42867, + "grad_norm": 1.4821726262263202, + "learning_rate": 0.003, + "loss": 4.032, + "step": 42867 + }, + { + "epoch": 0.42868, + "grad_norm": 1.584780183917517, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 42868 + }, + { + "epoch": 0.42869, + "grad_norm": 1.5461705169135898, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 42869 + }, + { + "epoch": 0.4287, + "grad_norm": 1.3990164043941606, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 42870 + }, + { + "epoch": 0.42871, + "grad_norm": 1.3304310315814762, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42871 + }, + { + "epoch": 0.42872, + "grad_norm": 1.5454938954938653, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 42872 + }, + { + "epoch": 0.42873, + "grad_norm": 1.2369051228132022, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 42873 + }, + { + "epoch": 0.42874, + "grad_norm": 1.831336245416106, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42874 + }, + { + "epoch": 0.42875, + "grad_norm": 1.2934246444442312, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 42875 + }, + { + "epoch": 0.42876, + "grad_norm": 1.5165755688639406, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 42876 + }, + { + "epoch": 0.42877, + "grad_norm": 1.545149505914604, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 42877 + }, + { + "epoch": 0.42878, + "grad_norm": 1.4251849153621947, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 42878 + }, + { + "epoch": 0.42879, + "grad_norm": 1.3466609994526006, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 42879 + }, + { + "epoch": 0.4288, + "grad_norm": 1.5587401229727036, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 42880 + }, + { + "epoch": 0.42881, + "grad_norm": 1.5271657443961113, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 42881 + }, + { + "epoch": 0.42882, + "grad_norm": 1.6775199335081348, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 42882 + }, + { + "epoch": 0.42883, + "grad_norm": 1.434112688701371, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 42883 + }, + { + "epoch": 0.42884, + "grad_norm": 1.3266209485254985, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 42884 + }, + { + "epoch": 0.42885, + "grad_norm": 1.7895744501968809, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 42885 + }, + { + "epoch": 0.42886, + "grad_norm": 1.3208329033659287, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 42886 + }, + { + "epoch": 0.42887, + "grad_norm": 1.751902265019173, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 42887 + }, + { + "epoch": 0.42888, + "grad_norm": 1.3967432040978727, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 42888 + }, + { + "epoch": 0.42889, + "grad_norm": 1.5450509917202988, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 42889 + }, + { + "epoch": 0.4289, + "grad_norm": 1.5491394973992925, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 42890 + }, + { + "epoch": 0.42891, + "grad_norm": 1.6173485237810257, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 42891 + }, + { + "epoch": 0.42892, + "grad_norm": 1.4280753427646662, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 42892 + }, + { + "epoch": 0.42893, + "grad_norm": 1.5986504854588837, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 42893 + }, + { + "epoch": 0.42894, + "grad_norm": 1.5963913799750808, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 42894 + }, + { + "epoch": 0.42895, + "grad_norm": 1.5754138587102877, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 42895 + }, + { + "epoch": 0.42896, + "grad_norm": 1.5284083897944485, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 42896 + }, + { + "epoch": 0.42897, + "grad_norm": 1.8687730442339732, + "learning_rate": 0.003, + "loss": 4.051, + "step": 42897 + }, + { + "epoch": 0.42898, + "grad_norm": 1.5456136076785054, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 42898 + }, + { + "epoch": 0.42899, + "grad_norm": 1.5118044346683874, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42899 + }, + { + "epoch": 0.429, + "grad_norm": 1.3863938037995733, + "learning_rate": 0.003, + "loss": 4.04, + "step": 42900 + }, + { + "epoch": 0.42901, + "grad_norm": 1.7370945369764597, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42901 + }, + { + "epoch": 0.42902, + "grad_norm": 1.2956044106842393, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 42902 + }, + { + "epoch": 0.42903, + "grad_norm": 1.4467962371992913, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 42903 + }, + { + "epoch": 0.42904, + "grad_norm": 1.3656629804249023, + "learning_rate": 0.003, + "loss": 4.041, + "step": 42904 + }, + { + "epoch": 0.42905, + "grad_norm": 1.7348389728084639, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 42905 + }, + { + "epoch": 0.42906, + "grad_norm": 1.4387618674432696, + "learning_rate": 0.003, + "loss": 4.038, + "step": 42906 + }, + { + "epoch": 0.42907, + "grad_norm": 1.6360057520856137, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 42907 + }, + { + "epoch": 0.42908, + "grad_norm": 1.8318754493700404, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 42908 + }, + { + "epoch": 0.42909, + "grad_norm": 1.2928222288027726, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 42909 + }, + { + "epoch": 0.4291, + "grad_norm": 1.6943982624345193, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 42910 + }, + { + "epoch": 0.42911, + "grad_norm": 1.2624018452531611, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 42911 + }, + { + "epoch": 0.42912, + "grad_norm": 1.803861751144249, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 42912 + }, + { + "epoch": 0.42913, + "grad_norm": 1.3640521901557447, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 42913 + }, + { + "epoch": 0.42914, + "grad_norm": 1.509716142613031, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 42914 + }, + { + "epoch": 0.42915, + "grad_norm": 1.3987951584258609, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 42915 + }, + { + "epoch": 0.42916, + "grad_norm": 1.3756586299665785, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 42916 + }, + { + "epoch": 0.42917, + "grad_norm": 1.7614455751803695, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 42917 + }, + { + "epoch": 0.42918, + "grad_norm": 1.3808269067285477, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 42918 + }, + { + "epoch": 0.42919, + "grad_norm": 1.6509086979107397, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 42919 + }, + { + "epoch": 0.4292, + "grad_norm": 1.4092819794998745, + "learning_rate": 0.003, + "loss": 4.025, + "step": 42920 + }, + { + "epoch": 0.42921, + "grad_norm": 1.9712125540175622, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 42921 + }, + { + "epoch": 0.42922, + "grad_norm": 1.7533629054197222, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 42922 + }, + { + "epoch": 0.42923, + "grad_norm": 1.4860529097020416, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 42923 + }, + { + "epoch": 0.42924, + "grad_norm": 1.4311678671828407, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 42924 + }, + { + "epoch": 0.42925, + "grad_norm": 1.5895520706968957, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 42925 + }, + { + "epoch": 0.42926, + "grad_norm": 1.4643785718708011, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 42926 + }, + { + "epoch": 0.42927, + "grad_norm": 1.5159270111299121, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 42927 + }, + { + "epoch": 0.42928, + "grad_norm": 1.8598012036472702, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 42928 + }, + { + "epoch": 0.42929, + "grad_norm": 1.086664782759617, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 42929 + }, + { + "epoch": 0.4293, + "grad_norm": 1.6309055239448182, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42930 + }, + { + "epoch": 0.42931, + "grad_norm": 1.8619620843262559, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 42931 + }, + { + "epoch": 0.42932, + "grad_norm": 1.4012885817128469, + "learning_rate": 0.003, + "loss": 4.031, + "step": 42932 + }, + { + "epoch": 0.42933, + "grad_norm": 1.8178684969459173, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 42933 + }, + { + "epoch": 0.42934, + "grad_norm": 1.171685134865146, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 42934 + }, + { + "epoch": 0.42935, + "grad_norm": 1.9154351144869315, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 42935 + }, + { + "epoch": 0.42936, + "grad_norm": 1.3964315380671513, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 42936 + }, + { + "epoch": 0.42937, + "grad_norm": 1.4948919338431004, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 42937 + }, + { + "epoch": 0.42938, + "grad_norm": 1.5878730313952218, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 42938 + }, + { + "epoch": 0.42939, + "grad_norm": 1.533326986036277, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 42939 + }, + { + "epoch": 0.4294, + "grad_norm": 1.6717479641073498, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 42940 + }, + { + "epoch": 0.42941, + "grad_norm": 1.3613225657356733, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42941 + }, + { + "epoch": 0.42942, + "grad_norm": 1.624155507975367, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 42942 + }, + { + "epoch": 0.42943, + "grad_norm": 1.4412064681987555, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 42943 + }, + { + "epoch": 0.42944, + "grad_norm": 1.6611355791450482, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 42944 + }, + { + "epoch": 0.42945, + "grad_norm": 1.3910405302672215, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 42945 + }, + { + "epoch": 0.42946, + "grad_norm": 1.5706350205445812, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 42946 + }, + { + "epoch": 0.42947, + "grad_norm": 1.2619168227166366, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 42947 + }, + { + "epoch": 0.42948, + "grad_norm": 1.5436929996874291, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 42948 + }, + { + "epoch": 0.42949, + "grad_norm": 1.398340420371895, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 42949 + }, + { + "epoch": 0.4295, + "grad_norm": 1.7816225909133836, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 42950 + }, + { + "epoch": 0.42951, + "grad_norm": 1.6574832268095316, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42951 + }, + { + "epoch": 0.42952, + "grad_norm": 1.6455683739100304, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42952 + }, + { + "epoch": 0.42953, + "grad_norm": 1.6289969787815035, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 42953 + }, + { + "epoch": 0.42954, + "grad_norm": 1.4157834586215576, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 42954 + }, + { + "epoch": 0.42955, + "grad_norm": 1.5690025373058731, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 42955 + }, + { + "epoch": 0.42956, + "grad_norm": 1.692869721344787, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 42956 + }, + { + "epoch": 0.42957, + "grad_norm": 1.2733003661977873, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 42957 + }, + { + "epoch": 0.42958, + "grad_norm": 1.8486545559391945, + "learning_rate": 0.003, + "loss": 4.073, + "step": 42958 + }, + { + "epoch": 0.42959, + "grad_norm": 1.4391812003772935, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 42959 + }, + { + "epoch": 0.4296, + "grad_norm": 1.5860312842545257, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 42960 + }, + { + "epoch": 0.42961, + "grad_norm": 1.2219570829257622, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 42961 + }, + { + "epoch": 0.42962, + "grad_norm": 1.6192309105324492, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 42962 + }, + { + "epoch": 0.42963, + "grad_norm": 1.7987648178284186, + "learning_rate": 0.003, + "loss": 4.053, + "step": 42963 + }, + { + "epoch": 0.42964, + "grad_norm": 1.724963195993634, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 42964 + }, + { + "epoch": 0.42965, + "grad_norm": 1.554262968434103, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 42965 + }, + { + "epoch": 0.42966, + "grad_norm": 1.7255875779240528, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 42966 + }, + { + "epoch": 0.42967, + "grad_norm": 1.5720595079031934, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 42967 + }, + { + "epoch": 0.42968, + "grad_norm": 1.4888893533480272, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 42968 + }, + { + "epoch": 0.42969, + "grad_norm": 1.5310450740077959, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 42969 + }, + { + "epoch": 0.4297, + "grad_norm": 1.752650615236425, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 42970 + }, + { + "epoch": 0.42971, + "grad_norm": 1.5221921533600937, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 42971 + }, + { + "epoch": 0.42972, + "grad_norm": 1.5112247734521274, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 42972 + }, + { + "epoch": 0.42973, + "grad_norm": 1.4045686331075442, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 42973 + }, + { + "epoch": 0.42974, + "grad_norm": 1.6536902796331665, + "learning_rate": 0.003, + "loss": 4.031, + "step": 42974 + }, + { + "epoch": 0.42975, + "grad_norm": 1.3735237849172488, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42975 + }, + { + "epoch": 0.42976, + "grad_norm": 1.2249485797071504, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 42976 + }, + { + "epoch": 0.42977, + "grad_norm": 1.9365932868248463, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 42977 + }, + { + "epoch": 0.42978, + "grad_norm": 1.502803565883409, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 42978 + }, + { + "epoch": 0.42979, + "grad_norm": 1.9592127077442039, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 42979 + }, + { + "epoch": 0.4298, + "grad_norm": 1.2160912602261125, + "learning_rate": 0.003, + "loss": 4.026, + "step": 42980 + }, + { + "epoch": 0.42981, + "grad_norm": 1.688472873462847, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 42981 + }, + { + "epoch": 0.42982, + "grad_norm": 1.5754838041736163, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 42982 + }, + { + "epoch": 0.42983, + "grad_norm": 1.584365583111445, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 42983 + }, + { + "epoch": 0.42984, + "grad_norm": 1.6228207268203958, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 42984 + }, + { + "epoch": 0.42985, + "grad_norm": 1.3690255218128873, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42985 + }, + { + "epoch": 0.42986, + "grad_norm": 1.549193723369444, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 42986 + }, + { + "epoch": 0.42987, + "grad_norm": 1.5429311715566982, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 42987 + }, + { + "epoch": 0.42988, + "grad_norm": 1.5271724528970363, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 42988 + }, + { + "epoch": 0.42989, + "grad_norm": 1.5097673291685652, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 42989 + }, + { + "epoch": 0.4299, + "grad_norm": 1.2807761185945, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 42990 + }, + { + "epoch": 0.42991, + "grad_norm": 1.425834759406081, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 42991 + }, + { + "epoch": 0.42992, + "grad_norm": 1.3872250716396672, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 42992 + }, + { + "epoch": 0.42993, + "grad_norm": 1.635282818505464, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 42993 + }, + { + "epoch": 0.42994, + "grad_norm": 1.89980966349495, + "learning_rate": 0.003, + "loss": 4.052, + "step": 42994 + }, + { + "epoch": 0.42995, + "grad_norm": 1.2341128609627183, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 42995 + }, + { + "epoch": 0.42996, + "grad_norm": 1.6649151283028372, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 42996 + }, + { + "epoch": 0.42997, + "grad_norm": 1.341366829965389, + "learning_rate": 0.003, + "loss": 4.053, + "step": 42997 + }, + { + "epoch": 0.42998, + "grad_norm": 1.74474640913941, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 42998 + }, + { + "epoch": 0.42999, + "grad_norm": 1.4069207739380565, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 42999 + }, + { + "epoch": 0.43, + "grad_norm": 1.570959549570409, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43000 + }, + { + "epoch": 0.43001, + "grad_norm": 1.5204994050519587, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 43001 + }, + { + "epoch": 0.43002, + "grad_norm": 1.5740667897228604, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 43002 + }, + { + "epoch": 0.43003, + "grad_norm": 1.3549466374510313, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 43003 + }, + { + "epoch": 0.43004, + "grad_norm": 1.616531283404618, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43004 + }, + { + "epoch": 0.43005, + "grad_norm": 1.3391479577011276, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 43005 + }, + { + "epoch": 0.43006, + "grad_norm": 1.7360460444812078, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 43006 + }, + { + "epoch": 0.43007, + "grad_norm": 1.4436768033470857, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43007 + }, + { + "epoch": 0.43008, + "grad_norm": 1.9888776677008202, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 43008 + }, + { + "epoch": 0.43009, + "grad_norm": 1.2140108395199187, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 43009 + }, + { + "epoch": 0.4301, + "grad_norm": 1.4968069387216665, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 43010 + }, + { + "epoch": 0.43011, + "grad_norm": 1.462049777124018, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 43011 + }, + { + "epoch": 0.43012, + "grad_norm": 1.4558439796382616, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 43012 + }, + { + "epoch": 0.43013, + "grad_norm": 1.891766275887408, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 43013 + }, + { + "epoch": 0.43014, + "grad_norm": 1.5709540465128435, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 43014 + }, + { + "epoch": 0.43015, + "grad_norm": 1.2815520190636291, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 43015 + }, + { + "epoch": 0.43016, + "grad_norm": 1.5265600661254446, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 43016 + }, + { + "epoch": 0.43017, + "grad_norm": 1.5349396198405543, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43017 + }, + { + "epoch": 0.43018, + "grad_norm": 1.509813483841224, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 43018 + }, + { + "epoch": 0.43019, + "grad_norm": 1.7180814632675083, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 43019 + }, + { + "epoch": 0.4302, + "grad_norm": 1.5125324553749007, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 43020 + }, + { + "epoch": 0.43021, + "grad_norm": 1.7201595209316043, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 43021 + }, + { + "epoch": 0.43022, + "grad_norm": 1.7297664016265517, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43022 + }, + { + "epoch": 0.43023, + "grad_norm": 1.2655425534324287, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 43023 + }, + { + "epoch": 0.43024, + "grad_norm": 1.5550525887035465, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43024 + }, + { + "epoch": 0.43025, + "grad_norm": 1.577875390283871, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 43025 + }, + { + "epoch": 0.43026, + "grad_norm": 1.48773635479244, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 43026 + }, + { + "epoch": 0.43027, + "grad_norm": 1.583704361489017, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43027 + }, + { + "epoch": 0.43028, + "grad_norm": 1.3830946605303742, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43028 + }, + { + "epoch": 0.43029, + "grad_norm": 1.7177512201488652, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 43029 + }, + { + "epoch": 0.4303, + "grad_norm": 1.5570518080958164, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 43030 + }, + { + "epoch": 0.43031, + "grad_norm": 1.6199881198929538, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 43031 + }, + { + "epoch": 0.43032, + "grad_norm": 1.6527121516401946, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 43032 + }, + { + "epoch": 0.43033, + "grad_norm": 1.8065190365222445, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 43033 + }, + { + "epoch": 0.43034, + "grad_norm": 1.4582025762509387, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 43034 + }, + { + "epoch": 0.43035, + "grad_norm": 1.7730071681757409, + "learning_rate": 0.003, + "loss": 4.076, + "step": 43035 + }, + { + "epoch": 0.43036, + "grad_norm": 1.7265236009495448, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 43036 + }, + { + "epoch": 0.43037, + "grad_norm": 1.3026713283109261, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 43037 + }, + { + "epoch": 0.43038, + "grad_norm": 1.6484578466654622, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 43038 + }, + { + "epoch": 0.43039, + "grad_norm": 1.404082168618179, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 43039 + }, + { + "epoch": 0.4304, + "grad_norm": 1.6463632243041222, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 43040 + }, + { + "epoch": 0.43041, + "grad_norm": 1.4407112822231536, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 43041 + }, + { + "epoch": 0.43042, + "grad_norm": 1.5308922791697592, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 43042 + }, + { + "epoch": 0.43043, + "grad_norm": 1.9135303446083336, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 43043 + }, + { + "epoch": 0.43044, + "grad_norm": 1.1218119427768691, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 43044 + }, + { + "epoch": 0.43045, + "grad_norm": 2.0147673032801388, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 43045 + }, + { + "epoch": 0.43046, + "grad_norm": 1.43234005891958, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 43046 + }, + { + "epoch": 0.43047, + "grad_norm": 1.4019217538524922, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43047 + }, + { + "epoch": 0.43048, + "grad_norm": 1.6870135792336995, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43048 + }, + { + "epoch": 0.43049, + "grad_norm": 1.361295123782907, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 43049 + }, + { + "epoch": 0.4305, + "grad_norm": 1.6201604962718545, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 43050 + }, + { + "epoch": 0.43051, + "grad_norm": 1.5013295146984693, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43051 + }, + { + "epoch": 0.43052, + "grad_norm": 1.5149537120843768, + "learning_rate": 0.003, + "loss": 4.006, + "step": 43052 + }, + { + "epoch": 0.43053, + "grad_norm": 1.6580058958703598, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 43053 + }, + { + "epoch": 0.43054, + "grad_norm": 1.4724971743088315, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43054 + }, + { + "epoch": 0.43055, + "grad_norm": 1.4170776084041983, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 43055 + }, + { + "epoch": 0.43056, + "grad_norm": 1.6743643289925105, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 43056 + }, + { + "epoch": 0.43057, + "grad_norm": 1.4545392474698937, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 43057 + }, + { + "epoch": 0.43058, + "grad_norm": 1.5869667904901361, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 43058 + }, + { + "epoch": 0.43059, + "grad_norm": 1.5239379111978513, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 43059 + }, + { + "epoch": 0.4306, + "grad_norm": 1.6865962607016662, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 43060 + }, + { + "epoch": 0.43061, + "grad_norm": 1.191495640858537, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 43061 + }, + { + "epoch": 0.43062, + "grad_norm": 1.6254902404408762, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43062 + }, + { + "epoch": 0.43063, + "grad_norm": 1.5633477225836454, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 43063 + }, + { + "epoch": 0.43064, + "grad_norm": 1.430786497497595, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 43064 + }, + { + "epoch": 0.43065, + "grad_norm": 1.9547557224563594, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 43065 + }, + { + "epoch": 0.43066, + "grad_norm": 1.6192835428035703, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43066 + }, + { + "epoch": 0.43067, + "grad_norm": 1.4262794777751568, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 43067 + }, + { + "epoch": 0.43068, + "grad_norm": 1.4378406515968203, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 43068 + }, + { + "epoch": 0.43069, + "grad_norm": 1.7518878519396215, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 43069 + }, + { + "epoch": 0.4307, + "grad_norm": 1.247643915344837, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 43070 + }, + { + "epoch": 0.43071, + "grad_norm": 1.9218494554415926, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 43071 + }, + { + "epoch": 0.43072, + "grad_norm": 1.1612638332554879, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 43072 + }, + { + "epoch": 0.43073, + "grad_norm": 1.2997478471049564, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 43073 + }, + { + "epoch": 0.43074, + "grad_norm": 2.0497839058357963, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 43074 + }, + { + "epoch": 0.43075, + "grad_norm": 1.5116898027453318, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43075 + }, + { + "epoch": 0.43076, + "grad_norm": 1.4142517529423617, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 43076 + }, + { + "epoch": 0.43077, + "grad_norm": 1.4441967448208415, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 43077 + }, + { + "epoch": 0.43078, + "grad_norm": 1.70934810906289, + "learning_rate": 0.003, + "loss": 4.038, + "step": 43078 + }, + { + "epoch": 0.43079, + "grad_norm": 1.412892093780238, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 43079 + }, + { + "epoch": 0.4308, + "grad_norm": 1.7101976643770265, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43080 + }, + { + "epoch": 0.43081, + "grad_norm": 1.3176906269685873, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43081 + }, + { + "epoch": 0.43082, + "grad_norm": 1.740694341731293, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 43082 + }, + { + "epoch": 0.43083, + "grad_norm": 1.2063659051796416, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 43083 + }, + { + "epoch": 0.43084, + "grad_norm": 1.5768142952863893, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 43084 + }, + { + "epoch": 0.43085, + "grad_norm": 1.2627438204147565, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 43085 + }, + { + "epoch": 0.43086, + "grad_norm": 1.593097625436047, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 43086 + }, + { + "epoch": 0.43087, + "grad_norm": 1.8466206755209107, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 43087 + }, + { + "epoch": 0.43088, + "grad_norm": 1.7507042613607093, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 43088 + }, + { + "epoch": 0.43089, + "grad_norm": 1.244784561538406, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 43089 + }, + { + "epoch": 0.4309, + "grad_norm": 1.576302379282591, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 43090 + }, + { + "epoch": 0.43091, + "grad_norm": 1.9266687422251658, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 43091 + }, + { + "epoch": 0.43092, + "grad_norm": 1.1144745231743824, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 43092 + }, + { + "epoch": 0.43093, + "grad_norm": 1.7355392092571944, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 43093 + }, + { + "epoch": 0.43094, + "grad_norm": 1.4462059435283408, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 43094 + }, + { + "epoch": 0.43095, + "grad_norm": 1.493110826832055, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 43095 + }, + { + "epoch": 0.43096, + "grad_norm": 1.4493032685794038, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 43096 + }, + { + "epoch": 0.43097, + "grad_norm": 1.4747576971643337, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 43097 + }, + { + "epoch": 0.43098, + "grad_norm": 1.3158765350862436, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 43098 + }, + { + "epoch": 0.43099, + "grad_norm": 1.7904354999141916, + "learning_rate": 0.003, + "loss": 4.082, + "step": 43099 + }, + { + "epoch": 0.431, + "grad_norm": 1.6273759463457844, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43100 + }, + { + "epoch": 0.43101, + "grad_norm": 1.4055581798018104, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 43101 + }, + { + "epoch": 0.43102, + "grad_norm": 1.725021190761588, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 43102 + }, + { + "epoch": 0.43103, + "grad_norm": 1.5269565525228492, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 43103 + }, + { + "epoch": 0.43104, + "grad_norm": 1.1992303704545486, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 43104 + }, + { + "epoch": 0.43105, + "grad_norm": 1.9693645017337267, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 43105 + }, + { + "epoch": 0.43106, + "grad_norm": 1.3117801451328148, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 43106 + }, + { + "epoch": 0.43107, + "grad_norm": 1.6832646972246141, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43107 + }, + { + "epoch": 0.43108, + "grad_norm": 1.4061562900966182, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 43108 + }, + { + "epoch": 0.43109, + "grad_norm": 1.6378997065258423, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43109 + }, + { + "epoch": 0.4311, + "grad_norm": 1.8200697182686363, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 43110 + }, + { + "epoch": 0.43111, + "grad_norm": 1.3296176734499936, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43111 + }, + { + "epoch": 0.43112, + "grad_norm": 1.408875457920476, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43112 + }, + { + "epoch": 0.43113, + "grad_norm": 1.4577488578147095, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 43113 + }, + { + "epoch": 0.43114, + "grad_norm": 1.4853820744457276, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 43114 + }, + { + "epoch": 0.43115, + "grad_norm": 1.336043308082033, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 43115 + }, + { + "epoch": 0.43116, + "grad_norm": 1.719505071940957, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 43116 + }, + { + "epoch": 0.43117, + "grad_norm": 1.5966243972642606, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43117 + }, + { + "epoch": 0.43118, + "grad_norm": 1.6507708765505307, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 43118 + }, + { + "epoch": 0.43119, + "grad_norm": 1.4278597715726198, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 43119 + }, + { + "epoch": 0.4312, + "grad_norm": 1.6562367557330329, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 43120 + }, + { + "epoch": 0.43121, + "grad_norm": 1.3384570918530836, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43121 + }, + { + "epoch": 0.43122, + "grad_norm": 1.6902334893037048, + "learning_rate": 0.003, + "loss": 4.067, + "step": 43122 + }, + { + "epoch": 0.43123, + "grad_norm": 1.7460494827095012, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 43123 + }, + { + "epoch": 0.43124, + "grad_norm": 1.5342384075081956, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 43124 + }, + { + "epoch": 0.43125, + "grad_norm": 1.9520244333248955, + "learning_rate": 0.003, + "loss": 4.024, + "step": 43125 + }, + { + "epoch": 0.43126, + "grad_norm": 1.3444315754463856, + "learning_rate": 0.003, + "loss": 4.043, + "step": 43126 + }, + { + "epoch": 0.43127, + "grad_norm": 1.7824212001481659, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 43127 + }, + { + "epoch": 0.43128, + "grad_norm": 1.248659576465008, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 43128 + }, + { + "epoch": 0.43129, + "grad_norm": 1.6417267357146168, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 43129 + }, + { + "epoch": 0.4313, + "grad_norm": 1.323719632755784, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 43130 + }, + { + "epoch": 0.43131, + "grad_norm": 1.4324003810196435, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 43131 + }, + { + "epoch": 0.43132, + "grad_norm": 1.588031375444644, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43132 + }, + { + "epoch": 0.43133, + "grad_norm": 1.569923579829526, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 43133 + }, + { + "epoch": 0.43134, + "grad_norm": 1.4828791910115937, + "learning_rate": 0.003, + "loss": 4.063, + "step": 43134 + }, + { + "epoch": 0.43135, + "grad_norm": 1.5430272649693746, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 43135 + }, + { + "epoch": 0.43136, + "grad_norm": 1.3916850296783518, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 43136 + }, + { + "epoch": 0.43137, + "grad_norm": 1.9113044081049104, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 43137 + }, + { + "epoch": 0.43138, + "grad_norm": 1.2897024948109468, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43138 + }, + { + "epoch": 0.43139, + "grad_norm": 1.8809740439686258, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 43139 + }, + { + "epoch": 0.4314, + "grad_norm": 1.625107956469322, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 43140 + }, + { + "epoch": 0.43141, + "grad_norm": 1.4089380553326676, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 43141 + }, + { + "epoch": 0.43142, + "grad_norm": 1.4864414134051163, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 43142 + }, + { + "epoch": 0.43143, + "grad_norm": 1.4564226639863367, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 43143 + }, + { + "epoch": 0.43144, + "grad_norm": 1.2970352264359832, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43144 + }, + { + "epoch": 0.43145, + "grad_norm": 1.4623645101063576, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 43145 + }, + { + "epoch": 0.43146, + "grad_norm": 1.513559225038742, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 43146 + }, + { + "epoch": 0.43147, + "grad_norm": 1.4635774744086814, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 43147 + }, + { + "epoch": 0.43148, + "grad_norm": 1.4823509175412946, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43148 + }, + { + "epoch": 0.43149, + "grad_norm": 1.8339547371203548, + "learning_rate": 0.003, + "loss": 4.062, + "step": 43149 + }, + { + "epoch": 0.4315, + "grad_norm": 1.5922082620848528, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 43150 + }, + { + "epoch": 0.43151, + "grad_norm": 1.5832118482371536, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 43151 + }, + { + "epoch": 0.43152, + "grad_norm": 1.2839405495347451, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 43152 + }, + { + "epoch": 0.43153, + "grad_norm": 1.7839854877306618, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43153 + }, + { + "epoch": 0.43154, + "grad_norm": 1.3649753640448696, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 43154 + }, + { + "epoch": 0.43155, + "grad_norm": 1.815548055358805, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43155 + }, + { + "epoch": 0.43156, + "grad_norm": 1.6307481606905478, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43156 + }, + { + "epoch": 0.43157, + "grad_norm": 1.4964033590698453, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43157 + }, + { + "epoch": 0.43158, + "grad_norm": 1.3247083246836036, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 43158 + }, + { + "epoch": 0.43159, + "grad_norm": 1.7591603533659534, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 43159 + }, + { + "epoch": 0.4316, + "grad_norm": 1.4659843760534643, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 43160 + }, + { + "epoch": 0.43161, + "grad_norm": 1.5446950358738771, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43161 + }, + { + "epoch": 0.43162, + "grad_norm": 1.4652133874408617, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 43162 + }, + { + "epoch": 0.43163, + "grad_norm": 1.4459518809182514, + "learning_rate": 0.003, + "loss": 4.049, + "step": 43163 + }, + { + "epoch": 0.43164, + "grad_norm": 1.6974957573576435, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 43164 + }, + { + "epoch": 0.43165, + "grad_norm": 1.2003684202972398, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 43165 + }, + { + "epoch": 0.43166, + "grad_norm": 1.9457096039716837, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 43166 + }, + { + "epoch": 0.43167, + "grad_norm": 1.2945929904685984, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43167 + }, + { + "epoch": 0.43168, + "grad_norm": 1.8994329625781727, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 43168 + }, + { + "epoch": 0.43169, + "grad_norm": 1.4779572712547802, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 43169 + }, + { + "epoch": 0.4317, + "grad_norm": 1.7965650501490988, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43170 + }, + { + "epoch": 0.43171, + "grad_norm": 1.7169213660943712, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43171 + }, + { + "epoch": 0.43172, + "grad_norm": 1.3438961957560835, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43172 + }, + { + "epoch": 0.43173, + "grad_norm": 1.6980539305172413, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43173 + }, + { + "epoch": 0.43174, + "grad_norm": 1.6644639892214048, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 43174 + }, + { + "epoch": 0.43175, + "grad_norm": 1.5405132935281154, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 43175 + }, + { + "epoch": 0.43176, + "grad_norm": 1.5554329904751074, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 43176 + }, + { + "epoch": 0.43177, + "grad_norm": 1.8012802584934178, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 43177 + }, + { + "epoch": 0.43178, + "grad_norm": 1.4483933215348168, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 43178 + }, + { + "epoch": 0.43179, + "grad_norm": 1.7563391250371534, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43179 + }, + { + "epoch": 0.4318, + "grad_norm": 1.1470692844792114, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43180 + }, + { + "epoch": 0.43181, + "grad_norm": 1.5835369810044067, + "learning_rate": 0.003, + "loss": 4.054, + "step": 43181 + }, + { + "epoch": 0.43182, + "grad_norm": 1.3221645665966755, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 43182 + }, + { + "epoch": 0.43183, + "grad_norm": 1.6989131012209542, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 43183 + }, + { + "epoch": 0.43184, + "grad_norm": 1.4929056378407655, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 43184 + }, + { + "epoch": 0.43185, + "grad_norm": 1.6489255191694245, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43185 + }, + { + "epoch": 0.43186, + "grad_norm": 1.6091888263855523, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43186 + }, + { + "epoch": 0.43187, + "grad_norm": 1.5412143516791372, + "learning_rate": 0.003, + "loss": 4.02, + "step": 43187 + }, + { + "epoch": 0.43188, + "grad_norm": 1.6462354774828423, + "learning_rate": 0.003, + "loss": 4.006, + "step": 43188 + }, + { + "epoch": 0.43189, + "grad_norm": 1.3379550519787748, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 43189 + }, + { + "epoch": 0.4319, + "grad_norm": 1.7231753146408657, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 43190 + }, + { + "epoch": 0.43191, + "grad_norm": 1.423821247196092, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 43191 + }, + { + "epoch": 0.43192, + "grad_norm": 1.5834300662467033, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43192 + }, + { + "epoch": 0.43193, + "grad_norm": 1.3502209153805902, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 43193 + }, + { + "epoch": 0.43194, + "grad_norm": 1.8440289621341779, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 43194 + }, + { + "epoch": 0.43195, + "grad_norm": 1.6287308892585448, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43195 + }, + { + "epoch": 0.43196, + "grad_norm": 1.4252946378789761, + "learning_rate": 0.003, + "loss": 4.038, + "step": 43196 + }, + { + "epoch": 0.43197, + "grad_norm": 1.6200289611079912, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 43197 + }, + { + "epoch": 0.43198, + "grad_norm": 1.42110099801592, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43198 + }, + { + "epoch": 0.43199, + "grad_norm": 1.490474224016913, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43199 + }, + { + "epoch": 0.432, + "grad_norm": 1.3252949803473542, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43200 + }, + { + "epoch": 0.43201, + "grad_norm": 1.5003845646263587, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 43201 + }, + { + "epoch": 0.43202, + "grad_norm": 1.3551742641912403, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 43202 + }, + { + "epoch": 0.43203, + "grad_norm": 1.6177500297243494, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 43203 + }, + { + "epoch": 0.43204, + "grad_norm": 1.6539542778774585, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43204 + }, + { + "epoch": 0.43205, + "grad_norm": 1.5210465244107298, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43205 + }, + { + "epoch": 0.43206, + "grad_norm": 1.6797119632447497, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 43206 + }, + { + "epoch": 0.43207, + "grad_norm": 1.4527957704556689, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 43207 + }, + { + "epoch": 0.43208, + "grad_norm": 1.5620990419186764, + "learning_rate": 0.003, + "loss": 4.026, + "step": 43208 + }, + { + "epoch": 0.43209, + "grad_norm": 1.4251595336016833, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43209 + }, + { + "epoch": 0.4321, + "grad_norm": 1.5186738939868492, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43210 + }, + { + "epoch": 0.43211, + "grad_norm": 1.8596877193028025, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 43211 + }, + { + "epoch": 0.43212, + "grad_norm": 1.2791364230539901, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 43212 + }, + { + "epoch": 0.43213, + "grad_norm": 1.7565430194699936, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43213 + }, + { + "epoch": 0.43214, + "grad_norm": 1.2732018968799834, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43214 + }, + { + "epoch": 0.43215, + "grad_norm": 1.9173980738804512, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 43215 + }, + { + "epoch": 0.43216, + "grad_norm": 1.6615096915762044, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 43216 + }, + { + "epoch": 0.43217, + "grad_norm": 1.4698718446764862, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 43217 + }, + { + "epoch": 0.43218, + "grad_norm": 1.8020031662765386, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43218 + }, + { + "epoch": 0.43219, + "grad_norm": 1.3429941482430516, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 43219 + }, + { + "epoch": 0.4322, + "grad_norm": 1.4062880189257043, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 43220 + }, + { + "epoch": 0.43221, + "grad_norm": 1.6650676865090954, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 43221 + }, + { + "epoch": 0.43222, + "grad_norm": 1.6990046837407946, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43222 + }, + { + "epoch": 0.43223, + "grad_norm": 1.1501529341076981, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43223 + }, + { + "epoch": 0.43224, + "grad_norm": 1.62441207076946, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 43224 + }, + { + "epoch": 0.43225, + "grad_norm": 1.4981348619797732, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 43225 + }, + { + "epoch": 0.43226, + "grad_norm": 1.4056939685228285, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 43226 + }, + { + "epoch": 0.43227, + "grad_norm": 1.5487429790829834, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 43227 + }, + { + "epoch": 0.43228, + "grad_norm": 1.4432361995406737, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43228 + }, + { + "epoch": 0.43229, + "grad_norm": 1.9296067834649766, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43229 + }, + { + "epoch": 0.4323, + "grad_norm": 1.3286871440206636, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 43230 + }, + { + "epoch": 0.43231, + "grad_norm": 1.7701924232519786, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 43231 + }, + { + "epoch": 0.43232, + "grad_norm": 1.4700432636126526, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43232 + }, + { + "epoch": 0.43233, + "grad_norm": 1.511661773021028, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 43233 + }, + { + "epoch": 0.43234, + "grad_norm": 1.716888384953085, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 43234 + }, + { + "epoch": 0.43235, + "grad_norm": 1.6144163048273021, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 43235 + }, + { + "epoch": 0.43236, + "grad_norm": 1.4118172450754862, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43236 + }, + { + "epoch": 0.43237, + "grad_norm": 1.8495373419120276, + "learning_rate": 0.003, + "loss": 4.048, + "step": 43237 + }, + { + "epoch": 0.43238, + "grad_norm": 1.2692247860950934, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 43238 + }, + { + "epoch": 0.43239, + "grad_norm": 1.7627804453627884, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 43239 + }, + { + "epoch": 0.4324, + "grad_norm": 1.3643082637494715, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 43240 + }, + { + "epoch": 0.43241, + "grad_norm": 1.5725925242424392, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 43241 + }, + { + "epoch": 0.43242, + "grad_norm": 1.5197910731188011, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 43242 + }, + { + "epoch": 0.43243, + "grad_norm": 1.6449561713420673, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 43243 + }, + { + "epoch": 0.43244, + "grad_norm": 1.463366315945229, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 43244 + }, + { + "epoch": 0.43245, + "grad_norm": 1.6014449732940137, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 43245 + }, + { + "epoch": 0.43246, + "grad_norm": 1.4943863648150573, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 43246 + }, + { + "epoch": 0.43247, + "grad_norm": 1.4928054968872264, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 43247 + }, + { + "epoch": 0.43248, + "grad_norm": 1.6902504207540174, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 43248 + }, + { + "epoch": 0.43249, + "grad_norm": 1.2659547663759234, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 43249 + }, + { + "epoch": 0.4325, + "grad_norm": 1.6528208897420118, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 43250 + }, + { + "epoch": 0.43251, + "grad_norm": 1.467224199479159, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 43251 + }, + { + "epoch": 0.43252, + "grad_norm": 1.7183364795482698, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 43252 + }, + { + "epoch": 0.43253, + "grad_norm": 1.4974217390442104, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43253 + }, + { + "epoch": 0.43254, + "grad_norm": 1.5921383187229825, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 43254 + }, + { + "epoch": 0.43255, + "grad_norm": 1.3466080153986342, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 43255 + }, + { + "epoch": 0.43256, + "grad_norm": 1.5582775493251344, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43256 + }, + { + "epoch": 0.43257, + "grad_norm": 1.2885946524111938, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43257 + }, + { + "epoch": 0.43258, + "grad_norm": 1.9862998373708451, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 43258 + }, + { + "epoch": 0.43259, + "grad_norm": 1.3646595712406375, + "learning_rate": 0.003, + "loss": 4.023, + "step": 43259 + }, + { + "epoch": 0.4326, + "grad_norm": 1.7232118436219543, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 43260 + }, + { + "epoch": 0.43261, + "grad_norm": 1.6325671862997129, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43261 + }, + { + "epoch": 0.43262, + "grad_norm": 1.4257599700110217, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 43262 + }, + { + "epoch": 0.43263, + "grad_norm": 1.933831496132254, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 43263 + }, + { + "epoch": 0.43264, + "grad_norm": 1.4132369030043814, + "learning_rate": 0.003, + "loss": 4.054, + "step": 43264 + }, + { + "epoch": 0.43265, + "grad_norm": 1.613159455495719, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 43265 + }, + { + "epoch": 0.43266, + "grad_norm": 1.664549956195813, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 43266 + }, + { + "epoch": 0.43267, + "grad_norm": 1.7407704572547333, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43267 + }, + { + "epoch": 0.43268, + "grad_norm": 1.3619768358787103, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 43268 + }, + { + "epoch": 0.43269, + "grad_norm": 1.7031411988400151, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 43269 + }, + { + "epoch": 0.4327, + "grad_norm": 1.3518105224332815, + "learning_rate": 0.003, + "loss": 4.041, + "step": 43270 + }, + { + "epoch": 0.43271, + "grad_norm": 1.7081584184978, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43271 + }, + { + "epoch": 0.43272, + "grad_norm": 1.4793110527161464, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 43272 + }, + { + "epoch": 0.43273, + "grad_norm": 1.5475847484740712, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 43273 + }, + { + "epoch": 0.43274, + "grad_norm": 1.5123010979557494, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 43274 + }, + { + "epoch": 0.43275, + "grad_norm": 1.5357817217368253, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 43275 + }, + { + "epoch": 0.43276, + "grad_norm": 1.6906642395345524, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 43276 + }, + { + "epoch": 0.43277, + "grad_norm": 1.3545094745769592, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43277 + }, + { + "epoch": 0.43278, + "grad_norm": 1.8798952884688291, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 43278 + }, + { + "epoch": 0.43279, + "grad_norm": 1.2022455971546615, + "learning_rate": 0.003, + "loss": 4.02, + "step": 43279 + }, + { + "epoch": 0.4328, + "grad_norm": 1.6283497285410444, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 43280 + }, + { + "epoch": 0.43281, + "grad_norm": 1.6143604856845586, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 43281 + }, + { + "epoch": 0.43282, + "grad_norm": 1.7707652782483723, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 43282 + }, + { + "epoch": 0.43283, + "grad_norm": 1.5641839886427922, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 43283 + }, + { + "epoch": 0.43284, + "grad_norm": 1.4455876397129415, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43284 + }, + { + "epoch": 0.43285, + "grad_norm": 1.657480852523109, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 43285 + }, + { + "epoch": 0.43286, + "grad_norm": 1.3281998033228337, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 43286 + }, + { + "epoch": 0.43287, + "grad_norm": 1.6835436948030034, + "learning_rate": 0.003, + "loss": 4.066, + "step": 43287 + }, + { + "epoch": 0.43288, + "grad_norm": 1.3620445915196624, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 43288 + }, + { + "epoch": 0.43289, + "grad_norm": 1.8121050024533183, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 43289 + }, + { + "epoch": 0.4329, + "grad_norm": 1.434784942359001, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 43290 + }, + { + "epoch": 0.43291, + "grad_norm": 1.7228785190388511, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 43291 + }, + { + "epoch": 0.43292, + "grad_norm": 1.5778899234517665, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 43292 + }, + { + "epoch": 0.43293, + "grad_norm": 1.7724976236110481, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 43293 + }, + { + "epoch": 0.43294, + "grad_norm": 1.4193469517976989, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 43294 + }, + { + "epoch": 0.43295, + "grad_norm": 1.6421363092720218, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 43295 + }, + { + "epoch": 0.43296, + "grad_norm": 1.323887552777388, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43296 + }, + { + "epoch": 0.43297, + "grad_norm": 1.8467433552399577, + "learning_rate": 0.003, + "loss": 4.064, + "step": 43297 + }, + { + "epoch": 0.43298, + "grad_norm": 1.217895044334679, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 43298 + }, + { + "epoch": 0.43299, + "grad_norm": 1.648838212781819, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43299 + }, + { + "epoch": 0.433, + "grad_norm": 1.7264024080284384, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 43300 + }, + { + "epoch": 0.43301, + "grad_norm": 1.4560892000926897, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43301 + }, + { + "epoch": 0.43302, + "grad_norm": 1.9546157621819307, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 43302 + }, + { + "epoch": 0.43303, + "grad_norm": 1.18472375425437, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43303 + }, + { + "epoch": 0.43304, + "grad_norm": 1.2947712892141625, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43304 + }, + { + "epoch": 0.43305, + "grad_norm": 1.7132970900456699, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 43305 + }, + { + "epoch": 0.43306, + "grad_norm": 1.5992509517678044, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 43306 + }, + { + "epoch": 0.43307, + "grad_norm": 1.3130156751520832, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 43307 + }, + { + "epoch": 0.43308, + "grad_norm": 1.939458453861989, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43308 + }, + { + "epoch": 0.43309, + "grad_norm": 1.3470388417544474, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 43309 + }, + { + "epoch": 0.4331, + "grad_norm": 1.4087465169482793, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 43310 + }, + { + "epoch": 0.43311, + "grad_norm": 1.782910434729818, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 43311 + }, + { + "epoch": 0.43312, + "grad_norm": 1.4658960727917092, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 43312 + }, + { + "epoch": 0.43313, + "grad_norm": 1.4933481003275102, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 43313 + }, + { + "epoch": 0.43314, + "grad_norm": 1.6470981506772293, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 43314 + }, + { + "epoch": 0.43315, + "grad_norm": 1.356625489865832, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 43315 + }, + { + "epoch": 0.43316, + "grad_norm": 1.6476328467633494, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43316 + }, + { + "epoch": 0.43317, + "grad_norm": 1.3778699160644674, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43317 + }, + { + "epoch": 0.43318, + "grad_norm": 1.6609146874943628, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43318 + }, + { + "epoch": 0.43319, + "grad_norm": 1.5185116785582722, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 43319 + }, + { + "epoch": 0.4332, + "grad_norm": 1.6348107856419791, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 43320 + }, + { + "epoch": 0.43321, + "grad_norm": 1.6084689068088867, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43321 + }, + { + "epoch": 0.43322, + "grad_norm": 1.4882449298402265, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43322 + }, + { + "epoch": 0.43323, + "grad_norm": 1.5856566580212166, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43323 + }, + { + "epoch": 0.43324, + "grad_norm": 1.573854936045617, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 43324 + }, + { + "epoch": 0.43325, + "grad_norm": 1.4321506102834463, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 43325 + }, + { + "epoch": 0.43326, + "grad_norm": 1.619222438348933, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43326 + }, + { + "epoch": 0.43327, + "grad_norm": 1.3234084861502045, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43327 + }, + { + "epoch": 0.43328, + "grad_norm": 1.6629228732288377, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 43328 + }, + { + "epoch": 0.43329, + "grad_norm": 1.3129069492923595, + "learning_rate": 0.003, + "loss": 4.07, + "step": 43329 + }, + { + "epoch": 0.4333, + "grad_norm": 1.6343925805499389, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43330 + }, + { + "epoch": 0.43331, + "grad_norm": 1.4463290171438572, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 43331 + }, + { + "epoch": 0.43332, + "grad_norm": 1.5543359438415236, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 43332 + }, + { + "epoch": 0.43333, + "grad_norm": 1.4986327485800246, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 43333 + }, + { + "epoch": 0.43334, + "grad_norm": 1.6572086590282957, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 43334 + }, + { + "epoch": 0.43335, + "grad_norm": 1.2817055352178532, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 43335 + }, + { + "epoch": 0.43336, + "grad_norm": 1.4427154011209626, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 43336 + }, + { + "epoch": 0.43337, + "grad_norm": 1.7253078336975822, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 43337 + }, + { + "epoch": 0.43338, + "grad_norm": 1.7909572112780394, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 43338 + }, + { + "epoch": 0.43339, + "grad_norm": 1.4058265796452554, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 43339 + }, + { + "epoch": 0.4334, + "grad_norm": 1.557277886073931, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 43340 + }, + { + "epoch": 0.43341, + "grad_norm": 1.6564883882532202, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43341 + }, + { + "epoch": 0.43342, + "grad_norm": 1.749431744067087, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 43342 + }, + { + "epoch": 0.43343, + "grad_norm": 1.5122536007150935, + "learning_rate": 0.003, + "loss": 4.051, + "step": 43343 + }, + { + "epoch": 0.43344, + "grad_norm": 1.541150769329685, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 43344 + }, + { + "epoch": 0.43345, + "grad_norm": 1.4296966578324488, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43345 + }, + { + "epoch": 0.43346, + "grad_norm": 1.5924816289189385, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 43346 + }, + { + "epoch": 0.43347, + "grad_norm": 1.73858550554351, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 43347 + }, + { + "epoch": 0.43348, + "grad_norm": 1.6186552100902056, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43348 + }, + { + "epoch": 0.43349, + "grad_norm": 1.455422975444279, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 43349 + }, + { + "epoch": 0.4335, + "grad_norm": 1.686690844454224, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43350 + }, + { + "epoch": 0.43351, + "grad_norm": 1.43101248592871, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 43351 + }, + { + "epoch": 0.43352, + "grad_norm": 1.5380880911728234, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 43352 + }, + { + "epoch": 0.43353, + "grad_norm": 1.6986235380993049, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 43353 + }, + { + "epoch": 0.43354, + "grad_norm": 1.4021588481158613, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43354 + }, + { + "epoch": 0.43355, + "grad_norm": 1.6223623018351672, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 43355 + }, + { + "epoch": 0.43356, + "grad_norm": 1.3346879915598169, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 43356 + }, + { + "epoch": 0.43357, + "grad_norm": 1.652927095414591, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43357 + }, + { + "epoch": 0.43358, + "grad_norm": 1.420650058572768, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43358 + }, + { + "epoch": 0.43359, + "grad_norm": 1.3725315980340151, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43359 + }, + { + "epoch": 0.4336, + "grad_norm": 1.5111916667515084, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 43360 + }, + { + "epoch": 0.43361, + "grad_norm": 1.4356981951690926, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 43361 + }, + { + "epoch": 0.43362, + "grad_norm": 1.7656980460210419, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 43362 + }, + { + "epoch": 0.43363, + "grad_norm": 1.2435713398319612, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 43363 + }, + { + "epoch": 0.43364, + "grad_norm": 1.38415598272005, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43364 + }, + { + "epoch": 0.43365, + "grad_norm": 1.5705428220770628, + "learning_rate": 0.003, + "loss": 4.059, + "step": 43365 + }, + { + "epoch": 0.43366, + "grad_norm": 1.6521246271782444, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 43366 + }, + { + "epoch": 0.43367, + "grad_norm": 1.837261052420493, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43367 + }, + { + "epoch": 0.43368, + "grad_norm": 1.3978458920494758, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 43368 + }, + { + "epoch": 0.43369, + "grad_norm": 1.7966856157893762, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 43369 + }, + { + "epoch": 0.4337, + "grad_norm": 1.811572617133702, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 43370 + }, + { + "epoch": 0.43371, + "grad_norm": 1.2612113769130955, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43371 + }, + { + "epoch": 0.43372, + "grad_norm": 1.549862937527303, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 43372 + }, + { + "epoch": 0.43373, + "grad_norm": 1.5174060886696723, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 43373 + }, + { + "epoch": 0.43374, + "grad_norm": 1.8410080698634994, + "learning_rate": 0.003, + "loss": 4.076, + "step": 43374 + }, + { + "epoch": 0.43375, + "grad_norm": 1.2878251280646138, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 43375 + }, + { + "epoch": 0.43376, + "grad_norm": 1.7571277306102482, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 43376 + }, + { + "epoch": 0.43377, + "grad_norm": 1.5715368166758206, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 43377 + }, + { + "epoch": 0.43378, + "grad_norm": 1.4739660660925966, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 43378 + }, + { + "epoch": 0.43379, + "grad_norm": 1.437109385913058, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 43379 + }, + { + "epoch": 0.4338, + "grad_norm": 1.5217266086097399, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 43380 + }, + { + "epoch": 0.43381, + "grad_norm": 1.3651892197266935, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43381 + }, + { + "epoch": 0.43382, + "grad_norm": 1.8360805237092117, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 43382 + }, + { + "epoch": 0.43383, + "grad_norm": 1.3582479540833365, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43383 + }, + { + "epoch": 0.43384, + "grad_norm": 1.8008166772473144, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 43384 + }, + { + "epoch": 0.43385, + "grad_norm": 1.7844678693908638, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43385 + }, + { + "epoch": 0.43386, + "grad_norm": 1.4866883538494693, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 43386 + }, + { + "epoch": 0.43387, + "grad_norm": 1.43987123286629, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 43387 + }, + { + "epoch": 0.43388, + "grad_norm": 1.597280019877736, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 43388 + }, + { + "epoch": 0.43389, + "grad_norm": 1.4473490304681558, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 43389 + }, + { + "epoch": 0.4339, + "grad_norm": 1.5040388593581115, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43390 + }, + { + "epoch": 0.43391, + "grad_norm": 1.7810271109015652, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43391 + }, + { + "epoch": 0.43392, + "grad_norm": 1.4484975387869712, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 43392 + }, + { + "epoch": 0.43393, + "grad_norm": 1.6767833792039775, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 43393 + }, + { + "epoch": 0.43394, + "grad_norm": 1.576579051205189, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 43394 + }, + { + "epoch": 0.43395, + "grad_norm": 1.8323426423740574, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 43395 + }, + { + "epoch": 0.43396, + "grad_norm": 1.2497357498596495, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 43396 + }, + { + "epoch": 0.43397, + "grad_norm": 1.7285065153196892, + "learning_rate": 0.003, + "loss": 4.018, + "step": 43397 + }, + { + "epoch": 0.43398, + "grad_norm": 1.3520822750204966, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 43398 + }, + { + "epoch": 0.43399, + "grad_norm": 1.7815365365379727, + "learning_rate": 0.003, + "loss": 4.059, + "step": 43399 + }, + { + "epoch": 0.434, + "grad_norm": 1.421101674912577, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43400 + }, + { + "epoch": 0.43401, + "grad_norm": 1.7541730287170212, + "learning_rate": 0.003, + "loss": 4.055, + "step": 43401 + }, + { + "epoch": 0.43402, + "grad_norm": 1.3955885258578073, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43402 + }, + { + "epoch": 0.43403, + "grad_norm": 1.6471647185568705, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 43403 + }, + { + "epoch": 0.43404, + "grad_norm": 1.7250399212207408, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 43404 + }, + { + "epoch": 0.43405, + "grad_norm": 1.452907463857625, + "learning_rate": 0.003, + "loss": 4.027, + "step": 43405 + }, + { + "epoch": 0.43406, + "grad_norm": 1.666278798502987, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 43406 + }, + { + "epoch": 0.43407, + "grad_norm": 1.56640191998055, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 43407 + }, + { + "epoch": 0.43408, + "grad_norm": 1.5583761526401083, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43408 + }, + { + "epoch": 0.43409, + "grad_norm": 1.4121761273108302, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43409 + }, + { + "epoch": 0.4341, + "grad_norm": 1.4226905542461707, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43410 + }, + { + "epoch": 0.43411, + "grad_norm": 1.3003037953486147, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 43411 + }, + { + "epoch": 0.43412, + "grad_norm": 1.5939140175275206, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43412 + }, + { + "epoch": 0.43413, + "grad_norm": 1.587925513391551, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 43413 + }, + { + "epoch": 0.43414, + "grad_norm": 1.3823298956020975, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43414 + }, + { + "epoch": 0.43415, + "grad_norm": 1.8166232593846443, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 43415 + }, + { + "epoch": 0.43416, + "grad_norm": 1.3066970314177, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 43416 + }, + { + "epoch": 0.43417, + "grad_norm": 1.7385662979688092, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43417 + }, + { + "epoch": 0.43418, + "grad_norm": 1.3403562713043555, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 43418 + }, + { + "epoch": 0.43419, + "grad_norm": 1.5358224397713394, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43419 + }, + { + "epoch": 0.4342, + "grad_norm": 1.390248363847066, + "learning_rate": 0.003, + "loss": 4.018, + "step": 43420 + }, + { + "epoch": 0.43421, + "grad_norm": 1.3780450388341727, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 43421 + }, + { + "epoch": 0.43422, + "grad_norm": 1.4859017144724926, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 43422 + }, + { + "epoch": 0.43423, + "grad_norm": 1.6321012861936348, + "learning_rate": 0.003, + "loss": 4.031, + "step": 43423 + }, + { + "epoch": 0.43424, + "grad_norm": 1.8524777192925874, + "learning_rate": 0.003, + "loss": 4.041, + "step": 43424 + }, + { + "epoch": 0.43425, + "grad_norm": 1.9864407912964366, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43425 + }, + { + "epoch": 0.43426, + "grad_norm": 1.3392950850379053, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 43426 + }, + { + "epoch": 0.43427, + "grad_norm": 1.411550911904768, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 43427 + }, + { + "epoch": 0.43428, + "grad_norm": 1.7200466113614485, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43428 + }, + { + "epoch": 0.43429, + "grad_norm": 1.4427351629572744, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 43429 + }, + { + "epoch": 0.4343, + "grad_norm": 1.6367777294474481, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 43430 + }, + { + "epoch": 0.43431, + "grad_norm": 1.4212622184290236, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 43431 + }, + { + "epoch": 0.43432, + "grad_norm": 1.8720722914853145, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 43432 + }, + { + "epoch": 0.43433, + "grad_norm": 1.4665194464269973, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 43433 + }, + { + "epoch": 0.43434, + "grad_norm": 1.7873243281316147, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 43434 + }, + { + "epoch": 0.43435, + "grad_norm": 1.5493020994094766, + "learning_rate": 0.003, + "loss": 4.043, + "step": 43435 + }, + { + "epoch": 0.43436, + "grad_norm": 1.6006867048467828, + "learning_rate": 0.003, + "loss": 4.031, + "step": 43436 + }, + { + "epoch": 0.43437, + "grad_norm": 1.4896745407686105, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 43437 + }, + { + "epoch": 0.43438, + "grad_norm": 1.6945019804323622, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43438 + }, + { + "epoch": 0.43439, + "grad_norm": 1.3175929555113766, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 43439 + }, + { + "epoch": 0.4344, + "grad_norm": 1.4771483166822839, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 43440 + }, + { + "epoch": 0.43441, + "grad_norm": 1.7021048024397305, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43441 + }, + { + "epoch": 0.43442, + "grad_norm": 1.3264716511304984, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 43442 + }, + { + "epoch": 0.43443, + "grad_norm": 1.4946528904501724, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 43443 + }, + { + "epoch": 0.43444, + "grad_norm": 1.518783242679207, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 43444 + }, + { + "epoch": 0.43445, + "grad_norm": 1.5582591120724665, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43445 + }, + { + "epoch": 0.43446, + "grad_norm": 1.5193705705036127, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 43446 + }, + { + "epoch": 0.43447, + "grad_norm": 1.3322660053702582, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 43447 + }, + { + "epoch": 0.43448, + "grad_norm": 1.7157456452918813, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 43448 + }, + { + "epoch": 0.43449, + "grad_norm": 1.7053215278782268, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 43449 + }, + { + "epoch": 0.4345, + "grad_norm": 1.337150298847912, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43450 + }, + { + "epoch": 0.43451, + "grad_norm": 2.1488410274533245, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 43451 + }, + { + "epoch": 0.43452, + "grad_norm": 1.3557822366625067, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 43452 + }, + { + "epoch": 0.43453, + "grad_norm": 1.9558796434698942, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43453 + }, + { + "epoch": 0.43454, + "grad_norm": 1.710768779343557, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43454 + }, + { + "epoch": 0.43455, + "grad_norm": 1.2027337247784604, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43455 + }, + { + "epoch": 0.43456, + "grad_norm": 1.7456587578718366, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 43456 + }, + { + "epoch": 0.43457, + "grad_norm": 1.620011027723755, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 43457 + }, + { + "epoch": 0.43458, + "grad_norm": 1.3679468617817927, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 43458 + }, + { + "epoch": 0.43459, + "grad_norm": 1.651311154270144, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 43459 + }, + { + "epoch": 0.4346, + "grad_norm": 1.5903732122225562, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43460 + }, + { + "epoch": 0.43461, + "grad_norm": 1.6410655807031194, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 43461 + }, + { + "epoch": 0.43462, + "grad_norm": 1.3417163812107988, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 43462 + }, + { + "epoch": 0.43463, + "grad_norm": 1.671474037073577, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 43463 + }, + { + "epoch": 0.43464, + "grad_norm": 1.4224051419882213, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 43464 + }, + { + "epoch": 0.43465, + "grad_norm": 2.001917721895832, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 43465 + }, + { + "epoch": 0.43466, + "grad_norm": 1.4699355965032121, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 43466 + }, + { + "epoch": 0.43467, + "grad_norm": 1.8018019497725943, + "learning_rate": 0.003, + "loss": 4.056, + "step": 43467 + }, + { + "epoch": 0.43468, + "grad_norm": 1.7221086499226457, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 43468 + }, + { + "epoch": 0.43469, + "grad_norm": 1.6327397552483958, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43469 + }, + { + "epoch": 0.4347, + "grad_norm": 1.8660243148051407, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 43470 + }, + { + "epoch": 0.43471, + "grad_norm": 1.3210793303954091, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 43471 + }, + { + "epoch": 0.43472, + "grad_norm": 1.5584942979056444, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43472 + }, + { + "epoch": 0.43473, + "grad_norm": 1.2400563779584883, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 43473 + }, + { + "epoch": 0.43474, + "grad_norm": 1.5437194711669253, + "learning_rate": 0.003, + "loss": 4.028, + "step": 43474 + }, + { + "epoch": 0.43475, + "grad_norm": 1.773893735256746, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 43475 + }, + { + "epoch": 0.43476, + "grad_norm": 1.4858222060041455, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43476 + }, + { + "epoch": 0.43477, + "grad_norm": 1.511788935507152, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 43477 + }, + { + "epoch": 0.43478, + "grad_norm": 1.4352581923210843, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43478 + }, + { + "epoch": 0.43479, + "grad_norm": 1.4944800882313838, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43479 + }, + { + "epoch": 0.4348, + "grad_norm": 1.3734412892147498, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 43480 + }, + { + "epoch": 0.43481, + "grad_norm": 1.470512400668709, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 43481 + }, + { + "epoch": 0.43482, + "grad_norm": 1.466551686286597, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 43482 + }, + { + "epoch": 0.43483, + "grad_norm": 1.3696887899577672, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 43483 + }, + { + "epoch": 0.43484, + "grad_norm": 1.5577045549768112, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 43484 + }, + { + "epoch": 0.43485, + "grad_norm": 1.2833174533862068, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 43485 + }, + { + "epoch": 0.43486, + "grad_norm": 1.9334876146327256, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 43486 + }, + { + "epoch": 0.43487, + "grad_norm": 1.2717517470454303, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 43487 + }, + { + "epoch": 0.43488, + "grad_norm": 2.280855223822504, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 43488 + }, + { + "epoch": 0.43489, + "grad_norm": 1.6277063083634904, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43489 + }, + { + "epoch": 0.4349, + "grad_norm": 1.4306840985788762, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 43490 + }, + { + "epoch": 0.43491, + "grad_norm": 1.59519100055235, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 43491 + }, + { + "epoch": 0.43492, + "grad_norm": 1.6841871176802823, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43492 + }, + { + "epoch": 0.43493, + "grad_norm": 1.3422541063876263, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 43493 + }, + { + "epoch": 0.43494, + "grad_norm": 1.7826582585598938, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 43494 + }, + { + "epoch": 0.43495, + "grad_norm": 1.2549221041373488, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 43495 + }, + { + "epoch": 0.43496, + "grad_norm": 1.5062076704176777, + "learning_rate": 0.003, + "loss": 4.049, + "step": 43496 + }, + { + "epoch": 0.43497, + "grad_norm": 1.5681942032840333, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 43497 + }, + { + "epoch": 0.43498, + "grad_norm": 1.505579073443114, + "learning_rate": 0.003, + "loss": 4.044, + "step": 43498 + }, + { + "epoch": 0.43499, + "grad_norm": 1.8672729375087416, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 43499 + }, + { + "epoch": 0.435, + "grad_norm": 1.563529662083144, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 43500 + }, + { + "epoch": 0.43501, + "grad_norm": 1.6884213057701307, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 43501 + }, + { + "epoch": 0.43502, + "grad_norm": 1.3099777971768625, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 43502 + }, + { + "epoch": 0.43503, + "grad_norm": 1.6087963199672353, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43503 + }, + { + "epoch": 0.43504, + "grad_norm": 1.2483757149386097, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43504 + }, + { + "epoch": 0.43505, + "grad_norm": 1.5790409370344036, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 43505 + }, + { + "epoch": 0.43506, + "grad_norm": 1.222761215935391, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 43506 + }, + { + "epoch": 0.43507, + "grad_norm": 1.8015517243881738, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 43507 + }, + { + "epoch": 0.43508, + "grad_norm": 1.3506916307565955, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 43508 + }, + { + "epoch": 0.43509, + "grad_norm": 1.6042404144677591, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 43509 + }, + { + "epoch": 0.4351, + "grad_norm": 1.627495048194674, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 43510 + }, + { + "epoch": 0.43511, + "grad_norm": 1.4480701009294399, + "learning_rate": 0.003, + "loss": 4.044, + "step": 43511 + }, + { + "epoch": 0.43512, + "grad_norm": 1.923367288896827, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 43512 + }, + { + "epoch": 0.43513, + "grad_norm": 1.3453841596793148, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 43513 + }, + { + "epoch": 0.43514, + "grad_norm": 1.6215044392543094, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 43514 + }, + { + "epoch": 0.43515, + "grad_norm": 1.5818711572700117, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 43515 + }, + { + "epoch": 0.43516, + "grad_norm": 1.85041065001155, + "learning_rate": 0.003, + "loss": 4.028, + "step": 43516 + }, + { + "epoch": 0.43517, + "grad_norm": 1.321691752398167, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 43517 + }, + { + "epoch": 0.43518, + "grad_norm": 1.813515565405518, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 43518 + }, + { + "epoch": 0.43519, + "grad_norm": 1.3979303536106824, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 43519 + }, + { + "epoch": 0.4352, + "grad_norm": 1.6355660425181322, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 43520 + }, + { + "epoch": 0.43521, + "grad_norm": 1.793856171993557, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43521 + }, + { + "epoch": 0.43522, + "grad_norm": 1.3082544309224509, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 43522 + }, + { + "epoch": 0.43523, + "grad_norm": 1.5898115413609548, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 43523 + }, + { + "epoch": 0.43524, + "grad_norm": 1.2589937218773735, + "learning_rate": 0.003, + "loss": 4.021, + "step": 43524 + }, + { + "epoch": 0.43525, + "grad_norm": 1.4030236135484666, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 43525 + }, + { + "epoch": 0.43526, + "grad_norm": 1.3230062326169174, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 43526 + }, + { + "epoch": 0.43527, + "grad_norm": 1.6810223036510097, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 43527 + }, + { + "epoch": 0.43528, + "grad_norm": 1.5580910453742118, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 43528 + }, + { + "epoch": 0.43529, + "grad_norm": 1.6364406761478951, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 43529 + }, + { + "epoch": 0.4353, + "grad_norm": 1.4421203925854928, + "learning_rate": 0.003, + "loss": 4.046, + "step": 43530 + }, + { + "epoch": 0.43531, + "grad_norm": 2.081365835477371, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 43531 + }, + { + "epoch": 0.43532, + "grad_norm": 1.2565377919625564, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 43532 + }, + { + "epoch": 0.43533, + "grad_norm": 1.6940510756692366, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 43533 + }, + { + "epoch": 0.43534, + "grad_norm": 1.4085273317867342, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43534 + }, + { + "epoch": 0.43535, + "grad_norm": 1.6174223100234448, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 43535 + }, + { + "epoch": 0.43536, + "grad_norm": 1.4752734708336137, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 43536 + }, + { + "epoch": 0.43537, + "grad_norm": 1.5531187405664704, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 43537 + }, + { + "epoch": 0.43538, + "grad_norm": 1.5367803738326766, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 43538 + }, + { + "epoch": 0.43539, + "grad_norm": 1.450999913741856, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43539 + }, + { + "epoch": 0.4354, + "grad_norm": 1.3696149024660478, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 43540 + }, + { + "epoch": 0.43541, + "grad_norm": 1.7192686270373772, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 43541 + }, + { + "epoch": 0.43542, + "grad_norm": 1.572145406443524, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43542 + }, + { + "epoch": 0.43543, + "grad_norm": 1.6745085734628065, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 43543 + }, + { + "epoch": 0.43544, + "grad_norm": 1.493492338491915, + "learning_rate": 0.003, + "loss": 4.001, + "step": 43544 + }, + { + "epoch": 0.43545, + "grad_norm": 1.586796771266783, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 43545 + }, + { + "epoch": 0.43546, + "grad_norm": 1.3991474909341903, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 43546 + }, + { + "epoch": 0.43547, + "grad_norm": 1.578407174994117, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 43547 + }, + { + "epoch": 0.43548, + "grad_norm": 1.3888767522772745, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 43548 + }, + { + "epoch": 0.43549, + "grad_norm": 1.9167426067187696, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 43549 + }, + { + "epoch": 0.4355, + "grad_norm": 1.4261308205560561, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 43550 + }, + { + "epoch": 0.43551, + "grad_norm": 1.8173118188043866, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 43551 + }, + { + "epoch": 0.43552, + "grad_norm": 1.50712092638188, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 43552 + }, + { + "epoch": 0.43553, + "grad_norm": 1.5380979630407574, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 43553 + }, + { + "epoch": 0.43554, + "grad_norm": 1.1500877264058693, + "learning_rate": 0.003, + "loss": 4.023, + "step": 43554 + }, + { + "epoch": 0.43555, + "grad_norm": 1.579885855260568, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 43555 + }, + { + "epoch": 0.43556, + "grad_norm": 1.3350554014924678, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 43556 + }, + { + "epoch": 0.43557, + "grad_norm": 1.6061401811108476, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 43557 + }, + { + "epoch": 0.43558, + "grad_norm": 1.3815905339230963, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 43558 + }, + { + "epoch": 0.43559, + "grad_norm": 1.570009869243284, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 43559 + }, + { + "epoch": 0.4356, + "grad_norm": 1.6822902216943534, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 43560 + }, + { + "epoch": 0.43561, + "grad_norm": 1.5102595498182791, + "learning_rate": 0.003, + "loss": 4.006, + "step": 43561 + }, + { + "epoch": 0.43562, + "grad_norm": 1.814072541290204, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43562 + }, + { + "epoch": 0.43563, + "grad_norm": 1.3312183929531365, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 43563 + }, + { + "epoch": 0.43564, + "grad_norm": 1.5198321370000327, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 43564 + }, + { + "epoch": 0.43565, + "grad_norm": 1.6830625293258823, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43565 + }, + { + "epoch": 0.43566, + "grad_norm": 1.2480830002724541, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 43566 + }, + { + "epoch": 0.43567, + "grad_norm": 1.9356329376717205, + "learning_rate": 0.003, + "loss": 4.038, + "step": 43567 + }, + { + "epoch": 0.43568, + "grad_norm": 1.1903821876976224, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 43568 + }, + { + "epoch": 0.43569, + "grad_norm": 1.8438555373936965, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 43569 + }, + { + "epoch": 0.4357, + "grad_norm": 1.5149951930899046, + "learning_rate": 0.003, + "loss": 4.065, + "step": 43570 + }, + { + "epoch": 0.43571, + "grad_norm": 1.7529432238823939, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43571 + }, + { + "epoch": 0.43572, + "grad_norm": 1.5745255808019891, + "learning_rate": 0.003, + "loss": 4.056, + "step": 43572 + }, + { + "epoch": 0.43573, + "grad_norm": 1.4901935148150174, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 43573 + }, + { + "epoch": 0.43574, + "grad_norm": 1.395385119194324, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43574 + }, + { + "epoch": 0.43575, + "grad_norm": 1.2892961619894285, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43575 + }, + { + "epoch": 0.43576, + "grad_norm": 1.7944930572028999, + "learning_rate": 0.003, + "loss": 4.063, + "step": 43576 + }, + { + "epoch": 0.43577, + "grad_norm": 1.5038880268461763, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 43577 + }, + { + "epoch": 0.43578, + "grad_norm": 1.6199585384333715, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 43578 + }, + { + "epoch": 0.43579, + "grad_norm": 1.5538906023621057, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 43579 + }, + { + "epoch": 0.4358, + "grad_norm": 1.8131783426862718, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 43580 + }, + { + "epoch": 0.43581, + "grad_norm": 1.1647400124086182, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 43581 + }, + { + "epoch": 0.43582, + "grad_norm": 1.839697069763491, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 43582 + }, + { + "epoch": 0.43583, + "grad_norm": 1.41532292879719, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 43583 + }, + { + "epoch": 0.43584, + "grad_norm": 1.9362018300840642, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 43584 + }, + { + "epoch": 0.43585, + "grad_norm": 1.090319877866824, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 43585 + }, + { + "epoch": 0.43586, + "grad_norm": 1.6822429154964875, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43586 + }, + { + "epoch": 0.43587, + "grad_norm": 1.4517409633474865, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 43587 + }, + { + "epoch": 0.43588, + "grad_norm": 1.9742368773814416, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 43588 + }, + { + "epoch": 0.43589, + "grad_norm": 1.1730583320882773, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 43589 + }, + { + "epoch": 0.4359, + "grad_norm": 1.7738570349200704, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 43590 + }, + { + "epoch": 0.43591, + "grad_norm": 1.4799939828034647, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 43591 + }, + { + "epoch": 0.43592, + "grad_norm": 1.4757756822434418, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43592 + }, + { + "epoch": 0.43593, + "grad_norm": 1.7621762916868238, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43593 + }, + { + "epoch": 0.43594, + "grad_norm": 1.6697144982938183, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43594 + }, + { + "epoch": 0.43595, + "grad_norm": 1.1954464911758897, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 43595 + }, + { + "epoch": 0.43596, + "grad_norm": 2.0216289417153326, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 43596 + }, + { + "epoch": 0.43597, + "grad_norm": 1.1356206710641823, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43597 + }, + { + "epoch": 0.43598, + "grad_norm": 1.561380603607594, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 43598 + }, + { + "epoch": 0.43599, + "grad_norm": 1.472232292030812, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 43599 + }, + { + "epoch": 0.436, + "grad_norm": 1.7430258157711096, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 43600 + }, + { + "epoch": 0.43601, + "grad_norm": 1.5183615538400066, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 43601 + }, + { + "epoch": 0.43602, + "grad_norm": 1.6655344229175757, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 43602 + }, + { + "epoch": 0.43603, + "grad_norm": 1.7185239468156654, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 43603 + }, + { + "epoch": 0.43604, + "grad_norm": 1.5673319661442868, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 43604 + }, + { + "epoch": 0.43605, + "grad_norm": 1.3028164570297966, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 43605 + }, + { + "epoch": 0.43606, + "grad_norm": 1.5727170706992508, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43606 + }, + { + "epoch": 0.43607, + "grad_norm": 1.3935961298605204, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 43607 + }, + { + "epoch": 0.43608, + "grad_norm": 1.4735782593563074, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 43608 + }, + { + "epoch": 0.43609, + "grad_norm": 1.5688002440339366, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 43609 + }, + { + "epoch": 0.4361, + "grad_norm": 1.6733558849048493, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 43610 + }, + { + "epoch": 0.43611, + "grad_norm": 1.328934456301142, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 43611 + }, + { + "epoch": 0.43612, + "grad_norm": 1.585934259993639, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 43612 + }, + { + "epoch": 0.43613, + "grad_norm": 1.6302660995205962, + "learning_rate": 0.003, + "loss": 4.031, + "step": 43613 + }, + { + "epoch": 0.43614, + "grad_norm": 1.7769194119897067, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43614 + }, + { + "epoch": 0.43615, + "grad_norm": 1.4479260824691833, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43615 + }, + { + "epoch": 0.43616, + "grad_norm": 1.1761169946636412, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 43616 + }, + { + "epoch": 0.43617, + "grad_norm": 1.6701461121450318, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 43617 + }, + { + "epoch": 0.43618, + "grad_norm": 1.1315839572999542, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 43618 + }, + { + "epoch": 0.43619, + "grad_norm": 1.829100409320443, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 43619 + }, + { + "epoch": 0.4362, + "grad_norm": 1.2293965305759522, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 43620 + }, + { + "epoch": 0.43621, + "grad_norm": 1.8351944042690684, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 43621 + }, + { + "epoch": 0.43622, + "grad_norm": 1.651699972115226, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 43622 + }, + { + "epoch": 0.43623, + "grad_norm": 1.7830805345335168, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43623 + }, + { + "epoch": 0.43624, + "grad_norm": 1.4931144811477073, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 43624 + }, + { + "epoch": 0.43625, + "grad_norm": 1.3795402358115902, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 43625 + }, + { + "epoch": 0.43626, + "grad_norm": 1.4716173519938756, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 43626 + }, + { + "epoch": 0.43627, + "grad_norm": 1.5074146092088663, + "learning_rate": 0.003, + "loss": 4.029, + "step": 43627 + }, + { + "epoch": 0.43628, + "grad_norm": 1.3034573678790333, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 43628 + }, + { + "epoch": 0.43629, + "grad_norm": 1.41525997243601, + "learning_rate": 0.003, + "loss": 4.016, + "step": 43629 + }, + { + "epoch": 0.4363, + "grad_norm": 1.5645085368687086, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 43630 + }, + { + "epoch": 0.43631, + "grad_norm": 1.5760150435514255, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 43631 + }, + { + "epoch": 0.43632, + "grad_norm": 1.669546360444416, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43632 + }, + { + "epoch": 0.43633, + "grad_norm": 1.3240936807150947, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 43633 + }, + { + "epoch": 0.43634, + "grad_norm": 1.9052047810121167, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 43634 + }, + { + "epoch": 0.43635, + "grad_norm": 1.4461550543152422, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43635 + }, + { + "epoch": 0.43636, + "grad_norm": 1.3838148691118863, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 43636 + }, + { + "epoch": 0.43637, + "grad_norm": 1.6453109624099944, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 43637 + }, + { + "epoch": 0.43638, + "grad_norm": 1.4203937697438993, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 43638 + }, + { + "epoch": 0.43639, + "grad_norm": 1.8382833383060613, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 43639 + }, + { + "epoch": 0.4364, + "grad_norm": 1.4719474979910914, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 43640 + }, + { + "epoch": 0.43641, + "grad_norm": 1.8265279997037358, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43641 + }, + { + "epoch": 0.43642, + "grad_norm": 1.424173475947151, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 43642 + }, + { + "epoch": 0.43643, + "grad_norm": 1.6060752375160057, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 43643 + }, + { + "epoch": 0.43644, + "grad_norm": 1.6648526414253972, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 43644 + }, + { + "epoch": 0.43645, + "grad_norm": 1.595094474011574, + "learning_rate": 0.003, + "loss": 3.999, + "step": 43645 + }, + { + "epoch": 0.43646, + "grad_norm": 1.5141734903916444, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43646 + }, + { + "epoch": 0.43647, + "grad_norm": 1.6018776161675976, + "learning_rate": 0.003, + "loss": 4.025, + "step": 43647 + }, + { + "epoch": 0.43648, + "grad_norm": 1.3910535281542047, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 43648 + }, + { + "epoch": 0.43649, + "grad_norm": 1.6281977791334936, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 43649 + }, + { + "epoch": 0.4365, + "grad_norm": 1.330808303506724, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 43650 + }, + { + "epoch": 0.43651, + "grad_norm": 1.8201678790006712, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 43651 + }, + { + "epoch": 0.43652, + "grad_norm": 1.1910854143223635, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 43652 + }, + { + "epoch": 0.43653, + "grad_norm": 1.6070538331376572, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43653 + }, + { + "epoch": 0.43654, + "grad_norm": 1.4583165213619318, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43654 + }, + { + "epoch": 0.43655, + "grad_norm": 1.7812415092719354, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 43655 + }, + { + "epoch": 0.43656, + "grad_norm": 1.2533123881166073, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43656 + }, + { + "epoch": 0.43657, + "grad_norm": 1.783536918532318, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 43657 + }, + { + "epoch": 0.43658, + "grad_norm": 1.45215576982044, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 43658 + }, + { + "epoch": 0.43659, + "grad_norm": 1.817000346981522, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 43659 + }, + { + "epoch": 0.4366, + "grad_norm": 1.6594188877085274, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 43660 + }, + { + "epoch": 0.43661, + "grad_norm": 1.429905074160767, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 43661 + }, + { + "epoch": 0.43662, + "grad_norm": 1.523651359556832, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43662 + }, + { + "epoch": 0.43663, + "grad_norm": 1.9422840092588223, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43663 + }, + { + "epoch": 0.43664, + "grad_norm": 1.267342748036905, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43664 + }, + { + "epoch": 0.43665, + "grad_norm": 1.5600753960970077, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 43665 + }, + { + "epoch": 0.43666, + "grad_norm": 1.5371054905355068, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 43666 + }, + { + "epoch": 0.43667, + "grad_norm": 1.5277598160200374, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 43667 + }, + { + "epoch": 0.43668, + "grad_norm": 1.4717974220652195, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43668 + }, + { + "epoch": 0.43669, + "grad_norm": 1.8107769263290046, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 43669 + }, + { + "epoch": 0.4367, + "grad_norm": 1.4855052763030736, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 43670 + }, + { + "epoch": 0.43671, + "grad_norm": 1.6886882721696475, + "learning_rate": 0.003, + "loss": 4.063, + "step": 43671 + }, + { + "epoch": 0.43672, + "grad_norm": 1.3616388811713056, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 43672 + }, + { + "epoch": 0.43673, + "grad_norm": 1.6028809373981692, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 43673 + }, + { + "epoch": 0.43674, + "grad_norm": 1.4367284862407224, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 43674 + }, + { + "epoch": 0.43675, + "grad_norm": 1.9200830121786887, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 43675 + }, + { + "epoch": 0.43676, + "grad_norm": 1.2558456152238633, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 43676 + }, + { + "epoch": 0.43677, + "grad_norm": 1.9198115667601834, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 43677 + }, + { + "epoch": 0.43678, + "grad_norm": 1.3433719993247517, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43678 + }, + { + "epoch": 0.43679, + "grad_norm": 1.7833503311696015, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 43679 + }, + { + "epoch": 0.4368, + "grad_norm": 1.612571833743844, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 43680 + }, + { + "epoch": 0.43681, + "grad_norm": 1.4807341742464455, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 43681 + }, + { + "epoch": 0.43682, + "grad_norm": 1.5608625528504136, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 43682 + }, + { + "epoch": 0.43683, + "grad_norm": 1.2677903963381825, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43683 + }, + { + "epoch": 0.43684, + "grad_norm": 1.7377186683303767, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 43684 + }, + { + "epoch": 0.43685, + "grad_norm": 1.2973883459059647, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 43685 + }, + { + "epoch": 0.43686, + "grad_norm": 1.738336112371878, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 43686 + }, + { + "epoch": 0.43687, + "grad_norm": 1.3469038269643654, + "learning_rate": 0.003, + "loss": 4.051, + "step": 43687 + }, + { + "epoch": 0.43688, + "grad_norm": 1.510501453976268, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43688 + }, + { + "epoch": 0.43689, + "grad_norm": 1.48906064792801, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43689 + }, + { + "epoch": 0.4369, + "grad_norm": 1.8386344311261533, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 43690 + }, + { + "epoch": 0.43691, + "grad_norm": 1.4318601730793494, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 43691 + }, + { + "epoch": 0.43692, + "grad_norm": 1.9156697622628716, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43692 + }, + { + "epoch": 0.43693, + "grad_norm": 1.5216193684078574, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 43693 + }, + { + "epoch": 0.43694, + "grad_norm": 1.4582655545199477, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 43694 + }, + { + "epoch": 0.43695, + "grad_norm": 1.4693862327769391, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 43695 + }, + { + "epoch": 0.43696, + "grad_norm": 1.4376410443739216, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 43696 + }, + { + "epoch": 0.43697, + "grad_norm": 1.6076205133202952, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43697 + }, + { + "epoch": 0.43698, + "grad_norm": 1.489473641837189, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 43698 + }, + { + "epoch": 0.43699, + "grad_norm": 1.7019301215498774, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43699 + }, + { + "epoch": 0.437, + "grad_norm": 1.3123759230701124, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 43700 + }, + { + "epoch": 0.43701, + "grad_norm": 1.8265504365798926, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 43701 + }, + { + "epoch": 0.43702, + "grad_norm": 1.5293496109799065, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 43702 + }, + { + "epoch": 0.43703, + "grad_norm": 1.6760167584755727, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 43703 + }, + { + "epoch": 0.43704, + "grad_norm": 1.4267367511189564, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 43704 + }, + { + "epoch": 0.43705, + "grad_norm": 1.601813416744293, + "learning_rate": 0.003, + "loss": 4.033, + "step": 43705 + }, + { + "epoch": 0.43706, + "grad_norm": 1.658807854246021, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 43706 + }, + { + "epoch": 0.43707, + "grad_norm": 1.3840278330260853, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 43707 + }, + { + "epoch": 0.43708, + "grad_norm": 1.73356918843672, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43708 + }, + { + "epoch": 0.43709, + "grad_norm": 1.3631097019311076, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 43709 + }, + { + "epoch": 0.4371, + "grad_norm": 1.6547625513151842, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 43710 + }, + { + "epoch": 0.43711, + "grad_norm": 1.3327454751303294, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 43711 + }, + { + "epoch": 0.43712, + "grad_norm": 1.849695113795042, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 43712 + }, + { + "epoch": 0.43713, + "grad_norm": 1.3758181224939476, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43713 + }, + { + "epoch": 0.43714, + "grad_norm": 2.130181664844866, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43714 + }, + { + "epoch": 0.43715, + "grad_norm": 1.4106901536093057, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43715 + }, + { + "epoch": 0.43716, + "grad_norm": 1.623334809826056, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 43716 + }, + { + "epoch": 0.43717, + "grad_norm": 1.4278073359776624, + "learning_rate": 0.003, + "loss": 4.057, + "step": 43717 + }, + { + "epoch": 0.43718, + "grad_norm": 1.5540117200423058, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 43718 + }, + { + "epoch": 0.43719, + "grad_norm": 1.5731538606822228, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 43719 + }, + { + "epoch": 0.4372, + "grad_norm": 1.7121220271897841, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 43720 + }, + { + "epoch": 0.43721, + "grad_norm": 1.457331375236253, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 43721 + }, + { + "epoch": 0.43722, + "grad_norm": 1.6168186506149902, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 43722 + }, + { + "epoch": 0.43723, + "grad_norm": 1.5710566467212377, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 43723 + }, + { + "epoch": 0.43724, + "grad_norm": 1.4940452697977926, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 43724 + }, + { + "epoch": 0.43725, + "grad_norm": 1.365769533418073, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 43725 + }, + { + "epoch": 0.43726, + "grad_norm": 1.7054176960879883, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 43726 + }, + { + "epoch": 0.43727, + "grad_norm": 1.5168172118283116, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 43727 + }, + { + "epoch": 0.43728, + "grad_norm": 1.5884718321213667, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 43728 + }, + { + "epoch": 0.43729, + "grad_norm": 1.5650643824944086, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 43729 + }, + { + "epoch": 0.4373, + "grad_norm": 1.428051886526076, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 43730 + }, + { + "epoch": 0.43731, + "grad_norm": 1.5477005418378842, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 43731 + }, + { + "epoch": 0.43732, + "grad_norm": 1.3061789286839387, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 43732 + }, + { + "epoch": 0.43733, + "grad_norm": 1.5408270539107496, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 43733 + }, + { + "epoch": 0.43734, + "grad_norm": 1.4412394550341814, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 43734 + }, + { + "epoch": 0.43735, + "grad_norm": 1.6601216545087971, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 43735 + }, + { + "epoch": 0.43736, + "grad_norm": 1.3976795980672327, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 43736 + }, + { + "epoch": 0.43737, + "grad_norm": 1.5578719847633566, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43737 + }, + { + "epoch": 0.43738, + "grad_norm": 1.3595077402026667, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 43738 + }, + { + "epoch": 0.43739, + "grad_norm": 1.848786336897087, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 43739 + }, + { + "epoch": 0.4374, + "grad_norm": 1.1335788286106105, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 43740 + }, + { + "epoch": 0.43741, + "grad_norm": 2.039149926994399, + "learning_rate": 0.003, + "loss": 4.015, + "step": 43741 + }, + { + "epoch": 0.43742, + "grad_norm": 1.0991022527055991, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 43742 + }, + { + "epoch": 0.43743, + "grad_norm": 2.0541317205281375, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 43743 + }, + { + "epoch": 0.43744, + "grad_norm": 1.397007457796485, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 43744 + }, + { + "epoch": 0.43745, + "grad_norm": 1.4776597532261901, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 43745 + }, + { + "epoch": 0.43746, + "grad_norm": 1.9074327122198962, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 43746 + }, + { + "epoch": 0.43747, + "grad_norm": 1.5744410654940928, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 43747 + }, + { + "epoch": 0.43748, + "grad_norm": 1.7525723826373476, + "learning_rate": 0.003, + "loss": 4.08, + "step": 43748 + }, + { + "epoch": 0.43749, + "grad_norm": 1.3574001847965527, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 43749 + }, + { + "epoch": 0.4375, + "grad_norm": 1.8925449747608605, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 43750 + }, + { + "epoch": 0.43751, + "grad_norm": 1.3493633858169523, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43751 + }, + { + "epoch": 0.43752, + "grad_norm": 1.6891532188349576, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 43752 + }, + { + "epoch": 0.43753, + "grad_norm": 1.22682034470732, + "learning_rate": 0.003, + "loss": 4.034, + "step": 43753 + }, + { + "epoch": 0.43754, + "grad_norm": 1.683122034114432, + "learning_rate": 0.003, + "loss": 4.029, + "step": 43754 + }, + { + "epoch": 0.43755, + "grad_norm": 1.8070774848628046, + "learning_rate": 0.003, + "loss": 4.031, + "step": 43755 + }, + { + "epoch": 0.43756, + "grad_norm": 1.42511878060772, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 43756 + }, + { + "epoch": 0.43757, + "grad_norm": 1.4232912383839673, + "learning_rate": 0.003, + "loss": 4.037, + "step": 43757 + }, + { + "epoch": 0.43758, + "grad_norm": 1.6864965865779082, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 43758 + }, + { + "epoch": 0.43759, + "grad_norm": 1.2897910885311616, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 43759 + }, + { + "epoch": 0.4376, + "grad_norm": 1.380078072572213, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 43760 + }, + { + "epoch": 0.43761, + "grad_norm": 1.5893759381190544, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 43761 + }, + { + "epoch": 0.43762, + "grad_norm": 1.371869980090427, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 43762 + }, + { + "epoch": 0.43763, + "grad_norm": 1.5009340952262573, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 43763 + }, + { + "epoch": 0.43764, + "grad_norm": 1.5570427229107837, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43764 + }, + { + "epoch": 0.43765, + "grad_norm": 1.5797826349661654, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 43765 + }, + { + "epoch": 0.43766, + "grad_norm": 1.6893399989590123, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43766 + }, + { + "epoch": 0.43767, + "grad_norm": 1.3079520976439953, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 43767 + }, + { + "epoch": 0.43768, + "grad_norm": 1.8454740462202202, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 43768 + }, + { + "epoch": 0.43769, + "grad_norm": 1.2214799430670806, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 43769 + }, + { + "epoch": 0.4377, + "grad_norm": 1.9500288765788985, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 43770 + }, + { + "epoch": 0.43771, + "grad_norm": 1.4731672656900099, + "learning_rate": 0.003, + "loss": 4.003, + "step": 43771 + }, + { + "epoch": 0.43772, + "grad_norm": 1.824443070571514, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 43772 + }, + { + "epoch": 0.43773, + "grad_norm": 1.340405287098564, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 43773 + }, + { + "epoch": 0.43774, + "grad_norm": 1.6096589011852653, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 43774 + }, + { + "epoch": 0.43775, + "grad_norm": 1.6527019243923189, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 43775 + }, + { + "epoch": 0.43776, + "grad_norm": 1.496381406465129, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 43776 + }, + { + "epoch": 0.43777, + "grad_norm": 1.6148298463968018, + "learning_rate": 0.003, + "loss": 4.041, + "step": 43777 + }, + { + "epoch": 0.43778, + "grad_norm": 1.3419013576307737, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 43778 + }, + { + "epoch": 0.43779, + "grad_norm": 1.5628984489731366, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 43779 + }, + { + "epoch": 0.4378, + "grad_norm": 1.2229278752158221, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 43780 + }, + { + "epoch": 0.43781, + "grad_norm": 1.6933183118372164, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 43781 + }, + { + "epoch": 0.43782, + "grad_norm": 1.2156663924479651, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 43782 + }, + { + "epoch": 0.43783, + "grad_norm": 2.0498589785618697, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 43783 + }, + { + "epoch": 0.43784, + "grad_norm": 1.4958552161783887, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 43784 + }, + { + "epoch": 0.43785, + "grad_norm": 1.6827680850577336, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 43785 + }, + { + "epoch": 0.43786, + "grad_norm": 1.6863053396872232, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 43786 + }, + { + "epoch": 0.43787, + "grad_norm": 1.3431957828911474, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43787 + }, + { + "epoch": 0.43788, + "grad_norm": 1.9049061277800512, + "learning_rate": 0.003, + "loss": 4.038, + "step": 43788 + }, + { + "epoch": 0.43789, + "grad_norm": 1.4494277001012816, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 43789 + }, + { + "epoch": 0.4379, + "grad_norm": 1.3955682710393247, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 43790 + }, + { + "epoch": 0.43791, + "grad_norm": 1.4669426160900498, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43791 + }, + { + "epoch": 0.43792, + "grad_norm": 1.465346723236352, + "learning_rate": 0.003, + "loss": 4.045, + "step": 43792 + }, + { + "epoch": 0.43793, + "grad_norm": 1.4554763499193732, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 43793 + }, + { + "epoch": 0.43794, + "grad_norm": 1.3524817264307014, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 43794 + }, + { + "epoch": 0.43795, + "grad_norm": 1.4879703651824223, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 43795 + }, + { + "epoch": 0.43796, + "grad_norm": 1.3515420724922387, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43796 + }, + { + "epoch": 0.43797, + "grad_norm": 1.9937803843864677, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 43797 + }, + { + "epoch": 0.43798, + "grad_norm": 1.3578993362074385, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 43798 + }, + { + "epoch": 0.43799, + "grad_norm": 1.430867291974336, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 43799 + }, + { + "epoch": 0.438, + "grad_norm": 1.9490981518993302, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43800 + }, + { + "epoch": 0.43801, + "grad_norm": 1.572762256885497, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43801 + }, + { + "epoch": 0.43802, + "grad_norm": 1.6566289002830445, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 43802 + }, + { + "epoch": 0.43803, + "grad_norm": 1.4051663831615597, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 43803 + }, + { + "epoch": 0.43804, + "grad_norm": 1.554026272859859, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 43804 + }, + { + "epoch": 0.43805, + "grad_norm": 1.5510924934338812, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 43805 + }, + { + "epoch": 0.43806, + "grad_norm": 1.6721110116374995, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 43806 + }, + { + "epoch": 0.43807, + "grad_norm": 1.4035418513299562, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 43807 + }, + { + "epoch": 0.43808, + "grad_norm": 1.4335892932019583, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 43808 + }, + { + "epoch": 0.43809, + "grad_norm": 1.398807477687299, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 43809 + }, + { + "epoch": 0.4381, + "grad_norm": 1.7210228263598137, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 43810 + }, + { + "epoch": 0.43811, + "grad_norm": 1.2758435214609125, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 43811 + }, + { + "epoch": 0.43812, + "grad_norm": 1.617579764716922, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 43812 + }, + { + "epoch": 0.43813, + "grad_norm": 1.2482196812392652, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 43813 + }, + { + "epoch": 0.43814, + "grad_norm": 1.5623106377596554, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 43814 + }, + { + "epoch": 0.43815, + "grad_norm": 1.4235110539644173, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 43815 + }, + { + "epoch": 0.43816, + "grad_norm": 1.8170465628464194, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 43816 + }, + { + "epoch": 0.43817, + "grad_norm": 1.7521458628627442, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43817 + }, + { + "epoch": 0.43818, + "grad_norm": 1.5797888721518767, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43818 + }, + { + "epoch": 0.43819, + "grad_norm": 1.6559679109731558, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 43819 + }, + { + "epoch": 0.4382, + "grad_norm": 1.3811111385434263, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 43820 + }, + { + "epoch": 0.43821, + "grad_norm": 1.6870555955880224, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 43821 + }, + { + "epoch": 0.43822, + "grad_norm": 1.440302143734132, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43822 + }, + { + "epoch": 0.43823, + "grad_norm": 1.6625961457246992, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 43823 + }, + { + "epoch": 0.43824, + "grad_norm": 1.5197312615082244, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 43824 + }, + { + "epoch": 0.43825, + "grad_norm": 1.2997662361342892, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 43825 + }, + { + "epoch": 0.43826, + "grad_norm": 1.5779677647021084, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 43826 + }, + { + "epoch": 0.43827, + "grad_norm": 1.4153580809300654, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43827 + }, + { + "epoch": 0.43828, + "grad_norm": 1.4617721192803836, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 43828 + }, + { + "epoch": 0.43829, + "grad_norm": 1.3792913007684005, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 43829 + }, + { + "epoch": 0.4383, + "grad_norm": 1.6340665022005894, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 43830 + }, + { + "epoch": 0.43831, + "grad_norm": 1.5262306160577706, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43831 + }, + { + "epoch": 0.43832, + "grad_norm": 1.4167738314386635, + "learning_rate": 0.003, + "loss": 4.04, + "step": 43832 + }, + { + "epoch": 0.43833, + "grad_norm": 1.660386071885439, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43833 + }, + { + "epoch": 0.43834, + "grad_norm": 1.5759168975111593, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43834 + }, + { + "epoch": 0.43835, + "grad_norm": 1.7393251062193436, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 43835 + }, + { + "epoch": 0.43836, + "grad_norm": 1.4743721007750723, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 43836 + }, + { + "epoch": 0.43837, + "grad_norm": 1.5430353032329593, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 43837 + }, + { + "epoch": 0.43838, + "grad_norm": 1.4985912116935327, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 43838 + }, + { + "epoch": 0.43839, + "grad_norm": 1.4030963190571388, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 43839 + }, + { + "epoch": 0.4384, + "grad_norm": 1.67502502611668, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 43840 + }, + { + "epoch": 0.43841, + "grad_norm": 1.3960738495420317, + "learning_rate": 0.003, + "loss": 4.054, + "step": 43841 + }, + { + "epoch": 0.43842, + "grad_norm": 1.4017888208527012, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 43842 + }, + { + "epoch": 0.43843, + "grad_norm": 1.6477189929310858, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 43843 + }, + { + "epoch": 0.43844, + "grad_norm": 1.3869084093926267, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 43844 + }, + { + "epoch": 0.43845, + "grad_norm": 1.8309130772528261, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 43845 + }, + { + "epoch": 0.43846, + "grad_norm": 1.1827749536536387, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 43846 + }, + { + "epoch": 0.43847, + "grad_norm": 1.9745509033321678, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43847 + }, + { + "epoch": 0.43848, + "grad_norm": 1.4501277733244868, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 43848 + }, + { + "epoch": 0.43849, + "grad_norm": 1.64765641550045, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 43849 + }, + { + "epoch": 0.4385, + "grad_norm": 1.6048930029986708, + "learning_rate": 0.003, + "loss": 4.048, + "step": 43850 + }, + { + "epoch": 0.43851, + "grad_norm": 1.5555860186643935, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 43851 + }, + { + "epoch": 0.43852, + "grad_norm": 1.8025974955057202, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 43852 + }, + { + "epoch": 0.43853, + "grad_norm": 1.2451470388135801, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43853 + }, + { + "epoch": 0.43854, + "grad_norm": 1.4483758670704685, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 43854 + }, + { + "epoch": 0.43855, + "grad_norm": 1.348485094412669, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 43855 + }, + { + "epoch": 0.43856, + "grad_norm": 1.7555725207941408, + "learning_rate": 0.003, + "loss": 4.036, + "step": 43856 + }, + { + "epoch": 0.43857, + "grad_norm": 1.236100045842435, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 43857 + }, + { + "epoch": 0.43858, + "grad_norm": 1.9496952951180075, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 43858 + }, + { + "epoch": 0.43859, + "grad_norm": 1.54628316501303, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 43859 + }, + { + "epoch": 0.4386, + "grad_norm": 1.9226008433252335, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 43860 + }, + { + "epoch": 0.43861, + "grad_norm": 1.3790056805773854, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 43861 + }, + { + "epoch": 0.43862, + "grad_norm": 1.507847142344138, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 43862 + }, + { + "epoch": 0.43863, + "grad_norm": 1.3712067802986738, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 43863 + }, + { + "epoch": 0.43864, + "grad_norm": 1.585217286655112, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 43864 + }, + { + "epoch": 0.43865, + "grad_norm": 1.5206496960330174, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 43865 + }, + { + "epoch": 0.43866, + "grad_norm": 1.5038161530572687, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 43866 + }, + { + "epoch": 0.43867, + "grad_norm": 1.8364553340540597, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 43867 + }, + { + "epoch": 0.43868, + "grad_norm": 1.3751975828879248, + "learning_rate": 0.003, + "loss": 4.033, + "step": 43868 + }, + { + "epoch": 0.43869, + "grad_norm": 1.4720545838371533, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 43869 + }, + { + "epoch": 0.4387, + "grad_norm": 1.6480432333553643, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 43870 + }, + { + "epoch": 0.43871, + "grad_norm": 1.3485365385830372, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 43871 + }, + { + "epoch": 0.43872, + "grad_norm": 1.75659085911108, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 43872 + }, + { + "epoch": 0.43873, + "grad_norm": 1.6294200158548375, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 43873 + }, + { + "epoch": 0.43874, + "grad_norm": 1.4819781246367134, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 43874 + }, + { + "epoch": 0.43875, + "grad_norm": 1.6132061348107662, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 43875 + }, + { + "epoch": 0.43876, + "grad_norm": 1.3709479915027427, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 43876 + }, + { + "epoch": 0.43877, + "grad_norm": 1.5115638405263698, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 43877 + }, + { + "epoch": 0.43878, + "grad_norm": 1.6879181692012495, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43878 + }, + { + "epoch": 0.43879, + "grad_norm": 1.5380459226465717, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 43879 + }, + { + "epoch": 0.4388, + "grad_norm": 1.3275931349343528, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 43880 + }, + { + "epoch": 0.43881, + "grad_norm": 1.827713311399863, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43881 + }, + { + "epoch": 0.43882, + "grad_norm": 1.2016559550109758, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43882 + }, + { + "epoch": 0.43883, + "grad_norm": 1.6536789960805753, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 43883 + }, + { + "epoch": 0.43884, + "grad_norm": 1.3004668763416063, + "learning_rate": 0.003, + "loss": 4.029, + "step": 43884 + }, + { + "epoch": 0.43885, + "grad_norm": 1.9632917058572057, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 43885 + }, + { + "epoch": 0.43886, + "grad_norm": 1.287098456627065, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 43886 + }, + { + "epoch": 0.43887, + "grad_norm": 1.5588405860383492, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 43887 + }, + { + "epoch": 0.43888, + "grad_norm": 1.9522555776439607, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 43888 + }, + { + "epoch": 0.43889, + "grad_norm": 1.1767166204298785, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 43889 + }, + { + "epoch": 0.4389, + "grad_norm": 1.6774875459975631, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 43890 + }, + { + "epoch": 0.43891, + "grad_norm": 1.6445431100169454, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 43891 + }, + { + "epoch": 0.43892, + "grad_norm": 1.64056689065256, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 43892 + }, + { + "epoch": 0.43893, + "grad_norm": 1.5497374731680134, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 43893 + }, + { + "epoch": 0.43894, + "grad_norm": 1.3705429938639635, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 43894 + }, + { + "epoch": 0.43895, + "grad_norm": 1.7214585043622679, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 43895 + }, + { + "epoch": 0.43896, + "grad_norm": 1.6637592406184314, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43896 + }, + { + "epoch": 0.43897, + "grad_norm": 1.525736816646373, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43897 + }, + { + "epoch": 0.43898, + "grad_norm": 1.604994025435856, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 43898 + }, + { + "epoch": 0.43899, + "grad_norm": 1.4839767001603295, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 43899 + }, + { + "epoch": 0.439, + "grad_norm": 1.3368221003070944, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 43900 + }, + { + "epoch": 0.43901, + "grad_norm": 1.5659746786506663, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 43901 + }, + { + "epoch": 0.43902, + "grad_norm": 1.3686838492558584, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43902 + }, + { + "epoch": 0.43903, + "grad_norm": 1.4404005492747594, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 43903 + }, + { + "epoch": 0.43904, + "grad_norm": 1.5767665259698933, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 43904 + }, + { + "epoch": 0.43905, + "grad_norm": 1.3681802199065132, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 43905 + }, + { + "epoch": 0.43906, + "grad_norm": 1.4854048670153053, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 43906 + }, + { + "epoch": 0.43907, + "grad_norm": 1.8013260897533134, + "learning_rate": 0.003, + "loss": 4.039, + "step": 43907 + }, + { + "epoch": 0.43908, + "grad_norm": 1.2935324721278683, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 43908 + }, + { + "epoch": 0.43909, + "grad_norm": 2.1057578945213407, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 43909 + }, + { + "epoch": 0.4391, + "grad_norm": 1.165899853581495, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 43910 + }, + { + "epoch": 0.43911, + "grad_norm": 1.7784854945898094, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 43911 + }, + { + "epoch": 0.43912, + "grad_norm": 1.5196144692086266, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 43912 + }, + { + "epoch": 0.43913, + "grad_norm": 1.6329700668157694, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 43913 + }, + { + "epoch": 0.43914, + "grad_norm": 1.9228700220405623, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 43914 + }, + { + "epoch": 0.43915, + "grad_norm": 1.2262249723319179, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 43915 + }, + { + "epoch": 0.43916, + "grad_norm": 1.5635914646293732, + "learning_rate": 0.003, + "loss": 4.018, + "step": 43916 + }, + { + "epoch": 0.43917, + "grad_norm": 1.6272753505823037, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 43917 + }, + { + "epoch": 0.43918, + "grad_norm": 1.5129733214712668, + "learning_rate": 0.003, + "loss": 4.032, + "step": 43918 + }, + { + "epoch": 0.43919, + "grad_norm": 1.4116929557229756, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 43919 + }, + { + "epoch": 0.4392, + "grad_norm": 1.4588250439053372, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 43920 + }, + { + "epoch": 0.43921, + "grad_norm": 1.5495608839726551, + "learning_rate": 0.003, + "loss": 4.055, + "step": 43921 + }, + { + "epoch": 0.43922, + "grad_norm": 1.3802502369340537, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 43922 + }, + { + "epoch": 0.43923, + "grad_norm": 1.4505981861006922, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 43923 + }, + { + "epoch": 0.43924, + "grad_norm": 1.255385582403879, + "learning_rate": 0.003, + "loss": 4.041, + "step": 43924 + }, + { + "epoch": 0.43925, + "grad_norm": 1.6484538222190803, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 43925 + }, + { + "epoch": 0.43926, + "grad_norm": 1.3715701516044614, + "learning_rate": 0.003, + "loss": 4.049, + "step": 43926 + }, + { + "epoch": 0.43927, + "grad_norm": 1.7828095135123438, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 43927 + }, + { + "epoch": 0.43928, + "grad_norm": 1.3925868798382652, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 43928 + }, + { + "epoch": 0.43929, + "grad_norm": 1.5583770634038443, + "learning_rate": 0.003, + "loss": 4.043, + "step": 43929 + }, + { + "epoch": 0.4393, + "grad_norm": 1.4123312411677786, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 43930 + }, + { + "epoch": 0.43931, + "grad_norm": 1.567677211636236, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 43931 + }, + { + "epoch": 0.43932, + "grad_norm": 1.6727751583200328, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 43932 + }, + { + "epoch": 0.43933, + "grad_norm": 2.0342453094046267, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 43933 + }, + { + "epoch": 0.43934, + "grad_norm": 1.7275667899767189, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 43934 + }, + { + "epoch": 0.43935, + "grad_norm": 1.4691032247166678, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 43935 + }, + { + "epoch": 0.43936, + "grad_norm": 1.7043009833071292, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 43936 + }, + { + "epoch": 0.43937, + "grad_norm": 1.5961724129219645, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 43937 + }, + { + "epoch": 0.43938, + "grad_norm": 1.6869266419670745, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 43938 + }, + { + "epoch": 0.43939, + "grad_norm": 1.156600677159466, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 43939 + }, + { + "epoch": 0.4394, + "grad_norm": 1.5609974273835594, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 43940 + }, + { + "epoch": 0.43941, + "grad_norm": 1.4638302489968615, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 43941 + }, + { + "epoch": 0.43942, + "grad_norm": 1.556297650761005, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 43942 + }, + { + "epoch": 0.43943, + "grad_norm": 1.3953162295225194, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 43943 + }, + { + "epoch": 0.43944, + "grad_norm": 1.566785057841995, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 43944 + }, + { + "epoch": 0.43945, + "grad_norm": 1.5280332573231032, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 43945 + }, + { + "epoch": 0.43946, + "grad_norm": 1.5716105170049073, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 43946 + }, + { + "epoch": 0.43947, + "grad_norm": 1.6044740247406284, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 43947 + }, + { + "epoch": 0.43948, + "grad_norm": 1.60066509836064, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 43948 + }, + { + "epoch": 0.43949, + "grad_norm": 1.4226070124992805, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 43949 + }, + { + "epoch": 0.4395, + "grad_norm": 1.3764223311544859, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 43950 + }, + { + "epoch": 0.43951, + "grad_norm": 1.8038892620677625, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 43951 + }, + { + "epoch": 0.43952, + "grad_norm": 1.4292314169534013, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 43952 + }, + { + "epoch": 0.43953, + "grad_norm": 1.4617260915716148, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43953 + }, + { + "epoch": 0.43954, + "grad_norm": 1.5564349661800612, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 43954 + }, + { + "epoch": 0.43955, + "grad_norm": 1.5999803757131115, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 43955 + }, + { + "epoch": 0.43956, + "grad_norm": 1.332836443277618, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 43956 + }, + { + "epoch": 0.43957, + "grad_norm": 1.5213020488702516, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 43957 + }, + { + "epoch": 0.43958, + "grad_norm": 1.4861133947020857, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43958 + }, + { + "epoch": 0.43959, + "grad_norm": 1.5499971614865935, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 43959 + }, + { + "epoch": 0.4396, + "grad_norm": 1.6511123825050913, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 43960 + }, + { + "epoch": 0.43961, + "grad_norm": 1.3796987159106924, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 43961 + }, + { + "epoch": 0.43962, + "grad_norm": 1.3809120185788664, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 43962 + }, + { + "epoch": 0.43963, + "grad_norm": 1.5439082082724396, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 43963 + }, + { + "epoch": 0.43964, + "grad_norm": 1.96833462481753, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 43964 + }, + { + "epoch": 0.43965, + "grad_norm": 1.5720317674156181, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 43965 + }, + { + "epoch": 0.43966, + "grad_norm": 1.7813684328600128, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 43966 + }, + { + "epoch": 0.43967, + "grad_norm": 1.597921343808223, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 43967 + }, + { + "epoch": 0.43968, + "grad_norm": 1.4992807845338194, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 43968 + }, + { + "epoch": 0.43969, + "grad_norm": 1.396043036076702, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 43969 + }, + { + "epoch": 0.4397, + "grad_norm": 1.5318229007323942, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 43970 + }, + { + "epoch": 0.43971, + "grad_norm": 1.289735892696438, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 43971 + }, + { + "epoch": 0.43972, + "grad_norm": 1.5313136134159269, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 43972 + }, + { + "epoch": 0.43973, + "grad_norm": 1.404819694742181, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 43973 + }, + { + "epoch": 0.43974, + "grad_norm": 1.6070594546593158, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 43974 + }, + { + "epoch": 0.43975, + "grad_norm": 1.4756303316714123, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 43975 + }, + { + "epoch": 0.43976, + "grad_norm": 1.881373717879176, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 43976 + }, + { + "epoch": 0.43977, + "grad_norm": 1.6382338914740475, + "learning_rate": 0.003, + "loss": 4.047, + "step": 43977 + }, + { + "epoch": 0.43978, + "grad_norm": 1.5565180783909671, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 43978 + }, + { + "epoch": 0.43979, + "grad_norm": 1.673650702937547, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43979 + }, + { + "epoch": 0.4398, + "grad_norm": 1.6108011249105776, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 43980 + }, + { + "epoch": 0.43981, + "grad_norm": 1.385412502901085, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 43981 + }, + { + "epoch": 0.43982, + "grad_norm": 1.4971898096884442, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 43982 + }, + { + "epoch": 0.43983, + "grad_norm": 1.3652049131343549, + "learning_rate": 0.003, + "loss": 4.044, + "step": 43983 + }, + { + "epoch": 0.43984, + "grad_norm": 1.5479848685259914, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43984 + }, + { + "epoch": 0.43985, + "grad_norm": 1.3518284246298387, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 43985 + }, + { + "epoch": 0.43986, + "grad_norm": 1.5515883459883335, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 43986 + }, + { + "epoch": 0.43987, + "grad_norm": 1.7111060552207402, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 43987 + }, + { + "epoch": 0.43988, + "grad_norm": 1.46518140695274, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 43988 + }, + { + "epoch": 0.43989, + "grad_norm": 1.8769133971716123, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 43989 + }, + { + "epoch": 0.4399, + "grad_norm": 1.5278724780158215, + "learning_rate": 0.003, + "loss": 4.028, + "step": 43990 + }, + { + "epoch": 0.43991, + "grad_norm": 1.8473109109402555, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 43991 + }, + { + "epoch": 0.43992, + "grad_norm": 1.5836828694745892, + "learning_rate": 0.003, + "loss": 4.021, + "step": 43992 + }, + { + "epoch": 0.43993, + "grad_norm": 1.4679156708429035, + "learning_rate": 0.003, + "loss": 4.028, + "step": 43993 + }, + { + "epoch": 0.43994, + "grad_norm": 1.261531349669501, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 43994 + }, + { + "epoch": 0.43995, + "grad_norm": 1.81185254211551, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 43995 + }, + { + "epoch": 0.43996, + "grad_norm": 1.30121143116946, + "learning_rate": 0.003, + "loss": 4.042, + "step": 43996 + }, + { + "epoch": 0.43997, + "grad_norm": 1.6125375282712353, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 43997 + }, + { + "epoch": 0.43998, + "grad_norm": 1.3410365728313351, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 43998 + }, + { + "epoch": 0.43999, + "grad_norm": 1.7392196110251716, + "learning_rate": 0.003, + "loss": 4.03, + "step": 43999 + }, + { + "epoch": 0.44, + "grad_norm": 1.5648642960946397, + "learning_rate": 0.003, + "loss": 4.052, + "step": 44000 + }, + { + "epoch": 0.44001, + "grad_norm": 1.7044882502830685, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 44001 + }, + { + "epoch": 0.44002, + "grad_norm": 1.3230818182057704, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 44002 + }, + { + "epoch": 0.44003, + "grad_norm": 1.8191156898480063, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 44003 + }, + { + "epoch": 0.44004, + "grad_norm": 1.609061156802599, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 44004 + }, + { + "epoch": 0.44005, + "grad_norm": 1.8438364721629874, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44005 + }, + { + "epoch": 0.44006, + "grad_norm": 1.3687163325381175, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 44006 + }, + { + "epoch": 0.44007, + "grad_norm": 1.5347948869359402, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 44007 + }, + { + "epoch": 0.44008, + "grad_norm": 1.6548567614679146, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 44008 + }, + { + "epoch": 0.44009, + "grad_norm": 1.5322529384132841, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 44009 + }, + { + "epoch": 0.4401, + "grad_norm": 1.4134843587538806, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 44010 + }, + { + "epoch": 0.44011, + "grad_norm": 1.487765853539569, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 44011 + }, + { + "epoch": 0.44012, + "grad_norm": 1.4752330208029973, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 44012 + }, + { + "epoch": 0.44013, + "grad_norm": 1.3665941735313814, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44013 + }, + { + "epoch": 0.44014, + "grad_norm": 1.372723963801556, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44014 + }, + { + "epoch": 0.44015, + "grad_norm": 1.64721039093273, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 44015 + }, + { + "epoch": 0.44016, + "grad_norm": 1.356399677587439, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 44016 + }, + { + "epoch": 0.44017, + "grad_norm": 1.5901479700382284, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44017 + }, + { + "epoch": 0.44018, + "grad_norm": 1.6715888840857827, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 44018 + }, + { + "epoch": 0.44019, + "grad_norm": 1.7178376798914265, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44019 + }, + { + "epoch": 0.4402, + "grad_norm": 1.2362284109993826, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 44020 + }, + { + "epoch": 0.44021, + "grad_norm": 1.754200400523183, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 44021 + }, + { + "epoch": 0.44022, + "grad_norm": 1.5551284923481106, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 44022 + }, + { + "epoch": 0.44023, + "grad_norm": 1.434547705895715, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44023 + }, + { + "epoch": 0.44024, + "grad_norm": 1.6558868954448633, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44024 + }, + { + "epoch": 0.44025, + "grad_norm": 1.4070619669961266, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44025 + }, + { + "epoch": 0.44026, + "grad_norm": 1.6514139637019767, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44026 + }, + { + "epoch": 0.44027, + "grad_norm": 1.280272820768764, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 44027 + }, + { + "epoch": 0.44028, + "grad_norm": 1.8026994933191443, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 44028 + }, + { + "epoch": 0.44029, + "grad_norm": 1.2937511728495685, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44029 + }, + { + "epoch": 0.4403, + "grad_norm": 1.6350701949245827, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 44030 + }, + { + "epoch": 0.44031, + "grad_norm": 1.8431918283767763, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 44031 + }, + { + "epoch": 0.44032, + "grad_norm": 1.3577505914799732, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44032 + }, + { + "epoch": 0.44033, + "grad_norm": 1.8196084731033746, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 44033 + }, + { + "epoch": 0.44034, + "grad_norm": 1.4563905634308811, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44034 + }, + { + "epoch": 0.44035, + "grad_norm": 1.425642231551269, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44035 + }, + { + "epoch": 0.44036, + "grad_norm": 1.6641965945407347, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 44036 + }, + { + "epoch": 0.44037, + "grad_norm": 1.3802679004123832, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 44037 + }, + { + "epoch": 0.44038, + "grad_norm": 1.5042945556831626, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44038 + }, + { + "epoch": 0.44039, + "grad_norm": 1.2176049856623974, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 44039 + }, + { + "epoch": 0.4404, + "grad_norm": 1.6398311369480096, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44040 + }, + { + "epoch": 0.44041, + "grad_norm": 1.6731243714267476, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44041 + }, + { + "epoch": 0.44042, + "grad_norm": 1.5910959210919169, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 44042 + }, + { + "epoch": 0.44043, + "grad_norm": 1.7829869320678808, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44043 + }, + { + "epoch": 0.44044, + "grad_norm": 1.6803254529617349, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 44044 + }, + { + "epoch": 0.44045, + "grad_norm": 1.1114497740947054, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 44045 + }, + { + "epoch": 0.44046, + "grad_norm": 1.929344560158812, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 44046 + }, + { + "epoch": 0.44047, + "grad_norm": 1.3236479654200555, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 44047 + }, + { + "epoch": 0.44048, + "grad_norm": 1.697984449270736, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 44048 + }, + { + "epoch": 0.44049, + "grad_norm": 1.4431105035025504, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 44049 + }, + { + "epoch": 0.4405, + "grad_norm": 1.8751742378174805, + "learning_rate": 0.003, + "loss": 4.054, + "step": 44050 + }, + { + "epoch": 0.44051, + "grad_norm": 1.0893394053946996, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44051 + }, + { + "epoch": 0.44052, + "grad_norm": 1.8928015047724658, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44052 + }, + { + "epoch": 0.44053, + "grad_norm": 1.8096099624304838, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 44053 + }, + { + "epoch": 0.44054, + "grad_norm": 1.2046193970544685, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 44054 + }, + { + "epoch": 0.44055, + "grad_norm": 1.6031845511828071, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 44055 + }, + { + "epoch": 0.44056, + "grad_norm": 1.3385248117736264, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 44056 + }, + { + "epoch": 0.44057, + "grad_norm": 1.5708912588425135, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44057 + }, + { + "epoch": 0.44058, + "grad_norm": 1.6788697010581581, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44058 + }, + { + "epoch": 0.44059, + "grad_norm": 1.4327960779890243, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 44059 + }, + { + "epoch": 0.4406, + "grad_norm": 1.8732384830000917, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 44060 + }, + { + "epoch": 0.44061, + "grad_norm": 1.8752470915920725, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 44061 + }, + { + "epoch": 0.44062, + "grad_norm": 1.121749132851671, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 44062 + }, + { + "epoch": 0.44063, + "grad_norm": 1.8578090767689708, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 44063 + }, + { + "epoch": 0.44064, + "grad_norm": 1.5050833126574532, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 44064 + }, + { + "epoch": 0.44065, + "grad_norm": 1.695134935120066, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44065 + }, + { + "epoch": 0.44066, + "grad_norm": 1.227932701004856, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 44066 + }, + { + "epoch": 0.44067, + "grad_norm": 1.662798325911603, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 44067 + }, + { + "epoch": 0.44068, + "grad_norm": 1.619376333501473, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 44068 + }, + { + "epoch": 0.44069, + "grad_norm": 1.396863817634558, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 44069 + }, + { + "epoch": 0.4407, + "grad_norm": 1.7429288458202299, + "learning_rate": 0.003, + "loss": 4.034, + "step": 44070 + }, + { + "epoch": 0.44071, + "grad_norm": 1.3209842462300023, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 44071 + }, + { + "epoch": 0.44072, + "grad_norm": 1.6973207657476435, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44072 + }, + { + "epoch": 0.44073, + "grad_norm": 1.4306527241230353, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 44073 + }, + { + "epoch": 0.44074, + "grad_norm": 1.5814822259683032, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 44074 + }, + { + "epoch": 0.44075, + "grad_norm": 1.2007708028086324, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 44075 + }, + { + "epoch": 0.44076, + "grad_norm": 1.7068613485517348, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 44076 + }, + { + "epoch": 0.44077, + "grad_norm": 1.5566149098303825, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 44077 + }, + { + "epoch": 0.44078, + "grad_norm": 1.3397245091573955, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 44078 + }, + { + "epoch": 0.44079, + "grad_norm": 1.700546494340298, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44079 + }, + { + "epoch": 0.4408, + "grad_norm": 1.285906610262858, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 44080 + }, + { + "epoch": 0.44081, + "grad_norm": 1.5051885600337687, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 44081 + }, + { + "epoch": 0.44082, + "grad_norm": 1.5452240009131315, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44082 + }, + { + "epoch": 0.44083, + "grad_norm": 1.4151238008693179, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44083 + }, + { + "epoch": 0.44084, + "grad_norm": 2.5300299534394592, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 44084 + }, + { + "epoch": 0.44085, + "grad_norm": 1.6871375464226288, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 44085 + }, + { + "epoch": 0.44086, + "grad_norm": 1.5383405413314413, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44086 + }, + { + "epoch": 0.44087, + "grad_norm": 1.5045434739837151, + "learning_rate": 0.003, + "loss": 4.069, + "step": 44087 + }, + { + "epoch": 0.44088, + "grad_norm": 1.2989536477956525, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44088 + }, + { + "epoch": 0.44089, + "grad_norm": 1.747133413301221, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44089 + }, + { + "epoch": 0.4409, + "grad_norm": 1.4642134898443246, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44090 + }, + { + "epoch": 0.44091, + "grad_norm": 1.7663339237195332, + "learning_rate": 0.003, + "loss": 4.088, + "step": 44091 + }, + { + "epoch": 0.44092, + "grad_norm": 1.501115308206439, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44092 + }, + { + "epoch": 0.44093, + "grad_norm": 1.6054053943839657, + "learning_rate": 0.003, + "loss": 4.024, + "step": 44093 + }, + { + "epoch": 0.44094, + "grad_norm": 1.2761865228162332, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44094 + }, + { + "epoch": 0.44095, + "grad_norm": 1.5063543302954427, + "learning_rate": 0.003, + "loss": 4.054, + "step": 44095 + }, + { + "epoch": 0.44096, + "grad_norm": 1.3133055477034188, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44096 + }, + { + "epoch": 0.44097, + "grad_norm": 1.8437091097404326, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 44097 + }, + { + "epoch": 0.44098, + "grad_norm": 1.3961503098451022, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44098 + }, + { + "epoch": 0.44099, + "grad_norm": 1.7744892080429, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 44099 + }, + { + "epoch": 0.441, + "grad_norm": 1.3590921177961734, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 44100 + }, + { + "epoch": 0.44101, + "grad_norm": 1.6689368224471666, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 44101 + }, + { + "epoch": 0.44102, + "grad_norm": 1.2813994425302184, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 44102 + }, + { + "epoch": 0.44103, + "grad_norm": 1.697951475181288, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 44103 + }, + { + "epoch": 0.44104, + "grad_norm": 1.4983501558060932, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 44104 + }, + { + "epoch": 0.44105, + "grad_norm": 1.6562140646990198, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 44105 + }, + { + "epoch": 0.44106, + "grad_norm": 1.6471301408770127, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 44106 + }, + { + "epoch": 0.44107, + "grad_norm": 1.387968184824774, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 44107 + }, + { + "epoch": 0.44108, + "grad_norm": 1.426275238787068, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 44108 + }, + { + "epoch": 0.44109, + "grad_norm": 1.902537100026212, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 44109 + }, + { + "epoch": 0.4411, + "grad_norm": 1.3222512766074979, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 44110 + }, + { + "epoch": 0.44111, + "grad_norm": 1.6275927918996893, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44111 + }, + { + "epoch": 0.44112, + "grad_norm": 1.3914363416239857, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 44112 + }, + { + "epoch": 0.44113, + "grad_norm": 1.7295156228169497, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 44113 + }, + { + "epoch": 0.44114, + "grad_norm": 1.3422430832411163, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 44114 + }, + { + "epoch": 0.44115, + "grad_norm": 1.5240122834460743, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 44115 + }, + { + "epoch": 0.44116, + "grad_norm": 1.5455627075845497, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44116 + }, + { + "epoch": 0.44117, + "grad_norm": 1.7533875492751296, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 44117 + }, + { + "epoch": 0.44118, + "grad_norm": 1.4052341173015694, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 44118 + }, + { + "epoch": 0.44119, + "grad_norm": 1.504586805158427, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 44119 + }, + { + "epoch": 0.4412, + "grad_norm": 1.4545480318590602, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 44120 + }, + { + "epoch": 0.44121, + "grad_norm": 1.4487848955505889, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 44121 + }, + { + "epoch": 0.44122, + "grad_norm": 1.382881277294752, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 44122 + }, + { + "epoch": 0.44123, + "grad_norm": 1.7058257247681117, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 44123 + }, + { + "epoch": 0.44124, + "grad_norm": 1.2669786243419932, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 44124 + }, + { + "epoch": 0.44125, + "grad_norm": 1.8569336709823607, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44125 + }, + { + "epoch": 0.44126, + "grad_norm": 1.3872937762346897, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 44126 + }, + { + "epoch": 0.44127, + "grad_norm": 1.6534119523842352, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 44127 + }, + { + "epoch": 0.44128, + "grad_norm": 1.8703302316105404, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44128 + }, + { + "epoch": 0.44129, + "grad_norm": 1.57897106557851, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 44129 + }, + { + "epoch": 0.4413, + "grad_norm": 1.4195254739529986, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44130 + }, + { + "epoch": 0.44131, + "grad_norm": 1.4738780058404324, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 44131 + }, + { + "epoch": 0.44132, + "grad_norm": 1.4387687877697104, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44132 + }, + { + "epoch": 0.44133, + "grad_norm": 1.3550669696282063, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44133 + }, + { + "epoch": 0.44134, + "grad_norm": 1.6437381807601132, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44134 + }, + { + "epoch": 0.44135, + "grad_norm": 1.3478148222603608, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 44135 + }, + { + "epoch": 0.44136, + "grad_norm": 1.5534620857913761, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44136 + }, + { + "epoch": 0.44137, + "grad_norm": 1.564728291588565, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 44137 + }, + { + "epoch": 0.44138, + "grad_norm": 1.6146553484691404, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 44138 + }, + { + "epoch": 0.44139, + "grad_norm": 1.752485577456658, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 44139 + }, + { + "epoch": 0.4414, + "grad_norm": 1.569927351492724, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44140 + }, + { + "epoch": 0.44141, + "grad_norm": 1.78963029915545, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 44141 + }, + { + "epoch": 0.44142, + "grad_norm": 1.6062986935321912, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 44142 + }, + { + "epoch": 0.44143, + "grad_norm": 1.5658104639539199, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 44143 + }, + { + "epoch": 0.44144, + "grad_norm": 1.4423283229896935, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44144 + }, + { + "epoch": 0.44145, + "grad_norm": 1.2181089502428202, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 44145 + }, + { + "epoch": 0.44146, + "grad_norm": 1.754779050039515, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44146 + }, + { + "epoch": 0.44147, + "grad_norm": 1.1060994350894473, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 44147 + }, + { + "epoch": 0.44148, + "grad_norm": 1.9121928824013599, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 44148 + }, + { + "epoch": 0.44149, + "grad_norm": 1.109750551817656, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44149 + }, + { + "epoch": 0.4415, + "grad_norm": 1.966657947143526, + "learning_rate": 0.003, + "loss": 4.054, + "step": 44150 + }, + { + "epoch": 0.44151, + "grad_norm": 1.2809650275075661, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 44151 + }, + { + "epoch": 0.44152, + "grad_norm": 1.8131355597879921, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 44152 + }, + { + "epoch": 0.44153, + "grad_norm": 1.8629016346936165, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 44153 + }, + { + "epoch": 0.44154, + "grad_norm": 1.3204035928386522, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 44154 + }, + { + "epoch": 0.44155, + "grad_norm": 1.6176656750093088, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 44155 + }, + { + "epoch": 0.44156, + "grad_norm": 1.4094831214064047, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 44156 + }, + { + "epoch": 0.44157, + "grad_norm": 1.6425010229317614, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 44157 + }, + { + "epoch": 0.44158, + "grad_norm": 1.4873587642641193, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44158 + }, + { + "epoch": 0.44159, + "grad_norm": 1.5598879472997775, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 44159 + }, + { + "epoch": 0.4416, + "grad_norm": 1.545620960229535, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 44160 + }, + { + "epoch": 0.44161, + "grad_norm": 1.5982438214227428, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 44161 + }, + { + "epoch": 0.44162, + "grad_norm": 1.5576648611439485, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 44162 + }, + { + "epoch": 0.44163, + "grad_norm": 1.6805573908136877, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 44163 + }, + { + "epoch": 0.44164, + "grad_norm": 1.510120457657537, + "learning_rate": 0.003, + "loss": 4.039, + "step": 44164 + }, + { + "epoch": 0.44165, + "grad_norm": 1.3342450149073297, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 44165 + }, + { + "epoch": 0.44166, + "grad_norm": 1.634269370651213, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44166 + }, + { + "epoch": 0.44167, + "grad_norm": 1.337924717438283, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 44167 + }, + { + "epoch": 0.44168, + "grad_norm": 1.7448995097612054, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 44168 + }, + { + "epoch": 0.44169, + "grad_norm": 1.335043720448673, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 44169 + }, + { + "epoch": 0.4417, + "grad_norm": 1.5850627652877174, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44170 + }, + { + "epoch": 0.44171, + "grad_norm": 1.5171457279857627, + "learning_rate": 0.003, + "loss": 4.047, + "step": 44171 + }, + { + "epoch": 0.44172, + "grad_norm": 1.6028555634297574, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 44172 + }, + { + "epoch": 0.44173, + "grad_norm": 1.6696468838412084, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 44173 + }, + { + "epoch": 0.44174, + "grad_norm": 1.4210368157477165, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 44174 + }, + { + "epoch": 0.44175, + "grad_norm": 1.5204652158395486, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 44175 + }, + { + "epoch": 0.44176, + "grad_norm": 1.4275926475776088, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 44176 + }, + { + "epoch": 0.44177, + "grad_norm": 1.2510052674142904, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 44177 + }, + { + "epoch": 0.44178, + "grad_norm": 1.915484182316528, + "learning_rate": 0.003, + "loss": 4.011, + "step": 44178 + }, + { + "epoch": 0.44179, + "grad_norm": 1.5091275536664885, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44179 + }, + { + "epoch": 0.4418, + "grad_norm": 1.7928693448294737, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 44180 + }, + { + "epoch": 0.44181, + "grad_norm": 1.6612388142088077, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44181 + }, + { + "epoch": 0.44182, + "grad_norm": 1.5889767340918737, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 44182 + }, + { + "epoch": 0.44183, + "grad_norm": 1.403197704546103, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44183 + }, + { + "epoch": 0.44184, + "grad_norm": 1.3959974188663662, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 44184 + }, + { + "epoch": 0.44185, + "grad_norm": 1.6987229237024941, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44185 + }, + { + "epoch": 0.44186, + "grad_norm": 1.3672883504340523, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44186 + }, + { + "epoch": 0.44187, + "grad_norm": 1.5286173970590613, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 44187 + }, + { + "epoch": 0.44188, + "grad_norm": 1.6447418156267781, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 44188 + }, + { + "epoch": 0.44189, + "grad_norm": 1.3502038642757903, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 44189 + }, + { + "epoch": 0.4419, + "grad_norm": 1.4430255035697277, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 44190 + }, + { + "epoch": 0.44191, + "grad_norm": 1.5302737927010304, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 44191 + }, + { + "epoch": 0.44192, + "grad_norm": 1.7301046704635412, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 44192 + }, + { + "epoch": 0.44193, + "grad_norm": 1.219569787175654, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 44193 + }, + { + "epoch": 0.44194, + "grad_norm": 1.6699488331671288, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 44194 + }, + { + "epoch": 0.44195, + "grad_norm": 1.4899191182992502, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44195 + }, + { + "epoch": 0.44196, + "grad_norm": 1.5500176452136016, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 44196 + }, + { + "epoch": 0.44197, + "grad_norm": 1.334910542203066, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 44197 + }, + { + "epoch": 0.44198, + "grad_norm": 1.7200555113276028, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 44198 + }, + { + "epoch": 0.44199, + "grad_norm": 1.4586162996559027, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 44199 + }, + { + "epoch": 0.442, + "grad_norm": 1.494253170497588, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44200 + }, + { + "epoch": 0.44201, + "grad_norm": 1.4364436616986864, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44201 + }, + { + "epoch": 0.44202, + "grad_norm": 1.5425170892896918, + "learning_rate": 0.003, + "loss": 4.04, + "step": 44202 + }, + { + "epoch": 0.44203, + "grad_norm": 1.5421313575415732, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 44203 + }, + { + "epoch": 0.44204, + "grad_norm": 1.2378988929195367, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 44204 + }, + { + "epoch": 0.44205, + "grad_norm": 1.8088156061139256, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 44205 + }, + { + "epoch": 0.44206, + "grad_norm": 1.4440289808697013, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 44206 + }, + { + "epoch": 0.44207, + "grad_norm": 2.0422303108566373, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 44207 + }, + { + "epoch": 0.44208, + "grad_norm": 1.0174791049428324, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44208 + }, + { + "epoch": 0.44209, + "grad_norm": 1.9228635270995373, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 44209 + }, + { + "epoch": 0.4421, + "grad_norm": 1.340024734826501, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 44210 + }, + { + "epoch": 0.44211, + "grad_norm": 1.6897050663606354, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 44211 + }, + { + "epoch": 0.44212, + "grad_norm": 1.7696876899573413, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44212 + }, + { + "epoch": 0.44213, + "grad_norm": 1.519466001560456, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 44213 + }, + { + "epoch": 0.44214, + "grad_norm": 1.6871968365201837, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44214 + }, + { + "epoch": 0.44215, + "grad_norm": 1.587361074886343, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44215 + }, + { + "epoch": 0.44216, + "grad_norm": 1.3823230523327374, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 44216 + }, + { + "epoch": 0.44217, + "grad_norm": 1.4586354931748475, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 44217 + }, + { + "epoch": 0.44218, + "grad_norm": 1.7474467595115375, + "learning_rate": 0.003, + "loss": 4.065, + "step": 44218 + }, + { + "epoch": 0.44219, + "grad_norm": 1.2840607671939988, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44219 + }, + { + "epoch": 0.4422, + "grad_norm": 1.6984852357364433, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44220 + }, + { + "epoch": 0.44221, + "grad_norm": 1.3078910175687664, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44221 + }, + { + "epoch": 0.44222, + "grad_norm": 1.7992338171658573, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 44222 + }, + { + "epoch": 0.44223, + "grad_norm": 1.4807565007423673, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 44223 + }, + { + "epoch": 0.44224, + "grad_norm": 1.7330337659511195, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 44224 + }, + { + "epoch": 0.44225, + "grad_norm": 1.6307264013697118, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 44225 + }, + { + "epoch": 0.44226, + "grad_norm": 1.5807576139985549, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 44226 + }, + { + "epoch": 0.44227, + "grad_norm": 1.5755601068274494, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 44227 + }, + { + "epoch": 0.44228, + "grad_norm": 1.2597774674464939, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 44228 + }, + { + "epoch": 0.44229, + "grad_norm": 1.6636282418825765, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 44229 + }, + { + "epoch": 0.4423, + "grad_norm": 1.108893540167111, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 44230 + }, + { + "epoch": 0.44231, + "grad_norm": 1.9817671385457287, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 44231 + }, + { + "epoch": 0.44232, + "grad_norm": 1.4667876235752981, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 44232 + }, + { + "epoch": 0.44233, + "grad_norm": 1.4713594045164737, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44233 + }, + { + "epoch": 0.44234, + "grad_norm": 1.3863275553602898, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44234 + }, + { + "epoch": 0.44235, + "grad_norm": 1.7008330881067049, + "learning_rate": 0.003, + "loss": 4.07, + "step": 44235 + }, + { + "epoch": 0.44236, + "grad_norm": 1.3904647881157173, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44236 + }, + { + "epoch": 0.44237, + "grad_norm": 2.150431055558988, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 44237 + }, + { + "epoch": 0.44238, + "grad_norm": 1.2862650370747517, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 44238 + }, + { + "epoch": 0.44239, + "grad_norm": 1.6006125270912184, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 44239 + }, + { + "epoch": 0.4424, + "grad_norm": 1.8216344260366977, + "learning_rate": 0.003, + "loss": 4.062, + "step": 44240 + }, + { + "epoch": 0.44241, + "grad_norm": 1.531938214128822, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44241 + }, + { + "epoch": 0.44242, + "grad_norm": 1.6206949557088959, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 44242 + }, + { + "epoch": 0.44243, + "grad_norm": 1.7643086543794722, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 44243 + }, + { + "epoch": 0.44244, + "grad_norm": 1.5672398311361528, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 44244 + }, + { + "epoch": 0.44245, + "grad_norm": 1.2575795154826392, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 44245 + }, + { + "epoch": 0.44246, + "grad_norm": 1.8948253138012185, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44246 + }, + { + "epoch": 0.44247, + "grad_norm": 1.4713623127415627, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 44247 + }, + { + "epoch": 0.44248, + "grad_norm": 1.7232216473018953, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44248 + }, + { + "epoch": 0.44249, + "grad_norm": 1.364700635037782, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44249 + }, + { + "epoch": 0.4425, + "grad_norm": 1.6027010289832198, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44250 + }, + { + "epoch": 0.44251, + "grad_norm": 1.3450536770907873, + "learning_rate": 0.003, + "loss": 4.007, + "step": 44251 + }, + { + "epoch": 0.44252, + "grad_norm": 1.421680263966293, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44252 + }, + { + "epoch": 0.44253, + "grad_norm": 1.5725977267387143, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 44253 + }, + { + "epoch": 0.44254, + "grad_norm": 1.7757933052291757, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44254 + }, + { + "epoch": 0.44255, + "grad_norm": 1.0817324564525677, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 44255 + }, + { + "epoch": 0.44256, + "grad_norm": 1.913331095716244, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44256 + }, + { + "epoch": 0.44257, + "grad_norm": 1.2959068989236981, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 44257 + }, + { + "epoch": 0.44258, + "grad_norm": 1.8303411305822606, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 44258 + }, + { + "epoch": 0.44259, + "grad_norm": 1.320227568388163, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 44259 + }, + { + "epoch": 0.4426, + "grad_norm": 1.59195332746725, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 44260 + }, + { + "epoch": 0.44261, + "grad_norm": 1.709318546874493, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 44261 + }, + { + "epoch": 0.44262, + "grad_norm": 1.4020113185915264, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44262 + }, + { + "epoch": 0.44263, + "grad_norm": 1.725012285123423, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 44263 + }, + { + "epoch": 0.44264, + "grad_norm": 1.449943456751571, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 44264 + }, + { + "epoch": 0.44265, + "grad_norm": 1.3333766740143451, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 44265 + }, + { + "epoch": 0.44266, + "grad_norm": 1.69224995773277, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44266 + }, + { + "epoch": 0.44267, + "grad_norm": 1.4517621364710966, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 44267 + }, + { + "epoch": 0.44268, + "grad_norm": 1.554189515670281, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44268 + }, + { + "epoch": 0.44269, + "grad_norm": 1.5462502689703606, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 44269 + }, + { + "epoch": 0.4427, + "grad_norm": 1.478882571658463, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 44270 + }, + { + "epoch": 0.44271, + "grad_norm": 1.2325116714365767, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 44271 + }, + { + "epoch": 0.44272, + "grad_norm": 1.6052397659660773, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44272 + }, + { + "epoch": 0.44273, + "grad_norm": 1.3071775429321693, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 44273 + }, + { + "epoch": 0.44274, + "grad_norm": 1.6639401604695012, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44274 + }, + { + "epoch": 0.44275, + "grad_norm": 1.2911190264713073, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 44275 + }, + { + "epoch": 0.44276, + "grad_norm": 1.6822260561389906, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 44276 + }, + { + "epoch": 0.44277, + "grad_norm": 1.6247172875836804, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 44277 + }, + { + "epoch": 0.44278, + "grad_norm": 2.0065713431471326, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44278 + }, + { + "epoch": 0.44279, + "grad_norm": 1.3809412384178765, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 44279 + }, + { + "epoch": 0.4428, + "grad_norm": 1.8364300797078905, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 44280 + }, + { + "epoch": 0.44281, + "grad_norm": 1.1786568697286304, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 44281 + }, + { + "epoch": 0.44282, + "grad_norm": 1.6265417017111121, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 44282 + }, + { + "epoch": 0.44283, + "grad_norm": 1.4905982207815527, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 44283 + }, + { + "epoch": 0.44284, + "grad_norm": 1.5503506059413457, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 44284 + }, + { + "epoch": 0.44285, + "grad_norm": 1.5989354847872776, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 44285 + }, + { + "epoch": 0.44286, + "grad_norm": 1.4510909732192077, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 44286 + }, + { + "epoch": 0.44287, + "grad_norm": 1.5700153767109957, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 44287 + }, + { + "epoch": 0.44288, + "grad_norm": 1.4536091516219352, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 44288 + }, + { + "epoch": 0.44289, + "grad_norm": 1.5505581860249256, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44289 + }, + { + "epoch": 0.4429, + "grad_norm": 1.3431032477902383, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 44290 + }, + { + "epoch": 0.44291, + "grad_norm": 1.4696791318410918, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 44291 + }, + { + "epoch": 0.44292, + "grad_norm": 1.8509929473669517, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 44292 + }, + { + "epoch": 0.44293, + "grad_norm": 1.2864652906101974, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 44293 + }, + { + "epoch": 0.44294, + "grad_norm": 1.6842926974469512, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 44294 + }, + { + "epoch": 0.44295, + "grad_norm": 1.440956071398466, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 44295 + }, + { + "epoch": 0.44296, + "grad_norm": 1.561375733074418, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 44296 + }, + { + "epoch": 0.44297, + "grad_norm": 1.6481471339777196, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 44297 + }, + { + "epoch": 0.44298, + "grad_norm": 1.637981173371049, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44298 + }, + { + "epoch": 0.44299, + "grad_norm": 1.608022542422629, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 44299 + }, + { + "epoch": 0.443, + "grad_norm": 1.536467724297639, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 44300 + }, + { + "epoch": 0.44301, + "grad_norm": 1.658330866258377, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 44301 + }, + { + "epoch": 0.44302, + "grad_norm": 1.423609586055798, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 44302 + }, + { + "epoch": 0.44303, + "grad_norm": 1.632459016614295, + "learning_rate": 0.003, + "loss": 4.051, + "step": 44303 + }, + { + "epoch": 0.44304, + "grad_norm": 1.593344502223283, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44304 + }, + { + "epoch": 0.44305, + "grad_norm": 1.2791198085044402, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44305 + }, + { + "epoch": 0.44306, + "grad_norm": 1.8447986305088309, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 44306 + }, + { + "epoch": 0.44307, + "grad_norm": 1.4291488278198559, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44307 + }, + { + "epoch": 0.44308, + "grad_norm": 1.7461225409818921, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 44308 + }, + { + "epoch": 0.44309, + "grad_norm": 1.4639727360226362, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 44309 + }, + { + "epoch": 0.4431, + "grad_norm": 1.6596305830642073, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 44310 + }, + { + "epoch": 0.44311, + "grad_norm": 1.4282760570031678, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44311 + }, + { + "epoch": 0.44312, + "grad_norm": 1.6188928444209736, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 44312 + }, + { + "epoch": 0.44313, + "grad_norm": 1.3852831963360466, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 44313 + }, + { + "epoch": 0.44314, + "grad_norm": 1.6028097887535462, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44314 + }, + { + "epoch": 0.44315, + "grad_norm": 1.4302889603404971, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 44315 + }, + { + "epoch": 0.44316, + "grad_norm": 1.4689110043550055, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 44316 + }, + { + "epoch": 0.44317, + "grad_norm": 1.602914247481703, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 44317 + }, + { + "epoch": 0.44318, + "grad_norm": 1.4852320201395472, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 44318 + }, + { + "epoch": 0.44319, + "grad_norm": 1.3256412194242277, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44319 + }, + { + "epoch": 0.4432, + "grad_norm": 1.5504458998222332, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44320 + }, + { + "epoch": 0.44321, + "grad_norm": 1.5230084681291336, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44321 + }, + { + "epoch": 0.44322, + "grad_norm": 1.7201725735662254, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 44322 + }, + { + "epoch": 0.44323, + "grad_norm": 1.3637255973499198, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44323 + }, + { + "epoch": 0.44324, + "grad_norm": 1.6196357955533347, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 44324 + }, + { + "epoch": 0.44325, + "grad_norm": 1.5724786398857167, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 44325 + }, + { + "epoch": 0.44326, + "grad_norm": 1.592994645648659, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44326 + }, + { + "epoch": 0.44327, + "grad_norm": 1.7245523914287864, + "learning_rate": 0.003, + "loss": 4.059, + "step": 44327 + }, + { + "epoch": 0.44328, + "grad_norm": 1.7599883484349164, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 44328 + }, + { + "epoch": 0.44329, + "grad_norm": 1.4543453724907218, + "learning_rate": 0.003, + "loss": 4.05, + "step": 44329 + }, + { + "epoch": 0.4433, + "grad_norm": 1.5158641984270427, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 44330 + }, + { + "epoch": 0.44331, + "grad_norm": 1.7111236397640752, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44331 + }, + { + "epoch": 0.44332, + "grad_norm": 1.7829845942958549, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44332 + }, + { + "epoch": 0.44333, + "grad_norm": 1.1779556890863154, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44333 + }, + { + "epoch": 0.44334, + "grad_norm": 1.8815390965861118, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 44334 + }, + { + "epoch": 0.44335, + "grad_norm": 1.2840861428817152, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44335 + }, + { + "epoch": 0.44336, + "grad_norm": 1.6352307694124162, + "learning_rate": 0.003, + "loss": 4.035, + "step": 44336 + }, + { + "epoch": 0.44337, + "grad_norm": 1.2897451934452835, + "learning_rate": 0.003, + "loss": 4.038, + "step": 44337 + }, + { + "epoch": 0.44338, + "grad_norm": 1.7408714209340574, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 44338 + }, + { + "epoch": 0.44339, + "grad_norm": 1.312626394596343, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 44339 + }, + { + "epoch": 0.4434, + "grad_norm": 1.6662017953555677, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44340 + }, + { + "epoch": 0.44341, + "grad_norm": 1.4573037566499274, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 44341 + }, + { + "epoch": 0.44342, + "grad_norm": 1.378792172236714, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44342 + }, + { + "epoch": 0.44343, + "grad_norm": 1.421677554362959, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 44343 + }, + { + "epoch": 0.44344, + "grad_norm": 1.5987123974454982, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 44344 + }, + { + "epoch": 0.44345, + "grad_norm": 1.499081157114091, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44345 + }, + { + "epoch": 0.44346, + "grad_norm": 1.4957658011918396, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 44346 + }, + { + "epoch": 0.44347, + "grad_norm": 1.714988795860525, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 44347 + }, + { + "epoch": 0.44348, + "grad_norm": 1.4485172723971835, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44348 + }, + { + "epoch": 0.44349, + "grad_norm": 1.7123493172852124, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 44349 + }, + { + "epoch": 0.4435, + "grad_norm": 1.730442293808094, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 44350 + }, + { + "epoch": 0.44351, + "grad_norm": 2.1286595320822754, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 44351 + }, + { + "epoch": 0.44352, + "grad_norm": 1.4402694919726453, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 44352 + }, + { + "epoch": 0.44353, + "grad_norm": 1.6123529792683193, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44353 + }, + { + "epoch": 0.44354, + "grad_norm": 1.7182308607745793, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 44354 + }, + { + "epoch": 0.44355, + "grad_norm": 1.3024079205541195, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 44355 + }, + { + "epoch": 0.44356, + "grad_norm": 1.7901206793212698, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44356 + }, + { + "epoch": 0.44357, + "grad_norm": 1.4910112417002706, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 44357 + }, + { + "epoch": 0.44358, + "grad_norm": 1.3742996462672985, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44358 + }, + { + "epoch": 0.44359, + "grad_norm": 1.3541649648109033, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 44359 + }, + { + "epoch": 0.4436, + "grad_norm": 1.727078481908645, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 44360 + }, + { + "epoch": 0.44361, + "grad_norm": 1.2003127221679324, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 44361 + }, + { + "epoch": 0.44362, + "grad_norm": 1.5461179859371597, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 44362 + }, + { + "epoch": 0.44363, + "grad_norm": 1.4566391346138394, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 44363 + }, + { + "epoch": 0.44364, + "grad_norm": 1.7464490751794015, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 44364 + }, + { + "epoch": 0.44365, + "grad_norm": 1.605230259263547, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 44365 + }, + { + "epoch": 0.44366, + "grad_norm": 1.6890089012302938, + "learning_rate": 0.003, + "loss": 4.098, + "step": 44366 + }, + { + "epoch": 0.44367, + "grad_norm": 1.6261553845494625, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 44367 + }, + { + "epoch": 0.44368, + "grad_norm": 1.6951165375283996, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44368 + }, + { + "epoch": 0.44369, + "grad_norm": 1.2468116067579484, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 44369 + }, + { + "epoch": 0.4437, + "grad_norm": 1.8901026274000556, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 44370 + }, + { + "epoch": 0.44371, + "grad_norm": 1.2331095141786024, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 44371 + }, + { + "epoch": 0.44372, + "grad_norm": 1.544131058592881, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 44372 + }, + { + "epoch": 0.44373, + "grad_norm": 1.5627513996476852, + "learning_rate": 0.003, + "loss": 4.048, + "step": 44373 + }, + { + "epoch": 0.44374, + "grad_norm": 1.3015962316128393, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44374 + }, + { + "epoch": 0.44375, + "grad_norm": 1.4287500444553805, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44375 + }, + { + "epoch": 0.44376, + "grad_norm": 1.4928969970071062, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 44376 + }, + { + "epoch": 0.44377, + "grad_norm": 1.45944873660154, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44377 + }, + { + "epoch": 0.44378, + "grad_norm": 1.5146275635972535, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 44378 + }, + { + "epoch": 0.44379, + "grad_norm": 1.519068638622139, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 44379 + }, + { + "epoch": 0.4438, + "grad_norm": 1.3816439115883339, + "learning_rate": 0.003, + "loss": 4.024, + "step": 44380 + }, + { + "epoch": 0.44381, + "grad_norm": 1.465180574167026, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44381 + }, + { + "epoch": 0.44382, + "grad_norm": 1.4481114242554833, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 44382 + }, + { + "epoch": 0.44383, + "grad_norm": 1.62089389348253, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 44383 + }, + { + "epoch": 0.44384, + "grad_norm": 1.4809893882657155, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 44384 + }, + { + "epoch": 0.44385, + "grad_norm": 1.5677366948068252, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44385 + }, + { + "epoch": 0.44386, + "grad_norm": 1.536698022774581, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 44386 + }, + { + "epoch": 0.44387, + "grad_norm": 1.6185652359023583, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 44387 + }, + { + "epoch": 0.44388, + "grad_norm": 1.3353940725111304, + "learning_rate": 0.003, + "loss": 4.035, + "step": 44388 + }, + { + "epoch": 0.44389, + "grad_norm": 1.9947797646083438, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44389 + }, + { + "epoch": 0.4439, + "grad_norm": 1.3694581940495432, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 44390 + }, + { + "epoch": 0.44391, + "grad_norm": 1.4961166015466636, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 44391 + }, + { + "epoch": 0.44392, + "grad_norm": 1.4669233694637693, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 44392 + }, + { + "epoch": 0.44393, + "grad_norm": 1.4348203489859768, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44393 + }, + { + "epoch": 0.44394, + "grad_norm": 1.8165219117899427, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 44394 + }, + { + "epoch": 0.44395, + "grad_norm": 1.4817221782187624, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44395 + }, + { + "epoch": 0.44396, + "grad_norm": 1.589417436784586, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 44396 + }, + { + "epoch": 0.44397, + "grad_norm": 1.563640119259455, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 44397 + }, + { + "epoch": 0.44398, + "grad_norm": 1.3976416637453615, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44398 + }, + { + "epoch": 0.44399, + "grad_norm": 1.3451014514780217, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 44399 + }, + { + "epoch": 0.444, + "grad_norm": 1.5232124330621741, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 44400 + }, + { + "epoch": 0.44401, + "grad_norm": 1.6028674907796534, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44401 + }, + { + "epoch": 0.44402, + "grad_norm": 1.638403752574772, + "learning_rate": 0.003, + "loss": 4.039, + "step": 44402 + }, + { + "epoch": 0.44403, + "grad_norm": 1.5283465809607746, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44403 + }, + { + "epoch": 0.44404, + "grad_norm": 1.5204700514088014, + "learning_rate": 0.003, + "loss": 4.034, + "step": 44404 + }, + { + "epoch": 0.44405, + "grad_norm": 1.4942684601897496, + "learning_rate": 0.003, + "loss": 4.006, + "step": 44405 + }, + { + "epoch": 0.44406, + "grad_norm": 1.537659417590596, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44406 + }, + { + "epoch": 0.44407, + "grad_norm": 1.3947036781453641, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 44407 + }, + { + "epoch": 0.44408, + "grad_norm": 1.5382811240136947, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 44408 + }, + { + "epoch": 0.44409, + "grad_norm": 1.3630662985195543, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44409 + }, + { + "epoch": 0.4441, + "grad_norm": 1.7208920900316271, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 44410 + }, + { + "epoch": 0.44411, + "grad_norm": 1.7595170303560534, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 44411 + }, + { + "epoch": 0.44412, + "grad_norm": 1.451525596223124, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 44412 + }, + { + "epoch": 0.44413, + "grad_norm": 1.5212143123892723, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 44413 + }, + { + "epoch": 0.44414, + "grad_norm": 1.753193097502194, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 44414 + }, + { + "epoch": 0.44415, + "grad_norm": 1.3569596014210055, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 44415 + }, + { + "epoch": 0.44416, + "grad_norm": 1.5365678610555253, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44416 + }, + { + "epoch": 0.44417, + "grad_norm": 1.497547353072197, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44417 + }, + { + "epoch": 0.44418, + "grad_norm": 1.6460045783179673, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44418 + }, + { + "epoch": 0.44419, + "grad_norm": 1.3811116677387707, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44419 + }, + { + "epoch": 0.4442, + "grad_norm": 1.5652257802529903, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 44420 + }, + { + "epoch": 0.44421, + "grad_norm": 1.5398846583687054, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 44421 + }, + { + "epoch": 0.44422, + "grad_norm": 1.578315366174135, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 44422 + }, + { + "epoch": 0.44423, + "grad_norm": 1.6363091312570675, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44423 + }, + { + "epoch": 0.44424, + "grad_norm": 1.647852748771324, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 44424 + }, + { + "epoch": 0.44425, + "grad_norm": 1.3479402955626423, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 44425 + }, + { + "epoch": 0.44426, + "grad_norm": 1.7396820609121204, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 44426 + }, + { + "epoch": 0.44427, + "grad_norm": 1.489842949998206, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 44427 + }, + { + "epoch": 0.44428, + "grad_norm": 1.7750019053624198, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44428 + }, + { + "epoch": 0.44429, + "grad_norm": 1.3851577971327886, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 44429 + }, + { + "epoch": 0.4443, + "grad_norm": 1.4103477672599345, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 44430 + }, + { + "epoch": 0.44431, + "grad_norm": 1.6558713792741573, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44431 + }, + { + "epoch": 0.44432, + "grad_norm": 1.625399696407926, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 44432 + }, + { + "epoch": 0.44433, + "grad_norm": 1.979428551440574, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 44433 + }, + { + "epoch": 0.44434, + "grad_norm": 1.2725902063712424, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 44434 + }, + { + "epoch": 0.44435, + "grad_norm": 1.7137458079358963, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 44435 + }, + { + "epoch": 0.44436, + "grad_norm": 1.4982392148160206, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 44436 + }, + { + "epoch": 0.44437, + "grad_norm": 1.4296936721032132, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 44437 + }, + { + "epoch": 0.44438, + "grad_norm": 1.5970577174639138, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 44438 + }, + { + "epoch": 0.44439, + "grad_norm": 1.323266055588009, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 44439 + }, + { + "epoch": 0.4444, + "grad_norm": 1.684750764711484, + "learning_rate": 0.003, + "loss": 4.058, + "step": 44440 + }, + { + "epoch": 0.44441, + "grad_norm": 1.3879787344723407, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 44441 + }, + { + "epoch": 0.44442, + "grad_norm": 1.7139921299398198, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 44442 + }, + { + "epoch": 0.44443, + "grad_norm": 1.9535162184921742, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 44443 + }, + { + "epoch": 0.44444, + "grad_norm": 1.1212655729625474, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44444 + }, + { + "epoch": 0.44445, + "grad_norm": 1.7972845731861076, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 44445 + }, + { + "epoch": 0.44446, + "grad_norm": 1.8603635906685796, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44446 + }, + { + "epoch": 0.44447, + "grad_norm": 1.0802982382285178, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 44447 + }, + { + "epoch": 0.44448, + "grad_norm": 1.7882448355893161, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 44448 + }, + { + "epoch": 0.44449, + "grad_norm": 1.2266908549440596, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44449 + }, + { + "epoch": 0.4445, + "grad_norm": 1.476429338212874, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44450 + }, + { + "epoch": 0.44451, + "grad_norm": 1.3631911453433294, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44451 + }, + { + "epoch": 0.44452, + "grad_norm": 1.6546373130976206, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 44452 + }, + { + "epoch": 0.44453, + "grad_norm": 1.5331755508445672, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 44453 + }, + { + "epoch": 0.44454, + "grad_norm": 1.513969229091968, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 44454 + }, + { + "epoch": 0.44455, + "grad_norm": 1.8191172829229834, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 44455 + }, + { + "epoch": 0.44456, + "grad_norm": 1.4951511415275018, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44456 + }, + { + "epoch": 0.44457, + "grad_norm": 1.731597292158209, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44457 + }, + { + "epoch": 0.44458, + "grad_norm": 1.5582900116070704, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 44458 + }, + { + "epoch": 0.44459, + "grad_norm": 1.3478782789443173, + "learning_rate": 0.003, + "loss": 4.007, + "step": 44459 + }, + { + "epoch": 0.4446, + "grad_norm": 1.4982937811394983, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 44460 + }, + { + "epoch": 0.44461, + "grad_norm": 1.4434877445631111, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 44461 + }, + { + "epoch": 0.44462, + "grad_norm": 1.3715918563658323, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44462 + }, + { + "epoch": 0.44463, + "grad_norm": 1.579018086659429, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44463 + }, + { + "epoch": 0.44464, + "grad_norm": 1.518039463377832, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 44464 + }, + { + "epoch": 0.44465, + "grad_norm": 1.5520425851617563, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 44465 + }, + { + "epoch": 0.44466, + "grad_norm": 1.723348553692129, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44466 + }, + { + "epoch": 0.44467, + "grad_norm": 1.6291735955200852, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 44467 + }, + { + "epoch": 0.44468, + "grad_norm": 1.3971425062883307, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44468 + }, + { + "epoch": 0.44469, + "grad_norm": 1.3755779858528627, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44469 + }, + { + "epoch": 0.4447, + "grad_norm": 1.6443564808886075, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 44470 + }, + { + "epoch": 0.44471, + "grad_norm": 1.6317815884160718, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 44471 + }, + { + "epoch": 0.44472, + "grad_norm": 1.916043655657533, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 44472 + }, + { + "epoch": 0.44473, + "grad_norm": 1.6655271463797756, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44473 + }, + { + "epoch": 0.44474, + "grad_norm": 1.6381303201593234, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 44474 + }, + { + "epoch": 0.44475, + "grad_norm": 1.3619575834185749, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44475 + }, + { + "epoch": 0.44476, + "grad_norm": 1.5782437792986344, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 44476 + }, + { + "epoch": 0.44477, + "grad_norm": 1.4314746173302453, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44477 + }, + { + "epoch": 0.44478, + "grad_norm": 1.5931039847578696, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 44478 + }, + { + "epoch": 0.44479, + "grad_norm": 1.4810119523523577, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44479 + }, + { + "epoch": 0.4448, + "grad_norm": 1.497046853848041, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 44480 + }, + { + "epoch": 0.44481, + "grad_norm": 1.363793613355275, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 44481 + }, + { + "epoch": 0.44482, + "grad_norm": 1.6976233290368945, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44482 + }, + { + "epoch": 0.44483, + "grad_norm": 1.3275052821652125, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 44483 + }, + { + "epoch": 0.44484, + "grad_norm": 1.4251963062940747, + "learning_rate": 0.003, + "loss": 4.069, + "step": 44484 + }, + { + "epoch": 0.44485, + "grad_norm": 1.532138711994114, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 44485 + }, + { + "epoch": 0.44486, + "grad_norm": 1.590573634552236, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 44486 + }, + { + "epoch": 0.44487, + "grad_norm": 1.7172377571849928, + "learning_rate": 0.003, + "loss": 4.042, + "step": 44487 + }, + { + "epoch": 0.44488, + "grad_norm": 1.6509734424293487, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 44488 + }, + { + "epoch": 0.44489, + "grad_norm": 1.928508024464437, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44489 + }, + { + "epoch": 0.4449, + "grad_norm": 1.8627789874108083, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44490 + }, + { + "epoch": 0.44491, + "grad_norm": 1.1848106246827053, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44491 + }, + { + "epoch": 0.44492, + "grad_norm": 1.9559338960792003, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 44492 + }, + { + "epoch": 0.44493, + "grad_norm": 1.3888833957803612, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 44493 + }, + { + "epoch": 0.44494, + "grad_norm": 1.431359474641335, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 44494 + }, + { + "epoch": 0.44495, + "grad_norm": 1.751270387301438, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 44495 + }, + { + "epoch": 0.44496, + "grad_norm": 1.710001624760132, + "learning_rate": 0.003, + "loss": 4.044, + "step": 44496 + }, + { + "epoch": 0.44497, + "grad_norm": 1.256787802128839, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 44497 + }, + { + "epoch": 0.44498, + "grad_norm": 1.8152967757212817, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 44498 + }, + { + "epoch": 0.44499, + "grad_norm": 1.2779759632100516, + "learning_rate": 0.003, + "loss": 4.001, + "step": 44499 + }, + { + "epoch": 0.445, + "grad_norm": 1.5283168162501046, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 44500 + }, + { + "epoch": 0.44501, + "grad_norm": 1.7626928994239643, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 44501 + }, + { + "epoch": 0.44502, + "grad_norm": 1.39360813784524, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44502 + }, + { + "epoch": 0.44503, + "grad_norm": 1.6002890412916515, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44503 + }, + { + "epoch": 0.44504, + "grad_norm": 1.538435302219328, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 44504 + }, + { + "epoch": 0.44505, + "grad_norm": 1.3864870116496775, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 44505 + }, + { + "epoch": 0.44506, + "grad_norm": 1.4533165093281255, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 44506 + }, + { + "epoch": 0.44507, + "grad_norm": 1.7081552718891417, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 44507 + }, + { + "epoch": 0.44508, + "grad_norm": 1.2799047976525428, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 44508 + }, + { + "epoch": 0.44509, + "grad_norm": 1.5869914074799805, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 44509 + }, + { + "epoch": 0.4451, + "grad_norm": 1.5189206281423129, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 44510 + }, + { + "epoch": 0.44511, + "grad_norm": 1.5114115538035917, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44511 + }, + { + "epoch": 0.44512, + "grad_norm": 1.520857925764908, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 44512 + }, + { + "epoch": 0.44513, + "grad_norm": 1.520817251088135, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 44513 + }, + { + "epoch": 0.44514, + "grad_norm": 1.229757522142017, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 44514 + }, + { + "epoch": 0.44515, + "grad_norm": 1.7167161656448684, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 44515 + }, + { + "epoch": 0.44516, + "grad_norm": 1.569502256800442, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 44516 + }, + { + "epoch": 0.44517, + "grad_norm": 1.3644913811796644, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44517 + }, + { + "epoch": 0.44518, + "grad_norm": 1.6246151861279758, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 44518 + }, + { + "epoch": 0.44519, + "grad_norm": 1.2213975542688005, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44519 + }, + { + "epoch": 0.4452, + "grad_norm": 1.8952756176354277, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 44520 + }, + { + "epoch": 0.44521, + "grad_norm": 1.2539476395838187, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 44521 + }, + { + "epoch": 0.44522, + "grad_norm": 2.1626133838789023, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44522 + }, + { + "epoch": 0.44523, + "grad_norm": 1.3935741787921407, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 44523 + }, + { + "epoch": 0.44524, + "grad_norm": 1.6025375648501057, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 44524 + }, + { + "epoch": 0.44525, + "grad_norm": 1.613037265638934, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 44525 + }, + { + "epoch": 0.44526, + "grad_norm": 1.5840967238141566, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 44526 + }, + { + "epoch": 0.44527, + "grad_norm": 1.631788923365154, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 44527 + }, + { + "epoch": 0.44528, + "grad_norm": 1.5578719725217456, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44528 + }, + { + "epoch": 0.44529, + "grad_norm": 1.6625672068627178, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44529 + }, + { + "epoch": 0.4453, + "grad_norm": 1.4411544823859546, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 44530 + }, + { + "epoch": 0.44531, + "grad_norm": 1.4320629980538901, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 44531 + }, + { + "epoch": 0.44532, + "grad_norm": 1.227961379647342, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44532 + }, + { + "epoch": 0.44533, + "grad_norm": 1.5885393710991351, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 44533 + }, + { + "epoch": 0.44534, + "grad_norm": 1.2707465803016578, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 44534 + }, + { + "epoch": 0.44535, + "grad_norm": 1.9716309289471279, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 44535 + }, + { + "epoch": 0.44536, + "grad_norm": 1.5698446237738803, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 44536 + }, + { + "epoch": 0.44537, + "grad_norm": 1.5534938090360477, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 44537 + }, + { + "epoch": 0.44538, + "grad_norm": 1.4187561838441445, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 44538 + }, + { + "epoch": 0.44539, + "grad_norm": 1.738365823586243, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 44539 + }, + { + "epoch": 0.4454, + "grad_norm": 1.1892681324964713, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44540 + }, + { + "epoch": 0.44541, + "grad_norm": 1.7908531710873574, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 44541 + }, + { + "epoch": 0.44542, + "grad_norm": 1.353518172315116, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 44542 + }, + { + "epoch": 0.44543, + "grad_norm": 1.5022373902880435, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 44543 + }, + { + "epoch": 0.44544, + "grad_norm": 1.5649882157796313, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 44544 + }, + { + "epoch": 0.44545, + "grad_norm": 1.6508102523045054, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 44545 + }, + { + "epoch": 0.44546, + "grad_norm": 1.4768377923904672, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 44546 + }, + { + "epoch": 0.44547, + "grad_norm": 1.4972161119146123, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44547 + }, + { + "epoch": 0.44548, + "grad_norm": 1.533204282583116, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 44548 + }, + { + "epoch": 0.44549, + "grad_norm": 1.5886137760890626, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 44549 + }, + { + "epoch": 0.4455, + "grad_norm": 1.4540051365679871, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 44550 + }, + { + "epoch": 0.44551, + "grad_norm": 1.6515981079841433, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44551 + }, + { + "epoch": 0.44552, + "grad_norm": 1.1173777250059151, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 44552 + }, + { + "epoch": 0.44553, + "grad_norm": 1.85161933150897, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 44553 + }, + { + "epoch": 0.44554, + "grad_norm": 1.344431189262562, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44554 + }, + { + "epoch": 0.44555, + "grad_norm": 1.9576352254931184, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44555 + }, + { + "epoch": 0.44556, + "grad_norm": 1.6742073451723007, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 44556 + }, + { + "epoch": 0.44557, + "grad_norm": 1.4899705650089523, + "learning_rate": 0.003, + "loss": 4.015, + "step": 44557 + }, + { + "epoch": 0.44558, + "grad_norm": 1.6251052864202644, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 44558 + }, + { + "epoch": 0.44559, + "grad_norm": 1.69545152403381, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 44559 + }, + { + "epoch": 0.4456, + "grad_norm": 1.3212454539232663, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 44560 + }, + { + "epoch": 0.44561, + "grad_norm": 1.400687931176328, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 44561 + }, + { + "epoch": 0.44562, + "grad_norm": 1.6458275747976616, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 44562 + }, + { + "epoch": 0.44563, + "grad_norm": 1.4273179763086878, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 44563 + }, + { + "epoch": 0.44564, + "grad_norm": 1.2879830298909376, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44564 + }, + { + "epoch": 0.44565, + "grad_norm": 1.7695845893184194, + "learning_rate": 0.003, + "loss": 4.045, + "step": 44565 + }, + { + "epoch": 0.44566, + "grad_norm": 1.4155298568811434, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44566 + }, + { + "epoch": 0.44567, + "grad_norm": 1.9796393623181339, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44567 + }, + { + "epoch": 0.44568, + "grad_norm": 1.3154963223947533, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44568 + }, + { + "epoch": 0.44569, + "grad_norm": 1.5537755744673936, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 44569 + }, + { + "epoch": 0.4457, + "grad_norm": 1.5411697602958798, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 44570 + }, + { + "epoch": 0.44571, + "grad_norm": 1.5488281524507088, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44571 + }, + { + "epoch": 0.44572, + "grad_norm": 1.6300469674840212, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 44572 + }, + { + "epoch": 0.44573, + "grad_norm": 1.66670242240406, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44573 + }, + { + "epoch": 0.44574, + "grad_norm": 1.5514468531282122, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 44574 + }, + { + "epoch": 0.44575, + "grad_norm": 1.567018459950307, + "learning_rate": 0.003, + "loss": 4.022, + "step": 44575 + }, + { + "epoch": 0.44576, + "grad_norm": 1.4222134366891879, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44576 + }, + { + "epoch": 0.44577, + "grad_norm": 1.8418994157020498, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44577 + }, + { + "epoch": 0.44578, + "grad_norm": 1.317332353959941, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 44578 + }, + { + "epoch": 0.44579, + "grad_norm": 1.5468473135841068, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 44579 + }, + { + "epoch": 0.4458, + "grad_norm": 1.5329906980972088, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 44580 + }, + { + "epoch": 0.44581, + "grad_norm": 1.784030327329986, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 44581 + }, + { + "epoch": 0.44582, + "grad_norm": 1.2809937891205083, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 44582 + }, + { + "epoch": 0.44583, + "grad_norm": 1.8673320854638775, + "learning_rate": 0.003, + "loss": 4.019, + "step": 44583 + }, + { + "epoch": 0.44584, + "grad_norm": 1.4369098400684566, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 44584 + }, + { + "epoch": 0.44585, + "grad_norm": 1.6080812106335403, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 44585 + }, + { + "epoch": 0.44586, + "grad_norm": 1.246022368224844, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44586 + }, + { + "epoch": 0.44587, + "grad_norm": 1.6583497919208334, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 44587 + }, + { + "epoch": 0.44588, + "grad_norm": 1.333596828182111, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 44588 + }, + { + "epoch": 0.44589, + "grad_norm": 1.8587861055294157, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44589 + }, + { + "epoch": 0.4459, + "grad_norm": 1.320333134550277, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 44590 + }, + { + "epoch": 0.44591, + "grad_norm": 1.8576129359977192, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 44591 + }, + { + "epoch": 0.44592, + "grad_norm": 1.381633222875759, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 44592 + }, + { + "epoch": 0.44593, + "grad_norm": 1.7517380436412682, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 44593 + }, + { + "epoch": 0.44594, + "grad_norm": 1.7748474683925408, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 44594 + }, + { + "epoch": 0.44595, + "grad_norm": 1.3769123419401532, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 44595 + }, + { + "epoch": 0.44596, + "grad_norm": 1.4793708152595162, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 44596 + }, + { + "epoch": 0.44597, + "grad_norm": 1.3659438801917372, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 44597 + }, + { + "epoch": 0.44598, + "grad_norm": 1.5811386669509835, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 44598 + }, + { + "epoch": 0.44599, + "grad_norm": 1.5378744700820932, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 44599 + }, + { + "epoch": 0.446, + "grad_norm": 1.6089040328093582, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 44600 + }, + { + "epoch": 0.44601, + "grad_norm": 1.4565647739758627, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 44601 + }, + { + "epoch": 0.44602, + "grad_norm": 1.496661343974984, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 44602 + }, + { + "epoch": 0.44603, + "grad_norm": 1.610738213447312, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 44603 + }, + { + "epoch": 0.44604, + "grad_norm": 1.3021456921151104, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 44604 + }, + { + "epoch": 0.44605, + "grad_norm": 1.854457366193177, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 44605 + }, + { + "epoch": 0.44606, + "grad_norm": 1.3147191902287279, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 44606 + }, + { + "epoch": 0.44607, + "grad_norm": 1.9217119558789009, + "learning_rate": 0.003, + "loss": 4.025, + "step": 44607 + }, + { + "epoch": 0.44608, + "grad_norm": 1.2847649341872007, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 44608 + }, + { + "epoch": 0.44609, + "grad_norm": 1.4924596134922548, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 44609 + }, + { + "epoch": 0.4461, + "grad_norm": 1.5643219024362525, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 44610 + }, + { + "epoch": 0.44611, + "grad_norm": 1.5310883249937441, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44611 + }, + { + "epoch": 0.44612, + "grad_norm": 1.6489497241393383, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 44612 + }, + { + "epoch": 0.44613, + "grad_norm": 1.488099315579587, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 44613 + }, + { + "epoch": 0.44614, + "grad_norm": 1.5618536201221938, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 44614 + }, + { + "epoch": 0.44615, + "grad_norm": 1.3823508327495901, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 44615 + }, + { + "epoch": 0.44616, + "grad_norm": 1.8805217679556372, + "learning_rate": 0.003, + "loss": 4.04, + "step": 44616 + }, + { + "epoch": 0.44617, + "grad_norm": 1.9112115434724497, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 44617 + }, + { + "epoch": 0.44618, + "grad_norm": 1.605541754676354, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44618 + }, + { + "epoch": 0.44619, + "grad_norm": 1.421082722453455, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 44619 + }, + { + "epoch": 0.4462, + "grad_norm": 1.4160734645333162, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 44620 + }, + { + "epoch": 0.44621, + "grad_norm": 1.5409854656284356, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 44621 + }, + { + "epoch": 0.44622, + "grad_norm": 1.3572895964634057, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 44622 + }, + { + "epoch": 0.44623, + "grad_norm": 1.4088960377865862, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 44623 + }, + { + "epoch": 0.44624, + "grad_norm": 1.689896647603429, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 44624 + }, + { + "epoch": 0.44625, + "grad_norm": 1.5382353613831257, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 44625 + }, + { + "epoch": 0.44626, + "grad_norm": 1.721220090984284, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 44626 + }, + { + "epoch": 0.44627, + "grad_norm": 1.3127598614193727, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44627 + }, + { + "epoch": 0.44628, + "grad_norm": 1.7970781508569422, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44628 + }, + { + "epoch": 0.44629, + "grad_norm": 1.4001610105819098, + "learning_rate": 0.003, + "loss": 4.036, + "step": 44629 + }, + { + "epoch": 0.4463, + "grad_norm": 1.2202584143833697, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 44630 + }, + { + "epoch": 0.44631, + "grad_norm": 1.8145786284600383, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 44631 + }, + { + "epoch": 0.44632, + "grad_norm": 1.5082811998823893, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 44632 + }, + { + "epoch": 0.44633, + "grad_norm": 1.5051971319793132, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 44633 + }, + { + "epoch": 0.44634, + "grad_norm": 1.5720400255881888, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 44634 + }, + { + "epoch": 0.44635, + "grad_norm": 1.647819327860018, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 44635 + }, + { + "epoch": 0.44636, + "grad_norm": 1.6585781172837692, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 44636 + }, + { + "epoch": 0.44637, + "grad_norm": 1.4946929888572837, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 44637 + }, + { + "epoch": 0.44638, + "grad_norm": 1.432661799692589, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 44638 + }, + { + "epoch": 0.44639, + "grad_norm": 1.5912712101255964, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 44639 + }, + { + "epoch": 0.4464, + "grad_norm": 1.2563144405960638, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 44640 + }, + { + "epoch": 0.44641, + "grad_norm": 1.694282036858328, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44641 + }, + { + "epoch": 0.44642, + "grad_norm": 1.3192862732791366, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 44642 + }, + { + "epoch": 0.44643, + "grad_norm": 1.9965344358730202, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 44643 + }, + { + "epoch": 0.44644, + "grad_norm": 1.4084762456527056, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 44644 + }, + { + "epoch": 0.44645, + "grad_norm": 1.9155622219194355, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 44645 + }, + { + "epoch": 0.44646, + "grad_norm": 1.3602364252681654, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44646 + }, + { + "epoch": 0.44647, + "grad_norm": 1.816909295922154, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44647 + }, + { + "epoch": 0.44648, + "grad_norm": 1.4441833152595371, + "learning_rate": 0.003, + "loss": 4.038, + "step": 44648 + }, + { + "epoch": 0.44649, + "grad_norm": 1.6879141183973392, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 44649 + }, + { + "epoch": 0.4465, + "grad_norm": 1.7277852289880855, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44650 + }, + { + "epoch": 0.44651, + "grad_norm": 1.327923366494027, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 44651 + }, + { + "epoch": 0.44652, + "grad_norm": 1.4750706921180299, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 44652 + }, + { + "epoch": 0.44653, + "grad_norm": 1.7649047306408072, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 44653 + }, + { + "epoch": 0.44654, + "grad_norm": 1.5352108138829936, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 44654 + }, + { + "epoch": 0.44655, + "grad_norm": 1.580164766531279, + "learning_rate": 0.003, + "loss": 4.04, + "step": 44655 + }, + { + "epoch": 0.44656, + "grad_norm": 1.6339324468211234, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 44656 + }, + { + "epoch": 0.44657, + "grad_norm": 1.2659996138101703, + "learning_rate": 0.003, + "loss": 4.013, + "step": 44657 + }, + { + "epoch": 0.44658, + "grad_norm": 1.5970586820260861, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 44658 + }, + { + "epoch": 0.44659, + "grad_norm": 1.4590807815549511, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 44659 + }, + { + "epoch": 0.4466, + "grad_norm": 1.9684191474138546, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 44660 + }, + { + "epoch": 0.44661, + "grad_norm": 1.262857160908861, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 44661 + }, + { + "epoch": 0.44662, + "grad_norm": 1.8942983964738584, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 44662 + }, + { + "epoch": 0.44663, + "grad_norm": 1.3616315739606495, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 44663 + }, + { + "epoch": 0.44664, + "grad_norm": 1.4380915846608053, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 44664 + }, + { + "epoch": 0.44665, + "grad_norm": 1.7981418649038636, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 44665 + }, + { + "epoch": 0.44666, + "grad_norm": 1.6010782132473373, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 44666 + }, + { + "epoch": 0.44667, + "grad_norm": 1.5875060087102957, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 44667 + }, + { + "epoch": 0.44668, + "grad_norm": 1.6293618415740985, + "learning_rate": 0.003, + "loss": 4.06, + "step": 44668 + }, + { + "epoch": 0.44669, + "grad_norm": 1.5374037615660077, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 44669 + }, + { + "epoch": 0.4467, + "grad_norm": 1.5166418971321791, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 44670 + }, + { + "epoch": 0.44671, + "grad_norm": 1.5508528886166049, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44671 + }, + { + "epoch": 0.44672, + "grad_norm": 1.3597006888971233, + "learning_rate": 0.003, + "loss": 4.066, + "step": 44672 + }, + { + "epoch": 0.44673, + "grad_norm": 1.8165684685868022, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 44673 + }, + { + "epoch": 0.44674, + "grad_norm": 1.1845718194745096, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44674 + }, + { + "epoch": 0.44675, + "grad_norm": 1.7352631732238144, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 44675 + }, + { + "epoch": 0.44676, + "grad_norm": 1.58124068892223, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 44676 + }, + { + "epoch": 0.44677, + "grad_norm": 1.7115716385444335, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 44677 + }, + { + "epoch": 0.44678, + "grad_norm": 1.4918157161284602, + "learning_rate": 0.003, + "loss": 4.051, + "step": 44678 + }, + { + "epoch": 0.44679, + "grad_norm": 1.589702349310606, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 44679 + }, + { + "epoch": 0.4468, + "grad_norm": 1.4126691701144731, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44680 + }, + { + "epoch": 0.44681, + "grad_norm": 1.7172180933377104, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 44681 + }, + { + "epoch": 0.44682, + "grad_norm": 1.4550395645729812, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 44682 + }, + { + "epoch": 0.44683, + "grad_norm": 1.8450392227608254, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 44683 + }, + { + "epoch": 0.44684, + "grad_norm": 1.523482162238819, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 44684 + }, + { + "epoch": 0.44685, + "grad_norm": 1.4600232429521405, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 44685 + }, + { + "epoch": 0.44686, + "grad_norm": 1.5820540073858975, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 44686 + }, + { + "epoch": 0.44687, + "grad_norm": 1.5756608386905728, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44687 + }, + { + "epoch": 0.44688, + "grad_norm": 1.6609680000646931, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 44688 + }, + { + "epoch": 0.44689, + "grad_norm": 1.2569713705283871, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 44689 + }, + { + "epoch": 0.4469, + "grad_norm": 1.6759902906852946, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44690 + }, + { + "epoch": 0.44691, + "grad_norm": 1.78241010017489, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44691 + }, + { + "epoch": 0.44692, + "grad_norm": 1.3365314940442092, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 44692 + }, + { + "epoch": 0.44693, + "grad_norm": 1.772340915731479, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44693 + }, + { + "epoch": 0.44694, + "grad_norm": 1.642235931076375, + "learning_rate": 0.003, + "loss": 4.032, + "step": 44694 + }, + { + "epoch": 0.44695, + "grad_norm": 1.5147361444937497, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 44695 + }, + { + "epoch": 0.44696, + "grad_norm": 1.4579394260998493, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 44696 + }, + { + "epoch": 0.44697, + "grad_norm": 1.3718864640154815, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 44697 + }, + { + "epoch": 0.44698, + "grad_norm": 1.5270475986290364, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44698 + }, + { + "epoch": 0.44699, + "grad_norm": 1.3557409729259775, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 44699 + }, + { + "epoch": 0.447, + "grad_norm": 1.444530900687755, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 44700 + }, + { + "epoch": 0.44701, + "grad_norm": 1.2812263809391036, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 44701 + }, + { + "epoch": 0.44702, + "grad_norm": 1.8865337366991652, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 44702 + }, + { + "epoch": 0.44703, + "grad_norm": 1.2697984544241787, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 44703 + }, + { + "epoch": 0.44704, + "grad_norm": 1.5235702416368566, + "learning_rate": 0.003, + "loss": 4.046, + "step": 44704 + }, + { + "epoch": 0.44705, + "grad_norm": 1.4029559044993323, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 44705 + }, + { + "epoch": 0.44706, + "grad_norm": 2.1307282534406182, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 44706 + }, + { + "epoch": 0.44707, + "grad_norm": 1.4434757927793938, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 44707 + }, + { + "epoch": 0.44708, + "grad_norm": 1.9133799099418658, + "learning_rate": 0.003, + "loss": 4.033, + "step": 44708 + }, + { + "epoch": 0.44709, + "grad_norm": 1.4577050202880606, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 44709 + }, + { + "epoch": 0.4471, + "grad_norm": 1.46723396855444, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 44710 + }, + { + "epoch": 0.44711, + "grad_norm": 1.6098733272043728, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 44711 + }, + { + "epoch": 0.44712, + "grad_norm": 1.7206385397437205, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 44712 + }, + { + "epoch": 0.44713, + "grad_norm": 1.322571522620078, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 44713 + }, + { + "epoch": 0.44714, + "grad_norm": 1.632741785356706, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 44714 + }, + { + "epoch": 0.44715, + "grad_norm": 1.3630368397143853, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 44715 + }, + { + "epoch": 0.44716, + "grad_norm": 1.3915052830287575, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 44716 + }, + { + "epoch": 0.44717, + "grad_norm": 1.8323472809153243, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 44717 + }, + { + "epoch": 0.44718, + "grad_norm": 1.3202487601873316, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 44718 + }, + { + "epoch": 0.44719, + "grad_norm": 1.42438169256135, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 44719 + }, + { + "epoch": 0.4472, + "grad_norm": 1.4820540226378607, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 44720 + }, + { + "epoch": 0.44721, + "grad_norm": 1.873253389534575, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44721 + }, + { + "epoch": 0.44722, + "grad_norm": 1.801675150345499, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 44722 + }, + { + "epoch": 0.44723, + "grad_norm": 1.4418076056477247, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 44723 + }, + { + "epoch": 0.44724, + "grad_norm": 1.4327566384477868, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 44724 + }, + { + "epoch": 0.44725, + "grad_norm": 1.9435950793805576, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44725 + }, + { + "epoch": 0.44726, + "grad_norm": 1.0592974991567665, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 44726 + }, + { + "epoch": 0.44727, + "grad_norm": 1.797616817268142, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 44727 + }, + { + "epoch": 0.44728, + "grad_norm": 1.5731957586761016, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 44728 + }, + { + "epoch": 0.44729, + "grad_norm": 1.7377644646484975, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 44729 + }, + { + "epoch": 0.4473, + "grad_norm": 1.6614061087875878, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 44730 + }, + { + "epoch": 0.44731, + "grad_norm": 1.7565520638667416, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 44731 + }, + { + "epoch": 0.44732, + "grad_norm": 1.4412095414084167, + "learning_rate": 0.003, + "loss": 4.064, + "step": 44732 + }, + { + "epoch": 0.44733, + "grad_norm": 1.8829683607043237, + "learning_rate": 0.003, + "loss": 4.037, + "step": 44733 + }, + { + "epoch": 0.44734, + "grad_norm": 1.125343555740035, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 44734 + }, + { + "epoch": 0.44735, + "grad_norm": 1.898756759275749, + "learning_rate": 0.003, + "loss": 4.044, + "step": 44735 + }, + { + "epoch": 0.44736, + "grad_norm": 1.4669661052293907, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 44736 + }, + { + "epoch": 0.44737, + "grad_norm": 1.6525425129678455, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44737 + }, + { + "epoch": 0.44738, + "grad_norm": 1.478447137891282, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 44738 + }, + { + "epoch": 0.44739, + "grad_norm": 1.4503667417833037, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 44739 + }, + { + "epoch": 0.4474, + "grad_norm": 1.4674632829397962, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 44740 + }, + { + "epoch": 0.44741, + "grad_norm": 1.4378883572778294, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44741 + }, + { + "epoch": 0.44742, + "grad_norm": 1.445704080744737, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 44742 + }, + { + "epoch": 0.44743, + "grad_norm": 1.3730691306524931, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44743 + }, + { + "epoch": 0.44744, + "grad_norm": 1.717979023439242, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 44744 + }, + { + "epoch": 0.44745, + "grad_norm": 1.548685587573099, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 44745 + }, + { + "epoch": 0.44746, + "grad_norm": 1.3296774954553274, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 44746 + }, + { + "epoch": 0.44747, + "grad_norm": 1.713574196907768, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 44747 + }, + { + "epoch": 0.44748, + "grad_norm": 1.4556311913108173, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 44748 + }, + { + "epoch": 0.44749, + "grad_norm": 1.5955245192729481, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 44749 + }, + { + "epoch": 0.4475, + "grad_norm": 1.4712070388188347, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 44750 + }, + { + "epoch": 0.44751, + "grad_norm": 1.527708963233963, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 44751 + }, + { + "epoch": 0.44752, + "grad_norm": 1.4106264778688031, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 44752 + }, + { + "epoch": 0.44753, + "grad_norm": 1.4442041393444087, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 44753 + }, + { + "epoch": 0.44754, + "grad_norm": 1.7803767829828194, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 44754 + }, + { + "epoch": 0.44755, + "grad_norm": 1.4541517996710431, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44755 + }, + { + "epoch": 0.44756, + "grad_norm": 1.7257420777873063, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44756 + }, + { + "epoch": 0.44757, + "grad_norm": 1.4781517452276292, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 44757 + }, + { + "epoch": 0.44758, + "grad_norm": 1.811736832304607, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44758 + }, + { + "epoch": 0.44759, + "grad_norm": 1.3197853363963072, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 44759 + }, + { + "epoch": 0.4476, + "grad_norm": 1.6900281136803608, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 44760 + }, + { + "epoch": 0.44761, + "grad_norm": 1.6092219917585582, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 44761 + }, + { + "epoch": 0.44762, + "grad_norm": 1.7513760948474058, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 44762 + }, + { + "epoch": 0.44763, + "grad_norm": 1.5125169717448197, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 44763 + }, + { + "epoch": 0.44764, + "grad_norm": 1.6048684355325822, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 44764 + }, + { + "epoch": 0.44765, + "grad_norm": 1.3666300173922878, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 44765 + }, + { + "epoch": 0.44766, + "grad_norm": 1.6029042143702883, + "learning_rate": 0.003, + "loss": 4.052, + "step": 44766 + }, + { + "epoch": 0.44767, + "grad_norm": 1.4249595699797872, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 44767 + }, + { + "epoch": 0.44768, + "grad_norm": 1.7385252974461225, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 44768 + }, + { + "epoch": 0.44769, + "grad_norm": 1.097202558240984, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 44769 + }, + { + "epoch": 0.4477, + "grad_norm": 2.0616725965603258, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 44770 + }, + { + "epoch": 0.44771, + "grad_norm": 1.4527923657338362, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 44771 + }, + { + "epoch": 0.44772, + "grad_norm": 1.8647520591867568, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 44772 + }, + { + "epoch": 0.44773, + "grad_norm": 1.6570741255594632, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44773 + }, + { + "epoch": 0.44774, + "grad_norm": 1.1595576356402222, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 44774 + }, + { + "epoch": 0.44775, + "grad_norm": 1.9082272585011668, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 44775 + }, + { + "epoch": 0.44776, + "grad_norm": 1.3874065113790526, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 44776 + }, + { + "epoch": 0.44777, + "grad_norm": 1.5616187118098217, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 44777 + }, + { + "epoch": 0.44778, + "grad_norm": 1.9311111361090778, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 44778 + }, + { + "epoch": 0.44779, + "grad_norm": 1.278375510698264, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 44779 + }, + { + "epoch": 0.4478, + "grad_norm": 1.6524354255114353, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 44780 + }, + { + "epoch": 0.44781, + "grad_norm": 1.1455213453444395, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 44781 + }, + { + "epoch": 0.44782, + "grad_norm": 1.51309920312087, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 44782 + }, + { + "epoch": 0.44783, + "grad_norm": 1.12824177831646, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 44783 + }, + { + "epoch": 0.44784, + "grad_norm": 1.6740141868825906, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44784 + }, + { + "epoch": 0.44785, + "grad_norm": 1.328417334425566, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 44785 + }, + { + "epoch": 0.44786, + "grad_norm": 1.598431155355486, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 44786 + }, + { + "epoch": 0.44787, + "grad_norm": 1.3201159468598707, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44787 + }, + { + "epoch": 0.44788, + "grad_norm": 1.7104356894875246, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 44788 + }, + { + "epoch": 0.44789, + "grad_norm": 1.5974405483597447, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44789 + }, + { + "epoch": 0.4479, + "grad_norm": 1.8629697427300977, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 44790 + }, + { + "epoch": 0.44791, + "grad_norm": 1.5221557440603535, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 44791 + }, + { + "epoch": 0.44792, + "grad_norm": 1.436337183585485, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 44792 + }, + { + "epoch": 0.44793, + "grad_norm": 1.7150462290473298, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 44793 + }, + { + "epoch": 0.44794, + "grad_norm": 1.245943031148978, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 44794 + }, + { + "epoch": 0.44795, + "grad_norm": 1.8508134730158237, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44795 + }, + { + "epoch": 0.44796, + "grad_norm": 1.3654631783995446, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 44796 + }, + { + "epoch": 0.44797, + "grad_norm": 1.8235665784271393, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 44797 + }, + { + "epoch": 0.44798, + "grad_norm": 1.435141276137711, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 44798 + }, + { + "epoch": 0.44799, + "grad_norm": 1.795972197690132, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 44799 + }, + { + "epoch": 0.448, + "grad_norm": 1.4973057168285921, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 44800 + }, + { + "epoch": 0.44801, + "grad_norm": 1.3974468523128178, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 44801 + }, + { + "epoch": 0.44802, + "grad_norm": 1.5501444374551419, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44802 + }, + { + "epoch": 0.44803, + "grad_norm": 1.6149466557649814, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 44803 + }, + { + "epoch": 0.44804, + "grad_norm": 1.5843347284128817, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 44804 + }, + { + "epoch": 0.44805, + "grad_norm": 1.4700688398808313, + "learning_rate": 0.003, + "loss": 4.014, + "step": 44805 + }, + { + "epoch": 0.44806, + "grad_norm": 1.7797029585588826, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 44806 + }, + { + "epoch": 0.44807, + "grad_norm": 1.3791109144278988, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 44807 + }, + { + "epoch": 0.44808, + "grad_norm": 1.6256747127174078, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44808 + }, + { + "epoch": 0.44809, + "grad_norm": 1.295010143695008, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 44809 + }, + { + "epoch": 0.4481, + "grad_norm": 1.741552218382768, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 44810 + }, + { + "epoch": 0.44811, + "grad_norm": 1.3156216545100639, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 44811 + }, + { + "epoch": 0.44812, + "grad_norm": 2.202066997043912, + "learning_rate": 0.003, + "loss": 4.044, + "step": 44812 + }, + { + "epoch": 0.44813, + "grad_norm": 1.3766401333646339, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 44813 + }, + { + "epoch": 0.44814, + "grad_norm": 1.6838667057434225, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 44814 + }, + { + "epoch": 0.44815, + "grad_norm": 1.9168677961409504, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 44815 + }, + { + "epoch": 0.44816, + "grad_norm": 1.512984276548589, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 44816 + }, + { + "epoch": 0.44817, + "grad_norm": 1.301978817591856, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44817 + }, + { + "epoch": 0.44818, + "grad_norm": 1.7069613371595733, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 44818 + }, + { + "epoch": 0.44819, + "grad_norm": 1.2285314075689229, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44819 + }, + { + "epoch": 0.4482, + "grad_norm": 1.7407761890036126, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 44820 + }, + { + "epoch": 0.44821, + "grad_norm": 1.3826220176108108, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 44821 + }, + { + "epoch": 0.44822, + "grad_norm": 1.5889642411017542, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 44822 + }, + { + "epoch": 0.44823, + "grad_norm": 1.2900746185569507, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 44823 + }, + { + "epoch": 0.44824, + "grad_norm": 1.5287906002544496, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 44824 + }, + { + "epoch": 0.44825, + "grad_norm": 1.3824005133621589, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 44825 + }, + { + "epoch": 0.44826, + "grad_norm": 1.887410632245554, + "learning_rate": 0.003, + "loss": 4.027, + "step": 44826 + }, + { + "epoch": 0.44827, + "grad_norm": 1.6730297133212326, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 44827 + }, + { + "epoch": 0.44828, + "grad_norm": 1.797608149452127, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44828 + }, + { + "epoch": 0.44829, + "grad_norm": 1.420947841768727, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44829 + }, + { + "epoch": 0.4483, + "grad_norm": 1.8884556624291164, + "learning_rate": 0.003, + "loss": 4.028, + "step": 44830 + }, + { + "epoch": 0.44831, + "grad_norm": 1.3029043048106623, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44831 + }, + { + "epoch": 0.44832, + "grad_norm": 1.964589369853273, + "learning_rate": 0.003, + "loss": 4.029, + "step": 44832 + }, + { + "epoch": 0.44833, + "grad_norm": 1.4148516408129603, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 44833 + }, + { + "epoch": 0.44834, + "grad_norm": 1.30300288515301, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 44834 + }, + { + "epoch": 0.44835, + "grad_norm": 1.5603665153955026, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 44835 + }, + { + "epoch": 0.44836, + "grad_norm": 1.6237648225866923, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44836 + }, + { + "epoch": 0.44837, + "grad_norm": 1.4916917467595208, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 44837 + }, + { + "epoch": 0.44838, + "grad_norm": 1.8112770234404603, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 44838 + }, + { + "epoch": 0.44839, + "grad_norm": 1.2233492114384965, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 44839 + }, + { + "epoch": 0.4484, + "grad_norm": 1.6454610120365083, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 44840 + }, + { + "epoch": 0.44841, + "grad_norm": 1.4378624409464884, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 44841 + }, + { + "epoch": 0.44842, + "grad_norm": 1.6078379483438643, + "learning_rate": 0.003, + "loss": 4.051, + "step": 44842 + }, + { + "epoch": 0.44843, + "grad_norm": 1.6184691048317303, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 44843 + }, + { + "epoch": 0.44844, + "grad_norm": 1.2989286809464786, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44844 + }, + { + "epoch": 0.44845, + "grad_norm": 1.328566046144954, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 44845 + }, + { + "epoch": 0.44846, + "grad_norm": 1.4837321660657357, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 44846 + }, + { + "epoch": 0.44847, + "grad_norm": 1.5027912478444807, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 44847 + }, + { + "epoch": 0.44848, + "grad_norm": 1.373751669258319, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 44848 + }, + { + "epoch": 0.44849, + "grad_norm": 1.5427204762865077, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 44849 + }, + { + "epoch": 0.4485, + "grad_norm": 1.6656268100077372, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 44850 + }, + { + "epoch": 0.44851, + "grad_norm": 1.9002721114815107, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 44851 + }, + { + "epoch": 0.44852, + "grad_norm": 1.2512381628864129, + "learning_rate": 0.003, + "loss": 4.043, + "step": 44852 + }, + { + "epoch": 0.44853, + "grad_norm": 1.777932798743558, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44853 + }, + { + "epoch": 0.44854, + "grad_norm": 1.7085703602579398, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 44854 + }, + { + "epoch": 0.44855, + "grad_norm": 1.5841534630876044, + "learning_rate": 0.003, + "loss": 4.058, + "step": 44855 + }, + { + "epoch": 0.44856, + "grad_norm": 1.4928638784551236, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 44856 + }, + { + "epoch": 0.44857, + "grad_norm": 1.5173004354164572, + "learning_rate": 0.003, + "loss": 4.03, + "step": 44857 + }, + { + "epoch": 0.44858, + "grad_norm": 1.354892526410123, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44858 + }, + { + "epoch": 0.44859, + "grad_norm": 1.7270199914362265, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 44859 + }, + { + "epoch": 0.4486, + "grad_norm": 1.4072422657543329, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 44860 + }, + { + "epoch": 0.44861, + "grad_norm": 1.7376861748740469, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 44861 + }, + { + "epoch": 0.44862, + "grad_norm": 1.5480736818289225, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 44862 + }, + { + "epoch": 0.44863, + "grad_norm": 1.4307854257089756, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44863 + }, + { + "epoch": 0.44864, + "grad_norm": 1.5841714434470073, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 44864 + }, + { + "epoch": 0.44865, + "grad_norm": 1.350317160809079, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 44865 + }, + { + "epoch": 0.44866, + "grad_norm": 1.6902847135842707, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 44866 + }, + { + "epoch": 0.44867, + "grad_norm": 1.319896486532983, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 44867 + }, + { + "epoch": 0.44868, + "grad_norm": 1.572065426385339, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 44868 + }, + { + "epoch": 0.44869, + "grad_norm": 1.3164341709099434, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 44869 + }, + { + "epoch": 0.4487, + "grad_norm": 1.544629708435219, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 44870 + }, + { + "epoch": 0.44871, + "grad_norm": 1.7690010030349996, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 44871 + }, + { + "epoch": 0.44872, + "grad_norm": 1.7014863607276478, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44872 + }, + { + "epoch": 0.44873, + "grad_norm": 1.5196364730888372, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44873 + }, + { + "epoch": 0.44874, + "grad_norm": 1.4580236813553185, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 44874 + }, + { + "epoch": 0.44875, + "grad_norm": 1.5834474067770627, + "learning_rate": 0.003, + "loss": 3.997, + "step": 44875 + }, + { + "epoch": 0.44876, + "grad_norm": 1.7970906828137692, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44876 + }, + { + "epoch": 0.44877, + "grad_norm": 1.4888303973964767, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 44877 + }, + { + "epoch": 0.44878, + "grad_norm": 1.5195050602920397, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 44878 + }, + { + "epoch": 0.44879, + "grad_norm": 1.398812839305381, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 44879 + }, + { + "epoch": 0.4488, + "grad_norm": 1.6444166523294856, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 44880 + }, + { + "epoch": 0.44881, + "grad_norm": 1.6229816822060712, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 44881 + }, + { + "epoch": 0.44882, + "grad_norm": 1.7109196544818621, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44882 + }, + { + "epoch": 0.44883, + "grad_norm": 1.6111260631278923, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44883 + }, + { + "epoch": 0.44884, + "grad_norm": 1.5890836959641221, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 44884 + }, + { + "epoch": 0.44885, + "grad_norm": 1.682198542526924, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 44885 + }, + { + "epoch": 0.44886, + "grad_norm": 1.4601160384001037, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44886 + }, + { + "epoch": 0.44887, + "grad_norm": 1.5932532924862224, + "learning_rate": 0.003, + "loss": 4.012, + "step": 44887 + }, + { + "epoch": 0.44888, + "grad_norm": 1.5959479362844327, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 44888 + }, + { + "epoch": 0.44889, + "grad_norm": 1.3638452249334498, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 44889 + }, + { + "epoch": 0.4489, + "grad_norm": 1.5023222678658397, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 44890 + }, + { + "epoch": 0.44891, + "grad_norm": 1.4503269442566142, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 44891 + }, + { + "epoch": 0.44892, + "grad_norm": 1.3125827123890317, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 44892 + }, + { + "epoch": 0.44893, + "grad_norm": 1.6778099621031628, + "learning_rate": 0.003, + "loss": 4.06, + "step": 44893 + }, + { + "epoch": 0.44894, + "grad_norm": 1.3000660564573618, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44894 + }, + { + "epoch": 0.44895, + "grad_norm": 1.6857943463983802, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 44895 + }, + { + "epoch": 0.44896, + "grad_norm": 1.5339170678453344, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 44896 + }, + { + "epoch": 0.44897, + "grad_norm": 1.8398172970286737, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 44897 + }, + { + "epoch": 0.44898, + "grad_norm": 1.7351182177782853, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 44898 + }, + { + "epoch": 0.44899, + "grad_norm": 1.3467033216034487, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 44899 + }, + { + "epoch": 0.449, + "grad_norm": 1.724887933683231, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 44900 + }, + { + "epoch": 0.44901, + "grad_norm": 1.483479469827046, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44901 + }, + { + "epoch": 0.44902, + "grad_norm": 1.4360868800960152, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 44902 + }, + { + "epoch": 0.44903, + "grad_norm": 1.560990697205524, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 44903 + }, + { + "epoch": 0.44904, + "grad_norm": 1.856926499529379, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 44904 + }, + { + "epoch": 0.44905, + "grad_norm": 1.4050459067778445, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 44905 + }, + { + "epoch": 0.44906, + "grad_norm": 1.5907177090988924, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 44906 + }, + { + "epoch": 0.44907, + "grad_norm": 1.4975074925779346, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 44907 + }, + { + "epoch": 0.44908, + "grad_norm": 1.7488650564047152, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 44908 + }, + { + "epoch": 0.44909, + "grad_norm": 1.305139176186865, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 44909 + }, + { + "epoch": 0.4491, + "grad_norm": 1.604377065556934, + "learning_rate": 0.003, + "loss": 4.037, + "step": 44910 + }, + { + "epoch": 0.44911, + "grad_norm": 1.4690277591571494, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 44911 + }, + { + "epoch": 0.44912, + "grad_norm": 1.8216459728623833, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 44912 + }, + { + "epoch": 0.44913, + "grad_norm": 1.50129024734267, + "learning_rate": 0.003, + "loss": 4.049, + "step": 44913 + }, + { + "epoch": 0.44914, + "grad_norm": 1.4502194252364804, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 44914 + }, + { + "epoch": 0.44915, + "grad_norm": 1.7250634630038333, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 44915 + }, + { + "epoch": 0.44916, + "grad_norm": 1.5676556302115852, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 44916 + }, + { + "epoch": 0.44917, + "grad_norm": 1.5165108589748761, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 44917 + }, + { + "epoch": 0.44918, + "grad_norm": 1.5720672375956715, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 44918 + }, + { + "epoch": 0.44919, + "grad_norm": 1.4556745313488244, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44919 + }, + { + "epoch": 0.4492, + "grad_norm": 1.6539772532801478, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 44920 + }, + { + "epoch": 0.44921, + "grad_norm": 1.8221114988164633, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 44921 + }, + { + "epoch": 0.44922, + "grad_norm": 1.4563457379828286, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44922 + }, + { + "epoch": 0.44923, + "grad_norm": 1.5410527108461418, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 44923 + }, + { + "epoch": 0.44924, + "grad_norm": 1.3603887338253513, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44924 + }, + { + "epoch": 0.44925, + "grad_norm": 1.6143886445827735, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 44925 + }, + { + "epoch": 0.44926, + "grad_norm": 1.3564303277827265, + "learning_rate": 0.003, + "loss": 4.049, + "step": 44926 + }, + { + "epoch": 0.44927, + "grad_norm": 1.61347677900417, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44927 + }, + { + "epoch": 0.44928, + "grad_norm": 1.3182249181834926, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 44928 + }, + { + "epoch": 0.44929, + "grad_norm": 1.486932352829978, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 44929 + }, + { + "epoch": 0.4493, + "grad_norm": 1.535167350586372, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 44930 + }, + { + "epoch": 0.44931, + "grad_norm": 1.740531618090155, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 44931 + }, + { + "epoch": 0.44932, + "grad_norm": 1.8050388259338452, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 44932 + }, + { + "epoch": 0.44933, + "grad_norm": 1.3168146529564533, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 44933 + }, + { + "epoch": 0.44934, + "grad_norm": 1.6252368986851482, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 44934 + }, + { + "epoch": 0.44935, + "grad_norm": 1.4660311548079372, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 44935 + }, + { + "epoch": 0.44936, + "grad_norm": 1.5301251827198696, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 44936 + }, + { + "epoch": 0.44937, + "grad_norm": 1.9878469584112537, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 44937 + }, + { + "epoch": 0.44938, + "grad_norm": 1.2528096782967577, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 44938 + }, + { + "epoch": 0.44939, + "grad_norm": 1.7208989199893654, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 44939 + }, + { + "epoch": 0.4494, + "grad_norm": 1.4704780743849053, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 44940 + }, + { + "epoch": 0.44941, + "grad_norm": 1.7216833524516897, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 44941 + }, + { + "epoch": 0.44942, + "grad_norm": 1.6431969507026918, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 44942 + }, + { + "epoch": 0.44943, + "grad_norm": 1.406216229191327, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 44943 + }, + { + "epoch": 0.44944, + "grad_norm": 1.7142054410867966, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44944 + }, + { + "epoch": 0.44945, + "grad_norm": 1.4596866973923766, + "learning_rate": 0.003, + "loss": 4.006, + "step": 44945 + }, + { + "epoch": 0.44946, + "grad_norm": 1.7131939506390617, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 44946 + }, + { + "epoch": 0.44947, + "grad_norm": 1.29602633993662, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 44947 + }, + { + "epoch": 0.44948, + "grad_norm": 1.7220554434240551, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 44948 + }, + { + "epoch": 0.44949, + "grad_norm": 1.2788126532923663, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 44949 + }, + { + "epoch": 0.4495, + "grad_norm": 1.6650802932713273, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 44950 + }, + { + "epoch": 0.44951, + "grad_norm": 1.504282565728189, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 44951 + }, + { + "epoch": 0.44952, + "grad_norm": 1.4782161921133457, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 44952 + }, + { + "epoch": 0.44953, + "grad_norm": 1.6760977259272574, + "learning_rate": 0.003, + "loss": 4.005, + "step": 44953 + }, + { + "epoch": 0.44954, + "grad_norm": 1.3929736318445876, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 44954 + }, + { + "epoch": 0.44955, + "grad_norm": 1.9092748020570096, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 44955 + }, + { + "epoch": 0.44956, + "grad_norm": 1.1293505719741304, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 44956 + }, + { + "epoch": 0.44957, + "grad_norm": 1.6744337864135224, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 44957 + }, + { + "epoch": 0.44958, + "grad_norm": 1.5617169471804175, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 44958 + }, + { + "epoch": 0.44959, + "grad_norm": 1.493057755953311, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 44959 + }, + { + "epoch": 0.4496, + "grad_norm": 1.6446321869645615, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 44960 + }, + { + "epoch": 0.44961, + "grad_norm": 1.3316597358424322, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 44961 + }, + { + "epoch": 0.44962, + "grad_norm": 1.6672992988716147, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 44962 + }, + { + "epoch": 0.44963, + "grad_norm": 1.4481827506856153, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 44963 + }, + { + "epoch": 0.44964, + "grad_norm": 1.3437603086203962, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 44964 + }, + { + "epoch": 0.44965, + "grad_norm": 1.5146085771718254, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 44965 + }, + { + "epoch": 0.44966, + "grad_norm": 1.4170631529883193, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 44966 + }, + { + "epoch": 0.44967, + "grad_norm": 1.4335345633816743, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 44967 + }, + { + "epoch": 0.44968, + "grad_norm": 1.7445966713954721, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 44968 + }, + { + "epoch": 0.44969, + "grad_norm": 1.4702475174415617, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 44969 + }, + { + "epoch": 0.4497, + "grad_norm": 1.4037910425096916, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 44970 + }, + { + "epoch": 0.44971, + "grad_norm": 1.2814621436507714, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 44971 + }, + { + "epoch": 0.44972, + "grad_norm": 1.6590053895747094, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 44972 + }, + { + "epoch": 0.44973, + "grad_norm": 1.492558320175927, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 44973 + }, + { + "epoch": 0.44974, + "grad_norm": 2.2619125957462676, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 44974 + }, + { + "epoch": 0.44975, + "grad_norm": 1.4257431886060674, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 44975 + }, + { + "epoch": 0.44976, + "grad_norm": 1.4539794583759316, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 44976 + }, + { + "epoch": 0.44977, + "grad_norm": 2.0575768893351656, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 44977 + }, + { + "epoch": 0.44978, + "grad_norm": 1.018327033197792, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44978 + }, + { + "epoch": 0.44979, + "grad_norm": 1.9027733143355408, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 44979 + }, + { + "epoch": 0.4498, + "grad_norm": 1.4736446970504329, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 44980 + }, + { + "epoch": 0.44981, + "grad_norm": 1.7356919690103592, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 44981 + }, + { + "epoch": 0.44982, + "grad_norm": 1.5187966868294702, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 44982 + }, + { + "epoch": 0.44983, + "grad_norm": 1.3560804715204766, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 44983 + }, + { + "epoch": 0.44984, + "grad_norm": 1.506806948374962, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 44984 + }, + { + "epoch": 0.44985, + "grad_norm": 1.542445165324976, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 44985 + }, + { + "epoch": 0.44986, + "grad_norm": 1.399304424526552, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 44986 + }, + { + "epoch": 0.44987, + "grad_norm": 1.6351346786191763, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 44987 + }, + { + "epoch": 0.44988, + "grad_norm": 1.5613101557024256, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 44988 + }, + { + "epoch": 0.44989, + "grad_norm": 1.622221283942386, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 44989 + }, + { + "epoch": 0.4499, + "grad_norm": 1.3983177768325035, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 44990 + }, + { + "epoch": 0.44991, + "grad_norm": 1.7312070116146598, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 44991 + }, + { + "epoch": 0.44992, + "grad_norm": 1.628579519301574, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 44992 + }, + { + "epoch": 0.44993, + "grad_norm": 1.5829523132078314, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 44993 + }, + { + "epoch": 0.44994, + "grad_norm": 1.430958179768539, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 44994 + }, + { + "epoch": 0.44995, + "grad_norm": 1.3184708003793384, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 44995 + }, + { + "epoch": 0.44996, + "grad_norm": 1.8079080212770642, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 44996 + }, + { + "epoch": 0.44997, + "grad_norm": 1.8171289201213836, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 44997 + }, + { + "epoch": 0.44998, + "grad_norm": 1.2928197149256067, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 44998 + }, + { + "epoch": 0.44999, + "grad_norm": 1.55935913780341, + "learning_rate": 0.003, + "loss": 4.019, + "step": 44999 + }, + { + "epoch": 0.45, + "grad_norm": 1.2587729632830624, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45000 + }, + { + "epoch": 0.45001, + "grad_norm": 1.6311473362493392, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 45001 + }, + { + "epoch": 0.45002, + "grad_norm": 1.442668946295592, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45002 + }, + { + "epoch": 0.45003, + "grad_norm": 1.764483096564823, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45003 + }, + { + "epoch": 0.45004, + "grad_norm": 1.2465448563926294, + "learning_rate": 0.003, + "loss": 4.048, + "step": 45004 + }, + { + "epoch": 0.45005, + "grad_norm": 1.6869753529117395, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 45005 + }, + { + "epoch": 0.45006, + "grad_norm": 1.185088712106774, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 45006 + }, + { + "epoch": 0.45007, + "grad_norm": 1.7878906085240263, + "learning_rate": 0.003, + "loss": 4.066, + "step": 45007 + }, + { + "epoch": 0.45008, + "grad_norm": 1.541236992649454, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45008 + }, + { + "epoch": 0.45009, + "grad_norm": 1.9015825355915408, + "learning_rate": 0.003, + "loss": 4.021, + "step": 45009 + }, + { + "epoch": 0.4501, + "grad_norm": 1.4218572703759782, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 45010 + }, + { + "epoch": 0.45011, + "grad_norm": 1.3504899149436655, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 45011 + }, + { + "epoch": 0.45012, + "grad_norm": 1.4853842784445035, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 45012 + }, + { + "epoch": 0.45013, + "grad_norm": 1.7462173067369398, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45013 + }, + { + "epoch": 0.45014, + "grad_norm": 1.2513675661470889, + "learning_rate": 0.003, + "loss": 4.049, + "step": 45014 + }, + { + "epoch": 0.45015, + "grad_norm": 2.001099831048338, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 45015 + }, + { + "epoch": 0.45016, + "grad_norm": 1.2740039198881725, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45016 + }, + { + "epoch": 0.45017, + "grad_norm": 1.9947579378712514, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 45017 + }, + { + "epoch": 0.45018, + "grad_norm": 1.5141635135681695, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 45018 + }, + { + "epoch": 0.45019, + "grad_norm": 1.6767739130524013, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 45019 + }, + { + "epoch": 0.4502, + "grad_norm": 1.1237382126517725, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45020 + }, + { + "epoch": 0.45021, + "grad_norm": 1.9358966359139798, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 45021 + }, + { + "epoch": 0.45022, + "grad_norm": 1.2526737409005093, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 45022 + }, + { + "epoch": 0.45023, + "grad_norm": 1.7897132363496495, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 45023 + }, + { + "epoch": 0.45024, + "grad_norm": 1.382783062528921, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 45024 + }, + { + "epoch": 0.45025, + "grad_norm": 1.6033611525854792, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 45025 + }, + { + "epoch": 0.45026, + "grad_norm": 1.4066937275559017, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 45026 + }, + { + "epoch": 0.45027, + "grad_norm": 1.721381146407328, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 45027 + }, + { + "epoch": 0.45028, + "grad_norm": 1.3493232829373711, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 45028 + }, + { + "epoch": 0.45029, + "grad_norm": 1.800610699274367, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45029 + }, + { + "epoch": 0.4503, + "grad_norm": 1.5228520455224908, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 45030 + }, + { + "epoch": 0.45031, + "grad_norm": 1.5776086553416646, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 45031 + }, + { + "epoch": 0.45032, + "grad_norm": 1.351882064383698, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45032 + }, + { + "epoch": 0.45033, + "grad_norm": 1.6367022889036809, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45033 + }, + { + "epoch": 0.45034, + "grad_norm": 1.5844447547690315, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45034 + }, + { + "epoch": 0.45035, + "grad_norm": 1.5389801117886626, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 45035 + }, + { + "epoch": 0.45036, + "grad_norm": 1.1932155508622735, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 45036 + }, + { + "epoch": 0.45037, + "grad_norm": 1.8639642631576867, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 45037 + }, + { + "epoch": 0.45038, + "grad_norm": 1.1312912557004424, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45038 + }, + { + "epoch": 0.45039, + "grad_norm": 1.887105672596334, + "learning_rate": 0.003, + "loss": 4.062, + "step": 45039 + }, + { + "epoch": 0.4504, + "grad_norm": 1.326140518696186, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 45040 + }, + { + "epoch": 0.45041, + "grad_norm": 1.4193285543100567, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 45041 + }, + { + "epoch": 0.45042, + "grad_norm": 1.5681321742597811, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 45042 + }, + { + "epoch": 0.45043, + "grad_norm": 1.8392851000445465, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 45043 + }, + { + "epoch": 0.45044, + "grad_norm": 1.4043432443297217, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 45044 + }, + { + "epoch": 0.45045, + "grad_norm": 1.4575015817932107, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45045 + }, + { + "epoch": 0.45046, + "grad_norm": 1.8707769129305785, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 45046 + }, + { + "epoch": 0.45047, + "grad_norm": 1.5620539780458536, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 45047 + }, + { + "epoch": 0.45048, + "grad_norm": 1.3958710832272143, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45048 + }, + { + "epoch": 0.45049, + "grad_norm": 1.5801765442355538, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 45049 + }, + { + "epoch": 0.4505, + "grad_norm": 1.241215975897547, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 45050 + }, + { + "epoch": 0.45051, + "grad_norm": 1.393942985224251, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 45051 + }, + { + "epoch": 0.45052, + "grad_norm": 1.654141361225673, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 45052 + }, + { + "epoch": 0.45053, + "grad_norm": 1.4785151774562708, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 45053 + }, + { + "epoch": 0.45054, + "grad_norm": 1.689101791707478, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 45054 + }, + { + "epoch": 0.45055, + "grad_norm": 2.085243489621638, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 45055 + }, + { + "epoch": 0.45056, + "grad_norm": 1.2135022327386495, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 45056 + }, + { + "epoch": 0.45057, + "grad_norm": 1.726441851952747, + "learning_rate": 0.003, + "loss": 4.044, + "step": 45057 + }, + { + "epoch": 0.45058, + "grad_norm": 1.7141959911538849, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 45058 + }, + { + "epoch": 0.45059, + "grad_norm": 1.4562439913010927, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 45059 + }, + { + "epoch": 0.4506, + "grad_norm": 1.5148022504073997, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 45060 + }, + { + "epoch": 0.45061, + "grad_norm": 1.5126128013633535, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45061 + }, + { + "epoch": 0.45062, + "grad_norm": 1.568087773029704, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45062 + }, + { + "epoch": 0.45063, + "grad_norm": 1.208996351003204, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 45063 + }, + { + "epoch": 0.45064, + "grad_norm": 1.8405302733355353, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 45064 + }, + { + "epoch": 0.45065, + "grad_norm": 1.027931698396841, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 45065 + }, + { + "epoch": 0.45066, + "grad_norm": 2.0426767504715326, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 45066 + }, + { + "epoch": 0.45067, + "grad_norm": 1.4713863492928796, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 45067 + }, + { + "epoch": 0.45068, + "grad_norm": 1.696595120401832, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 45068 + }, + { + "epoch": 0.45069, + "grad_norm": 1.723241917239435, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 45069 + }, + { + "epoch": 0.4507, + "grad_norm": 1.7624400214113547, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 45070 + }, + { + "epoch": 0.45071, + "grad_norm": 1.5271014164100818, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 45071 + }, + { + "epoch": 0.45072, + "grad_norm": 1.396096035144493, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45072 + }, + { + "epoch": 0.45073, + "grad_norm": 1.4000551481343333, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 45073 + }, + { + "epoch": 0.45074, + "grad_norm": 1.6908228696916134, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 45074 + }, + { + "epoch": 0.45075, + "grad_norm": 1.2613106922223893, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45075 + }, + { + "epoch": 0.45076, + "grad_norm": 1.819526104884243, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 45076 + }, + { + "epoch": 0.45077, + "grad_norm": 1.159957460454931, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45077 + }, + { + "epoch": 0.45078, + "grad_norm": 1.8719168895487819, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 45078 + }, + { + "epoch": 0.45079, + "grad_norm": 1.4936064875221635, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 45079 + }, + { + "epoch": 0.4508, + "grad_norm": 2.0911100227160757, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45080 + }, + { + "epoch": 0.45081, + "grad_norm": 1.4239061001505071, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 45081 + }, + { + "epoch": 0.45082, + "grad_norm": 1.546968755581241, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 45082 + }, + { + "epoch": 0.45083, + "grad_norm": 1.8564726574198716, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45083 + }, + { + "epoch": 0.45084, + "grad_norm": 1.2130428173034935, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 45084 + }, + { + "epoch": 0.45085, + "grad_norm": 1.6497290454165592, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45085 + }, + { + "epoch": 0.45086, + "grad_norm": 1.317238024853342, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 45086 + }, + { + "epoch": 0.45087, + "grad_norm": 1.3552193450025594, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 45087 + }, + { + "epoch": 0.45088, + "grad_norm": 1.7097867410935537, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 45088 + }, + { + "epoch": 0.45089, + "grad_norm": 1.367316376877461, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 45089 + }, + { + "epoch": 0.4509, + "grad_norm": 1.5839243820306124, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 45090 + }, + { + "epoch": 0.45091, + "grad_norm": 1.307368405058803, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45091 + }, + { + "epoch": 0.45092, + "grad_norm": 1.4192572270636536, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 45092 + }, + { + "epoch": 0.45093, + "grad_norm": 1.3624910790968945, + "learning_rate": 0.003, + "loss": 4.041, + "step": 45093 + }, + { + "epoch": 0.45094, + "grad_norm": 1.531403950238321, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 45094 + }, + { + "epoch": 0.45095, + "grad_norm": 1.8241679541217708, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 45095 + }, + { + "epoch": 0.45096, + "grad_norm": 1.6320021910902152, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 45096 + }, + { + "epoch": 0.45097, + "grad_norm": 1.8845326803605003, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 45097 + }, + { + "epoch": 0.45098, + "grad_norm": 1.5488128578532148, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 45098 + }, + { + "epoch": 0.45099, + "grad_norm": 1.4797841418980795, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 45099 + }, + { + "epoch": 0.451, + "grad_norm": 1.6653599511432082, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45100 + }, + { + "epoch": 0.45101, + "grad_norm": 1.3702253909990065, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 45101 + }, + { + "epoch": 0.45102, + "grad_norm": 1.3404268767217042, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45102 + }, + { + "epoch": 0.45103, + "grad_norm": 2.017157925551258, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45103 + }, + { + "epoch": 0.45104, + "grad_norm": 1.284800345133153, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 45104 + }, + { + "epoch": 0.45105, + "grad_norm": 1.7178203195331452, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 45105 + }, + { + "epoch": 0.45106, + "grad_norm": 1.4539090051664054, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 45106 + }, + { + "epoch": 0.45107, + "grad_norm": 1.544256716515676, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 45107 + }, + { + "epoch": 0.45108, + "grad_norm": 1.397378800765754, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45108 + }, + { + "epoch": 0.45109, + "grad_norm": 1.666208686344773, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 45109 + }, + { + "epoch": 0.4511, + "grad_norm": 1.459575000018783, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45110 + }, + { + "epoch": 0.45111, + "grad_norm": 1.358486102258725, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 45111 + }, + { + "epoch": 0.45112, + "grad_norm": 1.343145411008069, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 45112 + }, + { + "epoch": 0.45113, + "grad_norm": 1.5358957166888816, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 45113 + }, + { + "epoch": 0.45114, + "grad_norm": 1.7734762338637888, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 45114 + }, + { + "epoch": 0.45115, + "grad_norm": 1.6441041991995031, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45115 + }, + { + "epoch": 0.45116, + "grad_norm": 1.5253495253285532, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 45116 + }, + { + "epoch": 0.45117, + "grad_norm": 1.412545735008927, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 45117 + }, + { + "epoch": 0.45118, + "grad_norm": 1.4947787736493692, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45118 + }, + { + "epoch": 0.45119, + "grad_norm": 1.505663146869972, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45119 + }, + { + "epoch": 0.4512, + "grad_norm": 1.48483406763116, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45120 + }, + { + "epoch": 0.45121, + "grad_norm": 1.4261889883184151, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 45121 + }, + { + "epoch": 0.45122, + "grad_norm": 1.7823200523743838, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 45122 + }, + { + "epoch": 0.45123, + "grad_norm": 1.465481360555586, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45123 + }, + { + "epoch": 0.45124, + "grad_norm": 1.690182616148757, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 45124 + }, + { + "epoch": 0.45125, + "grad_norm": 1.6492125642306903, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 45125 + }, + { + "epoch": 0.45126, + "grad_norm": 1.7803422149030155, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 45126 + }, + { + "epoch": 0.45127, + "grad_norm": 1.2751409694918392, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 45127 + }, + { + "epoch": 0.45128, + "grad_norm": 1.62148714860407, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 45128 + }, + { + "epoch": 0.45129, + "grad_norm": 1.5867282924713795, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 45129 + }, + { + "epoch": 0.4513, + "grad_norm": 1.5833317201272892, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45130 + }, + { + "epoch": 0.45131, + "grad_norm": 1.3694551033454128, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45131 + }, + { + "epoch": 0.45132, + "grad_norm": 1.5230693101150108, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45132 + }, + { + "epoch": 0.45133, + "grad_norm": 1.3397752189995293, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45133 + }, + { + "epoch": 0.45134, + "grad_norm": 1.5409328354575236, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45134 + }, + { + "epoch": 0.45135, + "grad_norm": 1.475933019617754, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 45135 + }, + { + "epoch": 0.45136, + "grad_norm": 1.787895297682561, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 45136 + }, + { + "epoch": 0.45137, + "grad_norm": 1.3488794317094597, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45137 + }, + { + "epoch": 0.45138, + "grad_norm": 1.7992188689436368, + "learning_rate": 0.003, + "loss": 4.035, + "step": 45138 + }, + { + "epoch": 0.45139, + "grad_norm": 1.800575658346675, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 45139 + }, + { + "epoch": 0.4514, + "grad_norm": 1.3337793244565002, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 45140 + }, + { + "epoch": 0.45141, + "grad_norm": 1.6786151363120385, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 45141 + }, + { + "epoch": 0.45142, + "grad_norm": 1.6310419647891352, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 45142 + }, + { + "epoch": 0.45143, + "grad_norm": 1.5068489228445576, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 45143 + }, + { + "epoch": 0.45144, + "grad_norm": 1.490461217113405, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45144 + }, + { + "epoch": 0.45145, + "grad_norm": 1.5803279184031946, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45145 + }, + { + "epoch": 0.45146, + "grad_norm": 1.3575813421230456, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 45146 + }, + { + "epoch": 0.45147, + "grad_norm": 1.7568196993634853, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 45147 + }, + { + "epoch": 0.45148, + "grad_norm": 1.3365847444120689, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 45148 + }, + { + "epoch": 0.45149, + "grad_norm": 1.911838164314606, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45149 + }, + { + "epoch": 0.4515, + "grad_norm": 1.4377006579843243, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 45150 + }, + { + "epoch": 0.45151, + "grad_norm": 1.6365563527364726, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 45151 + }, + { + "epoch": 0.45152, + "grad_norm": 1.473608802090768, + "learning_rate": 0.003, + "loss": 4.044, + "step": 45152 + }, + { + "epoch": 0.45153, + "grad_norm": 1.4423680125146199, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 45153 + }, + { + "epoch": 0.45154, + "grad_norm": 1.5289618120459079, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 45154 + }, + { + "epoch": 0.45155, + "grad_norm": 1.5900786023690885, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 45155 + }, + { + "epoch": 0.45156, + "grad_norm": 1.44008463582468, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 45156 + }, + { + "epoch": 0.45157, + "grad_norm": 1.4206537057113946, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45157 + }, + { + "epoch": 0.45158, + "grad_norm": 1.5601091124169821, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 45158 + }, + { + "epoch": 0.45159, + "grad_norm": 1.6223068011253228, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45159 + }, + { + "epoch": 0.4516, + "grad_norm": 1.8838682715614419, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 45160 + }, + { + "epoch": 0.45161, + "grad_norm": 1.2640903998716444, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 45161 + }, + { + "epoch": 0.45162, + "grad_norm": 1.7004759460985166, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 45162 + }, + { + "epoch": 0.45163, + "grad_norm": 1.637125684604008, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45163 + }, + { + "epoch": 0.45164, + "grad_norm": 1.5191512799477545, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 45164 + }, + { + "epoch": 0.45165, + "grad_norm": 1.4339802849193237, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45165 + }, + { + "epoch": 0.45166, + "grad_norm": 1.4839477015196236, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 45166 + }, + { + "epoch": 0.45167, + "grad_norm": 1.5134243811974188, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 45167 + }, + { + "epoch": 0.45168, + "grad_norm": 1.4556880343133294, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 45168 + }, + { + "epoch": 0.45169, + "grad_norm": 1.600508998277885, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 45169 + }, + { + "epoch": 0.4517, + "grad_norm": 1.4844268487447374, + "learning_rate": 0.003, + "loss": 4.048, + "step": 45170 + }, + { + "epoch": 0.45171, + "grad_norm": 1.4770412390622114, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 45171 + }, + { + "epoch": 0.45172, + "grad_norm": 1.733137645757142, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 45172 + }, + { + "epoch": 0.45173, + "grad_norm": 1.5999232628424602, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 45173 + }, + { + "epoch": 0.45174, + "grad_norm": 1.5292609571965197, + "learning_rate": 0.003, + "loss": 4.057, + "step": 45174 + }, + { + "epoch": 0.45175, + "grad_norm": 1.4643001790306296, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 45175 + }, + { + "epoch": 0.45176, + "grad_norm": 1.417908997811547, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 45176 + }, + { + "epoch": 0.45177, + "grad_norm": 1.7424376936860788, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 45177 + }, + { + "epoch": 0.45178, + "grad_norm": 1.4174321963674577, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 45178 + }, + { + "epoch": 0.45179, + "grad_norm": 1.8112541767290105, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 45179 + }, + { + "epoch": 0.4518, + "grad_norm": 1.2468848012459341, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45180 + }, + { + "epoch": 0.45181, + "grad_norm": 1.6616011897733132, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 45181 + }, + { + "epoch": 0.45182, + "grad_norm": 1.5126868483701292, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45182 + }, + { + "epoch": 0.45183, + "grad_norm": 1.5471749320388997, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 45183 + }, + { + "epoch": 0.45184, + "grad_norm": 1.760265247679297, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 45184 + }, + { + "epoch": 0.45185, + "grad_norm": 1.253151049493427, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45185 + }, + { + "epoch": 0.45186, + "grad_norm": 1.9575130062158688, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 45186 + }, + { + "epoch": 0.45187, + "grad_norm": 1.3669532350673397, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45187 + }, + { + "epoch": 0.45188, + "grad_norm": 1.867247473060493, + "learning_rate": 0.003, + "loss": 4.055, + "step": 45188 + }, + { + "epoch": 0.45189, + "grad_norm": 1.2906058342550535, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45189 + }, + { + "epoch": 0.4519, + "grad_norm": 1.5624655274950014, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45190 + }, + { + "epoch": 0.45191, + "grad_norm": 1.7598502868126797, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 45191 + }, + { + "epoch": 0.45192, + "grad_norm": 1.4432033734402325, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 45192 + }, + { + "epoch": 0.45193, + "grad_norm": 1.4971528219073231, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 45193 + }, + { + "epoch": 0.45194, + "grad_norm": 1.6440511885057119, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 45194 + }, + { + "epoch": 0.45195, + "grad_norm": 1.5032278274096977, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45195 + }, + { + "epoch": 0.45196, + "grad_norm": 1.6621802061705204, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45196 + }, + { + "epoch": 0.45197, + "grad_norm": 1.7120746371972908, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 45197 + }, + { + "epoch": 0.45198, + "grad_norm": 1.3977737721020347, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 45198 + }, + { + "epoch": 0.45199, + "grad_norm": 1.3716868071348667, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45199 + }, + { + "epoch": 0.452, + "grad_norm": 1.8179413963023403, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45200 + }, + { + "epoch": 0.45201, + "grad_norm": 1.358784692265395, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 45201 + }, + { + "epoch": 0.45202, + "grad_norm": 1.76891103885333, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45202 + }, + { + "epoch": 0.45203, + "grad_norm": 1.6181306175214358, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45203 + }, + { + "epoch": 0.45204, + "grad_norm": 1.2527521293881008, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 45204 + }, + { + "epoch": 0.45205, + "grad_norm": 1.440619473253261, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 45205 + }, + { + "epoch": 0.45206, + "grad_norm": 1.5598935306026782, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45206 + }, + { + "epoch": 0.45207, + "grad_norm": 1.5058785010004765, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45207 + }, + { + "epoch": 0.45208, + "grad_norm": 1.5656347373840147, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 45208 + }, + { + "epoch": 0.45209, + "grad_norm": 1.6041166675972516, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45209 + }, + { + "epoch": 0.4521, + "grad_norm": 1.9593686513522457, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 45210 + }, + { + "epoch": 0.45211, + "grad_norm": 1.0329871674032818, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 45211 + }, + { + "epoch": 0.45212, + "grad_norm": 1.6413173478626768, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45212 + }, + { + "epoch": 0.45213, + "grad_norm": 1.3958037243134533, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 45213 + }, + { + "epoch": 0.45214, + "grad_norm": 1.8551032844614352, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 45214 + }, + { + "epoch": 0.45215, + "grad_norm": 1.6226286004782275, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 45215 + }, + { + "epoch": 0.45216, + "grad_norm": 1.306335759401064, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45216 + }, + { + "epoch": 0.45217, + "grad_norm": 1.5790060133346153, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45217 + }, + { + "epoch": 0.45218, + "grad_norm": 1.4192918481271735, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45218 + }, + { + "epoch": 0.45219, + "grad_norm": 1.3582053778387047, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 45219 + }, + { + "epoch": 0.4522, + "grad_norm": 1.731921789933817, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 45220 + }, + { + "epoch": 0.45221, + "grad_norm": 1.6151598830374398, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 45221 + }, + { + "epoch": 0.45222, + "grad_norm": 1.8365025129559107, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 45222 + }, + { + "epoch": 0.45223, + "grad_norm": 1.4112390280444984, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 45223 + }, + { + "epoch": 0.45224, + "grad_norm": 1.5210659914249145, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45224 + }, + { + "epoch": 0.45225, + "grad_norm": 1.3973106524305914, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 45225 + }, + { + "epoch": 0.45226, + "grad_norm": 1.6675302837286088, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 45226 + }, + { + "epoch": 0.45227, + "grad_norm": 1.3186756359777958, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 45227 + }, + { + "epoch": 0.45228, + "grad_norm": 1.6292682805217216, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 45228 + }, + { + "epoch": 0.45229, + "grad_norm": 1.5225284954583413, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 45229 + }, + { + "epoch": 0.4523, + "grad_norm": 1.5995859542829138, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 45230 + }, + { + "epoch": 0.45231, + "grad_norm": 1.443097960274921, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45231 + }, + { + "epoch": 0.45232, + "grad_norm": 1.7810148914647062, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 45232 + }, + { + "epoch": 0.45233, + "grad_norm": 1.7431598973710294, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 45233 + }, + { + "epoch": 0.45234, + "grad_norm": 1.6768685352666484, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 45234 + }, + { + "epoch": 0.45235, + "grad_norm": 1.68679097835591, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 45235 + }, + { + "epoch": 0.45236, + "grad_norm": 1.5114069283286606, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 45236 + }, + { + "epoch": 0.45237, + "grad_norm": 1.308491136476397, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 45237 + }, + { + "epoch": 0.45238, + "grad_norm": 1.524510483715716, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 45238 + }, + { + "epoch": 0.45239, + "grad_norm": 1.3630184248668211, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 45239 + }, + { + "epoch": 0.4524, + "grad_norm": 1.5527998986806724, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45240 + }, + { + "epoch": 0.45241, + "grad_norm": 1.5866971588231944, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 45241 + }, + { + "epoch": 0.45242, + "grad_norm": 1.4732113270842508, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 45242 + }, + { + "epoch": 0.45243, + "grad_norm": 1.6652330275010254, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 45243 + }, + { + "epoch": 0.45244, + "grad_norm": 1.6237578516774367, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 45244 + }, + { + "epoch": 0.45245, + "grad_norm": 1.2634560965648456, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 45245 + }, + { + "epoch": 0.45246, + "grad_norm": 1.8383933406109212, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 45246 + }, + { + "epoch": 0.45247, + "grad_norm": 1.2076640202660296, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 45247 + }, + { + "epoch": 0.45248, + "grad_norm": 1.9453662678429355, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 45248 + }, + { + "epoch": 0.45249, + "grad_norm": 1.4823646581585574, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45249 + }, + { + "epoch": 0.4525, + "grad_norm": 1.6685722564364038, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 45250 + }, + { + "epoch": 0.45251, + "grad_norm": 1.3999333588345964, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 45251 + }, + { + "epoch": 0.45252, + "grad_norm": 1.4807170264538398, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 45252 + }, + { + "epoch": 0.45253, + "grad_norm": 1.3669455062143532, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 45253 + }, + { + "epoch": 0.45254, + "grad_norm": 1.3558550610404512, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 45254 + }, + { + "epoch": 0.45255, + "grad_norm": 1.4502966640036383, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 45255 + }, + { + "epoch": 0.45256, + "grad_norm": 1.3287649503811023, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45256 + }, + { + "epoch": 0.45257, + "grad_norm": 1.7232894590434278, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 45257 + }, + { + "epoch": 0.45258, + "grad_norm": 1.5498357116934973, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 45258 + }, + { + "epoch": 0.45259, + "grad_norm": 1.7124347267433568, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45259 + }, + { + "epoch": 0.4526, + "grad_norm": 1.526808843244016, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45260 + }, + { + "epoch": 0.45261, + "grad_norm": 1.381465721345365, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 45261 + }, + { + "epoch": 0.45262, + "grad_norm": 1.6916348129082934, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 45262 + }, + { + "epoch": 0.45263, + "grad_norm": 1.7463273812960123, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 45263 + }, + { + "epoch": 0.45264, + "grad_norm": 1.3721147951366994, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 45264 + }, + { + "epoch": 0.45265, + "grad_norm": 1.6053795408713165, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45265 + }, + { + "epoch": 0.45266, + "grad_norm": 1.4382107352038174, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 45266 + }, + { + "epoch": 0.45267, + "grad_norm": 1.8033758433477765, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 45267 + }, + { + "epoch": 0.45268, + "grad_norm": 1.4680122084703888, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 45268 + }, + { + "epoch": 0.45269, + "grad_norm": 1.557822377066003, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 45269 + }, + { + "epoch": 0.4527, + "grad_norm": 1.4987601502669596, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 45270 + }, + { + "epoch": 0.45271, + "grad_norm": 1.4902003606206657, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 45271 + }, + { + "epoch": 0.45272, + "grad_norm": 1.5456698442105303, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45272 + }, + { + "epoch": 0.45273, + "grad_norm": 1.4206538520546905, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 45273 + }, + { + "epoch": 0.45274, + "grad_norm": 1.426270194217412, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 45274 + }, + { + "epoch": 0.45275, + "grad_norm": 1.8059188665515764, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 45275 + }, + { + "epoch": 0.45276, + "grad_norm": 1.4886790961840808, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 45276 + }, + { + "epoch": 0.45277, + "grad_norm": 1.4732207494607592, + "learning_rate": 0.003, + "loss": 4.026, + "step": 45277 + }, + { + "epoch": 0.45278, + "grad_norm": 1.765922777125414, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 45278 + }, + { + "epoch": 0.45279, + "grad_norm": 1.2707636486910265, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 45279 + }, + { + "epoch": 0.4528, + "grad_norm": 2.1529215943164757, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 45280 + }, + { + "epoch": 0.45281, + "grad_norm": 1.1784833879247407, + "learning_rate": 0.003, + "loss": 4.054, + "step": 45281 + }, + { + "epoch": 0.45282, + "grad_norm": 2.0773971246361795, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 45282 + }, + { + "epoch": 0.45283, + "grad_norm": 1.4093838619107792, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 45283 + }, + { + "epoch": 0.45284, + "grad_norm": 1.9841283579648343, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 45284 + }, + { + "epoch": 0.45285, + "grad_norm": 1.662508739799041, + "learning_rate": 0.003, + "loss": 4.057, + "step": 45285 + }, + { + "epoch": 0.45286, + "grad_norm": 1.4726952043396682, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 45286 + }, + { + "epoch": 0.45287, + "grad_norm": 1.548731307344163, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 45287 + }, + { + "epoch": 0.45288, + "grad_norm": 1.413707587130502, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 45288 + }, + { + "epoch": 0.45289, + "grad_norm": 1.5174325318835382, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 45289 + }, + { + "epoch": 0.4529, + "grad_norm": 1.4462363789027162, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45290 + }, + { + "epoch": 0.45291, + "grad_norm": 1.573785445429943, + "learning_rate": 0.003, + "loss": 4.074, + "step": 45291 + }, + { + "epoch": 0.45292, + "grad_norm": 1.3718000713535283, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45292 + }, + { + "epoch": 0.45293, + "grad_norm": 1.4334788550257993, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45293 + }, + { + "epoch": 0.45294, + "grad_norm": 1.620956189783868, + "learning_rate": 0.003, + "loss": 4.018, + "step": 45294 + }, + { + "epoch": 0.45295, + "grad_norm": 2.16210767083711, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45295 + }, + { + "epoch": 0.45296, + "grad_norm": 1.4542029884153689, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45296 + }, + { + "epoch": 0.45297, + "grad_norm": 1.3823605159879853, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45297 + }, + { + "epoch": 0.45298, + "grad_norm": 1.8513751467015425, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45298 + }, + { + "epoch": 0.45299, + "grad_norm": 1.247547007288254, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 45299 + }, + { + "epoch": 0.453, + "grad_norm": 1.4523866847489768, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 45300 + }, + { + "epoch": 0.45301, + "grad_norm": 1.7603061240185904, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 45301 + }, + { + "epoch": 0.45302, + "grad_norm": 1.3226637184286152, + "learning_rate": 0.003, + "loss": 4.056, + "step": 45302 + }, + { + "epoch": 0.45303, + "grad_norm": 1.5015300510545282, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 45303 + }, + { + "epoch": 0.45304, + "grad_norm": 1.6829600788332242, + "learning_rate": 0.003, + "loss": 4.039, + "step": 45304 + }, + { + "epoch": 0.45305, + "grad_norm": 1.5383571014227626, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 45305 + }, + { + "epoch": 0.45306, + "grad_norm": 1.8635150498423856, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 45306 + }, + { + "epoch": 0.45307, + "grad_norm": 1.453279102392935, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 45307 + }, + { + "epoch": 0.45308, + "grad_norm": 1.5191903237718616, + "learning_rate": 0.003, + "loss": 4.051, + "step": 45308 + }, + { + "epoch": 0.45309, + "grad_norm": 1.233745874819195, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 45309 + }, + { + "epoch": 0.4531, + "grad_norm": 1.6971127331587403, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 45310 + }, + { + "epoch": 0.45311, + "grad_norm": 1.2696328218073252, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 45311 + }, + { + "epoch": 0.45312, + "grad_norm": 1.5232081326347022, + "learning_rate": 0.003, + "loss": 4.03, + "step": 45312 + }, + { + "epoch": 0.45313, + "grad_norm": 1.405316876054064, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 45313 + }, + { + "epoch": 0.45314, + "grad_norm": 1.4850572095355363, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 45314 + }, + { + "epoch": 0.45315, + "grad_norm": 1.7781235965374416, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 45315 + }, + { + "epoch": 0.45316, + "grad_norm": 1.5009924235457888, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45316 + }, + { + "epoch": 0.45317, + "grad_norm": 1.46458527042522, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 45317 + }, + { + "epoch": 0.45318, + "grad_norm": 1.4785199706885999, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45318 + }, + { + "epoch": 0.45319, + "grad_norm": 1.6805131659608765, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 45319 + }, + { + "epoch": 0.4532, + "grad_norm": 1.445813745339767, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 45320 + }, + { + "epoch": 0.45321, + "grad_norm": 1.4465634981792654, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45321 + }, + { + "epoch": 0.45322, + "grad_norm": 1.6884159119731166, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 45322 + }, + { + "epoch": 0.45323, + "grad_norm": 1.7778365039341488, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 45323 + }, + { + "epoch": 0.45324, + "grad_norm": 1.5454413789246084, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45324 + }, + { + "epoch": 0.45325, + "grad_norm": 1.5757739271292694, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 45325 + }, + { + "epoch": 0.45326, + "grad_norm": 1.5341861394827787, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 45326 + }, + { + "epoch": 0.45327, + "grad_norm": 1.356913344323924, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45327 + }, + { + "epoch": 0.45328, + "grad_norm": 1.6208645987804748, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 45328 + }, + { + "epoch": 0.45329, + "grad_norm": 1.3636009807690956, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45329 + }, + { + "epoch": 0.4533, + "grad_norm": 1.593664833069025, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 45330 + }, + { + "epoch": 0.45331, + "grad_norm": 1.6952224021074078, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 45331 + }, + { + "epoch": 0.45332, + "grad_norm": 1.6059453609216428, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 45332 + }, + { + "epoch": 0.45333, + "grad_norm": 1.543956053948877, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 45333 + }, + { + "epoch": 0.45334, + "grad_norm": 1.542537107168781, + "learning_rate": 0.003, + "loss": 3.99, + "step": 45334 + }, + { + "epoch": 0.45335, + "grad_norm": 1.4391394842600098, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45335 + }, + { + "epoch": 0.45336, + "grad_norm": 1.4232717190841158, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 45336 + }, + { + "epoch": 0.45337, + "grad_norm": 1.52248474368029, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 45337 + }, + { + "epoch": 0.45338, + "grad_norm": 1.3820110554879852, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45338 + }, + { + "epoch": 0.45339, + "grad_norm": 1.5414258879207199, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 45339 + }, + { + "epoch": 0.4534, + "grad_norm": 1.7360765189244927, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45340 + }, + { + "epoch": 0.45341, + "grad_norm": 2.0156364455558733, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 45341 + }, + { + "epoch": 0.45342, + "grad_norm": 1.5301385668709113, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45342 + }, + { + "epoch": 0.45343, + "grad_norm": 1.5247359122591013, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 45343 + }, + { + "epoch": 0.45344, + "grad_norm": 1.667552942329734, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45344 + }, + { + "epoch": 0.45345, + "grad_norm": 1.4986622073161893, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45345 + }, + { + "epoch": 0.45346, + "grad_norm": 1.5193093610632695, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 45346 + }, + { + "epoch": 0.45347, + "grad_norm": 1.3466853776519072, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 45347 + }, + { + "epoch": 0.45348, + "grad_norm": 1.5529450021733349, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 45348 + }, + { + "epoch": 0.45349, + "grad_norm": 1.4500663045934123, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45349 + }, + { + "epoch": 0.4535, + "grad_norm": 1.6590475422846431, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45350 + }, + { + "epoch": 0.45351, + "grad_norm": 1.503770908783963, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 45351 + }, + { + "epoch": 0.45352, + "grad_norm": 1.6175246965131196, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 45352 + }, + { + "epoch": 0.45353, + "grad_norm": 1.460040379680283, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45353 + }, + { + "epoch": 0.45354, + "grad_norm": 1.4438852062644387, + "learning_rate": 0.003, + "loss": 4.006, + "step": 45354 + }, + { + "epoch": 0.45355, + "grad_norm": 1.5031265492234664, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 45355 + }, + { + "epoch": 0.45356, + "grad_norm": 1.5587153883591442, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 45356 + }, + { + "epoch": 0.45357, + "grad_norm": 1.9339564509000202, + "learning_rate": 0.003, + "loss": 4.055, + "step": 45357 + }, + { + "epoch": 0.45358, + "grad_norm": 1.2891914210213113, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45358 + }, + { + "epoch": 0.45359, + "grad_norm": 1.5108216876388643, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 45359 + }, + { + "epoch": 0.4536, + "grad_norm": 1.655726787375827, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 45360 + }, + { + "epoch": 0.45361, + "grad_norm": 1.5146607950635682, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45361 + }, + { + "epoch": 0.45362, + "grad_norm": 1.2640588111286652, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 45362 + }, + { + "epoch": 0.45363, + "grad_norm": 1.3779457848220054, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45363 + }, + { + "epoch": 0.45364, + "grad_norm": 1.718345730763226, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45364 + }, + { + "epoch": 0.45365, + "grad_norm": 1.5488566999950957, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45365 + }, + { + "epoch": 0.45366, + "grad_norm": 1.788330352341223, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 45366 + }, + { + "epoch": 0.45367, + "grad_norm": 1.4235962436104137, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 45367 + }, + { + "epoch": 0.45368, + "grad_norm": 1.9211706495627865, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45368 + }, + { + "epoch": 0.45369, + "grad_norm": 1.3731411888359824, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45369 + }, + { + "epoch": 0.4537, + "grad_norm": 1.7523189897294902, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 45370 + }, + { + "epoch": 0.45371, + "grad_norm": 1.531328320910873, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 45371 + }, + { + "epoch": 0.45372, + "grad_norm": 1.1403850767405843, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 45372 + }, + { + "epoch": 0.45373, + "grad_norm": 1.8251985279641683, + "learning_rate": 0.003, + "loss": 4.003, + "step": 45373 + }, + { + "epoch": 0.45374, + "grad_norm": 1.4766242526462636, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45374 + }, + { + "epoch": 0.45375, + "grad_norm": 1.5604873109045785, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45375 + }, + { + "epoch": 0.45376, + "grad_norm": 1.9144520824199855, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 45376 + }, + { + "epoch": 0.45377, + "grad_norm": 1.3437617109042959, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45377 + }, + { + "epoch": 0.45378, + "grad_norm": 1.3091189884490069, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45378 + }, + { + "epoch": 0.45379, + "grad_norm": 1.5397235253746515, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 45379 + }, + { + "epoch": 0.4538, + "grad_norm": 1.3502968402332878, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 45380 + }, + { + "epoch": 0.45381, + "grad_norm": 1.724003908156055, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 45381 + }, + { + "epoch": 0.45382, + "grad_norm": 1.5431162228708277, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 45382 + }, + { + "epoch": 0.45383, + "grad_norm": 1.5888532965371958, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45383 + }, + { + "epoch": 0.45384, + "grad_norm": 1.976923975123469, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 45384 + }, + { + "epoch": 0.45385, + "grad_norm": 1.3549937304340194, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45385 + }, + { + "epoch": 0.45386, + "grad_norm": 1.7017452808439801, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45386 + }, + { + "epoch": 0.45387, + "grad_norm": 1.5939614758871827, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 45387 + }, + { + "epoch": 0.45388, + "grad_norm": 1.5788916522213716, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 45388 + }, + { + "epoch": 0.45389, + "grad_norm": 1.5180040458642872, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 45389 + }, + { + "epoch": 0.4539, + "grad_norm": 1.5752082884923808, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45390 + }, + { + "epoch": 0.45391, + "grad_norm": 1.59296997252543, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 45391 + }, + { + "epoch": 0.45392, + "grad_norm": 1.3319660914123788, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 45392 + }, + { + "epoch": 0.45393, + "grad_norm": 1.8208606289974847, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 45393 + }, + { + "epoch": 0.45394, + "grad_norm": 1.2819935866010392, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45394 + }, + { + "epoch": 0.45395, + "grad_norm": 1.8263511560890737, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45395 + }, + { + "epoch": 0.45396, + "grad_norm": 1.4795265519489633, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 45396 + }, + { + "epoch": 0.45397, + "grad_norm": 1.778291281442754, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45397 + }, + { + "epoch": 0.45398, + "grad_norm": 1.5833081885612272, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 45398 + }, + { + "epoch": 0.45399, + "grad_norm": 1.5417734768066245, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 45399 + }, + { + "epoch": 0.454, + "grad_norm": 1.5135386295441569, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45400 + }, + { + "epoch": 0.45401, + "grad_norm": 1.4456374061935244, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 45401 + }, + { + "epoch": 0.45402, + "grad_norm": 1.5273220576250532, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 45402 + }, + { + "epoch": 0.45403, + "grad_norm": 1.4093185481641863, + "learning_rate": 0.003, + "loss": 4.054, + "step": 45403 + }, + { + "epoch": 0.45404, + "grad_norm": 1.4761121086595028, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45404 + }, + { + "epoch": 0.45405, + "grad_norm": 1.7686279589754554, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 45405 + }, + { + "epoch": 0.45406, + "grad_norm": 1.358236508028355, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45406 + }, + { + "epoch": 0.45407, + "grad_norm": 1.8438019655334763, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 45407 + }, + { + "epoch": 0.45408, + "grad_norm": 1.560428911775955, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45408 + }, + { + "epoch": 0.45409, + "grad_norm": 1.3285525361951716, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 45409 + }, + { + "epoch": 0.4541, + "grad_norm": 1.5854235898189284, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 45410 + }, + { + "epoch": 0.45411, + "grad_norm": 1.4945955019543247, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 45411 + }, + { + "epoch": 0.45412, + "grad_norm": 1.4249523312359131, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 45412 + }, + { + "epoch": 0.45413, + "grad_norm": 1.5790753524422514, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 45413 + }, + { + "epoch": 0.45414, + "grad_norm": 1.4750817042062594, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45414 + }, + { + "epoch": 0.45415, + "grad_norm": 1.597149282920156, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 45415 + }, + { + "epoch": 0.45416, + "grad_norm": 1.557747488980984, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45416 + }, + { + "epoch": 0.45417, + "grad_norm": 1.652109304385766, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 45417 + }, + { + "epoch": 0.45418, + "grad_norm": 1.5287848912903368, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 45418 + }, + { + "epoch": 0.45419, + "grad_norm": 1.4504616186822796, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45419 + }, + { + "epoch": 0.4542, + "grad_norm": 1.6444855167455241, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45420 + }, + { + "epoch": 0.45421, + "grad_norm": 1.4230365276459866, + "learning_rate": 0.003, + "loss": 4.046, + "step": 45421 + }, + { + "epoch": 0.45422, + "grad_norm": 1.6432840430984827, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45422 + }, + { + "epoch": 0.45423, + "grad_norm": 1.6621845125966945, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 45423 + }, + { + "epoch": 0.45424, + "grad_norm": 1.5321727754733845, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45424 + }, + { + "epoch": 0.45425, + "grad_norm": 1.5868651263542999, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45425 + }, + { + "epoch": 0.45426, + "grad_norm": 1.5294400653734448, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 45426 + }, + { + "epoch": 0.45427, + "grad_norm": 1.5939298109305873, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 45427 + }, + { + "epoch": 0.45428, + "grad_norm": 1.8386664903645278, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 45428 + }, + { + "epoch": 0.45429, + "grad_norm": 1.396040943458957, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 45429 + }, + { + "epoch": 0.4543, + "grad_norm": 1.716738809310895, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 45430 + }, + { + "epoch": 0.45431, + "grad_norm": 1.41965956563418, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 45431 + }, + { + "epoch": 0.45432, + "grad_norm": 1.6391059149692748, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 45432 + }, + { + "epoch": 0.45433, + "grad_norm": 1.2576748944665033, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 45433 + }, + { + "epoch": 0.45434, + "grad_norm": 1.6706227080639329, + "learning_rate": 0.003, + "loss": 4.013, + "step": 45434 + }, + { + "epoch": 0.45435, + "grad_norm": 1.5441181526340835, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 45435 + }, + { + "epoch": 0.45436, + "grad_norm": 2.0178267980889277, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 45436 + }, + { + "epoch": 0.45437, + "grad_norm": 1.4981555841820338, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 45437 + }, + { + "epoch": 0.45438, + "grad_norm": 1.30228839704151, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 45438 + }, + { + "epoch": 0.45439, + "grad_norm": 1.5731951206474808, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 45439 + }, + { + "epoch": 0.4544, + "grad_norm": 1.6828549016753522, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45440 + }, + { + "epoch": 0.45441, + "grad_norm": 1.4412129304191066, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 45441 + }, + { + "epoch": 0.45442, + "grad_norm": 1.6466127665425316, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45442 + }, + { + "epoch": 0.45443, + "grad_norm": 1.21039434486272, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 45443 + }, + { + "epoch": 0.45444, + "grad_norm": 1.9137913717887356, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 45444 + }, + { + "epoch": 0.45445, + "grad_norm": 1.0249457405432774, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 45445 + }, + { + "epoch": 0.45446, + "grad_norm": 1.838731077730803, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 45446 + }, + { + "epoch": 0.45447, + "grad_norm": 1.3496798024526842, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45447 + }, + { + "epoch": 0.45448, + "grad_norm": 1.600034053051042, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45448 + }, + { + "epoch": 0.45449, + "grad_norm": 1.5980785687890964, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 45449 + }, + { + "epoch": 0.4545, + "grad_norm": 1.4562123126401691, + "learning_rate": 0.003, + "loss": 4.061, + "step": 45450 + }, + { + "epoch": 0.45451, + "grad_norm": 1.4208284488503655, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 45451 + }, + { + "epoch": 0.45452, + "grad_norm": 1.5392203832624547, + "learning_rate": 0.003, + "loss": 4.047, + "step": 45452 + }, + { + "epoch": 0.45453, + "grad_norm": 1.4696631097090929, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 45453 + }, + { + "epoch": 0.45454, + "grad_norm": 1.4272592663590116, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 45454 + }, + { + "epoch": 0.45455, + "grad_norm": 1.6816672721350323, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 45455 + }, + { + "epoch": 0.45456, + "grad_norm": 1.5155814310652191, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45456 + }, + { + "epoch": 0.45457, + "grad_norm": 1.7615840750414786, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 45457 + }, + { + "epoch": 0.45458, + "grad_norm": 1.5497389992126418, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 45458 + }, + { + "epoch": 0.45459, + "grad_norm": 1.6592878008270942, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 45459 + }, + { + "epoch": 0.4546, + "grad_norm": 1.549465190234134, + "learning_rate": 0.003, + "loss": 4.036, + "step": 45460 + }, + { + "epoch": 0.45461, + "grad_norm": 1.7222546364512976, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 45461 + }, + { + "epoch": 0.45462, + "grad_norm": 1.5075126707441233, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 45462 + }, + { + "epoch": 0.45463, + "grad_norm": 1.5391127359397938, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 45463 + }, + { + "epoch": 0.45464, + "grad_norm": 1.5953077919798417, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45464 + }, + { + "epoch": 0.45465, + "grad_norm": 1.3224340867603501, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45465 + }, + { + "epoch": 0.45466, + "grad_norm": 1.4810928978803422, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45466 + }, + { + "epoch": 0.45467, + "grad_norm": 1.4944868088763923, + "learning_rate": 0.003, + "loss": 4.057, + "step": 45467 + }, + { + "epoch": 0.45468, + "grad_norm": 1.5501614166409112, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 45468 + }, + { + "epoch": 0.45469, + "grad_norm": 1.407209971511285, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45469 + }, + { + "epoch": 0.4547, + "grad_norm": 1.6261412461456306, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 45470 + }, + { + "epoch": 0.45471, + "grad_norm": 1.318829293814701, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45471 + }, + { + "epoch": 0.45472, + "grad_norm": 1.497364937260828, + "learning_rate": 0.003, + "loss": 4.022, + "step": 45472 + }, + { + "epoch": 0.45473, + "grad_norm": 1.6486304198176027, + "learning_rate": 0.003, + "loss": 4.053, + "step": 45473 + }, + { + "epoch": 0.45474, + "grad_norm": 1.5059466868496054, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 45474 + }, + { + "epoch": 0.45475, + "grad_norm": 1.642923421079973, + "learning_rate": 0.003, + "loss": 4.051, + "step": 45475 + }, + { + "epoch": 0.45476, + "grad_norm": 1.5368688751446378, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 45476 + }, + { + "epoch": 0.45477, + "grad_norm": 1.6157201590437216, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 45477 + }, + { + "epoch": 0.45478, + "grad_norm": 1.3170310653836124, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 45478 + }, + { + "epoch": 0.45479, + "grad_norm": 1.9365589853266323, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45479 + }, + { + "epoch": 0.4548, + "grad_norm": 1.2770960542785594, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 45480 + }, + { + "epoch": 0.45481, + "grad_norm": 1.7905877048050722, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 45481 + }, + { + "epoch": 0.45482, + "grad_norm": 1.3629578948947652, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 45482 + }, + { + "epoch": 0.45483, + "grad_norm": 1.589787897276419, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 45483 + }, + { + "epoch": 0.45484, + "grad_norm": 1.673540385286027, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 45484 + }, + { + "epoch": 0.45485, + "grad_norm": 1.240277720089457, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 45485 + }, + { + "epoch": 0.45486, + "grad_norm": 1.5963243656400639, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 45486 + }, + { + "epoch": 0.45487, + "grad_norm": 1.5723460101475868, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 45487 + }, + { + "epoch": 0.45488, + "grad_norm": 1.202468810750441, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 45488 + }, + { + "epoch": 0.45489, + "grad_norm": 1.8678327998773452, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45489 + }, + { + "epoch": 0.4549, + "grad_norm": 1.3899326493909094, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45490 + }, + { + "epoch": 0.45491, + "grad_norm": 1.4205500469400465, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 45491 + }, + { + "epoch": 0.45492, + "grad_norm": 1.6769874862755585, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 45492 + }, + { + "epoch": 0.45493, + "grad_norm": 1.6144655267376258, + "learning_rate": 0.003, + "loss": 4.053, + "step": 45493 + }, + { + "epoch": 0.45494, + "grad_norm": 1.5447507527106663, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45494 + }, + { + "epoch": 0.45495, + "grad_norm": 1.4764763341428906, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 45495 + }, + { + "epoch": 0.45496, + "grad_norm": 1.5996514266354518, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 45496 + }, + { + "epoch": 0.45497, + "grad_norm": 1.4352397622613338, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 45497 + }, + { + "epoch": 0.45498, + "grad_norm": 2.096689512639585, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45498 + }, + { + "epoch": 0.45499, + "grad_norm": 1.4757718751117366, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 45499 + }, + { + "epoch": 0.455, + "grad_norm": 1.5919644026227668, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 45500 + }, + { + "epoch": 0.45501, + "grad_norm": 1.3388042589740279, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 45501 + }, + { + "epoch": 0.45502, + "grad_norm": 1.952336081616809, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45502 + }, + { + "epoch": 0.45503, + "grad_norm": 1.3786045662678543, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 45503 + }, + { + "epoch": 0.45504, + "grad_norm": 1.780524474234939, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 45504 + }, + { + "epoch": 0.45505, + "grad_norm": 1.5806784781762224, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 45505 + }, + { + "epoch": 0.45506, + "grad_norm": 1.512617918997123, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45506 + }, + { + "epoch": 0.45507, + "grad_norm": 1.6878023366631079, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 45507 + }, + { + "epoch": 0.45508, + "grad_norm": 1.3861752539974106, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 45508 + }, + { + "epoch": 0.45509, + "grad_norm": 1.7099384788371121, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45509 + }, + { + "epoch": 0.4551, + "grad_norm": 1.6631415846966704, + "learning_rate": 0.003, + "loss": 4.045, + "step": 45510 + }, + { + "epoch": 0.45511, + "grad_norm": 1.183718375300197, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 45511 + }, + { + "epoch": 0.45512, + "grad_norm": 2.0193931191470744, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 45512 + }, + { + "epoch": 0.45513, + "grad_norm": 1.136172286770183, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 45513 + }, + { + "epoch": 0.45514, + "grad_norm": 1.6392137800859579, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45514 + }, + { + "epoch": 0.45515, + "grad_norm": 1.4789571566783586, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 45515 + }, + { + "epoch": 0.45516, + "grad_norm": 1.6622834762665637, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 45516 + }, + { + "epoch": 0.45517, + "grad_norm": 1.6656511197199215, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 45517 + }, + { + "epoch": 0.45518, + "grad_norm": 1.5137491172816002, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 45518 + }, + { + "epoch": 0.45519, + "grad_norm": 1.3913024773463154, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45519 + }, + { + "epoch": 0.4552, + "grad_norm": 1.446585224435918, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 45520 + }, + { + "epoch": 0.45521, + "grad_norm": 1.5236327545485289, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 45521 + }, + { + "epoch": 0.45522, + "grad_norm": 1.4030059461333761, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 45522 + }, + { + "epoch": 0.45523, + "grad_norm": 1.7386358448946766, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 45523 + }, + { + "epoch": 0.45524, + "grad_norm": 1.314605110081023, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 45524 + }, + { + "epoch": 0.45525, + "grad_norm": 2.012476696973375, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 45525 + }, + { + "epoch": 0.45526, + "grad_norm": 1.6136279447831152, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 45526 + }, + { + "epoch": 0.45527, + "grad_norm": 1.783958751410358, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 45527 + }, + { + "epoch": 0.45528, + "grad_norm": 1.4472115375298402, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 45528 + }, + { + "epoch": 0.45529, + "grad_norm": 1.6949014223532206, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 45529 + }, + { + "epoch": 0.4553, + "grad_norm": 1.1740863380698985, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 45530 + }, + { + "epoch": 0.45531, + "grad_norm": 2.028428507337515, + "learning_rate": 0.003, + "loss": 4.036, + "step": 45531 + }, + { + "epoch": 0.45532, + "grad_norm": 1.2331099933099336, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 45532 + }, + { + "epoch": 0.45533, + "grad_norm": 1.7436218229907388, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45533 + }, + { + "epoch": 0.45534, + "grad_norm": 1.4847419693349415, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 45534 + }, + { + "epoch": 0.45535, + "grad_norm": 1.7043961257347666, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 45535 + }, + { + "epoch": 0.45536, + "grad_norm": 1.4409067212015445, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 45536 + }, + { + "epoch": 0.45537, + "grad_norm": 1.6496097651941, + "learning_rate": 0.003, + "loss": 4.039, + "step": 45537 + }, + { + "epoch": 0.45538, + "grad_norm": 1.4694957561215078, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 45538 + }, + { + "epoch": 0.45539, + "grad_norm": 1.6336955980299523, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 45539 + }, + { + "epoch": 0.4554, + "grad_norm": 1.3942213335658764, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 45540 + }, + { + "epoch": 0.45541, + "grad_norm": 1.633585868147465, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45541 + }, + { + "epoch": 0.45542, + "grad_norm": 1.415896771659197, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45542 + }, + { + "epoch": 0.45543, + "grad_norm": 2.0887597301524865, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45543 + }, + { + "epoch": 0.45544, + "grad_norm": 1.6422149292274855, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 45544 + }, + { + "epoch": 0.45545, + "grad_norm": 1.4123172985396277, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 45545 + }, + { + "epoch": 0.45546, + "grad_norm": 1.4649666495201257, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45546 + }, + { + "epoch": 0.45547, + "grad_norm": 1.5233534423802066, + "learning_rate": 0.003, + "loss": 4.014, + "step": 45547 + }, + { + "epoch": 0.45548, + "grad_norm": 1.433170072714577, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 45548 + }, + { + "epoch": 0.45549, + "grad_norm": 1.4663933536070535, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 45549 + }, + { + "epoch": 0.4555, + "grad_norm": 1.5898214989058772, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 45550 + }, + { + "epoch": 0.45551, + "grad_norm": 1.5089862258926967, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 45551 + }, + { + "epoch": 0.45552, + "grad_norm": 1.5646655110547247, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 45552 + }, + { + "epoch": 0.45553, + "grad_norm": 1.5774561323456682, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 45553 + }, + { + "epoch": 0.45554, + "grad_norm": 1.6722759668828455, + "learning_rate": 0.003, + "loss": 4.027, + "step": 45554 + }, + { + "epoch": 0.45555, + "grad_norm": 1.4563609870143528, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 45555 + }, + { + "epoch": 0.45556, + "grad_norm": 1.4955825351528576, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45556 + }, + { + "epoch": 0.45557, + "grad_norm": 1.5793973435619462, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 45557 + }, + { + "epoch": 0.45558, + "grad_norm": 1.93915478632198, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 45558 + }, + { + "epoch": 0.45559, + "grad_norm": 1.115520433840827, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 45559 + }, + { + "epoch": 0.4556, + "grad_norm": 1.814844078405523, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 45560 + }, + { + "epoch": 0.45561, + "grad_norm": 1.730900398629409, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 45561 + }, + { + "epoch": 0.45562, + "grad_norm": 1.3294096018618533, + "learning_rate": 0.003, + "loss": 4.043, + "step": 45562 + }, + { + "epoch": 0.45563, + "grad_norm": 1.7389737710752817, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45563 + }, + { + "epoch": 0.45564, + "grad_norm": 1.6977537740960373, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 45564 + }, + { + "epoch": 0.45565, + "grad_norm": 1.4614910133794663, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45565 + }, + { + "epoch": 0.45566, + "grad_norm": 1.425989892919937, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 45566 + }, + { + "epoch": 0.45567, + "grad_norm": 1.4179899291511975, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 45567 + }, + { + "epoch": 0.45568, + "grad_norm": 1.663905475784115, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 45568 + }, + { + "epoch": 0.45569, + "grad_norm": 1.6140377779925077, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45569 + }, + { + "epoch": 0.4557, + "grad_norm": 1.4749848615394427, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 45570 + }, + { + "epoch": 0.45571, + "grad_norm": 1.5294218879750128, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45571 + }, + { + "epoch": 0.45572, + "grad_norm": 1.6653455989406258, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 45572 + }, + { + "epoch": 0.45573, + "grad_norm": 1.472100067350233, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 45573 + }, + { + "epoch": 0.45574, + "grad_norm": 1.6072710642286288, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 45574 + }, + { + "epoch": 0.45575, + "grad_norm": 1.591534811754109, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 45575 + }, + { + "epoch": 0.45576, + "grad_norm": 1.7083311060386523, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 45576 + }, + { + "epoch": 0.45577, + "grad_norm": 1.6669248618874386, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 45577 + }, + { + "epoch": 0.45578, + "grad_norm": 1.484003707916371, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 45578 + }, + { + "epoch": 0.45579, + "grad_norm": 1.4885598321951525, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 45579 + }, + { + "epoch": 0.4558, + "grad_norm": 1.5807426182877669, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 45580 + }, + { + "epoch": 0.45581, + "grad_norm": 1.4944748308209772, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 45581 + }, + { + "epoch": 0.45582, + "grad_norm": 1.4239600406169513, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 45582 + }, + { + "epoch": 0.45583, + "grad_norm": 1.6092536499609573, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 45583 + }, + { + "epoch": 0.45584, + "grad_norm": 1.7659153037554591, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 45584 + }, + { + "epoch": 0.45585, + "grad_norm": 1.4591581039952388, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45585 + }, + { + "epoch": 0.45586, + "grad_norm": 1.5677925438424065, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45586 + }, + { + "epoch": 0.45587, + "grad_norm": 1.6979894745703799, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 45587 + }, + { + "epoch": 0.45588, + "grad_norm": 1.3602328122375518, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 45588 + }, + { + "epoch": 0.45589, + "grad_norm": 1.5659372353053713, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45589 + }, + { + "epoch": 0.4559, + "grad_norm": 1.363291268060385, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45590 + }, + { + "epoch": 0.45591, + "grad_norm": 1.5568236379879705, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 45591 + }, + { + "epoch": 0.45592, + "grad_norm": 1.6011003777407615, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 45592 + }, + { + "epoch": 0.45593, + "grad_norm": 1.891410325372179, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 45593 + }, + { + "epoch": 0.45594, + "grad_norm": 1.3064034797614734, + "learning_rate": 0.003, + "loss": 4.016, + "step": 45594 + }, + { + "epoch": 0.45595, + "grad_norm": 1.9171244567708003, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 45595 + }, + { + "epoch": 0.45596, + "grad_norm": 1.3766196498337, + "learning_rate": 0.003, + "loss": 4.017, + "step": 45596 + }, + { + "epoch": 0.45597, + "grad_norm": 1.6270848538137286, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 45597 + }, + { + "epoch": 0.45598, + "grad_norm": 1.545574853125002, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 45598 + }, + { + "epoch": 0.45599, + "grad_norm": 1.676570686031065, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 45599 + }, + { + "epoch": 0.456, + "grad_norm": 1.7349495626662654, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 45600 + }, + { + "epoch": 0.45601, + "grad_norm": 1.4530229461349389, + "learning_rate": 0.003, + "loss": 4.01, + "step": 45601 + }, + { + "epoch": 0.45602, + "grad_norm": 1.649189642465852, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45602 + }, + { + "epoch": 0.45603, + "grad_norm": 1.495331535434307, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 45603 + }, + { + "epoch": 0.45604, + "grad_norm": 1.332922742250593, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45604 + }, + { + "epoch": 0.45605, + "grad_norm": 1.4872684298963208, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 45605 + }, + { + "epoch": 0.45606, + "grad_norm": 1.2422507002541598, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45606 + }, + { + "epoch": 0.45607, + "grad_norm": 1.6626644473257899, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45607 + }, + { + "epoch": 0.45608, + "grad_norm": 1.5243221741389725, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 45608 + }, + { + "epoch": 0.45609, + "grad_norm": 1.4925617176255441, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 45609 + }, + { + "epoch": 0.4561, + "grad_norm": 1.414400274035746, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 45610 + }, + { + "epoch": 0.45611, + "grad_norm": 1.6088098660251955, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 45611 + }, + { + "epoch": 0.45612, + "grad_norm": 1.214892637876799, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45612 + }, + { + "epoch": 0.45613, + "grad_norm": 1.426949799407304, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45613 + }, + { + "epoch": 0.45614, + "grad_norm": 1.3223426282067756, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 45614 + }, + { + "epoch": 0.45615, + "grad_norm": 1.7223500440181254, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 45615 + }, + { + "epoch": 0.45616, + "grad_norm": 1.7427258678833306, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 45616 + }, + { + "epoch": 0.45617, + "grad_norm": 2.1082792271497164, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 45617 + }, + { + "epoch": 0.45618, + "grad_norm": 1.4707750710720355, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 45618 + }, + { + "epoch": 0.45619, + "grad_norm": 1.8336158370106488, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 45619 + }, + { + "epoch": 0.4562, + "grad_norm": 1.4219918352813241, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45620 + }, + { + "epoch": 0.45621, + "grad_norm": 1.5548783556616532, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45621 + }, + { + "epoch": 0.45622, + "grad_norm": 1.6239530407425606, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 45622 + }, + { + "epoch": 0.45623, + "grad_norm": 1.5882316122086617, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 45623 + }, + { + "epoch": 0.45624, + "grad_norm": 1.8084813312015426, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 45624 + }, + { + "epoch": 0.45625, + "grad_norm": 1.5293152257421472, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 45625 + }, + { + "epoch": 0.45626, + "grad_norm": 1.5111681520696114, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 45626 + }, + { + "epoch": 0.45627, + "grad_norm": 1.4244554933239957, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 45627 + }, + { + "epoch": 0.45628, + "grad_norm": 1.4507798466411506, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 45628 + }, + { + "epoch": 0.45629, + "grad_norm": 1.7271051148305954, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 45629 + }, + { + "epoch": 0.4563, + "grad_norm": 1.5870393512852186, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 45630 + }, + { + "epoch": 0.45631, + "grad_norm": 1.7018972673475863, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 45631 + }, + { + "epoch": 0.45632, + "grad_norm": 1.4658739824045335, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 45632 + }, + { + "epoch": 0.45633, + "grad_norm": 1.539103047914378, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45633 + }, + { + "epoch": 0.45634, + "grad_norm": 1.3830454132758982, + "learning_rate": 0.003, + "loss": 4.028, + "step": 45634 + }, + { + "epoch": 0.45635, + "grad_norm": 1.7017906670184717, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 45635 + }, + { + "epoch": 0.45636, + "grad_norm": 1.251341287025374, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 45636 + }, + { + "epoch": 0.45637, + "grad_norm": 1.6792977035084478, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 45637 + }, + { + "epoch": 0.45638, + "grad_norm": 1.441542742905614, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 45638 + }, + { + "epoch": 0.45639, + "grad_norm": 1.5522386270696926, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 45639 + }, + { + "epoch": 0.4564, + "grad_norm": 1.6030758732523125, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 45640 + }, + { + "epoch": 0.45641, + "grad_norm": 1.434790462235457, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45641 + }, + { + "epoch": 0.45642, + "grad_norm": 1.7547016785236496, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 45642 + }, + { + "epoch": 0.45643, + "grad_norm": 1.2712964869150352, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 45643 + }, + { + "epoch": 0.45644, + "grad_norm": 1.7017548328493688, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 45644 + }, + { + "epoch": 0.45645, + "grad_norm": 1.5349420371576887, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 45645 + }, + { + "epoch": 0.45646, + "grad_norm": 1.5469909151169716, + "learning_rate": 0.003, + "loss": 3.993, + "step": 45646 + }, + { + "epoch": 0.45647, + "grad_norm": 1.479790357095668, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 45647 + }, + { + "epoch": 0.45648, + "grad_norm": 1.908364404189591, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45648 + }, + { + "epoch": 0.45649, + "grad_norm": 1.1380044382607017, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 45649 + }, + { + "epoch": 0.4565, + "grad_norm": 1.7877799534396641, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45650 + }, + { + "epoch": 0.45651, + "grad_norm": 1.5414367799402446, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 45651 + }, + { + "epoch": 0.45652, + "grad_norm": 1.5201820405510027, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 45652 + }, + { + "epoch": 0.45653, + "grad_norm": 1.3940456115264768, + "learning_rate": 0.003, + "loss": 4.021, + "step": 45653 + }, + { + "epoch": 0.45654, + "grad_norm": 1.5442505013887444, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 45654 + }, + { + "epoch": 0.45655, + "grad_norm": 1.71091188642207, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 45655 + }, + { + "epoch": 0.45656, + "grad_norm": 1.6665891368859547, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 45656 + }, + { + "epoch": 0.45657, + "grad_norm": 1.4443136631831248, + "learning_rate": 0.003, + "loss": 4.02, + "step": 45657 + }, + { + "epoch": 0.45658, + "grad_norm": 2.0329642219639945, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 45658 + }, + { + "epoch": 0.45659, + "grad_norm": 1.2081084526124837, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45659 + }, + { + "epoch": 0.4566, + "grad_norm": 1.9581690324714032, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45660 + }, + { + "epoch": 0.45661, + "grad_norm": 1.7596804115261055, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 45661 + }, + { + "epoch": 0.45662, + "grad_norm": 1.8269908895635274, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 45662 + }, + { + "epoch": 0.45663, + "grad_norm": 1.5314467675895955, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 45663 + }, + { + "epoch": 0.45664, + "grad_norm": 1.6615932146880228, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 45664 + }, + { + "epoch": 0.45665, + "grad_norm": 1.7029750834792772, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 45665 + }, + { + "epoch": 0.45666, + "grad_norm": 1.4009364669798616, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 45666 + }, + { + "epoch": 0.45667, + "grad_norm": 1.593778550457657, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 45667 + }, + { + "epoch": 0.45668, + "grad_norm": 1.4904840250598796, + "learning_rate": 0.003, + "loss": 4.048, + "step": 45668 + }, + { + "epoch": 0.45669, + "grad_norm": 1.6169677022947204, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45669 + }, + { + "epoch": 0.4567, + "grad_norm": 1.368924106199119, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 45670 + }, + { + "epoch": 0.45671, + "grad_norm": 1.7918607445805297, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45671 + }, + { + "epoch": 0.45672, + "grad_norm": 1.1207145033337351, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 45672 + }, + { + "epoch": 0.45673, + "grad_norm": 1.6352302831928096, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 45673 + }, + { + "epoch": 0.45674, + "grad_norm": 1.331940548852418, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 45674 + }, + { + "epoch": 0.45675, + "grad_norm": 1.593753953387351, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45675 + }, + { + "epoch": 0.45676, + "grad_norm": 1.4664784741670638, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 45676 + }, + { + "epoch": 0.45677, + "grad_norm": 1.712135871622735, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45677 + }, + { + "epoch": 0.45678, + "grad_norm": 1.7636295325285616, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 45678 + }, + { + "epoch": 0.45679, + "grad_norm": 1.5305176996609555, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 45679 + }, + { + "epoch": 0.4568, + "grad_norm": 1.3512168509595837, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 45680 + }, + { + "epoch": 0.45681, + "grad_norm": 1.4188877326769187, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45681 + }, + { + "epoch": 0.45682, + "grad_norm": 1.6270833988298945, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 45682 + }, + { + "epoch": 0.45683, + "grad_norm": 1.5334582262732763, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 45683 + }, + { + "epoch": 0.45684, + "grad_norm": 1.6527821343430698, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 45684 + }, + { + "epoch": 0.45685, + "grad_norm": 1.3269785351520413, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45685 + }, + { + "epoch": 0.45686, + "grad_norm": 1.508424530965725, + "learning_rate": 0.003, + "loss": 4.023, + "step": 45686 + }, + { + "epoch": 0.45687, + "grad_norm": 1.3499276080521232, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45687 + }, + { + "epoch": 0.45688, + "grad_norm": 1.5082378494888307, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 45688 + }, + { + "epoch": 0.45689, + "grad_norm": 1.5965548519852892, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45689 + }, + { + "epoch": 0.4569, + "grad_norm": 1.6545945385903313, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 45690 + }, + { + "epoch": 0.45691, + "grad_norm": 1.5836317934279902, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45691 + }, + { + "epoch": 0.45692, + "grad_norm": 1.8019394009298262, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 45692 + }, + { + "epoch": 0.45693, + "grad_norm": 1.5283796358711548, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 45693 + }, + { + "epoch": 0.45694, + "grad_norm": 1.4042635334980766, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 45694 + }, + { + "epoch": 0.45695, + "grad_norm": 1.6203040348386935, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 45695 + }, + { + "epoch": 0.45696, + "grad_norm": 1.4980924368875124, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 45696 + }, + { + "epoch": 0.45697, + "grad_norm": 1.7494801242888767, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 45697 + }, + { + "epoch": 0.45698, + "grad_norm": 1.4800273254078276, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45698 + }, + { + "epoch": 0.45699, + "grad_norm": 1.83276515714791, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 45699 + }, + { + "epoch": 0.457, + "grad_norm": 1.4357328991992604, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 45700 + }, + { + "epoch": 0.45701, + "grad_norm": 1.83885292873574, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 45701 + }, + { + "epoch": 0.45702, + "grad_norm": 1.3963271084857136, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45702 + }, + { + "epoch": 0.45703, + "grad_norm": 1.7128216516783958, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 45703 + }, + { + "epoch": 0.45704, + "grad_norm": 1.7097556398472162, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45704 + }, + { + "epoch": 0.45705, + "grad_norm": 1.2601417786276858, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45705 + }, + { + "epoch": 0.45706, + "grad_norm": 1.8151967003385312, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45706 + }, + { + "epoch": 0.45707, + "grad_norm": 1.4664854450733216, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 45707 + }, + { + "epoch": 0.45708, + "grad_norm": 1.3700751532651894, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 45708 + }, + { + "epoch": 0.45709, + "grad_norm": 1.5276885453349325, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 45709 + }, + { + "epoch": 0.4571, + "grad_norm": 1.626942050408857, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45710 + }, + { + "epoch": 0.45711, + "grad_norm": 1.5842692396859297, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 45711 + }, + { + "epoch": 0.45712, + "grad_norm": 1.6175768438137919, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 45712 + }, + { + "epoch": 0.45713, + "grad_norm": 1.673594678012788, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 45713 + }, + { + "epoch": 0.45714, + "grad_norm": 1.4561639981014776, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 45714 + }, + { + "epoch": 0.45715, + "grad_norm": 1.510371864617892, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 45715 + }, + { + "epoch": 0.45716, + "grad_norm": 1.6031906829725495, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 45716 + }, + { + "epoch": 0.45717, + "grad_norm": 1.5765857242175263, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 45717 + }, + { + "epoch": 0.45718, + "grad_norm": 1.6176904309367548, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 45718 + }, + { + "epoch": 0.45719, + "grad_norm": 1.623375176467106, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 45719 + }, + { + "epoch": 0.4572, + "grad_norm": 1.6169243454777253, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 45720 + }, + { + "epoch": 0.45721, + "grad_norm": 1.7042144028622952, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 45721 + }, + { + "epoch": 0.45722, + "grad_norm": 1.2842077415222393, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 45722 + }, + { + "epoch": 0.45723, + "grad_norm": 1.7809948874927637, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 45723 + }, + { + "epoch": 0.45724, + "grad_norm": 1.4830747464585812, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45724 + }, + { + "epoch": 0.45725, + "grad_norm": 1.4639551998129243, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 45725 + }, + { + "epoch": 0.45726, + "grad_norm": 1.614117258563871, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 45726 + }, + { + "epoch": 0.45727, + "grad_norm": 1.601192501748927, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 45727 + }, + { + "epoch": 0.45728, + "grad_norm": 1.4591240059235835, + "learning_rate": 0.003, + "loss": 4.074, + "step": 45728 + }, + { + "epoch": 0.45729, + "grad_norm": 1.4928309982259027, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 45729 + }, + { + "epoch": 0.4573, + "grad_norm": 1.7595902295838248, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 45730 + }, + { + "epoch": 0.45731, + "grad_norm": 1.4669577281881345, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 45731 + }, + { + "epoch": 0.45732, + "grad_norm": 1.5038483712585562, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 45732 + }, + { + "epoch": 0.45733, + "grad_norm": 1.425497292193393, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 45733 + }, + { + "epoch": 0.45734, + "grad_norm": 1.537566189668161, + "learning_rate": 0.003, + "loss": 4.029, + "step": 45734 + }, + { + "epoch": 0.45735, + "grad_norm": 1.4528830439291545, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 45735 + }, + { + "epoch": 0.45736, + "grad_norm": 1.4905460686499645, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45736 + }, + { + "epoch": 0.45737, + "grad_norm": 1.4128188169359113, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 45737 + }, + { + "epoch": 0.45738, + "grad_norm": 1.554625513476405, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 45738 + }, + { + "epoch": 0.45739, + "grad_norm": 1.507998611088151, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 45739 + }, + { + "epoch": 0.4574, + "grad_norm": 1.7711030932904945, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 45740 + }, + { + "epoch": 0.45741, + "grad_norm": 1.4906956944815264, + "learning_rate": 0.003, + "loss": 4.032, + "step": 45741 + }, + { + "epoch": 0.45742, + "grad_norm": 1.6105907360692278, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 45742 + }, + { + "epoch": 0.45743, + "grad_norm": 1.6385016106366548, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 45743 + }, + { + "epoch": 0.45744, + "grad_norm": 1.2437196182186854, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 45744 + }, + { + "epoch": 0.45745, + "grad_norm": 1.8814357086734819, + "learning_rate": 0.003, + "loss": 4.025, + "step": 45745 + }, + { + "epoch": 0.45746, + "grad_norm": 1.402692896872694, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 45746 + }, + { + "epoch": 0.45747, + "grad_norm": 1.6741102070180338, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45747 + }, + { + "epoch": 0.45748, + "grad_norm": 1.5188259219228781, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 45748 + }, + { + "epoch": 0.45749, + "grad_norm": 1.718901067546898, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 45749 + }, + { + "epoch": 0.4575, + "grad_norm": 1.6628537972589146, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 45750 + }, + { + "epoch": 0.45751, + "grad_norm": 1.3247729897209581, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 45751 + }, + { + "epoch": 0.45752, + "grad_norm": 1.6296339834716727, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 45752 + }, + { + "epoch": 0.45753, + "grad_norm": 1.1214586136556781, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 45753 + }, + { + "epoch": 0.45754, + "grad_norm": 1.8767413665824064, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 45754 + }, + { + "epoch": 0.45755, + "grad_norm": 1.2876658509057413, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 45755 + }, + { + "epoch": 0.45756, + "grad_norm": 1.9649327122158329, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 45756 + }, + { + "epoch": 0.45757, + "grad_norm": 1.5988058878585263, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 45757 + }, + { + "epoch": 0.45758, + "grad_norm": 1.7249147384827277, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 45758 + }, + { + "epoch": 0.45759, + "grad_norm": 1.566138516305827, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 45759 + }, + { + "epoch": 0.4576, + "grad_norm": 1.7076705137862664, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 45760 + }, + { + "epoch": 0.45761, + "grad_norm": 1.6006529644650187, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 45761 + }, + { + "epoch": 0.45762, + "grad_norm": 1.7589091591291681, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45762 + }, + { + "epoch": 0.45763, + "grad_norm": 1.814791145403577, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 45763 + }, + { + "epoch": 0.45764, + "grad_norm": 1.3351903696234408, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 45764 + }, + { + "epoch": 0.45765, + "grad_norm": 1.6196624830535467, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 45765 + }, + { + "epoch": 0.45766, + "grad_norm": 1.5427627237858903, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 45766 + }, + { + "epoch": 0.45767, + "grad_norm": 1.3818536724938768, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 45767 + }, + { + "epoch": 0.45768, + "grad_norm": 1.4852417657120334, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 45768 + }, + { + "epoch": 0.45769, + "grad_norm": 1.7573392074484835, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 45769 + }, + { + "epoch": 0.4577, + "grad_norm": 1.3381670673290842, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 45770 + }, + { + "epoch": 0.45771, + "grad_norm": 1.5520079584466295, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 45771 + }, + { + "epoch": 0.45772, + "grad_norm": 1.4245674082970115, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 45772 + }, + { + "epoch": 0.45773, + "grad_norm": 1.3481898754922175, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 45773 + }, + { + "epoch": 0.45774, + "grad_norm": 1.6252392462676002, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 45774 + }, + { + "epoch": 0.45775, + "grad_norm": 1.715165930889185, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 45775 + }, + { + "epoch": 0.45776, + "grad_norm": 1.4027561421976826, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 45776 + }, + { + "epoch": 0.45777, + "grad_norm": 1.539133431981006, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 45777 + }, + { + "epoch": 0.45778, + "grad_norm": 1.4579375279956663, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 45778 + }, + { + "epoch": 0.45779, + "grad_norm": 1.7635449087526118, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 45779 + }, + { + "epoch": 0.4578, + "grad_norm": 1.8044323932694968, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 45780 + }, + { + "epoch": 0.45781, + "grad_norm": 1.33620813633146, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45781 + }, + { + "epoch": 0.45782, + "grad_norm": 1.7484948764898676, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 45782 + }, + { + "epoch": 0.45783, + "grad_norm": 1.4188046826455876, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 45783 + }, + { + "epoch": 0.45784, + "grad_norm": 1.7304300979757816, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 45784 + }, + { + "epoch": 0.45785, + "grad_norm": 1.3003120007544389, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 45785 + }, + { + "epoch": 0.45786, + "grad_norm": 1.6430892019739822, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 45786 + }, + { + "epoch": 0.45787, + "grad_norm": 1.5322591013389828, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 45787 + }, + { + "epoch": 0.45788, + "grad_norm": 1.748919336412856, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 45788 + }, + { + "epoch": 0.45789, + "grad_norm": 1.6873396841228305, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 45789 + }, + { + "epoch": 0.4579, + "grad_norm": 1.3202483857981377, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 45790 + }, + { + "epoch": 0.45791, + "grad_norm": 1.8252157978613905, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45791 + }, + { + "epoch": 0.45792, + "grad_norm": 1.2664593572991822, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45792 + }, + { + "epoch": 0.45793, + "grad_norm": 1.5794355751060003, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 45793 + }, + { + "epoch": 0.45794, + "grad_norm": 1.464723237941011, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 45794 + }, + { + "epoch": 0.45795, + "grad_norm": 1.50857940109633, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 45795 + }, + { + "epoch": 0.45796, + "grad_norm": 1.6510046960657345, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 45796 + }, + { + "epoch": 0.45797, + "grad_norm": 1.3343537520555129, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 45797 + }, + { + "epoch": 0.45798, + "grad_norm": 1.9135530926187023, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 45798 + }, + { + "epoch": 0.45799, + "grad_norm": 1.5635387299171302, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45799 + }, + { + "epoch": 0.458, + "grad_norm": 1.3539938468717732, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 45800 + }, + { + "epoch": 0.45801, + "grad_norm": 1.5621839981550876, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45801 + }, + { + "epoch": 0.45802, + "grad_norm": 1.789979689393018, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 45802 + }, + { + "epoch": 0.45803, + "grad_norm": 1.25547837375303, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45803 + }, + { + "epoch": 0.45804, + "grad_norm": 2.062792617571255, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 45804 + }, + { + "epoch": 0.45805, + "grad_norm": 1.3363367061085965, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45805 + }, + { + "epoch": 0.45806, + "grad_norm": 1.8057665688370228, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45806 + }, + { + "epoch": 0.45807, + "grad_norm": 1.2983195494206672, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45807 + }, + { + "epoch": 0.45808, + "grad_norm": 1.9750091618355268, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 45808 + }, + { + "epoch": 0.45809, + "grad_norm": 1.4214284434562925, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 45809 + }, + { + "epoch": 0.4581, + "grad_norm": 1.646990419310457, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45810 + }, + { + "epoch": 0.45811, + "grad_norm": 1.4887206379832836, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 45811 + }, + { + "epoch": 0.45812, + "grad_norm": 1.494041116269255, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 45812 + }, + { + "epoch": 0.45813, + "grad_norm": 1.7024132172876836, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 45813 + }, + { + "epoch": 0.45814, + "grad_norm": 1.5021152063726457, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 45814 + }, + { + "epoch": 0.45815, + "grad_norm": 1.6374147656104632, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 45815 + }, + { + "epoch": 0.45816, + "grad_norm": 1.9354692886944938, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 45816 + }, + { + "epoch": 0.45817, + "grad_norm": 1.257623746725321, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 45817 + }, + { + "epoch": 0.45818, + "grad_norm": 1.441706564798279, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 45818 + }, + { + "epoch": 0.45819, + "grad_norm": 1.8308065669148688, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45819 + }, + { + "epoch": 0.4582, + "grad_norm": 1.3414504593436245, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45820 + }, + { + "epoch": 0.45821, + "grad_norm": 1.5667097942940054, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45821 + }, + { + "epoch": 0.45822, + "grad_norm": 1.6480458319682807, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 45822 + }, + { + "epoch": 0.45823, + "grad_norm": 1.5230687387937398, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45823 + }, + { + "epoch": 0.45824, + "grad_norm": 1.4029237269874733, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 45824 + }, + { + "epoch": 0.45825, + "grad_norm": 1.650353859036279, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45825 + }, + { + "epoch": 0.45826, + "grad_norm": 1.5685450974694384, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 45826 + }, + { + "epoch": 0.45827, + "grad_norm": 1.4342251473448193, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 45827 + }, + { + "epoch": 0.45828, + "grad_norm": 1.711959071809548, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 45828 + }, + { + "epoch": 0.45829, + "grad_norm": 1.4165677834130612, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 45829 + }, + { + "epoch": 0.4583, + "grad_norm": 1.5603793956804026, + "learning_rate": 0.003, + "loss": 4.003, + "step": 45830 + }, + { + "epoch": 0.45831, + "grad_norm": 1.3358224963340957, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 45831 + }, + { + "epoch": 0.45832, + "grad_norm": 1.7816010811806369, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 45832 + }, + { + "epoch": 0.45833, + "grad_norm": 1.3783992438155688, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 45833 + }, + { + "epoch": 0.45834, + "grad_norm": 1.7955035813713711, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 45834 + }, + { + "epoch": 0.45835, + "grad_norm": 1.5379732849998466, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 45835 + }, + { + "epoch": 0.45836, + "grad_norm": 1.4573945821787537, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45836 + }, + { + "epoch": 0.45837, + "grad_norm": 1.3421468026237326, + "learning_rate": 0.003, + "loss": 4.018, + "step": 45837 + }, + { + "epoch": 0.45838, + "grad_norm": 1.7438457822083853, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 45838 + }, + { + "epoch": 0.45839, + "grad_norm": 1.3777985657884142, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 45839 + }, + { + "epoch": 0.4584, + "grad_norm": 1.588762375647492, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 45840 + }, + { + "epoch": 0.45841, + "grad_norm": 1.4394276748748578, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 45841 + }, + { + "epoch": 0.45842, + "grad_norm": 1.6659983279001476, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 45842 + }, + { + "epoch": 0.45843, + "grad_norm": 1.5812531138380557, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 45843 + }, + { + "epoch": 0.45844, + "grad_norm": 1.307907651595337, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 45844 + }, + { + "epoch": 0.45845, + "grad_norm": 1.80960088150284, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 45845 + }, + { + "epoch": 0.45846, + "grad_norm": 1.3641573008486396, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 45846 + }, + { + "epoch": 0.45847, + "grad_norm": 1.769663975307166, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 45847 + }, + { + "epoch": 0.45848, + "grad_norm": 1.4800774167509045, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 45848 + }, + { + "epoch": 0.45849, + "grad_norm": 1.727193707732097, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 45849 + }, + { + "epoch": 0.4585, + "grad_norm": 1.2923941013047848, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 45850 + }, + { + "epoch": 0.45851, + "grad_norm": 1.7919794789764858, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45851 + }, + { + "epoch": 0.45852, + "grad_norm": 1.3015819313173642, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 45852 + }, + { + "epoch": 0.45853, + "grad_norm": 1.7871177558939755, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 45853 + }, + { + "epoch": 0.45854, + "grad_norm": 1.567553858713762, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 45854 + }, + { + "epoch": 0.45855, + "grad_norm": 1.7964125433406803, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45855 + }, + { + "epoch": 0.45856, + "grad_norm": 1.5595903896291567, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 45856 + }, + { + "epoch": 0.45857, + "grad_norm": 1.2255425144296348, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 45857 + }, + { + "epoch": 0.45858, + "grad_norm": 1.7170443660651666, + "learning_rate": 0.003, + "loss": 4.01, + "step": 45858 + }, + { + "epoch": 0.45859, + "grad_norm": 1.5889083775007815, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 45859 + }, + { + "epoch": 0.4586, + "grad_norm": 1.4734983835579158, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 45860 + }, + { + "epoch": 0.45861, + "grad_norm": 1.9896039973638042, + "learning_rate": 0.003, + "loss": 4.035, + "step": 45861 + }, + { + "epoch": 0.45862, + "grad_norm": 1.294400842424426, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 45862 + }, + { + "epoch": 0.45863, + "grad_norm": 1.5967693947126365, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 45863 + }, + { + "epoch": 0.45864, + "grad_norm": 1.8311584176391376, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45864 + }, + { + "epoch": 0.45865, + "grad_norm": 1.7858060594576075, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 45865 + }, + { + "epoch": 0.45866, + "grad_norm": 1.2561485588894226, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 45866 + }, + { + "epoch": 0.45867, + "grad_norm": 1.7231648989960846, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45867 + }, + { + "epoch": 0.45868, + "grad_norm": 1.5282926002049546, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45868 + }, + { + "epoch": 0.45869, + "grad_norm": 1.63796946425151, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 45869 + }, + { + "epoch": 0.4587, + "grad_norm": 1.329663156447888, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 45870 + }, + { + "epoch": 0.45871, + "grad_norm": 1.6704034200917086, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 45871 + }, + { + "epoch": 0.45872, + "grad_norm": 1.2736428903521029, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 45872 + }, + { + "epoch": 0.45873, + "grad_norm": 1.5025201300058113, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45873 + }, + { + "epoch": 0.45874, + "grad_norm": 1.5312368782217767, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 45874 + }, + { + "epoch": 0.45875, + "grad_norm": 1.7567963044327928, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 45875 + }, + { + "epoch": 0.45876, + "grad_norm": 1.3685327796188385, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 45876 + }, + { + "epoch": 0.45877, + "grad_norm": 1.7474499039007256, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 45877 + }, + { + "epoch": 0.45878, + "grad_norm": 1.515660267612354, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 45878 + }, + { + "epoch": 0.45879, + "grad_norm": 1.6629763645069338, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 45879 + }, + { + "epoch": 0.4588, + "grad_norm": 1.5846857563987904, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 45880 + }, + { + "epoch": 0.45881, + "grad_norm": 1.429667413577395, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 45881 + }, + { + "epoch": 0.45882, + "grad_norm": 1.6572994683255398, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 45882 + }, + { + "epoch": 0.45883, + "grad_norm": 1.420349048880594, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 45883 + }, + { + "epoch": 0.45884, + "grad_norm": 1.6315941910261127, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 45884 + }, + { + "epoch": 0.45885, + "grad_norm": 1.3459183486807018, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 45885 + }, + { + "epoch": 0.45886, + "grad_norm": 1.6190545409610893, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 45886 + }, + { + "epoch": 0.45887, + "grad_norm": 1.7455934247204687, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 45887 + }, + { + "epoch": 0.45888, + "grad_norm": 1.6119902800568815, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 45888 + }, + { + "epoch": 0.45889, + "grad_norm": 1.694809752656845, + "learning_rate": 0.003, + "loss": 4.055, + "step": 45889 + }, + { + "epoch": 0.4589, + "grad_norm": 1.3718876139830007, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 45890 + }, + { + "epoch": 0.45891, + "grad_norm": 1.6497300767722387, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 45891 + }, + { + "epoch": 0.45892, + "grad_norm": 1.697164406080703, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 45892 + }, + { + "epoch": 0.45893, + "grad_norm": 1.5136985279377144, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 45893 + }, + { + "epoch": 0.45894, + "grad_norm": 1.5443350521042036, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 45894 + }, + { + "epoch": 0.45895, + "grad_norm": 1.4531289286078353, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 45895 + }, + { + "epoch": 0.45896, + "grad_norm": 1.4206732171008505, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45896 + }, + { + "epoch": 0.45897, + "grad_norm": 1.786715126933887, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45897 + }, + { + "epoch": 0.45898, + "grad_norm": 1.4684711592475026, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 45898 + }, + { + "epoch": 0.45899, + "grad_norm": 1.4767450827377262, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 45899 + }, + { + "epoch": 0.459, + "grad_norm": 1.6169381108288015, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 45900 + }, + { + "epoch": 0.45901, + "grad_norm": 1.3134764478401237, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 45901 + }, + { + "epoch": 0.45902, + "grad_norm": 1.704801605220661, + "learning_rate": 0.003, + "loss": 4.04, + "step": 45902 + }, + { + "epoch": 0.45903, + "grad_norm": 1.5060332553704143, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 45903 + }, + { + "epoch": 0.45904, + "grad_norm": 1.4875505410109284, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 45904 + }, + { + "epoch": 0.45905, + "grad_norm": 1.331333652003841, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 45905 + }, + { + "epoch": 0.45906, + "grad_norm": 1.804687640868009, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 45906 + }, + { + "epoch": 0.45907, + "grad_norm": 1.3617327091466351, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 45907 + }, + { + "epoch": 0.45908, + "grad_norm": 1.721278411283771, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 45908 + }, + { + "epoch": 0.45909, + "grad_norm": 1.808563001345621, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45909 + }, + { + "epoch": 0.4591, + "grad_norm": 1.6355841990397368, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 45910 + }, + { + "epoch": 0.45911, + "grad_norm": 1.5624413129049857, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 45911 + }, + { + "epoch": 0.45912, + "grad_norm": 1.2869983133815817, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 45912 + }, + { + "epoch": 0.45913, + "grad_norm": 1.6030256201790758, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 45913 + }, + { + "epoch": 0.45914, + "grad_norm": 1.2898338320930136, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 45914 + }, + { + "epoch": 0.45915, + "grad_norm": 1.739073537263199, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 45915 + }, + { + "epoch": 0.45916, + "grad_norm": 1.341033482649948, + "learning_rate": 0.003, + "loss": 4.035, + "step": 45916 + }, + { + "epoch": 0.45917, + "grad_norm": 1.7673532594913877, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 45917 + }, + { + "epoch": 0.45918, + "grad_norm": 1.438379418052368, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 45918 + }, + { + "epoch": 0.45919, + "grad_norm": 1.5570311861649966, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 45919 + }, + { + "epoch": 0.4592, + "grad_norm": 1.5077046487312786, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 45920 + }, + { + "epoch": 0.45921, + "grad_norm": 1.468403795629734, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 45921 + }, + { + "epoch": 0.45922, + "grad_norm": 1.6626829316521676, + "learning_rate": 0.003, + "loss": 4.033, + "step": 45922 + }, + { + "epoch": 0.45923, + "grad_norm": 1.5693631278617761, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 45923 + }, + { + "epoch": 0.45924, + "grad_norm": 1.609731047259362, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 45924 + }, + { + "epoch": 0.45925, + "grad_norm": 1.6097766783127105, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 45925 + }, + { + "epoch": 0.45926, + "grad_norm": 1.4272499257442546, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 45926 + }, + { + "epoch": 0.45927, + "grad_norm": 1.418071496452845, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 45927 + }, + { + "epoch": 0.45928, + "grad_norm": 1.3826698193354392, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 45928 + }, + { + "epoch": 0.45929, + "grad_norm": 1.85863994622607, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 45929 + }, + { + "epoch": 0.4593, + "grad_norm": 1.4695003923959986, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 45930 + }, + { + "epoch": 0.45931, + "grad_norm": 1.4444433079904115, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 45931 + }, + { + "epoch": 0.45932, + "grad_norm": 1.5016287517813351, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 45932 + }, + { + "epoch": 0.45933, + "grad_norm": 1.7034794114754277, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 45933 + }, + { + "epoch": 0.45934, + "grad_norm": 1.7067483628054192, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 45934 + }, + { + "epoch": 0.45935, + "grad_norm": 1.3353059523816988, + "learning_rate": 0.003, + "loss": 4.022, + "step": 45935 + }, + { + "epoch": 0.45936, + "grad_norm": 1.7226088723614357, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 45936 + }, + { + "epoch": 0.45937, + "grad_norm": 1.333106582602541, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45937 + }, + { + "epoch": 0.45938, + "grad_norm": 1.6011692950186374, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 45938 + }, + { + "epoch": 0.45939, + "grad_norm": 1.6207327123755761, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 45939 + }, + { + "epoch": 0.4594, + "grad_norm": 1.6761746574187915, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 45940 + }, + { + "epoch": 0.45941, + "grad_norm": 1.6787699799597566, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 45941 + }, + { + "epoch": 0.45942, + "grad_norm": 1.953852094649409, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 45942 + }, + { + "epoch": 0.45943, + "grad_norm": 1.1872706002695925, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 45943 + }, + { + "epoch": 0.45944, + "grad_norm": 1.8875953348926136, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 45944 + }, + { + "epoch": 0.45945, + "grad_norm": 1.4748261428373255, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 45945 + }, + { + "epoch": 0.45946, + "grad_norm": 1.5608476188370735, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 45946 + }, + { + "epoch": 0.45947, + "grad_norm": 1.634224563925811, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 45947 + }, + { + "epoch": 0.45948, + "grad_norm": 1.5892581406105712, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 45948 + }, + { + "epoch": 0.45949, + "grad_norm": 1.6389085549010693, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 45949 + }, + { + "epoch": 0.4595, + "grad_norm": 1.3009863652298783, + "learning_rate": 0.003, + "loss": 4.034, + "step": 45950 + }, + { + "epoch": 0.45951, + "grad_norm": 1.7825744267303245, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 45951 + }, + { + "epoch": 0.45952, + "grad_norm": 1.3826264881469537, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 45952 + }, + { + "epoch": 0.45953, + "grad_norm": 1.4581534722635832, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 45953 + }, + { + "epoch": 0.45954, + "grad_norm": 1.483422334435401, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 45954 + }, + { + "epoch": 0.45955, + "grad_norm": 1.7728972179901907, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 45955 + }, + { + "epoch": 0.45956, + "grad_norm": 1.5068471505130787, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 45956 + }, + { + "epoch": 0.45957, + "grad_norm": 1.9836641373748478, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 45957 + }, + { + "epoch": 0.45958, + "grad_norm": 1.50430758458846, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 45958 + }, + { + "epoch": 0.45959, + "grad_norm": 1.7115868664484346, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 45959 + }, + { + "epoch": 0.4596, + "grad_norm": 1.3179338207374882, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 45960 + }, + { + "epoch": 0.45961, + "grad_norm": 1.4968142511366889, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 45961 + }, + { + "epoch": 0.45962, + "grad_norm": 1.6342597695191936, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 45962 + }, + { + "epoch": 0.45963, + "grad_norm": 1.5714687545216617, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 45963 + }, + { + "epoch": 0.45964, + "grad_norm": 1.6591651294122796, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 45964 + }, + { + "epoch": 0.45965, + "grad_norm": 1.3292586481720468, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 45965 + }, + { + "epoch": 0.45966, + "grad_norm": 1.4875748958289197, + "learning_rate": 0.003, + "loss": 4.042, + "step": 45966 + }, + { + "epoch": 0.45967, + "grad_norm": 1.533299223291712, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 45967 + }, + { + "epoch": 0.45968, + "grad_norm": 1.568024742096088, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 45968 + }, + { + "epoch": 0.45969, + "grad_norm": 1.5646150647266959, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 45969 + }, + { + "epoch": 0.4597, + "grad_norm": 1.6523374255713927, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 45970 + }, + { + "epoch": 0.45971, + "grad_norm": 1.6603075914330896, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 45971 + }, + { + "epoch": 0.45972, + "grad_norm": 1.5036197789179067, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 45972 + }, + { + "epoch": 0.45973, + "grad_norm": 1.4857132566036326, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 45973 + }, + { + "epoch": 0.45974, + "grad_norm": 1.8177868238233008, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 45974 + }, + { + "epoch": 0.45975, + "grad_norm": 1.1430414264823812, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 45975 + }, + { + "epoch": 0.45976, + "grad_norm": 1.995473255966013, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 45976 + }, + { + "epoch": 0.45977, + "grad_norm": 1.2592353878075055, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 45977 + }, + { + "epoch": 0.45978, + "grad_norm": 1.7688925935201378, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 45978 + }, + { + "epoch": 0.45979, + "grad_norm": 1.2834363794223556, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 45979 + }, + { + "epoch": 0.4598, + "grad_norm": 1.5199988514961968, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45980 + }, + { + "epoch": 0.45981, + "grad_norm": 1.3730746506519178, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 45981 + }, + { + "epoch": 0.45982, + "grad_norm": 1.6847374871504315, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 45982 + }, + { + "epoch": 0.45983, + "grad_norm": 1.4315337455836807, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 45983 + }, + { + "epoch": 0.45984, + "grad_norm": 1.8490013498980702, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 45984 + }, + { + "epoch": 0.45985, + "grad_norm": 1.934493163527718, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45985 + }, + { + "epoch": 0.45986, + "grad_norm": 1.5352558222518493, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 45986 + }, + { + "epoch": 0.45987, + "grad_norm": 1.9098884212106328, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 45987 + }, + { + "epoch": 0.45988, + "grad_norm": 1.6604215281325982, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 45988 + }, + { + "epoch": 0.45989, + "grad_norm": 1.3256301595953581, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45989 + }, + { + "epoch": 0.4599, + "grad_norm": 1.491540397579353, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 45990 + }, + { + "epoch": 0.45991, + "grad_norm": 1.663158738024037, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 45991 + }, + { + "epoch": 0.45992, + "grad_norm": 1.2548414687531464, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 45992 + }, + { + "epoch": 0.45993, + "grad_norm": 1.5957710521027035, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 45993 + }, + { + "epoch": 0.45994, + "grad_norm": 1.7106859376260575, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 45994 + }, + { + "epoch": 0.45995, + "grad_norm": 1.5653587433172171, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 45995 + }, + { + "epoch": 0.45996, + "grad_norm": 1.7571840909441354, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 45996 + }, + { + "epoch": 0.45997, + "grad_norm": 1.3841631085442785, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 45997 + }, + { + "epoch": 0.45998, + "grad_norm": 2.076958244353218, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 45998 + }, + { + "epoch": 0.45999, + "grad_norm": 1.5682195274518929, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 45999 + }, + { + "epoch": 0.46, + "grad_norm": 1.64580374781735, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46000 + }, + { + "epoch": 0.46001, + "grad_norm": 1.797502256816952, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46001 + }, + { + "epoch": 0.46002, + "grad_norm": 1.1735536848243742, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46002 + }, + { + "epoch": 0.46003, + "grad_norm": 1.5966061641843567, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46003 + }, + { + "epoch": 0.46004, + "grad_norm": 1.2148204543106562, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 46004 + }, + { + "epoch": 0.46005, + "grad_norm": 1.6913328028834185, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 46005 + }, + { + "epoch": 0.46006, + "grad_norm": 1.3565841426346648, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46006 + }, + { + "epoch": 0.46007, + "grad_norm": 1.4453624362124817, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46007 + }, + { + "epoch": 0.46008, + "grad_norm": 1.773518514638327, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 46008 + }, + { + "epoch": 0.46009, + "grad_norm": 2.0677903594465152, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 46009 + }, + { + "epoch": 0.4601, + "grad_norm": 1.4175374803244802, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 46010 + }, + { + "epoch": 0.46011, + "grad_norm": 1.5575390743949067, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 46011 + }, + { + "epoch": 0.46012, + "grad_norm": 1.5960410311851099, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 46012 + }, + { + "epoch": 0.46013, + "grad_norm": 1.4505624150772463, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46013 + }, + { + "epoch": 0.46014, + "grad_norm": 1.3897331502925796, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46014 + }, + { + "epoch": 0.46015, + "grad_norm": 1.5265660610335527, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 46015 + }, + { + "epoch": 0.46016, + "grad_norm": 1.4110119243176877, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 46016 + }, + { + "epoch": 0.46017, + "grad_norm": 1.6163838080165382, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 46017 + }, + { + "epoch": 0.46018, + "grad_norm": 1.514399235019992, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46018 + }, + { + "epoch": 0.46019, + "grad_norm": 2.0537105614712754, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 46019 + }, + { + "epoch": 0.4602, + "grad_norm": 1.2759527171493268, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46020 + }, + { + "epoch": 0.46021, + "grad_norm": 1.5265668300514736, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 46021 + }, + { + "epoch": 0.46022, + "grad_norm": 1.9057551898787677, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 46022 + }, + { + "epoch": 0.46023, + "grad_norm": 1.3298345314508528, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 46023 + }, + { + "epoch": 0.46024, + "grad_norm": 1.7623581394388037, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 46024 + }, + { + "epoch": 0.46025, + "grad_norm": 1.5342771980321122, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46025 + }, + { + "epoch": 0.46026, + "grad_norm": 1.5545881140637998, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 46026 + }, + { + "epoch": 0.46027, + "grad_norm": 1.4927515285351216, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46027 + }, + { + "epoch": 0.46028, + "grad_norm": 1.7330548286827092, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 46028 + }, + { + "epoch": 0.46029, + "grad_norm": 1.0616498396420742, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 46029 + }, + { + "epoch": 0.4603, + "grad_norm": 1.5043290534968077, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 46030 + }, + { + "epoch": 0.46031, + "grad_norm": 1.3555454519135608, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46031 + }, + { + "epoch": 0.46032, + "grad_norm": 1.6122195246324624, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 46032 + }, + { + "epoch": 0.46033, + "grad_norm": 1.3650881178000358, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 46033 + }, + { + "epoch": 0.46034, + "grad_norm": 1.5299919332465692, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46034 + }, + { + "epoch": 0.46035, + "grad_norm": 1.7288865976882921, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 46035 + }, + { + "epoch": 0.46036, + "grad_norm": 1.818929213896501, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46036 + }, + { + "epoch": 0.46037, + "grad_norm": 1.468383006156581, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 46037 + }, + { + "epoch": 0.46038, + "grad_norm": 1.5494945790609422, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 46038 + }, + { + "epoch": 0.46039, + "grad_norm": 1.8803036541029086, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 46039 + }, + { + "epoch": 0.4604, + "grad_norm": 1.5599460369992233, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 46040 + }, + { + "epoch": 0.46041, + "grad_norm": 1.4031900659296914, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 46041 + }, + { + "epoch": 0.46042, + "grad_norm": 1.538113552980141, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 46042 + }, + { + "epoch": 0.46043, + "grad_norm": 1.3208040756540849, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 46043 + }, + { + "epoch": 0.46044, + "grad_norm": 1.9405830102591315, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 46044 + }, + { + "epoch": 0.46045, + "grad_norm": 1.3595832593225834, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 46045 + }, + { + "epoch": 0.46046, + "grad_norm": 1.4455756773574158, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 46046 + }, + { + "epoch": 0.46047, + "grad_norm": 1.5974383845858608, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 46047 + }, + { + "epoch": 0.46048, + "grad_norm": 1.482768220918537, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46048 + }, + { + "epoch": 0.46049, + "grad_norm": 1.464752913573088, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 46049 + }, + { + "epoch": 0.4605, + "grad_norm": 1.4612732301960645, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 46050 + }, + { + "epoch": 0.46051, + "grad_norm": 1.8516882705517757, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46051 + }, + { + "epoch": 0.46052, + "grad_norm": 1.4285062795274663, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 46052 + }, + { + "epoch": 0.46053, + "grad_norm": 1.8990544882228324, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46053 + }, + { + "epoch": 0.46054, + "grad_norm": 1.4763235047476992, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46054 + }, + { + "epoch": 0.46055, + "grad_norm": 1.5223295780476864, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 46055 + }, + { + "epoch": 0.46056, + "grad_norm": 1.5848386200991975, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46056 + }, + { + "epoch": 0.46057, + "grad_norm": 1.5168359755018506, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 46057 + }, + { + "epoch": 0.46058, + "grad_norm": 1.522831456517262, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46058 + }, + { + "epoch": 0.46059, + "grad_norm": 1.3226097913629018, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 46059 + }, + { + "epoch": 0.4606, + "grad_norm": 2.003234738124594, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46060 + }, + { + "epoch": 0.46061, + "grad_norm": 1.1250907220594721, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 46061 + }, + { + "epoch": 0.46062, + "grad_norm": 1.7044686849568742, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46062 + }, + { + "epoch": 0.46063, + "grad_norm": 1.3442390954687853, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 46063 + }, + { + "epoch": 0.46064, + "grad_norm": 1.8636258037213236, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 46064 + }, + { + "epoch": 0.46065, + "grad_norm": 1.7351124293723323, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 46065 + }, + { + "epoch": 0.46066, + "grad_norm": 1.4474834357760191, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 46066 + }, + { + "epoch": 0.46067, + "grad_norm": 1.8049009133291902, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 46067 + }, + { + "epoch": 0.46068, + "grad_norm": 1.566448566228982, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 46068 + }, + { + "epoch": 0.46069, + "grad_norm": 1.3876399458796163, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 46069 + }, + { + "epoch": 0.4607, + "grad_norm": 1.8396416142066436, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46070 + }, + { + "epoch": 0.46071, + "grad_norm": 1.4781465697334077, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 46071 + }, + { + "epoch": 0.46072, + "grad_norm": 1.559698679341154, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46072 + }, + { + "epoch": 0.46073, + "grad_norm": 1.8829803005695953, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 46073 + }, + { + "epoch": 0.46074, + "grad_norm": 1.38959608824922, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 46074 + }, + { + "epoch": 0.46075, + "grad_norm": 1.623778359403943, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 46075 + }, + { + "epoch": 0.46076, + "grad_norm": 1.3958436427617014, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 46076 + }, + { + "epoch": 0.46077, + "grad_norm": 1.5684093402123753, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 46077 + }, + { + "epoch": 0.46078, + "grad_norm": 1.2682565687909033, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 46078 + }, + { + "epoch": 0.46079, + "grad_norm": 1.4839361487776352, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 46079 + }, + { + "epoch": 0.4608, + "grad_norm": 1.595287234892385, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 46080 + }, + { + "epoch": 0.46081, + "grad_norm": 1.3626336248217743, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 46081 + }, + { + "epoch": 0.46082, + "grad_norm": 1.7020973531294212, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 46082 + }, + { + "epoch": 0.46083, + "grad_norm": 1.4081140222333404, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46083 + }, + { + "epoch": 0.46084, + "grad_norm": 1.6779749495475607, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46084 + }, + { + "epoch": 0.46085, + "grad_norm": 1.7531895035435248, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 46085 + }, + { + "epoch": 0.46086, + "grad_norm": 1.6947111353762396, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 46086 + }, + { + "epoch": 0.46087, + "grad_norm": 1.7608832667466274, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46087 + }, + { + "epoch": 0.46088, + "grad_norm": 1.3771391448480208, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 46088 + }, + { + "epoch": 0.46089, + "grad_norm": 1.4005842574374796, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 46089 + }, + { + "epoch": 0.4609, + "grad_norm": 1.9324489784897467, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 46090 + }, + { + "epoch": 0.46091, + "grad_norm": 1.3075361086652848, + "learning_rate": 0.003, + "loss": 4.043, + "step": 46091 + }, + { + "epoch": 0.46092, + "grad_norm": 1.593286875508453, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 46092 + }, + { + "epoch": 0.46093, + "grad_norm": 1.3032097754856344, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 46093 + }, + { + "epoch": 0.46094, + "grad_norm": 1.6813359998481543, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 46094 + }, + { + "epoch": 0.46095, + "grad_norm": 1.5659788203569378, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 46095 + }, + { + "epoch": 0.46096, + "grad_norm": 1.8061182216295941, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46096 + }, + { + "epoch": 0.46097, + "grad_norm": 1.7499752598852174, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 46097 + }, + { + "epoch": 0.46098, + "grad_norm": 1.3442660611988326, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 46098 + }, + { + "epoch": 0.46099, + "grad_norm": 1.584380510377278, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46099 + }, + { + "epoch": 0.461, + "grad_norm": 1.5935936854056627, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 46100 + }, + { + "epoch": 0.46101, + "grad_norm": 1.630383334074375, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 46101 + }, + { + "epoch": 0.46102, + "grad_norm": 1.3676541436008915, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 46102 + }, + { + "epoch": 0.46103, + "grad_norm": 1.6803782935585916, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 46103 + }, + { + "epoch": 0.46104, + "grad_norm": 1.4786970951555225, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46104 + }, + { + "epoch": 0.46105, + "grad_norm": 1.5509961088451905, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46105 + }, + { + "epoch": 0.46106, + "grad_norm": 1.4506217081223687, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46106 + }, + { + "epoch": 0.46107, + "grad_norm": 1.6596764167490374, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 46107 + }, + { + "epoch": 0.46108, + "grad_norm": 1.272270258446616, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 46108 + }, + { + "epoch": 0.46109, + "grad_norm": 1.9650110980080866, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 46109 + }, + { + "epoch": 0.4611, + "grad_norm": 1.8554515655161818, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 46110 + }, + { + "epoch": 0.46111, + "grad_norm": 1.410267545629851, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 46111 + }, + { + "epoch": 0.46112, + "grad_norm": 1.4755575482499221, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46112 + }, + { + "epoch": 0.46113, + "grad_norm": 1.7763913622031915, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46113 + }, + { + "epoch": 0.46114, + "grad_norm": 1.3645591025552788, + "learning_rate": 0.003, + "loss": 4.056, + "step": 46114 + }, + { + "epoch": 0.46115, + "grad_norm": 1.5349537870361447, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 46115 + }, + { + "epoch": 0.46116, + "grad_norm": 1.5397791657147153, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 46116 + }, + { + "epoch": 0.46117, + "grad_norm": 1.4508207673084428, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46117 + }, + { + "epoch": 0.46118, + "grad_norm": 1.3486386332713587, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 46118 + }, + { + "epoch": 0.46119, + "grad_norm": 1.7044976900678195, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 46119 + }, + { + "epoch": 0.4612, + "grad_norm": 1.4169289022114473, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 46120 + }, + { + "epoch": 0.46121, + "grad_norm": 1.746496612810878, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 46121 + }, + { + "epoch": 0.46122, + "grad_norm": 1.6186385681891073, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 46122 + }, + { + "epoch": 0.46123, + "grad_norm": 1.9009888566923203, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 46123 + }, + { + "epoch": 0.46124, + "grad_norm": 1.2105048735969375, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46124 + }, + { + "epoch": 0.46125, + "grad_norm": 1.4839473934016616, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 46125 + }, + { + "epoch": 0.46126, + "grad_norm": 1.5767230344630885, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 46126 + }, + { + "epoch": 0.46127, + "grad_norm": 1.4561890394381827, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46127 + }, + { + "epoch": 0.46128, + "grad_norm": 1.4154337982720335, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 46128 + }, + { + "epoch": 0.46129, + "grad_norm": 1.7989927753969974, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 46129 + }, + { + "epoch": 0.4613, + "grad_norm": 1.513642846417088, + "learning_rate": 0.003, + "loss": 4.069, + "step": 46130 + }, + { + "epoch": 0.46131, + "grad_norm": 1.4386806646390844, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 46131 + }, + { + "epoch": 0.46132, + "grad_norm": 1.6497285824105985, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46132 + }, + { + "epoch": 0.46133, + "grad_norm": 1.8975706301427913, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 46133 + }, + { + "epoch": 0.46134, + "grad_norm": 1.2194199644168777, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 46134 + }, + { + "epoch": 0.46135, + "grad_norm": 1.5952729373706966, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 46135 + }, + { + "epoch": 0.46136, + "grad_norm": 1.6764860366339094, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46136 + }, + { + "epoch": 0.46137, + "grad_norm": 1.4537924419264352, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46137 + }, + { + "epoch": 0.46138, + "grad_norm": 1.8568003062166698, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 46138 + }, + { + "epoch": 0.46139, + "grad_norm": 1.4799429405117384, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 46139 + }, + { + "epoch": 0.4614, + "grad_norm": 1.6524225624506788, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 46140 + }, + { + "epoch": 0.46141, + "grad_norm": 1.4837888846651996, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46141 + }, + { + "epoch": 0.46142, + "grad_norm": 1.6695295265402141, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46142 + }, + { + "epoch": 0.46143, + "grad_norm": 1.6137988039334954, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 46143 + }, + { + "epoch": 0.46144, + "grad_norm": 1.2789790732105957, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 46144 + }, + { + "epoch": 0.46145, + "grad_norm": 1.5619167838863721, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 46145 + }, + { + "epoch": 0.46146, + "grad_norm": 1.695557542160647, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 46146 + }, + { + "epoch": 0.46147, + "grad_norm": 1.4121243058325348, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 46147 + }, + { + "epoch": 0.46148, + "grad_norm": 1.6017159495996551, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 46148 + }, + { + "epoch": 0.46149, + "grad_norm": 1.22571982965878, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46149 + }, + { + "epoch": 0.4615, + "grad_norm": 2.0198280774084774, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 46150 + }, + { + "epoch": 0.46151, + "grad_norm": 1.31641938850358, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 46151 + }, + { + "epoch": 0.46152, + "grad_norm": 2.0473420757887144, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 46152 + }, + { + "epoch": 0.46153, + "grad_norm": 1.2760002423544161, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 46153 + }, + { + "epoch": 0.46154, + "grad_norm": 1.7770842129193123, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46154 + }, + { + "epoch": 0.46155, + "grad_norm": 1.5184070735649309, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 46155 + }, + { + "epoch": 0.46156, + "grad_norm": 1.607707888701143, + "learning_rate": 0.003, + "loss": 4.049, + "step": 46156 + }, + { + "epoch": 0.46157, + "grad_norm": 1.787394972485966, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 46157 + }, + { + "epoch": 0.46158, + "grad_norm": 1.3687510093234034, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 46158 + }, + { + "epoch": 0.46159, + "grad_norm": 1.589402488968297, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 46159 + }, + { + "epoch": 0.4616, + "grad_norm": 1.510295417692623, + "learning_rate": 0.003, + "loss": 4.013, + "step": 46160 + }, + { + "epoch": 0.46161, + "grad_norm": 1.4012027711976633, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 46161 + }, + { + "epoch": 0.46162, + "grad_norm": 1.4237451643589525, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 46162 + }, + { + "epoch": 0.46163, + "grad_norm": 1.5789876566435324, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46163 + }, + { + "epoch": 0.46164, + "grad_norm": 1.7299144589999667, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 46164 + }, + { + "epoch": 0.46165, + "grad_norm": 1.3515235202168092, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 46165 + }, + { + "epoch": 0.46166, + "grad_norm": 1.6818142653269579, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 46166 + }, + { + "epoch": 0.46167, + "grad_norm": 1.3591411753374778, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 46167 + }, + { + "epoch": 0.46168, + "grad_norm": 1.5023719500282087, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 46168 + }, + { + "epoch": 0.46169, + "grad_norm": 1.6119161059188145, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 46169 + }, + { + "epoch": 0.4617, + "grad_norm": 1.5916812071206219, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 46170 + }, + { + "epoch": 0.46171, + "grad_norm": 1.5073127117875846, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 46171 + }, + { + "epoch": 0.46172, + "grad_norm": 1.710346330796125, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 46172 + }, + { + "epoch": 0.46173, + "grad_norm": 1.372202583892393, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 46173 + }, + { + "epoch": 0.46174, + "grad_norm": 1.8470124944383783, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 46174 + }, + { + "epoch": 0.46175, + "grad_norm": 1.4025580557840076, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46175 + }, + { + "epoch": 0.46176, + "grad_norm": 1.9970365103745036, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 46176 + }, + { + "epoch": 0.46177, + "grad_norm": 1.544999946858141, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 46177 + }, + { + "epoch": 0.46178, + "grad_norm": 1.280115826161388, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 46178 + }, + { + "epoch": 0.46179, + "grad_norm": 1.6039961498077302, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 46179 + }, + { + "epoch": 0.4618, + "grad_norm": 1.3439331424227343, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 46180 + }, + { + "epoch": 0.46181, + "grad_norm": 1.668080748741396, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 46181 + }, + { + "epoch": 0.46182, + "grad_norm": 1.5390191825275292, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46182 + }, + { + "epoch": 0.46183, + "grad_norm": 1.686499745324358, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46183 + }, + { + "epoch": 0.46184, + "grad_norm": 1.5897541017642263, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46184 + }, + { + "epoch": 0.46185, + "grad_norm": 1.5623864104548892, + "learning_rate": 0.003, + "loss": 4.047, + "step": 46185 + }, + { + "epoch": 0.46186, + "grad_norm": 1.5426585214409771, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46186 + }, + { + "epoch": 0.46187, + "grad_norm": 1.7692817447430136, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46187 + }, + { + "epoch": 0.46188, + "grad_norm": 1.142130923705563, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 46188 + }, + { + "epoch": 0.46189, + "grad_norm": 1.719262355142803, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 46189 + }, + { + "epoch": 0.4619, + "grad_norm": 1.3679404285565446, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46190 + }, + { + "epoch": 0.46191, + "grad_norm": 1.4713398086399623, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46191 + }, + { + "epoch": 0.46192, + "grad_norm": 1.392305098566295, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 46192 + }, + { + "epoch": 0.46193, + "grad_norm": 1.7210202353161737, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 46193 + }, + { + "epoch": 0.46194, + "grad_norm": 1.3323943500216837, + "learning_rate": 0.003, + "loss": 4.025, + "step": 46194 + }, + { + "epoch": 0.46195, + "grad_norm": 2.1850370042973775, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 46195 + }, + { + "epoch": 0.46196, + "grad_norm": 1.6117466882540301, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 46196 + }, + { + "epoch": 0.46197, + "grad_norm": 1.3892190199114889, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 46197 + }, + { + "epoch": 0.46198, + "grad_norm": 1.6271372633475858, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 46198 + }, + { + "epoch": 0.46199, + "grad_norm": 1.796595789946628, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 46199 + }, + { + "epoch": 0.462, + "grad_norm": 1.2059830035982082, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 46200 + }, + { + "epoch": 0.46201, + "grad_norm": 1.7325231355494428, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 46201 + }, + { + "epoch": 0.46202, + "grad_norm": 1.0617368123695674, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46202 + }, + { + "epoch": 0.46203, + "grad_norm": 1.9264866909915437, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46203 + }, + { + "epoch": 0.46204, + "grad_norm": 1.3463261653142276, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 46204 + }, + { + "epoch": 0.46205, + "grad_norm": 1.927017620340688, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 46205 + }, + { + "epoch": 0.46206, + "grad_norm": 1.5714542992617493, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 46206 + }, + { + "epoch": 0.46207, + "grad_norm": 1.5399848890981553, + "learning_rate": 0.003, + "loss": 4.052, + "step": 46207 + }, + { + "epoch": 0.46208, + "grad_norm": 1.531197579798436, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 46208 + }, + { + "epoch": 0.46209, + "grad_norm": 1.524273412670347, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46209 + }, + { + "epoch": 0.4621, + "grad_norm": 1.463806258553445, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 46210 + }, + { + "epoch": 0.46211, + "grad_norm": 1.4389020087503646, + "learning_rate": 0.003, + "loss": 4.049, + "step": 46211 + }, + { + "epoch": 0.46212, + "grad_norm": 1.6339664295929228, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 46212 + }, + { + "epoch": 0.46213, + "grad_norm": 1.439828863442244, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46213 + }, + { + "epoch": 0.46214, + "grad_norm": 1.6852984937251914, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 46214 + }, + { + "epoch": 0.46215, + "grad_norm": 1.3993312799825732, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46215 + }, + { + "epoch": 0.46216, + "grad_norm": 1.5541273306830523, + "learning_rate": 0.003, + "loss": 4.061, + "step": 46216 + }, + { + "epoch": 0.46217, + "grad_norm": 1.6343884536317188, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 46217 + }, + { + "epoch": 0.46218, + "grad_norm": 1.8107728397570328, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46218 + }, + { + "epoch": 0.46219, + "grad_norm": 1.2872407676138744, + "learning_rate": 0.003, + "loss": 4.051, + "step": 46219 + }, + { + "epoch": 0.4622, + "grad_norm": 1.6255270367900487, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 46220 + }, + { + "epoch": 0.46221, + "grad_norm": 1.4817847658125474, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46221 + }, + { + "epoch": 0.46222, + "grad_norm": 2.012519785804525, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 46222 + }, + { + "epoch": 0.46223, + "grad_norm": 1.5770069372501796, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 46223 + }, + { + "epoch": 0.46224, + "grad_norm": 1.6478076807836037, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 46224 + }, + { + "epoch": 0.46225, + "grad_norm": 1.6750501768098325, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46225 + }, + { + "epoch": 0.46226, + "grad_norm": 1.2884085162695298, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 46226 + }, + { + "epoch": 0.46227, + "grad_norm": 1.712429540925729, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46227 + }, + { + "epoch": 0.46228, + "grad_norm": 1.4040379478510312, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 46228 + }, + { + "epoch": 0.46229, + "grad_norm": 1.3965993799920111, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 46229 + }, + { + "epoch": 0.4623, + "grad_norm": 1.6310126433747096, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 46230 + }, + { + "epoch": 0.46231, + "grad_norm": 1.5952259377493008, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46231 + }, + { + "epoch": 0.46232, + "grad_norm": 1.5375027421859606, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 46232 + }, + { + "epoch": 0.46233, + "grad_norm": 1.7917358210461602, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 46233 + }, + { + "epoch": 0.46234, + "grad_norm": 1.7943975686957137, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46234 + }, + { + "epoch": 0.46235, + "grad_norm": 1.5845704541940757, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46235 + }, + { + "epoch": 0.46236, + "grad_norm": 1.6342307938736873, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 46236 + }, + { + "epoch": 0.46237, + "grad_norm": 1.7520592356164717, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 46237 + }, + { + "epoch": 0.46238, + "grad_norm": 1.1285998458277147, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 46238 + }, + { + "epoch": 0.46239, + "grad_norm": 1.951503752572313, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46239 + }, + { + "epoch": 0.4624, + "grad_norm": 1.3841580103416928, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46240 + }, + { + "epoch": 0.46241, + "grad_norm": 1.8804545452449704, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46241 + }, + { + "epoch": 0.46242, + "grad_norm": 1.369708485861899, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 46242 + }, + { + "epoch": 0.46243, + "grad_norm": 1.6977868032713086, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 46243 + }, + { + "epoch": 0.46244, + "grad_norm": 1.3850884304530537, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 46244 + }, + { + "epoch": 0.46245, + "grad_norm": 1.5173852452093457, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 46245 + }, + { + "epoch": 0.46246, + "grad_norm": 1.3759188104633313, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 46246 + }, + { + "epoch": 0.46247, + "grad_norm": 1.79379913887442, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46247 + }, + { + "epoch": 0.46248, + "grad_norm": 1.4533246803918267, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46248 + }, + { + "epoch": 0.46249, + "grad_norm": 1.6775226358899162, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 46249 + }, + { + "epoch": 0.4625, + "grad_norm": 1.55820443776254, + "learning_rate": 0.003, + "loss": 4.056, + "step": 46250 + }, + { + "epoch": 0.46251, + "grad_norm": 2.3067171257820482, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46251 + }, + { + "epoch": 0.46252, + "grad_norm": 1.509151093543349, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46252 + }, + { + "epoch": 0.46253, + "grad_norm": 1.6164743750860346, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 46253 + }, + { + "epoch": 0.46254, + "grad_norm": 1.7097355189986196, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 46254 + }, + { + "epoch": 0.46255, + "grad_norm": 1.457426866553841, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 46255 + }, + { + "epoch": 0.46256, + "grad_norm": 1.6262100075961345, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 46256 + }, + { + "epoch": 0.46257, + "grad_norm": 1.4018938359469313, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 46257 + }, + { + "epoch": 0.46258, + "grad_norm": 1.455725719523373, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46258 + }, + { + "epoch": 0.46259, + "grad_norm": 1.6053277701506585, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 46259 + }, + { + "epoch": 0.4626, + "grad_norm": 1.794285938397021, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 46260 + }, + { + "epoch": 0.46261, + "grad_norm": 1.3724387590995766, + "learning_rate": 0.003, + "loss": 4.027, + "step": 46261 + }, + { + "epoch": 0.46262, + "grad_norm": 1.5341696299050658, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46262 + }, + { + "epoch": 0.46263, + "grad_norm": 1.5665613665150975, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 46263 + }, + { + "epoch": 0.46264, + "grad_norm": 1.6668199144159457, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 46264 + }, + { + "epoch": 0.46265, + "grad_norm": 1.3759596484020173, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 46265 + }, + { + "epoch": 0.46266, + "grad_norm": 1.668938782987044, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46266 + }, + { + "epoch": 0.46267, + "grad_norm": 1.401918621817006, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 46267 + }, + { + "epoch": 0.46268, + "grad_norm": 1.607715660282667, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 46268 + }, + { + "epoch": 0.46269, + "grad_norm": 1.6092895860419292, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 46269 + }, + { + "epoch": 0.4627, + "grad_norm": 1.5962541110638067, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46270 + }, + { + "epoch": 0.46271, + "grad_norm": 1.5991814807341296, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 46271 + }, + { + "epoch": 0.46272, + "grad_norm": 1.6787459368200213, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 46272 + }, + { + "epoch": 0.46273, + "grad_norm": 1.4703229436893002, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46273 + }, + { + "epoch": 0.46274, + "grad_norm": 1.635271120192994, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 46274 + }, + { + "epoch": 0.46275, + "grad_norm": 1.6797037346628272, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 46275 + }, + { + "epoch": 0.46276, + "grad_norm": 1.5280119563446977, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 46276 + }, + { + "epoch": 0.46277, + "grad_norm": 1.6038057812256146, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46277 + }, + { + "epoch": 0.46278, + "grad_norm": 1.7370527859354505, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 46278 + }, + { + "epoch": 0.46279, + "grad_norm": 1.3674606168011938, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 46279 + }, + { + "epoch": 0.4628, + "grad_norm": 1.6366100877938405, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 46280 + }, + { + "epoch": 0.46281, + "grad_norm": 1.4960304473295085, + "learning_rate": 0.003, + "loss": 4.015, + "step": 46281 + }, + { + "epoch": 0.46282, + "grad_norm": 1.610685917070838, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 46282 + }, + { + "epoch": 0.46283, + "grad_norm": 1.2851134626742373, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 46283 + }, + { + "epoch": 0.46284, + "grad_norm": 1.8074055620081693, + "learning_rate": 0.003, + "loss": 4.027, + "step": 46284 + }, + { + "epoch": 0.46285, + "grad_norm": 1.2121874022757415, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 46285 + }, + { + "epoch": 0.46286, + "grad_norm": 1.5006836886938542, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 46286 + }, + { + "epoch": 0.46287, + "grad_norm": 1.495258062080903, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46287 + }, + { + "epoch": 0.46288, + "grad_norm": 2.0356028513978672, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46288 + }, + { + "epoch": 0.46289, + "grad_norm": 1.5403177272163069, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 46289 + }, + { + "epoch": 0.4629, + "grad_norm": 1.2352033888778022, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 46290 + }, + { + "epoch": 0.46291, + "grad_norm": 1.7459959138017327, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 46291 + }, + { + "epoch": 0.46292, + "grad_norm": 1.1705270773080987, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 46292 + }, + { + "epoch": 0.46293, + "grad_norm": 1.7187537621221662, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 46293 + }, + { + "epoch": 0.46294, + "grad_norm": 1.417960829263547, + "learning_rate": 0.003, + "loss": 4.025, + "step": 46294 + }, + { + "epoch": 0.46295, + "grad_norm": 2.086426067371065, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 46295 + }, + { + "epoch": 0.46296, + "grad_norm": 1.2752375170806078, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 46296 + }, + { + "epoch": 0.46297, + "grad_norm": 1.4378494278908178, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 46297 + }, + { + "epoch": 0.46298, + "grad_norm": 1.7258358429268859, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 46298 + }, + { + "epoch": 0.46299, + "grad_norm": 1.9383425913884724, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 46299 + }, + { + "epoch": 0.463, + "grad_norm": 1.477009535738738, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 46300 + }, + { + "epoch": 0.46301, + "grad_norm": 1.7469869773386324, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 46301 + }, + { + "epoch": 0.46302, + "grad_norm": 1.5412385351476232, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 46302 + }, + { + "epoch": 0.46303, + "grad_norm": 1.4589087865857135, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 46303 + }, + { + "epoch": 0.46304, + "grad_norm": 1.431602483374739, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46304 + }, + { + "epoch": 0.46305, + "grad_norm": 1.7148326769408593, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 46305 + }, + { + "epoch": 0.46306, + "grad_norm": 1.4402680321646193, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 46306 + }, + { + "epoch": 0.46307, + "grad_norm": 1.6671800358884252, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 46307 + }, + { + "epoch": 0.46308, + "grad_norm": 1.4236549538776622, + "learning_rate": 0.003, + "loss": 4.066, + "step": 46308 + }, + { + "epoch": 0.46309, + "grad_norm": 1.5788352168720738, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46309 + }, + { + "epoch": 0.4631, + "grad_norm": 1.2789650923881095, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 46310 + }, + { + "epoch": 0.46311, + "grad_norm": 1.6383086686438286, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 46311 + }, + { + "epoch": 0.46312, + "grad_norm": 1.908528416603557, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46312 + }, + { + "epoch": 0.46313, + "grad_norm": 1.6488308068519402, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 46313 + }, + { + "epoch": 0.46314, + "grad_norm": 1.5113788410744189, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46314 + }, + { + "epoch": 0.46315, + "grad_norm": 1.2924589731297746, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46315 + }, + { + "epoch": 0.46316, + "grad_norm": 1.9347265970239176, + "learning_rate": 0.003, + "loss": 4.022, + "step": 46316 + }, + { + "epoch": 0.46317, + "grad_norm": 1.347167186893312, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 46317 + }, + { + "epoch": 0.46318, + "grad_norm": 1.5591051712248138, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 46318 + }, + { + "epoch": 0.46319, + "grad_norm": 1.384663827043162, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46319 + }, + { + "epoch": 0.4632, + "grad_norm": 1.7228152264590795, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 46320 + }, + { + "epoch": 0.46321, + "grad_norm": 1.4085260794613905, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 46321 + }, + { + "epoch": 0.46322, + "grad_norm": 1.390398548316592, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 46322 + }, + { + "epoch": 0.46323, + "grad_norm": 1.6934659475123466, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 46323 + }, + { + "epoch": 0.46324, + "grad_norm": 1.22562115281648, + "learning_rate": 0.003, + "loss": 4.04, + "step": 46324 + }, + { + "epoch": 0.46325, + "grad_norm": 1.7134031765127244, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 46325 + }, + { + "epoch": 0.46326, + "grad_norm": 1.2858706599610135, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46326 + }, + { + "epoch": 0.46327, + "grad_norm": 1.8367254943650757, + "learning_rate": 0.003, + "loss": 4.015, + "step": 46327 + }, + { + "epoch": 0.46328, + "grad_norm": 1.1542422850709075, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 46328 + }, + { + "epoch": 0.46329, + "grad_norm": 1.8215809391174105, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 46329 + }, + { + "epoch": 0.4633, + "grad_norm": 1.4130790697222293, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 46330 + }, + { + "epoch": 0.46331, + "grad_norm": 1.6240829796052427, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 46331 + }, + { + "epoch": 0.46332, + "grad_norm": 1.6539710222137318, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46332 + }, + { + "epoch": 0.46333, + "grad_norm": 1.651514356345737, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 46333 + }, + { + "epoch": 0.46334, + "grad_norm": 1.6636401881961553, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46334 + }, + { + "epoch": 0.46335, + "grad_norm": 1.6681774863080172, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46335 + }, + { + "epoch": 0.46336, + "grad_norm": 1.3087008708550707, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 46336 + }, + { + "epoch": 0.46337, + "grad_norm": 1.8925739908095116, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 46337 + }, + { + "epoch": 0.46338, + "grad_norm": 1.3337952853637474, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 46338 + }, + { + "epoch": 0.46339, + "grad_norm": 1.6504660323102898, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 46339 + }, + { + "epoch": 0.4634, + "grad_norm": 1.3556572483571263, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46340 + }, + { + "epoch": 0.46341, + "grad_norm": 1.6002399874943531, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46341 + }, + { + "epoch": 0.46342, + "grad_norm": 1.561964173050749, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46342 + }, + { + "epoch": 0.46343, + "grad_norm": 1.6868622824169313, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 46343 + }, + { + "epoch": 0.46344, + "grad_norm": 1.5049557497276407, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46344 + }, + { + "epoch": 0.46345, + "grad_norm": 1.5216892202743468, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 46345 + }, + { + "epoch": 0.46346, + "grad_norm": 1.5358939793832813, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46346 + }, + { + "epoch": 0.46347, + "grad_norm": 1.3585135237601147, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46347 + }, + { + "epoch": 0.46348, + "grad_norm": 1.6396790841794888, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 46348 + }, + { + "epoch": 0.46349, + "grad_norm": 1.6068897978425012, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46349 + }, + { + "epoch": 0.4635, + "grad_norm": 1.4759279588268017, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 46350 + }, + { + "epoch": 0.46351, + "grad_norm": 1.6715596156870105, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46351 + }, + { + "epoch": 0.46352, + "grad_norm": 1.4893166279948025, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 46352 + }, + { + "epoch": 0.46353, + "grad_norm": 1.6231193906834953, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46353 + }, + { + "epoch": 0.46354, + "grad_norm": 1.829489628290719, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 46354 + }, + { + "epoch": 0.46355, + "grad_norm": 1.5945724094595812, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 46355 + }, + { + "epoch": 0.46356, + "grad_norm": 1.1453424686431362, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 46356 + }, + { + "epoch": 0.46357, + "grad_norm": 1.7769531166604275, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46357 + }, + { + "epoch": 0.46358, + "grad_norm": 1.1126170429566866, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 46358 + }, + { + "epoch": 0.46359, + "grad_norm": 1.9233019455139344, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 46359 + }, + { + "epoch": 0.4636, + "grad_norm": 1.4973799919907862, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46360 + }, + { + "epoch": 0.46361, + "grad_norm": 1.5159795373204672, + "learning_rate": 0.003, + "loss": 4.055, + "step": 46361 + }, + { + "epoch": 0.46362, + "grad_norm": 1.6392641347569326, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 46362 + }, + { + "epoch": 0.46363, + "grad_norm": 1.5228406727791333, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 46363 + }, + { + "epoch": 0.46364, + "grad_norm": 1.2862647149926574, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 46364 + }, + { + "epoch": 0.46365, + "grad_norm": 1.4496127845796976, + "learning_rate": 0.003, + "loss": 4.053, + "step": 46365 + }, + { + "epoch": 0.46366, + "grad_norm": 1.4611873012244343, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 46366 + }, + { + "epoch": 0.46367, + "grad_norm": 1.6645337525097958, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 46367 + }, + { + "epoch": 0.46368, + "grad_norm": 1.5460814404256653, + "learning_rate": 0.003, + "loss": 4.037, + "step": 46368 + }, + { + "epoch": 0.46369, + "grad_norm": 1.7438034431295513, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 46369 + }, + { + "epoch": 0.4637, + "grad_norm": 1.5362007924926966, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 46370 + }, + { + "epoch": 0.46371, + "grad_norm": 1.6072308573669405, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 46371 + }, + { + "epoch": 0.46372, + "grad_norm": 1.5904048590003999, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 46372 + }, + { + "epoch": 0.46373, + "grad_norm": 1.6903362880214243, + "learning_rate": 0.003, + "loss": 4.047, + "step": 46373 + }, + { + "epoch": 0.46374, + "grad_norm": 1.4542797283027438, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 46374 + }, + { + "epoch": 0.46375, + "grad_norm": 1.938572509344994, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 46375 + }, + { + "epoch": 0.46376, + "grad_norm": 1.6052432719067877, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 46376 + }, + { + "epoch": 0.46377, + "grad_norm": 1.3578646838782726, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 46377 + }, + { + "epoch": 0.46378, + "grad_norm": 1.4069890223610475, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46378 + }, + { + "epoch": 0.46379, + "grad_norm": 1.7777793060356029, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 46379 + }, + { + "epoch": 0.4638, + "grad_norm": 1.4143332352339495, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 46380 + }, + { + "epoch": 0.46381, + "grad_norm": 1.6016395068656164, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 46381 + }, + { + "epoch": 0.46382, + "grad_norm": 1.343254479385994, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 46382 + }, + { + "epoch": 0.46383, + "grad_norm": 1.4832370898354434, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 46383 + }, + { + "epoch": 0.46384, + "grad_norm": 1.6847796138901248, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 46384 + }, + { + "epoch": 0.46385, + "grad_norm": 1.2882501621336235, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 46385 + }, + { + "epoch": 0.46386, + "grad_norm": 1.5266413060939377, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46386 + }, + { + "epoch": 0.46387, + "grad_norm": 1.4578876643237315, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 46387 + }, + { + "epoch": 0.46388, + "grad_norm": 1.4268784260636065, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46388 + }, + { + "epoch": 0.46389, + "grad_norm": 1.2942568083351482, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 46389 + }, + { + "epoch": 0.4639, + "grad_norm": 1.588598961414253, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46390 + }, + { + "epoch": 0.46391, + "grad_norm": 1.7168407098471326, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 46391 + }, + { + "epoch": 0.46392, + "grad_norm": 1.7808081696241753, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 46392 + }, + { + "epoch": 0.46393, + "grad_norm": 1.7985309382958543, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 46393 + }, + { + "epoch": 0.46394, + "grad_norm": 1.6045959413906319, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 46394 + }, + { + "epoch": 0.46395, + "grad_norm": 1.5091454864589346, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46395 + }, + { + "epoch": 0.46396, + "grad_norm": 1.8650458239647631, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 46396 + }, + { + "epoch": 0.46397, + "grad_norm": 1.2897052903894244, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 46397 + }, + { + "epoch": 0.46398, + "grad_norm": 1.5442778189706245, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46398 + }, + { + "epoch": 0.46399, + "grad_norm": 1.3887189916862326, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 46399 + }, + { + "epoch": 0.464, + "grad_norm": 1.4789016830902295, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 46400 + }, + { + "epoch": 0.46401, + "grad_norm": 1.6912480132600793, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 46401 + }, + { + "epoch": 0.46402, + "grad_norm": 1.301458333841317, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 46402 + }, + { + "epoch": 0.46403, + "grad_norm": 1.5624164389061062, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46403 + }, + { + "epoch": 0.46404, + "grad_norm": 1.3980774637762647, + "learning_rate": 0.003, + "loss": 4.03, + "step": 46404 + }, + { + "epoch": 0.46405, + "grad_norm": 1.732472714694337, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 46405 + }, + { + "epoch": 0.46406, + "grad_norm": 1.5455450403049573, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 46406 + }, + { + "epoch": 0.46407, + "grad_norm": 1.5303398998799038, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 46407 + }, + { + "epoch": 0.46408, + "grad_norm": 1.7323075764790885, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 46408 + }, + { + "epoch": 0.46409, + "grad_norm": 1.4095415700071956, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 46409 + }, + { + "epoch": 0.4641, + "grad_norm": 1.7240199436866679, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 46410 + }, + { + "epoch": 0.46411, + "grad_norm": 1.6127399301253385, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46411 + }, + { + "epoch": 0.46412, + "grad_norm": 1.5547456038666332, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46412 + }, + { + "epoch": 0.46413, + "grad_norm": 1.5558814521853768, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 46413 + }, + { + "epoch": 0.46414, + "grad_norm": 1.651969580992259, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 46414 + }, + { + "epoch": 0.46415, + "grad_norm": 1.567976697072954, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 46415 + }, + { + "epoch": 0.46416, + "grad_norm": 1.7116383111295423, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 46416 + }, + { + "epoch": 0.46417, + "grad_norm": 1.2847062018836775, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 46417 + }, + { + "epoch": 0.46418, + "grad_norm": 1.5086659577739487, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 46418 + }, + { + "epoch": 0.46419, + "grad_norm": 1.5710397217848686, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 46419 + }, + { + "epoch": 0.4642, + "grad_norm": 1.7205851322043546, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 46420 + }, + { + "epoch": 0.46421, + "grad_norm": 1.407922274784457, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46421 + }, + { + "epoch": 0.46422, + "grad_norm": 1.5753416935836049, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 46422 + }, + { + "epoch": 0.46423, + "grad_norm": 1.4596928808243264, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 46423 + }, + { + "epoch": 0.46424, + "grad_norm": 1.5282665383228804, + "learning_rate": 0.003, + "loss": 4.022, + "step": 46424 + }, + { + "epoch": 0.46425, + "grad_norm": 1.2982158481317003, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46425 + }, + { + "epoch": 0.46426, + "grad_norm": 1.9609776928803089, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 46426 + }, + { + "epoch": 0.46427, + "grad_norm": 1.1392374569157373, + "learning_rate": 0.003, + "loss": 4.038, + "step": 46427 + }, + { + "epoch": 0.46428, + "grad_norm": 2.0124227139680877, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 46428 + }, + { + "epoch": 0.46429, + "grad_norm": 1.5573654374625774, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46429 + }, + { + "epoch": 0.4643, + "grad_norm": 1.6838920305663863, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 46430 + }, + { + "epoch": 0.46431, + "grad_norm": 1.5126151074241367, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 46431 + }, + { + "epoch": 0.46432, + "grad_norm": 1.6313219850272735, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46432 + }, + { + "epoch": 0.46433, + "grad_norm": 1.6743589459101647, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 46433 + }, + { + "epoch": 0.46434, + "grad_norm": 1.6751378998821893, + "learning_rate": 0.003, + "loss": 4.043, + "step": 46434 + }, + { + "epoch": 0.46435, + "grad_norm": 1.4775094428393973, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 46435 + }, + { + "epoch": 0.46436, + "grad_norm": 1.7713194383992303, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 46436 + }, + { + "epoch": 0.46437, + "grad_norm": 1.35879958780016, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 46437 + }, + { + "epoch": 0.46438, + "grad_norm": 1.5105058982138528, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46438 + }, + { + "epoch": 0.46439, + "grad_norm": 1.6313774975237347, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 46439 + }, + { + "epoch": 0.4644, + "grad_norm": 1.6022225365409881, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 46440 + }, + { + "epoch": 0.46441, + "grad_norm": 1.5787381519732007, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46441 + }, + { + "epoch": 0.46442, + "grad_norm": 1.6558021115892338, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 46442 + }, + { + "epoch": 0.46443, + "grad_norm": 1.3002586730815093, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 46443 + }, + { + "epoch": 0.46444, + "grad_norm": 1.6327359981006915, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46444 + }, + { + "epoch": 0.46445, + "grad_norm": 1.1539310234889795, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46445 + }, + { + "epoch": 0.46446, + "grad_norm": 1.5595504543990735, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 46446 + }, + { + "epoch": 0.46447, + "grad_norm": 1.3124377650542274, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46447 + }, + { + "epoch": 0.46448, + "grad_norm": 1.651389045499618, + "learning_rate": 0.003, + "loss": 4.064, + "step": 46448 + }, + { + "epoch": 0.46449, + "grad_norm": 1.802602260821239, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 46449 + }, + { + "epoch": 0.4645, + "grad_norm": 1.5839917424401988, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46450 + }, + { + "epoch": 0.46451, + "grad_norm": 1.8443437039091535, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 46451 + }, + { + "epoch": 0.46452, + "grad_norm": 1.5922494851468012, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46452 + }, + { + "epoch": 0.46453, + "grad_norm": 1.5632009943822305, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46453 + }, + { + "epoch": 0.46454, + "grad_norm": 1.4928978990375967, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 46454 + }, + { + "epoch": 0.46455, + "grad_norm": 1.4710779397794942, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46455 + }, + { + "epoch": 0.46456, + "grad_norm": 1.5079775182728983, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 46456 + }, + { + "epoch": 0.46457, + "grad_norm": 1.7814822775828045, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 46457 + }, + { + "epoch": 0.46458, + "grad_norm": 1.472288002734184, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46458 + }, + { + "epoch": 0.46459, + "grad_norm": 1.6385817995537213, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46459 + }, + { + "epoch": 0.4646, + "grad_norm": 1.4220762384170174, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46460 + }, + { + "epoch": 0.46461, + "grad_norm": 2.0039645745791748, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 46461 + }, + { + "epoch": 0.46462, + "grad_norm": 1.4709478079957765, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 46462 + }, + { + "epoch": 0.46463, + "grad_norm": 1.6695177754645354, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46463 + }, + { + "epoch": 0.46464, + "grad_norm": 1.531350766148128, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46464 + }, + { + "epoch": 0.46465, + "grad_norm": 1.6917924626936691, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 46465 + }, + { + "epoch": 0.46466, + "grad_norm": 1.2661879434908325, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 46466 + }, + { + "epoch": 0.46467, + "grad_norm": 1.6490605373477345, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 46467 + }, + { + "epoch": 0.46468, + "grad_norm": 1.3977284115314237, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46468 + }, + { + "epoch": 0.46469, + "grad_norm": 1.3781515590705324, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46469 + }, + { + "epoch": 0.4647, + "grad_norm": 1.663776324157351, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 46470 + }, + { + "epoch": 0.46471, + "grad_norm": 1.4551300657460353, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46471 + }, + { + "epoch": 0.46472, + "grad_norm": 1.755809034322323, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 46472 + }, + { + "epoch": 0.46473, + "grad_norm": 1.6547717211303095, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 46473 + }, + { + "epoch": 0.46474, + "grad_norm": 1.5782892961043833, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 46474 + }, + { + "epoch": 0.46475, + "grad_norm": 1.5362639399266649, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 46475 + }, + { + "epoch": 0.46476, + "grad_norm": 1.729407600263797, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46476 + }, + { + "epoch": 0.46477, + "grad_norm": 1.4264970600522318, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46477 + }, + { + "epoch": 0.46478, + "grad_norm": 1.285608248694344, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 46478 + }, + { + "epoch": 0.46479, + "grad_norm": 1.5320962610544924, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 46479 + }, + { + "epoch": 0.4648, + "grad_norm": 1.3487725500234835, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 46480 + }, + { + "epoch": 0.46481, + "grad_norm": 1.6325163630289288, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 46481 + }, + { + "epoch": 0.46482, + "grad_norm": 1.2711545168835852, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 46482 + }, + { + "epoch": 0.46483, + "grad_norm": 1.6503419196781242, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 46483 + }, + { + "epoch": 0.46484, + "grad_norm": 1.4966791215767956, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46484 + }, + { + "epoch": 0.46485, + "grad_norm": 2.1311045470578054, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 46485 + }, + { + "epoch": 0.46486, + "grad_norm": 1.2495028641036352, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 46486 + }, + { + "epoch": 0.46487, + "grad_norm": 1.5807261960237098, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 46487 + }, + { + "epoch": 0.46488, + "grad_norm": 1.5053570903470261, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 46488 + }, + { + "epoch": 0.46489, + "grad_norm": 1.90241657129031, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 46489 + }, + { + "epoch": 0.4649, + "grad_norm": 1.365482366888161, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 46490 + }, + { + "epoch": 0.46491, + "grad_norm": 1.4716829383481531, + "learning_rate": 0.003, + "loss": 4.042, + "step": 46491 + }, + { + "epoch": 0.46492, + "grad_norm": 1.5922490315953521, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46492 + }, + { + "epoch": 0.46493, + "grad_norm": 1.3796566049509644, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46493 + }, + { + "epoch": 0.46494, + "grad_norm": 1.5033114720489071, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 46494 + }, + { + "epoch": 0.46495, + "grad_norm": 1.6056045606349776, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 46495 + }, + { + "epoch": 0.46496, + "grad_norm": 1.3665523921680043, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 46496 + }, + { + "epoch": 0.46497, + "grad_norm": 1.5647396148550003, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 46497 + }, + { + "epoch": 0.46498, + "grad_norm": 1.9633021891767282, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46498 + }, + { + "epoch": 0.46499, + "grad_norm": 1.5673229707630127, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 46499 + }, + { + "epoch": 0.465, + "grad_norm": 1.3808094638390993, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 46500 + }, + { + "epoch": 0.46501, + "grad_norm": 1.6345306964327784, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 46501 + }, + { + "epoch": 0.46502, + "grad_norm": 1.4713829645401453, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 46502 + }, + { + "epoch": 0.46503, + "grad_norm": 1.4006083996215635, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 46503 + }, + { + "epoch": 0.46504, + "grad_norm": 1.7390810677572617, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 46504 + }, + { + "epoch": 0.46505, + "grad_norm": 1.3347239280630812, + "learning_rate": 0.003, + "loss": 4.019, + "step": 46505 + }, + { + "epoch": 0.46506, + "grad_norm": 1.653017735662681, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 46506 + }, + { + "epoch": 0.46507, + "grad_norm": 1.3905118558414333, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46507 + }, + { + "epoch": 0.46508, + "grad_norm": 1.9885718559801535, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 46508 + }, + { + "epoch": 0.46509, + "grad_norm": 1.721200646468149, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 46509 + }, + { + "epoch": 0.4651, + "grad_norm": 1.3355851369135634, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 46510 + }, + { + "epoch": 0.46511, + "grad_norm": 1.3283741533219193, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46511 + }, + { + "epoch": 0.46512, + "grad_norm": 1.5443109827162667, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 46512 + }, + { + "epoch": 0.46513, + "grad_norm": 1.4932229380049238, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 46513 + }, + { + "epoch": 0.46514, + "grad_norm": 1.510878639147062, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 46514 + }, + { + "epoch": 0.46515, + "grad_norm": 1.4946648462392393, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 46515 + }, + { + "epoch": 0.46516, + "grad_norm": 1.670210900251731, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 46516 + }, + { + "epoch": 0.46517, + "grad_norm": 1.4131773648636432, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 46517 + }, + { + "epoch": 0.46518, + "grad_norm": 1.7416088223325927, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 46518 + }, + { + "epoch": 0.46519, + "grad_norm": 1.8289854597924047, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46519 + }, + { + "epoch": 0.4652, + "grad_norm": 1.3849499434096653, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 46520 + }, + { + "epoch": 0.46521, + "grad_norm": 1.6887482819214967, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46521 + }, + { + "epoch": 0.46522, + "grad_norm": 1.3164813294822353, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46522 + }, + { + "epoch": 0.46523, + "grad_norm": 2.17985777545075, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46523 + }, + { + "epoch": 0.46524, + "grad_norm": 1.4290336365818888, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 46524 + }, + { + "epoch": 0.46525, + "grad_norm": 1.7412099848837934, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46525 + }, + { + "epoch": 0.46526, + "grad_norm": 1.281969873005756, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46526 + }, + { + "epoch": 0.46527, + "grad_norm": 1.4904839195264359, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 46527 + }, + { + "epoch": 0.46528, + "grad_norm": 1.6544409936573115, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 46528 + }, + { + "epoch": 0.46529, + "grad_norm": 1.3917811074770723, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 46529 + }, + { + "epoch": 0.4653, + "grad_norm": 1.5151637144726844, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46530 + }, + { + "epoch": 0.46531, + "grad_norm": 1.714906514966583, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 46531 + }, + { + "epoch": 0.46532, + "grad_norm": 1.483790877665326, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 46532 + }, + { + "epoch": 0.46533, + "grad_norm": 1.6383093370012303, + "learning_rate": 0.003, + "loss": 4.012, + "step": 46533 + }, + { + "epoch": 0.46534, + "grad_norm": 1.3751691008366662, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 46534 + }, + { + "epoch": 0.46535, + "grad_norm": 1.42836224078544, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 46535 + }, + { + "epoch": 0.46536, + "grad_norm": 1.7708677726905941, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 46536 + }, + { + "epoch": 0.46537, + "grad_norm": 1.605325457014362, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 46537 + }, + { + "epoch": 0.46538, + "grad_norm": 1.3329540430953517, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 46538 + }, + { + "epoch": 0.46539, + "grad_norm": 1.4952093891279512, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46539 + }, + { + "epoch": 0.4654, + "grad_norm": 1.535663740992907, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 46540 + }, + { + "epoch": 0.46541, + "grad_norm": 1.7374221519349293, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 46541 + }, + { + "epoch": 0.46542, + "grad_norm": 1.2364144095794678, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 46542 + }, + { + "epoch": 0.46543, + "grad_norm": 1.820969774711078, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 46543 + }, + { + "epoch": 0.46544, + "grad_norm": 1.3088985445105181, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 46544 + }, + { + "epoch": 0.46545, + "grad_norm": 1.512798655756137, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 46545 + }, + { + "epoch": 0.46546, + "grad_norm": 1.6294200921578894, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 46546 + }, + { + "epoch": 0.46547, + "grad_norm": 1.717325853578517, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 46547 + }, + { + "epoch": 0.46548, + "grad_norm": 2.301857080088638, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 46548 + }, + { + "epoch": 0.46549, + "grad_norm": 1.088136183736347, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46549 + }, + { + "epoch": 0.4655, + "grad_norm": 1.5317473625323317, + "learning_rate": 0.003, + "loss": 4.029, + "step": 46550 + }, + { + "epoch": 0.46551, + "grad_norm": 1.687626579837917, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 46551 + }, + { + "epoch": 0.46552, + "grad_norm": 1.6811513298996046, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46552 + }, + { + "epoch": 0.46553, + "grad_norm": 1.238249120314396, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46553 + }, + { + "epoch": 0.46554, + "grad_norm": 1.8646216908468014, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46554 + }, + { + "epoch": 0.46555, + "grad_norm": 1.2352229179695817, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 46555 + }, + { + "epoch": 0.46556, + "grad_norm": 1.550300684146595, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46556 + }, + { + "epoch": 0.46557, + "grad_norm": 1.4309218147916674, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 46557 + }, + { + "epoch": 0.46558, + "grad_norm": 1.537995511788297, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 46558 + }, + { + "epoch": 0.46559, + "grad_norm": 1.388911219009678, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 46559 + }, + { + "epoch": 0.4656, + "grad_norm": 1.8076076140024508, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 46560 + }, + { + "epoch": 0.46561, + "grad_norm": 1.3658138317672268, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46561 + }, + { + "epoch": 0.46562, + "grad_norm": 1.5496246032700085, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 46562 + }, + { + "epoch": 0.46563, + "grad_norm": 1.443208376819789, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 46563 + }, + { + "epoch": 0.46564, + "grad_norm": 1.4860748711506948, + "learning_rate": 0.003, + "loss": 4.06, + "step": 46564 + }, + { + "epoch": 0.46565, + "grad_norm": 1.3018717842877687, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 46565 + }, + { + "epoch": 0.46566, + "grad_norm": 1.8968334440319248, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 46566 + }, + { + "epoch": 0.46567, + "grad_norm": 1.2073862692391446, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 46567 + }, + { + "epoch": 0.46568, + "grad_norm": 1.8127033983117484, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 46568 + }, + { + "epoch": 0.46569, + "grad_norm": 1.702484362838902, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 46569 + }, + { + "epoch": 0.4657, + "grad_norm": 1.866578439917, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 46570 + }, + { + "epoch": 0.46571, + "grad_norm": 1.5312125345290466, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 46571 + }, + { + "epoch": 0.46572, + "grad_norm": 1.5444515235059102, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 46572 + }, + { + "epoch": 0.46573, + "grad_norm": 1.781285401969841, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 46573 + }, + { + "epoch": 0.46574, + "grad_norm": 1.70316135665044, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 46574 + }, + { + "epoch": 0.46575, + "grad_norm": 1.2992845842478724, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 46575 + }, + { + "epoch": 0.46576, + "grad_norm": 1.8608922361545932, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46576 + }, + { + "epoch": 0.46577, + "grad_norm": 1.216853283727067, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46577 + }, + { + "epoch": 0.46578, + "grad_norm": 1.6652312984211026, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46578 + }, + { + "epoch": 0.46579, + "grad_norm": 1.5925569188936506, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 46579 + }, + { + "epoch": 0.4658, + "grad_norm": 1.6732371045596075, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 46580 + }, + { + "epoch": 0.46581, + "grad_norm": 1.3435857715669461, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46581 + }, + { + "epoch": 0.46582, + "grad_norm": 1.6240669558174767, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 46582 + }, + { + "epoch": 0.46583, + "grad_norm": 1.5802994800935093, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46583 + }, + { + "epoch": 0.46584, + "grad_norm": 1.685219669012027, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 46584 + }, + { + "epoch": 0.46585, + "grad_norm": 1.3423663986952241, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 46585 + }, + { + "epoch": 0.46586, + "grad_norm": 1.6428460622744339, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 46586 + }, + { + "epoch": 0.46587, + "grad_norm": 1.5690699674343374, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 46587 + }, + { + "epoch": 0.46588, + "grad_norm": 1.4713796717842504, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 46588 + }, + { + "epoch": 0.46589, + "grad_norm": 1.3014508598495809, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 46589 + }, + { + "epoch": 0.4659, + "grad_norm": 1.820064250550378, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46590 + }, + { + "epoch": 0.46591, + "grad_norm": 1.4112432730261468, + "learning_rate": 0.003, + "loss": 4.036, + "step": 46591 + }, + { + "epoch": 0.46592, + "grad_norm": 1.591570948318371, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 46592 + }, + { + "epoch": 0.46593, + "grad_norm": 1.5435862035119559, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46593 + }, + { + "epoch": 0.46594, + "grad_norm": 1.264271904239835, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 46594 + }, + { + "epoch": 0.46595, + "grad_norm": 1.667476884850219, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 46595 + }, + { + "epoch": 0.46596, + "grad_norm": 1.4572706153950956, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46596 + }, + { + "epoch": 0.46597, + "grad_norm": 1.6798394053277332, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46597 + }, + { + "epoch": 0.46598, + "grad_norm": 1.4989108233911963, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 46598 + }, + { + "epoch": 0.46599, + "grad_norm": 1.864849493124066, + "learning_rate": 0.003, + "loss": 4.011, + "step": 46599 + }, + { + "epoch": 0.466, + "grad_norm": 1.0721818308993796, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 46600 + }, + { + "epoch": 0.46601, + "grad_norm": 1.7892827698213567, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 46601 + }, + { + "epoch": 0.46602, + "grad_norm": 1.615849618413558, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 46602 + }, + { + "epoch": 0.46603, + "grad_norm": 1.6617210111068026, + "learning_rate": 0.003, + "loss": 4.034, + "step": 46603 + }, + { + "epoch": 0.46604, + "grad_norm": 1.6358805743711948, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 46604 + }, + { + "epoch": 0.46605, + "grad_norm": 1.6303162057100304, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46605 + }, + { + "epoch": 0.46606, + "grad_norm": 1.571784118007437, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 46606 + }, + { + "epoch": 0.46607, + "grad_norm": 1.5916341509978795, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 46607 + }, + { + "epoch": 0.46608, + "grad_norm": 1.4683794264935872, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 46608 + }, + { + "epoch": 0.46609, + "grad_norm": 1.4671508909710365, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 46609 + }, + { + "epoch": 0.4661, + "grad_norm": 1.5634065035145546, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 46610 + }, + { + "epoch": 0.46611, + "grad_norm": 1.566942577021743, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 46611 + }, + { + "epoch": 0.46612, + "grad_norm": 1.8863677648084618, + "learning_rate": 0.003, + "loss": 4.041, + "step": 46612 + }, + { + "epoch": 0.46613, + "grad_norm": 1.28273877381802, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 46613 + }, + { + "epoch": 0.46614, + "grad_norm": 1.9073755381205566, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 46614 + }, + { + "epoch": 0.46615, + "grad_norm": 1.4371594202433289, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 46615 + }, + { + "epoch": 0.46616, + "grad_norm": 1.4203690156563766, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 46616 + }, + { + "epoch": 0.46617, + "grad_norm": 1.8226696020217952, + "learning_rate": 0.003, + "loss": 4.042, + "step": 46617 + }, + { + "epoch": 0.46618, + "grad_norm": 1.2450020750113187, + "learning_rate": 0.003, + "loss": 4.039, + "step": 46618 + }, + { + "epoch": 0.46619, + "grad_norm": 1.643527812750747, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 46619 + }, + { + "epoch": 0.4662, + "grad_norm": 1.4423630442682829, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 46620 + }, + { + "epoch": 0.46621, + "grad_norm": 1.742744066147402, + "learning_rate": 0.003, + "loss": 4.059, + "step": 46621 + }, + { + "epoch": 0.46622, + "grad_norm": 1.3547143094001308, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46622 + }, + { + "epoch": 0.46623, + "grad_norm": 1.623606978352556, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 46623 + }, + { + "epoch": 0.46624, + "grad_norm": 1.4977628134957168, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 46624 + }, + { + "epoch": 0.46625, + "grad_norm": 1.3290905203358825, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 46625 + }, + { + "epoch": 0.46626, + "grad_norm": 1.700921862390392, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 46626 + }, + { + "epoch": 0.46627, + "grad_norm": 1.4185660641746602, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46627 + }, + { + "epoch": 0.46628, + "grad_norm": 1.6388384245603238, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46628 + }, + { + "epoch": 0.46629, + "grad_norm": 1.4183159603535878, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46629 + }, + { + "epoch": 0.4663, + "grad_norm": 1.4229873030908415, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 46630 + }, + { + "epoch": 0.46631, + "grad_norm": 1.7325087739275618, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46631 + }, + { + "epoch": 0.46632, + "grad_norm": 1.5726508334760743, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46632 + }, + { + "epoch": 0.46633, + "grad_norm": 1.7377470365173628, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 46633 + }, + { + "epoch": 0.46634, + "grad_norm": 2.030946486817744, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46634 + }, + { + "epoch": 0.46635, + "grad_norm": 1.4734378371300256, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 46635 + }, + { + "epoch": 0.46636, + "grad_norm": 1.4511916891474137, + "learning_rate": 0.003, + "loss": 4.049, + "step": 46636 + }, + { + "epoch": 0.46637, + "grad_norm": 1.4551939465202768, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46637 + }, + { + "epoch": 0.46638, + "grad_norm": 1.520863118612993, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 46638 + }, + { + "epoch": 0.46639, + "grad_norm": 1.4934578139467192, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 46639 + }, + { + "epoch": 0.4664, + "grad_norm": 1.696936520616188, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 46640 + }, + { + "epoch": 0.46641, + "grad_norm": 1.3795568946308343, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 46641 + }, + { + "epoch": 0.46642, + "grad_norm": 1.7042863395583836, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 46642 + }, + { + "epoch": 0.46643, + "grad_norm": 1.3963772882517138, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 46643 + }, + { + "epoch": 0.46644, + "grad_norm": 1.8452459933613272, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 46644 + }, + { + "epoch": 0.46645, + "grad_norm": 1.4381667932957602, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 46645 + }, + { + "epoch": 0.46646, + "grad_norm": 1.7619660846122827, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 46646 + }, + { + "epoch": 0.46647, + "grad_norm": 1.5936636745012007, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 46647 + }, + { + "epoch": 0.46648, + "grad_norm": 1.645848938411843, + "learning_rate": 0.003, + "loss": 4.056, + "step": 46648 + }, + { + "epoch": 0.46649, + "grad_norm": 1.4562036164092798, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46649 + }, + { + "epoch": 0.4665, + "grad_norm": 1.414521727545302, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 46650 + }, + { + "epoch": 0.46651, + "grad_norm": 1.8364625149901197, + "learning_rate": 0.003, + "loss": 4.026, + "step": 46651 + }, + { + "epoch": 0.46652, + "grad_norm": 1.276208638167298, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 46652 + }, + { + "epoch": 0.46653, + "grad_norm": 1.7779936076779932, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46653 + }, + { + "epoch": 0.46654, + "grad_norm": 1.4530263958849132, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 46654 + }, + { + "epoch": 0.46655, + "grad_norm": 1.6705590797941163, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 46655 + }, + { + "epoch": 0.46656, + "grad_norm": 1.4612098085950505, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46656 + }, + { + "epoch": 0.46657, + "grad_norm": 1.6913195309252895, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 46657 + }, + { + "epoch": 0.46658, + "grad_norm": 1.4937325720361652, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 46658 + }, + { + "epoch": 0.46659, + "grad_norm": 1.5695907477887319, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 46659 + }, + { + "epoch": 0.4666, + "grad_norm": 1.4221785245681293, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 46660 + }, + { + "epoch": 0.46661, + "grad_norm": 1.6225677155656915, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 46661 + }, + { + "epoch": 0.46662, + "grad_norm": 1.678172572904485, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46662 + }, + { + "epoch": 0.46663, + "grad_norm": 1.5775989512389843, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46663 + }, + { + "epoch": 0.46664, + "grad_norm": 1.6636676413202622, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 46664 + }, + { + "epoch": 0.46665, + "grad_norm": 1.4443226055130785, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 46665 + }, + { + "epoch": 0.46666, + "grad_norm": 1.6704678105809676, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46666 + }, + { + "epoch": 0.46667, + "grad_norm": 1.489313730387749, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46667 + }, + { + "epoch": 0.46668, + "grad_norm": 1.5073643466755744, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 46668 + }, + { + "epoch": 0.46669, + "grad_norm": 1.5797026166322947, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 46669 + }, + { + "epoch": 0.4667, + "grad_norm": 1.3727270318888503, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 46670 + }, + { + "epoch": 0.46671, + "grad_norm": 1.4530972179688693, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 46671 + }, + { + "epoch": 0.46672, + "grad_norm": 1.338916179908441, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46672 + }, + { + "epoch": 0.46673, + "grad_norm": 1.580788928594122, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 46673 + }, + { + "epoch": 0.46674, + "grad_norm": 1.579601545975026, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 46674 + }, + { + "epoch": 0.46675, + "grad_norm": 1.8129788043283457, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 46675 + }, + { + "epoch": 0.46676, + "grad_norm": 1.3901824583159597, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 46676 + }, + { + "epoch": 0.46677, + "grad_norm": 1.573327195050391, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 46677 + }, + { + "epoch": 0.46678, + "grad_norm": 1.4332196499871233, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46678 + }, + { + "epoch": 0.46679, + "grad_norm": 1.6811487061452224, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 46679 + }, + { + "epoch": 0.4668, + "grad_norm": 1.544264657729749, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 46680 + }, + { + "epoch": 0.46681, + "grad_norm": 1.921131938856084, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46681 + }, + { + "epoch": 0.46682, + "grad_norm": 1.4035060585104178, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 46682 + }, + { + "epoch": 0.46683, + "grad_norm": 1.560326986469608, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 46683 + }, + { + "epoch": 0.46684, + "grad_norm": 1.7911012770774377, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 46684 + }, + { + "epoch": 0.46685, + "grad_norm": 1.463778218130816, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46685 + }, + { + "epoch": 0.46686, + "grad_norm": 1.4795491337692366, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 46686 + }, + { + "epoch": 0.46687, + "grad_norm": 1.619628382176246, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 46687 + }, + { + "epoch": 0.46688, + "grad_norm": 1.4569045611849496, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46688 + }, + { + "epoch": 0.46689, + "grad_norm": 1.5762944801679204, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 46689 + }, + { + "epoch": 0.4669, + "grad_norm": 1.3962128937404625, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 46690 + }, + { + "epoch": 0.46691, + "grad_norm": 1.8722372106540894, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 46691 + }, + { + "epoch": 0.46692, + "grad_norm": 0.9741592545347689, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 46692 + }, + { + "epoch": 0.46693, + "grad_norm": 1.7534512558324142, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46693 + }, + { + "epoch": 0.46694, + "grad_norm": 1.429276877749549, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46694 + }, + { + "epoch": 0.46695, + "grad_norm": 1.69142487108773, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 46695 + }, + { + "epoch": 0.46696, + "grad_norm": 1.5946027299623307, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 46696 + }, + { + "epoch": 0.46697, + "grad_norm": 1.4797215078473867, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 46697 + }, + { + "epoch": 0.46698, + "grad_norm": 1.7764778463314082, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 46698 + }, + { + "epoch": 0.46699, + "grad_norm": 1.687248126072142, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 46699 + }, + { + "epoch": 0.467, + "grad_norm": 1.54854205482997, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 46700 + }, + { + "epoch": 0.46701, + "grad_norm": 1.4687691656699502, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 46701 + }, + { + "epoch": 0.46702, + "grad_norm": 2.141624565084847, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 46702 + }, + { + "epoch": 0.46703, + "grad_norm": 1.2757486635986657, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46703 + }, + { + "epoch": 0.46704, + "grad_norm": 1.4414509651238836, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 46704 + }, + { + "epoch": 0.46705, + "grad_norm": 2.004701124028883, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 46705 + }, + { + "epoch": 0.46706, + "grad_norm": 1.4380606323280791, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 46706 + }, + { + "epoch": 0.46707, + "grad_norm": 1.466357083090928, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 46707 + }, + { + "epoch": 0.46708, + "grad_norm": 1.6412261676662767, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 46708 + }, + { + "epoch": 0.46709, + "grad_norm": 1.440900534797587, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 46709 + }, + { + "epoch": 0.4671, + "grad_norm": 1.5691146148571429, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 46710 + }, + { + "epoch": 0.46711, + "grad_norm": 1.482146073810264, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 46711 + }, + { + "epoch": 0.46712, + "grad_norm": 1.5759842171056266, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46712 + }, + { + "epoch": 0.46713, + "grad_norm": 1.4852812589070448, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 46713 + }, + { + "epoch": 0.46714, + "grad_norm": 1.5561951988909164, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 46714 + }, + { + "epoch": 0.46715, + "grad_norm": 1.775359701790875, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46715 + }, + { + "epoch": 0.46716, + "grad_norm": 1.4532469081047275, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 46716 + }, + { + "epoch": 0.46717, + "grad_norm": 1.8707290883093575, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 46717 + }, + { + "epoch": 0.46718, + "grad_norm": 1.3129951726066, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 46718 + }, + { + "epoch": 0.46719, + "grad_norm": 1.6059856981798293, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 46719 + }, + { + "epoch": 0.4672, + "grad_norm": 1.438378942034145, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 46720 + }, + { + "epoch": 0.46721, + "grad_norm": 1.5925004096791027, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 46721 + }, + { + "epoch": 0.46722, + "grad_norm": 1.3577800398118212, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 46722 + }, + { + "epoch": 0.46723, + "grad_norm": 1.6238279028670692, + "learning_rate": 0.003, + "loss": 4.038, + "step": 46723 + }, + { + "epoch": 0.46724, + "grad_norm": 1.417960242709985, + "learning_rate": 0.003, + "loss": 4.037, + "step": 46724 + }, + { + "epoch": 0.46725, + "grad_norm": 1.4668114116088866, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 46725 + }, + { + "epoch": 0.46726, + "grad_norm": 1.5984245504384158, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46726 + }, + { + "epoch": 0.46727, + "grad_norm": 1.637437686431324, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46727 + }, + { + "epoch": 0.46728, + "grad_norm": 2.1675740842984736, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 46728 + }, + { + "epoch": 0.46729, + "grad_norm": 1.0843853904562792, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46729 + }, + { + "epoch": 0.4673, + "grad_norm": 1.8627746191006624, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 46730 + }, + { + "epoch": 0.46731, + "grad_norm": 1.8990611099260308, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 46731 + }, + { + "epoch": 0.46732, + "grad_norm": 1.7576289246141878, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 46732 + }, + { + "epoch": 0.46733, + "grad_norm": 1.4655350233307916, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 46733 + }, + { + "epoch": 0.46734, + "grad_norm": 1.3929674975613076, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 46734 + }, + { + "epoch": 0.46735, + "grad_norm": 1.6662361378932984, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 46735 + }, + { + "epoch": 0.46736, + "grad_norm": 1.2739652495794844, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46736 + }, + { + "epoch": 0.46737, + "grad_norm": 2.1100853799383614, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 46737 + }, + { + "epoch": 0.46738, + "grad_norm": 1.3476020857998599, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 46738 + }, + { + "epoch": 0.46739, + "grad_norm": 1.7933305328971916, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 46739 + }, + { + "epoch": 0.4674, + "grad_norm": 1.797270496222689, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 46740 + }, + { + "epoch": 0.46741, + "grad_norm": 1.2065462575852415, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46741 + }, + { + "epoch": 0.46742, + "grad_norm": 1.5702741819234936, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 46742 + }, + { + "epoch": 0.46743, + "grad_norm": 1.4081083617379102, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 46743 + }, + { + "epoch": 0.46744, + "grad_norm": 1.5525628463402035, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 46744 + }, + { + "epoch": 0.46745, + "grad_norm": 1.9286006078780988, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 46745 + }, + { + "epoch": 0.46746, + "grad_norm": 1.2670585312004912, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 46746 + }, + { + "epoch": 0.46747, + "grad_norm": 1.7184177449178473, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 46747 + }, + { + "epoch": 0.46748, + "grad_norm": 1.3544910133995243, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46748 + }, + { + "epoch": 0.46749, + "grad_norm": 1.5368920452212067, + "learning_rate": 0.003, + "loss": 4.068, + "step": 46749 + }, + { + "epoch": 0.4675, + "grad_norm": 1.5377592554094976, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 46750 + }, + { + "epoch": 0.46751, + "grad_norm": 1.5672431608942516, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 46751 + }, + { + "epoch": 0.46752, + "grad_norm": 1.4749029972323149, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46752 + }, + { + "epoch": 0.46753, + "grad_norm": 1.4850581879668203, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46753 + }, + { + "epoch": 0.46754, + "grad_norm": 1.718063782051419, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 46754 + }, + { + "epoch": 0.46755, + "grad_norm": 1.3034958981938363, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 46755 + }, + { + "epoch": 0.46756, + "grad_norm": 1.5282340743441152, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 46756 + }, + { + "epoch": 0.46757, + "grad_norm": 1.6287996867820886, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46757 + }, + { + "epoch": 0.46758, + "grad_norm": 1.4515082155236927, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 46758 + }, + { + "epoch": 0.46759, + "grad_norm": 1.5731532918492286, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 46759 + }, + { + "epoch": 0.4676, + "grad_norm": 1.596854089414919, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 46760 + }, + { + "epoch": 0.46761, + "grad_norm": 1.343411500667571, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 46761 + }, + { + "epoch": 0.46762, + "grad_norm": 1.691256094139627, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 46762 + }, + { + "epoch": 0.46763, + "grad_norm": 1.4185661472002091, + "learning_rate": 0.003, + "loss": 4.023, + "step": 46763 + }, + { + "epoch": 0.46764, + "grad_norm": 1.7289958209624943, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 46764 + }, + { + "epoch": 0.46765, + "grad_norm": 1.2566314056968708, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 46765 + }, + { + "epoch": 0.46766, + "grad_norm": 2.223949314035341, + "learning_rate": 0.003, + "loss": 4.031, + "step": 46766 + }, + { + "epoch": 0.46767, + "grad_norm": 1.4074053606914159, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46767 + }, + { + "epoch": 0.46768, + "grad_norm": 1.6412596054942952, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 46768 + }, + { + "epoch": 0.46769, + "grad_norm": 1.3692322494666853, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 46769 + }, + { + "epoch": 0.4677, + "grad_norm": 1.9204620196116144, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 46770 + }, + { + "epoch": 0.46771, + "grad_norm": 1.3060557665095043, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 46771 + }, + { + "epoch": 0.46772, + "grad_norm": 1.3959943480964756, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 46772 + }, + { + "epoch": 0.46773, + "grad_norm": 1.5714744618283674, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 46773 + }, + { + "epoch": 0.46774, + "grad_norm": 1.6408069342375138, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 46774 + }, + { + "epoch": 0.46775, + "grad_norm": 1.6406924746970122, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 46775 + }, + { + "epoch": 0.46776, + "grad_norm": 1.7748821622394315, + "learning_rate": 0.003, + "loss": 4.044, + "step": 46776 + }, + { + "epoch": 0.46777, + "grad_norm": 1.780478553682775, + "learning_rate": 0.003, + "loss": 4.03, + "step": 46777 + }, + { + "epoch": 0.46778, + "grad_norm": 1.302881762886237, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 46778 + }, + { + "epoch": 0.46779, + "grad_norm": 1.4732360227943193, + "learning_rate": 0.003, + "loss": 4.041, + "step": 46779 + }, + { + "epoch": 0.4678, + "grad_norm": 1.6097247751218167, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46780 + }, + { + "epoch": 0.46781, + "grad_norm": 1.4114875580570347, + "learning_rate": 0.003, + "loss": 4.0, + "step": 46781 + }, + { + "epoch": 0.46782, + "grad_norm": 1.4922387121700051, + "learning_rate": 0.003, + "loss": 4.022, + "step": 46782 + }, + { + "epoch": 0.46783, + "grad_norm": 1.4049756718992459, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46783 + }, + { + "epoch": 0.46784, + "grad_norm": 1.6468700686668813, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46784 + }, + { + "epoch": 0.46785, + "grad_norm": 1.524224382657056, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46785 + }, + { + "epoch": 0.46786, + "grad_norm": 1.6485988137314314, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 46786 + }, + { + "epoch": 0.46787, + "grad_norm": 1.5137985629488069, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 46787 + }, + { + "epoch": 0.46788, + "grad_norm": 1.4709820816130712, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 46788 + }, + { + "epoch": 0.46789, + "grad_norm": 1.347713916041374, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 46789 + }, + { + "epoch": 0.4679, + "grad_norm": 1.7209276574894046, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46790 + }, + { + "epoch": 0.46791, + "grad_norm": 1.315435666169726, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 46791 + }, + { + "epoch": 0.46792, + "grad_norm": 1.901431411891877, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 46792 + }, + { + "epoch": 0.46793, + "grad_norm": 1.4613701145647926, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 46793 + }, + { + "epoch": 0.46794, + "grad_norm": 1.5750108316339735, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 46794 + }, + { + "epoch": 0.46795, + "grad_norm": 1.3811408763319502, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 46795 + }, + { + "epoch": 0.46796, + "grad_norm": 1.9265222153284167, + "learning_rate": 0.003, + "loss": 4.055, + "step": 46796 + }, + { + "epoch": 0.46797, + "grad_norm": 1.5026016883192044, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46797 + }, + { + "epoch": 0.46798, + "grad_norm": 1.596004745496696, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46798 + }, + { + "epoch": 0.46799, + "grad_norm": 1.5519896775990596, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 46799 + }, + { + "epoch": 0.468, + "grad_norm": 1.534224815488405, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 46800 + }, + { + "epoch": 0.46801, + "grad_norm": 1.5538754868953804, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 46801 + }, + { + "epoch": 0.46802, + "grad_norm": 1.389698269240988, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 46802 + }, + { + "epoch": 0.46803, + "grad_norm": 1.5672857559197226, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 46803 + }, + { + "epoch": 0.46804, + "grad_norm": 1.7089420090576677, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 46804 + }, + { + "epoch": 0.46805, + "grad_norm": 1.7131614270061086, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 46805 + }, + { + "epoch": 0.46806, + "grad_norm": 1.5759002575267205, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46806 + }, + { + "epoch": 0.46807, + "grad_norm": 1.5182219346927164, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 46807 + }, + { + "epoch": 0.46808, + "grad_norm": 1.3155146305233318, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 46808 + }, + { + "epoch": 0.46809, + "grad_norm": 1.6683558059980952, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 46809 + }, + { + "epoch": 0.4681, + "grad_norm": 1.297463577393893, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 46810 + }, + { + "epoch": 0.46811, + "grad_norm": 1.7267502725657604, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 46811 + }, + { + "epoch": 0.46812, + "grad_norm": 1.4807900361743649, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 46812 + }, + { + "epoch": 0.46813, + "grad_norm": 1.728760474666455, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 46813 + }, + { + "epoch": 0.46814, + "grad_norm": 1.2911544129311183, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 46814 + }, + { + "epoch": 0.46815, + "grad_norm": 1.8547830590572314, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 46815 + }, + { + "epoch": 0.46816, + "grad_norm": 1.3044354440166483, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 46816 + }, + { + "epoch": 0.46817, + "grad_norm": 1.617617773791657, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 46817 + }, + { + "epoch": 0.46818, + "grad_norm": 1.6850388874544315, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46818 + }, + { + "epoch": 0.46819, + "grad_norm": 1.527383218283905, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 46819 + }, + { + "epoch": 0.4682, + "grad_norm": 1.4944617656144108, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 46820 + }, + { + "epoch": 0.46821, + "grad_norm": 1.5857051302668006, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 46821 + }, + { + "epoch": 0.46822, + "grad_norm": 1.3908180449043948, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 46822 + }, + { + "epoch": 0.46823, + "grad_norm": 1.5626089417671158, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46823 + }, + { + "epoch": 0.46824, + "grad_norm": 1.5006163871275866, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 46824 + }, + { + "epoch": 0.46825, + "grad_norm": 1.562461008876681, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 46825 + }, + { + "epoch": 0.46826, + "grad_norm": 1.7397511414490652, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 46826 + }, + { + "epoch": 0.46827, + "grad_norm": 1.5612108471322665, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 46827 + }, + { + "epoch": 0.46828, + "grad_norm": 1.4127625999716296, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 46828 + }, + { + "epoch": 0.46829, + "grad_norm": 1.7885783789324223, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 46829 + }, + { + "epoch": 0.4683, + "grad_norm": 1.1936339152937092, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 46830 + }, + { + "epoch": 0.46831, + "grad_norm": 1.9222297922416183, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 46831 + }, + { + "epoch": 0.46832, + "grad_norm": 1.7460224165727043, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 46832 + }, + { + "epoch": 0.46833, + "grad_norm": 1.3702204134581188, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 46833 + }, + { + "epoch": 0.46834, + "grad_norm": 1.4178762607753188, + "learning_rate": 0.003, + "loss": 4.032, + "step": 46834 + }, + { + "epoch": 0.46835, + "grad_norm": 1.6392396743961275, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46835 + }, + { + "epoch": 0.46836, + "grad_norm": 1.5201400573974246, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 46836 + }, + { + "epoch": 0.46837, + "grad_norm": 1.5169828833808057, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 46837 + }, + { + "epoch": 0.46838, + "grad_norm": 1.7856274721541274, + "learning_rate": 0.003, + "loss": 4.03, + "step": 46838 + }, + { + "epoch": 0.46839, + "grad_norm": 1.7120791846221848, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 46839 + }, + { + "epoch": 0.4684, + "grad_norm": 1.3668440769272954, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 46840 + }, + { + "epoch": 0.46841, + "grad_norm": 1.7955313493416776, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 46841 + }, + { + "epoch": 0.46842, + "grad_norm": 1.2831706439957975, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 46842 + }, + { + "epoch": 0.46843, + "grad_norm": 1.8258966413528657, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 46843 + }, + { + "epoch": 0.46844, + "grad_norm": 1.6113412864887955, + "learning_rate": 0.003, + "loss": 4.061, + "step": 46844 + }, + { + "epoch": 0.46845, + "grad_norm": 1.6025288583818713, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46845 + }, + { + "epoch": 0.46846, + "grad_norm": 1.4412264990125347, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 46846 + }, + { + "epoch": 0.46847, + "grad_norm": 1.7747295564385563, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 46847 + }, + { + "epoch": 0.46848, + "grad_norm": 1.0583446844093343, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46848 + }, + { + "epoch": 0.46849, + "grad_norm": 1.6829538798760577, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 46849 + }, + { + "epoch": 0.4685, + "grad_norm": 1.3768676422721113, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 46850 + }, + { + "epoch": 0.46851, + "grad_norm": 2.07678416165702, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 46851 + }, + { + "epoch": 0.46852, + "grad_norm": 1.2899696300597836, + "learning_rate": 0.003, + "loss": 4.052, + "step": 46852 + }, + { + "epoch": 0.46853, + "grad_norm": 2.0323293641932767, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 46853 + }, + { + "epoch": 0.46854, + "grad_norm": 1.6281449448556475, + "learning_rate": 0.003, + "loss": 4.048, + "step": 46854 + }, + { + "epoch": 0.46855, + "grad_norm": 1.3974826940453942, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 46855 + }, + { + "epoch": 0.46856, + "grad_norm": 1.502282578158052, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 46856 + }, + { + "epoch": 0.46857, + "grad_norm": 1.5835879423376602, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 46857 + }, + { + "epoch": 0.46858, + "grad_norm": 1.6512832389829943, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 46858 + }, + { + "epoch": 0.46859, + "grad_norm": 1.5525200010577638, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 46859 + }, + { + "epoch": 0.4686, + "grad_norm": 1.5589473764379351, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 46860 + }, + { + "epoch": 0.46861, + "grad_norm": 1.4888872678403935, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46861 + }, + { + "epoch": 0.46862, + "grad_norm": 1.5218746276045807, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 46862 + }, + { + "epoch": 0.46863, + "grad_norm": 1.3291469771062816, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 46863 + }, + { + "epoch": 0.46864, + "grad_norm": 1.534777133156119, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46864 + }, + { + "epoch": 0.46865, + "grad_norm": 1.399080794237576, + "learning_rate": 0.003, + "loss": 4.024, + "step": 46865 + }, + { + "epoch": 0.46866, + "grad_norm": 1.492490714092847, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 46866 + }, + { + "epoch": 0.46867, + "grad_norm": 1.9524679886390508, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 46867 + }, + { + "epoch": 0.46868, + "grad_norm": 1.3027502608167212, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 46868 + }, + { + "epoch": 0.46869, + "grad_norm": 1.6201798750271263, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 46869 + }, + { + "epoch": 0.4687, + "grad_norm": 1.7218375595534812, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 46870 + }, + { + "epoch": 0.46871, + "grad_norm": 1.432245851707845, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 46871 + }, + { + "epoch": 0.46872, + "grad_norm": 1.5649409203288127, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46872 + }, + { + "epoch": 0.46873, + "grad_norm": 1.5770150440398611, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 46873 + }, + { + "epoch": 0.46874, + "grad_norm": 1.6065286822293876, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46874 + }, + { + "epoch": 0.46875, + "grad_norm": 1.4269946733743257, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 46875 + }, + { + "epoch": 0.46876, + "grad_norm": 1.7781723032123176, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46876 + }, + { + "epoch": 0.46877, + "grad_norm": 1.5332827056174825, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46877 + }, + { + "epoch": 0.46878, + "grad_norm": 1.6178744595352224, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 46878 + }, + { + "epoch": 0.46879, + "grad_norm": 1.3475233562285418, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 46879 + }, + { + "epoch": 0.4688, + "grad_norm": 1.4209090009383805, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 46880 + }, + { + "epoch": 0.46881, + "grad_norm": 1.584548542164196, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 46881 + }, + { + "epoch": 0.46882, + "grad_norm": 1.5562784796062812, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 46882 + }, + { + "epoch": 0.46883, + "grad_norm": 1.465989343301044, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 46883 + }, + { + "epoch": 0.46884, + "grad_norm": 1.755693308352029, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46884 + }, + { + "epoch": 0.46885, + "grad_norm": 1.568891728392585, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 46885 + }, + { + "epoch": 0.46886, + "grad_norm": 1.4813968794290702, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 46886 + }, + { + "epoch": 0.46887, + "grad_norm": 1.5784706249421507, + "learning_rate": 0.003, + "loss": 4.018, + "step": 46887 + }, + { + "epoch": 0.46888, + "grad_norm": 1.5010232979659095, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 46888 + }, + { + "epoch": 0.46889, + "grad_norm": 1.4288552378285853, + "learning_rate": 0.003, + "loss": 4.071, + "step": 46889 + }, + { + "epoch": 0.4689, + "grad_norm": 1.5888541966797975, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 46890 + }, + { + "epoch": 0.46891, + "grad_norm": 1.302268631524604, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 46891 + }, + { + "epoch": 0.46892, + "grad_norm": 2.0374417218795706, + "learning_rate": 0.003, + "loss": 4.064, + "step": 46892 + }, + { + "epoch": 0.46893, + "grad_norm": 1.9022172616670405, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 46893 + }, + { + "epoch": 0.46894, + "grad_norm": 1.9537243658806476, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 46894 + }, + { + "epoch": 0.46895, + "grad_norm": 1.3163482228516739, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 46895 + }, + { + "epoch": 0.46896, + "grad_norm": 1.611199127216489, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46896 + }, + { + "epoch": 0.46897, + "grad_norm": 1.7654631342838598, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 46897 + }, + { + "epoch": 0.46898, + "grad_norm": 1.316964386604733, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46898 + }, + { + "epoch": 0.46899, + "grad_norm": 1.7423099160149798, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46899 + }, + { + "epoch": 0.469, + "grad_norm": 1.6795908448113388, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 46900 + }, + { + "epoch": 0.46901, + "grad_norm": 1.4397109648640272, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 46901 + }, + { + "epoch": 0.46902, + "grad_norm": 1.5767846212100676, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 46902 + }, + { + "epoch": 0.46903, + "grad_norm": 1.6541235449552223, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46903 + }, + { + "epoch": 0.46904, + "grad_norm": 1.5656946674231313, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 46904 + }, + { + "epoch": 0.46905, + "grad_norm": 1.5682129974498702, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 46905 + }, + { + "epoch": 0.46906, + "grad_norm": 1.6852058159187167, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 46906 + }, + { + "epoch": 0.46907, + "grad_norm": 1.3121488075372711, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 46907 + }, + { + "epoch": 0.46908, + "grad_norm": 1.4545948946555227, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 46908 + }, + { + "epoch": 0.46909, + "grad_norm": 1.3634331121096321, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 46909 + }, + { + "epoch": 0.4691, + "grad_norm": 1.598890276093339, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 46910 + }, + { + "epoch": 0.46911, + "grad_norm": 1.4764914846269737, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 46911 + }, + { + "epoch": 0.46912, + "grad_norm": 1.6981895537576892, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 46912 + }, + { + "epoch": 0.46913, + "grad_norm": 1.5001748468202623, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 46913 + }, + { + "epoch": 0.46914, + "grad_norm": 1.5109381082927809, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 46914 + }, + { + "epoch": 0.46915, + "grad_norm": 1.8221291069199894, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 46915 + }, + { + "epoch": 0.46916, + "grad_norm": 1.3392430048239112, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 46916 + }, + { + "epoch": 0.46917, + "grad_norm": 1.5797908945184058, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 46917 + }, + { + "epoch": 0.46918, + "grad_norm": 1.731213206652154, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46918 + }, + { + "epoch": 0.46919, + "grad_norm": 1.358893079714294, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 46919 + }, + { + "epoch": 0.4692, + "grad_norm": 1.5633234255869664, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 46920 + }, + { + "epoch": 0.46921, + "grad_norm": 1.6871072246351786, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 46921 + }, + { + "epoch": 0.46922, + "grad_norm": 1.5573017819327686, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 46922 + }, + { + "epoch": 0.46923, + "grad_norm": 1.6109294544982224, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 46923 + }, + { + "epoch": 0.46924, + "grad_norm": 1.5596787757806176, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 46924 + }, + { + "epoch": 0.46925, + "grad_norm": 1.5164996927808632, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46925 + }, + { + "epoch": 0.46926, + "grad_norm": 1.5169947068526544, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 46926 + }, + { + "epoch": 0.46927, + "grad_norm": 1.3479097354275356, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 46927 + }, + { + "epoch": 0.46928, + "grad_norm": 1.61381522390965, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 46928 + }, + { + "epoch": 0.46929, + "grad_norm": 1.4265877576363906, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 46929 + }, + { + "epoch": 0.4693, + "grad_norm": 1.5135928325665025, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 46930 + }, + { + "epoch": 0.46931, + "grad_norm": 1.3980748406033896, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46931 + }, + { + "epoch": 0.46932, + "grad_norm": 1.478341944598953, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 46932 + }, + { + "epoch": 0.46933, + "grad_norm": 1.4371655123802807, + "learning_rate": 0.003, + "loss": 4.047, + "step": 46933 + }, + { + "epoch": 0.46934, + "grad_norm": 1.5246676630407319, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 46934 + }, + { + "epoch": 0.46935, + "grad_norm": 1.509645715371353, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46935 + }, + { + "epoch": 0.46936, + "grad_norm": 1.5472698199968975, + "learning_rate": 0.003, + "loss": 4.04, + "step": 46936 + }, + { + "epoch": 0.46937, + "grad_norm": 1.7890711039050677, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 46937 + }, + { + "epoch": 0.46938, + "grad_norm": 2.030896432056988, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 46938 + }, + { + "epoch": 0.46939, + "grad_norm": 1.4684031511314282, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46939 + }, + { + "epoch": 0.4694, + "grad_norm": 1.5355943496257722, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 46940 + }, + { + "epoch": 0.46941, + "grad_norm": 1.8954515516480601, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 46941 + }, + { + "epoch": 0.46942, + "grad_norm": 1.3575881665659506, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 46942 + }, + { + "epoch": 0.46943, + "grad_norm": 1.7657875757109476, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 46943 + }, + { + "epoch": 0.46944, + "grad_norm": 1.3133563892785638, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 46944 + }, + { + "epoch": 0.46945, + "grad_norm": 1.8163810776984104, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 46945 + }, + { + "epoch": 0.46946, + "grad_norm": 1.3536085869155872, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 46946 + }, + { + "epoch": 0.46947, + "grad_norm": 1.7274991248049698, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 46947 + }, + { + "epoch": 0.46948, + "grad_norm": 1.317440972520842, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 46948 + }, + { + "epoch": 0.46949, + "grad_norm": 1.631695498128023, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 46949 + }, + { + "epoch": 0.4695, + "grad_norm": 1.5682195319284378, + "learning_rate": 0.003, + "loss": 4.042, + "step": 46950 + }, + { + "epoch": 0.46951, + "grad_norm": 1.6468474303059546, + "learning_rate": 0.003, + "loss": 4.039, + "step": 46951 + }, + { + "epoch": 0.46952, + "grad_norm": 1.475833824106731, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 46952 + }, + { + "epoch": 0.46953, + "grad_norm": 1.362813568659657, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 46953 + }, + { + "epoch": 0.46954, + "grad_norm": 1.5600402633453911, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 46954 + }, + { + "epoch": 0.46955, + "grad_norm": 1.3450340048611642, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 46955 + }, + { + "epoch": 0.46956, + "grad_norm": 1.8245110814628167, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 46956 + }, + { + "epoch": 0.46957, + "grad_norm": 1.4449629940773208, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 46957 + }, + { + "epoch": 0.46958, + "grad_norm": 1.5811533783734657, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 46958 + }, + { + "epoch": 0.46959, + "grad_norm": 1.3694244220461318, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 46959 + }, + { + "epoch": 0.4696, + "grad_norm": 1.876501403428261, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 46960 + }, + { + "epoch": 0.46961, + "grad_norm": 1.2895713513071894, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 46961 + }, + { + "epoch": 0.46962, + "grad_norm": 1.6555756498823941, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 46962 + }, + { + "epoch": 0.46963, + "grad_norm": 1.37706937982276, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 46963 + }, + { + "epoch": 0.46964, + "grad_norm": 1.630357383580264, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 46964 + }, + { + "epoch": 0.46965, + "grad_norm": 1.347614709137042, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 46965 + }, + { + "epoch": 0.46966, + "grad_norm": 1.9771629513011892, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 46966 + }, + { + "epoch": 0.46967, + "grad_norm": 1.4358397209949048, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 46967 + }, + { + "epoch": 0.46968, + "grad_norm": 1.5993598217923235, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 46968 + }, + { + "epoch": 0.46969, + "grad_norm": 1.4370493045568669, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 46969 + }, + { + "epoch": 0.4697, + "grad_norm": 1.7601163181826798, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 46970 + }, + { + "epoch": 0.46971, + "grad_norm": 1.5428756875560454, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 46971 + }, + { + "epoch": 0.46972, + "grad_norm": 1.443321159496594, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 46972 + }, + { + "epoch": 0.46973, + "grad_norm": 1.4992678507144155, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46973 + }, + { + "epoch": 0.46974, + "grad_norm": 1.5416457661609757, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 46974 + }, + { + "epoch": 0.46975, + "grad_norm": 1.4381728129373275, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 46975 + }, + { + "epoch": 0.46976, + "grad_norm": 1.652499444776981, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 46976 + }, + { + "epoch": 0.46977, + "grad_norm": 1.5074633188892559, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 46977 + }, + { + "epoch": 0.46978, + "grad_norm": 1.4658663791372926, + "learning_rate": 0.003, + "loss": 4.046, + "step": 46978 + }, + { + "epoch": 0.46979, + "grad_norm": 1.3766848904711553, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 46979 + }, + { + "epoch": 0.4698, + "grad_norm": 1.424432199421683, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 46980 + }, + { + "epoch": 0.46981, + "grad_norm": 1.592951300592699, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 46981 + }, + { + "epoch": 0.46982, + "grad_norm": 1.80442141244117, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 46982 + }, + { + "epoch": 0.46983, + "grad_norm": 1.6875290986229023, + "learning_rate": 0.003, + "loss": 4.035, + "step": 46983 + }, + { + "epoch": 0.46984, + "grad_norm": 1.2837916500355708, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 46984 + }, + { + "epoch": 0.46985, + "grad_norm": 1.8719383719679346, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 46985 + }, + { + "epoch": 0.46986, + "grad_norm": 1.48659565507934, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 46986 + }, + { + "epoch": 0.46987, + "grad_norm": 1.5010488566927565, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 46987 + }, + { + "epoch": 0.46988, + "grad_norm": 1.4598924863638187, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 46988 + }, + { + "epoch": 0.46989, + "grad_norm": 1.8951580654771474, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 46989 + }, + { + "epoch": 0.4699, + "grad_norm": 1.3075330719008778, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 46990 + }, + { + "epoch": 0.46991, + "grad_norm": 1.7103065905961834, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 46991 + }, + { + "epoch": 0.46992, + "grad_norm": 1.802428843966264, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 46992 + }, + { + "epoch": 0.46993, + "grad_norm": 1.6635490000010489, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 46993 + }, + { + "epoch": 0.46994, + "grad_norm": 2.030029004601237, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 46994 + }, + { + "epoch": 0.46995, + "grad_norm": 1.2476903523901979, + "learning_rate": 0.003, + "loss": 4.021, + "step": 46995 + }, + { + "epoch": 0.46996, + "grad_norm": 1.6589272379684483, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 46996 + }, + { + "epoch": 0.46997, + "grad_norm": 1.5988184223784951, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 46997 + }, + { + "epoch": 0.46998, + "grad_norm": 1.3000637219965, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 46998 + }, + { + "epoch": 0.46999, + "grad_norm": 1.5802041117163705, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 46999 + }, + { + "epoch": 0.47, + "grad_norm": 1.644879027851485, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47000 + }, + { + "epoch": 0.47001, + "grad_norm": 1.215636115670267, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 47001 + }, + { + "epoch": 0.47002, + "grad_norm": 1.5044577418258713, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 47002 + }, + { + "epoch": 0.47003, + "grad_norm": 1.6522738156705057, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47003 + }, + { + "epoch": 0.47004, + "grad_norm": 1.5445741360228955, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 47004 + }, + { + "epoch": 0.47005, + "grad_norm": 1.8007986539462966, + "learning_rate": 0.003, + "loss": 4.012, + "step": 47005 + }, + { + "epoch": 0.47006, + "grad_norm": 1.3610929206527718, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 47006 + }, + { + "epoch": 0.47007, + "grad_norm": 1.6295057986194823, + "learning_rate": 0.003, + "loss": 4.034, + "step": 47007 + }, + { + "epoch": 0.47008, + "grad_norm": 1.475263190369588, + "learning_rate": 0.003, + "loss": 4.012, + "step": 47008 + }, + { + "epoch": 0.47009, + "grad_norm": 1.4280200979970763, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 47009 + }, + { + "epoch": 0.4701, + "grad_norm": 1.8823180767492558, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 47010 + }, + { + "epoch": 0.47011, + "grad_norm": 1.4040211855088354, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 47011 + }, + { + "epoch": 0.47012, + "grad_norm": 1.6997533169055186, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 47012 + }, + { + "epoch": 0.47013, + "grad_norm": 1.4235382946102124, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47013 + }, + { + "epoch": 0.47014, + "grad_norm": 1.9249653532463846, + "learning_rate": 0.003, + "loss": 4.073, + "step": 47014 + }, + { + "epoch": 0.47015, + "grad_norm": 1.3278918479915145, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 47015 + }, + { + "epoch": 0.47016, + "grad_norm": 1.4777479210887519, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 47016 + }, + { + "epoch": 0.47017, + "grad_norm": 1.528799562229412, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 47017 + }, + { + "epoch": 0.47018, + "grad_norm": 1.8497475357401891, + "learning_rate": 0.003, + "loss": 4.067, + "step": 47018 + }, + { + "epoch": 0.47019, + "grad_norm": 1.5464065409677945, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47019 + }, + { + "epoch": 0.4702, + "grad_norm": 1.5197915123057342, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 47020 + }, + { + "epoch": 0.47021, + "grad_norm": 1.5508096616971951, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47021 + }, + { + "epoch": 0.47022, + "grad_norm": 1.4322553234890807, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47022 + }, + { + "epoch": 0.47023, + "grad_norm": 1.6278424682529034, + "learning_rate": 0.003, + "loss": 4.05, + "step": 47023 + }, + { + "epoch": 0.47024, + "grad_norm": 1.4230271934240182, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 47024 + }, + { + "epoch": 0.47025, + "grad_norm": 1.76737251749765, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47025 + }, + { + "epoch": 0.47026, + "grad_norm": 1.3379013953946226, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 47026 + }, + { + "epoch": 0.47027, + "grad_norm": 1.866534102221444, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 47027 + }, + { + "epoch": 0.47028, + "grad_norm": 1.4284022400220102, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47028 + }, + { + "epoch": 0.47029, + "grad_norm": 1.6267414864886987, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 47029 + }, + { + "epoch": 0.4703, + "grad_norm": 1.3565348271631559, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47030 + }, + { + "epoch": 0.47031, + "grad_norm": 1.5002460522140717, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 47031 + }, + { + "epoch": 0.47032, + "grad_norm": 2.003380983480116, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 47032 + }, + { + "epoch": 0.47033, + "grad_norm": 1.5304263242651293, + "learning_rate": 0.003, + "loss": 4.07, + "step": 47033 + }, + { + "epoch": 0.47034, + "grad_norm": 2.161027422400591, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 47034 + }, + { + "epoch": 0.47035, + "grad_norm": 1.6652836115147962, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 47035 + }, + { + "epoch": 0.47036, + "grad_norm": 1.5000267264500695, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47036 + }, + { + "epoch": 0.47037, + "grad_norm": 1.7361469136508565, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 47037 + }, + { + "epoch": 0.47038, + "grad_norm": 1.5444099679088519, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 47038 + }, + { + "epoch": 0.47039, + "grad_norm": 1.3561010243171785, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 47039 + }, + { + "epoch": 0.4704, + "grad_norm": 1.491027699437816, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 47040 + }, + { + "epoch": 0.47041, + "grad_norm": 1.6248061935143083, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 47041 + }, + { + "epoch": 0.47042, + "grad_norm": 1.6471459763807053, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 47042 + }, + { + "epoch": 0.47043, + "grad_norm": 1.3353595283211859, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 47043 + }, + { + "epoch": 0.47044, + "grad_norm": 1.6675779681243017, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 47044 + }, + { + "epoch": 0.47045, + "grad_norm": 1.6735687289292045, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 47045 + }, + { + "epoch": 0.47046, + "grad_norm": 1.441801837924226, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 47046 + }, + { + "epoch": 0.47047, + "grad_norm": 1.626797221895776, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 47047 + }, + { + "epoch": 0.47048, + "grad_norm": 1.3654271892117837, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 47048 + }, + { + "epoch": 0.47049, + "grad_norm": 1.719842743171978, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 47049 + }, + { + "epoch": 0.4705, + "grad_norm": 1.398731727620464, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 47050 + }, + { + "epoch": 0.47051, + "grad_norm": 1.6556358351478988, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47051 + }, + { + "epoch": 0.47052, + "grad_norm": 1.7049658218967638, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47052 + }, + { + "epoch": 0.47053, + "grad_norm": 1.3543243769105895, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 47053 + }, + { + "epoch": 0.47054, + "grad_norm": 1.5497703586650804, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47054 + }, + { + "epoch": 0.47055, + "grad_norm": 1.2872142268803495, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 47055 + }, + { + "epoch": 0.47056, + "grad_norm": 2.1301965860431737, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 47056 + }, + { + "epoch": 0.47057, + "grad_norm": 1.3627231861321671, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 47057 + }, + { + "epoch": 0.47058, + "grad_norm": 1.595902357904538, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 47058 + }, + { + "epoch": 0.47059, + "grad_norm": 1.807390444236991, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47059 + }, + { + "epoch": 0.4706, + "grad_norm": 1.6240532332164466, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47060 + }, + { + "epoch": 0.47061, + "grad_norm": 1.349588569231089, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 47061 + }, + { + "epoch": 0.47062, + "grad_norm": 1.7551999901584938, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 47062 + }, + { + "epoch": 0.47063, + "grad_norm": 1.336311311626797, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47063 + }, + { + "epoch": 0.47064, + "grad_norm": 1.625602693951272, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 47064 + }, + { + "epoch": 0.47065, + "grad_norm": 1.526871321981314, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 47065 + }, + { + "epoch": 0.47066, + "grad_norm": 1.5467480080368232, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 47066 + }, + { + "epoch": 0.47067, + "grad_norm": 1.599048617083512, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 47067 + }, + { + "epoch": 0.47068, + "grad_norm": 1.583081410752437, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 47068 + }, + { + "epoch": 0.47069, + "grad_norm": 1.4529059382430345, + "learning_rate": 0.003, + "loss": 4.049, + "step": 47069 + }, + { + "epoch": 0.4707, + "grad_norm": 1.6134619780454935, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 47070 + }, + { + "epoch": 0.47071, + "grad_norm": 1.4675709436829487, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 47071 + }, + { + "epoch": 0.47072, + "grad_norm": 1.3151912846686207, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47072 + }, + { + "epoch": 0.47073, + "grad_norm": 1.6484747631427883, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47073 + }, + { + "epoch": 0.47074, + "grad_norm": 1.4673047276486697, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 47074 + }, + { + "epoch": 0.47075, + "grad_norm": 1.9315129104486821, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47075 + }, + { + "epoch": 0.47076, + "grad_norm": 1.268240339807947, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47076 + }, + { + "epoch": 0.47077, + "grad_norm": 2.1060917051231702, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47077 + }, + { + "epoch": 0.47078, + "grad_norm": 1.2251795757909116, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 47078 + }, + { + "epoch": 0.47079, + "grad_norm": 1.77766079213556, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 47079 + }, + { + "epoch": 0.4708, + "grad_norm": 1.4605305253789282, + "learning_rate": 0.003, + "loss": 4.046, + "step": 47080 + }, + { + "epoch": 0.47081, + "grad_norm": 1.596117893981391, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 47081 + }, + { + "epoch": 0.47082, + "grad_norm": 1.6249911088424653, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 47082 + }, + { + "epoch": 0.47083, + "grad_norm": 1.3964114797030196, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47083 + }, + { + "epoch": 0.47084, + "grad_norm": 1.5426969332747198, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 47084 + }, + { + "epoch": 0.47085, + "grad_norm": 1.6852545381471986, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47085 + }, + { + "epoch": 0.47086, + "grad_norm": 1.6348984527295471, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 47086 + }, + { + "epoch": 0.47087, + "grad_norm": 1.5975405926348751, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 47087 + }, + { + "epoch": 0.47088, + "grad_norm": 1.4303833468312814, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 47088 + }, + { + "epoch": 0.47089, + "grad_norm": 1.6296964961658176, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 47089 + }, + { + "epoch": 0.4709, + "grad_norm": 1.3459552120473808, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 47090 + }, + { + "epoch": 0.47091, + "grad_norm": 1.6484292997389942, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47091 + }, + { + "epoch": 0.47092, + "grad_norm": 1.5742744668701796, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 47092 + }, + { + "epoch": 0.47093, + "grad_norm": 1.7603571162582512, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 47093 + }, + { + "epoch": 0.47094, + "grad_norm": 1.3662281406703696, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47094 + }, + { + "epoch": 0.47095, + "grad_norm": 1.6549824851190094, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 47095 + }, + { + "epoch": 0.47096, + "grad_norm": 1.4100543760166415, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 47096 + }, + { + "epoch": 0.47097, + "grad_norm": 1.8503569982421413, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 47097 + }, + { + "epoch": 0.47098, + "grad_norm": 1.666126976941358, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47098 + }, + { + "epoch": 0.47099, + "grad_norm": 1.4830679293471396, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 47099 + }, + { + "epoch": 0.471, + "grad_norm": 1.754615502756301, + "learning_rate": 0.003, + "loss": 4.042, + "step": 47100 + }, + { + "epoch": 0.47101, + "grad_norm": 1.2893631336230473, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 47101 + }, + { + "epoch": 0.47102, + "grad_norm": 1.9772516730736893, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 47102 + }, + { + "epoch": 0.47103, + "grad_norm": 1.4020039410104645, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47103 + }, + { + "epoch": 0.47104, + "grad_norm": 1.475277076558276, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 47104 + }, + { + "epoch": 0.47105, + "grad_norm": 1.4671847333997288, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 47105 + }, + { + "epoch": 0.47106, + "grad_norm": 1.505687304522194, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 47106 + }, + { + "epoch": 0.47107, + "grad_norm": 1.5528135127869402, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 47107 + }, + { + "epoch": 0.47108, + "grad_norm": 1.5115155708172534, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 47108 + }, + { + "epoch": 0.47109, + "grad_norm": 1.4138202504430832, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47109 + }, + { + "epoch": 0.4711, + "grad_norm": 1.6727061182919598, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 47110 + }, + { + "epoch": 0.47111, + "grad_norm": 1.7415129823970426, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 47111 + }, + { + "epoch": 0.47112, + "grad_norm": 1.3255802690284537, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 47112 + }, + { + "epoch": 0.47113, + "grad_norm": 1.8983068281149436, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47113 + }, + { + "epoch": 0.47114, + "grad_norm": 1.3025600208002928, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47114 + }, + { + "epoch": 0.47115, + "grad_norm": 1.6921537542493148, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 47115 + }, + { + "epoch": 0.47116, + "grad_norm": 1.6004291837815243, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 47116 + }, + { + "epoch": 0.47117, + "grad_norm": 1.6464045338336228, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 47117 + }, + { + "epoch": 0.47118, + "grad_norm": 1.3990827699734028, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 47118 + }, + { + "epoch": 0.47119, + "grad_norm": 1.6492747017332983, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 47119 + }, + { + "epoch": 0.4712, + "grad_norm": 1.6153793476858134, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 47120 + }, + { + "epoch": 0.47121, + "grad_norm": 1.2629183548278944, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 47121 + }, + { + "epoch": 0.47122, + "grad_norm": 1.5698605206370408, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 47122 + }, + { + "epoch": 0.47123, + "grad_norm": 1.3161967764173865, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 47123 + }, + { + "epoch": 0.47124, + "grad_norm": 1.7450044276893302, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 47124 + }, + { + "epoch": 0.47125, + "grad_norm": 1.5846110226782792, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 47125 + }, + { + "epoch": 0.47126, + "grad_norm": 1.7701704449096725, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 47126 + }, + { + "epoch": 0.47127, + "grad_norm": 1.5929105096716765, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 47127 + }, + { + "epoch": 0.47128, + "grad_norm": 1.6526675178486023, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 47128 + }, + { + "epoch": 0.47129, + "grad_norm": 1.5995623696812713, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 47129 + }, + { + "epoch": 0.4713, + "grad_norm": 1.6449722502979292, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 47130 + }, + { + "epoch": 0.47131, + "grad_norm": 1.7748129430140773, + "learning_rate": 0.003, + "loss": 4.042, + "step": 47131 + }, + { + "epoch": 0.47132, + "grad_norm": 1.3095823598521517, + "learning_rate": 0.003, + "loss": 4.032, + "step": 47132 + }, + { + "epoch": 0.47133, + "grad_norm": 1.7668532480257146, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 47133 + }, + { + "epoch": 0.47134, + "grad_norm": 1.5115385120243987, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47134 + }, + { + "epoch": 0.47135, + "grad_norm": 1.4927351707794776, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47135 + }, + { + "epoch": 0.47136, + "grad_norm": 1.7436014827540423, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 47136 + }, + { + "epoch": 0.47137, + "grad_norm": 1.5930631125727814, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 47137 + }, + { + "epoch": 0.47138, + "grad_norm": 1.5413494986289868, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 47138 + }, + { + "epoch": 0.47139, + "grad_norm": 1.626847358280709, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 47139 + }, + { + "epoch": 0.4714, + "grad_norm": 1.613259255125468, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 47140 + }, + { + "epoch": 0.47141, + "grad_norm": 1.1899644577472568, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 47141 + }, + { + "epoch": 0.47142, + "grad_norm": 1.8313333755557295, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47142 + }, + { + "epoch": 0.47143, + "grad_norm": 1.3021435301113635, + "learning_rate": 0.003, + "loss": 4.052, + "step": 47143 + }, + { + "epoch": 0.47144, + "grad_norm": 1.5338772021736327, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47144 + }, + { + "epoch": 0.47145, + "grad_norm": 1.4329203319321013, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 47145 + }, + { + "epoch": 0.47146, + "grad_norm": 1.6690359547874547, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 47146 + }, + { + "epoch": 0.47147, + "grad_norm": 1.5026336441175872, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 47147 + }, + { + "epoch": 0.47148, + "grad_norm": 1.682319786688538, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 47148 + }, + { + "epoch": 0.47149, + "grad_norm": 1.6115179231577108, + "learning_rate": 0.003, + "loss": 4.024, + "step": 47149 + }, + { + "epoch": 0.4715, + "grad_norm": 1.2916459343695195, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 47150 + }, + { + "epoch": 0.47151, + "grad_norm": 1.5915465821077677, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 47151 + }, + { + "epoch": 0.47152, + "grad_norm": 1.3695456433639879, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 47152 + }, + { + "epoch": 0.47153, + "grad_norm": 1.7603240873628203, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 47153 + }, + { + "epoch": 0.47154, + "grad_norm": 1.76329800505215, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47154 + }, + { + "epoch": 0.47155, + "grad_norm": 1.4963341810856816, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 47155 + }, + { + "epoch": 0.47156, + "grad_norm": 1.659386129794793, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 47156 + }, + { + "epoch": 0.47157, + "grad_norm": 1.4849175410365234, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 47157 + }, + { + "epoch": 0.47158, + "grad_norm": 1.750354362065256, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 47158 + }, + { + "epoch": 0.47159, + "grad_norm": 1.3563980813042926, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 47159 + }, + { + "epoch": 0.4716, + "grad_norm": 1.6754402444698349, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 47160 + }, + { + "epoch": 0.47161, + "grad_norm": 1.449974117608757, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 47161 + }, + { + "epoch": 0.47162, + "grad_norm": 1.22370748900266, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 47162 + }, + { + "epoch": 0.47163, + "grad_norm": 1.6538901977487397, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 47163 + }, + { + "epoch": 0.47164, + "grad_norm": 1.375694863545289, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 47164 + }, + { + "epoch": 0.47165, + "grad_norm": 1.822839428970755, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 47165 + }, + { + "epoch": 0.47166, + "grad_norm": 1.4720543177991725, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 47166 + }, + { + "epoch": 0.47167, + "grad_norm": 1.5592636377712645, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 47167 + }, + { + "epoch": 0.47168, + "grad_norm": 1.6436116198783688, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 47168 + }, + { + "epoch": 0.47169, + "grad_norm": 1.569547574674721, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 47169 + }, + { + "epoch": 0.4717, + "grad_norm": 1.7921166770061703, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 47170 + }, + { + "epoch": 0.47171, + "grad_norm": 1.3746285714473034, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 47171 + }, + { + "epoch": 0.47172, + "grad_norm": 1.4569269555206485, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47172 + }, + { + "epoch": 0.47173, + "grad_norm": 1.6755689604274544, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 47173 + }, + { + "epoch": 0.47174, + "grad_norm": 1.8415375540238261, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47174 + }, + { + "epoch": 0.47175, + "grad_norm": 1.539370078110406, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 47175 + }, + { + "epoch": 0.47176, + "grad_norm": 1.5888736417185692, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47176 + }, + { + "epoch": 0.47177, + "grad_norm": 1.5196979724915962, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 47177 + }, + { + "epoch": 0.47178, + "grad_norm": 1.4231073238635845, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47178 + }, + { + "epoch": 0.47179, + "grad_norm": 1.5743906211771237, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 47179 + }, + { + "epoch": 0.4718, + "grad_norm": 1.4870913962469257, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 47180 + }, + { + "epoch": 0.47181, + "grad_norm": 1.4365538162934344, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 47181 + }, + { + "epoch": 0.47182, + "grad_norm": 1.6746643688671514, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47182 + }, + { + "epoch": 0.47183, + "grad_norm": 1.483291940130947, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47183 + }, + { + "epoch": 0.47184, + "grad_norm": 1.8031455063673003, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 47184 + }, + { + "epoch": 0.47185, + "grad_norm": 1.5928036338857277, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 47185 + }, + { + "epoch": 0.47186, + "grad_norm": 1.6013763448892773, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 47186 + }, + { + "epoch": 0.47187, + "grad_norm": 1.3812155535582027, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 47187 + }, + { + "epoch": 0.47188, + "grad_norm": 1.7653878050964567, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 47188 + }, + { + "epoch": 0.47189, + "grad_norm": 1.3318425254556054, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 47189 + }, + { + "epoch": 0.4719, + "grad_norm": 1.818995054941289, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 47190 + }, + { + "epoch": 0.47191, + "grad_norm": 1.432687120840342, + "learning_rate": 0.003, + "loss": 4.039, + "step": 47191 + }, + { + "epoch": 0.47192, + "grad_norm": 1.6608814520223543, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 47192 + }, + { + "epoch": 0.47193, + "grad_norm": 1.5936182150730396, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47193 + }, + { + "epoch": 0.47194, + "grad_norm": 1.6277301680831842, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 47194 + }, + { + "epoch": 0.47195, + "grad_norm": 1.735302212103019, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 47195 + }, + { + "epoch": 0.47196, + "grad_norm": 1.28673914144365, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 47196 + }, + { + "epoch": 0.47197, + "grad_norm": 1.8720577541520622, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 47197 + }, + { + "epoch": 0.47198, + "grad_norm": 1.635380875932246, + "learning_rate": 0.003, + "loss": 4.041, + "step": 47198 + }, + { + "epoch": 0.47199, + "grad_norm": 1.8795032091999602, + "learning_rate": 0.003, + "loss": 4.048, + "step": 47199 + }, + { + "epoch": 0.472, + "grad_norm": 1.6072679625743875, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 47200 + }, + { + "epoch": 0.47201, + "grad_norm": 1.403064978159799, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 47201 + }, + { + "epoch": 0.47202, + "grad_norm": 1.8498948163838584, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47202 + }, + { + "epoch": 0.47203, + "grad_norm": 1.5155921854766927, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47203 + }, + { + "epoch": 0.47204, + "grad_norm": 1.4158003815748101, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 47204 + }, + { + "epoch": 0.47205, + "grad_norm": 1.6012517154871162, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47205 + }, + { + "epoch": 0.47206, + "grad_norm": 1.4851975576158032, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 47206 + }, + { + "epoch": 0.47207, + "grad_norm": 1.434256557236428, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47207 + }, + { + "epoch": 0.47208, + "grad_norm": 1.7142968257873203, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47208 + }, + { + "epoch": 0.47209, + "grad_norm": 1.5172767690357023, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47209 + }, + { + "epoch": 0.4721, + "grad_norm": 1.735997248437428, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 47210 + }, + { + "epoch": 0.47211, + "grad_norm": 1.4237653434414774, + "learning_rate": 0.003, + "loss": 4.0, + "step": 47211 + }, + { + "epoch": 0.47212, + "grad_norm": 1.5389608804909578, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47212 + }, + { + "epoch": 0.47213, + "grad_norm": 1.3127059599972635, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 47213 + }, + { + "epoch": 0.47214, + "grad_norm": 1.5346802967267565, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47214 + }, + { + "epoch": 0.47215, + "grad_norm": 1.740718604980609, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47215 + }, + { + "epoch": 0.47216, + "grad_norm": 1.5461406505919835, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 47216 + }, + { + "epoch": 0.47217, + "grad_norm": 1.542023238637298, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 47217 + }, + { + "epoch": 0.47218, + "grad_norm": 1.5653282667835333, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 47218 + }, + { + "epoch": 0.47219, + "grad_norm": 1.6152998723265888, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 47219 + }, + { + "epoch": 0.4722, + "grad_norm": 1.3202572325560296, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47220 + }, + { + "epoch": 0.47221, + "grad_norm": 1.5111143736512098, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47221 + }, + { + "epoch": 0.47222, + "grad_norm": 1.4487467575240438, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 47222 + }, + { + "epoch": 0.47223, + "grad_norm": 1.4843480119613082, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 47223 + }, + { + "epoch": 0.47224, + "grad_norm": 1.419744732183863, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 47224 + }, + { + "epoch": 0.47225, + "grad_norm": 1.5318908572263241, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47225 + }, + { + "epoch": 0.47226, + "grad_norm": 1.5423116813009468, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 47226 + }, + { + "epoch": 0.47227, + "grad_norm": 1.6745846613291322, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 47227 + }, + { + "epoch": 0.47228, + "grad_norm": 1.7111734380198464, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 47228 + }, + { + "epoch": 0.47229, + "grad_norm": 2.072552913098275, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 47229 + }, + { + "epoch": 0.4723, + "grad_norm": 1.3133403887504373, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 47230 + }, + { + "epoch": 0.47231, + "grad_norm": 1.7954475035532662, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 47231 + }, + { + "epoch": 0.47232, + "grad_norm": 1.6426746374437338, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 47232 + }, + { + "epoch": 0.47233, + "grad_norm": 1.508541953996758, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47233 + }, + { + "epoch": 0.47234, + "grad_norm": 1.5839745784382644, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 47234 + }, + { + "epoch": 0.47235, + "grad_norm": 1.6248996489824452, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47235 + }, + { + "epoch": 0.47236, + "grad_norm": 1.23675408544752, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 47236 + }, + { + "epoch": 0.47237, + "grad_norm": 1.5776856798592698, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 47237 + }, + { + "epoch": 0.47238, + "grad_norm": 1.435146839156874, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 47238 + }, + { + "epoch": 0.47239, + "grad_norm": 1.7430797932722302, + "learning_rate": 0.003, + "loss": 4.061, + "step": 47239 + }, + { + "epoch": 0.4724, + "grad_norm": 1.9110135496885088, + "learning_rate": 0.003, + "loss": 4.011, + "step": 47240 + }, + { + "epoch": 0.47241, + "grad_norm": 1.6749596376173288, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 47241 + }, + { + "epoch": 0.47242, + "grad_norm": 1.593335723242075, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 47242 + }, + { + "epoch": 0.47243, + "grad_norm": 1.368845495199755, + "learning_rate": 0.003, + "loss": 4.027, + "step": 47243 + }, + { + "epoch": 0.47244, + "grad_norm": 1.5097684257473545, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47244 + }, + { + "epoch": 0.47245, + "grad_norm": 1.5519450890024158, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 47245 + }, + { + "epoch": 0.47246, + "grad_norm": 1.375297976532605, + "learning_rate": 0.003, + "loss": 4.026, + "step": 47246 + }, + { + "epoch": 0.47247, + "grad_norm": 1.6165273825639719, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 47247 + }, + { + "epoch": 0.47248, + "grad_norm": 1.3810754401998597, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 47248 + }, + { + "epoch": 0.47249, + "grad_norm": 1.6160737957101217, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 47249 + }, + { + "epoch": 0.4725, + "grad_norm": 1.6551147071118548, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47250 + }, + { + "epoch": 0.47251, + "grad_norm": 1.868133703018866, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47251 + }, + { + "epoch": 0.47252, + "grad_norm": 1.5060796743947897, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47252 + }, + { + "epoch": 0.47253, + "grad_norm": 1.3872216038731735, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47253 + }, + { + "epoch": 0.47254, + "grad_norm": 1.73187309739761, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 47254 + }, + { + "epoch": 0.47255, + "grad_norm": 1.460235622192705, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 47255 + }, + { + "epoch": 0.47256, + "grad_norm": 1.5129791248264182, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 47256 + }, + { + "epoch": 0.47257, + "grad_norm": 1.7119178931277668, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47257 + }, + { + "epoch": 0.47258, + "grad_norm": 1.485606599421258, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 47258 + }, + { + "epoch": 0.47259, + "grad_norm": 1.5149332992555018, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 47259 + }, + { + "epoch": 0.4726, + "grad_norm": 1.5990083757526554, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47260 + }, + { + "epoch": 0.47261, + "grad_norm": 1.5508229178189519, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 47261 + }, + { + "epoch": 0.47262, + "grad_norm": 1.266775646670913, + "learning_rate": 0.003, + "loss": 4.051, + "step": 47262 + }, + { + "epoch": 0.47263, + "grad_norm": 1.5801199471391714, + "learning_rate": 0.003, + "loss": 4.028, + "step": 47263 + }, + { + "epoch": 0.47264, + "grad_norm": 1.6489897502838524, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 47264 + }, + { + "epoch": 0.47265, + "grad_norm": 1.5364007249961786, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 47265 + }, + { + "epoch": 0.47266, + "grad_norm": 2.0695169699061684, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 47266 + }, + { + "epoch": 0.47267, + "grad_norm": 1.5073413630180321, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 47267 + }, + { + "epoch": 0.47268, + "grad_norm": 1.6864914545303735, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 47268 + }, + { + "epoch": 0.47269, + "grad_norm": 1.4326628504872105, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 47269 + }, + { + "epoch": 0.4727, + "grad_norm": 1.5273499965470125, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 47270 + }, + { + "epoch": 0.47271, + "grad_norm": 1.4919317236339829, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 47271 + }, + { + "epoch": 0.47272, + "grad_norm": 1.2744017371222651, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 47272 + }, + { + "epoch": 0.47273, + "grad_norm": 1.7387461266397934, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47273 + }, + { + "epoch": 0.47274, + "grad_norm": 1.392692343809255, + "learning_rate": 0.003, + "loss": 4.023, + "step": 47274 + }, + { + "epoch": 0.47275, + "grad_norm": 1.7521375853826195, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 47275 + }, + { + "epoch": 0.47276, + "grad_norm": 1.4789081157409645, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 47276 + }, + { + "epoch": 0.47277, + "grad_norm": 1.8687454478215257, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 47277 + }, + { + "epoch": 0.47278, + "grad_norm": 1.2940910803423102, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 47278 + }, + { + "epoch": 0.47279, + "grad_norm": 1.7799367092236906, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 47279 + }, + { + "epoch": 0.4728, + "grad_norm": 1.3825349982884778, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 47280 + }, + { + "epoch": 0.47281, + "grad_norm": 1.7251840476391729, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 47281 + }, + { + "epoch": 0.47282, + "grad_norm": 1.4248972000412439, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 47282 + }, + { + "epoch": 0.47283, + "grad_norm": 1.877560610552191, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47283 + }, + { + "epoch": 0.47284, + "grad_norm": 1.5103870755548716, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 47284 + }, + { + "epoch": 0.47285, + "grad_norm": 1.510471468290066, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 47285 + }, + { + "epoch": 0.47286, + "grad_norm": 1.5695869724557323, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 47286 + }, + { + "epoch": 0.47287, + "grad_norm": 1.3980095316581622, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 47287 + }, + { + "epoch": 0.47288, + "grad_norm": 2.0580430199304076, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47288 + }, + { + "epoch": 0.47289, + "grad_norm": 1.0601973923421655, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 47289 + }, + { + "epoch": 0.4729, + "grad_norm": 1.833331886260321, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 47290 + }, + { + "epoch": 0.47291, + "grad_norm": 1.5262040050600867, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 47291 + }, + { + "epoch": 0.47292, + "grad_norm": 1.5296978638129899, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 47292 + }, + { + "epoch": 0.47293, + "grad_norm": 1.637993364756696, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 47293 + }, + { + "epoch": 0.47294, + "grad_norm": 1.6453703882909372, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 47294 + }, + { + "epoch": 0.47295, + "grad_norm": 1.3771686420620997, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 47295 + }, + { + "epoch": 0.47296, + "grad_norm": 1.4709126702663617, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 47296 + }, + { + "epoch": 0.47297, + "grad_norm": 1.5054986106737385, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47297 + }, + { + "epoch": 0.47298, + "grad_norm": 1.5753094242242025, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 47298 + }, + { + "epoch": 0.47299, + "grad_norm": 1.6435654411742273, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 47299 + }, + { + "epoch": 0.473, + "grad_norm": 1.270577323650144, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 47300 + }, + { + "epoch": 0.47301, + "grad_norm": 1.62842741693039, + "learning_rate": 0.003, + "loss": 4.054, + "step": 47301 + }, + { + "epoch": 0.47302, + "grad_norm": 1.2595669356060086, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 47302 + }, + { + "epoch": 0.47303, + "grad_norm": 1.7971751816480788, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 47303 + }, + { + "epoch": 0.47304, + "grad_norm": 1.650129964169119, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 47304 + }, + { + "epoch": 0.47305, + "grad_norm": 1.6185723347815721, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 47305 + }, + { + "epoch": 0.47306, + "grad_norm": 1.644426795237853, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 47306 + }, + { + "epoch": 0.47307, + "grad_norm": 1.3527772591967233, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 47307 + }, + { + "epoch": 0.47308, + "grad_norm": 1.6975324387577428, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 47308 + }, + { + "epoch": 0.47309, + "grad_norm": 1.6265976866518999, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 47309 + }, + { + "epoch": 0.4731, + "grad_norm": 1.2385526129113429, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47310 + }, + { + "epoch": 0.47311, + "grad_norm": 1.773261402148357, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 47311 + }, + { + "epoch": 0.47312, + "grad_norm": 1.4070386191447792, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 47312 + }, + { + "epoch": 0.47313, + "grad_norm": 1.7018207876274827, + "learning_rate": 0.003, + "loss": 4.044, + "step": 47313 + }, + { + "epoch": 0.47314, + "grad_norm": 1.2183708132648596, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47314 + }, + { + "epoch": 0.47315, + "grad_norm": 1.835234055312929, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 47315 + }, + { + "epoch": 0.47316, + "grad_norm": 1.1514857066337156, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 47316 + }, + { + "epoch": 0.47317, + "grad_norm": 1.7709804474716544, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47317 + }, + { + "epoch": 0.47318, + "grad_norm": 1.225012986709841, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 47318 + }, + { + "epoch": 0.47319, + "grad_norm": 1.6743057717454655, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 47319 + }, + { + "epoch": 0.4732, + "grad_norm": 1.4912547709764756, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 47320 + }, + { + "epoch": 0.47321, + "grad_norm": 1.6839538483257206, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 47321 + }, + { + "epoch": 0.47322, + "grad_norm": 1.4485539479619898, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 47322 + }, + { + "epoch": 0.47323, + "grad_norm": 1.868404986988782, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 47323 + }, + { + "epoch": 0.47324, + "grad_norm": 1.3782225855955688, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 47324 + }, + { + "epoch": 0.47325, + "grad_norm": 1.9891940510776736, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 47325 + }, + { + "epoch": 0.47326, + "grad_norm": 1.9003027307605767, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 47326 + }, + { + "epoch": 0.47327, + "grad_norm": 1.5521724485368815, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 47327 + }, + { + "epoch": 0.47328, + "grad_norm": 1.6674622409389879, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 47328 + }, + { + "epoch": 0.47329, + "grad_norm": 1.6449584136146782, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 47329 + }, + { + "epoch": 0.4733, + "grad_norm": 1.2464183278332972, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 47330 + }, + { + "epoch": 0.47331, + "grad_norm": 1.344349367824683, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47331 + }, + { + "epoch": 0.47332, + "grad_norm": 1.318521986576844, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47332 + }, + { + "epoch": 0.47333, + "grad_norm": 1.6239372767487286, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 47333 + }, + { + "epoch": 0.47334, + "grad_norm": 1.4761734347007123, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47334 + }, + { + "epoch": 0.47335, + "grad_norm": 1.9907771665459635, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47335 + }, + { + "epoch": 0.47336, + "grad_norm": 1.241355186762704, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 47336 + }, + { + "epoch": 0.47337, + "grad_norm": 1.629716393622126, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 47337 + }, + { + "epoch": 0.47338, + "grad_norm": 2.0031921595175386, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 47338 + }, + { + "epoch": 0.47339, + "grad_norm": 1.4421515567690295, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 47339 + }, + { + "epoch": 0.4734, + "grad_norm": 1.5993594156306778, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 47340 + }, + { + "epoch": 0.47341, + "grad_norm": 1.2252790342933357, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 47341 + }, + { + "epoch": 0.47342, + "grad_norm": 1.6366119166055266, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 47342 + }, + { + "epoch": 0.47343, + "grad_norm": 1.3825089743571493, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 47343 + }, + { + "epoch": 0.47344, + "grad_norm": 1.7460435613007954, + "learning_rate": 0.003, + "loss": 4.039, + "step": 47344 + }, + { + "epoch": 0.47345, + "grad_norm": 1.3085103567298448, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 47345 + }, + { + "epoch": 0.47346, + "grad_norm": 1.7411912025035592, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47346 + }, + { + "epoch": 0.47347, + "grad_norm": 1.6069739143744532, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 47347 + }, + { + "epoch": 0.47348, + "grad_norm": 1.7486040119576673, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 47348 + }, + { + "epoch": 0.47349, + "grad_norm": 1.4458604062428164, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47349 + }, + { + "epoch": 0.4735, + "grad_norm": 1.4388687074274307, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47350 + }, + { + "epoch": 0.47351, + "grad_norm": 1.6824351684161492, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 47351 + }, + { + "epoch": 0.47352, + "grad_norm": 1.2878207089852067, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47352 + }, + { + "epoch": 0.47353, + "grad_norm": 1.7272459091412693, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 47353 + }, + { + "epoch": 0.47354, + "grad_norm": 1.641039051249822, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 47354 + }, + { + "epoch": 0.47355, + "grad_norm": 1.4889969507251837, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 47355 + }, + { + "epoch": 0.47356, + "grad_norm": 1.5841591070535166, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 47356 + }, + { + "epoch": 0.47357, + "grad_norm": 1.4725497070382498, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 47357 + }, + { + "epoch": 0.47358, + "grad_norm": 1.6058902252676055, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 47358 + }, + { + "epoch": 0.47359, + "grad_norm": 1.524394670752091, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 47359 + }, + { + "epoch": 0.4736, + "grad_norm": 1.3951668524562635, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 47360 + }, + { + "epoch": 0.47361, + "grad_norm": 1.66911283673974, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 47361 + }, + { + "epoch": 0.47362, + "grad_norm": 1.6818972229190254, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 47362 + }, + { + "epoch": 0.47363, + "grad_norm": 1.817121336300465, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 47363 + }, + { + "epoch": 0.47364, + "grad_norm": 1.4265631701629002, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47364 + }, + { + "epoch": 0.47365, + "grad_norm": 1.5643515024461063, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 47365 + }, + { + "epoch": 0.47366, + "grad_norm": 1.569157536376408, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47366 + }, + { + "epoch": 0.47367, + "grad_norm": 1.543192928596898, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 47367 + }, + { + "epoch": 0.47368, + "grad_norm": 1.5324745048408537, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 47368 + }, + { + "epoch": 0.47369, + "grad_norm": 1.3504100522651323, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47369 + }, + { + "epoch": 0.4737, + "grad_norm": 1.6836492741925666, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 47370 + }, + { + "epoch": 0.47371, + "grad_norm": 1.1770015818293142, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47371 + }, + { + "epoch": 0.47372, + "grad_norm": 1.9926957164830494, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 47372 + }, + { + "epoch": 0.47373, + "grad_norm": 1.5441309028785393, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 47373 + }, + { + "epoch": 0.47374, + "grad_norm": 2.0309788806974, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 47374 + }, + { + "epoch": 0.47375, + "grad_norm": 1.5424919516189013, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 47375 + }, + { + "epoch": 0.47376, + "grad_norm": 1.3019057383451489, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 47376 + }, + { + "epoch": 0.47377, + "grad_norm": 1.9665669854930727, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 47377 + }, + { + "epoch": 0.47378, + "grad_norm": 1.12630242528944, + "learning_rate": 0.003, + "loss": 4.05, + "step": 47378 + }, + { + "epoch": 0.47379, + "grad_norm": 1.6268044820147909, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 47379 + }, + { + "epoch": 0.4738, + "grad_norm": 1.4964527869217603, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47380 + }, + { + "epoch": 0.47381, + "grad_norm": 2.060033984042775, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 47381 + }, + { + "epoch": 0.47382, + "grad_norm": 1.1037759319054132, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 47382 + }, + { + "epoch": 0.47383, + "grad_norm": 1.9315123957694438, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47383 + }, + { + "epoch": 0.47384, + "grad_norm": 1.3290861250236883, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 47384 + }, + { + "epoch": 0.47385, + "grad_norm": 1.6478671846525363, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 47385 + }, + { + "epoch": 0.47386, + "grad_norm": 1.382562138955217, + "learning_rate": 0.003, + "loss": 4.047, + "step": 47386 + }, + { + "epoch": 0.47387, + "grad_norm": 1.9449436862730278, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 47387 + }, + { + "epoch": 0.47388, + "grad_norm": 1.3189348013544757, + "learning_rate": 0.003, + "loss": 4.074, + "step": 47388 + }, + { + "epoch": 0.47389, + "grad_norm": 1.6895306018889296, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 47389 + }, + { + "epoch": 0.4739, + "grad_norm": 1.6434433786168543, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 47390 + }, + { + "epoch": 0.47391, + "grad_norm": 1.5702248976144793, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 47391 + }, + { + "epoch": 0.47392, + "grad_norm": 1.5633343959505261, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 47392 + }, + { + "epoch": 0.47393, + "grad_norm": 1.474610105775382, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 47393 + }, + { + "epoch": 0.47394, + "grad_norm": 1.603734650710486, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 47394 + }, + { + "epoch": 0.47395, + "grad_norm": 1.436631747825815, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 47395 + }, + { + "epoch": 0.47396, + "grad_norm": 1.7417772381052712, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 47396 + }, + { + "epoch": 0.47397, + "grad_norm": 1.2675152569499306, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47397 + }, + { + "epoch": 0.47398, + "grad_norm": 2.0222775680987217, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47398 + }, + { + "epoch": 0.47399, + "grad_norm": 1.3687878646399407, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 47399 + }, + { + "epoch": 0.474, + "grad_norm": 1.5985350407232248, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 47400 + }, + { + "epoch": 0.47401, + "grad_norm": 1.6301203547417122, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47401 + }, + { + "epoch": 0.47402, + "grad_norm": 1.472414405440076, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 47402 + }, + { + "epoch": 0.47403, + "grad_norm": 1.6966941727647609, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 47403 + }, + { + "epoch": 0.47404, + "grad_norm": 1.4179370609007291, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 47404 + }, + { + "epoch": 0.47405, + "grad_norm": 1.5343168578117188, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 47405 + }, + { + "epoch": 0.47406, + "grad_norm": 1.410443577429476, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47406 + }, + { + "epoch": 0.47407, + "grad_norm": 1.5110254445103086, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 47407 + }, + { + "epoch": 0.47408, + "grad_norm": 1.4311899815365448, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 47408 + }, + { + "epoch": 0.47409, + "grad_norm": 1.5209975905032933, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47409 + }, + { + "epoch": 0.4741, + "grad_norm": 1.4728171350707802, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 47410 + }, + { + "epoch": 0.47411, + "grad_norm": 1.527934615397672, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47411 + }, + { + "epoch": 0.47412, + "grad_norm": 1.6122296578815223, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 47412 + }, + { + "epoch": 0.47413, + "grad_norm": 1.6114552085730853, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47413 + }, + { + "epoch": 0.47414, + "grad_norm": 1.6152971383136032, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 47414 + }, + { + "epoch": 0.47415, + "grad_norm": 1.7722191899930386, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 47415 + }, + { + "epoch": 0.47416, + "grad_norm": 1.3405231164550344, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 47416 + }, + { + "epoch": 0.47417, + "grad_norm": 1.8474319984328906, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 47417 + }, + { + "epoch": 0.47418, + "grad_norm": 1.2771156710284932, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 47418 + }, + { + "epoch": 0.47419, + "grad_norm": 1.6993543246796623, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 47419 + }, + { + "epoch": 0.4742, + "grad_norm": 1.4767402008811503, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47420 + }, + { + "epoch": 0.47421, + "grad_norm": 1.6834239461842961, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 47421 + }, + { + "epoch": 0.47422, + "grad_norm": 1.6241629201416918, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 47422 + }, + { + "epoch": 0.47423, + "grad_norm": 1.3121542733903686, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47423 + }, + { + "epoch": 0.47424, + "grad_norm": 1.6215278820291417, + "learning_rate": 0.003, + "loss": 4.013, + "step": 47424 + }, + { + "epoch": 0.47425, + "grad_norm": 1.5645985969946021, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 47425 + }, + { + "epoch": 0.47426, + "grad_norm": 1.611497277488591, + "learning_rate": 0.003, + "loss": 4.041, + "step": 47426 + }, + { + "epoch": 0.47427, + "grad_norm": 1.4665628214818094, + "learning_rate": 0.003, + "loss": 4.046, + "step": 47427 + }, + { + "epoch": 0.47428, + "grad_norm": 1.6943562696367411, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 47428 + }, + { + "epoch": 0.47429, + "grad_norm": 1.616325862849119, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 47429 + }, + { + "epoch": 0.4743, + "grad_norm": 1.5442182411761214, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 47430 + }, + { + "epoch": 0.47431, + "grad_norm": 1.4822986053671452, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47431 + }, + { + "epoch": 0.47432, + "grad_norm": 1.7112216197371126, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 47432 + }, + { + "epoch": 0.47433, + "grad_norm": 1.3600732865324818, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 47433 + }, + { + "epoch": 0.47434, + "grad_norm": 1.4469009410274305, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 47434 + }, + { + "epoch": 0.47435, + "grad_norm": 1.4478568720314031, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47435 + }, + { + "epoch": 0.47436, + "grad_norm": 1.6772927829181883, + "learning_rate": 0.003, + "loss": 4.018, + "step": 47436 + }, + { + "epoch": 0.47437, + "grad_norm": 1.4288107215496217, + "learning_rate": 0.003, + "loss": 4.037, + "step": 47437 + }, + { + "epoch": 0.47438, + "grad_norm": 1.5896166095638695, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 47438 + }, + { + "epoch": 0.47439, + "grad_norm": 1.4177165448405755, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 47439 + }, + { + "epoch": 0.4744, + "grad_norm": 1.2444812055381356, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 47440 + }, + { + "epoch": 0.47441, + "grad_norm": 1.8202267753299637, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 47441 + }, + { + "epoch": 0.47442, + "grad_norm": 1.4724564409304823, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 47442 + }, + { + "epoch": 0.47443, + "grad_norm": 1.6939753501027603, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 47443 + }, + { + "epoch": 0.47444, + "grad_norm": 1.484790024406281, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 47444 + }, + { + "epoch": 0.47445, + "grad_norm": 1.656040163819399, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 47445 + }, + { + "epoch": 0.47446, + "grad_norm": 1.8498279568312788, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 47446 + }, + { + "epoch": 0.47447, + "grad_norm": 1.4851248701477966, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 47447 + }, + { + "epoch": 0.47448, + "grad_norm": 1.5643843566289688, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 47448 + }, + { + "epoch": 0.47449, + "grad_norm": 1.663454452412723, + "learning_rate": 0.003, + "loss": 4.01, + "step": 47449 + }, + { + "epoch": 0.4745, + "grad_norm": 1.4280443322343612, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 47450 + }, + { + "epoch": 0.47451, + "grad_norm": 1.675829571179076, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 47451 + }, + { + "epoch": 0.47452, + "grad_norm": 1.469974143346149, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 47452 + }, + { + "epoch": 0.47453, + "grad_norm": 1.5155810627151252, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 47453 + }, + { + "epoch": 0.47454, + "grad_norm": 1.2897184571985907, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47454 + }, + { + "epoch": 0.47455, + "grad_norm": 1.8475822883484525, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 47455 + }, + { + "epoch": 0.47456, + "grad_norm": 1.3108671631651991, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 47456 + }, + { + "epoch": 0.47457, + "grad_norm": 1.6529381924865583, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 47457 + }, + { + "epoch": 0.47458, + "grad_norm": 1.607961910671596, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 47458 + }, + { + "epoch": 0.47459, + "grad_norm": 1.8349175786337584, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 47459 + }, + { + "epoch": 0.4746, + "grad_norm": 1.6199549128347586, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47460 + }, + { + "epoch": 0.47461, + "grad_norm": 1.7182991365711944, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 47461 + }, + { + "epoch": 0.47462, + "grad_norm": 1.837174968096592, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47462 + }, + { + "epoch": 0.47463, + "grad_norm": 1.122656079090116, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 47463 + }, + { + "epoch": 0.47464, + "grad_norm": 2.0246199423926896, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 47464 + }, + { + "epoch": 0.47465, + "grad_norm": 1.556882268947096, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 47465 + }, + { + "epoch": 0.47466, + "grad_norm": 1.5174054792974276, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 47466 + }, + { + "epoch": 0.47467, + "grad_norm": 1.6195111425698177, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 47467 + }, + { + "epoch": 0.47468, + "grad_norm": 1.47695284127008, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47468 + }, + { + "epoch": 0.47469, + "grad_norm": 1.5497555497864228, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 47469 + }, + { + "epoch": 0.4747, + "grad_norm": 1.3937162516630972, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 47470 + }, + { + "epoch": 0.47471, + "grad_norm": 1.7465160000861903, + "learning_rate": 0.003, + "loss": 4.037, + "step": 47471 + }, + { + "epoch": 0.47472, + "grad_norm": 1.2696347638344838, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 47472 + }, + { + "epoch": 0.47473, + "grad_norm": 1.51443924219371, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47473 + }, + { + "epoch": 0.47474, + "grad_norm": 1.5505138497087676, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 47474 + }, + { + "epoch": 0.47475, + "grad_norm": 1.3954367780806256, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 47475 + }, + { + "epoch": 0.47476, + "grad_norm": 1.4154777879229918, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 47476 + }, + { + "epoch": 0.47477, + "grad_norm": 1.5854972919411618, + "learning_rate": 0.003, + "loss": 4.03, + "step": 47477 + }, + { + "epoch": 0.47478, + "grad_norm": 1.7508311247797541, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 47478 + }, + { + "epoch": 0.47479, + "grad_norm": 1.5410050840834, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 47479 + }, + { + "epoch": 0.4748, + "grad_norm": 1.685399876832067, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 47480 + }, + { + "epoch": 0.47481, + "grad_norm": 1.559510491081981, + "learning_rate": 0.003, + "loss": 4.006, + "step": 47481 + }, + { + "epoch": 0.47482, + "grad_norm": 1.96445530865934, + "learning_rate": 0.003, + "loss": 4.045, + "step": 47482 + }, + { + "epoch": 0.47483, + "grad_norm": 1.3015708452831585, + "learning_rate": 0.003, + "loss": 4.002, + "step": 47483 + }, + { + "epoch": 0.47484, + "grad_norm": 1.6722620875007004, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 47484 + }, + { + "epoch": 0.47485, + "grad_norm": 1.8580470673971763, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 47485 + }, + { + "epoch": 0.47486, + "grad_norm": 1.6800510636281936, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47486 + }, + { + "epoch": 0.47487, + "grad_norm": 1.2911594979670118, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 47487 + }, + { + "epoch": 0.47488, + "grad_norm": 1.8209440037628568, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 47488 + }, + { + "epoch": 0.47489, + "grad_norm": 1.3539505660402638, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 47489 + }, + { + "epoch": 0.4749, + "grad_norm": 1.9100228357794857, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 47490 + }, + { + "epoch": 0.47491, + "grad_norm": 1.1241660407377245, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47491 + }, + { + "epoch": 0.47492, + "grad_norm": 1.7585168855363498, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 47492 + }, + { + "epoch": 0.47493, + "grad_norm": 1.894876812992782, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 47493 + }, + { + "epoch": 0.47494, + "grad_norm": 1.789938964081114, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47494 + }, + { + "epoch": 0.47495, + "grad_norm": 1.2583289673280678, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 47495 + }, + { + "epoch": 0.47496, + "grad_norm": 1.958232187411614, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 47496 + }, + { + "epoch": 0.47497, + "grad_norm": 1.1671369237813922, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 47497 + }, + { + "epoch": 0.47498, + "grad_norm": 1.7295127267432582, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 47498 + }, + { + "epoch": 0.47499, + "grad_norm": 1.3326946987984605, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47499 + }, + { + "epoch": 0.475, + "grad_norm": 1.8813224672980993, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 47500 + }, + { + "epoch": 0.47501, + "grad_norm": 1.9296081220373222, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 47501 + }, + { + "epoch": 0.47502, + "grad_norm": 1.2550309304842233, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 47502 + }, + { + "epoch": 0.47503, + "grad_norm": 1.4946692009368652, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 47503 + }, + { + "epoch": 0.47504, + "grad_norm": 1.7754656065632521, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 47504 + }, + { + "epoch": 0.47505, + "grad_norm": 1.220410877914042, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 47505 + }, + { + "epoch": 0.47506, + "grad_norm": 1.7553453736351161, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 47506 + }, + { + "epoch": 0.47507, + "grad_norm": 1.3570541242255467, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 47507 + }, + { + "epoch": 0.47508, + "grad_norm": 1.5744636600069117, + "learning_rate": 0.003, + "loss": 4.047, + "step": 47508 + }, + { + "epoch": 0.47509, + "grad_norm": 1.3513746505660489, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 47509 + }, + { + "epoch": 0.4751, + "grad_norm": 1.674821916573295, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 47510 + }, + { + "epoch": 0.47511, + "grad_norm": 2.0625534918306916, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47511 + }, + { + "epoch": 0.47512, + "grad_norm": 1.2631824045039097, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 47512 + }, + { + "epoch": 0.47513, + "grad_norm": 1.708499473787316, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47513 + }, + { + "epoch": 0.47514, + "grad_norm": 1.5351721926708068, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 47514 + }, + { + "epoch": 0.47515, + "grad_norm": 1.6903405665328017, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47515 + }, + { + "epoch": 0.47516, + "grad_norm": 1.4083788846649443, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47516 + }, + { + "epoch": 0.47517, + "grad_norm": 1.6922614730153531, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 47517 + }, + { + "epoch": 0.47518, + "grad_norm": 1.47863027853563, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 47518 + }, + { + "epoch": 0.47519, + "grad_norm": 1.5163258811700242, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 47519 + }, + { + "epoch": 0.4752, + "grad_norm": 1.6714058701543228, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 47520 + }, + { + "epoch": 0.47521, + "grad_norm": 1.9981122479078273, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 47521 + }, + { + "epoch": 0.47522, + "grad_norm": 1.244055889654045, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 47522 + }, + { + "epoch": 0.47523, + "grad_norm": 1.5482133659070811, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 47523 + }, + { + "epoch": 0.47524, + "grad_norm": 1.7680359211276597, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 47524 + }, + { + "epoch": 0.47525, + "grad_norm": 1.5417007364684716, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 47525 + }, + { + "epoch": 0.47526, + "grad_norm": 1.5667359444326965, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 47526 + }, + { + "epoch": 0.47527, + "grad_norm": 1.6628825583119935, + "learning_rate": 0.003, + "loss": 4.053, + "step": 47527 + }, + { + "epoch": 0.47528, + "grad_norm": 1.1481686591529707, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 47528 + }, + { + "epoch": 0.47529, + "grad_norm": 1.6413120385148263, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 47529 + }, + { + "epoch": 0.4753, + "grad_norm": 1.2973136580730606, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 47530 + }, + { + "epoch": 0.47531, + "grad_norm": 1.7881147691254677, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 47531 + }, + { + "epoch": 0.47532, + "grad_norm": 1.2221913745730661, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 47532 + }, + { + "epoch": 0.47533, + "grad_norm": 1.825296195097463, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 47533 + }, + { + "epoch": 0.47534, + "grad_norm": 1.2347186782249473, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 47534 + }, + { + "epoch": 0.47535, + "grad_norm": 1.858342599444278, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 47535 + }, + { + "epoch": 0.47536, + "grad_norm": 1.6444606147234384, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 47536 + }, + { + "epoch": 0.47537, + "grad_norm": 1.8298317709991334, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 47537 + }, + { + "epoch": 0.47538, + "grad_norm": 1.6829861890497233, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 47538 + }, + { + "epoch": 0.47539, + "grad_norm": 1.6365644513639408, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47539 + }, + { + "epoch": 0.4754, + "grad_norm": 1.7243281046441434, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 47540 + }, + { + "epoch": 0.47541, + "grad_norm": 1.3437408011803755, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 47541 + }, + { + "epoch": 0.47542, + "grad_norm": 1.6649234035696174, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47542 + }, + { + "epoch": 0.47543, + "grad_norm": 1.3865199743974788, + "learning_rate": 0.003, + "loss": 4.04, + "step": 47543 + }, + { + "epoch": 0.47544, + "grad_norm": 1.506800554411912, + "learning_rate": 0.003, + "loss": 4.044, + "step": 47544 + }, + { + "epoch": 0.47545, + "grad_norm": 1.6893204567702889, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 47545 + }, + { + "epoch": 0.47546, + "grad_norm": 1.5897700573926221, + "learning_rate": 0.003, + "loss": 4.038, + "step": 47546 + }, + { + "epoch": 0.47547, + "grad_norm": 1.8792454433944787, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 47547 + }, + { + "epoch": 0.47548, + "grad_norm": 1.1464531319356943, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 47548 + }, + { + "epoch": 0.47549, + "grad_norm": 1.8368948596249586, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 47549 + }, + { + "epoch": 0.4755, + "grad_norm": 1.1085227641199964, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 47550 + }, + { + "epoch": 0.47551, + "grad_norm": 1.8211405587848026, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 47551 + }, + { + "epoch": 0.47552, + "grad_norm": 1.4347409524723287, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 47552 + }, + { + "epoch": 0.47553, + "grad_norm": 1.6060283354577305, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47553 + }, + { + "epoch": 0.47554, + "grad_norm": 1.6537762632599693, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 47554 + }, + { + "epoch": 0.47555, + "grad_norm": 1.364505458115764, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47555 + }, + { + "epoch": 0.47556, + "grad_norm": 1.6345477520846747, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 47556 + }, + { + "epoch": 0.47557, + "grad_norm": 1.515432239553937, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 47557 + }, + { + "epoch": 0.47558, + "grad_norm": 1.4870184274512799, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47558 + }, + { + "epoch": 0.47559, + "grad_norm": 1.4823006084702937, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 47559 + }, + { + "epoch": 0.4756, + "grad_norm": 1.8792617297526373, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 47560 + }, + { + "epoch": 0.47561, + "grad_norm": 1.2675122461556916, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 47561 + }, + { + "epoch": 0.47562, + "grad_norm": 1.6635176974527135, + "learning_rate": 0.003, + "loss": 4.066, + "step": 47562 + }, + { + "epoch": 0.47563, + "grad_norm": 1.2623716012980095, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 47563 + }, + { + "epoch": 0.47564, + "grad_norm": 1.9978611961874362, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 47564 + }, + { + "epoch": 0.47565, + "grad_norm": 1.381381961715941, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 47565 + }, + { + "epoch": 0.47566, + "grad_norm": 1.3900299260286715, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 47566 + }, + { + "epoch": 0.47567, + "grad_norm": 1.590741685030553, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 47567 + }, + { + "epoch": 0.47568, + "grad_norm": 1.629831538454841, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 47568 + }, + { + "epoch": 0.47569, + "grad_norm": 1.2880225044890259, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 47569 + }, + { + "epoch": 0.4757, + "grad_norm": 1.5528417038335758, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 47570 + }, + { + "epoch": 0.47571, + "grad_norm": 1.3332198261208117, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47571 + }, + { + "epoch": 0.47572, + "grad_norm": 1.575611593444353, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47572 + }, + { + "epoch": 0.47573, + "grad_norm": 1.5529324233761121, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 47573 + }, + { + "epoch": 0.47574, + "grad_norm": 1.2550232677482303, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 47574 + }, + { + "epoch": 0.47575, + "grad_norm": 1.622473114701002, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 47575 + }, + { + "epoch": 0.47576, + "grad_norm": 1.4007522202387215, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47576 + }, + { + "epoch": 0.47577, + "grad_norm": 1.8377086883212381, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 47577 + }, + { + "epoch": 0.47578, + "grad_norm": 1.85167735181841, + "learning_rate": 0.003, + "loss": 4.034, + "step": 47578 + }, + { + "epoch": 0.47579, + "grad_norm": 1.3196079237721636, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 47579 + }, + { + "epoch": 0.4758, + "grad_norm": 1.3606847181179804, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 47580 + }, + { + "epoch": 0.47581, + "grad_norm": 1.7515682947216848, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 47581 + }, + { + "epoch": 0.47582, + "grad_norm": 1.664758609580425, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47582 + }, + { + "epoch": 0.47583, + "grad_norm": 1.27510826001471, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 47583 + }, + { + "epoch": 0.47584, + "grad_norm": 1.556878134957412, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 47584 + }, + { + "epoch": 0.47585, + "grad_norm": 1.5398692706828698, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 47585 + }, + { + "epoch": 0.47586, + "grad_norm": 1.5959443060274423, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 47586 + }, + { + "epoch": 0.47587, + "grad_norm": 1.6405704377939165, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 47587 + }, + { + "epoch": 0.47588, + "grad_norm": 1.7369702299084868, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 47588 + }, + { + "epoch": 0.47589, + "grad_norm": 1.4191216401705167, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47589 + }, + { + "epoch": 0.4759, + "grad_norm": 1.5093770752859264, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 47590 + }, + { + "epoch": 0.47591, + "grad_norm": 1.556904914078227, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47591 + }, + { + "epoch": 0.47592, + "grad_norm": 1.5444727904937479, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 47592 + }, + { + "epoch": 0.47593, + "grad_norm": 1.6930274010315884, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47593 + }, + { + "epoch": 0.47594, + "grad_norm": 1.4619639478392426, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 47594 + }, + { + "epoch": 0.47595, + "grad_norm": 1.689672583195372, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 47595 + }, + { + "epoch": 0.47596, + "grad_norm": 1.1743185025787344, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 47596 + }, + { + "epoch": 0.47597, + "grad_norm": 1.4992635631741769, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 47597 + }, + { + "epoch": 0.47598, + "grad_norm": 1.393577273477895, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 47598 + }, + { + "epoch": 0.47599, + "grad_norm": 1.9304043715458408, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 47599 + }, + { + "epoch": 0.476, + "grad_norm": 1.5904850662339665, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 47600 + }, + { + "epoch": 0.47601, + "grad_norm": 1.7189075606449664, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 47601 + }, + { + "epoch": 0.47602, + "grad_norm": 1.7291640039011336, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 47602 + }, + { + "epoch": 0.47603, + "grad_norm": 1.6818468902752293, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 47603 + }, + { + "epoch": 0.47604, + "grad_norm": 1.2062639955021954, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 47604 + }, + { + "epoch": 0.47605, + "grad_norm": 1.6956921637187894, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 47605 + }, + { + "epoch": 0.47606, + "grad_norm": 1.4870076053020032, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 47606 + }, + { + "epoch": 0.47607, + "grad_norm": 1.3024444325146651, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 47607 + }, + { + "epoch": 0.47608, + "grad_norm": 1.5446174522703306, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47608 + }, + { + "epoch": 0.47609, + "grad_norm": 1.5146402099759235, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 47609 + }, + { + "epoch": 0.4761, + "grad_norm": 1.6781731342051556, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 47610 + }, + { + "epoch": 0.47611, + "grad_norm": 1.2229455838596133, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47611 + }, + { + "epoch": 0.47612, + "grad_norm": 1.902314033058465, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 47612 + }, + { + "epoch": 0.47613, + "grad_norm": 1.418964162586612, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 47613 + }, + { + "epoch": 0.47614, + "grad_norm": 1.6163401954534367, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 47614 + }, + { + "epoch": 0.47615, + "grad_norm": 1.4833059679392637, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 47615 + }, + { + "epoch": 0.47616, + "grad_norm": 1.8202001134123138, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 47616 + }, + { + "epoch": 0.47617, + "grad_norm": 1.6300121028255663, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 47617 + }, + { + "epoch": 0.47618, + "grad_norm": 1.3660614124381458, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47618 + }, + { + "epoch": 0.47619, + "grad_norm": 1.8415454303323955, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 47619 + }, + { + "epoch": 0.4762, + "grad_norm": 1.2727021040622553, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 47620 + }, + { + "epoch": 0.47621, + "grad_norm": 1.7223796879639979, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 47621 + }, + { + "epoch": 0.47622, + "grad_norm": 1.4807209813293358, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 47622 + }, + { + "epoch": 0.47623, + "grad_norm": 1.5871965971107485, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 47623 + }, + { + "epoch": 0.47624, + "grad_norm": 1.7813175344177081, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 47624 + }, + { + "epoch": 0.47625, + "grad_norm": 1.2834041076791785, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 47625 + }, + { + "epoch": 0.47626, + "grad_norm": 1.4865548529871595, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 47626 + }, + { + "epoch": 0.47627, + "grad_norm": 1.4439304402269488, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 47627 + }, + { + "epoch": 0.47628, + "grad_norm": 1.643078870273023, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 47628 + }, + { + "epoch": 0.47629, + "grad_norm": 1.5278217420619395, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 47629 + }, + { + "epoch": 0.4763, + "grad_norm": 1.737082607306633, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 47630 + }, + { + "epoch": 0.47631, + "grad_norm": 1.703916579338637, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 47631 + }, + { + "epoch": 0.47632, + "grad_norm": 1.5233325513863052, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 47632 + }, + { + "epoch": 0.47633, + "grad_norm": 1.606891002185908, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 47633 + }, + { + "epoch": 0.47634, + "grad_norm": 1.6364965701433751, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47634 + }, + { + "epoch": 0.47635, + "grad_norm": 1.4324148982882967, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 47635 + }, + { + "epoch": 0.47636, + "grad_norm": 1.5049263090034335, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 47636 + }, + { + "epoch": 0.47637, + "grad_norm": 1.5038553993992352, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47637 + }, + { + "epoch": 0.47638, + "grad_norm": 1.862312127938741, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47638 + }, + { + "epoch": 0.47639, + "grad_norm": 1.6544594185060597, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 47639 + }, + { + "epoch": 0.4764, + "grad_norm": 1.5801160828372105, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 47640 + }, + { + "epoch": 0.47641, + "grad_norm": 1.7204344710173904, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 47641 + }, + { + "epoch": 0.47642, + "grad_norm": 1.382706470796316, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47642 + }, + { + "epoch": 0.47643, + "grad_norm": 1.5874189771191949, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47643 + }, + { + "epoch": 0.47644, + "grad_norm": 1.2864862930845258, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47644 + }, + { + "epoch": 0.47645, + "grad_norm": 1.5636653143674772, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 47645 + }, + { + "epoch": 0.47646, + "grad_norm": 1.2525239696656958, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 47646 + }, + { + "epoch": 0.47647, + "grad_norm": 1.4404175137284663, + "learning_rate": 0.003, + "loss": 4.009, + "step": 47647 + }, + { + "epoch": 0.47648, + "grad_norm": 1.5250476246414193, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 47648 + }, + { + "epoch": 0.47649, + "grad_norm": 1.3818108655108297, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 47649 + }, + { + "epoch": 0.4765, + "grad_norm": 1.5169410514530226, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 47650 + }, + { + "epoch": 0.47651, + "grad_norm": 1.7883651987261977, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47651 + }, + { + "epoch": 0.47652, + "grad_norm": 1.7212491879913234, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 47652 + }, + { + "epoch": 0.47653, + "grad_norm": 1.3358321774458917, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47653 + }, + { + "epoch": 0.47654, + "grad_norm": 1.6925275642795619, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 47654 + }, + { + "epoch": 0.47655, + "grad_norm": 1.5374927017664652, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 47655 + }, + { + "epoch": 0.47656, + "grad_norm": 1.4738700762281705, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 47656 + }, + { + "epoch": 0.47657, + "grad_norm": 1.7235863473370983, + "learning_rate": 0.003, + "loss": 4.056, + "step": 47657 + }, + { + "epoch": 0.47658, + "grad_norm": 1.4887687263440266, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47658 + }, + { + "epoch": 0.47659, + "grad_norm": 1.5993562930833765, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 47659 + }, + { + "epoch": 0.4766, + "grad_norm": 1.5125357466124958, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47660 + }, + { + "epoch": 0.47661, + "grad_norm": 1.693629971418108, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 47661 + }, + { + "epoch": 0.47662, + "grad_norm": 1.5864964991698978, + "learning_rate": 0.003, + "loss": 4.018, + "step": 47662 + }, + { + "epoch": 0.47663, + "grad_norm": 1.7549100371188724, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 47663 + }, + { + "epoch": 0.47664, + "grad_norm": 1.6773718804515099, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 47664 + }, + { + "epoch": 0.47665, + "grad_norm": 1.6415788487619178, + "learning_rate": 0.003, + "loss": 4.033, + "step": 47665 + }, + { + "epoch": 0.47666, + "grad_norm": 1.9706760970542092, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 47666 + }, + { + "epoch": 0.47667, + "grad_norm": 1.3135559032802422, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 47667 + }, + { + "epoch": 0.47668, + "grad_norm": 1.9317179689517472, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 47668 + }, + { + "epoch": 0.47669, + "grad_norm": 1.2937807579088216, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 47669 + }, + { + "epoch": 0.4767, + "grad_norm": 1.4237996630170882, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47670 + }, + { + "epoch": 0.47671, + "grad_norm": 1.580596276180276, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47671 + }, + { + "epoch": 0.47672, + "grad_norm": 1.4924885482330914, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 47672 + }, + { + "epoch": 0.47673, + "grad_norm": 1.5705084069014787, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 47673 + }, + { + "epoch": 0.47674, + "grad_norm": 1.7024189193579677, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 47674 + }, + { + "epoch": 0.47675, + "grad_norm": 1.6250720103030027, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 47675 + }, + { + "epoch": 0.47676, + "grad_norm": 1.3976880427356695, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47676 + }, + { + "epoch": 0.47677, + "grad_norm": 1.5915517824926175, + "learning_rate": 0.003, + "loss": 4.049, + "step": 47677 + }, + { + "epoch": 0.47678, + "grad_norm": 1.5769136723494588, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47678 + }, + { + "epoch": 0.47679, + "grad_norm": 1.4830522362097056, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 47679 + }, + { + "epoch": 0.4768, + "grad_norm": 1.6505296883447793, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 47680 + }, + { + "epoch": 0.47681, + "grad_norm": 1.3806704578955187, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47681 + }, + { + "epoch": 0.47682, + "grad_norm": 1.6635390000010652, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 47682 + }, + { + "epoch": 0.47683, + "grad_norm": 1.3789321620963146, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47683 + }, + { + "epoch": 0.47684, + "grad_norm": 1.635012235157159, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 47684 + }, + { + "epoch": 0.47685, + "grad_norm": 1.341454015684649, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 47685 + }, + { + "epoch": 0.47686, + "grad_norm": 1.7512626168154264, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 47686 + }, + { + "epoch": 0.47687, + "grad_norm": 1.5421173280822558, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 47687 + }, + { + "epoch": 0.47688, + "grad_norm": 1.4918881977546354, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 47688 + }, + { + "epoch": 0.47689, + "grad_norm": 1.548207100871589, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 47689 + }, + { + "epoch": 0.4769, + "grad_norm": 1.3080603218989462, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 47690 + }, + { + "epoch": 0.47691, + "grad_norm": 1.5078423404786363, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 47691 + }, + { + "epoch": 0.47692, + "grad_norm": 1.5613364717896865, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 47692 + }, + { + "epoch": 0.47693, + "grad_norm": 1.7638139627801244, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 47693 + }, + { + "epoch": 0.47694, + "grad_norm": 1.3242234950326341, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 47694 + }, + { + "epoch": 0.47695, + "grad_norm": 1.8283017672479691, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 47695 + }, + { + "epoch": 0.47696, + "grad_norm": 1.614129556110492, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 47696 + }, + { + "epoch": 0.47697, + "grad_norm": 1.9900679485707777, + "learning_rate": 0.003, + "loss": 4.065, + "step": 47697 + }, + { + "epoch": 0.47698, + "grad_norm": 1.5250187026152364, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47698 + }, + { + "epoch": 0.47699, + "grad_norm": 1.3716830281918033, + "learning_rate": 0.003, + "loss": 4.043, + "step": 47699 + }, + { + "epoch": 0.477, + "grad_norm": 1.825266803039483, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 47700 + }, + { + "epoch": 0.47701, + "grad_norm": 1.4103572382561276, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 47701 + }, + { + "epoch": 0.47702, + "grad_norm": 1.750970240864356, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47702 + }, + { + "epoch": 0.47703, + "grad_norm": 1.2524721210843752, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47703 + }, + { + "epoch": 0.47704, + "grad_norm": 1.6466606325456925, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 47704 + }, + { + "epoch": 0.47705, + "grad_norm": 1.359434465357206, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 47705 + }, + { + "epoch": 0.47706, + "grad_norm": 1.5955945775241578, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 47706 + }, + { + "epoch": 0.47707, + "grad_norm": 1.4466446995208309, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 47707 + }, + { + "epoch": 0.47708, + "grad_norm": 1.5399106794532973, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 47708 + }, + { + "epoch": 0.47709, + "grad_norm": 1.5274449576649427, + "learning_rate": 0.003, + "loss": 4.065, + "step": 47709 + }, + { + "epoch": 0.4771, + "grad_norm": 1.8026585790892198, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47710 + }, + { + "epoch": 0.47711, + "grad_norm": 1.7107736376111127, + "learning_rate": 0.003, + "loss": 4.053, + "step": 47711 + }, + { + "epoch": 0.47712, + "grad_norm": 1.2869765384955365, + "learning_rate": 0.003, + "loss": 4.016, + "step": 47712 + }, + { + "epoch": 0.47713, + "grad_norm": 2.1284005858240405, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 47713 + }, + { + "epoch": 0.47714, + "grad_norm": 1.3687203654473585, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 47714 + }, + { + "epoch": 0.47715, + "grad_norm": 1.5965640809317228, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 47715 + }, + { + "epoch": 0.47716, + "grad_norm": 1.7939978930942542, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 47716 + }, + { + "epoch": 0.47717, + "grad_norm": 1.3398833524724547, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 47717 + }, + { + "epoch": 0.47718, + "grad_norm": 1.4915213055768115, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 47718 + }, + { + "epoch": 0.47719, + "grad_norm": 1.5779275369888757, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 47719 + }, + { + "epoch": 0.4772, + "grad_norm": 1.4995331324167054, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 47720 + }, + { + "epoch": 0.47721, + "grad_norm": 1.553245301064251, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 47721 + }, + { + "epoch": 0.47722, + "grad_norm": 1.5731324532399436, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 47722 + }, + { + "epoch": 0.47723, + "grad_norm": 1.6707404100786474, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 47723 + }, + { + "epoch": 0.47724, + "grad_norm": 1.414672186335528, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 47724 + }, + { + "epoch": 0.47725, + "grad_norm": 1.6736146401662115, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 47725 + }, + { + "epoch": 0.47726, + "grad_norm": 1.203990530841588, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 47726 + }, + { + "epoch": 0.47727, + "grad_norm": 1.7817829455446823, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 47727 + }, + { + "epoch": 0.47728, + "grad_norm": 1.1659930060372397, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47728 + }, + { + "epoch": 0.47729, + "grad_norm": 1.8079887759919324, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 47729 + }, + { + "epoch": 0.4773, + "grad_norm": 1.586091778652921, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 47730 + }, + { + "epoch": 0.47731, + "grad_norm": 1.5065076995981896, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 47731 + }, + { + "epoch": 0.47732, + "grad_norm": 1.5651046890355687, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47732 + }, + { + "epoch": 0.47733, + "grad_norm": 1.6683306551443986, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 47733 + }, + { + "epoch": 0.47734, + "grad_norm": 1.6941736541324999, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 47734 + }, + { + "epoch": 0.47735, + "grad_norm": 1.228508751256155, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47735 + }, + { + "epoch": 0.47736, + "grad_norm": 1.5853210172206016, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 47736 + }, + { + "epoch": 0.47737, + "grad_norm": 1.8334166347765817, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 47737 + }, + { + "epoch": 0.47738, + "grad_norm": 1.4548847984760724, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 47738 + }, + { + "epoch": 0.47739, + "grad_norm": 1.637137670423325, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47739 + }, + { + "epoch": 0.4774, + "grad_norm": 1.3246894725140006, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 47740 + }, + { + "epoch": 0.47741, + "grad_norm": 1.640922010419156, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 47741 + }, + { + "epoch": 0.47742, + "grad_norm": 1.175971270653351, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 47742 + }, + { + "epoch": 0.47743, + "grad_norm": 1.9979021705627134, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 47743 + }, + { + "epoch": 0.47744, + "grad_norm": 1.789933528094988, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 47744 + }, + { + "epoch": 0.47745, + "grad_norm": 1.590588826979337, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 47745 + }, + { + "epoch": 0.47746, + "grad_norm": 1.7864573872056457, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 47746 + }, + { + "epoch": 0.47747, + "grad_norm": 1.3708924807722025, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 47747 + }, + { + "epoch": 0.47748, + "grad_norm": 1.689945807533722, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 47748 + }, + { + "epoch": 0.47749, + "grad_norm": 1.5886214164363364, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 47749 + }, + { + "epoch": 0.4775, + "grad_norm": 1.6057784206680055, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 47750 + }, + { + "epoch": 0.47751, + "grad_norm": 1.5627351703254815, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 47751 + }, + { + "epoch": 0.47752, + "grad_norm": 1.6779659466342352, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 47752 + }, + { + "epoch": 0.47753, + "grad_norm": 1.5863505875360646, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 47753 + }, + { + "epoch": 0.47754, + "grad_norm": 1.7623549096378868, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 47754 + }, + { + "epoch": 0.47755, + "grad_norm": 1.5989644688018625, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 47755 + }, + { + "epoch": 0.47756, + "grad_norm": 1.5551365498955905, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 47756 + }, + { + "epoch": 0.47757, + "grad_norm": 1.5808449092697825, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 47757 + }, + { + "epoch": 0.47758, + "grad_norm": 1.7586266894269627, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 47758 + }, + { + "epoch": 0.47759, + "grad_norm": 1.4377475183731065, + "learning_rate": 0.003, + "loss": 4.007, + "step": 47759 + }, + { + "epoch": 0.4776, + "grad_norm": 1.3981135788411034, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 47760 + }, + { + "epoch": 0.47761, + "grad_norm": 1.4367637479521478, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 47761 + }, + { + "epoch": 0.47762, + "grad_norm": 1.794871913661047, + "learning_rate": 0.003, + "loss": 4.03, + "step": 47762 + }, + { + "epoch": 0.47763, + "grad_norm": 1.2308204513664143, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47763 + }, + { + "epoch": 0.47764, + "grad_norm": 1.6780150324298881, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 47764 + }, + { + "epoch": 0.47765, + "grad_norm": 1.5312780084233837, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 47765 + }, + { + "epoch": 0.47766, + "grad_norm": 1.4994475339531508, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 47766 + }, + { + "epoch": 0.47767, + "grad_norm": 1.6705660106988727, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 47767 + }, + { + "epoch": 0.47768, + "grad_norm": 1.8044310106952643, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 47768 + }, + { + "epoch": 0.47769, + "grad_norm": 1.5809608242282824, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 47769 + }, + { + "epoch": 0.4777, + "grad_norm": 1.4301409571035766, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 47770 + }, + { + "epoch": 0.47771, + "grad_norm": 1.6876290368309887, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 47771 + }, + { + "epoch": 0.47772, + "grad_norm": 1.8039511283659961, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 47772 + }, + { + "epoch": 0.47773, + "grad_norm": 1.3204096240311503, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47773 + }, + { + "epoch": 0.47774, + "grad_norm": 1.9939485961961139, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 47774 + }, + { + "epoch": 0.47775, + "grad_norm": 0.8868168720423251, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 47775 + }, + { + "epoch": 0.47776, + "grad_norm": 1.5004177075362086, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 47776 + }, + { + "epoch": 0.47777, + "grad_norm": 1.803300235011932, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 47777 + }, + { + "epoch": 0.47778, + "grad_norm": 1.5874589428741193, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47778 + }, + { + "epoch": 0.47779, + "grad_norm": 1.485477887697305, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 47779 + }, + { + "epoch": 0.4778, + "grad_norm": 1.7749377650878317, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47780 + }, + { + "epoch": 0.47781, + "grad_norm": 1.4031198922463552, + "learning_rate": 0.003, + "loss": 4.059, + "step": 47781 + }, + { + "epoch": 0.47782, + "grad_norm": 1.5788736119536002, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 47782 + }, + { + "epoch": 0.47783, + "grad_norm": 1.5318647173904947, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 47783 + }, + { + "epoch": 0.47784, + "grad_norm": 1.8731588115414304, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 47784 + }, + { + "epoch": 0.47785, + "grad_norm": 1.5286516529271799, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 47785 + }, + { + "epoch": 0.47786, + "grad_norm": 1.7066955989247785, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 47786 + }, + { + "epoch": 0.47787, + "grad_norm": 1.373600281166987, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 47787 + }, + { + "epoch": 0.47788, + "grad_norm": 1.687249950281563, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 47788 + }, + { + "epoch": 0.47789, + "grad_norm": 1.314953739285257, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 47789 + }, + { + "epoch": 0.4779, + "grad_norm": 1.63088615459781, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 47790 + }, + { + "epoch": 0.47791, + "grad_norm": 1.3076583129267216, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 47791 + }, + { + "epoch": 0.47792, + "grad_norm": 1.7487418693494625, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 47792 + }, + { + "epoch": 0.47793, + "grad_norm": 1.4854533901278655, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47793 + }, + { + "epoch": 0.47794, + "grad_norm": 1.5335984486599348, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 47794 + }, + { + "epoch": 0.47795, + "grad_norm": 1.5023339928199178, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 47795 + }, + { + "epoch": 0.47796, + "grad_norm": 1.7100441282625656, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 47796 + }, + { + "epoch": 0.47797, + "grad_norm": 1.5873148172575071, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 47797 + }, + { + "epoch": 0.47798, + "grad_norm": 1.4090697928879046, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 47798 + }, + { + "epoch": 0.47799, + "grad_norm": 1.4727034757453268, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 47799 + }, + { + "epoch": 0.478, + "grad_norm": 1.7969151604964226, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 47800 + }, + { + "epoch": 0.47801, + "grad_norm": 1.588615311152153, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 47801 + }, + { + "epoch": 0.47802, + "grad_norm": 1.6924129160366292, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 47802 + }, + { + "epoch": 0.47803, + "grad_norm": 1.7799174381305158, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 47803 + }, + { + "epoch": 0.47804, + "grad_norm": 1.5220497543732716, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47804 + }, + { + "epoch": 0.47805, + "grad_norm": 1.2840937052203996, + "learning_rate": 0.003, + "loss": 4.033, + "step": 47805 + }, + { + "epoch": 0.47806, + "grad_norm": 1.7208335773955568, + "learning_rate": 0.003, + "loss": 4.044, + "step": 47806 + }, + { + "epoch": 0.47807, + "grad_norm": 1.195151281091043, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 47807 + }, + { + "epoch": 0.47808, + "grad_norm": 1.5364161608415412, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 47808 + }, + { + "epoch": 0.47809, + "grad_norm": 1.3734481494643063, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 47809 + }, + { + "epoch": 0.4781, + "grad_norm": 1.493620583348032, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 47810 + }, + { + "epoch": 0.47811, + "grad_norm": 1.9596103182417866, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 47811 + }, + { + "epoch": 0.47812, + "grad_norm": 1.2672633781744853, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 47812 + }, + { + "epoch": 0.47813, + "grad_norm": 1.8491011418003431, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 47813 + }, + { + "epoch": 0.47814, + "grad_norm": 1.7488565075195437, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 47814 + }, + { + "epoch": 0.47815, + "grad_norm": 1.2038342520650485, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 47815 + }, + { + "epoch": 0.47816, + "grad_norm": 1.8493691500680898, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 47816 + }, + { + "epoch": 0.47817, + "grad_norm": 1.5848421588485766, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 47817 + }, + { + "epoch": 0.47818, + "grad_norm": 1.8866378903137513, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 47818 + }, + { + "epoch": 0.47819, + "grad_norm": 1.7802518964290788, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 47819 + }, + { + "epoch": 0.4782, + "grad_norm": 1.327576453629854, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 47820 + }, + { + "epoch": 0.47821, + "grad_norm": 1.4117258512427138, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 47821 + }, + { + "epoch": 0.47822, + "grad_norm": 1.5908128634332022, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 47822 + }, + { + "epoch": 0.47823, + "grad_norm": 1.6395530219495427, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 47823 + }, + { + "epoch": 0.47824, + "grad_norm": 1.5052238847678026, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 47824 + }, + { + "epoch": 0.47825, + "grad_norm": 1.5297882947717885, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47825 + }, + { + "epoch": 0.47826, + "grad_norm": 1.460138907217596, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 47826 + }, + { + "epoch": 0.47827, + "grad_norm": 1.4745230788609878, + "learning_rate": 0.003, + "loss": 4.057, + "step": 47827 + }, + { + "epoch": 0.47828, + "grad_norm": 1.5190529292418764, + "learning_rate": 0.003, + "loss": 4.022, + "step": 47828 + }, + { + "epoch": 0.47829, + "grad_norm": 1.2897791029960173, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 47829 + }, + { + "epoch": 0.4783, + "grad_norm": 1.6137568714536683, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 47830 + }, + { + "epoch": 0.47831, + "grad_norm": 1.5283661082825997, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 47831 + }, + { + "epoch": 0.47832, + "grad_norm": 1.61356584147284, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 47832 + }, + { + "epoch": 0.47833, + "grad_norm": 2.0863136761900227, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 47833 + }, + { + "epoch": 0.47834, + "grad_norm": 1.5956204905204592, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 47834 + }, + { + "epoch": 0.47835, + "grad_norm": 1.397441864040406, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 47835 + }, + { + "epoch": 0.47836, + "grad_norm": 1.6312815431263257, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 47836 + }, + { + "epoch": 0.47837, + "grad_norm": 1.5894394882925347, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 47837 + }, + { + "epoch": 0.47838, + "grad_norm": 1.3091902814852112, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 47838 + }, + { + "epoch": 0.47839, + "grad_norm": 1.9193582364013158, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 47839 + }, + { + "epoch": 0.4784, + "grad_norm": 1.4121982567925366, + "learning_rate": 0.003, + "loss": 4.049, + "step": 47840 + }, + { + "epoch": 0.47841, + "grad_norm": 1.636024138532557, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 47841 + }, + { + "epoch": 0.47842, + "grad_norm": 1.5450952916431766, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 47842 + }, + { + "epoch": 0.47843, + "grad_norm": 1.5666216777710762, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 47843 + }, + { + "epoch": 0.47844, + "grad_norm": 1.6065743088316597, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 47844 + }, + { + "epoch": 0.47845, + "grad_norm": 1.677683506022645, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47845 + }, + { + "epoch": 0.47846, + "grad_norm": 1.4712396271775654, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47846 + }, + { + "epoch": 0.47847, + "grad_norm": 1.5795989786753892, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 47847 + }, + { + "epoch": 0.47848, + "grad_norm": 1.574756346155785, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 47848 + }, + { + "epoch": 0.47849, + "grad_norm": 1.5187377779959723, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 47849 + }, + { + "epoch": 0.4785, + "grad_norm": 1.4276630629753695, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 47850 + }, + { + "epoch": 0.47851, + "grad_norm": 1.4343031509492665, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 47851 + }, + { + "epoch": 0.47852, + "grad_norm": 1.9705460685700242, + "learning_rate": 0.003, + "loss": 4.014, + "step": 47852 + }, + { + "epoch": 0.47853, + "grad_norm": 1.2812742614815809, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 47853 + }, + { + "epoch": 0.47854, + "grad_norm": 1.7276596622205458, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 47854 + }, + { + "epoch": 0.47855, + "grad_norm": 1.405255386924618, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 47855 + }, + { + "epoch": 0.47856, + "grad_norm": 1.6242455491747811, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 47856 + }, + { + "epoch": 0.47857, + "grad_norm": 1.40703388327795, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 47857 + }, + { + "epoch": 0.47858, + "grad_norm": 1.7131235485995198, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 47858 + }, + { + "epoch": 0.47859, + "grad_norm": 1.3553975662016666, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 47859 + }, + { + "epoch": 0.4786, + "grad_norm": 1.786365096863782, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 47860 + }, + { + "epoch": 0.47861, + "grad_norm": 1.6331253368630025, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 47861 + }, + { + "epoch": 0.47862, + "grad_norm": 1.6244917989524896, + "learning_rate": 0.003, + "loss": 4.052, + "step": 47862 + }, + { + "epoch": 0.47863, + "grad_norm": 1.7784235512609656, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 47863 + }, + { + "epoch": 0.47864, + "grad_norm": 1.4872104777297037, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 47864 + }, + { + "epoch": 0.47865, + "grad_norm": 1.7760124120426846, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 47865 + }, + { + "epoch": 0.47866, + "grad_norm": 1.3798821253883842, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 47866 + }, + { + "epoch": 0.47867, + "grad_norm": 1.6210522813425419, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 47867 + }, + { + "epoch": 0.47868, + "grad_norm": 1.4385016521058187, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 47868 + }, + { + "epoch": 0.47869, + "grad_norm": 1.5520371078927127, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 47869 + }, + { + "epoch": 0.4787, + "grad_norm": 1.6021033966763865, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 47870 + }, + { + "epoch": 0.47871, + "grad_norm": 1.212524791486231, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 47871 + }, + { + "epoch": 0.47872, + "grad_norm": 1.6568243046771642, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 47872 + }, + { + "epoch": 0.47873, + "grad_norm": 1.319989857054961, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47873 + }, + { + "epoch": 0.47874, + "grad_norm": 1.932008692032148, + "learning_rate": 0.003, + "loss": 4.045, + "step": 47874 + }, + { + "epoch": 0.47875, + "grad_norm": 1.2358226789547355, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 47875 + }, + { + "epoch": 0.47876, + "grad_norm": 1.8507837713906368, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 47876 + }, + { + "epoch": 0.47877, + "grad_norm": 1.5484025282293685, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 47877 + }, + { + "epoch": 0.47878, + "grad_norm": 1.6610212267897118, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 47878 + }, + { + "epoch": 0.47879, + "grad_norm": 1.8582595409485991, + "learning_rate": 0.003, + "loss": 4.012, + "step": 47879 + }, + { + "epoch": 0.4788, + "grad_norm": 1.2148838590149147, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 47880 + }, + { + "epoch": 0.47881, + "grad_norm": 1.6222782873876322, + "learning_rate": 0.003, + "loss": 4.055, + "step": 47881 + }, + { + "epoch": 0.47882, + "grad_norm": 1.44226029282219, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 47882 + }, + { + "epoch": 0.47883, + "grad_norm": 1.7870534018166935, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47883 + }, + { + "epoch": 0.47884, + "grad_norm": 1.6093338506447503, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 47884 + }, + { + "epoch": 0.47885, + "grad_norm": 1.858435254364824, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 47885 + }, + { + "epoch": 0.47886, + "grad_norm": 1.1246998696926285, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47886 + }, + { + "epoch": 0.47887, + "grad_norm": 1.6520376280866884, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 47887 + }, + { + "epoch": 0.47888, + "grad_norm": 1.7937713409263045, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 47888 + }, + { + "epoch": 0.47889, + "grad_norm": 1.4359314500558482, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 47889 + }, + { + "epoch": 0.4789, + "grad_norm": 1.6670986392463827, + "learning_rate": 0.003, + "loss": 4.027, + "step": 47890 + }, + { + "epoch": 0.47891, + "grad_norm": 1.5825945368787995, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 47891 + }, + { + "epoch": 0.47892, + "grad_norm": 1.2830527447170357, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 47892 + }, + { + "epoch": 0.47893, + "grad_norm": 1.741745560299763, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 47893 + }, + { + "epoch": 0.47894, + "grad_norm": 1.4203918303838727, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 47894 + }, + { + "epoch": 0.47895, + "grad_norm": 1.6433825657489256, + "learning_rate": 0.003, + "loss": 4.044, + "step": 47895 + }, + { + "epoch": 0.47896, + "grad_norm": 1.40020879221034, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 47896 + }, + { + "epoch": 0.47897, + "grad_norm": 1.7329659595825984, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 47897 + }, + { + "epoch": 0.47898, + "grad_norm": 1.253587977867622, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 47898 + }, + { + "epoch": 0.47899, + "grad_norm": 2.0107932559175254, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 47899 + }, + { + "epoch": 0.479, + "grad_norm": 1.4056399482467437, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 47900 + }, + { + "epoch": 0.47901, + "grad_norm": 1.6780731002456406, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 47901 + }, + { + "epoch": 0.47902, + "grad_norm": 1.3568762798158587, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 47902 + }, + { + "epoch": 0.47903, + "grad_norm": 1.7655414307519324, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 47903 + }, + { + "epoch": 0.47904, + "grad_norm": 1.4643154156401057, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 47904 + }, + { + "epoch": 0.47905, + "grad_norm": 1.697663033795368, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 47905 + }, + { + "epoch": 0.47906, + "grad_norm": 1.6709258562244604, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 47906 + }, + { + "epoch": 0.47907, + "grad_norm": 1.4799208493714313, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 47907 + }, + { + "epoch": 0.47908, + "grad_norm": 1.3304875132154335, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 47908 + }, + { + "epoch": 0.47909, + "grad_norm": 1.5614286642845998, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 47909 + }, + { + "epoch": 0.4791, + "grad_norm": 1.5668757569888807, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 47910 + }, + { + "epoch": 0.47911, + "grad_norm": 4.15597848681671, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 47911 + }, + { + "epoch": 0.47912, + "grad_norm": 4.197161577337637, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 47912 + }, + { + "epoch": 0.47913, + "grad_norm": 1.8277222472976289, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 47913 + }, + { + "epoch": 0.47914, + "grad_norm": 2.1127469955200633, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 47914 + }, + { + "epoch": 0.47915, + "grad_norm": 1.846882517221966, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 47915 + }, + { + "epoch": 0.47916, + "grad_norm": 2.430223112734576, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 47916 + }, + { + "epoch": 0.47917, + "grad_norm": 1.5906950830628166, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 47917 + }, + { + "epoch": 0.47918, + "grad_norm": 1.9119808555186284, + "learning_rate": 0.003, + "loss": 4.143, + "step": 47918 + }, + { + "epoch": 0.47919, + "grad_norm": 1.8541487816583102, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 47919 + }, + { + "epoch": 0.4792, + "grad_norm": 1.879714914084642, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 47920 + }, + { + "epoch": 0.47921, + "grad_norm": 2.32252974595521, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 47921 + }, + { + "epoch": 0.47922, + "grad_norm": 1.8671724109120553, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 47922 + }, + { + "epoch": 0.47923, + "grad_norm": 1.841342638207654, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 47923 + }, + { + "epoch": 0.47924, + "grad_norm": 1.4860117648203972, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 47924 + }, + { + "epoch": 0.47925, + "grad_norm": 1.8546259827687772, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 47925 + }, + { + "epoch": 0.47926, + "grad_norm": 1.7815848563881407, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 47926 + }, + { + "epoch": 0.47927, + "grad_norm": 1.849458372309168, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 47927 + }, + { + "epoch": 0.47928, + "grad_norm": 1.820575565616045, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 47928 + }, + { + "epoch": 0.47929, + "grad_norm": 1.6400897171421176, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 47929 + }, + { + "epoch": 0.4793, + "grad_norm": 2.1110198569644645, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 47930 + }, + { + "epoch": 0.47931, + "grad_norm": 2.069974949008832, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 47931 + }, + { + "epoch": 0.47932, + "grad_norm": 2.439063786796243, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 47932 + }, + { + "epoch": 0.47933, + "grad_norm": 1.5898271299938027, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 47933 + }, + { + "epoch": 0.47934, + "grad_norm": 1.6521851485018295, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 47934 + }, + { + "epoch": 0.47935, + "grad_norm": 1.5157778276968676, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 47935 + }, + { + "epoch": 0.47936, + "grad_norm": 1.8533572930695774, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 47936 + }, + { + "epoch": 0.47937, + "grad_norm": 1.2760214036333413, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 47937 + }, + { + "epoch": 0.47938, + "grad_norm": 1.7958051686072443, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 47938 + }, + { + "epoch": 0.47939, + "grad_norm": 1.7024502450020815, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 47939 + }, + { + "epoch": 0.4794, + "grad_norm": 1.4150667960978005, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 47940 + }, + { + "epoch": 0.47941, + "grad_norm": 1.675215421762758, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 47941 + }, + { + "epoch": 0.47942, + "grad_norm": 1.4286484439506753, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 47942 + }, + { + "epoch": 0.47943, + "grad_norm": 1.9607235189893322, + "learning_rate": 0.003, + "loss": 4.106, + "step": 47943 + }, + { + "epoch": 0.47944, + "grad_norm": 1.3508212141995934, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 47944 + }, + { + "epoch": 0.47945, + "grad_norm": 1.941325263717958, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 47945 + }, + { + "epoch": 0.47946, + "grad_norm": 1.3071754477034447, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 47946 + }, + { + "epoch": 0.47947, + "grad_norm": 1.855899621972758, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 47947 + }, + { + "epoch": 0.47948, + "grad_norm": 1.5723916059479581, + "learning_rate": 0.003, + "loss": 4.072, + "step": 47948 + }, + { + "epoch": 0.47949, + "grad_norm": 2.1186473516324074, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 47949 + }, + { + "epoch": 0.4795, + "grad_norm": 1.5438831341043837, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 47950 + }, + { + "epoch": 0.47951, + "grad_norm": 1.3456238392901467, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 47951 + }, + { + "epoch": 0.47952, + "grad_norm": 1.8568502158004438, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 47952 + }, + { + "epoch": 0.47953, + "grad_norm": 1.568278856643827, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 47953 + }, + { + "epoch": 0.47954, + "grad_norm": 1.821528209048954, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 47954 + }, + { + "epoch": 0.47955, + "grad_norm": 1.280370815767927, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 47955 + }, + { + "epoch": 0.47956, + "grad_norm": 1.7157359293979642, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 47956 + }, + { + "epoch": 0.47957, + "grad_norm": 1.5585520971929216, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 47957 + }, + { + "epoch": 0.47958, + "grad_norm": 1.9171567182831624, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 47958 + }, + { + "epoch": 0.47959, + "grad_norm": 1.1705632138988078, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 47959 + }, + { + "epoch": 0.4796, + "grad_norm": 1.7452116448981643, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 47960 + }, + { + "epoch": 0.47961, + "grad_norm": 1.4652553100953367, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 47961 + }, + { + "epoch": 0.47962, + "grad_norm": 1.4657428256873648, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 47962 + }, + { + "epoch": 0.47963, + "grad_norm": 1.6455915131354555, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 47963 + }, + { + "epoch": 0.47964, + "grad_norm": 1.4383055636770838, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 47964 + }, + { + "epoch": 0.47965, + "grad_norm": 1.4877773927789641, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 47965 + }, + { + "epoch": 0.47966, + "grad_norm": 1.3644288204002963, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 47966 + }, + { + "epoch": 0.47967, + "grad_norm": 1.6019702419719457, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 47967 + }, + { + "epoch": 0.47968, + "grad_norm": 1.4849126970513813, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 47968 + }, + { + "epoch": 0.47969, + "grad_norm": 1.22044226555372, + "learning_rate": 0.003, + "loss": 4.045, + "step": 47969 + }, + { + "epoch": 0.4797, + "grad_norm": 1.550004229395818, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 47970 + }, + { + "epoch": 0.47971, + "grad_norm": 1.5148486464022497, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 47971 + }, + { + "epoch": 0.47972, + "grad_norm": 1.5113070777952626, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 47972 + }, + { + "epoch": 0.47973, + "grad_norm": 1.4663969767955936, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 47973 + }, + { + "epoch": 0.47974, + "grad_norm": 1.8532892952736373, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 47974 + }, + { + "epoch": 0.47975, + "grad_norm": 1.7864670704817105, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 47975 + }, + { + "epoch": 0.47976, + "grad_norm": 1.676949475516878, + "learning_rate": 0.003, + "loss": 4.075, + "step": 47976 + }, + { + "epoch": 0.47977, + "grad_norm": 1.4490188221522362, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 47977 + }, + { + "epoch": 0.47978, + "grad_norm": 1.355492890426102, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 47978 + }, + { + "epoch": 0.47979, + "grad_norm": 1.6516020212872562, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 47979 + }, + { + "epoch": 0.4798, + "grad_norm": 1.3571346065170604, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 47980 + }, + { + "epoch": 0.47981, + "grad_norm": 1.9006298181896544, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 47981 + }, + { + "epoch": 0.47982, + "grad_norm": 1.1168384047610822, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 47982 + }, + { + "epoch": 0.47983, + "grad_norm": 1.6161194969325103, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 47983 + }, + { + "epoch": 0.47984, + "grad_norm": 1.3626268037400944, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 47984 + }, + { + "epoch": 0.47985, + "grad_norm": 1.554258798302168, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 47985 + }, + { + "epoch": 0.47986, + "grad_norm": 1.2064995319909517, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 47986 + }, + { + "epoch": 0.47987, + "grad_norm": 1.4292147107954778, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 47987 + }, + { + "epoch": 0.47988, + "grad_norm": 1.5058378270645414, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 47988 + }, + { + "epoch": 0.47989, + "grad_norm": 1.317332667821146, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 47989 + }, + { + "epoch": 0.4799, + "grad_norm": 1.5642686218684776, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 47990 + }, + { + "epoch": 0.47991, + "grad_norm": 1.6797827346023302, + "learning_rate": 0.003, + "loss": 4.029, + "step": 47991 + }, + { + "epoch": 0.47992, + "grad_norm": 2.0307468963840503, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 47992 + }, + { + "epoch": 0.47993, + "grad_norm": 1.4550132670986582, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 47993 + }, + { + "epoch": 0.47994, + "grad_norm": 1.5933362796878772, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 47994 + }, + { + "epoch": 0.47995, + "grad_norm": 1.4074466191393709, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 47995 + }, + { + "epoch": 0.47996, + "grad_norm": 1.5877873531960622, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 47996 + }, + { + "epoch": 0.47997, + "grad_norm": 1.6186071520868015, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 47997 + }, + { + "epoch": 0.47998, + "grad_norm": 1.3943627576098905, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 47998 + }, + { + "epoch": 0.47999, + "grad_norm": 1.5526338275452798, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 47999 + }, + { + "epoch": 0.48, + "grad_norm": 1.6730204079223627, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 48000 + }, + { + "epoch": 0.48001, + "grad_norm": 1.2803395522232113, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 48001 + }, + { + "epoch": 0.48002, + "grad_norm": 1.9594926688287142, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48002 + }, + { + "epoch": 0.48003, + "grad_norm": 1.2190595834090243, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48003 + }, + { + "epoch": 0.48004, + "grad_norm": 1.7723592619690205, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 48004 + }, + { + "epoch": 0.48005, + "grad_norm": 1.5260551276247465, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 48005 + }, + { + "epoch": 0.48006, + "grad_norm": 1.4175263726693588, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48006 + }, + { + "epoch": 0.48007, + "grad_norm": 1.4274088326483265, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48007 + }, + { + "epoch": 0.48008, + "grad_norm": 1.6145933561411967, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 48008 + }, + { + "epoch": 0.48009, + "grad_norm": 1.3962921550748366, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 48009 + }, + { + "epoch": 0.4801, + "grad_norm": 1.6297043976637011, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 48010 + }, + { + "epoch": 0.48011, + "grad_norm": 1.433259440138183, + "learning_rate": 0.003, + "loss": 4.032, + "step": 48011 + }, + { + "epoch": 0.48012, + "grad_norm": 1.673120464664782, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 48012 + }, + { + "epoch": 0.48013, + "grad_norm": 1.3358278264544838, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 48013 + }, + { + "epoch": 0.48014, + "grad_norm": 1.546591035918268, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 48014 + }, + { + "epoch": 0.48015, + "grad_norm": 1.643110659633033, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 48015 + }, + { + "epoch": 0.48016, + "grad_norm": 1.6935138368817804, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 48016 + }, + { + "epoch": 0.48017, + "grad_norm": 1.4979848575691233, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 48017 + }, + { + "epoch": 0.48018, + "grad_norm": 1.4280095313272967, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 48018 + }, + { + "epoch": 0.48019, + "grad_norm": 1.6002949990351532, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 48019 + }, + { + "epoch": 0.4802, + "grad_norm": 1.3928741325075347, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48020 + }, + { + "epoch": 0.48021, + "grad_norm": 1.7238546885744959, + "learning_rate": 0.003, + "loss": 4.047, + "step": 48021 + }, + { + "epoch": 0.48022, + "grad_norm": 1.500538597903649, + "learning_rate": 0.003, + "loss": 4.027, + "step": 48022 + }, + { + "epoch": 0.48023, + "grad_norm": 1.4261147447018048, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48023 + }, + { + "epoch": 0.48024, + "grad_norm": 1.6792243085563554, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 48024 + }, + { + "epoch": 0.48025, + "grad_norm": 1.4426915184548303, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 48025 + }, + { + "epoch": 0.48026, + "grad_norm": 1.548209256801454, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48026 + }, + { + "epoch": 0.48027, + "grad_norm": 1.2964546240397348, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 48027 + }, + { + "epoch": 0.48028, + "grad_norm": 1.8563010364389663, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 48028 + }, + { + "epoch": 0.48029, + "grad_norm": 1.5712960522776205, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 48029 + }, + { + "epoch": 0.4803, + "grad_norm": 1.7540972964366242, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 48030 + }, + { + "epoch": 0.48031, + "grad_norm": 1.5357536522134838, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 48031 + }, + { + "epoch": 0.48032, + "grad_norm": 1.5729973764739114, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 48032 + }, + { + "epoch": 0.48033, + "grad_norm": 1.3301030928844026, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 48033 + }, + { + "epoch": 0.48034, + "grad_norm": 1.6209748459587494, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 48034 + }, + { + "epoch": 0.48035, + "grad_norm": 1.5124535763362112, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48035 + }, + { + "epoch": 0.48036, + "grad_norm": 1.7094105565271827, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 48036 + }, + { + "epoch": 0.48037, + "grad_norm": 1.090981734083127, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 48037 + }, + { + "epoch": 0.48038, + "grad_norm": 1.9129002900726764, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 48038 + }, + { + "epoch": 0.48039, + "grad_norm": 1.658374719858842, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48039 + }, + { + "epoch": 0.4804, + "grad_norm": 1.4221416830953384, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 48040 + }, + { + "epoch": 0.48041, + "grad_norm": 1.3753521659259034, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 48041 + }, + { + "epoch": 0.48042, + "grad_norm": 1.6714594111500454, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48042 + }, + { + "epoch": 0.48043, + "grad_norm": 1.4793586142205752, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48043 + }, + { + "epoch": 0.48044, + "grad_norm": 1.6456628372987032, + "learning_rate": 0.003, + "loss": 4.033, + "step": 48044 + }, + { + "epoch": 0.48045, + "grad_norm": 1.4316221918967693, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48045 + }, + { + "epoch": 0.48046, + "grad_norm": 1.8233242585424758, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 48046 + }, + { + "epoch": 0.48047, + "grad_norm": 1.2270261579913766, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48047 + }, + { + "epoch": 0.48048, + "grad_norm": 2.000668067912973, + "learning_rate": 0.003, + "loss": 4.049, + "step": 48048 + }, + { + "epoch": 0.48049, + "grad_norm": 1.3037635097114662, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 48049 + }, + { + "epoch": 0.4805, + "grad_norm": 1.635392295194611, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 48050 + }, + { + "epoch": 0.48051, + "grad_norm": 1.6944566388939313, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48051 + }, + { + "epoch": 0.48052, + "grad_norm": 1.6106925144203221, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 48052 + }, + { + "epoch": 0.48053, + "grad_norm": 1.3792124265639065, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48053 + }, + { + "epoch": 0.48054, + "grad_norm": 1.5937573995517391, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 48054 + }, + { + "epoch": 0.48055, + "grad_norm": 1.4263826252150777, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 48055 + }, + { + "epoch": 0.48056, + "grad_norm": 1.5314034639436982, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48056 + }, + { + "epoch": 0.48057, + "grad_norm": 1.3141553732513533, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 48057 + }, + { + "epoch": 0.48058, + "grad_norm": 1.626101177019928, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 48058 + }, + { + "epoch": 0.48059, + "grad_norm": 1.434218669145151, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48059 + }, + { + "epoch": 0.4806, + "grad_norm": 1.532808513254375, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48060 + }, + { + "epoch": 0.48061, + "grad_norm": 1.5872926827529779, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 48061 + }, + { + "epoch": 0.48062, + "grad_norm": 1.612273281229855, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 48062 + }, + { + "epoch": 0.48063, + "grad_norm": 1.8676716679720384, + "learning_rate": 0.003, + "loss": 4.048, + "step": 48063 + }, + { + "epoch": 0.48064, + "grad_norm": 1.646170675492355, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 48064 + }, + { + "epoch": 0.48065, + "grad_norm": 1.4401412828778382, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 48065 + }, + { + "epoch": 0.48066, + "grad_norm": 1.7197733358694356, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 48066 + }, + { + "epoch": 0.48067, + "grad_norm": 1.582423980413022, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48067 + }, + { + "epoch": 0.48068, + "grad_norm": 1.4505323423361782, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 48068 + }, + { + "epoch": 0.48069, + "grad_norm": 1.7252126469587132, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 48069 + }, + { + "epoch": 0.4807, + "grad_norm": 1.4423696889759268, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 48070 + }, + { + "epoch": 0.48071, + "grad_norm": 1.6198412275661231, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 48071 + }, + { + "epoch": 0.48072, + "grad_norm": 1.4257889658533744, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 48072 + }, + { + "epoch": 0.48073, + "grad_norm": 1.6417378614742462, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 48073 + }, + { + "epoch": 0.48074, + "grad_norm": 1.379507340339308, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48074 + }, + { + "epoch": 0.48075, + "grad_norm": 1.6821808994924317, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 48075 + }, + { + "epoch": 0.48076, + "grad_norm": 1.356128904506012, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48076 + }, + { + "epoch": 0.48077, + "grad_norm": 1.6231561201665803, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 48077 + }, + { + "epoch": 0.48078, + "grad_norm": 1.2209402470775157, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 48078 + }, + { + "epoch": 0.48079, + "grad_norm": 1.7217784805101006, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 48079 + }, + { + "epoch": 0.4808, + "grad_norm": 1.1507747375702007, + "learning_rate": 0.003, + "loss": 4.012, + "step": 48080 + }, + { + "epoch": 0.48081, + "grad_norm": 1.7110812571839333, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48081 + }, + { + "epoch": 0.48082, + "grad_norm": 1.123545998829436, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 48082 + }, + { + "epoch": 0.48083, + "grad_norm": 1.7643884061275434, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 48083 + }, + { + "epoch": 0.48084, + "grad_norm": 1.6932405369213965, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 48084 + }, + { + "epoch": 0.48085, + "grad_norm": 2.0577382623612555, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 48085 + }, + { + "epoch": 0.48086, + "grad_norm": 1.2310732671301923, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 48086 + }, + { + "epoch": 0.48087, + "grad_norm": 1.5561860642952319, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48087 + }, + { + "epoch": 0.48088, + "grad_norm": 1.801291288139202, + "learning_rate": 0.003, + "loss": 4.055, + "step": 48088 + }, + { + "epoch": 0.48089, + "grad_norm": 1.7710498210975607, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 48089 + }, + { + "epoch": 0.4809, + "grad_norm": 1.2057947810152654, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48090 + }, + { + "epoch": 0.48091, + "grad_norm": 1.622414175646674, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48091 + }, + { + "epoch": 0.48092, + "grad_norm": 1.6449885022307769, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48092 + }, + { + "epoch": 0.48093, + "grad_norm": 1.4192870743762869, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 48093 + }, + { + "epoch": 0.48094, + "grad_norm": 1.489791787911697, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 48094 + }, + { + "epoch": 0.48095, + "grad_norm": 1.443443499908824, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48095 + }, + { + "epoch": 0.48096, + "grad_norm": 1.9873090325725713, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48096 + }, + { + "epoch": 0.48097, + "grad_norm": 1.464347281698223, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 48097 + }, + { + "epoch": 0.48098, + "grad_norm": 1.6986565824317827, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48098 + }, + { + "epoch": 0.48099, + "grad_norm": 1.3625706662653936, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 48099 + }, + { + "epoch": 0.481, + "grad_norm": 1.5929859405611067, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 48100 + }, + { + "epoch": 0.48101, + "grad_norm": 1.4957587955167544, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 48101 + }, + { + "epoch": 0.48102, + "grad_norm": 1.3526163847548167, + "learning_rate": 0.003, + "loss": 4.036, + "step": 48102 + }, + { + "epoch": 0.48103, + "grad_norm": 1.4222436307788593, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48103 + }, + { + "epoch": 0.48104, + "grad_norm": 1.5166882136901303, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 48104 + }, + { + "epoch": 0.48105, + "grad_norm": 1.6121489215458553, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 48105 + }, + { + "epoch": 0.48106, + "grad_norm": 1.3338324955648444, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 48106 + }, + { + "epoch": 0.48107, + "grad_norm": 1.5923127315433245, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48107 + }, + { + "epoch": 0.48108, + "grad_norm": 1.4135732709567763, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48108 + }, + { + "epoch": 0.48109, + "grad_norm": 1.959646689171617, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 48109 + }, + { + "epoch": 0.4811, + "grad_norm": 1.1562328260720582, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 48110 + }, + { + "epoch": 0.48111, + "grad_norm": 1.6175262320745953, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 48111 + }, + { + "epoch": 0.48112, + "grad_norm": 1.9254903772149816, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 48112 + }, + { + "epoch": 0.48113, + "grad_norm": 1.5699588972673342, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 48113 + }, + { + "epoch": 0.48114, + "grad_norm": 1.314320681205029, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 48114 + }, + { + "epoch": 0.48115, + "grad_norm": 1.70162441792126, + "learning_rate": 0.003, + "loss": 4.053, + "step": 48115 + }, + { + "epoch": 0.48116, + "grad_norm": 1.4799553263796779, + "learning_rate": 0.003, + "loss": 4.064, + "step": 48116 + }, + { + "epoch": 0.48117, + "grad_norm": 1.6527692509450465, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 48117 + }, + { + "epoch": 0.48118, + "grad_norm": 1.337720911808876, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48118 + }, + { + "epoch": 0.48119, + "grad_norm": 1.6380534555613706, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 48119 + }, + { + "epoch": 0.4812, + "grad_norm": 1.8942808787229373, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 48120 + }, + { + "epoch": 0.48121, + "grad_norm": 1.5411395223292588, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48121 + }, + { + "epoch": 0.48122, + "grad_norm": 1.7655351686295218, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 48122 + }, + { + "epoch": 0.48123, + "grad_norm": 1.252513195356638, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 48123 + }, + { + "epoch": 0.48124, + "grad_norm": 1.7335962988960338, + "learning_rate": 0.003, + "loss": 4.072, + "step": 48124 + }, + { + "epoch": 0.48125, + "grad_norm": 1.3868937423467562, + "learning_rate": 0.003, + "loss": 4.037, + "step": 48125 + }, + { + "epoch": 0.48126, + "grad_norm": 1.5372734015703458, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48126 + }, + { + "epoch": 0.48127, + "grad_norm": 1.6184934404894638, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 48127 + }, + { + "epoch": 0.48128, + "grad_norm": 1.368737063179657, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48128 + }, + { + "epoch": 0.48129, + "grad_norm": 1.897233096991671, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48129 + }, + { + "epoch": 0.4813, + "grad_norm": 1.4636839338143617, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 48130 + }, + { + "epoch": 0.48131, + "grad_norm": 1.4984879595454887, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 48131 + }, + { + "epoch": 0.48132, + "grad_norm": 1.8389158105698014, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 48132 + }, + { + "epoch": 0.48133, + "grad_norm": 1.2530987183559772, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 48133 + }, + { + "epoch": 0.48134, + "grad_norm": 1.7370502609525575, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48134 + }, + { + "epoch": 0.48135, + "grad_norm": 1.3185272744229999, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 48135 + }, + { + "epoch": 0.48136, + "grad_norm": 1.8452037579206804, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48136 + }, + { + "epoch": 0.48137, + "grad_norm": 1.5106816094423414, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 48137 + }, + { + "epoch": 0.48138, + "grad_norm": 1.565403563872762, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 48138 + }, + { + "epoch": 0.48139, + "grad_norm": 1.3865989234571228, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48139 + }, + { + "epoch": 0.4814, + "grad_norm": 1.4829086582029871, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 48140 + }, + { + "epoch": 0.48141, + "grad_norm": 1.894531042916759, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 48141 + }, + { + "epoch": 0.48142, + "grad_norm": 1.3435796761416314, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 48142 + }, + { + "epoch": 0.48143, + "grad_norm": 1.7981401534597594, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 48143 + }, + { + "epoch": 0.48144, + "grad_norm": 1.4054905249778773, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48144 + }, + { + "epoch": 0.48145, + "grad_norm": 1.9529880497187957, + "learning_rate": 0.003, + "loss": 4.065, + "step": 48145 + }, + { + "epoch": 0.48146, + "grad_norm": 1.3651801277876903, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48146 + }, + { + "epoch": 0.48147, + "grad_norm": 1.6377546028354335, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48147 + }, + { + "epoch": 0.48148, + "grad_norm": 1.5531588589673844, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 48148 + }, + { + "epoch": 0.48149, + "grad_norm": 1.3030464856171253, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48149 + }, + { + "epoch": 0.4815, + "grad_norm": 1.5191752517782247, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 48150 + }, + { + "epoch": 0.48151, + "grad_norm": 1.5925779801543638, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 48151 + }, + { + "epoch": 0.48152, + "grad_norm": 1.505256887613549, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 48152 + }, + { + "epoch": 0.48153, + "grad_norm": 1.4793054787289956, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 48153 + }, + { + "epoch": 0.48154, + "grad_norm": 1.5692264399143003, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 48154 + }, + { + "epoch": 0.48155, + "grad_norm": 1.1787306298056197, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 48155 + }, + { + "epoch": 0.48156, + "grad_norm": 2.005896239605801, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 48156 + }, + { + "epoch": 0.48157, + "grad_norm": 1.2395281781195253, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 48157 + }, + { + "epoch": 0.48158, + "grad_norm": 2.0064949693033833, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 48158 + }, + { + "epoch": 0.48159, + "grad_norm": 1.366267289082987, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 48159 + }, + { + "epoch": 0.4816, + "grad_norm": 1.402451321148872, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 48160 + }, + { + "epoch": 0.48161, + "grad_norm": 1.8744410020526623, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 48161 + }, + { + "epoch": 0.48162, + "grad_norm": 1.4430664361592682, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 48162 + }, + { + "epoch": 0.48163, + "grad_norm": 1.4701178967378243, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48163 + }, + { + "epoch": 0.48164, + "grad_norm": 1.5019406297668716, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 48164 + }, + { + "epoch": 0.48165, + "grad_norm": 1.4458604344889152, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48165 + }, + { + "epoch": 0.48166, + "grad_norm": 1.3947190634128632, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 48166 + }, + { + "epoch": 0.48167, + "grad_norm": 1.6319100392256858, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48167 + }, + { + "epoch": 0.48168, + "grad_norm": 1.5041612284751855, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 48168 + }, + { + "epoch": 0.48169, + "grad_norm": 1.9008292882268005, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 48169 + }, + { + "epoch": 0.4817, + "grad_norm": 1.4291663343808314, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 48170 + }, + { + "epoch": 0.48171, + "grad_norm": 1.7674429337744832, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 48171 + }, + { + "epoch": 0.48172, + "grad_norm": 1.361848432240744, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48172 + }, + { + "epoch": 0.48173, + "grad_norm": 1.1270752186226791, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 48173 + }, + { + "epoch": 0.48174, + "grad_norm": 1.7388067215021528, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 48174 + }, + { + "epoch": 0.48175, + "grad_norm": 1.4515772948335626, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 48175 + }, + { + "epoch": 0.48176, + "grad_norm": 1.7221997033740852, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48176 + }, + { + "epoch": 0.48177, + "grad_norm": 1.5869615300930326, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 48177 + }, + { + "epoch": 0.48178, + "grad_norm": 1.5678614286586925, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48178 + }, + { + "epoch": 0.48179, + "grad_norm": 1.3232951540802818, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 48179 + }, + { + "epoch": 0.4818, + "grad_norm": 1.4928570718236862, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48180 + }, + { + "epoch": 0.48181, + "grad_norm": 1.5831255617729418, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 48181 + }, + { + "epoch": 0.48182, + "grad_norm": 1.4967292751388963, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 48182 + }, + { + "epoch": 0.48183, + "grad_norm": 1.569956374906785, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48183 + }, + { + "epoch": 0.48184, + "grad_norm": 1.9971823345333037, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 48184 + }, + { + "epoch": 0.48185, + "grad_norm": 1.4646791013939147, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 48185 + }, + { + "epoch": 0.48186, + "grad_norm": 1.6233540871498346, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 48186 + }, + { + "epoch": 0.48187, + "grad_norm": 1.7531154369628903, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 48187 + }, + { + "epoch": 0.48188, + "grad_norm": 1.3452578239900292, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 48188 + }, + { + "epoch": 0.48189, + "grad_norm": 1.574768251229144, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 48189 + }, + { + "epoch": 0.4819, + "grad_norm": 1.522045230160112, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 48190 + }, + { + "epoch": 0.48191, + "grad_norm": 1.4287518612336467, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 48191 + }, + { + "epoch": 0.48192, + "grad_norm": 1.451517458813355, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48192 + }, + { + "epoch": 0.48193, + "grad_norm": 1.3067128340490104, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 48193 + }, + { + "epoch": 0.48194, + "grad_norm": 2.0330335453841566, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 48194 + }, + { + "epoch": 0.48195, + "grad_norm": 1.1339754467626275, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 48195 + }, + { + "epoch": 0.48196, + "grad_norm": 1.861417063750313, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 48196 + }, + { + "epoch": 0.48197, + "grad_norm": 1.3136097537723126, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 48197 + }, + { + "epoch": 0.48198, + "grad_norm": 2.0602228579102007, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 48198 + }, + { + "epoch": 0.48199, + "grad_norm": 1.490990736069653, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 48199 + }, + { + "epoch": 0.482, + "grad_norm": 1.5394341890748702, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48200 + }, + { + "epoch": 0.48201, + "grad_norm": 1.4734880817777607, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 48201 + }, + { + "epoch": 0.48202, + "grad_norm": 1.7776673803016025, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 48202 + }, + { + "epoch": 0.48203, + "grad_norm": 1.2393882581207576, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48203 + }, + { + "epoch": 0.48204, + "grad_norm": 1.6346606834680912, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 48204 + }, + { + "epoch": 0.48205, + "grad_norm": 1.5159394568632543, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48205 + }, + { + "epoch": 0.48206, + "grad_norm": 1.2581010856041839, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 48206 + }, + { + "epoch": 0.48207, + "grad_norm": 1.72989987130978, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 48207 + }, + { + "epoch": 0.48208, + "grad_norm": 1.2742439975878992, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 48208 + }, + { + "epoch": 0.48209, + "grad_norm": 1.9598823512382817, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 48209 + }, + { + "epoch": 0.4821, + "grad_norm": 1.5858438681806277, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 48210 + }, + { + "epoch": 0.48211, + "grad_norm": 1.8943596595903063, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 48211 + }, + { + "epoch": 0.48212, + "grad_norm": 1.4463617544566407, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 48212 + }, + { + "epoch": 0.48213, + "grad_norm": 1.4911953697884348, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 48213 + }, + { + "epoch": 0.48214, + "grad_norm": 1.4933072825117824, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48214 + }, + { + "epoch": 0.48215, + "grad_norm": 1.3763038172199504, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 48215 + }, + { + "epoch": 0.48216, + "grad_norm": 1.4826701557704378, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48216 + }, + { + "epoch": 0.48217, + "grad_norm": 1.762141709006684, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 48217 + }, + { + "epoch": 0.48218, + "grad_norm": 1.4772092823336629, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 48218 + }, + { + "epoch": 0.48219, + "grad_norm": 2.2146390252498325, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 48219 + }, + { + "epoch": 0.4822, + "grad_norm": 1.0606292506611883, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 48220 + }, + { + "epoch": 0.48221, + "grad_norm": 1.6813471177894577, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 48221 + }, + { + "epoch": 0.48222, + "grad_norm": 1.4597879790321564, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 48222 + }, + { + "epoch": 0.48223, + "grad_norm": 1.7213829810889918, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 48223 + }, + { + "epoch": 0.48224, + "grad_norm": 1.3379081781065412, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 48224 + }, + { + "epoch": 0.48225, + "grad_norm": 1.7187757796085446, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 48225 + }, + { + "epoch": 0.48226, + "grad_norm": 1.3328671877430742, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 48226 + }, + { + "epoch": 0.48227, + "grad_norm": 1.8607205949819114, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 48227 + }, + { + "epoch": 0.48228, + "grad_norm": 1.429192203728896, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 48228 + }, + { + "epoch": 0.48229, + "grad_norm": 1.861075505969897, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48229 + }, + { + "epoch": 0.4823, + "grad_norm": 1.52966762881119, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 48230 + }, + { + "epoch": 0.48231, + "grad_norm": 1.4759021681351336, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48231 + }, + { + "epoch": 0.48232, + "grad_norm": 1.5986325816205198, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48232 + }, + { + "epoch": 0.48233, + "grad_norm": 1.3770899931619143, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48233 + }, + { + "epoch": 0.48234, + "grad_norm": 1.5870400896484258, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 48234 + }, + { + "epoch": 0.48235, + "grad_norm": 1.5225054563615803, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 48235 + }, + { + "epoch": 0.48236, + "grad_norm": 1.4928316398479122, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 48236 + }, + { + "epoch": 0.48237, + "grad_norm": 1.433517279134971, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48237 + }, + { + "epoch": 0.48238, + "grad_norm": 1.4560206574042636, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 48238 + }, + { + "epoch": 0.48239, + "grad_norm": 1.6361425202408142, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 48239 + }, + { + "epoch": 0.4824, + "grad_norm": 1.5128497482978889, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 48240 + }, + { + "epoch": 0.48241, + "grad_norm": 1.5546083742836665, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 48241 + }, + { + "epoch": 0.48242, + "grad_norm": 1.2818488939509072, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 48242 + }, + { + "epoch": 0.48243, + "grad_norm": 1.7817374047558086, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48243 + }, + { + "epoch": 0.48244, + "grad_norm": 1.4832745179481208, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 48244 + }, + { + "epoch": 0.48245, + "grad_norm": 1.4963838497906519, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 48245 + }, + { + "epoch": 0.48246, + "grad_norm": 1.3397137122480836, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48246 + }, + { + "epoch": 0.48247, + "grad_norm": 1.3977295655905617, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48247 + }, + { + "epoch": 0.48248, + "grad_norm": 1.4953480272428683, + "learning_rate": 0.003, + "loss": 4.062, + "step": 48248 + }, + { + "epoch": 0.48249, + "grad_norm": 1.5908435801862078, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 48249 + }, + { + "epoch": 0.4825, + "grad_norm": 1.8530932950853327, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 48250 + }, + { + "epoch": 0.48251, + "grad_norm": 1.7010213530007365, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 48251 + }, + { + "epoch": 0.48252, + "grad_norm": 1.6575907276984438, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 48252 + }, + { + "epoch": 0.48253, + "grad_norm": 1.5823479572235166, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 48253 + }, + { + "epoch": 0.48254, + "grad_norm": 1.7329498277583186, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48254 + }, + { + "epoch": 0.48255, + "grad_norm": 1.4865203276112324, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 48255 + }, + { + "epoch": 0.48256, + "grad_norm": 1.2638435758982434, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48256 + }, + { + "epoch": 0.48257, + "grad_norm": 1.72638995313187, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 48257 + }, + { + "epoch": 0.48258, + "grad_norm": 1.347019572314457, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 48258 + }, + { + "epoch": 0.48259, + "grad_norm": 1.6776999084916686, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 48259 + }, + { + "epoch": 0.4826, + "grad_norm": 1.8365440263621555, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48260 + }, + { + "epoch": 0.48261, + "grad_norm": 1.3754393956018889, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 48261 + }, + { + "epoch": 0.48262, + "grad_norm": 1.3004968292564472, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 48262 + }, + { + "epoch": 0.48263, + "grad_norm": 1.8128538873081461, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 48263 + }, + { + "epoch": 0.48264, + "grad_norm": 1.2589828979454827, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 48264 + }, + { + "epoch": 0.48265, + "grad_norm": 1.6055226238368476, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48265 + }, + { + "epoch": 0.48266, + "grad_norm": 1.400110207884125, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 48266 + }, + { + "epoch": 0.48267, + "grad_norm": 1.7066621663539114, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 48267 + }, + { + "epoch": 0.48268, + "grad_norm": 1.460309455457655, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48268 + }, + { + "epoch": 0.48269, + "grad_norm": 1.3954558580704848, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 48269 + }, + { + "epoch": 0.4827, + "grad_norm": 1.7098529874868313, + "learning_rate": 0.003, + "loss": 4.007, + "step": 48270 + }, + { + "epoch": 0.48271, + "grad_norm": 1.5078454670578259, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 48271 + }, + { + "epoch": 0.48272, + "grad_norm": 1.5995161554760082, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 48272 + }, + { + "epoch": 0.48273, + "grad_norm": 1.5169047577872041, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 48273 + }, + { + "epoch": 0.48274, + "grad_norm": 1.7348433797286555, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 48274 + }, + { + "epoch": 0.48275, + "grad_norm": 1.7732331780213866, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48275 + }, + { + "epoch": 0.48276, + "grad_norm": 1.332794400138887, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48276 + }, + { + "epoch": 0.48277, + "grad_norm": 2.0772368030203516, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48277 + }, + { + "epoch": 0.48278, + "grad_norm": 1.1122800087884102, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 48278 + }, + { + "epoch": 0.48279, + "grad_norm": 1.619257488316688, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48279 + }, + { + "epoch": 0.4828, + "grad_norm": 1.543554635114241, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48280 + }, + { + "epoch": 0.48281, + "grad_norm": 1.6602850080496614, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48281 + }, + { + "epoch": 0.48282, + "grad_norm": 1.8569271046249383, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 48282 + }, + { + "epoch": 0.48283, + "grad_norm": 1.4003621455979944, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 48283 + }, + { + "epoch": 0.48284, + "grad_norm": 1.7629882914732646, + "learning_rate": 0.003, + "loss": 4.007, + "step": 48284 + }, + { + "epoch": 0.48285, + "grad_norm": 1.288077111282725, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 48285 + }, + { + "epoch": 0.48286, + "grad_norm": 1.7186022983853293, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 48286 + }, + { + "epoch": 0.48287, + "grad_norm": 1.1233238884096077, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 48287 + }, + { + "epoch": 0.48288, + "grad_norm": 1.584007181341291, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48288 + }, + { + "epoch": 0.48289, + "grad_norm": 1.3993817507532047, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 48289 + }, + { + "epoch": 0.4829, + "grad_norm": 1.8312591444794701, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 48290 + }, + { + "epoch": 0.48291, + "grad_norm": 1.696830282288681, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 48291 + }, + { + "epoch": 0.48292, + "grad_norm": 1.9155336916565213, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48292 + }, + { + "epoch": 0.48293, + "grad_norm": 1.2482195639388718, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 48293 + }, + { + "epoch": 0.48294, + "grad_norm": 1.6432277504624702, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 48294 + }, + { + "epoch": 0.48295, + "grad_norm": 1.4271439177881056, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 48295 + }, + { + "epoch": 0.48296, + "grad_norm": 1.8421576428617017, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 48296 + }, + { + "epoch": 0.48297, + "grad_norm": 1.2919575519319972, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 48297 + }, + { + "epoch": 0.48298, + "grad_norm": 1.6322597507747783, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 48298 + }, + { + "epoch": 0.48299, + "grad_norm": 1.4569174559694116, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 48299 + }, + { + "epoch": 0.483, + "grad_norm": 1.6125707243118133, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 48300 + }, + { + "epoch": 0.48301, + "grad_norm": 1.5135878544757837, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 48301 + }, + { + "epoch": 0.48302, + "grad_norm": 1.437728508434514, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48302 + }, + { + "epoch": 0.48303, + "grad_norm": 1.4823364096092928, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 48303 + }, + { + "epoch": 0.48304, + "grad_norm": 1.789104848780784, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 48304 + }, + { + "epoch": 0.48305, + "grad_norm": 1.592381242325214, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 48305 + }, + { + "epoch": 0.48306, + "grad_norm": 1.8325496629955451, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 48306 + }, + { + "epoch": 0.48307, + "grad_norm": 1.528485901798695, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48307 + }, + { + "epoch": 0.48308, + "grad_norm": 1.490129460176433, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 48308 + }, + { + "epoch": 0.48309, + "grad_norm": 1.5286424828020464, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 48309 + }, + { + "epoch": 0.4831, + "grad_norm": 1.2098914812147814, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 48310 + }, + { + "epoch": 0.48311, + "grad_norm": 1.9360466158221536, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 48311 + }, + { + "epoch": 0.48312, + "grad_norm": 1.1264764202453865, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 48312 + }, + { + "epoch": 0.48313, + "grad_norm": 1.5679424978284044, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48313 + }, + { + "epoch": 0.48314, + "grad_norm": 1.5504631527647486, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 48314 + }, + { + "epoch": 0.48315, + "grad_norm": 1.395702599420995, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 48315 + }, + { + "epoch": 0.48316, + "grad_norm": 1.5884508682524194, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 48316 + }, + { + "epoch": 0.48317, + "grad_norm": 1.8384733430241351, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 48317 + }, + { + "epoch": 0.48318, + "grad_norm": 1.351789771739365, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48318 + }, + { + "epoch": 0.48319, + "grad_norm": 1.927923464591859, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 48319 + }, + { + "epoch": 0.4832, + "grad_norm": 1.6103068181162117, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 48320 + }, + { + "epoch": 0.48321, + "grad_norm": 1.5832962072979049, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 48321 + }, + { + "epoch": 0.48322, + "grad_norm": 1.276312043911831, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 48322 + }, + { + "epoch": 0.48323, + "grad_norm": 1.8276958524887619, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 48323 + }, + { + "epoch": 0.48324, + "grad_norm": 1.4221727623791485, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 48324 + }, + { + "epoch": 0.48325, + "grad_norm": 1.663315408506893, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 48325 + }, + { + "epoch": 0.48326, + "grad_norm": 1.8441734594409562, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 48326 + }, + { + "epoch": 0.48327, + "grad_norm": 1.1045122075300093, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48327 + }, + { + "epoch": 0.48328, + "grad_norm": 1.5776524977344957, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48328 + }, + { + "epoch": 0.48329, + "grad_norm": 1.3780801743012385, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 48329 + }, + { + "epoch": 0.4833, + "grad_norm": 1.767636342700922, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 48330 + }, + { + "epoch": 0.48331, + "grad_norm": 1.4657236337587423, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 48331 + }, + { + "epoch": 0.48332, + "grad_norm": 1.713832542980404, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 48332 + }, + { + "epoch": 0.48333, + "grad_norm": 1.346324253159675, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 48333 + }, + { + "epoch": 0.48334, + "grad_norm": 1.7836298069517387, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 48334 + }, + { + "epoch": 0.48335, + "grad_norm": 1.4889948535219701, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 48335 + }, + { + "epoch": 0.48336, + "grad_norm": 1.540819113426441, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 48336 + }, + { + "epoch": 0.48337, + "grad_norm": 1.6796299097726375, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 48337 + }, + { + "epoch": 0.48338, + "grad_norm": 1.4169531357234046, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 48338 + }, + { + "epoch": 0.48339, + "grad_norm": 1.765679522733508, + "learning_rate": 0.003, + "loss": 4.025, + "step": 48339 + }, + { + "epoch": 0.4834, + "grad_norm": 1.429797589113389, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 48340 + }, + { + "epoch": 0.48341, + "grad_norm": 1.516598667557201, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48341 + }, + { + "epoch": 0.48342, + "grad_norm": 1.5322497125073447, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48342 + }, + { + "epoch": 0.48343, + "grad_norm": 1.276340305503889, + "learning_rate": 0.003, + "loss": 4.009, + "step": 48343 + }, + { + "epoch": 0.48344, + "grad_norm": 1.6465477027236084, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 48344 + }, + { + "epoch": 0.48345, + "grad_norm": 1.3442565259121397, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 48345 + }, + { + "epoch": 0.48346, + "grad_norm": 1.5886859590852287, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 48346 + }, + { + "epoch": 0.48347, + "grad_norm": 1.4330871641973935, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48347 + }, + { + "epoch": 0.48348, + "grad_norm": 1.5428648917196803, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 48348 + }, + { + "epoch": 0.48349, + "grad_norm": 1.4780058206776263, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48349 + }, + { + "epoch": 0.4835, + "grad_norm": 1.5511711581084717, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 48350 + }, + { + "epoch": 0.48351, + "grad_norm": 1.5734032809863896, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 48351 + }, + { + "epoch": 0.48352, + "grad_norm": 1.5588129854495165, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48352 + }, + { + "epoch": 0.48353, + "grad_norm": 1.3148762001051781, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 48353 + }, + { + "epoch": 0.48354, + "grad_norm": 1.6513362826871787, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 48354 + }, + { + "epoch": 0.48355, + "grad_norm": 1.5440201669022129, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 48355 + }, + { + "epoch": 0.48356, + "grad_norm": 1.5595746862789168, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 48356 + }, + { + "epoch": 0.48357, + "grad_norm": 1.8041725009360632, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 48357 + }, + { + "epoch": 0.48358, + "grad_norm": 1.6196087057128903, + "learning_rate": 0.003, + "loss": 4.06, + "step": 48358 + }, + { + "epoch": 0.48359, + "grad_norm": 1.9350605332447446, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 48359 + }, + { + "epoch": 0.4836, + "grad_norm": 1.4395474443222531, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 48360 + }, + { + "epoch": 0.48361, + "grad_norm": 1.7558757021579465, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48361 + }, + { + "epoch": 0.48362, + "grad_norm": 1.5362333335864744, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 48362 + }, + { + "epoch": 0.48363, + "grad_norm": 1.8586926578612304, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 48363 + }, + { + "epoch": 0.48364, + "grad_norm": 1.5402678493526405, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 48364 + }, + { + "epoch": 0.48365, + "grad_norm": 1.4931551284298599, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48365 + }, + { + "epoch": 0.48366, + "grad_norm": 1.5038943766703572, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48366 + }, + { + "epoch": 0.48367, + "grad_norm": 1.4289064989258076, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 48367 + }, + { + "epoch": 0.48368, + "grad_norm": 1.8170678144798083, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 48368 + }, + { + "epoch": 0.48369, + "grad_norm": 1.1315151751706436, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 48369 + }, + { + "epoch": 0.4837, + "grad_norm": 1.6584982597268247, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 48370 + }, + { + "epoch": 0.48371, + "grad_norm": 1.2208232917873818, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 48371 + }, + { + "epoch": 0.48372, + "grad_norm": 1.850631464236691, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 48372 + }, + { + "epoch": 0.48373, + "grad_norm": 1.2990256136748088, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 48373 + }, + { + "epoch": 0.48374, + "grad_norm": 1.4579165633982238, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 48374 + }, + { + "epoch": 0.48375, + "grad_norm": 1.6126702956717174, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48375 + }, + { + "epoch": 0.48376, + "grad_norm": 1.3815679200423605, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 48376 + }, + { + "epoch": 0.48377, + "grad_norm": 1.51025412275154, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48377 + }, + { + "epoch": 0.48378, + "grad_norm": 1.82496343985361, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 48378 + }, + { + "epoch": 0.48379, + "grad_norm": 1.4393878218967422, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 48379 + }, + { + "epoch": 0.4838, + "grad_norm": 2.061731887068247, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 48380 + }, + { + "epoch": 0.48381, + "grad_norm": 1.2320048755362107, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 48381 + }, + { + "epoch": 0.48382, + "grad_norm": 1.6927324365720033, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 48382 + }, + { + "epoch": 0.48383, + "grad_norm": 1.6477572035403245, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 48383 + }, + { + "epoch": 0.48384, + "grad_norm": 1.3430135354671016, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48384 + }, + { + "epoch": 0.48385, + "grad_norm": 1.4238029364638036, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 48385 + }, + { + "epoch": 0.48386, + "grad_norm": 1.460763154420063, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 48386 + }, + { + "epoch": 0.48387, + "grad_norm": 1.5343587142272224, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 48387 + }, + { + "epoch": 0.48388, + "grad_norm": 1.3966374624046594, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 48388 + }, + { + "epoch": 0.48389, + "grad_norm": 1.7263563933781712, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 48389 + }, + { + "epoch": 0.4839, + "grad_norm": 1.309114564177188, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 48390 + }, + { + "epoch": 0.48391, + "grad_norm": 1.6298001306531547, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 48391 + }, + { + "epoch": 0.48392, + "grad_norm": 1.4690953120389283, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48392 + }, + { + "epoch": 0.48393, + "grad_norm": 1.510130261272311, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 48393 + }, + { + "epoch": 0.48394, + "grad_norm": 1.682797159156134, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 48394 + }, + { + "epoch": 0.48395, + "grad_norm": 1.5088586941484718, + "learning_rate": 0.003, + "loss": 4.045, + "step": 48395 + }, + { + "epoch": 0.48396, + "grad_norm": 1.9455415295631397, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 48396 + }, + { + "epoch": 0.48397, + "grad_norm": 1.424579647685922, + "learning_rate": 0.003, + "loss": 4.068, + "step": 48397 + }, + { + "epoch": 0.48398, + "grad_norm": 1.6643944045056598, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 48398 + }, + { + "epoch": 0.48399, + "grad_norm": 1.7491566792513835, + "learning_rate": 0.003, + "loss": 4.04, + "step": 48399 + }, + { + "epoch": 0.484, + "grad_norm": 1.6447337805802107, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 48400 + }, + { + "epoch": 0.48401, + "grad_norm": 1.4543621339064305, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 48401 + }, + { + "epoch": 0.48402, + "grad_norm": 1.5953244369581396, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 48402 + }, + { + "epoch": 0.48403, + "grad_norm": 1.5219045677468785, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48403 + }, + { + "epoch": 0.48404, + "grad_norm": 1.5455900661166349, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 48404 + }, + { + "epoch": 0.48405, + "grad_norm": 1.6253257114065578, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 48405 + }, + { + "epoch": 0.48406, + "grad_norm": 1.8574028908180855, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 48406 + }, + { + "epoch": 0.48407, + "grad_norm": 1.281082260255922, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 48407 + }, + { + "epoch": 0.48408, + "grad_norm": 1.679875646220918, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 48408 + }, + { + "epoch": 0.48409, + "grad_norm": 1.4057692962485897, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48409 + }, + { + "epoch": 0.4841, + "grad_norm": 1.652774290171629, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 48410 + }, + { + "epoch": 0.48411, + "grad_norm": 1.272337502311311, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48411 + }, + { + "epoch": 0.48412, + "grad_norm": 1.6116029440839499, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 48412 + }, + { + "epoch": 0.48413, + "grad_norm": 1.5167470134508851, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 48413 + }, + { + "epoch": 0.48414, + "grad_norm": 1.3379354068747518, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 48414 + }, + { + "epoch": 0.48415, + "grad_norm": 1.7149391152190419, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48415 + }, + { + "epoch": 0.48416, + "grad_norm": 1.3886808209429522, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 48416 + }, + { + "epoch": 0.48417, + "grad_norm": 1.671345505112948, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 48417 + }, + { + "epoch": 0.48418, + "grad_norm": 1.5511733178192963, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48418 + }, + { + "epoch": 0.48419, + "grad_norm": 1.8334127982766086, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 48419 + }, + { + "epoch": 0.4842, + "grad_norm": 1.6976919228373484, + "learning_rate": 0.003, + "loss": 4.044, + "step": 48420 + }, + { + "epoch": 0.48421, + "grad_norm": 1.2902511675644437, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48421 + }, + { + "epoch": 0.48422, + "grad_norm": 1.445529442860877, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 48422 + }, + { + "epoch": 0.48423, + "grad_norm": 1.6760013609821438, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 48423 + }, + { + "epoch": 0.48424, + "grad_norm": 1.6313037425439758, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 48424 + }, + { + "epoch": 0.48425, + "grad_norm": 1.7881623797994184, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 48425 + }, + { + "epoch": 0.48426, + "grad_norm": 1.2318407657180432, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 48426 + }, + { + "epoch": 0.48427, + "grad_norm": 1.6883818003003443, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48427 + }, + { + "epoch": 0.48428, + "grad_norm": 1.1959742266590927, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 48428 + }, + { + "epoch": 0.48429, + "grad_norm": 1.935752924628117, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 48429 + }, + { + "epoch": 0.4843, + "grad_norm": 1.3693276731912856, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 48430 + }, + { + "epoch": 0.48431, + "grad_norm": 1.6115341299102153, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48431 + }, + { + "epoch": 0.48432, + "grad_norm": 1.4591849029083068, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 48432 + }, + { + "epoch": 0.48433, + "grad_norm": 1.5666624090621573, + "learning_rate": 0.003, + "loss": 4.051, + "step": 48433 + }, + { + "epoch": 0.48434, + "grad_norm": 1.276535597759515, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 48434 + }, + { + "epoch": 0.48435, + "grad_norm": 1.727056124152773, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48435 + }, + { + "epoch": 0.48436, + "grad_norm": 1.5582545763642037, + "learning_rate": 0.003, + "loss": 4.034, + "step": 48436 + }, + { + "epoch": 0.48437, + "grad_norm": 1.5586300543290719, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48437 + }, + { + "epoch": 0.48438, + "grad_norm": 1.612369005478153, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 48438 + }, + { + "epoch": 0.48439, + "grad_norm": 1.8080279546072133, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 48439 + }, + { + "epoch": 0.4844, + "grad_norm": 1.3357459498054671, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 48440 + }, + { + "epoch": 0.48441, + "grad_norm": 1.7379144282118248, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 48441 + }, + { + "epoch": 0.48442, + "grad_norm": 1.1321789652431196, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 48442 + }, + { + "epoch": 0.48443, + "grad_norm": 1.5498010760209666, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 48443 + }, + { + "epoch": 0.48444, + "grad_norm": 1.4834240121952273, + "learning_rate": 0.003, + "loss": 4.057, + "step": 48444 + }, + { + "epoch": 0.48445, + "grad_norm": 1.700280483559068, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48445 + }, + { + "epoch": 0.48446, + "grad_norm": 1.6262866379678536, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 48446 + }, + { + "epoch": 0.48447, + "grad_norm": 1.6697237334888961, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 48447 + }, + { + "epoch": 0.48448, + "grad_norm": 1.566099193998736, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48448 + }, + { + "epoch": 0.48449, + "grad_norm": 1.4731814980709999, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 48449 + }, + { + "epoch": 0.4845, + "grad_norm": 1.453448259890368, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 48450 + }, + { + "epoch": 0.48451, + "grad_norm": 1.5330635626902975, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 48451 + }, + { + "epoch": 0.48452, + "grad_norm": 1.5081617219820773, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48452 + }, + { + "epoch": 0.48453, + "grad_norm": 1.4323706632349944, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 48453 + }, + { + "epoch": 0.48454, + "grad_norm": 1.4014886063816194, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48454 + }, + { + "epoch": 0.48455, + "grad_norm": 1.5883604613434181, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 48455 + }, + { + "epoch": 0.48456, + "grad_norm": 1.4680646207893058, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 48456 + }, + { + "epoch": 0.48457, + "grad_norm": 1.5394613102944932, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48457 + }, + { + "epoch": 0.48458, + "grad_norm": 1.3029771076069663, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 48458 + }, + { + "epoch": 0.48459, + "grad_norm": 1.5655771630670876, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 48459 + }, + { + "epoch": 0.4846, + "grad_norm": 1.4706439257749102, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 48460 + }, + { + "epoch": 0.48461, + "grad_norm": 1.6558418869584588, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 48461 + }, + { + "epoch": 0.48462, + "grad_norm": 1.674839291868912, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 48462 + }, + { + "epoch": 0.48463, + "grad_norm": 1.3864856531054155, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 48463 + }, + { + "epoch": 0.48464, + "grad_norm": 2.107884377206121, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 48464 + }, + { + "epoch": 0.48465, + "grad_norm": 1.3719635314810423, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48465 + }, + { + "epoch": 0.48466, + "grad_norm": 1.5597127711279384, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 48466 + }, + { + "epoch": 0.48467, + "grad_norm": 1.8055101416795174, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 48467 + }, + { + "epoch": 0.48468, + "grad_norm": 1.8966861953925367, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 48468 + }, + { + "epoch": 0.48469, + "grad_norm": 1.3001365997750556, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 48469 + }, + { + "epoch": 0.4847, + "grad_norm": 1.5164222204776825, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 48470 + }, + { + "epoch": 0.48471, + "grad_norm": 1.5578735665984658, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48471 + }, + { + "epoch": 0.48472, + "grad_norm": 1.6964321003116352, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 48472 + }, + { + "epoch": 0.48473, + "grad_norm": 1.6315046206677912, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 48473 + }, + { + "epoch": 0.48474, + "grad_norm": 1.7006424429489517, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 48474 + }, + { + "epoch": 0.48475, + "grad_norm": 1.3880610901001944, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 48475 + }, + { + "epoch": 0.48476, + "grad_norm": 1.5163195850849398, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 48476 + }, + { + "epoch": 0.48477, + "grad_norm": 1.6243126716923848, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 48477 + }, + { + "epoch": 0.48478, + "grad_norm": 1.3527359618699877, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 48478 + }, + { + "epoch": 0.48479, + "grad_norm": 1.5772646591191406, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48479 + }, + { + "epoch": 0.4848, + "grad_norm": 1.6027189191747082, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 48480 + }, + { + "epoch": 0.48481, + "grad_norm": 1.5698954061273604, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48481 + }, + { + "epoch": 0.48482, + "grad_norm": 1.6642394523862245, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 48482 + }, + { + "epoch": 0.48483, + "grad_norm": 1.3183789198172793, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 48483 + }, + { + "epoch": 0.48484, + "grad_norm": 1.6169794004715996, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 48484 + }, + { + "epoch": 0.48485, + "grad_norm": 1.6257795218702216, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 48485 + }, + { + "epoch": 0.48486, + "grad_norm": 1.5252060029890748, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 48486 + }, + { + "epoch": 0.48487, + "grad_norm": 1.670952100991521, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 48487 + }, + { + "epoch": 0.48488, + "grad_norm": 1.57422093944799, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 48488 + }, + { + "epoch": 0.48489, + "grad_norm": 1.6653213937528346, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 48489 + }, + { + "epoch": 0.4849, + "grad_norm": 1.3449566457143955, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 48490 + }, + { + "epoch": 0.48491, + "grad_norm": 1.672774371816897, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 48491 + }, + { + "epoch": 0.48492, + "grad_norm": 1.228563513887837, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48492 + }, + { + "epoch": 0.48493, + "grad_norm": 2.0673694315536753, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 48493 + }, + { + "epoch": 0.48494, + "grad_norm": 1.4245481617094518, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 48494 + }, + { + "epoch": 0.48495, + "grad_norm": 1.7205222265357922, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 48495 + }, + { + "epoch": 0.48496, + "grad_norm": 1.659060117968723, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48496 + }, + { + "epoch": 0.48497, + "grad_norm": 1.4301281507164343, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 48497 + }, + { + "epoch": 0.48498, + "grad_norm": 1.5938999217576244, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 48498 + }, + { + "epoch": 0.48499, + "grad_norm": 1.376030499380406, + "learning_rate": 0.003, + "loss": 4.012, + "step": 48499 + }, + { + "epoch": 0.485, + "grad_norm": 1.4767623252836923, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 48500 + }, + { + "epoch": 0.48501, + "grad_norm": 1.4532450854944086, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48501 + }, + { + "epoch": 0.48502, + "grad_norm": 1.6318820435968056, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48502 + }, + { + "epoch": 0.48503, + "grad_norm": 1.4818220884764324, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 48503 + }, + { + "epoch": 0.48504, + "grad_norm": 1.5147938839477024, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 48504 + }, + { + "epoch": 0.48505, + "grad_norm": 1.7675849803741333, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 48505 + }, + { + "epoch": 0.48506, + "grad_norm": 1.5085647679859913, + "learning_rate": 0.003, + "loss": 4.049, + "step": 48506 + }, + { + "epoch": 0.48507, + "grad_norm": 1.6473384951744927, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 48507 + }, + { + "epoch": 0.48508, + "grad_norm": 1.6048378344436873, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 48508 + }, + { + "epoch": 0.48509, + "grad_norm": 1.9626253620971386, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 48509 + }, + { + "epoch": 0.4851, + "grad_norm": 1.499695398502873, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48510 + }, + { + "epoch": 0.48511, + "grad_norm": 1.6939486012876201, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 48511 + }, + { + "epoch": 0.48512, + "grad_norm": 1.166225095436748, + "learning_rate": 0.003, + "loss": 4.0, + "step": 48512 + }, + { + "epoch": 0.48513, + "grad_norm": 1.7687909962561703, + "learning_rate": 0.003, + "loss": 4.048, + "step": 48513 + }, + { + "epoch": 0.48514, + "grad_norm": 1.5779628251547746, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 48514 + }, + { + "epoch": 0.48515, + "grad_norm": 1.4884428408033512, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 48515 + }, + { + "epoch": 0.48516, + "grad_norm": 1.639361919916547, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 48516 + }, + { + "epoch": 0.48517, + "grad_norm": 1.4254490400720965, + "learning_rate": 0.003, + "loss": 4.03, + "step": 48517 + }, + { + "epoch": 0.48518, + "grad_norm": 1.6349092387543198, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 48518 + }, + { + "epoch": 0.48519, + "grad_norm": 1.610245284200389, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48519 + }, + { + "epoch": 0.4852, + "grad_norm": 1.4107788060114925, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48520 + }, + { + "epoch": 0.48521, + "grad_norm": 1.7951512519425152, + "learning_rate": 0.003, + "loss": 4.056, + "step": 48521 + }, + { + "epoch": 0.48522, + "grad_norm": 1.3674826459378016, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48522 + }, + { + "epoch": 0.48523, + "grad_norm": 1.7719088433794448, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48523 + }, + { + "epoch": 0.48524, + "grad_norm": 1.4649328525866976, + "learning_rate": 0.003, + "loss": 4.017, + "step": 48524 + }, + { + "epoch": 0.48525, + "grad_norm": 1.5778269172854245, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 48525 + }, + { + "epoch": 0.48526, + "grad_norm": 1.643771116701024, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 48526 + }, + { + "epoch": 0.48527, + "grad_norm": 1.8420682086651217, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 48527 + }, + { + "epoch": 0.48528, + "grad_norm": 1.5076956307388498, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 48528 + }, + { + "epoch": 0.48529, + "grad_norm": 1.626328531550049, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 48529 + }, + { + "epoch": 0.4853, + "grad_norm": 1.3662836248426486, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 48530 + }, + { + "epoch": 0.48531, + "grad_norm": 1.7621679146628546, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 48531 + }, + { + "epoch": 0.48532, + "grad_norm": 1.3519285233239544, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 48532 + }, + { + "epoch": 0.48533, + "grad_norm": 1.868499798926569, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 48533 + }, + { + "epoch": 0.48534, + "grad_norm": 1.445513128292472, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 48534 + }, + { + "epoch": 0.48535, + "grad_norm": 1.6274362352593177, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 48535 + }, + { + "epoch": 0.48536, + "grad_norm": 1.2550907808297844, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48536 + }, + { + "epoch": 0.48537, + "grad_norm": 1.6533444686300784, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 48537 + }, + { + "epoch": 0.48538, + "grad_norm": 1.5559958700440129, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 48538 + }, + { + "epoch": 0.48539, + "grad_norm": 1.416234835167772, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 48539 + }, + { + "epoch": 0.4854, + "grad_norm": 1.6065737006679024, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 48540 + }, + { + "epoch": 0.48541, + "grad_norm": 1.3265936757627503, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 48541 + }, + { + "epoch": 0.48542, + "grad_norm": 1.7145107884502355, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 48542 + }, + { + "epoch": 0.48543, + "grad_norm": 1.4514802282443482, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 48543 + }, + { + "epoch": 0.48544, + "grad_norm": 1.9003171095183027, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 48544 + }, + { + "epoch": 0.48545, + "grad_norm": 1.8643515963977275, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48545 + }, + { + "epoch": 0.48546, + "grad_norm": 1.443029213084826, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48546 + }, + { + "epoch": 0.48547, + "grad_norm": 1.888563427503356, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 48547 + }, + { + "epoch": 0.48548, + "grad_norm": 1.4701419535055928, + "learning_rate": 0.003, + "loss": 4.032, + "step": 48548 + }, + { + "epoch": 0.48549, + "grad_norm": 1.644946794982446, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 48549 + }, + { + "epoch": 0.4855, + "grad_norm": 1.7533101716679516, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 48550 + }, + { + "epoch": 0.48551, + "grad_norm": 1.4928517403196315, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 48551 + }, + { + "epoch": 0.48552, + "grad_norm": 1.5961996759561567, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 48552 + }, + { + "epoch": 0.48553, + "grad_norm": 1.4636092059062766, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48553 + }, + { + "epoch": 0.48554, + "grad_norm": 1.4909829549873814, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 48554 + }, + { + "epoch": 0.48555, + "grad_norm": 1.4300751344127156, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 48555 + }, + { + "epoch": 0.48556, + "grad_norm": 1.6566979873830054, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48556 + }, + { + "epoch": 0.48557, + "grad_norm": 1.3352570706524156, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 48557 + }, + { + "epoch": 0.48558, + "grad_norm": 1.4203170598194907, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48558 + }, + { + "epoch": 0.48559, + "grad_norm": 1.4426977843883864, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 48559 + }, + { + "epoch": 0.4856, + "grad_norm": 1.607654113390088, + "learning_rate": 0.003, + "loss": 4.03, + "step": 48560 + }, + { + "epoch": 0.48561, + "grad_norm": 1.489155945508265, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 48561 + }, + { + "epoch": 0.48562, + "grad_norm": 2.1101066504414305, + "learning_rate": 0.003, + "loss": 4.012, + "step": 48562 + }, + { + "epoch": 0.48563, + "grad_norm": 1.3259974295877297, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 48563 + }, + { + "epoch": 0.48564, + "grad_norm": 1.6209622793273466, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 48564 + }, + { + "epoch": 0.48565, + "grad_norm": 1.7178034567758944, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 48565 + }, + { + "epoch": 0.48566, + "grad_norm": 1.6956406496269287, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 48566 + }, + { + "epoch": 0.48567, + "grad_norm": 1.3300186236892442, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 48567 + }, + { + "epoch": 0.48568, + "grad_norm": 1.9558839238869146, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48568 + }, + { + "epoch": 0.48569, + "grad_norm": 1.8738994220249159, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 48569 + }, + { + "epoch": 0.4857, + "grad_norm": 1.4323145877256487, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 48570 + }, + { + "epoch": 0.48571, + "grad_norm": 1.6158469935227302, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48571 + }, + { + "epoch": 0.48572, + "grad_norm": 1.7705601869464302, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 48572 + }, + { + "epoch": 0.48573, + "grad_norm": 1.2812127667130295, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48573 + }, + { + "epoch": 0.48574, + "grad_norm": 1.5233652183075972, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 48574 + }, + { + "epoch": 0.48575, + "grad_norm": 1.6223383352805534, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48575 + }, + { + "epoch": 0.48576, + "grad_norm": 1.7292726044206328, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 48576 + }, + { + "epoch": 0.48577, + "grad_norm": 1.2938215931647323, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 48577 + }, + { + "epoch": 0.48578, + "grad_norm": 1.639355037853137, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 48578 + }, + { + "epoch": 0.48579, + "grad_norm": 1.4731930903893344, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 48579 + }, + { + "epoch": 0.4858, + "grad_norm": 1.6401144629272917, + "learning_rate": 0.003, + "loss": 4.042, + "step": 48580 + }, + { + "epoch": 0.48581, + "grad_norm": 1.388994167016146, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48581 + }, + { + "epoch": 0.48582, + "grad_norm": 1.3318393888824223, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 48582 + }, + { + "epoch": 0.48583, + "grad_norm": 1.5660825386211321, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 48583 + }, + { + "epoch": 0.48584, + "grad_norm": 1.1667906880440966, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48584 + }, + { + "epoch": 0.48585, + "grad_norm": 1.7598413779515965, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 48585 + }, + { + "epoch": 0.48586, + "grad_norm": 1.3376417416334676, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48586 + }, + { + "epoch": 0.48587, + "grad_norm": 1.58485334617405, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 48587 + }, + { + "epoch": 0.48588, + "grad_norm": 1.8377221057715347, + "learning_rate": 0.003, + "loss": 4.056, + "step": 48588 + }, + { + "epoch": 0.48589, + "grad_norm": 2.17406691897357, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 48589 + }, + { + "epoch": 0.4859, + "grad_norm": 1.4783910156170543, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 48590 + }, + { + "epoch": 0.48591, + "grad_norm": 1.4256320192423855, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 48591 + }, + { + "epoch": 0.48592, + "grad_norm": 1.8737484287117296, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 48592 + }, + { + "epoch": 0.48593, + "grad_norm": 1.5917874637086962, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 48593 + }, + { + "epoch": 0.48594, + "grad_norm": 1.4945638075861287, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 48594 + }, + { + "epoch": 0.48595, + "grad_norm": 1.771149460764388, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 48595 + }, + { + "epoch": 0.48596, + "grad_norm": 1.236853308199134, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48596 + }, + { + "epoch": 0.48597, + "grad_norm": 1.5711745208899095, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 48597 + }, + { + "epoch": 0.48598, + "grad_norm": 1.5003195511562546, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 48598 + }, + { + "epoch": 0.48599, + "grad_norm": 1.642076485510399, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 48599 + }, + { + "epoch": 0.486, + "grad_norm": 1.2501165309694404, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 48600 + }, + { + "epoch": 0.48601, + "grad_norm": 1.380931660777328, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 48601 + }, + { + "epoch": 0.48602, + "grad_norm": 1.4711934519888685, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48602 + }, + { + "epoch": 0.48603, + "grad_norm": 1.4824496691077238, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 48603 + }, + { + "epoch": 0.48604, + "grad_norm": 1.2395206713698694, + "learning_rate": 0.003, + "loss": 4.018, + "step": 48604 + }, + { + "epoch": 0.48605, + "grad_norm": 1.826046241663329, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 48605 + }, + { + "epoch": 0.48606, + "grad_norm": 1.4538732261998284, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 48606 + }, + { + "epoch": 0.48607, + "grad_norm": 2.1959012475031554, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 48607 + }, + { + "epoch": 0.48608, + "grad_norm": 1.440129600285276, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 48608 + }, + { + "epoch": 0.48609, + "grad_norm": 1.3527467852512562, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 48609 + }, + { + "epoch": 0.4861, + "grad_norm": 1.5924465225067554, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 48610 + }, + { + "epoch": 0.48611, + "grad_norm": 2.08245053722677, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 48611 + }, + { + "epoch": 0.48612, + "grad_norm": 1.4785955244613302, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48612 + }, + { + "epoch": 0.48613, + "grad_norm": 1.8761681084544575, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 48613 + }, + { + "epoch": 0.48614, + "grad_norm": 1.6184604624183134, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 48614 + }, + { + "epoch": 0.48615, + "grad_norm": 1.4275518790049961, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 48615 + }, + { + "epoch": 0.48616, + "grad_norm": 1.8342659692263905, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 48616 + }, + { + "epoch": 0.48617, + "grad_norm": 1.5255030134868168, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 48617 + }, + { + "epoch": 0.48618, + "grad_norm": 1.6047312155696911, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 48618 + }, + { + "epoch": 0.48619, + "grad_norm": 1.6017552074030899, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48619 + }, + { + "epoch": 0.4862, + "grad_norm": 1.552839622141142, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 48620 + }, + { + "epoch": 0.48621, + "grad_norm": 1.6791734593807393, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 48621 + }, + { + "epoch": 0.48622, + "grad_norm": 1.5497994213852009, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 48622 + }, + { + "epoch": 0.48623, + "grad_norm": 1.2620915277354412, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 48623 + }, + { + "epoch": 0.48624, + "grad_norm": 1.4771174843429833, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 48624 + }, + { + "epoch": 0.48625, + "grad_norm": 1.4835088207971874, + "learning_rate": 0.003, + "loss": 4.067, + "step": 48625 + }, + { + "epoch": 0.48626, + "grad_norm": 1.3353013869747652, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 48626 + }, + { + "epoch": 0.48627, + "grad_norm": 1.3815647456182916, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48627 + }, + { + "epoch": 0.48628, + "grad_norm": 1.6654581602722274, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 48628 + }, + { + "epoch": 0.48629, + "grad_norm": 1.7670471185814776, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 48629 + }, + { + "epoch": 0.4863, + "grad_norm": 1.3315252956544585, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48630 + }, + { + "epoch": 0.48631, + "grad_norm": 1.563657006814701, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48631 + }, + { + "epoch": 0.48632, + "grad_norm": 1.452035674000135, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 48632 + }, + { + "epoch": 0.48633, + "grad_norm": 1.4569149780235795, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48633 + }, + { + "epoch": 0.48634, + "grad_norm": 1.6661013470156563, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 48634 + }, + { + "epoch": 0.48635, + "grad_norm": 1.562891880046528, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 48635 + }, + { + "epoch": 0.48636, + "grad_norm": 1.5739872924751197, + "learning_rate": 0.003, + "loss": 4.013, + "step": 48636 + }, + { + "epoch": 0.48637, + "grad_norm": 1.3405450634133251, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 48637 + }, + { + "epoch": 0.48638, + "grad_norm": 1.976012265025116, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 48638 + }, + { + "epoch": 0.48639, + "grad_norm": 1.1883465548360448, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 48639 + }, + { + "epoch": 0.4864, + "grad_norm": 1.7899268541694906, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 48640 + }, + { + "epoch": 0.48641, + "grad_norm": 1.2868488797870345, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 48641 + }, + { + "epoch": 0.48642, + "grad_norm": 1.5468199762809722, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 48642 + }, + { + "epoch": 0.48643, + "grad_norm": 1.6766042304043423, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48643 + }, + { + "epoch": 0.48644, + "grad_norm": 1.3785969011799741, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 48644 + }, + { + "epoch": 0.48645, + "grad_norm": 1.9720770073225657, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 48645 + }, + { + "epoch": 0.48646, + "grad_norm": 1.4205707285257276, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48646 + }, + { + "epoch": 0.48647, + "grad_norm": 1.401666846775171, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48647 + }, + { + "epoch": 0.48648, + "grad_norm": 1.5219535791633114, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 48648 + }, + { + "epoch": 0.48649, + "grad_norm": 1.2949909896462395, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48649 + }, + { + "epoch": 0.4865, + "grad_norm": 1.6578197262601992, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48650 + }, + { + "epoch": 0.48651, + "grad_norm": 1.8313250446964766, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 48651 + }, + { + "epoch": 0.48652, + "grad_norm": 1.4739339075831317, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 48652 + }, + { + "epoch": 0.48653, + "grad_norm": 1.5445612785839047, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 48653 + }, + { + "epoch": 0.48654, + "grad_norm": 1.58803032072235, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 48654 + }, + { + "epoch": 0.48655, + "grad_norm": 1.7222155819997054, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 48655 + }, + { + "epoch": 0.48656, + "grad_norm": 1.3640809361708015, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 48656 + }, + { + "epoch": 0.48657, + "grad_norm": 1.5809318013459726, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 48657 + }, + { + "epoch": 0.48658, + "grad_norm": 1.5618082624796035, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 48658 + }, + { + "epoch": 0.48659, + "grad_norm": 1.1291701808247752, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 48659 + }, + { + "epoch": 0.4866, + "grad_norm": 1.7733425406784624, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 48660 + }, + { + "epoch": 0.48661, + "grad_norm": 1.485624474479765, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48661 + }, + { + "epoch": 0.48662, + "grad_norm": 1.9502249755357324, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 48662 + }, + { + "epoch": 0.48663, + "grad_norm": 1.9537022095881695, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 48663 + }, + { + "epoch": 0.48664, + "grad_norm": 1.722587266387424, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 48664 + }, + { + "epoch": 0.48665, + "grad_norm": 1.3211566772370342, + "learning_rate": 0.003, + "loss": 4.033, + "step": 48665 + }, + { + "epoch": 0.48666, + "grad_norm": 1.5691141877986532, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 48666 + }, + { + "epoch": 0.48667, + "grad_norm": 1.595055340897436, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48667 + }, + { + "epoch": 0.48668, + "grad_norm": 1.51761137951269, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48668 + }, + { + "epoch": 0.48669, + "grad_norm": 1.3722406429275387, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 48669 + }, + { + "epoch": 0.4867, + "grad_norm": 1.431809571750227, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 48670 + }, + { + "epoch": 0.48671, + "grad_norm": 1.611289207462273, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 48671 + }, + { + "epoch": 0.48672, + "grad_norm": 1.2634513316710834, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 48672 + }, + { + "epoch": 0.48673, + "grad_norm": 1.5677349426850657, + "learning_rate": 0.003, + "loss": 4.061, + "step": 48673 + }, + { + "epoch": 0.48674, + "grad_norm": 1.256597880218091, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48674 + }, + { + "epoch": 0.48675, + "grad_norm": 1.6320511105966535, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 48675 + }, + { + "epoch": 0.48676, + "grad_norm": 1.523962173071937, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 48676 + }, + { + "epoch": 0.48677, + "grad_norm": 1.7220160707643222, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 48677 + }, + { + "epoch": 0.48678, + "grad_norm": 1.845630916262879, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 48678 + }, + { + "epoch": 0.48679, + "grad_norm": 1.8910377705359906, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 48679 + }, + { + "epoch": 0.4868, + "grad_norm": 1.2775822192511868, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 48680 + }, + { + "epoch": 0.48681, + "grad_norm": 1.795117098101426, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48681 + }, + { + "epoch": 0.48682, + "grad_norm": 1.5482483484189886, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 48682 + }, + { + "epoch": 0.48683, + "grad_norm": 1.5762345462547476, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 48683 + }, + { + "epoch": 0.48684, + "grad_norm": 1.3903324230941858, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 48684 + }, + { + "epoch": 0.48685, + "grad_norm": 1.8441587312087135, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 48685 + }, + { + "epoch": 0.48686, + "grad_norm": 1.3904842434143834, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48686 + }, + { + "epoch": 0.48687, + "grad_norm": 1.6300467844122042, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 48687 + }, + { + "epoch": 0.48688, + "grad_norm": 1.4141051490557954, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 48688 + }, + { + "epoch": 0.48689, + "grad_norm": 1.6541060306858442, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 48689 + }, + { + "epoch": 0.4869, + "grad_norm": 1.550296758129631, + "learning_rate": 0.003, + "loss": 4.062, + "step": 48690 + }, + { + "epoch": 0.48691, + "grad_norm": 1.3520477369825818, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48691 + }, + { + "epoch": 0.48692, + "grad_norm": 1.464872630811987, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 48692 + }, + { + "epoch": 0.48693, + "grad_norm": 2.0274384950239086, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 48693 + }, + { + "epoch": 0.48694, + "grad_norm": 1.1616188510324803, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 48694 + }, + { + "epoch": 0.48695, + "grad_norm": 1.584242697751826, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48695 + }, + { + "epoch": 0.48696, + "grad_norm": 1.478668153146816, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 48696 + }, + { + "epoch": 0.48697, + "grad_norm": 1.7529758903349337, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 48697 + }, + { + "epoch": 0.48698, + "grad_norm": 1.647307981168728, + "learning_rate": 0.003, + "loss": 4.035, + "step": 48698 + }, + { + "epoch": 0.48699, + "grad_norm": 1.4534093611440362, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 48699 + }, + { + "epoch": 0.487, + "grad_norm": 1.8564743684482337, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 48700 + }, + { + "epoch": 0.48701, + "grad_norm": 1.2609732584328106, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 48701 + }, + { + "epoch": 0.48702, + "grad_norm": 1.6671362068066273, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 48702 + }, + { + "epoch": 0.48703, + "grad_norm": 1.4582914637127589, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48703 + }, + { + "epoch": 0.48704, + "grad_norm": 1.7603449855945932, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 48704 + }, + { + "epoch": 0.48705, + "grad_norm": 1.6503322623728485, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 48705 + }, + { + "epoch": 0.48706, + "grad_norm": 1.3955566864254798, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 48706 + }, + { + "epoch": 0.48707, + "grad_norm": 1.5463985338517, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48707 + }, + { + "epoch": 0.48708, + "grad_norm": 1.375482508681824, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48708 + }, + { + "epoch": 0.48709, + "grad_norm": 1.679892041068507, + "learning_rate": 0.003, + "loss": 4.049, + "step": 48709 + }, + { + "epoch": 0.4871, + "grad_norm": 1.5700676152289463, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 48710 + }, + { + "epoch": 0.48711, + "grad_norm": 1.630279028712662, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 48711 + }, + { + "epoch": 0.48712, + "grad_norm": 1.4352737103397504, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 48712 + }, + { + "epoch": 0.48713, + "grad_norm": 1.693656332827687, + "learning_rate": 0.003, + "loss": 4.043, + "step": 48713 + }, + { + "epoch": 0.48714, + "grad_norm": 1.322070786285864, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48714 + }, + { + "epoch": 0.48715, + "grad_norm": 1.5543168814299868, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 48715 + }, + { + "epoch": 0.48716, + "grad_norm": 1.6348737069783623, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48716 + }, + { + "epoch": 0.48717, + "grad_norm": 1.5925403213866198, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 48717 + }, + { + "epoch": 0.48718, + "grad_norm": 1.7915726222798511, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 48718 + }, + { + "epoch": 0.48719, + "grad_norm": 1.6523880794954844, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 48719 + }, + { + "epoch": 0.4872, + "grad_norm": 1.4034121668109487, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 48720 + }, + { + "epoch": 0.48721, + "grad_norm": 1.6466528475282487, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48721 + }, + { + "epoch": 0.48722, + "grad_norm": 1.4608301042938723, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 48722 + }, + { + "epoch": 0.48723, + "grad_norm": 1.7324356390661855, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 48723 + }, + { + "epoch": 0.48724, + "grad_norm": 1.5759096796434249, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 48724 + }, + { + "epoch": 0.48725, + "grad_norm": 1.8234673946004827, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48725 + }, + { + "epoch": 0.48726, + "grad_norm": 1.4767930761903532, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 48726 + }, + { + "epoch": 0.48727, + "grad_norm": 1.5701973149448132, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 48727 + }, + { + "epoch": 0.48728, + "grad_norm": 1.4972715923874809, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 48728 + }, + { + "epoch": 0.48729, + "grad_norm": 1.4900092439829409, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 48729 + }, + { + "epoch": 0.4873, + "grad_norm": 1.4443224698324397, + "learning_rate": 0.003, + "loss": 4.064, + "step": 48730 + }, + { + "epoch": 0.48731, + "grad_norm": 1.4741367300159223, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 48731 + }, + { + "epoch": 0.48732, + "grad_norm": 1.5014748864003278, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 48732 + }, + { + "epoch": 0.48733, + "grad_norm": 1.4452889770368549, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 48733 + }, + { + "epoch": 0.48734, + "grad_norm": 1.8000942614764674, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 48734 + }, + { + "epoch": 0.48735, + "grad_norm": 1.9662782824255154, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 48735 + }, + { + "epoch": 0.48736, + "grad_norm": 1.707947120978949, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 48736 + }, + { + "epoch": 0.48737, + "grad_norm": 1.383480592379212, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 48737 + }, + { + "epoch": 0.48738, + "grad_norm": 1.7688609519524732, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48738 + }, + { + "epoch": 0.48739, + "grad_norm": 1.6023872357931654, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 48739 + }, + { + "epoch": 0.4874, + "grad_norm": 1.667212911589863, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 48740 + }, + { + "epoch": 0.48741, + "grad_norm": 1.3488252587686425, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 48741 + }, + { + "epoch": 0.48742, + "grad_norm": 1.6981994655625288, + "learning_rate": 0.003, + "loss": 4.012, + "step": 48742 + }, + { + "epoch": 0.48743, + "grad_norm": 1.5569080285190917, + "learning_rate": 0.003, + "loss": 4.024, + "step": 48743 + }, + { + "epoch": 0.48744, + "grad_norm": 1.3933515337928668, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48744 + }, + { + "epoch": 0.48745, + "grad_norm": 1.439906636045918, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 48745 + }, + { + "epoch": 0.48746, + "grad_norm": 1.6171967961286322, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 48746 + }, + { + "epoch": 0.48747, + "grad_norm": 1.6506659295629487, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48747 + }, + { + "epoch": 0.48748, + "grad_norm": 1.5106692703872733, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48748 + }, + { + "epoch": 0.48749, + "grad_norm": 1.3316517846910985, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 48749 + }, + { + "epoch": 0.4875, + "grad_norm": 1.5724225426577667, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 48750 + }, + { + "epoch": 0.48751, + "grad_norm": 1.7123259679762437, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 48751 + }, + { + "epoch": 0.48752, + "grad_norm": 1.4966248913699205, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48752 + }, + { + "epoch": 0.48753, + "grad_norm": 1.6516279590325909, + "learning_rate": 0.003, + "loss": 3.993, + "step": 48753 + }, + { + "epoch": 0.48754, + "grad_norm": 1.2668823516532457, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 48754 + }, + { + "epoch": 0.48755, + "grad_norm": 1.4734710751271534, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 48755 + }, + { + "epoch": 0.48756, + "grad_norm": 1.3112713461065666, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 48756 + }, + { + "epoch": 0.48757, + "grad_norm": 1.6130390439046844, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 48757 + }, + { + "epoch": 0.48758, + "grad_norm": 1.9471597667292895, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48758 + }, + { + "epoch": 0.48759, + "grad_norm": 1.7364791477019013, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 48759 + }, + { + "epoch": 0.4876, + "grad_norm": 1.6885780081481556, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 48760 + }, + { + "epoch": 0.48761, + "grad_norm": 1.6577967421258093, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48761 + }, + { + "epoch": 0.48762, + "grad_norm": 1.6083925637878285, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 48762 + }, + { + "epoch": 0.48763, + "grad_norm": 1.605652437301427, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 48763 + }, + { + "epoch": 0.48764, + "grad_norm": 1.3323578274086452, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 48764 + }, + { + "epoch": 0.48765, + "grad_norm": 1.733427031176839, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 48765 + }, + { + "epoch": 0.48766, + "grad_norm": 1.199140654570113, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48766 + }, + { + "epoch": 0.48767, + "grad_norm": 2.1081866833346394, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48767 + }, + { + "epoch": 0.48768, + "grad_norm": 1.120610853862747, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48768 + }, + { + "epoch": 0.48769, + "grad_norm": 1.7748874494032538, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 48769 + }, + { + "epoch": 0.4877, + "grad_norm": 1.5675555637837952, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 48770 + }, + { + "epoch": 0.48771, + "grad_norm": 1.8407109899453036, + "learning_rate": 0.003, + "loss": 4.064, + "step": 48771 + }, + { + "epoch": 0.48772, + "grad_norm": 1.6411200639900156, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 48772 + }, + { + "epoch": 0.48773, + "grad_norm": 1.6309037941825864, + "learning_rate": 0.003, + "loss": 4.054, + "step": 48773 + }, + { + "epoch": 0.48774, + "grad_norm": 1.707148256012695, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 48774 + }, + { + "epoch": 0.48775, + "grad_norm": 1.363860097537893, + "learning_rate": 0.003, + "loss": 4.04, + "step": 48775 + }, + { + "epoch": 0.48776, + "grad_norm": 1.5430147889480437, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 48776 + }, + { + "epoch": 0.48777, + "grad_norm": 1.8996963070237312, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48777 + }, + { + "epoch": 0.48778, + "grad_norm": 1.220714049109817, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 48778 + }, + { + "epoch": 0.48779, + "grad_norm": 1.681075898487584, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 48779 + }, + { + "epoch": 0.4878, + "grad_norm": 1.4930443802115079, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 48780 + }, + { + "epoch": 0.48781, + "grad_norm": 1.569356306214604, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 48781 + }, + { + "epoch": 0.48782, + "grad_norm": 1.5747269590551012, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48782 + }, + { + "epoch": 0.48783, + "grad_norm": 1.6353603594209023, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 48783 + }, + { + "epoch": 0.48784, + "grad_norm": 1.9810076160368943, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 48784 + }, + { + "epoch": 0.48785, + "grad_norm": 1.3975852982957193, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 48785 + }, + { + "epoch": 0.48786, + "grad_norm": 1.574690959571606, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 48786 + }, + { + "epoch": 0.48787, + "grad_norm": 1.427114125977779, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 48787 + }, + { + "epoch": 0.48788, + "grad_norm": 1.9000636061841358, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 48788 + }, + { + "epoch": 0.48789, + "grad_norm": 1.1566938257431898, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 48789 + }, + { + "epoch": 0.4879, + "grad_norm": 1.8436272857495632, + "learning_rate": 0.003, + "loss": 4.04, + "step": 48790 + }, + { + "epoch": 0.48791, + "grad_norm": 1.3155385942479931, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48791 + }, + { + "epoch": 0.48792, + "grad_norm": 1.5014941208741424, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48792 + }, + { + "epoch": 0.48793, + "grad_norm": 1.6648839091557786, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 48793 + }, + { + "epoch": 0.48794, + "grad_norm": 1.7497976508725352, + "learning_rate": 0.003, + "loss": 4.046, + "step": 48794 + }, + { + "epoch": 0.48795, + "grad_norm": 1.3926410638886073, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 48795 + }, + { + "epoch": 0.48796, + "grad_norm": 1.709716779957177, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 48796 + }, + { + "epoch": 0.48797, + "grad_norm": 1.344980460318478, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 48797 + }, + { + "epoch": 0.48798, + "grad_norm": 1.8376577242671273, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 48798 + }, + { + "epoch": 0.48799, + "grad_norm": 1.3401998997872786, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48799 + }, + { + "epoch": 0.488, + "grad_norm": 1.4160218605910289, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 48800 + }, + { + "epoch": 0.48801, + "grad_norm": 1.6791803767851456, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 48801 + }, + { + "epoch": 0.48802, + "grad_norm": 1.5835884693853304, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48802 + }, + { + "epoch": 0.48803, + "grad_norm": 1.6699845124536987, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48803 + }, + { + "epoch": 0.48804, + "grad_norm": 1.318120178309963, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 48804 + }, + { + "epoch": 0.48805, + "grad_norm": 1.6677509820775247, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 48805 + }, + { + "epoch": 0.48806, + "grad_norm": 1.4648769837621733, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 48806 + }, + { + "epoch": 0.48807, + "grad_norm": 1.763962396884583, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 48807 + }, + { + "epoch": 0.48808, + "grad_norm": 1.4228819729239324, + "learning_rate": 0.003, + "loss": 4.026, + "step": 48808 + }, + { + "epoch": 0.48809, + "grad_norm": 1.5223060167827767, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 48809 + }, + { + "epoch": 0.4881, + "grad_norm": 1.4855775077374975, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 48810 + }, + { + "epoch": 0.48811, + "grad_norm": 1.3954986204386202, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 48811 + }, + { + "epoch": 0.48812, + "grad_norm": 1.6647055025975077, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 48812 + }, + { + "epoch": 0.48813, + "grad_norm": 1.7464475321258615, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 48813 + }, + { + "epoch": 0.48814, + "grad_norm": 1.8550400182681197, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 48814 + }, + { + "epoch": 0.48815, + "grad_norm": 1.473447157156734, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48815 + }, + { + "epoch": 0.48816, + "grad_norm": 1.9285733833072, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 48816 + }, + { + "epoch": 0.48817, + "grad_norm": 1.4201102348997503, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 48817 + }, + { + "epoch": 0.48818, + "grad_norm": 1.6673660203041074, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 48818 + }, + { + "epoch": 0.48819, + "grad_norm": 1.4655530006960504, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 48819 + }, + { + "epoch": 0.4882, + "grad_norm": 1.400117257859199, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 48820 + }, + { + "epoch": 0.48821, + "grad_norm": 1.9146374804927035, + "learning_rate": 0.003, + "loss": 4.051, + "step": 48821 + }, + { + "epoch": 0.48822, + "grad_norm": 1.291162657182402, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 48822 + }, + { + "epoch": 0.48823, + "grad_norm": 1.7710552599334994, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 48823 + }, + { + "epoch": 0.48824, + "grad_norm": 1.5652055051487657, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 48824 + }, + { + "epoch": 0.48825, + "grad_norm": 1.2923781242974983, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48825 + }, + { + "epoch": 0.48826, + "grad_norm": 1.542604562111779, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 48826 + }, + { + "epoch": 0.48827, + "grad_norm": 1.2289563438155502, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 48827 + }, + { + "epoch": 0.48828, + "grad_norm": 1.6220941716993256, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 48828 + }, + { + "epoch": 0.48829, + "grad_norm": 1.292698263232523, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 48829 + }, + { + "epoch": 0.4883, + "grad_norm": 1.6064778494751124, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48830 + }, + { + "epoch": 0.48831, + "grad_norm": 1.5519851215591578, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 48831 + }, + { + "epoch": 0.48832, + "grad_norm": 1.7560870672508662, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 48832 + }, + { + "epoch": 0.48833, + "grad_norm": 1.7888060687321088, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48833 + }, + { + "epoch": 0.48834, + "grad_norm": 1.4001795137772146, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 48834 + }, + { + "epoch": 0.48835, + "grad_norm": 1.7935154290885902, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 48835 + }, + { + "epoch": 0.48836, + "grad_norm": 1.4544408651520382, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 48836 + }, + { + "epoch": 0.48837, + "grad_norm": 1.5166782282885038, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 48837 + }, + { + "epoch": 0.48838, + "grad_norm": 2.0160674659177342, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 48838 + }, + { + "epoch": 0.48839, + "grad_norm": 1.3707734127535864, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 48839 + }, + { + "epoch": 0.4884, + "grad_norm": 1.5393626611699824, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 48840 + }, + { + "epoch": 0.48841, + "grad_norm": 1.618072423454966, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 48841 + }, + { + "epoch": 0.48842, + "grad_norm": 1.527953985325178, + "learning_rate": 0.003, + "loss": 4.039, + "step": 48842 + }, + { + "epoch": 0.48843, + "grad_norm": 1.4407508723025062, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 48843 + }, + { + "epoch": 0.48844, + "grad_norm": 1.9536806712280517, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48844 + }, + { + "epoch": 0.48845, + "grad_norm": 1.339599011371187, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 48845 + }, + { + "epoch": 0.48846, + "grad_norm": 1.7331568241090325, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 48846 + }, + { + "epoch": 0.48847, + "grad_norm": 1.4253476354945074, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48847 + }, + { + "epoch": 0.48848, + "grad_norm": 1.4870325327397549, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 48848 + }, + { + "epoch": 0.48849, + "grad_norm": 1.4364922282465542, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 48849 + }, + { + "epoch": 0.4885, + "grad_norm": 1.7064659745171527, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 48850 + }, + { + "epoch": 0.48851, + "grad_norm": 1.3837279118887957, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 48851 + }, + { + "epoch": 0.48852, + "grad_norm": 1.7531218924435399, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 48852 + }, + { + "epoch": 0.48853, + "grad_norm": 1.3397960217665164, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 48853 + }, + { + "epoch": 0.48854, + "grad_norm": 1.9780980312890077, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48854 + }, + { + "epoch": 0.48855, + "grad_norm": 1.5530136310742373, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 48855 + }, + { + "epoch": 0.48856, + "grad_norm": 1.5845996982667383, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 48856 + }, + { + "epoch": 0.48857, + "grad_norm": 1.5114916495708741, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 48857 + }, + { + "epoch": 0.48858, + "grad_norm": 1.598783043805714, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 48858 + }, + { + "epoch": 0.48859, + "grad_norm": 1.6211333494042375, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 48859 + }, + { + "epoch": 0.4886, + "grad_norm": 1.2761521711306376, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 48860 + }, + { + "epoch": 0.48861, + "grad_norm": 1.5742173989385724, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 48861 + }, + { + "epoch": 0.48862, + "grad_norm": 1.6102676002946072, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 48862 + }, + { + "epoch": 0.48863, + "grad_norm": 1.450207038065518, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 48863 + }, + { + "epoch": 0.48864, + "grad_norm": 1.4562361546244746, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 48864 + }, + { + "epoch": 0.48865, + "grad_norm": 1.4278902156360758, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 48865 + }, + { + "epoch": 0.48866, + "grad_norm": 1.6894408574971422, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48866 + }, + { + "epoch": 0.48867, + "grad_norm": 1.2599197506372894, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48867 + }, + { + "epoch": 0.48868, + "grad_norm": 1.8068565424391725, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 48868 + }, + { + "epoch": 0.48869, + "grad_norm": 1.1977260703558434, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48869 + }, + { + "epoch": 0.4887, + "grad_norm": 2.002771591771087, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48870 + }, + { + "epoch": 0.48871, + "grad_norm": 1.4078960026951506, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 48871 + }, + { + "epoch": 0.48872, + "grad_norm": 2.2417687703383335, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 48872 + }, + { + "epoch": 0.48873, + "grad_norm": 1.1898358028374676, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 48873 + }, + { + "epoch": 0.48874, + "grad_norm": 1.595525686511785, + "learning_rate": 0.003, + "loss": 4.038, + "step": 48874 + }, + { + "epoch": 0.48875, + "grad_norm": 1.7664183277106689, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 48875 + }, + { + "epoch": 0.48876, + "grad_norm": 1.7490615093460826, + "learning_rate": 0.003, + "loss": 4.045, + "step": 48876 + }, + { + "epoch": 0.48877, + "grad_norm": 1.5317434632515217, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48877 + }, + { + "epoch": 0.48878, + "grad_norm": 1.5403584791949474, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 48878 + }, + { + "epoch": 0.48879, + "grad_norm": 1.517246748552563, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 48879 + }, + { + "epoch": 0.4888, + "grad_norm": 1.5269708027924493, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 48880 + }, + { + "epoch": 0.48881, + "grad_norm": 1.360721265841086, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 48881 + }, + { + "epoch": 0.48882, + "grad_norm": 1.5712081661986732, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48882 + }, + { + "epoch": 0.48883, + "grad_norm": 1.5540781254077805, + "learning_rate": 0.003, + "loss": 4.045, + "step": 48883 + }, + { + "epoch": 0.48884, + "grad_norm": 1.5150345205386204, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 48884 + }, + { + "epoch": 0.48885, + "grad_norm": 1.3613303466478779, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 48885 + }, + { + "epoch": 0.48886, + "grad_norm": 1.8536320148278167, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48886 + }, + { + "epoch": 0.48887, + "grad_norm": 1.3837634243284616, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 48887 + }, + { + "epoch": 0.48888, + "grad_norm": 1.3945628568078283, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48888 + }, + { + "epoch": 0.48889, + "grad_norm": 1.6237352407205814, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 48889 + }, + { + "epoch": 0.4889, + "grad_norm": 1.6250065026992973, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 48890 + }, + { + "epoch": 0.48891, + "grad_norm": 1.6557169164022612, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 48891 + }, + { + "epoch": 0.48892, + "grad_norm": 1.615951366776691, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 48892 + }, + { + "epoch": 0.48893, + "grad_norm": 1.9608709726487605, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48893 + }, + { + "epoch": 0.48894, + "grad_norm": 1.0528795063074217, + "learning_rate": 0.003, + "loss": 4.051, + "step": 48894 + }, + { + "epoch": 0.48895, + "grad_norm": 1.990688159804021, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 48895 + }, + { + "epoch": 0.48896, + "grad_norm": 1.2780463024083069, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 48896 + }, + { + "epoch": 0.48897, + "grad_norm": 1.6858663586238498, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 48897 + }, + { + "epoch": 0.48898, + "grad_norm": 1.3720897277200954, + "learning_rate": 0.003, + "loss": 4.051, + "step": 48898 + }, + { + "epoch": 0.48899, + "grad_norm": 1.265074940827962, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 48899 + }, + { + "epoch": 0.489, + "grad_norm": 1.68914610382917, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 48900 + }, + { + "epoch": 0.48901, + "grad_norm": 1.8492361569967652, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 48901 + }, + { + "epoch": 0.48902, + "grad_norm": 1.5587430872052945, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 48902 + }, + { + "epoch": 0.48903, + "grad_norm": 1.7181626636045935, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 48903 + }, + { + "epoch": 0.48904, + "grad_norm": 1.4835239044913506, + "learning_rate": 0.003, + "loss": 4.06, + "step": 48904 + }, + { + "epoch": 0.48905, + "grad_norm": 1.579935604566847, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48905 + }, + { + "epoch": 0.48906, + "grad_norm": 1.5078314351452091, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 48906 + }, + { + "epoch": 0.48907, + "grad_norm": 1.4520439519758914, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 48907 + }, + { + "epoch": 0.48908, + "grad_norm": 1.6851476696839185, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 48908 + }, + { + "epoch": 0.48909, + "grad_norm": 1.5068948447884356, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 48909 + }, + { + "epoch": 0.4891, + "grad_norm": 1.5786004074246405, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 48910 + }, + { + "epoch": 0.48911, + "grad_norm": 1.4415278332765806, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48911 + }, + { + "epoch": 0.48912, + "grad_norm": 1.6627762755791131, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 48912 + }, + { + "epoch": 0.48913, + "grad_norm": 1.6452023249602785, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 48913 + }, + { + "epoch": 0.48914, + "grad_norm": 1.2978612925743445, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 48914 + }, + { + "epoch": 0.48915, + "grad_norm": 1.941046677378365, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 48915 + }, + { + "epoch": 0.48916, + "grad_norm": 1.642479092153361, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48916 + }, + { + "epoch": 0.48917, + "grad_norm": 1.418523346361653, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 48917 + }, + { + "epoch": 0.48918, + "grad_norm": 1.5951289423782185, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 48918 + }, + { + "epoch": 0.48919, + "grad_norm": 1.938977619767491, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 48919 + }, + { + "epoch": 0.4892, + "grad_norm": 1.6673567348678004, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 48920 + }, + { + "epoch": 0.48921, + "grad_norm": 1.2611708464948477, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 48921 + }, + { + "epoch": 0.48922, + "grad_norm": 1.5902233685577165, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 48922 + }, + { + "epoch": 0.48923, + "grad_norm": 1.4945686187956804, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 48923 + }, + { + "epoch": 0.48924, + "grad_norm": 1.461577628462515, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 48924 + }, + { + "epoch": 0.48925, + "grad_norm": 1.3489415773672473, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 48925 + }, + { + "epoch": 0.48926, + "grad_norm": 1.3818872710193444, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 48926 + }, + { + "epoch": 0.48927, + "grad_norm": 1.5108697324026334, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 48927 + }, + { + "epoch": 0.48928, + "grad_norm": 1.3575780734360525, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 48928 + }, + { + "epoch": 0.48929, + "grad_norm": 1.956422866531749, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 48929 + }, + { + "epoch": 0.4893, + "grad_norm": 1.9310877024228348, + "learning_rate": 0.003, + "loss": 4.025, + "step": 48930 + }, + { + "epoch": 0.48931, + "grad_norm": 1.2855618655142254, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 48931 + }, + { + "epoch": 0.48932, + "grad_norm": 1.6343201080700762, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 48932 + }, + { + "epoch": 0.48933, + "grad_norm": 1.6623382848348003, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 48933 + }, + { + "epoch": 0.48934, + "grad_norm": 1.554749148979006, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 48934 + }, + { + "epoch": 0.48935, + "grad_norm": 1.422219433478786, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 48935 + }, + { + "epoch": 0.48936, + "grad_norm": 1.6261873510778109, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48936 + }, + { + "epoch": 0.48937, + "grad_norm": 1.4407050695594608, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 48937 + }, + { + "epoch": 0.48938, + "grad_norm": 1.280824827268597, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 48938 + }, + { + "epoch": 0.48939, + "grad_norm": 1.9993350540320647, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 48939 + }, + { + "epoch": 0.4894, + "grad_norm": 1.7472391391972144, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 48940 + }, + { + "epoch": 0.48941, + "grad_norm": 1.6479271416303367, + "learning_rate": 0.003, + "loss": 4.023, + "step": 48941 + }, + { + "epoch": 0.48942, + "grad_norm": 1.454540725542685, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 48942 + }, + { + "epoch": 0.48943, + "grad_norm": 1.6123780202494689, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 48943 + }, + { + "epoch": 0.48944, + "grad_norm": 1.5620735681013271, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 48944 + }, + { + "epoch": 0.48945, + "grad_norm": 1.4838897053696236, + "learning_rate": 0.003, + "loss": 4.053, + "step": 48945 + }, + { + "epoch": 0.48946, + "grad_norm": 1.6739041468319458, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 48946 + }, + { + "epoch": 0.48947, + "grad_norm": 1.583757555385808, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48947 + }, + { + "epoch": 0.48948, + "grad_norm": 1.44411255084854, + "learning_rate": 0.003, + "loss": 4.03, + "step": 48948 + }, + { + "epoch": 0.48949, + "grad_norm": 1.6130846144044904, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 48949 + }, + { + "epoch": 0.4895, + "grad_norm": 1.7564141750487174, + "learning_rate": 0.003, + "loss": 4.032, + "step": 48950 + }, + { + "epoch": 0.48951, + "grad_norm": 1.2284739882579285, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 48951 + }, + { + "epoch": 0.48952, + "grad_norm": 1.8461747005647051, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 48952 + }, + { + "epoch": 0.48953, + "grad_norm": 1.3052192805437506, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 48953 + }, + { + "epoch": 0.48954, + "grad_norm": 1.776270760949016, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 48954 + }, + { + "epoch": 0.48955, + "grad_norm": 1.4319684417987888, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 48955 + }, + { + "epoch": 0.48956, + "grad_norm": 1.7645127930968019, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 48956 + }, + { + "epoch": 0.48957, + "grad_norm": 1.8375030416962383, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 48957 + }, + { + "epoch": 0.48958, + "grad_norm": 1.5222908385345715, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 48958 + }, + { + "epoch": 0.48959, + "grad_norm": 1.5899991939762237, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 48959 + }, + { + "epoch": 0.4896, + "grad_norm": 1.5120624532322606, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 48960 + }, + { + "epoch": 0.48961, + "grad_norm": 1.406128585640311, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48961 + }, + { + "epoch": 0.48962, + "grad_norm": 1.7362350340813664, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 48962 + }, + { + "epoch": 0.48963, + "grad_norm": 1.4449712260058134, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48963 + }, + { + "epoch": 0.48964, + "grad_norm": 1.5325429430781026, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 48964 + }, + { + "epoch": 0.48965, + "grad_norm": 1.4181903136961538, + "learning_rate": 0.003, + "loss": 4.041, + "step": 48965 + }, + { + "epoch": 0.48966, + "grad_norm": 1.512065195845404, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 48966 + }, + { + "epoch": 0.48967, + "grad_norm": 1.6113028335276987, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 48967 + }, + { + "epoch": 0.48968, + "grad_norm": 1.5998834091869303, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 48968 + }, + { + "epoch": 0.48969, + "grad_norm": 1.6437390071945723, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 48969 + }, + { + "epoch": 0.4897, + "grad_norm": 1.539485481718926, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 48970 + }, + { + "epoch": 0.48971, + "grad_norm": 1.6509143733095384, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 48971 + }, + { + "epoch": 0.48972, + "grad_norm": 1.3444417235152362, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 48972 + }, + { + "epoch": 0.48973, + "grad_norm": 1.709090040742334, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 48973 + }, + { + "epoch": 0.48974, + "grad_norm": 1.2293478188372449, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 48974 + }, + { + "epoch": 0.48975, + "grad_norm": 1.870917110708863, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 48975 + }, + { + "epoch": 0.48976, + "grad_norm": 1.4414229121301254, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 48976 + }, + { + "epoch": 0.48977, + "grad_norm": 1.5001077978483568, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 48977 + }, + { + "epoch": 0.48978, + "grad_norm": 1.2895790773305447, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 48978 + }, + { + "epoch": 0.48979, + "grad_norm": 1.6249611062097118, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 48979 + }, + { + "epoch": 0.4898, + "grad_norm": 1.4457041926998315, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 48980 + }, + { + "epoch": 0.48981, + "grad_norm": 1.5153115559717254, + "learning_rate": 0.003, + "loss": 4.026, + "step": 48981 + }, + { + "epoch": 0.48982, + "grad_norm": 1.5496853481206883, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 48982 + }, + { + "epoch": 0.48983, + "grad_norm": 1.9187819181948629, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 48983 + }, + { + "epoch": 0.48984, + "grad_norm": 1.3667054598140689, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 48984 + }, + { + "epoch": 0.48985, + "grad_norm": 1.7765158027288996, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 48985 + }, + { + "epoch": 0.48986, + "grad_norm": 1.3983354588189303, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 48986 + }, + { + "epoch": 0.48987, + "grad_norm": 1.6800238890530668, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 48987 + }, + { + "epoch": 0.48988, + "grad_norm": 1.3078215994738904, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 48988 + }, + { + "epoch": 0.48989, + "grad_norm": 1.7100538426814749, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 48989 + }, + { + "epoch": 0.4899, + "grad_norm": 1.420172699226674, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 48990 + }, + { + "epoch": 0.48991, + "grad_norm": 1.8099853605464815, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 48991 + }, + { + "epoch": 0.48992, + "grad_norm": 1.3310237238531608, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 48992 + }, + { + "epoch": 0.48993, + "grad_norm": 2.233502100518912, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 48993 + }, + { + "epoch": 0.48994, + "grad_norm": 1.4154847330554705, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 48994 + }, + { + "epoch": 0.48995, + "grad_norm": 1.529320195846271, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 48995 + }, + { + "epoch": 0.48996, + "grad_norm": 1.8602717278687249, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 48996 + }, + { + "epoch": 0.48997, + "grad_norm": 1.7732808614629694, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 48997 + }, + { + "epoch": 0.48998, + "grad_norm": 1.261222516640609, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 48998 + }, + { + "epoch": 0.48999, + "grad_norm": 1.8497059726016338, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 48999 + }, + { + "epoch": 0.49, + "grad_norm": 1.7321376252773957, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 49000 + }, + { + "epoch": 0.49001, + "grad_norm": 1.2952445786636235, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 49001 + }, + { + "epoch": 0.49002, + "grad_norm": 1.7737902753105688, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 49002 + }, + { + "epoch": 0.49003, + "grad_norm": 1.3969386802106687, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 49003 + }, + { + "epoch": 0.49004, + "grad_norm": 1.380712388748665, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 49004 + }, + { + "epoch": 0.49005, + "grad_norm": 1.7076814174437, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49005 + }, + { + "epoch": 0.49006, + "grad_norm": 1.4148037345780218, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 49006 + }, + { + "epoch": 0.49007, + "grad_norm": 1.5070110920155533, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 49007 + }, + { + "epoch": 0.49008, + "grad_norm": 1.4302834691306816, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49008 + }, + { + "epoch": 0.49009, + "grad_norm": 1.4736137124176336, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 49009 + }, + { + "epoch": 0.4901, + "grad_norm": 1.7369348483480747, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 49010 + }, + { + "epoch": 0.49011, + "grad_norm": 1.481893434342438, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49011 + }, + { + "epoch": 0.49012, + "grad_norm": 1.6852320342974092, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 49012 + }, + { + "epoch": 0.49013, + "grad_norm": 1.4941476791680952, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 49013 + }, + { + "epoch": 0.49014, + "grad_norm": 1.8661170419178845, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 49014 + }, + { + "epoch": 0.49015, + "grad_norm": 1.293215728453002, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 49015 + }, + { + "epoch": 0.49016, + "grad_norm": 1.7372074738686212, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 49016 + }, + { + "epoch": 0.49017, + "grad_norm": 1.4676777179330622, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 49017 + }, + { + "epoch": 0.49018, + "grad_norm": 1.8342707622230336, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49018 + }, + { + "epoch": 0.49019, + "grad_norm": 1.3714609023624724, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 49019 + }, + { + "epoch": 0.4902, + "grad_norm": 1.611564046407005, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49020 + }, + { + "epoch": 0.49021, + "grad_norm": 1.3230300772031491, + "learning_rate": 0.003, + "loss": 4.067, + "step": 49021 + }, + { + "epoch": 0.49022, + "grad_norm": 1.73188677485774, + "learning_rate": 0.003, + "loss": 4.053, + "step": 49022 + }, + { + "epoch": 0.49023, + "grad_norm": 1.3564248822365657, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 49023 + }, + { + "epoch": 0.49024, + "grad_norm": 1.8814455482060926, + "learning_rate": 0.003, + "loss": 4.026, + "step": 49024 + }, + { + "epoch": 0.49025, + "grad_norm": 1.5904268185047548, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 49025 + }, + { + "epoch": 0.49026, + "grad_norm": 1.606911530737511, + "learning_rate": 0.003, + "loss": 4.021, + "step": 49026 + }, + { + "epoch": 0.49027, + "grad_norm": 1.4141487356228761, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 49027 + }, + { + "epoch": 0.49028, + "grad_norm": 1.7858356784593845, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 49028 + }, + { + "epoch": 0.49029, + "grad_norm": 1.321235954548852, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49029 + }, + { + "epoch": 0.4903, + "grad_norm": 1.8206515013634377, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 49030 + }, + { + "epoch": 0.49031, + "grad_norm": 1.8629298210878336, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 49031 + }, + { + "epoch": 0.49032, + "grad_norm": 1.261567290092621, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 49032 + }, + { + "epoch": 0.49033, + "grad_norm": 1.6650892917753677, + "learning_rate": 0.003, + "loss": 4.047, + "step": 49033 + }, + { + "epoch": 0.49034, + "grad_norm": 1.5791602523863628, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49034 + }, + { + "epoch": 0.49035, + "grad_norm": 1.5778886142008397, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49035 + }, + { + "epoch": 0.49036, + "grad_norm": 1.296461074892008, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 49036 + }, + { + "epoch": 0.49037, + "grad_norm": 1.6951692382784622, + "learning_rate": 0.003, + "loss": 4.013, + "step": 49037 + }, + { + "epoch": 0.49038, + "grad_norm": 1.238360490520008, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 49038 + }, + { + "epoch": 0.49039, + "grad_norm": 1.5544862280535168, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 49039 + }, + { + "epoch": 0.4904, + "grad_norm": 1.4932330045816617, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 49040 + }, + { + "epoch": 0.49041, + "grad_norm": 1.5293384449502063, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 49041 + }, + { + "epoch": 0.49042, + "grad_norm": 1.8895240824867112, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 49042 + }, + { + "epoch": 0.49043, + "grad_norm": 1.5945790747736006, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49043 + }, + { + "epoch": 0.49044, + "grad_norm": 1.7107677968341888, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49044 + }, + { + "epoch": 0.49045, + "grad_norm": 1.3999195205784694, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 49045 + }, + { + "epoch": 0.49046, + "grad_norm": 1.7622826922591501, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 49046 + }, + { + "epoch": 0.49047, + "grad_norm": 1.4086707164731926, + "learning_rate": 0.003, + "loss": 4.059, + "step": 49047 + }, + { + "epoch": 0.49048, + "grad_norm": 1.58380717544523, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 49048 + }, + { + "epoch": 0.49049, + "grad_norm": 1.639730101214041, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 49049 + }, + { + "epoch": 0.4905, + "grad_norm": 1.4458360343755303, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49050 + }, + { + "epoch": 0.49051, + "grad_norm": 1.872952612727916, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 49051 + }, + { + "epoch": 0.49052, + "grad_norm": 1.224706116552156, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 49052 + }, + { + "epoch": 0.49053, + "grad_norm": 1.584359821481752, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 49053 + }, + { + "epoch": 0.49054, + "grad_norm": 1.2979977241131309, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 49054 + }, + { + "epoch": 0.49055, + "grad_norm": 1.970196823822167, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 49055 + }, + { + "epoch": 0.49056, + "grad_norm": 1.4682139131919851, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49056 + }, + { + "epoch": 0.49057, + "grad_norm": 1.9098103053575386, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 49057 + }, + { + "epoch": 0.49058, + "grad_norm": 1.372827437728919, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 49058 + }, + { + "epoch": 0.49059, + "grad_norm": 1.7273829655422865, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 49059 + }, + { + "epoch": 0.4906, + "grad_norm": 1.1657895544492354, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49060 + }, + { + "epoch": 0.49061, + "grad_norm": 1.4718113856037545, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49061 + }, + { + "epoch": 0.49062, + "grad_norm": 1.7848343816952998, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 49062 + }, + { + "epoch": 0.49063, + "grad_norm": 1.616132732953973, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49063 + }, + { + "epoch": 0.49064, + "grad_norm": 1.449731162514878, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49064 + }, + { + "epoch": 0.49065, + "grad_norm": 1.7653623683404183, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 49065 + }, + { + "epoch": 0.49066, + "grad_norm": 1.7172883361536666, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 49066 + }, + { + "epoch": 0.49067, + "grad_norm": 1.2985089995720884, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 49067 + }, + { + "epoch": 0.49068, + "grad_norm": 1.6740597508294965, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 49068 + }, + { + "epoch": 0.49069, + "grad_norm": 1.2615366866414373, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49069 + }, + { + "epoch": 0.4907, + "grad_norm": 1.6695885613078307, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 49070 + }, + { + "epoch": 0.49071, + "grad_norm": 1.3780332678335214, + "learning_rate": 0.003, + "loss": 4.016, + "step": 49071 + }, + { + "epoch": 0.49072, + "grad_norm": 1.754935424017554, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 49072 + }, + { + "epoch": 0.49073, + "grad_norm": 1.3802653897577826, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 49073 + }, + { + "epoch": 0.49074, + "grad_norm": 1.7934975662660553, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 49074 + }, + { + "epoch": 0.49075, + "grad_norm": 1.4391221272215684, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 49075 + }, + { + "epoch": 0.49076, + "grad_norm": 1.696559829552851, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 49076 + }, + { + "epoch": 0.49077, + "grad_norm": 1.3795911145387125, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 49077 + }, + { + "epoch": 0.49078, + "grad_norm": 1.529743897132354, + "learning_rate": 0.003, + "loss": 4.015, + "step": 49078 + }, + { + "epoch": 0.49079, + "grad_norm": 1.6003005984650367, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 49079 + }, + { + "epoch": 0.4908, + "grad_norm": 1.6087993270131131, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 49080 + }, + { + "epoch": 0.49081, + "grad_norm": 1.7437690794356564, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 49081 + }, + { + "epoch": 0.49082, + "grad_norm": 1.3821814385461244, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 49082 + }, + { + "epoch": 0.49083, + "grad_norm": 1.8882192399643511, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 49083 + }, + { + "epoch": 0.49084, + "grad_norm": 1.3785231933797695, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49084 + }, + { + "epoch": 0.49085, + "grad_norm": 1.7191298645987423, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49085 + }, + { + "epoch": 0.49086, + "grad_norm": 1.4380314388333502, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 49086 + }, + { + "epoch": 0.49087, + "grad_norm": 1.4881749381717766, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 49087 + }, + { + "epoch": 0.49088, + "grad_norm": 1.397798751803643, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 49088 + }, + { + "epoch": 0.49089, + "grad_norm": 1.3887151322956672, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49089 + }, + { + "epoch": 0.4909, + "grad_norm": 1.6435093135264696, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49090 + }, + { + "epoch": 0.49091, + "grad_norm": 1.330942231073544, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 49091 + }, + { + "epoch": 0.49092, + "grad_norm": 1.8085854167659512, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 49092 + }, + { + "epoch": 0.49093, + "grad_norm": 1.6031095440749363, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 49093 + }, + { + "epoch": 0.49094, + "grad_norm": 1.4252269497161563, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49094 + }, + { + "epoch": 0.49095, + "grad_norm": 2.022653286360439, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 49095 + }, + { + "epoch": 0.49096, + "grad_norm": 1.296777573072589, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 49096 + }, + { + "epoch": 0.49097, + "grad_norm": 1.6159403303466686, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 49097 + }, + { + "epoch": 0.49098, + "grad_norm": 1.8135566413772382, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 49098 + }, + { + "epoch": 0.49099, + "grad_norm": 1.6523355656233512, + "learning_rate": 0.003, + "loss": 4.052, + "step": 49099 + }, + { + "epoch": 0.491, + "grad_norm": 1.6142819200471006, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 49100 + }, + { + "epoch": 0.49101, + "grad_norm": 1.2198839457366781, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 49101 + }, + { + "epoch": 0.49102, + "grad_norm": 1.582489705459609, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 49102 + }, + { + "epoch": 0.49103, + "grad_norm": 1.4974382969649986, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 49103 + }, + { + "epoch": 0.49104, + "grad_norm": 1.569829916714157, + "learning_rate": 0.003, + "loss": 4.046, + "step": 49104 + }, + { + "epoch": 0.49105, + "grad_norm": 1.5896872032593707, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 49105 + }, + { + "epoch": 0.49106, + "grad_norm": 1.6325369029149763, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 49106 + }, + { + "epoch": 0.49107, + "grad_norm": 1.6287147546560137, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49107 + }, + { + "epoch": 0.49108, + "grad_norm": 1.623818954464772, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49108 + }, + { + "epoch": 0.49109, + "grad_norm": 1.4052160503719717, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49109 + }, + { + "epoch": 0.4911, + "grad_norm": 1.7204382577066915, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 49110 + }, + { + "epoch": 0.49111, + "grad_norm": 1.432370629421403, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49111 + }, + { + "epoch": 0.49112, + "grad_norm": 1.4989734602329432, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49112 + }, + { + "epoch": 0.49113, + "grad_norm": 1.388666938523512, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 49113 + }, + { + "epoch": 0.49114, + "grad_norm": 1.5584781572341366, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 49114 + }, + { + "epoch": 0.49115, + "grad_norm": 1.8216833159356118, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49115 + }, + { + "epoch": 0.49116, + "grad_norm": 1.3995418271009021, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 49116 + }, + { + "epoch": 0.49117, + "grad_norm": 1.7473210682311597, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 49117 + }, + { + "epoch": 0.49118, + "grad_norm": 1.5970686292586884, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 49118 + }, + { + "epoch": 0.49119, + "grad_norm": 1.4934999784890968, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 49119 + }, + { + "epoch": 0.4912, + "grad_norm": 1.4874909216225356, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49120 + }, + { + "epoch": 0.49121, + "grad_norm": 1.5859504635452015, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 49121 + }, + { + "epoch": 0.49122, + "grad_norm": 1.6605425361799981, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49122 + }, + { + "epoch": 0.49123, + "grad_norm": 1.3545284813949643, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49123 + }, + { + "epoch": 0.49124, + "grad_norm": 1.789809231355729, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49124 + }, + { + "epoch": 0.49125, + "grad_norm": 1.8750620221945886, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 49125 + }, + { + "epoch": 0.49126, + "grad_norm": 1.282083979776384, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 49126 + }, + { + "epoch": 0.49127, + "grad_norm": 1.3746594924605942, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49127 + }, + { + "epoch": 0.49128, + "grad_norm": 1.57724060743948, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 49128 + }, + { + "epoch": 0.49129, + "grad_norm": 1.3599848043342913, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 49129 + }, + { + "epoch": 0.4913, + "grad_norm": 1.7303053536846225, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 49130 + }, + { + "epoch": 0.49131, + "grad_norm": 1.5204982235479663, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49131 + }, + { + "epoch": 0.49132, + "grad_norm": 1.646348349343423, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 49132 + }, + { + "epoch": 0.49133, + "grad_norm": 1.4437090311931482, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49133 + }, + { + "epoch": 0.49134, + "grad_norm": 1.6135027863267994, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49134 + }, + { + "epoch": 0.49135, + "grad_norm": 1.361585339136145, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 49135 + }, + { + "epoch": 0.49136, + "grad_norm": 1.615180454019271, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 49136 + }, + { + "epoch": 0.49137, + "grad_norm": 1.518298682334056, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 49137 + }, + { + "epoch": 0.49138, + "grad_norm": 1.5526115514684744, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 49138 + }, + { + "epoch": 0.49139, + "grad_norm": 1.493630494784672, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49139 + }, + { + "epoch": 0.4914, + "grad_norm": 1.479327181529004, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49140 + }, + { + "epoch": 0.49141, + "grad_norm": 1.7473464878233926, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 49141 + }, + { + "epoch": 0.49142, + "grad_norm": 1.62876342581325, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 49142 + }, + { + "epoch": 0.49143, + "grad_norm": 1.4001443205558024, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 49143 + }, + { + "epoch": 0.49144, + "grad_norm": 1.6295207756773855, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49144 + }, + { + "epoch": 0.49145, + "grad_norm": 1.4987393363583874, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 49145 + }, + { + "epoch": 0.49146, + "grad_norm": 1.5709435436427694, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 49146 + }, + { + "epoch": 0.49147, + "grad_norm": 1.5493862031247687, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49147 + }, + { + "epoch": 0.49148, + "grad_norm": 1.6463457070522458, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 49148 + }, + { + "epoch": 0.49149, + "grad_norm": 1.5811968941502705, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49149 + }, + { + "epoch": 0.4915, + "grad_norm": 1.4436145234474353, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 49150 + }, + { + "epoch": 0.49151, + "grad_norm": 1.5990709053566998, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 49151 + }, + { + "epoch": 0.49152, + "grad_norm": 1.6114769775530389, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 49152 + }, + { + "epoch": 0.49153, + "grad_norm": 1.3613825174207417, + "learning_rate": 0.003, + "loss": 4.037, + "step": 49153 + }, + { + "epoch": 0.49154, + "grad_norm": 1.5942889446409183, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49154 + }, + { + "epoch": 0.49155, + "grad_norm": 1.502628701540527, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 49155 + }, + { + "epoch": 0.49156, + "grad_norm": 1.7972256412836713, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49156 + }, + { + "epoch": 0.49157, + "grad_norm": 1.7434091537830427, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 49157 + }, + { + "epoch": 0.49158, + "grad_norm": 1.8134842534845843, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 49158 + }, + { + "epoch": 0.49159, + "grad_norm": 1.539739830591697, + "learning_rate": 0.003, + "loss": 4.034, + "step": 49159 + }, + { + "epoch": 0.4916, + "grad_norm": 1.6369173708556377, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 49160 + }, + { + "epoch": 0.49161, + "grad_norm": 1.4196586649137173, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49161 + }, + { + "epoch": 0.49162, + "grad_norm": 1.6394403507088253, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 49162 + }, + { + "epoch": 0.49163, + "grad_norm": 1.4303201306802502, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 49163 + }, + { + "epoch": 0.49164, + "grad_norm": 1.7539866142993599, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 49164 + }, + { + "epoch": 0.49165, + "grad_norm": 1.4081823937169546, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49165 + }, + { + "epoch": 0.49166, + "grad_norm": 1.5705421382680498, + "learning_rate": 0.003, + "loss": 4.029, + "step": 49166 + }, + { + "epoch": 0.49167, + "grad_norm": 1.4617723141175447, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 49167 + }, + { + "epoch": 0.49168, + "grad_norm": 1.65912265467856, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49168 + }, + { + "epoch": 0.49169, + "grad_norm": 1.4233703351208487, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49169 + }, + { + "epoch": 0.4917, + "grad_norm": 1.7587398821453264, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49170 + }, + { + "epoch": 0.49171, + "grad_norm": 1.4140100526318495, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 49171 + }, + { + "epoch": 0.49172, + "grad_norm": 1.677670832236797, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49172 + }, + { + "epoch": 0.49173, + "grad_norm": 1.321579287243517, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49173 + }, + { + "epoch": 0.49174, + "grad_norm": 1.6059683330065337, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 49174 + }, + { + "epoch": 0.49175, + "grad_norm": 1.717084480125822, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 49175 + }, + { + "epoch": 0.49176, + "grad_norm": 1.5049634058005072, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 49176 + }, + { + "epoch": 0.49177, + "grad_norm": 1.6635510691542454, + "learning_rate": 0.003, + "loss": 4.001, + "step": 49177 + }, + { + "epoch": 0.49178, + "grad_norm": 1.3362823715061927, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 49178 + }, + { + "epoch": 0.49179, + "grad_norm": 1.7873938704833734, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 49179 + }, + { + "epoch": 0.4918, + "grad_norm": 1.5447699838765274, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 49180 + }, + { + "epoch": 0.49181, + "grad_norm": 1.579752581009126, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 49181 + }, + { + "epoch": 0.49182, + "grad_norm": 2.1640848782286977, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 49182 + }, + { + "epoch": 0.49183, + "grad_norm": 1.0565193712537129, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 49183 + }, + { + "epoch": 0.49184, + "grad_norm": 1.792478550760464, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 49184 + }, + { + "epoch": 0.49185, + "grad_norm": 1.4260461370204207, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 49185 + }, + { + "epoch": 0.49186, + "grad_norm": 2.02695916753154, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49186 + }, + { + "epoch": 0.49187, + "grad_norm": 1.3181512812304381, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 49187 + }, + { + "epoch": 0.49188, + "grad_norm": 1.4648254768756093, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 49188 + }, + { + "epoch": 0.49189, + "grad_norm": 1.655874524605389, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 49189 + }, + { + "epoch": 0.4919, + "grad_norm": 1.6100341061546155, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 49190 + }, + { + "epoch": 0.49191, + "grad_norm": 1.5805608591568596, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 49191 + }, + { + "epoch": 0.49192, + "grad_norm": 1.4000913408930997, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 49192 + }, + { + "epoch": 0.49193, + "grad_norm": 1.5369893174893465, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49193 + }, + { + "epoch": 0.49194, + "grad_norm": 1.8808249845563765, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 49194 + }, + { + "epoch": 0.49195, + "grad_norm": 1.4920116817665408, + "learning_rate": 0.003, + "loss": 4.027, + "step": 49195 + }, + { + "epoch": 0.49196, + "grad_norm": 2.0152440684548134, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49196 + }, + { + "epoch": 0.49197, + "grad_norm": 1.6107963051300924, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 49197 + }, + { + "epoch": 0.49198, + "grad_norm": 1.3784882998251102, + "learning_rate": 0.003, + "loss": 4.01, + "step": 49198 + }, + { + "epoch": 0.49199, + "grad_norm": 1.4938957844501164, + "learning_rate": 0.003, + "loss": 4.041, + "step": 49199 + }, + { + "epoch": 0.492, + "grad_norm": 1.7350316688708298, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 49200 + }, + { + "epoch": 0.49201, + "grad_norm": 1.3566337129470183, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49201 + }, + { + "epoch": 0.49202, + "grad_norm": 1.644998437272512, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49202 + }, + { + "epoch": 0.49203, + "grad_norm": 1.399233690199824, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 49203 + }, + { + "epoch": 0.49204, + "grad_norm": 1.6867960389032555, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 49204 + }, + { + "epoch": 0.49205, + "grad_norm": 1.5271758642642517, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 49205 + }, + { + "epoch": 0.49206, + "grad_norm": 1.5646518206766686, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 49206 + }, + { + "epoch": 0.49207, + "grad_norm": 1.629779930731154, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 49207 + }, + { + "epoch": 0.49208, + "grad_norm": 1.8430443156513583, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 49208 + }, + { + "epoch": 0.49209, + "grad_norm": 1.2501643697950977, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49209 + }, + { + "epoch": 0.4921, + "grad_norm": 1.74369844404475, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49210 + }, + { + "epoch": 0.49211, + "grad_norm": 1.501964438533693, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 49211 + }, + { + "epoch": 0.49212, + "grad_norm": 1.6828441802204959, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 49212 + }, + { + "epoch": 0.49213, + "grad_norm": 1.3750744265772261, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49213 + }, + { + "epoch": 0.49214, + "grad_norm": 1.6007568595068045, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49214 + }, + { + "epoch": 0.49215, + "grad_norm": 1.487570004855052, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49215 + }, + { + "epoch": 0.49216, + "grad_norm": 1.481439166065025, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 49216 + }, + { + "epoch": 0.49217, + "grad_norm": 1.356671539201473, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49217 + }, + { + "epoch": 0.49218, + "grad_norm": 1.5708637845497442, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 49218 + }, + { + "epoch": 0.49219, + "grad_norm": 1.628160366868687, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 49219 + }, + { + "epoch": 0.4922, + "grad_norm": 1.4034218636648017, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 49220 + }, + { + "epoch": 0.49221, + "grad_norm": 1.785160726020042, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 49221 + }, + { + "epoch": 0.49222, + "grad_norm": 1.2839541276972029, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 49222 + }, + { + "epoch": 0.49223, + "grad_norm": 1.8666829390995814, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 49223 + }, + { + "epoch": 0.49224, + "grad_norm": 1.746369084651015, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 49224 + }, + { + "epoch": 0.49225, + "grad_norm": 1.5426867836183287, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 49225 + }, + { + "epoch": 0.49226, + "grad_norm": 1.4628015069121827, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49226 + }, + { + "epoch": 0.49227, + "grad_norm": 1.5442387281525274, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49227 + }, + { + "epoch": 0.49228, + "grad_norm": 1.698572475274461, + "learning_rate": 0.003, + "loss": 4.05, + "step": 49228 + }, + { + "epoch": 0.49229, + "grad_norm": 1.3161938158206388, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49229 + }, + { + "epoch": 0.4923, + "grad_norm": 1.8756988641717725, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 49230 + }, + { + "epoch": 0.49231, + "grad_norm": 1.6699722798279966, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 49231 + }, + { + "epoch": 0.49232, + "grad_norm": 1.5382133758312209, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49232 + }, + { + "epoch": 0.49233, + "grad_norm": 1.4417677257198396, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 49233 + }, + { + "epoch": 0.49234, + "grad_norm": 1.7310156247523225, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 49234 + }, + { + "epoch": 0.49235, + "grad_norm": 1.3177269746514872, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 49235 + }, + { + "epoch": 0.49236, + "grad_norm": 1.7945188713286966, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 49236 + }, + { + "epoch": 0.49237, + "grad_norm": 1.5674139582500888, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 49237 + }, + { + "epoch": 0.49238, + "grad_norm": 1.491781485433785, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 49238 + }, + { + "epoch": 0.49239, + "grad_norm": 1.3464762164485735, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49239 + }, + { + "epoch": 0.4924, + "grad_norm": 1.6930354418487872, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49240 + }, + { + "epoch": 0.49241, + "grad_norm": 1.672116851878767, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 49241 + }, + { + "epoch": 0.49242, + "grad_norm": 1.3055003366225135, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49242 + }, + { + "epoch": 0.49243, + "grad_norm": 1.840891540528248, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 49243 + }, + { + "epoch": 0.49244, + "grad_norm": 1.3182012875848788, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 49244 + }, + { + "epoch": 0.49245, + "grad_norm": 1.6892542052501658, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 49245 + }, + { + "epoch": 0.49246, + "grad_norm": 1.582241942482941, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 49246 + }, + { + "epoch": 0.49247, + "grad_norm": 1.7311366331001241, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 49247 + }, + { + "epoch": 0.49248, + "grad_norm": 1.4630844514418666, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49248 + }, + { + "epoch": 0.49249, + "grad_norm": 1.5193792362407001, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49249 + }, + { + "epoch": 0.4925, + "grad_norm": 1.7234827207894914, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 49250 + }, + { + "epoch": 0.49251, + "grad_norm": 1.2155083896393488, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 49251 + }, + { + "epoch": 0.49252, + "grad_norm": 1.5643282959650802, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 49252 + }, + { + "epoch": 0.49253, + "grad_norm": 1.5716685360475386, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49253 + }, + { + "epoch": 0.49254, + "grad_norm": 1.5597970285164555, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49254 + }, + { + "epoch": 0.49255, + "grad_norm": 1.5312555461521027, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 49255 + }, + { + "epoch": 0.49256, + "grad_norm": 1.6113459349429338, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 49256 + }, + { + "epoch": 0.49257, + "grad_norm": 1.4756762223177229, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 49257 + }, + { + "epoch": 0.49258, + "grad_norm": 1.4999343231605122, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49258 + }, + { + "epoch": 0.49259, + "grad_norm": 1.4218050408613878, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49259 + }, + { + "epoch": 0.4926, + "grad_norm": 1.7338162158087311, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 49260 + }, + { + "epoch": 0.49261, + "grad_norm": 1.3770822381257906, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 49261 + }, + { + "epoch": 0.49262, + "grad_norm": 1.6228021800570318, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 49262 + }, + { + "epoch": 0.49263, + "grad_norm": 1.6311405152381595, + "learning_rate": 0.003, + "loss": 4.041, + "step": 49263 + }, + { + "epoch": 0.49264, + "grad_norm": 1.614250657282579, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49264 + }, + { + "epoch": 0.49265, + "grad_norm": 1.7226354093780496, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49265 + }, + { + "epoch": 0.49266, + "grad_norm": 1.5954714752044536, + "learning_rate": 0.003, + "loss": 4.029, + "step": 49266 + }, + { + "epoch": 0.49267, + "grad_norm": 1.3650974812481398, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 49267 + }, + { + "epoch": 0.49268, + "grad_norm": 1.8121868891219128, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 49268 + }, + { + "epoch": 0.49269, + "grad_norm": 1.3640640987313752, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 49269 + }, + { + "epoch": 0.4927, + "grad_norm": 1.7813080990520465, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49270 + }, + { + "epoch": 0.49271, + "grad_norm": 1.2555306038015945, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 49271 + }, + { + "epoch": 0.49272, + "grad_norm": 1.9255524844401353, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 49272 + }, + { + "epoch": 0.49273, + "grad_norm": 1.2276871317772438, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49273 + }, + { + "epoch": 0.49274, + "grad_norm": 1.6400127878424762, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 49274 + }, + { + "epoch": 0.49275, + "grad_norm": 1.7901704767567834, + "learning_rate": 0.003, + "loss": 4.045, + "step": 49275 + }, + { + "epoch": 0.49276, + "grad_norm": 1.8381394261408486, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 49276 + }, + { + "epoch": 0.49277, + "grad_norm": 1.6795865119089737, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 49277 + }, + { + "epoch": 0.49278, + "grad_norm": 1.674017715174483, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 49278 + }, + { + "epoch": 0.49279, + "grad_norm": 1.4536774594832775, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 49279 + }, + { + "epoch": 0.4928, + "grad_norm": 1.3421948675714317, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 49280 + }, + { + "epoch": 0.49281, + "grad_norm": 1.7616499741011589, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49281 + }, + { + "epoch": 0.49282, + "grad_norm": 1.3582959095579508, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 49282 + }, + { + "epoch": 0.49283, + "grad_norm": 1.522606994978879, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 49283 + }, + { + "epoch": 0.49284, + "grad_norm": 1.490671417908837, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 49284 + }, + { + "epoch": 0.49285, + "grad_norm": 1.8682371674144884, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 49285 + }, + { + "epoch": 0.49286, + "grad_norm": 1.1960168672926432, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 49286 + }, + { + "epoch": 0.49287, + "grad_norm": 1.7631236232425462, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 49287 + }, + { + "epoch": 0.49288, + "grad_norm": 1.2678659244000288, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49288 + }, + { + "epoch": 0.49289, + "grad_norm": 1.9472759948054748, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49289 + }, + { + "epoch": 0.4929, + "grad_norm": 1.3640553331274632, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 49290 + }, + { + "epoch": 0.49291, + "grad_norm": 1.9909871236582213, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49291 + }, + { + "epoch": 0.49292, + "grad_norm": 1.4713124053126134, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 49292 + }, + { + "epoch": 0.49293, + "grad_norm": 1.6261842909543622, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 49293 + }, + { + "epoch": 0.49294, + "grad_norm": 1.4116131580347737, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49294 + }, + { + "epoch": 0.49295, + "grad_norm": 1.5732855963847123, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 49295 + }, + { + "epoch": 0.49296, + "grad_norm": 1.481411107121206, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 49296 + }, + { + "epoch": 0.49297, + "grad_norm": 1.7250487898413709, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 49297 + }, + { + "epoch": 0.49298, + "grad_norm": 1.5444063621774997, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 49298 + }, + { + "epoch": 0.49299, + "grad_norm": 1.4329628790539657, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 49299 + }, + { + "epoch": 0.493, + "grad_norm": 1.468419270644518, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 49300 + }, + { + "epoch": 0.49301, + "grad_norm": 1.5722772588525356, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 49301 + }, + { + "epoch": 0.49302, + "grad_norm": 1.354678433007208, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 49302 + }, + { + "epoch": 0.49303, + "grad_norm": 1.7554671983457548, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49303 + }, + { + "epoch": 0.49304, + "grad_norm": 1.7837213084014505, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 49304 + }, + { + "epoch": 0.49305, + "grad_norm": 1.9275784931863627, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49305 + }, + { + "epoch": 0.49306, + "grad_norm": 1.336908031997711, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 49306 + }, + { + "epoch": 0.49307, + "grad_norm": 1.505073501572239, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49307 + }, + { + "epoch": 0.49308, + "grad_norm": 1.3797367355294683, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49308 + }, + { + "epoch": 0.49309, + "grad_norm": 1.7336601000631087, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 49309 + }, + { + "epoch": 0.4931, + "grad_norm": 1.3674848937407218, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49310 + }, + { + "epoch": 0.49311, + "grad_norm": 1.5933275166462864, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 49311 + }, + { + "epoch": 0.49312, + "grad_norm": 1.4917810028547844, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49312 + }, + { + "epoch": 0.49313, + "grad_norm": 1.5091519815913128, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 49313 + }, + { + "epoch": 0.49314, + "grad_norm": 1.4490987230119237, + "learning_rate": 0.003, + "loss": 4.027, + "step": 49314 + }, + { + "epoch": 0.49315, + "grad_norm": 1.584137694807972, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 49315 + }, + { + "epoch": 0.49316, + "grad_norm": 1.4705615756266521, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 49316 + }, + { + "epoch": 0.49317, + "grad_norm": 1.6391540065165038, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 49317 + }, + { + "epoch": 0.49318, + "grad_norm": 1.4160622359337782, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 49318 + }, + { + "epoch": 0.49319, + "grad_norm": 1.5174306374740103, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 49319 + }, + { + "epoch": 0.4932, + "grad_norm": 1.5801373597804746, + "learning_rate": 0.003, + "loss": 4.0, + "step": 49320 + }, + { + "epoch": 0.49321, + "grad_norm": 1.608581133953969, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 49321 + }, + { + "epoch": 0.49322, + "grad_norm": 1.532145640762536, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49322 + }, + { + "epoch": 0.49323, + "grad_norm": 1.6770683849267936, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 49323 + }, + { + "epoch": 0.49324, + "grad_norm": 1.4945780000917472, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 49324 + }, + { + "epoch": 0.49325, + "grad_norm": 1.5793457699931415, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49325 + }, + { + "epoch": 0.49326, + "grad_norm": 1.4535313892400201, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 49326 + }, + { + "epoch": 0.49327, + "grad_norm": 1.677129947365577, + "learning_rate": 0.003, + "loss": 4.002, + "step": 49327 + }, + { + "epoch": 0.49328, + "grad_norm": 1.4783278995855027, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 49328 + }, + { + "epoch": 0.49329, + "grad_norm": 1.7631635148462699, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 49329 + }, + { + "epoch": 0.4933, + "grad_norm": 1.5904355115313278, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49330 + }, + { + "epoch": 0.49331, + "grad_norm": 1.5196313230088594, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49331 + }, + { + "epoch": 0.49332, + "grad_norm": 1.6625927413281252, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49332 + }, + { + "epoch": 0.49333, + "grad_norm": 1.3696103872617704, + "learning_rate": 0.003, + "loss": 4.039, + "step": 49333 + }, + { + "epoch": 0.49334, + "grad_norm": 1.6825203042408579, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49334 + }, + { + "epoch": 0.49335, + "grad_norm": 1.3669374397328862, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 49335 + }, + { + "epoch": 0.49336, + "grad_norm": 1.8710165699743828, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 49336 + }, + { + "epoch": 0.49337, + "grad_norm": 1.1881846310852464, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 49337 + }, + { + "epoch": 0.49338, + "grad_norm": 1.7980411686227795, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 49338 + }, + { + "epoch": 0.49339, + "grad_norm": 1.315157582731034, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 49339 + }, + { + "epoch": 0.4934, + "grad_norm": 1.9893204675869176, + "learning_rate": 0.003, + "loss": 4.048, + "step": 49340 + }, + { + "epoch": 0.49341, + "grad_norm": 1.272126522142672, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 49341 + }, + { + "epoch": 0.49342, + "grad_norm": 1.7526896448686984, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 49342 + }, + { + "epoch": 0.49343, + "grad_norm": 1.5495312404187778, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 49343 + }, + { + "epoch": 0.49344, + "grad_norm": 1.5453999464135026, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49344 + }, + { + "epoch": 0.49345, + "grad_norm": 1.703316613400098, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 49345 + }, + { + "epoch": 0.49346, + "grad_norm": 1.9273906146751987, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49346 + }, + { + "epoch": 0.49347, + "grad_norm": 1.2387046087029532, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49347 + }, + { + "epoch": 0.49348, + "grad_norm": 1.9563225177109715, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 49348 + }, + { + "epoch": 0.49349, + "grad_norm": 1.3109207493168311, + "learning_rate": 0.003, + "loss": 4.059, + "step": 49349 + }, + { + "epoch": 0.4935, + "grad_norm": 1.3392000293966122, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 49350 + }, + { + "epoch": 0.49351, + "grad_norm": 1.8567800417999145, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49351 + }, + { + "epoch": 0.49352, + "grad_norm": 1.4429839566004026, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 49352 + }, + { + "epoch": 0.49353, + "grad_norm": 1.2966654449130441, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 49353 + }, + { + "epoch": 0.49354, + "grad_norm": 1.708425207987573, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 49354 + }, + { + "epoch": 0.49355, + "grad_norm": 1.2292791670830439, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 49355 + }, + { + "epoch": 0.49356, + "grad_norm": 1.5282527791103664, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 49356 + }, + { + "epoch": 0.49357, + "grad_norm": 1.5037936331287112, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 49357 + }, + { + "epoch": 0.49358, + "grad_norm": 1.890113743588958, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49358 + }, + { + "epoch": 0.49359, + "grad_norm": 1.3964796148999108, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 49359 + }, + { + "epoch": 0.4936, + "grad_norm": 2.0643599851717824, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 49360 + }, + { + "epoch": 0.49361, + "grad_norm": 1.7262474803950518, + "learning_rate": 0.003, + "loss": 4.043, + "step": 49361 + }, + { + "epoch": 0.49362, + "grad_norm": 1.4344569044121482, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 49362 + }, + { + "epoch": 0.49363, + "grad_norm": 1.4560192237375547, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 49363 + }, + { + "epoch": 0.49364, + "grad_norm": 1.8385259745631368, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 49364 + }, + { + "epoch": 0.49365, + "grad_norm": 1.538087190055971, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 49365 + }, + { + "epoch": 0.49366, + "grad_norm": 1.3372463898696947, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 49366 + }, + { + "epoch": 0.49367, + "grad_norm": 1.806090344722627, + "learning_rate": 0.003, + "loss": 4.026, + "step": 49367 + }, + { + "epoch": 0.49368, + "grad_norm": 1.337116345119518, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49368 + }, + { + "epoch": 0.49369, + "grad_norm": 1.459841634020687, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49369 + }, + { + "epoch": 0.4937, + "grad_norm": 1.7401708077619922, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49370 + }, + { + "epoch": 0.49371, + "grad_norm": 1.3278556443857805, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49371 + }, + { + "epoch": 0.49372, + "grad_norm": 1.9412063291198751, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 49372 + }, + { + "epoch": 0.49373, + "grad_norm": 1.5672943364880227, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49373 + }, + { + "epoch": 0.49374, + "grad_norm": 1.9459349533062313, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 49374 + }, + { + "epoch": 0.49375, + "grad_norm": 1.6263976248826526, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 49375 + }, + { + "epoch": 0.49376, + "grad_norm": 1.6455285535322377, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 49376 + }, + { + "epoch": 0.49377, + "grad_norm": 1.4558604210273491, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 49377 + }, + { + "epoch": 0.49378, + "grad_norm": 1.338668007401027, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 49378 + }, + { + "epoch": 0.49379, + "grad_norm": 1.644821985168051, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 49379 + }, + { + "epoch": 0.4938, + "grad_norm": 1.3427958943917917, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 49380 + }, + { + "epoch": 0.49381, + "grad_norm": 1.4099955263789927, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 49381 + }, + { + "epoch": 0.49382, + "grad_norm": 1.64788813234917, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 49382 + }, + { + "epoch": 0.49383, + "grad_norm": 1.2660321152859264, + "learning_rate": 0.003, + "loss": 4.019, + "step": 49383 + }, + { + "epoch": 0.49384, + "grad_norm": 1.9286240291316936, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 49384 + }, + { + "epoch": 0.49385, + "grad_norm": 1.316735006689888, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 49385 + }, + { + "epoch": 0.49386, + "grad_norm": 1.9214471988722188, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 49386 + }, + { + "epoch": 0.49387, + "grad_norm": 1.686203449012379, + "learning_rate": 0.003, + "loss": 4.057, + "step": 49387 + }, + { + "epoch": 0.49388, + "grad_norm": 1.4392821989306115, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49388 + }, + { + "epoch": 0.49389, + "grad_norm": 1.852281281392314, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 49389 + }, + { + "epoch": 0.4939, + "grad_norm": 1.3651361959001027, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 49390 + }, + { + "epoch": 0.49391, + "grad_norm": 1.8255880836441583, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 49391 + }, + { + "epoch": 0.49392, + "grad_norm": 1.4581457851408093, + "learning_rate": 0.003, + "loss": 4.057, + "step": 49392 + }, + { + "epoch": 0.49393, + "grad_norm": 1.8177012450221546, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49393 + }, + { + "epoch": 0.49394, + "grad_norm": 1.478171019843922, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 49394 + }, + { + "epoch": 0.49395, + "grad_norm": 1.7902415644049903, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 49395 + }, + { + "epoch": 0.49396, + "grad_norm": 1.5778492632023662, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 49396 + }, + { + "epoch": 0.49397, + "grad_norm": 1.4852785518288885, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49397 + }, + { + "epoch": 0.49398, + "grad_norm": 1.461098621532123, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 49398 + }, + { + "epoch": 0.49399, + "grad_norm": 1.40031773666981, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 49399 + }, + { + "epoch": 0.494, + "grad_norm": 1.5074023758832353, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49400 + }, + { + "epoch": 0.49401, + "grad_norm": 1.3187852371184532, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49401 + }, + { + "epoch": 0.49402, + "grad_norm": 1.8247995818059366, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 49402 + }, + { + "epoch": 0.49403, + "grad_norm": 1.1676408768359017, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 49403 + }, + { + "epoch": 0.49404, + "grad_norm": 1.779201530864547, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 49404 + }, + { + "epoch": 0.49405, + "grad_norm": 1.4716008874806417, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 49405 + }, + { + "epoch": 0.49406, + "grad_norm": 1.7945936195398107, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49406 + }, + { + "epoch": 0.49407, + "grad_norm": 1.5691165770712656, + "learning_rate": 0.003, + "loss": 4.054, + "step": 49407 + }, + { + "epoch": 0.49408, + "grad_norm": 1.6505623271395546, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 49408 + }, + { + "epoch": 0.49409, + "grad_norm": 1.7130685500951237, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 49409 + }, + { + "epoch": 0.4941, + "grad_norm": 1.3273210298609317, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49410 + }, + { + "epoch": 0.49411, + "grad_norm": 1.5850469901490256, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 49411 + }, + { + "epoch": 0.49412, + "grad_norm": 1.3987331226586563, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 49412 + }, + { + "epoch": 0.49413, + "grad_norm": 1.8104039229358868, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 49413 + }, + { + "epoch": 0.49414, + "grad_norm": 1.752576103032679, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 49414 + }, + { + "epoch": 0.49415, + "grad_norm": 1.2968692500234646, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 49415 + }, + { + "epoch": 0.49416, + "grad_norm": 1.7479879861264527, + "learning_rate": 0.003, + "loss": 4.045, + "step": 49416 + }, + { + "epoch": 0.49417, + "grad_norm": 1.2650204528172886, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 49417 + }, + { + "epoch": 0.49418, + "grad_norm": 1.9849695077268814, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 49418 + }, + { + "epoch": 0.49419, + "grad_norm": 1.387948585018105, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 49419 + }, + { + "epoch": 0.4942, + "grad_norm": 1.8679631475485587, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49420 + }, + { + "epoch": 0.49421, + "grad_norm": 1.28717560974861, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 49421 + }, + { + "epoch": 0.49422, + "grad_norm": 1.7183349525829237, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 49422 + }, + { + "epoch": 0.49423, + "grad_norm": 1.3580421908494682, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 49423 + }, + { + "epoch": 0.49424, + "grad_norm": 1.6201027973109199, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49424 + }, + { + "epoch": 0.49425, + "grad_norm": 1.2659902590752647, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 49425 + }, + { + "epoch": 0.49426, + "grad_norm": 1.5673601380966529, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49426 + }, + { + "epoch": 0.49427, + "grad_norm": 1.8234816357115038, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 49427 + }, + { + "epoch": 0.49428, + "grad_norm": 1.3821128835063092, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 49428 + }, + { + "epoch": 0.49429, + "grad_norm": 1.60110030348031, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 49429 + }, + { + "epoch": 0.4943, + "grad_norm": 1.5645125201941508, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 49430 + }, + { + "epoch": 0.49431, + "grad_norm": 1.586818370947084, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 49431 + }, + { + "epoch": 0.49432, + "grad_norm": 1.4612194886994214, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 49432 + }, + { + "epoch": 0.49433, + "grad_norm": 1.8174579328190468, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 49433 + }, + { + "epoch": 0.49434, + "grad_norm": 1.4385660849492528, + "learning_rate": 0.003, + "loss": 4.044, + "step": 49434 + }, + { + "epoch": 0.49435, + "grad_norm": 1.7678761031489927, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49435 + }, + { + "epoch": 0.49436, + "grad_norm": 1.4374150991997878, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 49436 + }, + { + "epoch": 0.49437, + "grad_norm": 1.8584743387263025, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 49437 + }, + { + "epoch": 0.49438, + "grad_norm": 1.0814963331966354, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49438 + }, + { + "epoch": 0.49439, + "grad_norm": 1.7952886147853993, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 49439 + }, + { + "epoch": 0.4944, + "grad_norm": 1.5163977939452307, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 49440 + }, + { + "epoch": 0.49441, + "grad_norm": 1.5343961064738778, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 49441 + }, + { + "epoch": 0.49442, + "grad_norm": 1.8741622290269762, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49442 + }, + { + "epoch": 0.49443, + "grad_norm": 1.2764349061458684, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49443 + }, + { + "epoch": 0.49444, + "grad_norm": 1.3752259161470988, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 49444 + }, + { + "epoch": 0.49445, + "grad_norm": 1.566362703626003, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49445 + }, + { + "epoch": 0.49446, + "grad_norm": 1.3971271182036153, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 49446 + }, + { + "epoch": 0.49447, + "grad_norm": 1.7871951484580808, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 49447 + }, + { + "epoch": 0.49448, + "grad_norm": 1.6977066952185196, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 49448 + }, + { + "epoch": 0.49449, + "grad_norm": 1.5690812113082477, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49449 + }, + { + "epoch": 0.4945, + "grad_norm": 1.3512174364757115, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 49450 + }, + { + "epoch": 0.49451, + "grad_norm": 1.8199381217274035, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 49451 + }, + { + "epoch": 0.49452, + "grad_norm": 1.2369082949648842, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49452 + }, + { + "epoch": 0.49453, + "grad_norm": 2.089376804190773, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 49453 + }, + { + "epoch": 0.49454, + "grad_norm": 1.246250672309652, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 49454 + }, + { + "epoch": 0.49455, + "grad_norm": 1.7631002547839856, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49455 + }, + { + "epoch": 0.49456, + "grad_norm": 1.5439092942680064, + "learning_rate": 0.003, + "loss": 4.049, + "step": 49456 + }, + { + "epoch": 0.49457, + "grad_norm": 1.5844417637443071, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 49457 + }, + { + "epoch": 0.49458, + "grad_norm": 1.744575793431803, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 49458 + }, + { + "epoch": 0.49459, + "grad_norm": 1.5380899295754977, + "learning_rate": 0.003, + "loss": 4.048, + "step": 49459 + }, + { + "epoch": 0.4946, + "grad_norm": 1.526424313845692, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 49460 + }, + { + "epoch": 0.49461, + "grad_norm": 1.3706539661111456, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 49461 + }, + { + "epoch": 0.49462, + "grad_norm": 1.8669545719257628, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 49462 + }, + { + "epoch": 0.49463, + "grad_norm": 1.333835741537245, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49463 + }, + { + "epoch": 0.49464, + "grad_norm": 2.1293122980616346, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 49464 + }, + { + "epoch": 0.49465, + "grad_norm": 1.50401835040301, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 49465 + }, + { + "epoch": 0.49466, + "grad_norm": 1.922237691498774, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49466 + }, + { + "epoch": 0.49467, + "grad_norm": 1.4828831662305761, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 49467 + }, + { + "epoch": 0.49468, + "grad_norm": 1.3091869369614364, + "learning_rate": 0.003, + "loss": 4.015, + "step": 49468 + }, + { + "epoch": 0.49469, + "grad_norm": 1.7299056248681843, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 49469 + }, + { + "epoch": 0.4947, + "grad_norm": 1.273725610863215, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 49470 + }, + { + "epoch": 0.49471, + "grad_norm": 1.411015494756622, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 49471 + }, + { + "epoch": 0.49472, + "grad_norm": 1.9162583690618225, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49472 + }, + { + "epoch": 0.49473, + "grad_norm": 1.390382736702265, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49473 + }, + { + "epoch": 0.49474, + "grad_norm": 1.6836707298686164, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49474 + }, + { + "epoch": 0.49475, + "grad_norm": 1.5530327895584481, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 49475 + }, + { + "epoch": 0.49476, + "grad_norm": 2.0782848418683733, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49476 + }, + { + "epoch": 0.49477, + "grad_norm": 1.2388650339873362, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 49477 + }, + { + "epoch": 0.49478, + "grad_norm": 1.857951471066099, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49478 + }, + { + "epoch": 0.49479, + "grad_norm": 1.5351638127846712, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49479 + }, + { + "epoch": 0.4948, + "grad_norm": 1.4715516883412179, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 49480 + }, + { + "epoch": 0.49481, + "grad_norm": 1.9609508817180543, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 49481 + }, + { + "epoch": 0.49482, + "grad_norm": 1.4289326505034219, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 49482 + }, + { + "epoch": 0.49483, + "grad_norm": 1.5942110657020263, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 49483 + }, + { + "epoch": 0.49484, + "grad_norm": 1.6692904315088763, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 49484 + }, + { + "epoch": 0.49485, + "grad_norm": 1.605256492086723, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49485 + }, + { + "epoch": 0.49486, + "grad_norm": 1.7562889099328898, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49486 + }, + { + "epoch": 0.49487, + "grad_norm": 1.3627213371859817, + "learning_rate": 0.003, + "loss": 4.034, + "step": 49487 + }, + { + "epoch": 0.49488, + "grad_norm": 1.7542686682420372, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 49488 + }, + { + "epoch": 0.49489, + "grad_norm": 1.3499527357632863, + "learning_rate": 0.003, + "loss": 4.018, + "step": 49489 + }, + { + "epoch": 0.4949, + "grad_norm": 1.491639551355786, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 49490 + }, + { + "epoch": 0.49491, + "grad_norm": 1.5768732030447652, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 49491 + }, + { + "epoch": 0.49492, + "grad_norm": 1.5688081150701887, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 49492 + }, + { + "epoch": 0.49493, + "grad_norm": 1.5647435571207415, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 49493 + }, + { + "epoch": 0.49494, + "grad_norm": 1.6265746504037475, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 49494 + }, + { + "epoch": 0.49495, + "grad_norm": 1.6118307687654616, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 49495 + }, + { + "epoch": 0.49496, + "grad_norm": 1.6588474926173091, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 49496 + }, + { + "epoch": 0.49497, + "grad_norm": 1.4014409557002103, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 49497 + }, + { + "epoch": 0.49498, + "grad_norm": 1.7115066901752427, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 49498 + }, + { + "epoch": 0.49499, + "grad_norm": 1.3613630862990698, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 49499 + }, + { + "epoch": 0.495, + "grad_norm": 1.4486975383080876, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 49500 + }, + { + "epoch": 0.49501, + "grad_norm": 1.3105883670048926, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49501 + }, + { + "epoch": 0.49502, + "grad_norm": 1.7911328541885991, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 49502 + }, + { + "epoch": 0.49503, + "grad_norm": 1.381235223693128, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 49503 + }, + { + "epoch": 0.49504, + "grad_norm": 1.575713250455001, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 49504 + }, + { + "epoch": 0.49505, + "grad_norm": 1.4199500168271364, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 49505 + }, + { + "epoch": 0.49506, + "grad_norm": 1.813657219147838, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 49506 + }, + { + "epoch": 0.49507, + "grad_norm": 1.420010533194053, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 49507 + }, + { + "epoch": 0.49508, + "grad_norm": 1.4008088104775733, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 49508 + }, + { + "epoch": 0.49509, + "grad_norm": 1.8580998516030385, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 49509 + }, + { + "epoch": 0.4951, + "grad_norm": 1.492492445856756, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49510 + }, + { + "epoch": 0.49511, + "grad_norm": 1.735673625012027, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 49511 + }, + { + "epoch": 0.49512, + "grad_norm": 1.6660954860620352, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 49512 + }, + { + "epoch": 0.49513, + "grad_norm": 1.7486044463133263, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 49513 + }, + { + "epoch": 0.49514, + "grad_norm": 1.5259655511097074, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49514 + }, + { + "epoch": 0.49515, + "grad_norm": 1.5061880384435775, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 49515 + }, + { + "epoch": 0.49516, + "grad_norm": 1.632161101583817, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49516 + }, + { + "epoch": 0.49517, + "grad_norm": 1.5202017708343447, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49517 + }, + { + "epoch": 0.49518, + "grad_norm": 1.6113423314181516, + "learning_rate": 0.003, + "loss": 4.025, + "step": 49518 + }, + { + "epoch": 0.49519, + "grad_norm": 1.338707853405824, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49519 + }, + { + "epoch": 0.4952, + "grad_norm": 1.700243968413113, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 49520 + }, + { + "epoch": 0.49521, + "grad_norm": 1.3595133438770273, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49521 + }, + { + "epoch": 0.49522, + "grad_norm": 1.4524234443163047, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 49522 + }, + { + "epoch": 0.49523, + "grad_norm": 1.422477500302164, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 49523 + }, + { + "epoch": 0.49524, + "grad_norm": 1.7493820451396607, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 49524 + }, + { + "epoch": 0.49525, + "grad_norm": 2.0883646745274653, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 49525 + }, + { + "epoch": 0.49526, + "grad_norm": 1.794092797189719, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 49526 + }, + { + "epoch": 0.49527, + "grad_norm": 1.6647926058448452, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 49527 + }, + { + "epoch": 0.49528, + "grad_norm": 1.3151092044201935, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 49528 + }, + { + "epoch": 0.49529, + "grad_norm": 1.800252625760973, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49529 + }, + { + "epoch": 0.4953, + "grad_norm": 1.4764003324175954, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49530 + }, + { + "epoch": 0.49531, + "grad_norm": 1.4380791071808734, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 49531 + }, + { + "epoch": 0.49532, + "grad_norm": 1.5922419001593784, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 49532 + }, + { + "epoch": 0.49533, + "grad_norm": 1.3958618242349652, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 49533 + }, + { + "epoch": 0.49534, + "grad_norm": 1.6479469039696704, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49534 + }, + { + "epoch": 0.49535, + "grad_norm": 1.3229581242255641, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 49535 + }, + { + "epoch": 0.49536, + "grad_norm": 1.925966089079979, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 49536 + }, + { + "epoch": 0.49537, + "grad_norm": 1.29366756713965, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 49537 + }, + { + "epoch": 0.49538, + "grad_norm": 1.9016743742290667, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 49538 + }, + { + "epoch": 0.49539, + "grad_norm": 1.1698980905302545, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 49539 + }, + { + "epoch": 0.4954, + "grad_norm": 1.5682652193305924, + "learning_rate": 0.003, + "loss": 4.042, + "step": 49540 + }, + { + "epoch": 0.49541, + "grad_norm": 1.7034687570872384, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 49541 + }, + { + "epoch": 0.49542, + "grad_norm": 1.8583335872206976, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 49542 + }, + { + "epoch": 0.49543, + "grad_norm": 1.3280544662864586, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 49543 + }, + { + "epoch": 0.49544, + "grad_norm": 1.5286225342034363, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 49544 + }, + { + "epoch": 0.49545, + "grad_norm": 1.4447684528720208, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 49545 + }, + { + "epoch": 0.49546, + "grad_norm": 1.4235613789434216, + "learning_rate": 0.003, + "loss": 4.022, + "step": 49546 + }, + { + "epoch": 0.49547, + "grad_norm": 1.668804913748585, + "learning_rate": 0.003, + "loss": 3.988, + "step": 49547 + }, + { + "epoch": 0.49548, + "grad_norm": 1.57399175778241, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49548 + }, + { + "epoch": 0.49549, + "grad_norm": 1.5269213400706612, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49549 + }, + { + "epoch": 0.4955, + "grad_norm": 1.5562415780426648, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 49550 + }, + { + "epoch": 0.49551, + "grad_norm": 1.4004770877116404, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49551 + }, + { + "epoch": 0.49552, + "grad_norm": 1.2580034709748025, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 49552 + }, + { + "epoch": 0.49553, + "grad_norm": 1.4903554589596122, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49553 + }, + { + "epoch": 0.49554, + "grad_norm": 1.4472448526538537, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 49554 + }, + { + "epoch": 0.49555, + "grad_norm": 2.2915719785749404, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 49555 + }, + { + "epoch": 0.49556, + "grad_norm": 1.6041975405978017, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 49556 + }, + { + "epoch": 0.49557, + "grad_norm": 2.1074079998701944, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 49557 + }, + { + "epoch": 0.49558, + "grad_norm": 1.5990680560429924, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49558 + }, + { + "epoch": 0.49559, + "grad_norm": 1.381197056246101, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 49559 + }, + { + "epoch": 0.4956, + "grad_norm": 1.753477962349743, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 49560 + }, + { + "epoch": 0.49561, + "grad_norm": 1.5378394399245803, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 49561 + }, + { + "epoch": 0.49562, + "grad_norm": 1.5870420523512472, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 49562 + }, + { + "epoch": 0.49563, + "grad_norm": 1.671280178597629, + "learning_rate": 0.003, + "loss": 4.081, + "step": 49563 + }, + { + "epoch": 0.49564, + "grad_norm": 1.5319335875793352, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 49564 + }, + { + "epoch": 0.49565, + "grad_norm": 1.6367843998754483, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 49565 + }, + { + "epoch": 0.49566, + "grad_norm": 1.3870868051967467, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 49566 + }, + { + "epoch": 0.49567, + "grad_norm": 1.4597918233163396, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 49567 + }, + { + "epoch": 0.49568, + "grad_norm": 1.8735029956090978, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 49568 + }, + { + "epoch": 0.49569, + "grad_norm": 1.2738042832648182, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 49569 + }, + { + "epoch": 0.4957, + "grad_norm": 1.73068825326442, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49570 + }, + { + "epoch": 0.49571, + "grad_norm": 1.4140752421070208, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 49571 + }, + { + "epoch": 0.49572, + "grad_norm": 1.8410662773466748, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49572 + }, + { + "epoch": 0.49573, + "grad_norm": 1.5228812372900171, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 49573 + }, + { + "epoch": 0.49574, + "grad_norm": 1.3104534857301537, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49574 + }, + { + "epoch": 0.49575, + "grad_norm": 1.6039955874298573, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 49575 + }, + { + "epoch": 0.49576, + "grad_norm": 1.61914516576672, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 49576 + }, + { + "epoch": 0.49577, + "grad_norm": 1.3911453058336154, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 49577 + }, + { + "epoch": 0.49578, + "grad_norm": 1.6815992795936738, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49578 + }, + { + "epoch": 0.49579, + "grad_norm": 1.3540485923224044, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 49579 + }, + { + "epoch": 0.4958, + "grad_norm": 1.6076873928257058, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 49580 + }, + { + "epoch": 0.49581, + "grad_norm": 1.2851438882434574, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 49581 + }, + { + "epoch": 0.49582, + "grad_norm": 1.5851536265583006, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 49582 + }, + { + "epoch": 0.49583, + "grad_norm": 1.193080295144318, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49583 + }, + { + "epoch": 0.49584, + "grad_norm": 1.7109915602958723, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49584 + }, + { + "epoch": 0.49585, + "grad_norm": 1.4522078670770509, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 49585 + }, + { + "epoch": 0.49586, + "grad_norm": 1.4088062237710095, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 49586 + }, + { + "epoch": 0.49587, + "grad_norm": 1.6135505834113346, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 49587 + }, + { + "epoch": 0.49588, + "grad_norm": 1.7489977372119279, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 49588 + }, + { + "epoch": 0.49589, + "grad_norm": 1.6247944761713236, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 49589 + }, + { + "epoch": 0.4959, + "grad_norm": 1.535455583934332, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 49590 + }, + { + "epoch": 0.49591, + "grad_norm": 1.4584658919437328, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 49591 + }, + { + "epoch": 0.49592, + "grad_norm": 1.7041046484902274, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 49592 + }, + { + "epoch": 0.49593, + "grad_norm": 1.5725369799081623, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 49593 + }, + { + "epoch": 0.49594, + "grad_norm": 1.8997331965253976, + "learning_rate": 0.003, + "loss": 4.043, + "step": 49594 + }, + { + "epoch": 0.49595, + "grad_norm": 1.5513385479173316, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 49595 + }, + { + "epoch": 0.49596, + "grad_norm": 1.65387575373182, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 49596 + }, + { + "epoch": 0.49597, + "grad_norm": 1.4015191185886813, + "learning_rate": 0.003, + "loss": 4.033, + "step": 49597 + }, + { + "epoch": 0.49598, + "grad_norm": 1.6830866862680176, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 49598 + }, + { + "epoch": 0.49599, + "grad_norm": 1.3426352822652114, + "learning_rate": 0.003, + "loss": 4.058, + "step": 49599 + }, + { + "epoch": 0.496, + "grad_norm": 1.7513066600675584, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 49600 + }, + { + "epoch": 0.49601, + "grad_norm": 1.2272259770385088, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49601 + }, + { + "epoch": 0.49602, + "grad_norm": 1.7732780765601384, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 49602 + }, + { + "epoch": 0.49603, + "grad_norm": 1.1312828685752685, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49603 + }, + { + "epoch": 0.49604, + "grad_norm": 1.8631799780685365, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 49604 + }, + { + "epoch": 0.49605, + "grad_norm": 1.4281846485701166, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 49605 + }, + { + "epoch": 0.49606, + "grad_norm": 1.6110656608089646, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 49606 + }, + { + "epoch": 0.49607, + "grad_norm": 1.7578790270122941, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 49607 + }, + { + "epoch": 0.49608, + "grad_norm": 1.68323549625065, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 49608 + }, + { + "epoch": 0.49609, + "grad_norm": 1.5559909782628862, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 49609 + }, + { + "epoch": 0.4961, + "grad_norm": 1.7815959841465363, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49610 + }, + { + "epoch": 0.49611, + "grad_norm": 1.3206007656887808, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 49611 + }, + { + "epoch": 0.49612, + "grad_norm": 1.7265005627606975, + "learning_rate": 0.003, + "loss": 4.06, + "step": 49612 + }, + { + "epoch": 0.49613, + "grad_norm": 1.6048059390723568, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49613 + }, + { + "epoch": 0.49614, + "grad_norm": 1.3511370739458375, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 49614 + }, + { + "epoch": 0.49615, + "grad_norm": 1.8350048157034355, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 49615 + }, + { + "epoch": 0.49616, + "grad_norm": 1.407166396684376, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49616 + }, + { + "epoch": 0.49617, + "grad_norm": 1.597811320642859, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49617 + }, + { + "epoch": 0.49618, + "grad_norm": 1.2912503692684711, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 49618 + }, + { + "epoch": 0.49619, + "grad_norm": 1.570447324299475, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49619 + }, + { + "epoch": 0.4962, + "grad_norm": 1.2095896493414824, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49620 + }, + { + "epoch": 0.49621, + "grad_norm": 1.9341741041363543, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49621 + }, + { + "epoch": 0.49622, + "grad_norm": 1.2497201816035883, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 49622 + }, + { + "epoch": 0.49623, + "grad_norm": 1.5159343918797048, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 49623 + }, + { + "epoch": 0.49624, + "grad_norm": 1.798493765684458, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 49624 + }, + { + "epoch": 0.49625, + "grad_norm": 2.1262293358696747, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 49625 + }, + { + "epoch": 0.49626, + "grad_norm": 1.4390825175639976, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 49626 + }, + { + "epoch": 0.49627, + "grad_norm": 1.4345490049349805, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49627 + }, + { + "epoch": 0.49628, + "grad_norm": 1.4178683782407098, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 49628 + }, + { + "epoch": 0.49629, + "grad_norm": 1.657940366194447, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 49629 + }, + { + "epoch": 0.4963, + "grad_norm": 1.3797504895571944, + "learning_rate": 0.003, + "loss": 4.037, + "step": 49630 + }, + { + "epoch": 0.49631, + "grad_norm": 1.5127999080432744, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 49631 + }, + { + "epoch": 0.49632, + "grad_norm": 1.6552680608580221, + "learning_rate": 0.003, + "loss": 4.059, + "step": 49632 + }, + { + "epoch": 0.49633, + "grad_norm": 1.8249704720378164, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49633 + }, + { + "epoch": 0.49634, + "grad_norm": 1.2659581495053835, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 49634 + }, + { + "epoch": 0.49635, + "grad_norm": 1.9913895471654337, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 49635 + }, + { + "epoch": 0.49636, + "grad_norm": 1.3874011366900556, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 49636 + }, + { + "epoch": 0.49637, + "grad_norm": 1.6356675906822549, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 49637 + }, + { + "epoch": 0.49638, + "grad_norm": 1.4829519050451367, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 49638 + }, + { + "epoch": 0.49639, + "grad_norm": 1.4793994987656038, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 49639 + }, + { + "epoch": 0.4964, + "grad_norm": 1.4686113615272915, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 49640 + }, + { + "epoch": 0.49641, + "grad_norm": 1.3532145338511723, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 49641 + }, + { + "epoch": 0.49642, + "grad_norm": 1.894392193553453, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49642 + }, + { + "epoch": 0.49643, + "grad_norm": 1.3682494890034465, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49643 + }, + { + "epoch": 0.49644, + "grad_norm": 1.6540093906812305, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 49644 + }, + { + "epoch": 0.49645, + "grad_norm": 2.0512521009994313, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 49645 + }, + { + "epoch": 0.49646, + "grad_norm": 1.2087499468290808, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 49646 + }, + { + "epoch": 0.49647, + "grad_norm": 1.7348413309543134, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 49647 + }, + { + "epoch": 0.49648, + "grad_norm": 2.0425776814351435, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49648 + }, + { + "epoch": 0.49649, + "grad_norm": 1.2819249711837264, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 49649 + }, + { + "epoch": 0.4965, + "grad_norm": 1.7233938635505897, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 49650 + }, + { + "epoch": 0.49651, + "grad_norm": 1.6263942086856518, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 49651 + }, + { + "epoch": 0.49652, + "grad_norm": 1.2570545062697593, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 49652 + }, + { + "epoch": 0.49653, + "grad_norm": 1.6493404999979617, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 49653 + }, + { + "epoch": 0.49654, + "grad_norm": 1.4900380025571356, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49654 + }, + { + "epoch": 0.49655, + "grad_norm": 1.3304667237571772, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49655 + }, + { + "epoch": 0.49656, + "grad_norm": 1.6652723191893544, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 49656 + }, + { + "epoch": 0.49657, + "grad_norm": 1.2182311462901347, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 49657 + }, + { + "epoch": 0.49658, + "grad_norm": 1.857179694605211, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 49658 + }, + { + "epoch": 0.49659, + "grad_norm": 1.9119230756806354, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 49659 + }, + { + "epoch": 0.4966, + "grad_norm": 1.7794862362034134, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49660 + }, + { + "epoch": 0.49661, + "grad_norm": 1.601256842840564, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 49661 + }, + { + "epoch": 0.49662, + "grad_norm": 1.5886636596579098, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 49662 + }, + { + "epoch": 0.49663, + "grad_norm": 1.455148807987317, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 49663 + }, + { + "epoch": 0.49664, + "grad_norm": 1.5589076915302527, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 49664 + }, + { + "epoch": 0.49665, + "grad_norm": 1.7063088464964975, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 49665 + }, + { + "epoch": 0.49666, + "grad_norm": 1.366216131688519, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 49666 + }, + { + "epoch": 0.49667, + "grad_norm": 1.6001875398934748, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49667 + }, + { + "epoch": 0.49668, + "grad_norm": 1.4862081881958777, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 49668 + }, + { + "epoch": 0.49669, + "grad_norm": 1.490389464779498, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49669 + }, + { + "epoch": 0.4967, + "grad_norm": 1.3571609989347677, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49670 + }, + { + "epoch": 0.49671, + "grad_norm": 1.4734869811408933, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 49671 + }, + { + "epoch": 0.49672, + "grad_norm": 1.546087955092311, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49672 + }, + { + "epoch": 0.49673, + "grad_norm": 1.7813744062970593, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 49673 + }, + { + "epoch": 0.49674, + "grad_norm": 1.2987717744007123, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 49674 + }, + { + "epoch": 0.49675, + "grad_norm": 1.5310350130001522, + "learning_rate": 0.003, + "loss": 4.036, + "step": 49675 + }, + { + "epoch": 0.49676, + "grad_norm": 1.733749583843662, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 49676 + }, + { + "epoch": 0.49677, + "grad_norm": 1.2693527906097133, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 49677 + }, + { + "epoch": 0.49678, + "grad_norm": 1.6335035226683905, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 49678 + }, + { + "epoch": 0.49679, + "grad_norm": 1.394084607479136, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 49679 + }, + { + "epoch": 0.4968, + "grad_norm": 1.671474353100174, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 49680 + }, + { + "epoch": 0.49681, + "grad_norm": 1.6549686019620578, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 49681 + }, + { + "epoch": 0.49682, + "grad_norm": 1.3311797063210877, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 49682 + }, + { + "epoch": 0.49683, + "grad_norm": 2.0645797823901626, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 49683 + }, + { + "epoch": 0.49684, + "grad_norm": 1.4434662577736725, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 49684 + }, + { + "epoch": 0.49685, + "grad_norm": 1.9252485791851588, + "learning_rate": 0.003, + "loss": 4.046, + "step": 49685 + }, + { + "epoch": 0.49686, + "grad_norm": 1.403007412335722, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 49686 + }, + { + "epoch": 0.49687, + "grad_norm": 1.609770425035876, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 49687 + }, + { + "epoch": 0.49688, + "grad_norm": 1.6914432701864468, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 49688 + }, + { + "epoch": 0.49689, + "grad_norm": 1.3702356537313476, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 49689 + }, + { + "epoch": 0.4969, + "grad_norm": 1.808369435887882, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 49690 + }, + { + "epoch": 0.49691, + "grad_norm": 1.5382800617006416, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 49691 + }, + { + "epoch": 0.49692, + "grad_norm": 1.4423208169846617, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 49692 + }, + { + "epoch": 0.49693, + "grad_norm": 1.586177040613612, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49693 + }, + { + "epoch": 0.49694, + "grad_norm": 1.333503307615331, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 49694 + }, + { + "epoch": 0.49695, + "grad_norm": 1.7873032758327032, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 49695 + }, + { + "epoch": 0.49696, + "grad_norm": 1.3777205269538457, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 49696 + }, + { + "epoch": 0.49697, + "grad_norm": 1.595443701749055, + "learning_rate": 0.003, + "loss": 4.071, + "step": 49697 + }, + { + "epoch": 0.49698, + "grad_norm": 1.2131441954350548, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 49698 + }, + { + "epoch": 0.49699, + "grad_norm": 1.6018352600922838, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 49699 + }, + { + "epoch": 0.497, + "grad_norm": 1.2051304238692473, + "learning_rate": 0.003, + "loss": 4.032, + "step": 49700 + }, + { + "epoch": 0.49701, + "grad_norm": 1.7057788587865683, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 49701 + }, + { + "epoch": 0.49702, + "grad_norm": 1.5264626470218603, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49702 + }, + { + "epoch": 0.49703, + "grad_norm": 1.3474945209199616, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49703 + }, + { + "epoch": 0.49704, + "grad_norm": 1.8800450788632923, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 49704 + }, + { + "epoch": 0.49705, + "grad_norm": 1.6944510237733887, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 49705 + }, + { + "epoch": 0.49706, + "grad_norm": 1.5238422342334348, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 49706 + }, + { + "epoch": 0.49707, + "grad_norm": 1.8051946739301183, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 49707 + }, + { + "epoch": 0.49708, + "grad_norm": 1.660534324715588, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 49708 + }, + { + "epoch": 0.49709, + "grad_norm": 1.3917042377375937, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 49709 + }, + { + "epoch": 0.4971, + "grad_norm": 1.6631346043033532, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 49710 + }, + { + "epoch": 0.49711, + "grad_norm": 1.5856039083509865, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 49711 + }, + { + "epoch": 0.49712, + "grad_norm": 1.1729838760298574, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 49712 + }, + { + "epoch": 0.49713, + "grad_norm": 1.805125395360521, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 49713 + }, + { + "epoch": 0.49714, + "grad_norm": 1.4089309519481439, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 49714 + }, + { + "epoch": 0.49715, + "grad_norm": 1.9413389615466872, + "learning_rate": 0.003, + "loss": 4.041, + "step": 49715 + }, + { + "epoch": 0.49716, + "grad_norm": 1.4779428240468655, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 49716 + }, + { + "epoch": 0.49717, + "grad_norm": 1.5075990396359848, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49717 + }, + { + "epoch": 0.49718, + "grad_norm": 1.672284196534643, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 49718 + }, + { + "epoch": 0.49719, + "grad_norm": 1.3730543247870868, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 49719 + }, + { + "epoch": 0.4972, + "grad_norm": 1.4046335338692666, + "learning_rate": 0.003, + "loss": 4.033, + "step": 49720 + }, + { + "epoch": 0.49721, + "grad_norm": 1.5142321569947013, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 49721 + }, + { + "epoch": 0.49722, + "grad_norm": 1.6123869837701643, + "learning_rate": 0.003, + "loss": 4.057, + "step": 49722 + }, + { + "epoch": 0.49723, + "grad_norm": 1.4774463079306646, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49723 + }, + { + "epoch": 0.49724, + "grad_norm": 1.3624273151128787, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 49724 + }, + { + "epoch": 0.49725, + "grad_norm": 1.8265184686520275, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 49725 + }, + { + "epoch": 0.49726, + "grad_norm": 1.6250750432970191, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 49726 + }, + { + "epoch": 0.49727, + "grad_norm": 1.3121104159694366, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 49727 + }, + { + "epoch": 0.49728, + "grad_norm": 1.7410724846117416, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 49728 + }, + { + "epoch": 0.49729, + "grad_norm": 1.341301946780442, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 49729 + }, + { + "epoch": 0.4973, + "grad_norm": 1.6370213585493993, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49730 + }, + { + "epoch": 0.49731, + "grad_norm": 1.7850513461164554, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49731 + }, + { + "epoch": 0.49732, + "grad_norm": 1.6718733933889276, + "learning_rate": 0.003, + "loss": 4.048, + "step": 49732 + }, + { + "epoch": 0.49733, + "grad_norm": 1.8699351518545484, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 49733 + }, + { + "epoch": 0.49734, + "grad_norm": 1.1414630366477656, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 49734 + }, + { + "epoch": 0.49735, + "grad_norm": 1.8118791531943872, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 49735 + }, + { + "epoch": 0.49736, + "grad_norm": 1.490772379834969, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 49736 + }, + { + "epoch": 0.49737, + "grad_norm": 1.5462930147591754, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49737 + }, + { + "epoch": 0.49738, + "grad_norm": 1.5291937385055414, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 49738 + }, + { + "epoch": 0.49739, + "grad_norm": 1.4825579761603793, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 49739 + }, + { + "epoch": 0.4974, + "grad_norm": 1.8124738276287482, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 49740 + }, + { + "epoch": 0.49741, + "grad_norm": 1.3194251855800048, + "learning_rate": 0.003, + "loss": 4.034, + "step": 49741 + }, + { + "epoch": 0.49742, + "grad_norm": 1.691592898947211, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 49742 + }, + { + "epoch": 0.49743, + "grad_norm": 1.5002133475426518, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 49743 + }, + { + "epoch": 0.49744, + "grad_norm": 1.5539008917795383, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 49744 + }, + { + "epoch": 0.49745, + "grad_norm": 1.6709237156596235, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 49745 + }, + { + "epoch": 0.49746, + "grad_norm": 1.059949118620735, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 49746 + }, + { + "epoch": 0.49747, + "grad_norm": 2.014015060110305, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 49747 + }, + { + "epoch": 0.49748, + "grad_norm": 1.4590298719517665, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 49748 + }, + { + "epoch": 0.49749, + "grad_norm": 1.6255941066679533, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 49749 + }, + { + "epoch": 0.4975, + "grad_norm": 1.795592094500784, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 49750 + }, + { + "epoch": 0.49751, + "grad_norm": 1.2647895338668549, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49751 + }, + { + "epoch": 0.49752, + "grad_norm": 1.6083030259479913, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49752 + }, + { + "epoch": 0.49753, + "grad_norm": 1.3158335312505578, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 49753 + }, + { + "epoch": 0.49754, + "grad_norm": 1.8249831777715158, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49754 + }, + { + "epoch": 0.49755, + "grad_norm": 1.2050656712425734, + "learning_rate": 0.003, + "loss": 4.052, + "step": 49755 + }, + { + "epoch": 0.49756, + "grad_norm": 1.7507730015294507, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49756 + }, + { + "epoch": 0.49757, + "grad_norm": 1.5643764150374488, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 49757 + }, + { + "epoch": 0.49758, + "grad_norm": 1.2048541173982217, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 49758 + }, + { + "epoch": 0.49759, + "grad_norm": 1.6950354800150935, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 49759 + }, + { + "epoch": 0.4976, + "grad_norm": 1.3363329968234694, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 49760 + }, + { + "epoch": 0.49761, + "grad_norm": 1.6760676179051246, + "learning_rate": 0.003, + "loss": 4.035, + "step": 49761 + }, + { + "epoch": 0.49762, + "grad_norm": 1.9846983302009313, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 49762 + }, + { + "epoch": 0.49763, + "grad_norm": 1.475764700128327, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 49763 + }, + { + "epoch": 0.49764, + "grad_norm": 1.6481085358940366, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 49764 + }, + { + "epoch": 0.49765, + "grad_norm": 1.350689134026704, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49765 + }, + { + "epoch": 0.49766, + "grad_norm": 1.64120610552503, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 49766 + }, + { + "epoch": 0.49767, + "grad_norm": 1.361871395383905, + "learning_rate": 0.003, + "loss": 4.03, + "step": 49767 + }, + { + "epoch": 0.49768, + "grad_norm": 1.8125197781640923, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 49768 + }, + { + "epoch": 0.49769, + "grad_norm": 1.501141964499383, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49769 + }, + { + "epoch": 0.4977, + "grad_norm": 1.8020527953749699, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 49770 + }, + { + "epoch": 0.49771, + "grad_norm": 1.4036694778037038, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 49771 + }, + { + "epoch": 0.49772, + "grad_norm": 1.6416364075847065, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 49772 + }, + { + "epoch": 0.49773, + "grad_norm": 1.5291375758119716, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 49773 + }, + { + "epoch": 0.49774, + "grad_norm": 1.593198287773714, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 49774 + }, + { + "epoch": 0.49775, + "grad_norm": 1.6856668926936056, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 49775 + }, + { + "epoch": 0.49776, + "grad_norm": 1.487733470063649, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49776 + }, + { + "epoch": 0.49777, + "grad_norm": 1.2349086670658866, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 49777 + }, + { + "epoch": 0.49778, + "grad_norm": 1.5694955841136384, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 49778 + }, + { + "epoch": 0.49779, + "grad_norm": 1.3543692598929449, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49779 + }, + { + "epoch": 0.4978, + "grad_norm": 1.5402210754304204, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 49780 + }, + { + "epoch": 0.49781, + "grad_norm": 1.3849291203216736, + "learning_rate": 0.003, + "loss": 4.037, + "step": 49781 + }, + { + "epoch": 0.49782, + "grad_norm": 1.4421098711545024, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 49782 + }, + { + "epoch": 0.49783, + "grad_norm": 1.6617622336717806, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 49783 + }, + { + "epoch": 0.49784, + "grad_norm": 1.4807030423598193, + "learning_rate": 0.003, + "loss": 4.055, + "step": 49784 + }, + { + "epoch": 0.49785, + "grad_norm": 1.6716654596093126, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 49785 + }, + { + "epoch": 0.49786, + "grad_norm": 1.7994945193625609, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49786 + }, + { + "epoch": 0.49787, + "grad_norm": 1.294370010432009, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 49787 + }, + { + "epoch": 0.49788, + "grad_norm": 1.6475810228975631, + "learning_rate": 0.003, + "loss": 4.06, + "step": 49788 + }, + { + "epoch": 0.49789, + "grad_norm": 1.6095286399773459, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49789 + }, + { + "epoch": 0.4979, + "grad_norm": 1.2378253146987028, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 49790 + }, + { + "epoch": 0.49791, + "grad_norm": 2.0398871112521495, + "learning_rate": 0.003, + "loss": 4.025, + "step": 49791 + }, + { + "epoch": 0.49792, + "grad_norm": 1.2157868764902855, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 49792 + }, + { + "epoch": 0.49793, + "grad_norm": 1.726811629822402, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49793 + }, + { + "epoch": 0.49794, + "grad_norm": 1.4838705860146517, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 49794 + }, + { + "epoch": 0.49795, + "grad_norm": 1.8718289958890082, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49795 + }, + { + "epoch": 0.49796, + "grad_norm": 1.4418652333600146, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49796 + }, + { + "epoch": 0.49797, + "grad_norm": 1.522561059899633, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 49797 + }, + { + "epoch": 0.49798, + "grad_norm": 1.7076394271596358, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49798 + }, + { + "epoch": 0.49799, + "grad_norm": 1.5192196186489728, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 49799 + }, + { + "epoch": 0.498, + "grad_norm": 1.9590122790267381, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 49800 + }, + { + "epoch": 0.49801, + "grad_norm": 1.1110799136592682, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49801 + }, + { + "epoch": 0.49802, + "grad_norm": 1.5872153584212179, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 49802 + }, + { + "epoch": 0.49803, + "grad_norm": 1.733006163102854, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 49803 + }, + { + "epoch": 0.49804, + "grad_norm": 1.4570540361394735, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 49804 + }, + { + "epoch": 0.49805, + "grad_norm": 1.4306111740028484, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 49805 + }, + { + "epoch": 0.49806, + "grad_norm": 1.891656878221809, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 49806 + }, + { + "epoch": 0.49807, + "grad_norm": 1.5058373421417095, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49807 + }, + { + "epoch": 0.49808, + "grad_norm": 1.4797832489433698, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 49808 + }, + { + "epoch": 0.49809, + "grad_norm": 1.8569555180072155, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49809 + }, + { + "epoch": 0.4981, + "grad_norm": 1.3131290409230318, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 49810 + }, + { + "epoch": 0.49811, + "grad_norm": 1.682850353747834, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49811 + }, + { + "epoch": 0.49812, + "grad_norm": 1.3192008597823894, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 49812 + }, + { + "epoch": 0.49813, + "grad_norm": 1.7967646450032855, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49813 + }, + { + "epoch": 0.49814, + "grad_norm": 1.3287537598085581, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 49814 + }, + { + "epoch": 0.49815, + "grad_norm": 1.4767273205513078, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 49815 + }, + { + "epoch": 0.49816, + "grad_norm": 1.571032680065994, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 49816 + }, + { + "epoch": 0.49817, + "grad_norm": 1.5316677575758462, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 49817 + }, + { + "epoch": 0.49818, + "grad_norm": 1.5269199989589928, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 49818 + }, + { + "epoch": 0.49819, + "grad_norm": 1.6823166099368076, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 49819 + }, + { + "epoch": 0.4982, + "grad_norm": 1.69363925901305, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 49820 + }, + { + "epoch": 0.49821, + "grad_norm": 2.119910851371427, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 49821 + }, + { + "epoch": 0.49822, + "grad_norm": 1.608553174357192, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 49822 + }, + { + "epoch": 0.49823, + "grad_norm": 1.6453437211114073, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 49823 + }, + { + "epoch": 0.49824, + "grad_norm": 1.4967906342853872, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 49824 + }, + { + "epoch": 0.49825, + "grad_norm": 1.7678080112193402, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 49825 + }, + { + "epoch": 0.49826, + "grad_norm": 1.5385551261953212, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 49826 + }, + { + "epoch": 0.49827, + "grad_norm": 1.6433329649243043, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 49827 + }, + { + "epoch": 0.49828, + "grad_norm": 1.540836051988412, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 49828 + }, + { + "epoch": 0.49829, + "grad_norm": 1.4208818688018474, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 49829 + }, + { + "epoch": 0.4983, + "grad_norm": 1.5132091192714774, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 49830 + }, + { + "epoch": 0.49831, + "grad_norm": 1.4472513109165628, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 49831 + }, + { + "epoch": 0.49832, + "grad_norm": 1.6204171011036839, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 49832 + }, + { + "epoch": 0.49833, + "grad_norm": 1.4804892897289468, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 49833 + }, + { + "epoch": 0.49834, + "grad_norm": 1.454224362842993, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 49834 + }, + { + "epoch": 0.49835, + "grad_norm": 1.351118979531366, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 49835 + }, + { + "epoch": 0.49836, + "grad_norm": 1.3924769460540447, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 49836 + }, + { + "epoch": 0.49837, + "grad_norm": 1.451199806424407, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49837 + }, + { + "epoch": 0.49838, + "grad_norm": 1.6286822858710701, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 49838 + }, + { + "epoch": 0.49839, + "grad_norm": 1.3299816237596505, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49839 + }, + { + "epoch": 0.4984, + "grad_norm": 1.5717444886031937, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 49840 + }, + { + "epoch": 0.49841, + "grad_norm": 1.5026894004730968, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 49841 + }, + { + "epoch": 0.49842, + "grad_norm": 1.3989076140496417, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 49842 + }, + { + "epoch": 0.49843, + "grad_norm": 1.592394442800309, + "learning_rate": 0.003, + "loss": 4.041, + "step": 49843 + }, + { + "epoch": 0.49844, + "grad_norm": 1.9545212625146924, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 49844 + }, + { + "epoch": 0.49845, + "grad_norm": 1.7849017522235169, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 49845 + }, + { + "epoch": 0.49846, + "grad_norm": 1.6629468071328921, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 49846 + }, + { + "epoch": 0.49847, + "grad_norm": 1.6290247969421288, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49847 + }, + { + "epoch": 0.49848, + "grad_norm": 1.603714821691498, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 49848 + }, + { + "epoch": 0.49849, + "grad_norm": 1.6508872566079589, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 49849 + }, + { + "epoch": 0.4985, + "grad_norm": 1.4880905640817625, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49850 + }, + { + "epoch": 0.49851, + "grad_norm": 1.3769680035993836, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49851 + }, + { + "epoch": 0.49852, + "grad_norm": 1.6027994719552754, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 49852 + }, + { + "epoch": 0.49853, + "grad_norm": 1.3193625937905713, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49853 + }, + { + "epoch": 0.49854, + "grad_norm": 1.745936601345955, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49854 + }, + { + "epoch": 0.49855, + "grad_norm": 1.1518278762621794, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 49855 + }, + { + "epoch": 0.49856, + "grad_norm": 1.6347363636305352, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 49856 + }, + { + "epoch": 0.49857, + "grad_norm": 1.345809319016655, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49857 + }, + { + "epoch": 0.49858, + "grad_norm": 1.7983122917956023, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 49858 + }, + { + "epoch": 0.49859, + "grad_norm": 1.3759673280857923, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 49859 + }, + { + "epoch": 0.4986, + "grad_norm": 1.9739359840272632, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 49860 + }, + { + "epoch": 0.49861, + "grad_norm": 1.4332052087838607, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 49861 + }, + { + "epoch": 0.49862, + "grad_norm": 1.293919302366192, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 49862 + }, + { + "epoch": 0.49863, + "grad_norm": 1.883249643014218, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 49863 + }, + { + "epoch": 0.49864, + "grad_norm": 1.6437459672147712, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49864 + }, + { + "epoch": 0.49865, + "grad_norm": 1.8079448133239255, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 49865 + }, + { + "epoch": 0.49866, + "grad_norm": 1.5930946714977294, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 49866 + }, + { + "epoch": 0.49867, + "grad_norm": 1.5062435124761708, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 49867 + }, + { + "epoch": 0.49868, + "grad_norm": 1.5839818090638424, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49868 + }, + { + "epoch": 0.49869, + "grad_norm": 1.345108802310138, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 49869 + }, + { + "epoch": 0.4987, + "grad_norm": 1.3790707798877937, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 49870 + }, + { + "epoch": 0.49871, + "grad_norm": 1.4190990629318678, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49871 + }, + { + "epoch": 0.49872, + "grad_norm": 1.3994742549167363, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 49872 + }, + { + "epoch": 0.49873, + "grad_norm": 1.429519992178206, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 49873 + }, + { + "epoch": 0.49874, + "grad_norm": 1.6200659193059062, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 49874 + }, + { + "epoch": 0.49875, + "grad_norm": 1.4573522210036267, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 49875 + }, + { + "epoch": 0.49876, + "grad_norm": 1.9487497621627468, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 49876 + }, + { + "epoch": 0.49877, + "grad_norm": 1.9242910164162987, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 49877 + }, + { + "epoch": 0.49878, + "grad_norm": 1.263335789827381, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 49878 + }, + { + "epoch": 0.49879, + "grad_norm": 1.8133346662275114, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 49879 + }, + { + "epoch": 0.4988, + "grad_norm": 1.5560564040315883, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 49880 + }, + { + "epoch": 0.49881, + "grad_norm": 1.6799583238410147, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 49881 + }, + { + "epoch": 0.49882, + "grad_norm": 1.2363587379020906, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 49882 + }, + { + "epoch": 0.49883, + "grad_norm": 1.613546807191773, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 49883 + }, + { + "epoch": 0.49884, + "grad_norm": 1.4253490810036353, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 49884 + }, + { + "epoch": 0.49885, + "grad_norm": 1.6372770528390421, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 49885 + }, + { + "epoch": 0.49886, + "grad_norm": 1.772348595761264, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 49886 + }, + { + "epoch": 0.49887, + "grad_norm": 1.3909422344221296, + "learning_rate": 0.003, + "loss": 4.04, + "step": 49887 + }, + { + "epoch": 0.49888, + "grad_norm": 1.666341719827658, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 49888 + }, + { + "epoch": 0.49889, + "grad_norm": 1.5345108832527856, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 49889 + }, + { + "epoch": 0.4989, + "grad_norm": 1.5246943009064113, + "learning_rate": 0.003, + "loss": 4.04, + "step": 49890 + }, + { + "epoch": 0.49891, + "grad_norm": 1.4324783786003377, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 49891 + }, + { + "epoch": 0.49892, + "grad_norm": 1.6009868507949465, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 49892 + }, + { + "epoch": 0.49893, + "grad_norm": 1.2775714127408186, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 49893 + }, + { + "epoch": 0.49894, + "grad_norm": 1.7821104656071172, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49894 + }, + { + "epoch": 0.49895, + "grad_norm": 1.625881116634322, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 49895 + }, + { + "epoch": 0.49896, + "grad_norm": 1.4525868646589173, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 49896 + }, + { + "epoch": 0.49897, + "grad_norm": 1.4260784481602808, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49897 + }, + { + "epoch": 0.49898, + "grad_norm": 1.4675707892799936, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49898 + }, + { + "epoch": 0.49899, + "grad_norm": 1.548716329071915, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49899 + }, + { + "epoch": 0.499, + "grad_norm": 1.529802258963995, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 49900 + }, + { + "epoch": 0.49901, + "grad_norm": 1.554840150396701, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 49901 + }, + { + "epoch": 0.49902, + "grad_norm": 1.4346895781654787, + "learning_rate": 0.003, + "loss": 4.056, + "step": 49902 + }, + { + "epoch": 0.49903, + "grad_norm": 1.3997905597548916, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 49903 + }, + { + "epoch": 0.49904, + "grad_norm": 1.8945004296157033, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 49904 + }, + { + "epoch": 0.49905, + "grad_norm": 1.978208776818471, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 49905 + }, + { + "epoch": 0.49906, + "grad_norm": 1.5408271439922254, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 49906 + }, + { + "epoch": 0.49907, + "grad_norm": 1.4730225263813685, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 49907 + }, + { + "epoch": 0.49908, + "grad_norm": 1.9644377242525222, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 49908 + }, + { + "epoch": 0.49909, + "grad_norm": 1.1867153426798456, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 49909 + }, + { + "epoch": 0.4991, + "grad_norm": 1.6985240561821406, + "learning_rate": 0.003, + "loss": 4.041, + "step": 49910 + }, + { + "epoch": 0.49911, + "grad_norm": 1.5565644890031856, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 49911 + }, + { + "epoch": 0.49912, + "grad_norm": 1.3180550901334565, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 49912 + }, + { + "epoch": 0.49913, + "grad_norm": 1.7616817163680378, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 49913 + }, + { + "epoch": 0.49914, + "grad_norm": 1.337511450025037, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 49914 + }, + { + "epoch": 0.49915, + "grad_norm": 1.7276056092510117, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 49915 + }, + { + "epoch": 0.49916, + "grad_norm": 1.3736660446497422, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49916 + }, + { + "epoch": 0.49917, + "grad_norm": 1.8314753016807306, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 49917 + }, + { + "epoch": 0.49918, + "grad_norm": 1.3360740681555296, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 49918 + }, + { + "epoch": 0.49919, + "grad_norm": 1.6141065326279134, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 49919 + }, + { + "epoch": 0.4992, + "grad_norm": 1.3600248496640153, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 49920 + }, + { + "epoch": 0.49921, + "grad_norm": 1.9880264089951554, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 49921 + }, + { + "epoch": 0.49922, + "grad_norm": 1.5867886689399149, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 49922 + }, + { + "epoch": 0.49923, + "grad_norm": 1.7143867107408188, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 49923 + }, + { + "epoch": 0.49924, + "grad_norm": 1.5604412378139048, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 49924 + }, + { + "epoch": 0.49925, + "grad_norm": 1.7481753864091858, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49925 + }, + { + "epoch": 0.49926, + "grad_norm": 1.1450166174921892, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 49926 + }, + { + "epoch": 0.49927, + "grad_norm": 1.7107427576715075, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 49927 + }, + { + "epoch": 0.49928, + "grad_norm": 1.4937074727304576, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 49928 + }, + { + "epoch": 0.49929, + "grad_norm": 1.3810308439174628, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 49929 + }, + { + "epoch": 0.4993, + "grad_norm": 1.7833184396138586, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 49930 + }, + { + "epoch": 0.49931, + "grad_norm": 1.3759840301204296, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 49931 + }, + { + "epoch": 0.49932, + "grad_norm": 1.7925314070889966, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 49932 + }, + { + "epoch": 0.49933, + "grad_norm": 1.5423310205641512, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 49933 + }, + { + "epoch": 0.49934, + "grad_norm": 1.3821878441443396, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 49934 + }, + { + "epoch": 0.49935, + "grad_norm": 1.7543043438058232, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 49935 + }, + { + "epoch": 0.49936, + "grad_norm": 1.283002306674664, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 49936 + }, + { + "epoch": 0.49937, + "grad_norm": 1.8370281997228581, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 49937 + }, + { + "epoch": 0.49938, + "grad_norm": 1.4568798466746393, + "learning_rate": 0.003, + "loss": 4.04, + "step": 49938 + }, + { + "epoch": 0.49939, + "grad_norm": 1.5612120826128304, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 49939 + }, + { + "epoch": 0.4994, + "grad_norm": 1.456202386859332, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 49940 + }, + { + "epoch": 0.49941, + "grad_norm": 1.6018377915882818, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 49941 + }, + { + "epoch": 0.49942, + "grad_norm": 1.703559655720374, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 49942 + }, + { + "epoch": 0.49943, + "grad_norm": 1.3615911398585587, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 49943 + }, + { + "epoch": 0.49944, + "grad_norm": 1.5524246906858028, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 49944 + }, + { + "epoch": 0.49945, + "grad_norm": 1.5839301955230902, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 49945 + }, + { + "epoch": 0.49946, + "grad_norm": 1.488661242435424, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 49946 + }, + { + "epoch": 0.49947, + "grad_norm": 1.2980298465804607, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 49947 + }, + { + "epoch": 0.49948, + "grad_norm": 1.7357923634728247, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 49948 + }, + { + "epoch": 0.49949, + "grad_norm": 1.3944133155533278, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 49949 + }, + { + "epoch": 0.4995, + "grad_norm": 1.5902607833065774, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 49950 + }, + { + "epoch": 0.49951, + "grad_norm": 1.360737399889508, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 49951 + }, + { + "epoch": 0.49952, + "grad_norm": 2.019837903217998, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 49952 + }, + { + "epoch": 0.49953, + "grad_norm": 1.2031859477395752, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 49953 + }, + { + "epoch": 0.49954, + "grad_norm": 1.9208460844913646, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 49954 + }, + { + "epoch": 0.49955, + "grad_norm": 1.6223772933258824, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 49955 + }, + { + "epoch": 0.49956, + "grad_norm": 1.7786269284259681, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 49956 + }, + { + "epoch": 0.49957, + "grad_norm": 1.396918832491622, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 49957 + }, + { + "epoch": 0.49958, + "grad_norm": 1.5979397059405958, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 49958 + }, + { + "epoch": 0.49959, + "grad_norm": 1.7688220298893722, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 49959 + }, + { + "epoch": 0.4996, + "grad_norm": 1.3361806453859348, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 49960 + }, + { + "epoch": 0.49961, + "grad_norm": 1.79067044471536, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 49961 + }, + { + "epoch": 0.49962, + "grad_norm": 1.2492863351520438, + "learning_rate": 0.003, + "loss": 4.024, + "step": 49962 + }, + { + "epoch": 0.49963, + "grad_norm": 1.7463780355720593, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 49963 + }, + { + "epoch": 0.49964, + "grad_norm": 1.6086144217560479, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 49964 + }, + { + "epoch": 0.49965, + "grad_norm": 1.2340481494630222, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 49965 + }, + { + "epoch": 0.49966, + "grad_norm": 1.7738325768587473, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 49966 + }, + { + "epoch": 0.49967, + "grad_norm": 1.5251487897570366, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 49967 + }, + { + "epoch": 0.49968, + "grad_norm": 1.497033473319357, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 49968 + }, + { + "epoch": 0.49969, + "grad_norm": 1.5585601007406655, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 49969 + }, + { + "epoch": 0.4997, + "grad_norm": 1.4538402972137892, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 49970 + }, + { + "epoch": 0.49971, + "grad_norm": 1.6177774565256486, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 49971 + }, + { + "epoch": 0.49972, + "grad_norm": 1.8093384620392445, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 49972 + }, + { + "epoch": 0.49973, + "grad_norm": 1.275080876990211, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 49973 + }, + { + "epoch": 0.49974, + "grad_norm": 1.8316113805374714, + "learning_rate": 0.003, + "loss": 4.055, + "step": 49974 + }, + { + "epoch": 0.49975, + "grad_norm": 1.5045361230612655, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 49975 + }, + { + "epoch": 0.49976, + "grad_norm": 1.3323907169872733, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 49976 + }, + { + "epoch": 0.49977, + "grad_norm": 1.4875874503210629, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 49977 + }, + { + "epoch": 0.49978, + "grad_norm": 1.6639503368124684, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 49978 + }, + { + "epoch": 0.49979, + "grad_norm": 1.5393888780419114, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 49979 + }, + { + "epoch": 0.4998, + "grad_norm": 1.1787627504500917, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 49980 + }, + { + "epoch": 0.49981, + "grad_norm": 1.6838211446263631, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 49981 + }, + { + "epoch": 0.49982, + "grad_norm": 1.259481789683076, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 49982 + }, + { + "epoch": 0.49983, + "grad_norm": 1.7061580291393197, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 49983 + }, + { + "epoch": 0.49984, + "grad_norm": 1.65660260718547, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 49984 + }, + { + "epoch": 0.49985, + "grad_norm": 1.8618843637745062, + "learning_rate": 0.003, + "loss": 4.025, + "step": 49985 + }, + { + "epoch": 0.49986, + "grad_norm": 1.7179383320668487, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 49986 + }, + { + "epoch": 0.49987, + "grad_norm": 1.3685525959590887, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 49987 + }, + { + "epoch": 0.49988, + "grad_norm": 1.7263938417354359, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 49988 + }, + { + "epoch": 0.49989, + "grad_norm": 1.4080936547013523, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 49989 + }, + { + "epoch": 0.4999, + "grad_norm": 1.6044749313686446, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 49990 + }, + { + "epoch": 0.49991, + "grad_norm": 1.5018718433781089, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 49991 + }, + { + "epoch": 0.49992, + "grad_norm": 1.6337626425843235, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 49992 + }, + { + "epoch": 0.49993, + "grad_norm": 1.3693239774390553, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 49993 + }, + { + "epoch": 0.49994, + "grad_norm": 1.6216803466587482, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 49994 + }, + { + "epoch": 0.49995, + "grad_norm": 1.5493120091603345, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 49995 + }, + { + "epoch": 0.49996, + "grad_norm": 1.650900410143719, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 49996 + }, + { + "epoch": 0.49997, + "grad_norm": 1.493871025527561, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49997 + }, + { + "epoch": 0.49998, + "grad_norm": 1.569943576795394, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 49998 + }, + { + "epoch": 0.49999, + "grad_norm": 1.7778621354571853, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 49999 + }, + { + "epoch": 0.5, + "grad_norm": 1.83594982590937, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 50000 + }, + { + "epoch": 0.50001, + "grad_norm": 1.6527507917159538, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 50001 + }, + { + "epoch": 0.50002, + "grad_norm": 1.6585299126213258, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 50002 + }, + { + "epoch": 0.50003, + "grad_norm": 1.3375509249645368, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 50003 + }, + { + "epoch": 0.50004, + "grad_norm": 1.8376300985138967, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 50004 + }, + { + "epoch": 0.50005, + "grad_norm": 1.5663766257672531, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 50005 + }, + { + "epoch": 0.50006, + "grad_norm": 1.4314446782454133, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 50006 + }, + { + "epoch": 0.50007, + "grad_norm": 1.6196386961771345, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 50007 + }, + { + "epoch": 0.50008, + "grad_norm": 1.3184814115461116, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 50008 + }, + { + "epoch": 0.50009, + "grad_norm": 1.6527159352759764, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 50009 + }, + { + "epoch": 0.5001, + "grad_norm": 1.5136161845191676, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 50010 + }, + { + "epoch": 0.50011, + "grad_norm": 1.609759866833813, + "learning_rate": 0.003, + "loss": 4.018, + "step": 50011 + }, + { + "epoch": 0.50012, + "grad_norm": 1.4354019362705994, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 50012 + }, + { + "epoch": 0.50013, + "grad_norm": 1.5501911902674395, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 50013 + }, + { + "epoch": 0.50014, + "grad_norm": 1.815942860854829, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 50014 + }, + { + "epoch": 0.50015, + "grad_norm": 1.2317257133524087, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 50015 + }, + { + "epoch": 0.50016, + "grad_norm": 2.0133518797455716, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 50016 + }, + { + "epoch": 0.50017, + "grad_norm": 1.360936747219499, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 50017 + }, + { + "epoch": 0.50018, + "grad_norm": 1.6412794935575528, + "learning_rate": 0.003, + "loss": 4.03, + "step": 50018 + }, + { + "epoch": 0.50019, + "grad_norm": 1.4946713143071118, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50019 + }, + { + "epoch": 0.5002, + "grad_norm": 1.8825805686211976, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50020 + }, + { + "epoch": 0.50021, + "grad_norm": 1.5410717011767086, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 50021 + }, + { + "epoch": 0.50022, + "grad_norm": 1.7247102698086645, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50022 + }, + { + "epoch": 0.50023, + "grad_norm": 1.3305998294430734, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50023 + }, + { + "epoch": 0.50024, + "grad_norm": 1.608686260882074, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50024 + }, + { + "epoch": 0.50025, + "grad_norm": 1.2816460340807287, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 50025 + }, + { + "epoch": 0.50026, + "grad_norm": 1.6901033872350382, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 50026 + }, + { + "epoch": 0.50027, + "grad_norm": 1.3699872512943065, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50027 + }, + { + "epoch": 0.50028, + "grad_norm": 1.3883664168121506, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 50028 + }, + { + "epoch": 0.50029, + "grad_norm": 1.7237697386975315, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50029 + }, + { + "epoch": 0.5003, + "grad_norm": 1.7236012222722719, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 50030 + }, + { + "epoch": 0.50031, + "grad_norm": 1.8915549468216226, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50031 + }, + { + "epoch": 0.50032, + "grad_norm": 1.4361579811947136, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 50032 + }, + { + "epoch": 0.50033, + "grad_norm": 1.3608960852333154, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 50033 + }, + { + "epoch": 0.50034, + "grad_norm": 1.6691049063773076, + "learning_rate": 0.003, + "loss": 4.05, + "step": 50034 + }, + { + "epoch": 0.50035, + "grad_norm": 1.4542778515126766, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 50035 + }, + { + "epoch": 0.50036, + "grad_norm": 1.781208074669046, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50036 + }, + { + "epoch": 0.50037, + "grad_norm": 1.5337665732067831, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 50037 + }, + { + "epoch": 0.50038, + "grad_norm": 1.6794953362295242, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50038 + }, + { + "epoch": 0.50039, + "grad_norm": 1.5864330607309434, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50039 + }, + { + "epoch": 0.5004, + "grad_norm": 1.5751089623045043, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 50040 + }, + { + "epoch": 0.50041, + "grad_norm": 1.39279283673777, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 50041 + }, + { + "epoch": 0.50042, + "grad_norm": 1.2815303710680397, + "learning_rate": 0.003, + "loss": 4.048, + "step": 50042 + }, + { + "epoch": 0.50043, + "grad_norm": 1.4971347316514485, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 50043 + }, + { + "epoch": 0.50044, + "grad_norm": 1.412773645427403, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 50044 + }, + { + "epoch": 0.50045, + "grad_norm": 1.630097529430877, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50045 + }, + { + "epoch": 0.50046, + "grad_norm": 1.415443176581007, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 50046 + }, + { + "epoch": 0.50047, + "grad_norm": 1.7762447326122104, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 50047 + }, + { + "epoch": 0.50048, + "grad_norm": 1.7334046176408957, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 50048 + }, + { + "epoch": 0.50049, + "grad_norm": 1.6171755586762226, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 50049 + }, + { + "epoch": 0.5005, + "grad_norm": 1.8406315909090132, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 50050 + }, + { + "epoch": 0.50051, + "grad_norm": 1.5406355788318558, + "learning_rate": 0.003, + "loss": 4.038, + "step": 50051 + }, + { + "epoch": 0.50052, + "grad_norm": 1.362167964787047, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 50052 + }, + { + "epoch": 0.50053, + "grad_norm": 1.6073042949045333, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 50053 + }, + { + "epoch": 0.50054, + "grad_norm": 1.5214858134062894, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 50054 + }, + { + "epoch": 0.50055, + "grad_norm": 1.558188136075282, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50055 + }, + { + "epoch": 0.50056, + "grad_norm": 1.8134995212153577, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 50056 + }, + { + "epoch": 0.50057, + "grad_norm": 1.3745079537433658, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 50057 + }, + { + "epoch": 0.50058, + "grad_norm": 1.5757264435366405, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 50058 + }, + { + "epoch": 0.50059, + "grad_norm": 1.539255821853969, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 50059 + }, + { + "epoch": 0.5006, + "grad_norm": 1.489671121608878, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 50060 + }, + { + "epoch": 0.50061, + "grad_norm": 1.608001099834493, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 50061 + }, + { + "epoch": 0.50062, + "grad_norm": 1.4928066275248235, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50062 + }, + { + "epoch": 0.50063, + "grad_norm": 1.822099639668562, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50063 + }, + { + "epoch": 0.50064, + "grad_norm": 1.6834532643586082, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 50064 + }, + { + "epoch": 0.50065, + "grad_norm": 1.6050033815362863, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50065 + }, + { + "epoch": 0.50066, + "grad_norm": 1.7731625083109257, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 50066 + }, + { + "epoch": 0.50067, + "grad_norm": 1.487904625359932, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 50067 + }, + { + "epoch": 0.50068, + "grad_norm": 1.5145297589809308, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 50068 + }, + { + "epoch": 0.50069, + "grad_norm": 1.8301251705040478, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50069 + }, + { + "epoch": 0.5007, + "grad_norm": 1.4543394083481493, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50070 + }, + { + "epoch": 0.50071, + "grad_norm": 1.5927494439956877, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 50071 + }, + { + "epoch": 0.50072, + "grad_norm": 1.2306465226102667, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 50072 + }, + { + "epoch": 0.50073, + "grad_norm": 1.6405998552614895, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 50073 + }, + { + "epoch": 0.50074, + "grad_norm": 1.5843691110323548, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 50074 + }, + { + "epoch": 0.50075, + "grad_norm": 1.579630541337454, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 50075 + }, + { + "epoch": 0.50076, + "grad_norm": 2.174612648945289, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 50076 + }, + { + "epoch": 0.50077, + "grad_norm": 1.2222224623983418, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 50077 + }, + { + "epoch": 0.50078, + "grad_norm": 1.633500567816068, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 50078 + }, + { + "epoch": 0.50079, + "grad_norm": 1.6924336875201247, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 50079 + }, + { + "epoch": 0.5008, + "grad_norm": 1.6857901485509963, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 50080 + }, + { + "epoch": 0.50081, + "grad_norm": 1.295251908655567, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 50081 + }, + { + "epoch": 0.50082, + "grad_norm": 1.5577348679849656, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 50082 + }, + { + "epoch": 0.50083, + "grad_norm": 1.4059921946238156, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50083 + }, + { + "epoch": 0.50084, + "grad_norm": 1.5915795327511988, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 50084 + }, + { + "epoch": 0.50085, + "grad_norm": 1.4932175096739302, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 50085 + }, + { + "epoch": 0.50086, + "grad_norm": 1.6826706495141721, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 50086 + }, + { + "epoch": 0.50087, + "grad_norm": 1.3375223202953688, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 50087 + }, + { + "epoch": 0.50088, + "grad_norm": 1.44660177540386, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50088 + }, + { + "epoch": 0.50089, + "grad_norm": 1.88034722553256, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50089 + }, + { + "epoch": 0.5009, + "grad_norm": 1.5475854017134651, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50090 + }, + { + "epoch": 0.50091, + "grad_norm": 2.0330149814666725, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 50091 + }, + { + "epoch": 0.50092, + "grad_norm": 1.1173513557621644, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 50092 + }, + { + "epoch": 0.50093, + "grad_norm": 1.8383145956455624, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50093 + }, + { + "epoch": 0.50094, + "grad_norm": 1.8397052798431661, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50094 + }, + { + "epoch": 0.50095, + "grad_norm": 1.4668344480002975, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50095 + }, + { + "epoch": 0.50096, + "grad_norm": 1.413888500430641, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 50096 + }, + { + "epoch": 0.50097, + "grad_norm": 1.4832219125720374, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50097 + }, + { + "epoch": 0.50098, + "grad_norm": 1.5140264075639622, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 50098 + }, + { + "epoch": 0.50099, + "grad_norm": 1.3991220570613017, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 50099 + }, + { + "epoch": 0.501, + "grad_norm": 1.7141120705454675, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 50100 + }, + { + "epoch": 0.50101, + "grad_norm": 1.3854169161479384, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 50101 + }, + { + "epoch": 0.50102, + "grad_norm": 1.4561051328483507, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 50102 + }, + { + "epoch": 0.50103, + "grad_norm": 1.8931946365080248, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 50103 + }, + { + "epoch": 0.50104, + "grad_norm": 1.5948549518318873, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50104 + }, + { + "epoch": 0.50105, + "grad_norm": 1.6066143557299808, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50105 + }, + { + "epoch": 0.50106, + "grad_norm": 1.6155598892479812, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 50106 + }, + { + "epoch": 0.50107, + "grad_norm": 1.3709716160870673, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 50107 + }, + { + "epoch": 0.50108, + "grad_norm": 1.9510110288953113, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 50108 + }, + { + "epoch": 0.50109, + "grad_norm": 1.3208094046927452, + "learning_rate": 0.003, + "loss": 4.066, + "step": 50109 + }, + { + "epoch": 0.5011, + "grad_norm": 1.6656432771639955, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 50110 + }, + { + "epoch": 0.50111, + "grad_norm": 1.3820467619306225, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50111 + }, + { + "epoch": 0.50112, + "grad_norm": 1.8738974819074519, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 50112 + }, + { + "epoch": 0.50113, + "grad_norm": 1.5616916026668093, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50113 + }, + { + "epoch": 0.50114, + "grad_norm": 1.5152781264376827, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50114 + }, + { + "epoch": 0.50115, + "grad_norm": 1.7343949243671164, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50115 + }, + { + "epoch": 0.50116, + "grad_norm": 1.4898406377253746, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 50116 + }, + { + "epoch": 0.50117, + "grad_norm": 1.5689167198664384, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50117 + }, + { + "epoch": 0.50118, + "grad_norm": 1.678589115941911, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 50118 + }, + { + "epoch": 0.50119, + "grad_norm": 1.8200125396039857, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 50119 + }, + { + "epoch": 0.5012, + "grad_norm": 1.197891407745692, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50120 + }, + { + "epoch": 0.50121, + "grad_norm": 1.6141151292743567, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 50121 + }, + { + "epoch": 0.50122, + "grad_norm": 1.4400440230163667, + "learning_rate": 0.003, + "loss": 4.055, + "step": 50122 + }, + { + "epoch": 0.50123, + "grad_norm": 1.3777292167487436, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50123 + }, + { + "epoch": 0.50124, + "grad_norm": 1.8468489373455501, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 50124 + }, + { + "epoch": 0.50125, + "grad_norm": 1.1684460213390464, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 50125 + }, + { + "epoch": 0.50126, + "grad_norm": 2.124672715426535, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 50126 + }, + { + "epoch": 0.50127, + "grad_norm": 1.4452445240697802, + "learning_rate": 0.003, + "loss": 4.024, + "step": 50127 + }, + { + "epoch": 0.50128, + "grad_norm": 1.5886691764334837, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 50128 + }, + { + "epoch": 0.50129, + "grad_norm": 1.9411723586203804, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 50129 + }, + { + "epoch": 0.5013, + "grad_norm": 1.2470261953960358, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50130 + }, + { + "epoch": 0.50131, + "grad_norm": 1.7123772997346673, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 50131 + }, + { + "epoch": 0.50132, + "grad_norm": 1.4820918041313718, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 50132 + }, + { + "epoch": 0.50133, + "grad_norm": 1.5205611222566826, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 50133 + }, + { + "epoch": 0.50134, + "grad_norm": 1.8065595088570126, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 50134 + }, + { + "epoch": 0.50135, + "grad_norm": 1.2873449935352095, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50135 + }, + { + "epoch": 0.50136, + "grad_norm": 1.4689708303108704, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50136 + }, + { + "epoch": 0.50137, + "grad_norm": 1.805636996617877, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 50137 + }, + { + "epoch": 0.50138, + "grad_norm": 1.237715585626224, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 50138 + }, + { + "epoch": 0.50139, + "grad_norm": 1.693403891377815, + "learning_rate": 0.003, + "loss": 4.045, + "step": 50139 + }, + { + "epoch": 0.5014, + "grad_norm": 1.3244289637304472, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 50140 + }, + { + "epoch": 0.50141, + "grad_norm": 1.787208188849642, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 50141 + }, + { + "epoch": 0.50142, + "grad_norm": 1.9681970843263907, + "learning_rate": 0.003, + "loss": 4.02, + "step": 50142 + }, + { + "epoch": 0.50143, + "grad_norm": 1.135445529165955, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 50143 + }, + { + "epoch": 0.50144, + "grad_norm": 1.6839957959659286, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50144 + }, + { + "epoch": 0.50145, + "grad_norm": 1.6404151956568074, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50145 + }, + { + "epoch": 0.50146, + "grad_norm": 1.6190701241937873, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 50146 + }, + { + "epoch": 0.50147, + "grad_norm": 1.3544648900689191, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 50147 + }, + { + "epoch": 0.50148, + "grad_norm": 1.4742101015088658, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50148 + }, + { + "epoch": 0.50149, + "grad_norm": 1.464255279750956, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50149 + }, + { + "epoch": 0.5015, + "grad_norm": 1.568902435967188, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 50150 + }, + { + "epoch": 0.50151, + "grad_norm": 1.6877413741985903, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 50151 + }, + { + "epoch": 0.50152, + "grad_norm": 1.352001240029271, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50152 + }, + { + "epoch": 0.50153, + "grad_norm": 1.7814533415778946, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 50153 + }, + { + "epoch": 0.50154, + "grad_norm": 1.7755416289145778, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 50154 + }, + { + "epoch": 0.50155, + "grad_norm": 1.394727294473617, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 50155 + }, + { + "epoch": 0.50156, + "grad_norm": 1.8938607542029224, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 50156 + }, + { + "epoch": 0.50157, + "grad_norm": 1.6673983291783843, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 50157 + }, + { + "epoch": 0.50158, + "grad_norm": 1.3792226845339453, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 50158 + }, + { + "epoch": 0.50159, + "grad_norm": 1.8001822085032082, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 50159 + }, + { + "epoch": 0.5016, + "grad_norm": 1.5271284177196076, + "learning_rate": 0.003, + "loss": 4.038, + "step": 50160 + }, + { + "epoch": 0.50161, + "grad_norm": 1.672866406099217, + "learning_rate": 0.003, + "loss": 4.024, + "step": 50161 + }, + { + "epoch": 0.50162, + "grad_norm": 1.8123368677888514, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 50162 + }, + { + "epoch": 0.50163, + "grad_norm": 1.164925299210758, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 50163 + }, + { + "epoch": 0.50164, + "grad_norm": 1.618607878954855, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 50164 + }, + { + "epoch": 0.50165, + "grad_norm": 1.453069585810731, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 50165 + }, + { + "epoch": 0.50166, + "grad_norm": 1.5126810395728414, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50166 + }, + { + "epoch": 0.50167, + "grad_norm": 1.9597858749340873, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 50167 + }, + { + "epoch": 0.50168, + "grad_norm": 1.3351145976478491, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 50168 + }, + { + "epoch": 0.50169, + "grad_norm": 1.617450449193655, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 50169 + }, + { + "epoch": 0.5017, + "grad_norm": 1.4416697888495198, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50170 + }, + { + "epoch": 0.50171, + "grad_norm": 1.6732274609289723, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 50171 + }, + { + "epoch": 0.50172, + "grad_norm": 1.656148592688793, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 50172 + }, + { + "epoch": 0.50173, + "grad_norm": 1.4268350095293403, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50173 + }, + { + "epoch": 0.50174, + "grad_norm": 1.7109250195707493, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 50174 + }, + { + "epoch": 0.50175, + "grad_norm": 1.4844866271566137, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50175 + }, + { + "epoch": 0.50176, + "grad_norm": 1.493055151722354, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 50176 + }, + { + "epoch": 0.50177, + "grad_norm": 1.6798308629324887, + "learning_rate": 0.003, + "loss": 4.071, + "step": 50177 + }, + { + "epoch": 0.50178, + "grad_norm": 1.4382470254591107, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 50178 + }, + { + "epoch": 0.50179, + "grad_norm": 1.601641249860687, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 50179 + }, + { + "epoch": 0.5018, + "grad_norm": 1.397172973572369, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 50180 + }, + { + "epoch": 0.50181, + "grad_norm": 1.5482816539048647, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 50181 + }, + { + "epoch": 0.50182, + "grad_norm": 1.4560250492303075, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 50182 + }, + { + "epoch": 0.50183, + "grad_norm": 1.727908137400134, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 50183 + }, + { + "epoch": 0.50184, + "grad_norm": 1.842222500844111, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 50184 + }, + { + "epoch": 0.50185, + "grad_norm": 1.3551003579207415, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 50185 + }, + { + "epoch": 0.50186, + "grad_norm": 1.7873367999302596, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 50186 + }, + { + "epoch": 0.50187, + "grad_norm": 1.3973732232642535, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50187 + }, + { + "epoch": 0.50188, + "grad_norm": 1.7421865364249594, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 50188 + }, + { + "epoch": 0.50189, + "grad_norm": 1.3141889328874239, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 50189 + }, + { + "epoch": 0.5019, + "grad_norm": 1.5261733142492786, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 50190 + }, + { + "epoch": 0.50191, + "grad_norm": 1.4867167946816096, + "learning_rate": 0.003, + "loss": 4.067, + "step": 50191 + }, + { + "epoch": 0.50192, + "grad_norm": 1.7834797632433346, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 50192 + }, + { + "epoch": 0.50193, + "grad_norm": 1.3391517237516233, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50193 + }, + { + "epoch": 0.50194, + "grad_norm": 1.5408205620989168, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 50194 + }, + { + "epoch": 0.50195, + "grad_norm": 1.4227564499984846, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 50195 + }, + { + "epoch": 0.50196, + "grad_norm": 1.5278455075816955, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50196 + }, + { + "epoch": 0.50197, + "grad_norm": 1.4236880660017537, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50197 + }, + { + "epoch": 0.50198, + "grad_norm": 1.5972557716619489, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50198 + }, + { + "epoch": 0.50199, + "grad_norm": 1.37619608107933, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50199 + }, + { + "epoch": 0.502, + "grad_norm": 1.7534366123921203, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 50200 + }, + { + "epoch": 0.50201, + "grad_norm": 1.4028299604002756, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 50201 + }, + { + "epoch": 0.50202, + "grad_norm": 2.1617463085582354, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 50202 + }, + { + "epoch": 0.50203, + "grad_norm": 1.4947471679078643, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50203 + }, + { + "epoch": 0.50204, + "grad_norm": 1.5723493194225509, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 50204 + }, + { + "epoch": 0.50205, + "grad_norm": 1.4219580292778902, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50205 + }, + { + "epoch": 0.50206, + "grad_norm": 1.604428465898814, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 50206 + }, + { + "epoch": 0.50207, + "grad_norm": 1.6636418275868843, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50207 + }, + { + "epoch": 0.50208, + "grad_norm": 1.3238873419473853, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 50208 + }, + { + "epoch": 0.50209, + "grad_norm": 1.9263258000447994, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 50209 + }, + { + "epoch": 0.5021, + "grad_norm": 1.392562688063667, + "learning_rate": 0.003, + "loss": 4.009, + "step": 50210 + }, + { + "epoch": 0.50211, + "grad_norm": 1.621064200165496, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50211 + }, + { + "epoch": 0.50212, + "grad_norm": 1.6394538709724962, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50212 + }, + { + "epoch": 0.50213, + "grad_norm": 1.527359696668427, + "learning_rate": 0.003, + "loss": 4.007, + "step": 50213 + }, + { + "epoch": 0.50214, + "grad_norm": 1.4864611037701732, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 50214 + }, + { + "epoch": 0.50215, + "grad_norm": 1.6481241431143836, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50215 + }, + { + "epoch": 0.50216, + "grad_norm": 1.4720484140175625, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 50216 + }, + { + "epoch": 0.50217, + "grad_norm": 1.6491971824056775, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 50217 + }, + { + "epoch": 0.50218, + "grad_norm": 1.7428621310297399, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 50218 + }, + { + "epoch": 0.50219, + "grad_norm": 1.258713401549577, + "learning_rate": 0.003, + "loss": 4.013, + "step": 50219 + }, + { + "epoch": 0.5022, + "grad_norm": 1.5311314895961394, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 50220 + }, + { + "epoch": 0.50221, + "grad_norm": 1.6947444889431598, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 50221 + }, + { + "epoch": 0.50222, + "grad_norm": 1.587981784493932, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50222 + }, + { + "epoch": 0.50223, + "grad_norm": 1.6846200078223998, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 50223 + }, + { + "epoch": 0.50224, + "grad_norm": 1.6169903649958648, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50224 + }, + { + "epoch": 0.50225, + "grad_norm": 1.8830083797902415, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50225 + }, + { + "epoch": 0.50226, + "grad_norm": 1.5495922613894049, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 50226 + }, + { + "epoch": 0.50227, + "grad_norm": 1.8986914355818143, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 50227 + }, + { + "epoch": 0.50228, + "grad_norm": 1.4122466522778685, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 50228 + }, + { + "epoch": 0.50229, + "grad_norm": 1.486001321108198, + "learning_rate": 0.003, + "loss": 4.039, + "step": 50229 + }, + { + "epoch": 0.5023, + "grad_norm": 1.4829819749582427, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50230 + }, + { + "epoch": 0.50231, + "grad_norm": 1.4818489377976602, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 50231 + }, + { + "epoch": 0.50232, + "grad_norm": 1.4259320086218763, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50232 + }, + { + "epoch": 0.50233, + "grad_norm": 1.6961696914349527, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50233 + }, + { + "epoch": 0.50234, + "grad_norm": 1.4726986455851718, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 50234 + }, + { + "epoch": 0.50235, + "grad_norm": 1.6407870274522578, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 50235 + }, + { + "epoch": 0.50236, + "grad_norm": 1.6159480384645164, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 50236 + }, + { + "epoch": 0.50237, + "grad_norm": 1.8394429270272374, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 50237 + }, + { + "epoch": 0.50238, + "grad_norm": 1.777098986702753, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50238 + }, + { + "epoch": 0.50239, + "grad_norm": 1.2751516485822667, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 50239 + }, + { + "epoch": 0.5024, + "grad_norm": 1.689734783433608, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 50240 + }, + { + "epoch": 0.50241, + "grad_norm": 1.672899339209667, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50241 + }, + { + "epoch": 0.50242, + "grad_norm": 1.5955089476096755, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 50242 + }, + { + "epoch": 0.50243, + "grad_norm": 1.386003906498695, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 50243 + }, + { + "epoch": 0.50244, + "grad_norm": 1.6832411714137714, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 50244 + }, + { + "epoch": 0.50245, + "grad_norm": 1.2332466285275545, + "learning_rate": 0.003, + "loss": 4.014, + "step": 50245 + }, + { + "epoch": 0.50246, + "grad_norm": 1.8223824499973524, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 50246 + }, + { + "epoch": 0.50247, + "grad_norm": 1.2783134810637933, + "learning_rate": 0.003, + "loss": 4.018, + "step": 50247 + }, + { + "epoch": 0.50248, + "grad_norm": 1.5628761342167665, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 50248 + }, + { + "epoch": 0.50249, + "grad_norm": 1.4207049219500485, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 50249 + }, + { + "epoch": 0.5025, + "grad_norm": 1.634316415053952, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 50250 + }, + { + "epoch": 0.50251, + "grad_norm": 1.4012817285797885, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 50251 + }, + { + "epoch": 0.50252, + "grad_norm": 1.6877600065099692, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 50252 + }, + { + "epoch": 0.50253, + "grad_norm": 1.7008831095170023, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 50253 + }, + { + "epoch": 0.50254, + "grad_norm": 1.8179095023195704, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 50254 + }, + { + "epoch": 0.50255, + "grad_norm": 1.6597380654574678, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50255 + }, + { + "epoch": 0.50256, + "grad_norm": 1.7561331020874211, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50256 + }, + { + "epoch": 0.50257, + "grad_norm": 1.6597072393609011, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 50257 + }, + { + "epoch": 0.50258, + "grad_norm": 1.2963673074498567, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 50258 + }, + { + "epoch": 0.50259, + "grad_norm": 1.3944208512171277, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50259 + }, + { + "epoch": 0.5026, + "grad_norm": 1.6159660381784495, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 50260 + }, + { + "epoch": 0.50261, + "grad_norm": 1.5148674996396345, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50261 + }, + { + "epoch": 0.50262, + "grad_norm": 1.6704077374009656, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 50262 + }, + { + "epoch": 0.50263, + "grad_norm": 1.5575492260029014, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 50263 + }, + { + "epoch": 0.50264, + "grad_norm": 1.484715512459765, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 50264 + }, + { + "epoch": 0.50265, + "grad_norm": 1.3585124232239913, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 50265 + }, + { + "epoch": 0.50266, + "grad_norm": 1.7337509231026518, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50266 + }, + { + "epoch": 0.50267, + "grad_norm": 1.4675268968036204, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 50267 + }, + { + "epoch": 0.50268, + "grad_norm": 1.8470672495834355, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 50268 + }, + { + "epoch": 0.50269, + "grad_norm": 1.05663027623059, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 50269 + }, + { + "epoch": 0.5027, + "grad_norm": 1.6812058358787279, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 50270 + }, + { + "epoch": 0.50271, + "grad_norm": 1.3945113892095797, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 50271 + }, + { + "epoch": 0.50272, + "grad_norm": 1.7371457199784501, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 50272 + }, + { + "epoch": 0.50273, + "grad_norm": 1.5224149632058033, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 50273 + }, + { + "epoch": 0.50274, + "grad_norm": 1.7872063911666662, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 50274 + }, + { + "epoch": 0.50275, + "grad_norm": 1.2784156881050375, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50275 + }, + { + "epoch": 0.50276, + "grad_norm": 1.5256400391332166, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50276 + }, + { + "epoch": 0.50277, + "grad_norm": 1.7732088852163912, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 50277 + }, + { + "epoch": 0.50278, + "grad_norm": 1.7735874284471516, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 50278 + }, + { + "epoch": 0.50279, + "grad_norm": 1.6319349272873336, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 50279 + }, + { + "epoch": 0.5028, + "grad_norm": 1.5016899619616777, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50280 + }, + { + "epoch": 0.50281, + "grad_norm": 1.334502195089116, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50281 + }, + { + "epoch": 0.50282, + "grad_norm": 1.8917397651150265, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 50282 + }, + { + "epoch": 0.50283, + "grad_norm": 1.245910555150408, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 50283 + }, + { + "epoch": 0.50284, + "grad_norm": 1.974152615836742, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 50284 + }, + { + "epoch": 0.50285, + "grad_norm": 1.6358658167500397, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 50285 + }, + { + "epoch": 0.50286, + "grad_norm": 1.4045805127978555, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 50286 + }, + { + "epoch": 0.50287, + "grad_norm": 1.9223615632060451, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50287 + }, + { + "epoch": 0.50288, + "grad_norm": 1.4256394406493902, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 50288 + }, + { + "epoch": 0.50289, + "grad_norm": 1.7817912866969041, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 50289 + }, + { + "epoch": 0.5029, + "grad_norm": 1.703287381364225, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 50290 + }, + { + "epoch": 0.50291, + "grad_norm": 1.4384899411889305, + "learning_rate": 0.003, + "loss": 4.048, + "step": 50291 + }, + { + "epoch": 0.50292, + "grad_norm": 1.668965704415001, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 50292 + }, + { + "epoch": 0.50293, + "grad_norm": 1.6770362995013093, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 50293 + }, + { + "epoch": 0.50294, + "grad_norm": 1.3743384938917438, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 50294 + }, + { + "epoch": 0.50295, + "grad_norm": 1.7391092030850366, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 50295 + }, + { + "epoch": 0.50296, + "grad_norm": 1.4845745793504748, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 50296 + }, + { + "epoch": 0.50297, + "grad_norm": 1.6128449435131575, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50297 + }, + { + "epoch": 0.50298, + "grad_norm": 1.359364470828219, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 50298 + }, + { + "epoch": 0.50299, + "grad_norm": 1.4355838628983661, + "learning_rate": 0.003, + "loss": 4.056, + "step": 50299 + }, + { + "epoch": 0.503, + "grad_norm": 1.3818746521371912, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 50300 + }, + { + "epoch": 0.50301, + "grad_norm": 1.8683232269419048, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 50301 + }, + { + "epoch": 0.50302, + "grad_norm": 1.4749793744295585, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 50302 + }, + { + "epoch": 0.50303, + "grad_norm": 1.5271323938023265, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50303 + }, + { + "epoch": 0.50304, + "grad_norm": 1.4310732912305644, + "learning_rate": 0.003, + "loss": 4.025, + "step": 50304 + }, + { + "epoch": 0.50305, + "grad_norm": 1.8020330994163876, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 50305 + }, + { + "epoch": 0.50306, + "grad_norm": 1.481080144523302, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50306 + }, + { + "epoch": 0.50307, + "grad_norm": 1.7234290187723573, + "learning_rate": 0.003, + "loss": 4.035, + "step": 50307 + }, + { + "epoch": 0.50308, + "grad_norm": 1.397701262130646, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50308 + }, + { + "epoch": 0.50309, + "grad_norm": 1.7113587555355796, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 50309 + }, + { + "epoch": 0.5031, + "grad_norm": 1.3880017403874445, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 50310 + }, + { + "epoch": 0.50311, + "grad_norm": 1.7915129943928962, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 50311 + }, + { + "epoch": 0.50312, + "grad_norm": 1.3846453736586821, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 50312 + }, + { + "epoch": 0.50313, + "grad_norm": 1.7601403049724542, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50313 + }, + { + "epoch": 0.50314, + "grad_norm": 1.510265153091066, + "learning_rate": 0.003, + "loss": 4.053, + "step": 50314 + }, + { + "epoch": 0.50315, + "grad_norm": 1.4336257151854734, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 50315 + }, + { + "epoch": 0.50316, + "grad_norm": 1.9314282962044138, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 50316 + }, + { + "epoch": 0.50317, + "grad_norm": 1.183731945817707, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50317 + }, + { + "epoch": 0.50318, + "grad_norm": 1.7729043725053284, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50318 + }, + { + "epoch": 0.50319, + "grad_norm": 1.2596955750089496, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50319 + }, + { + "epoch": 0.5032, + "grad_norm": 1.9003249083054716, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 50320 + }, + { + "epoch": 0.50321, + "grad_norm": 1.4652306113598443, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50321 + }, + { + "epoch": 0.50322, + "grad_norm": 1.3024929473062188, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 50322 + }, + { + "epoch": 0.50323, + "grad_norm": 1.781430247652059, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 50323 + }, + { + "epoch": 0.50324, + "grad_norm": 1.4938630753821367, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 50324 + }, + { + "epoch": 0.50325, + "grad_norm": 1.507935561095421, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 50325 + }, + { + "epoch": 0.50326, + "grad_norm": 1.5324747075634018, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 50326 + }, + { + "epoch": 0.50327, + "grad_norm": 1.6002949708414254, + "learning_rate": 0.003, + "loss": 4.04, + "step": 50327 + }, + { + "epoch": 0.50328, + "grad_norm": 1.5831941406425192, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 50328 + }, + { + "epoch": 0.50329, + "grad_norm": 1.4293675306375782, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 50329 + }, + { + "epoch": 0.5033, + "grad_norm": 1.4865981878166343, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50330 + }, + { + "epoch": 0.50331, + "grad_norm": 1.636679288156554, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50331 + }, + { + "epoch": 0.50332, + "grad_norm": 1.3454396871683973, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 50332 + }, + { + "epoch": 0.50333, + "grad_norm": 1.520138049467506, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 50333 + }, + { + "epoch": 0.50334, + "grad_norm": 1.523485517860855, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 50334 + }, + { + "epoch": 0.50335, + "grad_norm": 1.7529598134405762, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50335 + }, + { + "epoch": 0.50336, + "grad_norm": 1.6071374174686095, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 50336 + }, + { + "epoch": 0.50337, + "grad_norm": 1.5249127038244141, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 50337 + }, + { + "epoch": 0.50338, + "grad_norm": 1.4241455700597585, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 50338 + }, + { + "epoch": 0.50339, + "grad_norm": 1.6557347574820358, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50339 + }, + { + "epoch": 0.5034, + "grad_norm": 1.5336388974138635, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50340 + }, + { + "epoch": 0.50341, + "grad_norm": 1.7484807610784707, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 50341 + }, + { + "epoch": 0.50342, + "grad_norm": 1.6208131616446957, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 50342 + }, + { + "epoch": 0.50343, + "grad_norm": 1.651926482215535, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50343 + }, + { + "epoch": 0.50344, + "grad_norm": 1.3687086101632535, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50344 + }, + { + "epoch": 0.50345, + "grad_norm": 1.7953391210456207, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 50345 + }, + { + "epoch": 0.50346, + "grad_norm": 1.31555422784707, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 50346 + }, + { + "epoch": 0.50347, + "grad_norm": 1.649776758675357, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50347 + }, + { + "epoch": 0.50348, + "grad_norm": 1.5231585600124513, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50348 + }, + { + "epoch": 0.50349, + "grad_norm": 1.4576904392470662, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 50349 + }, + { + "epoch": 0.5035, + "grad_norm": 1.6747969774249212, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50350 + }, + { + "epoch": 0.50351, + "grad_norm": 1.4163092123571726, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 50351 + }, + { + "epoch": 0.50352, + "grad_norm": 1.8243830517335013, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 50352 + }, + { + "epoch": 0.50353, + "grad_norm": 1.536739103700164, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 50353 + }, + { + "epoch": 0.50354, + "grad_norm": 1.8143937417143536, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50354 + }, + { + "epoch": 0.50355, + "grad_norm": 1.4148313656435865, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 50355 + }, + { + "epoch": 0.50356, + "grad_norm": 1.4233472609479711, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 50356 + }, + { + "epoch": 0.50357, + "grad_norm": 1.7458070689693796, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 50357 + }, + { + "epoch": 0.50358, + "grad_norm": 1.439202769987128, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 50358 + }, + { + "epoch": 0.50359, + "grad_norm": 1.6738161381104995, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 50359 + }, + { + "epoch": 0.5036, + "grad_norm": 1.409681546876824, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 50360 + }, + { + "epoch": 0.50361, + "grad_norm": 1.9058907569148165, + "learning_rate": 0.003, + "loss": 4.03, + "step": 50361 + }, + { + "epoch": 0.50362, + "grad_norm": 1.3400687737192223, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50362 + }, + { + "epoch": 0.50363, + "grad_norm": 2.1509729737218417, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 50363 + }, + { + "epoch": 0.50364, + "grad_norm": 1.746307667092257, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50364 + }, + { + "epoch": 0.50365, + "grad_norm": 1.4468824395808562, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50365 + }, + { + "epoch": 0.50366, + "grad_norm": 1.4244413046636712, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 50366 + }, + { + "epoch": 0.50367, + "grad_norm": 1.7975485993518758, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 50367 + }, + { + "epoch": 0.50368, + "grad_norm": 1.2639087530663355, + "learning_rate": 0.003, + "loss": 4.023, + "step": 50368 + }, + { + "epoch": 0.50369, + "grad_norm": 1.5655947638453462, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 50369 + }, + { + "epoch": 0.5037, + "grad_norm": 1.275757205300581, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 50370 + }, + { + "epoch": 0.50371, + "grad_norm": 1.6088909298153837, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 50371 + }, + { + "epoch": 0.50372, + "grad_norm": 1.5726396905244397, + "learning_rate": 0.003, + "loss": 4.022, + "step": 50372 + }, + { + "epoch": 0.50373, + "grad_norm": 1.865797676906721, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 50373 + }, + { + "epoch": 0.50374, + "grad_norm": 1.3354000145650162, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 50374 + }, + { + "epoch": 0.50375, + "grad_norm": 1.6047893034750753, + "learning_rate": 0.003, + "loss": 4.068, + "step": 50375 + }, + { + "epoch": 0.50376, + "grad_norm": 1.534046657292214, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 50376 + }, + { + "epoch": 0.50377, + "grad_norm": 1.5181612480701008, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50377 + }, + { + "epoch": 0.50378, + "grad_norm": 1.4482057658079501, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50378 + }, + { + "epoch": 0.50379, + "grad_norm": 1.4869882959247174, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 50379 + }, + { + "epoch": 0.5038, + "grad_norm": 1.8204688136911222, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 50380 + }, + { + "epoch": 0.50381, + "grad_norm": 1.6723250394592786, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 50381 + }, + { + "epoch": 0.50382, + "grad_norm": 1.881711694142433, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50382 + }, + { + "epoch": 0.50383, + "grad_norm": 1.676358339912805, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 50383 + }, + { + "epoch": 0.50384, + "grad_norm": 1.3736897662990817, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50384 + }, + { + "epoch": 0.50385, + "grad_norm": 1.5614447098313637, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50385 + }, + { + "epoch": 0.50386, + "grad_norm": 1.594539577827193, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50386 + }, + { + "epoch": 0.50387, + "grad_norm": 1.71521991449332, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 50387 + }, + { + "epoch": 0.50388, + "grad_norm": 1.853005601220802, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50388 + }, + { + "epoch": 0.50389, + "grad_norm": 1.3478518938790736, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 50389 + }, + { + "epoch": 0.5039, + "grad_norm": 1.5226312467333214, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 50390 + }, + { + "epoch": 0.50391, + "grad_norm": 1.7483445295960287, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 50391 + }, + { + "epoch": 0.50392, + "grad_norm": 1.6424469178151255, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 50392 + }, + { + "epoch": 0.50393, + "grad_norm": 1.377190265790559, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 50393 + }, + { + "epoch": 0.50394, + "grad_norm": 1.4095278938696216, + "learning_rate": 0.003, + "loss": 4.022, + "step": 50394 + }, + { + "epoch": 0.50395, + "grad_norm": 1.4676324388312816, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 50395 + }, + { + "epoch": 0.50396, + "grad_norm": 1.5648445219694265, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 50396 + }, + { + "epoch": 0.50397, + "grad_norm": 1.4520645546062556, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 50397 + }, + { + "epoch": 0.50398, + "grad_norm": 1.8056164174872202, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 50398 + }, + { + "epoch": 0.50399, + "grad_norm": 1.4195277141808278, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 50399 + }, + { + "epoch": 0.504, + "grad_norm": 1.9407340101691064, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 50400 + }, + { + "epoch": 0.50401, + "grad_norm": 2.0734065981828858, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 50401 + }, + { + "epoch": 0.50402, + "grad_norm": 1.4477873617007349, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50402 + }, + { + "epoch": 0.50403, + "grad_norm": 1.7290767632441535, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 50403 + }, + { + "epoch": 0.50404, + "grad_norm": 1.3654787853885964, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 50404 + }, + { + "epoch": 0.50405, + "grad_norm": 1.8481375698489149, + "learning_rate": 0.003, + "loss": 4.058, + "step": 50405 + }, + { + "epoch": 0.50406, + "grad_norm": 1.2249868274746452, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 50406 + }, + { + "epoch": 0.50407, + "grad_norm": 1.9814754379093156, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 50407 + }, + { + "epoch": 0.50408, + "grad_norm": 1.4688660923339727, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 50408 + }, + { + "epoch": 0.50409, + "grad_norm": 1.910665780889605, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 50409 + }, + { + "epoch": 0.5041, + "grad_norm": 1.50552998543289, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 50410 + }, + { + "epoch": 0.50411, + "grad_norm": 1.520614477402337, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 50411 + }, + { + "epoch": 0.50412, + "grad_norm": 1.4438524053691526, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 50412 + }, + { + "epoch": 0.50413, + "grad_norm": 1.581770130167233, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50413 + }, + { + "epoch": 0.50414, + "grad_norm": 1.6791309819767326, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 50414 + }, + { + "epoch": 0.50415, + "grad_norm": 1.48133450415821, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 50415 + }, + { + "epoch": 0.50416, + "grad_norm": 1.57482495074215, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 50416 + }, + { + "epoch": 0.50417, + "grad_norm": 1.627215854162745, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50417 + }, + { + "epoch": 0.50418, + "grad_norm": 1.3460599941347542, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 50418 + }, + { + "epoch": 0.50419, + "grad_norm": 1.6002836345053941, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 50419 + }, + { + "epoch": 0.5042, + "grad_norm": 1.2989793937551728, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50420 + }, + { + "epoch": 0.50421, + "grad_norm": 1.6530657532331776, + "learning_rate": 0.003, + "loss": 4.063, + "step": 50421 + }, + { + "epoch": 0.50422, + "grad_norm": 1.516745913047141, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 50422 + }, + { + "epoch": 0.50423, + "grad_norm": 1.3908691627749201, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 50423 + }, + { + "epoch": 0.50424, + "grad_norm": 1.6547294821130094, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 50424 + }, + { + "epoch": 0.50425, + "grad_norm": 1.3155126583050618, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 50425 + }, + { + "epoch": 0.50426, + "grad_norm": 2.0549271669479237, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 50426 + }, + { + "epoch": 0.50427, + "grad_norm": 1.670503858250757, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 50427 + }, + { + "epoch": 0.50428, + "grad_norm": 1.6493952624616424, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50428 + }, + { + "epoch": 0.50429, + "grad_norm": 1.7540258761677534, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 50429 + }, + { + "epoch": 0.5043, + "grad_norm": 1.356223342261774, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50430 + }, + { + "epoch": 0.50431, + "grad_norm": 1.6598356528075031, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 50431 + }, + { + "epoch": 0.50432, + "grad_norm": 1.9901352539713153, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 50432 + }, + { + "epoch": 0.50433, + "grad_norm": 1.700844169177833, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50433 + }, + { + "epoch": 0.50434, + "grad_norm": 1.2724785752103496, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50434 + }, + { + "epoch": 0.50435, + "grad_norm": 1.6824791594606852, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50435 + }, + { + "epoch": 0.50436, + "grad_norm": 1.4728258758415373, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 50436 + }, + { + "epoch": 0.50437, + "grad_norm": 1.4476358182442544, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 50437 + }, + { + "epoch": 0.50438, + "grad_norm": 1.4123890719798777, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 50438 + }, + { + "epoch": 0.50439, + "grad_norm": 1.838684063419268, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50439 + }, + { + "epoch": 0.5044, + "grad_norm": 1.3815796447649653, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50440 + }, + { + "epoch": 0.50441, + "grad_norm": 1.6834547185965527, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50441 + }, + { + "epoch": 0.50442, + "grad_norm": 1.2666749820223042, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 50442 + }, + { + "epoch": 0.50443, + "grad_norm": 1.8464468281816575, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 50443 + }, + { + "epoch": 0.50444, + "grad_norm": 1.5113607544822878, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 50444 + }, + { + "epoch": 0.50445, + "grad_norm": 1.6126804892613902, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 50445 + }, + { + "epoch": 0.50446, + "grad_norm": 1.6491494306420145, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 50446 + }, + { + "epoch": 0.50447, + "grad_norm": 1.3842463887656338, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50447 + }, + { + "epoch": 0.50448, + "grad_norm": 1.5507109785042215, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 50448 + }, + { + "epoch": 0.50449, + "grad_norm": 1.5861275856528407, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 50449 + }, + { + "epoch": 0.5045, + "grad_norm": 1.8013053824804044, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 50450 + }, + { + "epoch": 0.50451, + "grad_norm": 1.3762456768537281, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 50451 + }, + { + "epoch": 0.50452, + "grad_norm": 1.7116077706747328, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 50452 + }, + { + "epoch": 0.50453, + "grad_norm": 1.538017566532124, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50453 + }, + { + "epoch": 0.50454, + "grad_norm": 1.46754384779541, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50454 + }, + { + "epoch": 0.50455, + "grad_norm": 1.445371692398374, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 50455 + }, + { + "epoch": 0.50456, + "grad_norm": 1.5126897935841348, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 50456 + }, + { + "epoch": 0.50457, + "grad_norm": 1.3083229074521283, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 50457 + }, + { + "epoch": 0.50458, + "grad_norm": 1.7455701246766013, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 50458 + }, + { + "epoch": 0.50459, + "grad_norm": 1.6724466195452852, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 50459 + }, + { + "epoch": 0.5046, + "grad_norm": 1.622447365756669, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 50460 + }, + { + "epoch": 0.50461, + "grad_norm": 1.643237239283123, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 50461 + }, + { + "epoch": 0.50462, + "grad_norm": 1.3501019443701185, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 50462 + }, + { + "epoch": 0.50463, + "grad_norm": 1.663388958312888, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 50463 + }, + { + "epoch": 0.50464, + "grad_norm": 1.7442198796528354, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 50464 + }, + { + "epoch": 0.50465, + "grad_norm": 1.1755227999626146, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 50465 + }, + { + "epoch": 0.50466, + "grad_norm": 1.7743866848041479, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 50466 + }, + { + "epoch": 0.50467, + "grad_norm": 1.4090463521813243, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 50467 + }, + { + "epoch": 0.50468, + "grad_norm": 1.913123563194188, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50468 + }, + { + "epoch": 0.50469, + "grad_norm": 1.2062045715060437, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 50469 + }, + { + "epoch": 0.5047, + "grad_norm": 1.8819303881358076, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 50470 + }, + { + "epoch": 0.50471, + "grad_norm": 1.3488160804620768, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 50471 + }, + { + "epoch": 0.50472, + "grad_norm": 1.6974692682151697, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 50472 + }, + { + "epoch": 0.50473, + "grad_norm": 1.404567078475015, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50473 + }, + { + "epoch": 0.50474, + "grad_norm": 1.7447304809217583, + "learning_rate": 0.003, + "loss": 4.035, + "step": 50474 + }, + { + "epoch": 0.50475, + "grad_norm": 1.5476289339020288, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50475 + }, + { + "epoch": 0.50476, + "grad_norm": 1.5900727571704933, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50476 + }, + { + "epoch": 0.50477, + "grad_norm": 1.589690755894358, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 50477 + }, + { + "epoch": 0.50478, + "grad_norm": 1.342167063176385, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 50478 + }, + { + "epoch": 0.50479, + "grad_norm": 1.435648459302318, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50479 + }, + { + "epoch": 0.5048, + "grad_norm": 1.8619570758860287, + "learning_rate": 0.003, + "loss": 4.039, + "step": 50480 + }, + { + "epoch": 0.50481, + "grad_norm": 1.3956544694751594, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 50481 + }, + { + "epoch": 0.50482, + "grad_norm": 1.6568623461729528, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 50482 + }, + { + "epoch": 0.50483, + "grad_norm": 1.266734458461565, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 50483 + }, + { + "epoch": 0.50484, + "grad_norm": 2.2673308945909687, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 50484 + }, + { + "epoch": 0.50485, + "grad_norm": 1.4024411959182495, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 50485 + }, + { + "epoch": 0.50486, + "grad_norm": 1.239435229789683, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50486 + }, + { + "epoch": 0.50487, + "grad_norm": 1.5131799558112837, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 50487 + }, + { + "epoch": 0.50488, + "grad_norm": 1.8806199285409626, + "learning_rate": 0.003, + "loss": 4.039, + "step": 50488 + }, + { + "epoch": 0.50489, + "grad_norm": 1.7042564040103558, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 50489 + }, + { + "epoch": 0.5049, + "grad_norm": 1.26470765637656, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 50490 + }, + { + "epoch": 0.50491, + "grad_norm": 1.818766486931505, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50491 + }, + { + "epoch": 0.50492, + "grad_norm": 1.4527091860987982, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 50492 + }, + { + "epoch": 0.50493, + "grad_norm": 1.6892012783863215, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 50493 + }, + { + "epoch": 0.50494, + "grad_norm": 1.5166352721016483, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50494 + }, + { + "epoch": 0.50495, + "grad_norm": 1.4919222032992596, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 50495 + }, + { + "epoch": 0.50496, + "grad_norm": 1.5209611179239617, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 50496 + }, + { + "epoch": 0.50497, + "grad_norm": 1.7073526053056474, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 50497 + }, + { + "epoch": 0.50498, + "grad_norm": 1.5282561242892778, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 50498 + }, + { + "epoch": 0.50499, + "grad_norm": 1.6298135972290322, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 50499 + }, + { + "epoch": 0.505, + "grad_norm": 1.4365792399833375, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50500 + }, + { + "epoch": 0.50501, + "grad_norm": 1.4241356964955625, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 50501 + }, + { + "epoch": 0.50502, + "grad_norm": 1.7215522903513638, + "learning_rate": 0.003, + "loss": 4.06, + "step": 50502 + }, + { + "epoch": 0.50503, + "grad_norm": 1.2951724990476423, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 50503 + }, + { + "epoch": 0.50504, + "grad_norm": 1.9973127185649369, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 50504 + }, + { + "epoch": 0.50505, + "grad_norm": 1.5628135775726872, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50505 + }, + { + "epoch": 0.50506, + "grad_norm": 1.9919905426500288, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 50506 + }, + { + "epoch": 0.50507, + "grad_norm": 1.617222239893219, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50507 + }, + { + "epoch": 0.50508, + "grad_norm": 1.2550820567690253, + "learning_rate": 0.003, + "loss": 4.04, + "step": 50508 + }, + { + "epoch": 0.50509, + "grad_norm": 1.5584850089707154, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 50509 + }, + { + "epoch": 0.5051, + "grad_norm": 1.6234333415733158, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 50510 + }, + { + "epoch": 0.50511, + "grad_norm": 1.3065814033542626, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50511 + }, + { + "epoch": 0.50512, + "grad_norm": 1.6792963763824094, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 50512 + }, + { + "epoch": 0.50513, + "grad_norm": 1.5832528211727475, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 50513 + }, + { + "epoch": 0.50514, + "grad_norm": 1.4511371340751607, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50514 + }, + { + "epoch": 0.50515, + "grad_norm": 1.7166927134117225, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 50515 + }, + { + "epoch": 0.50516, + "grad_norm": 1.5540339615991487, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 50516 + }, + { + "epoch": 0.50517, + "grad_norm": 1.7407939761333346, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 50517 + }, + { + "epoch": 0.50518, + "grad_norm": 1.615376413132265, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 50518 + }, + { + "epoch": 0.50519, + "grad_norm": 1.6212716986227063, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50519 + }, + { + "epoch": 0.5052, + "grad_norm": 1.429797119751927, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 50520 + }, + { + "epoch": 0.50521, + "grad_norm": 1.2897193507473723, + "learning_rate": 0.003, + "loss": 4.029, + "step": 50521 + }, + { + "epoch": 0.50522, + "grad_norm": 1.5695341121018616, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 50522 + }, + { + "epoch": 0.50523, + "grad_norm": 1.409013377770723, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 50523 + }, + { + "epoch": 0.50524, + "grad_norm": 1.6146641599953506, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 50524 + }, + { + "epoch": 0.50525, + "grad_norm": 1.2239984478625285, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 50525 + }, + { + "epoch": 0.50526, + "grad_norm": 1.8761713028946267, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 50526 + }, + { + "epoch": 0.50527, + "grad_norm": 1.7127159372285528, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50527 + }, + { + "epoch": 0.50528, + "grad_norm": 1.5453155807258343, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 50528 + }, + { + "epoch": 0.50529, + "grad_norm": 1.6422377597079663, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 50529 + }, + { + "epoch": 0.5053, + "grad_norm": 1.4895837869367934, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50530 + }, + { + "epoch": 0.50531, + "grad_norm": 1.6761931852575915, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 50531 + }, + { + "epoch": 0.50532, + "grad_norm": 1.5621902579903941, + "learning_rate": 0.003, + "loss": 4.045, + "step": 50532 + }, + { + "epoch": 0.50533, + "grad_norm": 1.3843669271179813, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 50533 + }, + { + "epoch": 0.50534, + "grad_norm": 1.5595828338426592, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 50534 + }, + { + "epoch": 0.50535, + "grad_norm": 1.4354799516902714, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 50535 + }, + { + "epoch": 0.50536, + "grad_norm": 1.7843074959627516, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 50536 + }, + { + "epoch": 0.50537, + "grad_norm": 1.381561135572039, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 50537 + }, + { + "epoch": 0.50538, + "grad_norm": 1.764267094811188, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 50538 + }, + { + "epoch": 0.50539, + "grad_norm": 1.4835680353058602, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 50539 + }, + { + "epoch": 0.5054, + "grad_norm": 1.4831043161730206, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 50540 + }, + { + "epoch": 0.50541, + "grad_norm": 1.5612023438073444, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 50541 + }, + { + "epoch": 0.50542, + "grad_norm": 1.559458233310382, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50542 + }, + { + "epoch": 0.50543, + "grad_norm": 1.5383272386841225, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50543 + }, + { + "epoch": 0.50544, + "grad_norm": 1.5783526616238739, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50544 + }, + { + "epoch": 0.50545, + "grad_norm": 1.8069560043044148, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 50545 + }, + { + "epoch": 0.50546, + "grad_norm": 1.682955018174283, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 50546 + }, + { + "epoch": 0.50547, + "grad_norm": 1.983662381142729, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50547 + }, + { + "epoch": 0.50548, + "grad_norm": 1.2090395000287806, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 50548 + }, + { + "epoch": 0.50549, + "grad_norm": 1.3833479466885361, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 50549 + }, + { + "epoch": 0.5055, + "grad_norm": 1.5604141904022006, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 50550 + }, + { + "epoch": 0.50551, + "grad_norm": 1.2909181424364211, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 50551 + }, + { + "epoch": 0.50552, + "grad_norm": 2.0403685270502083, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50552 + }, + { + "epoch": 0.50553, + "grad_norm": 1.112185812107014, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 50553 + }, + { + "epoch": 0.50554, + "grad_norm": 1.9264599436355352, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 50554 + }, + { + "epoch": 0.50555, + "grad_norm": 1.6849676762041843, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 50555 + }, + { + "epoch": 0.50556, + "grad_norm": 1.5198838345578505, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 50556 + }, + { + "epoch": 0.50557, + "grad_norm": 1.867759324651947, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 50557 + }, + { + "epoch": 0.50558, + "grad_norm": 1.5244656575327258, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50558 + }, + { + "epoch": 0.50559, + "grad_norm": 1.3273096082573579, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 50559 + }, + { + "epoch": 0.5056, + "grad_norm": 1.8688852809134746, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50560 + }, + { + "epoch": 0.50561, + "grad_norm": 1.558819583764826, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 50561 + }, + { + "epoch": 0.50562, + "grad_norm": 1.5533639579068947, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 50562 + }, + { + "epoch": 0.50563, + "grad_norm": 1.415426616865262, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 50563 + }, + { + "epoch": 0.50564, + "grad_norm": 1.504369144776337, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 50564 + }, + { + "epoch": 0.50565, + "grad_norm": 1.1485609535919725, + "learning_rate": 0.003, + "loss": 4.013, + "step": 50565 + }, + { + "epoch": 0.50566, + "grad_norm": 1.8370916101326522, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50566 + }, + { + "epoch": 0.50567, + "grad_norm": 1.2806886958609727, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50567 + }, + { + "epoch": 0.50568, + "grad_norm": 1.9031784401484815, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 50568 + }, + { + "epoch": 0.50569, + "grad_norm": 2.0511281619015906, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 50569 + }, + { + "epoch": 0.5057, + "grad_norm": 1.263902731871148, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 50570 + }, + { + "epoch": 0.50571, + "grad_norm": 1.7553621282817495, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 50571 + }, + { + "epoch": 0.50572, + "grad_norm": 1.587465132961509, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 50572 + }, + { + "epoch": 0.50573, + "grad_norm": 1.780871379501946, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 50573 + }, + { + "epoch": 0.50574, + "grad_norm": 1.3620243602985804, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 50574 + }, + { + "epoch": 0.50575, + "grad_norm": 1.481241170390262, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 50575 + }, + { + "epoch": 0.50576, + "grad_norm": 1.7243838425698406, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 50576 + }, + { + "epoch": 0.50577, + "grad_norm": 1.1452535535049335, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 50577 + }, + { + "epoch": 0.50578, + "grad_norm": 1.8165235127718462, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 50578 + }, + { + "epoch": 0.50579, + "grad_norm": 1.5235691886651928, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 50579 + }, + { + "epoch": 0.5058, + "grad_norm": 1.562410946222286, + "learning_rate": 0.003, + "loss": 4.032, + "step": 50580 + }, + { + "epoch": 0.50581, + "grad_norm": 1.3353141080872004, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 50581 + }, + { + "epoch": 0.50582, + "grad_norm": 1.5912825398190282, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 50582 + }, + { + "epoch": 0.50583, + "grad_norm": 1.6499779675331419, + "learning_rate": 0.003, + "loss": 4.05, + "step": 50583 + }, + { + "epoch": 0.50584, + "grad_norm": 1.7336692149831994, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 50584 + }, + { + "epoch": 0.50585, + "grad_norm": 1.7637295112557627, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 50585 + }, + { + "epoch": 0.50586, + "grad_norm": 1.448011662226716, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50586 + }, + { + "epoch": 0.50587, + "grad_norm": 1.656052934927666, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 50587 + }, + { + "epoch": 0.50588, + "grad_norm": 1.453507417554678, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 50588 + }, + { + "epoch": 0.50589, + "grad_norm": 1.4208461288334546, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 50589 + }, + { + "epoch": 0.5059, + "grad_norm": 1.4878630658898293, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 50590 + }, + { + "epoch": 0.50591, + "grad_norm": 1.4322708197313532, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 50591 + }, + { + "epoch": 0.50592, + "grad_norm": 1.577310862506926, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 50592 + }, + { + "epoch": 0.50593, + "grad_norm": 1.4488481708248484, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 50593 + }, + { + "epoch": 0.50594, + "grad_norm": 1.7456564924587803, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 50594 + }, + { + "epoch": 0.50595, + "grad_norm": 1.5438852359152309, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50595 + }, + { + "epoch": 0.50596, + "grad_norm": 1.635880486914436, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50596 + }, + { + "epoch": 0.50597, + "grad_norm": 1.4639067898812006, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50597 + }, + { + "epoch": 0.50598, + "grad_norm": 1.6716034142182012, + "learning_rate": 0.003, + "loss": 4.056, + "step": 50598 + }, + { + "epoch": 0.50599, + "grad_norm": 1.1880920757949498, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 50599 + }, + { + "epoch": 0.506, + "grad_norm": 1.6124982917586153, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 50600 + }, + { + "epoch": 0.50601, + "grad_norm": 1.3250373628636347, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 50601 + }, + { + "epoch": 0.50602, + "grad_norm": 1.559917938014534, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 50602 + }, + { + "epoch": 0.50603, + "grad_norm": 1.4792239177581572, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 50603 + }, + { + "epoch": 0.50604, + "grad_norm": 1.913814700550118, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 50604 + }, + { + "epoch": 0.50605, + "grad_norm": 1.4005879851385223, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 50605 + }, + { + "epoch": 0.50606, + "grad_norm": 1.8623274845461901, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 50606 + }, + { + "epoch": 0.50607, + "grad_norm": 1.5431278789720544, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50607 + }, + { + "epoch": 0.50608, + "grad_norm": 1.4201753934152983, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 50608 + }, + { + "epoch": 0.50609, + "grad_norm": 1.5450855473473155, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50609 + }, + { + "epoch": 0.5061, + "grad_norm": 1.6641559653924543, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 50610 + }, + { + "epoch": 0.50611, + "grad_norm": 1.3886199154393604, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50611 + }, + { + "epoch": 0.50612, + "grad_norm": 1.7512988848697482, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50612 + }, + { + "epoch": 0.50613, + "grad_norm": 1.6942898035716623, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 50613 + }, + { + "epoch": 0.50614, + "grad_norm": 1.3007177348475385, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50614 + }, + { + "epoch": 0.50615, + "grad_norm": 1.5870987566415355, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 50615 + }, + { + "epoch": 0.50616, + "grad_norm": 1.502022461614554, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 50616 + }, + { + "epoch": 0.50617, + "grad_norm": 1.7306183647695834, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 50617 + }, + { + "epoch": 0.50618, + "grad_norm": 1.504552965370404, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 50618 + }, + { + "epoch": 0.50619, + "grad_norm": 1.6481404252181309, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50619 + }, + { + "epoch": 0.5062, + "grad_norm": 1.4682087614236792, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 50620 + }, + { + "epoch": 0.50621, + "grad_norm": 1.5928315278266942, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50621 + }, + { + "epoch": 0.50622, + "grad_norm": 1.563897009630123, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50622 + }, + { + "epoch": 0.50623, + "grad_norm": 1.7630254307964937, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 50623 + }, + { + "epoch": 0.50624, + "grad_norm": 1.2093165008409046, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 50624 + }, + { + "epoch": 0.50625, + "grad_norm": 1.6822710830526644, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50625 + }, + { + "epoch": 0.50626, + "grad_norm": 1.3932103421884157, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50626 + }, + { + "epoch": 0.50627, + "grad_norm": 1.5330731734463967, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 50627 + }, + { + "epoch": 0.50628, + "grad_norm": 1.615367000241624, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 50628 + }, + { + "epoch": 0.50629, + "grad_norm": 1.9449238458377254, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 50629 + }, + { + "epoch": 0.5063, + "grad_norm": 1.362053245511557, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 50630 + }, + { + "epoch": 0.50631, + "grad_norm": 1.4408607773669684, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 50631 + }, + { + "epoch": 0.50632, + "grad_norm": 1.6762516207461773, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 50632 + }, + { + "epoch": 0.50633, + "grad_norm": 1.5332185408386756, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 50633 + }, + { + "epoch": 0.50634, + "grad_norm": 1.3419786365508384, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50634 + }, + { + "epoch": 0.50635, + "grad_norm": 1.5397547628356727, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 50635 + }, + { + "epoch": 0.50636, + "grad_norm": 1.3465131814457936, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 50636 + }, + { + "epoch": 0.50637, + "grad_norm": 1.5877360633240052, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 50637 + }, + { + "epoch": 0.50638, + "grad_norm": 1.5458789953879786, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 50638 + }, + { + "epoch": 0.50639, + "grad_norm": 1.6998237246771686, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50639 + }, + { + "epoch": 0.5064, + "grad_norm": 1.649662345031614, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 50640 + }, + { + "epoch": 0.50641, + "grad_norm": 1.8297171322606256, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50641 + }, + { + "epoch": 0.50642, + "grad_norm": 1.402314479238666, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50642 + }, + { + "epoch": 0.50643, + "grad_norm": 1.5469867045067502, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 50643 + }, + { + "epoch": 0.50644, + "grad_norm": 1.865908981201366, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 50644 + }, + { + "epoch": 0.50645, + "grad_norm": 1.670039826209113, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 50645 + }, + { + "epoch": 0.50646, + "grad_norm": 1.5390867919542706, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 50646 + }, + { + "epoch": 0.50647, + "grad_norm": 1.6366001213135448, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 50647 + }, + { + "epoch": 0.50648, + "grad_norm": 1.3356215756643615, + "learning_rate": 0.003, + "loss": 4.003, + "step": 50648 + }, + { + "epoch": 0.50649, + "grad_norm": 1.5901969165014456, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 50649 + }, + { + "epoch": 0.5065, + "grad_norm": 1.1928182099929365, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 50650 + }, + { + "epoch": 0.50651, + "grad_norm": 1.8716074145508788, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50651 + }, + { + "epoch": 0.50652, + "grad_norm": 1.3183173088278224, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 50652 + }, + { + "epoch": 0.50653, + "grad_norm": 1.794388318918965, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 50653 + }, + { + "epoch": 0.50654, + "grad_norm": 1.573137776412791, + "learning_rate": 0.003, + "loss": 4.058, + "step": 50654 + }, + { + "epoch": 0.50655, + "grad_norm": 1.3802427369683543, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 50655 + }, + { + "epoch": 0.50656, + "grad_norm": 1.4186825074890832, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 50656 + }, + { + "epoch": 0.50657, + "grad_norm": 1.435647569044167, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 50657 + }, + { + "epoch": 0.50658, + "grad_norm": 1.7665406262975676, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50658 + }, + { + "epoch": 0.50659, + "grad_norm": 1.2870678534979259, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 50659 + }, + { + "epoch": 0.5066, + "grad_norm": 1.6550974799158924, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 50660 + }, + { + "epoch": 0.50661, + "grad_norm": 1.309782739245784, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50661 + }, + { + "epoch": 0.50662, + "grad_norm": 1.581488021729981, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50662 + }, + { + "epoch": 0.50663, + "grad_norm": 1.9411360673809148, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 50663 + }, + { + "epoch": 0.50664, + "grad_norm": 1.7368577321422294, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50664 + }, + { + "epoch": 0.50665, + "grad_norm": 1.6370982522118958, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50665 + }, + { + "epoch": 0.50666, + "grad_norm": 1.6692048291938792, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50666 + }, + { + "epoch": 0.50667, + "grad_norm": 1.55139142902489, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 50667 + }, + { + "epoch": 0.50668, + "grad_norm": 1.4480628066564543, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 50668 + }, + { + "epoch": 0.50669, + "grad_norm": 1.4459291931953926, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 50669 + }, + { + "epoch": 0.5067, + "grad_norm": 1.3927849546184055, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 50670 + }, + { + "epoch": 0.50671, + "grad_norm": 1.8056273474801816, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 50671 + }, + { + "epoch": 0.50672, + "grad_norm": 1.366964450168034, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50672 + }, + { + "epoch": 0.50673, + "grad_norm": 1.7554402362899493, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 50673 + }, + { + "epoch": 0.50674, + "grad_norm": 1.3256200294255827, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 50674 + }, + { + "epoch": 0.50675, + "grad_norm": 1.6798125516783895, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 50675 + }, + { + "epoch": 0.50676, + "grad_norm": 1.5455402510343343, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 50676 + }, + { + "epoch": 0.50677, + "grad_norm": 1.5391183254784386, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 50677 + }, + { + "epoch": 0.50678, + "grad_norm": 1.5664818432357768, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 50678 + }, + { + "epoch": 0.50679, + "grad_norm": 1.3830963967348002, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 50679 + }, + { + "epoch": 0.5068, + "grad_norm": 1.7069831884285371, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 50680 + }, + { + "epoch": 0.50681, + "grad_norm": 1.5045596472938532, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 50681 + }, + { + "epoch": 0.50682, + "grad_norm": 1.6499702569780772, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50682 + }, + { + "epoch": 0.50683, + "grad_norm": 1.4318916258416434, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 50683 + }, + { + "epoch": 0.50684, + "grad_norm": 1.9841541901556974, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 50684 + }, + { + "epoch": 0.50685, + "grad_norm": 1.2854906619105506, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 50685 + }, + { + "epoch": 0.50686, + "grad_norm": 1.7856182238430898, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 50686 + }, + { + "epoch": 0.50687, + "grad_norm": 1.7520248457349443, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 50687 + }, + { + "epoch": 0.50688, + "grad_norm": 1.4591862481320783, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50688 + }, + { + "epoch": 0.50689, + "grad_norm": 1.5465895184935943, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 50689 + }, + { + "epoch": 0.5069, + "grad_norm": 1.5372764585175143, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 50690 + }, + { + "epoch": 0.50691, + "grad_norm": 1.5145679356978428, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 50691 + }, + { + "epoch": 0.50692, + "grad_norm": 1.4600523544034147, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 50692 + }, + { + "epoch": 0.50693, + "grad_norm": 1.6034128459498855, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50693 + }, + { + "epoch": 0.50694, + "grad_norm": 1.6124848402090566, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 50694 + }, + { + "epoch": 0.50695, + "grad_norm": 1.657209175416345, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 50695 + }, + { + "epoch": 0.50696, + "grad_norm": 1.7085613543215075, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50696 + }, + { + "epoch": 0.50697, + "grad_norm": 1.24773390006356, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 50697 + }, + { + "epoch": 0.50698, + "grad_norm": 1.6150278114047305, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 50698 + }, + { + "epoch": 0.50699, + "grad_norm": 1.531701117069373, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 50699 + }, + { + "epoch": 0.507, + "grad_norm": 1.4118339712959935, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 50700 + }, + { + "epoch": 0.50701, + "grad_norm": 1.7720637696822663, + "learning_rate": 0.003, + "loss": 4.055, + "step": 50701 + }, + { + "epoch": 0.50702, + "grad_norm": 1.750255639483627, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 50702 + }, + { + "epoch": 0.50703, + "grad_norm": 1.3116095688039007, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50703 + }, + { + "epoch": 0.50704, + "grad_norm": 1.6902284080684626, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 50704 + }, + { + "epoch": 0.50705, + "grad_norm": 1.2917264881347896, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 50705 + }, + { + "epoch": 0.50706, + "grad_norm": 1.7051124984870774, + "learning_rate": 0.003, + "loss": 4.03, + "step": 50706 + }, + { + "epoch": 0.50707, + "grad_norm": 1.6059156406549084, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 50707 + }, + { + "epoch": 0.50708, + "grad_norm": 1.533287508952977, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50708 + }, + { + "epoch": 0.50709, + "grad_norm": 1.6045147083397873, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 50709 + }, + { + "epoch": 0.5071, + "grad_norm": 1.4625060665275345, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50710 + }, + { + "epoch": 0.50711, + "grad_norm": 1.4764695760438462, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 50711 + }, + { + "epoch": 0.50712, + "grad_norm": 1.4313604193950424, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 50712 + }, + { + "epoch": 0.50713, + "grad_norm": 1.5689024492833976, + "learning_rate": 0.003, + "loss": 3.995, + "step": 50713 + }, + { + "epoch": 0.50714, + "grad_norm": 1.6670154347121224, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50714 + }, + { + "epoch": 0.50715, + "grad_norm": 1.6047290861692216, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 50715 + }, + { + "epoch": 0.50716, + "grad_norm": 1.5203026806553386, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 50716 + }, + { + "epoch": 0.50717, + "grad_norm": 1.7673678236945929, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 50717 + }, + { + "epoch": 0.50718, + "grad_norm": 1.2993460248753579, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50718 + }, + { + "epoch": 0.50719, + "grad_norm": 1.7680105972495874, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 50719 + }, + { + "epoch": 0.5072, + "grad_norm": 1.8301124966535762, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 50720 + }, + { + "epoch": 0.50721, + "grad_norm": 1.4353945253894176, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50721 + }, + { + "epoch": 0.50722, + "grad_norm": 1.585470982689666, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 50722 + }, + { + "epoch": 0.50723, + "grad_norm": 1.7591536333297566, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 50723 + }, + { + "epoch": 0.50724, + "grad_norm": 1.2944091896471788, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 50724 + }, + { + "epoch": 0.50725, + "grad_norm": 1.7783026790596639, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 50725 + }, + { + "epoch": 0.50726, + "grad_norm": 1.4259384865424078, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 50726 + }, + { + "epoch": 0.50727, + "grad_norm": 1.714600270592148, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50727 + }, + { + "epoch": 0.50728, + "grad_norm": 1.5840245844133471, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50728 + }, + { + "epoch": 0.50729, + "grad_norm": 1.5963265823193167, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 50729 + }, + { + "epoch": 0.5073, + "grad_norm": 1.918068507110426, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 50730 + }, + { + "epoch": 0.50731, + "grad_norm": 1.370584574495381, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 50731 + }, + { + "epoch": 0.50732, + "grad_norm": 1.4203529943757096, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 50732 + }, + { + "epoch": 0.50733, + "grad_norm": 1.6782065405645816, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 50733 + }, + { + "epoch": 0.50734, + "grad_norm": 1.5123321525967437, + "learning_rate": 0.003, + "loss": 4.049, + "step": 50734 + }, + { + "epoch": 0.50735, + "grad_norm": 1.6087316775452318, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 50735 + }, + { + "epoch": 0.50736, + "grad_norm": 1.5558821215873158, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 50736 + }, + { + "epoch": 0.50737, + "grad_norm": 1.5811741226686313, + "learning_rate": 0.003, + "loss": 4.043, + "step": 50737 + }, + { + "epoch": 0.50738, + "grad_norm": 1.5075476848292233, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 50738 + }, + { + "epoch": 0.50739, + "grad_norm": 1.4403496656344261, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 50739 + }, + { + "epoch": 0.5074, + "grad_norm": 1.4159937824108748, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 50740 + }, + { + "epoch": 0.50741, + "grad_norm": 1.664717736934681, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 50741 + }, + { + "epoch": 0.50742, + "grad_norm": 1.5540912265779885, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 50742 + }, + { + "epoch": 0.50743, + "grad_norm": 1.6132628724848694, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 50743 + }, + { + "epoch": 0.50744, + "grad_norm": 1.4890125526127043, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 50744 + }, + { + "epoch": 0.50745, + "grad_norm": 1.7571683544415633, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50745 + }, + { + "epoch": 0.50746, + "grad_norm": 1.269593223292712, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 50746 + }, + { + "epoch": 0.50747, + "grad_norm": 2.0942518503569194, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 50747 + }, + { + "epoch": 0.50748, + "grad_norm": 1.3299951083979655, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 50748 + }, + { + "epoch": 0.50749, + "grad_norm": 2.151008647857236, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50749 + }, + { + "epoch": 0.5075, + "grad_norm": 1.3677175790984324, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 50750 + }, + { + "epoch": 0.50751, + "grad_norm": 1.490352103406778, + "learning_rate": 0.003, + "loss": 4.047, + "step": 50751 + }, + { + "epoch": 0.50752, + "grad_norm": 1.5985003525149726, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 50752 + }, + { + "epoch": 0.50753, + "grad_norm": 1.5550841599060206, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 50753 + }, + { + "epoch": 0.50754, + "grad_norm": 1.668345110459035, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 50754 + }, + { + "epoch": 0.50755, + "grad_norm": 1.2566178498967018, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 50755 + }, + { + "epoch": 0.50756, + "grad_norm": 1.5830967276063248, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 50756 + }, + { + "epoch": 0.50757, + "grad_norm": 1.6888698606951476, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 50757 + }, + { + "epoch": 0.50758, + "grad_norm": 1.3355584298709202, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50758 + }, + { + "epoch": 0.50759, + "grad_norm": 1.3704301025656522, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 50759 + }, + { + "epoch": 0.5076, + "grad_norm": 1.647517389649607, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50760 + }, + { + "epoch": 0.50761, + "grad_norm": 1.2340651172196417, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 50761 + }, + { + "epoch": 0.50762, + "grad_norm": 1.935206170122436, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 50762 + }, + { + "epoch": 0.50763, + "grad_norm": 1.2454723193500117, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50763 + }, + { + "epoch": 0.50764, + "grad_norm": 1.5028550095764774, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 50764 + }, + { + "epoch": 0.50765, + "grad_norm": 1.5714195077163229, + "learning_rate": 0.003, + "loss": 4.059, + "step": 50765 + }, + { + "epoch": 0.50766, + "grad_norm": 1.6114779583489733, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 50766 + }, + { + "epoch": 0.50767, + "grad_norm": 1.9008730072139282, + "learning_rate": 0.003, + "loss": 4.025, + "step": 50767 + }, + { + "epoch": 0.50768, + "grad_norm": 1.486570751074064, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 50768 + }, + { + "epoch": 0.50769, + "grad_norm": 1.5685429476427402, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 50769 + }, + { + "epoch": 0.5077, + "grad_norm": 1.6877515082779646, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 50770 + }, + { + "epoch": 0.50771, + "grad_norm": 1.5252923277371657, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 50771 + }, + { + "epoch": 0.50772, + "grad_norm": 1.7212073340697471, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50772 + }, + { + "epoch": 0.50773, + "grad_norm": 1.46706983591778, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 50773 + }, + { + "epoch": 0.50774, + "grad_norm": 1.3943444849106599, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 50774 + }, + { + "epoch": 0.50775, + "grad_norm": 1.5428858927183169, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50775 + }, + { + "epoch": 0.50776, + "grad_norm": 1.551837669381299, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 50776 + }, + { + "epoch": 0.50777, + "grad_norm": 1.5276079366956754, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 50777 + }, + { + "epoch": 0.50778, + "grad_norm": 1.5660182619394831, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 50778 + }, + { + "epoch": 0.50779, + "grad_norm": 1.581960639621708, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 50779 + }, + { + "epoch": 0.5078, + "grad_norm": 1.6563998744485238, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 50780 + }, + { + "epoch": 0.50781, + "grad_norm": 1.628659845133554, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50781 + }, + { + "epoch": 0.50782, + "grad_norm": 1.5997917256515446, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 50782 + }, + { + "epoch": 0.50783, + "grad_norm": 1.554736600052154, + "learning_rate": 0.003, + "loss": 4.024, + "step": 50783 + }, + { + "epoch": 0.50784, + "grad_norm": 1.754102031001505, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 50784 + }, + { + "epoch": 0.50785, + "grad_norm": 1.300754936992756, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 50785 + }, + { + "epoch": 0.50786, + "grad_norm": 1.4971228250722228, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 50786 + }, + { + "epoch": 0.50787, + "grad_norm": 1.4990129409549635, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 50787 + }, + { + "epoch": 0.50788, + "grad_norm": 1.609558258560869, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 50788 + }, + { + "epoch": 0.50789, + "grad_norm": 1.3947715126190485, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 50789 + }, + { + "epoch": 0.5079, + "grad_norm": 1.6266702430413007, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50790 + }, + { + "epoch": 0.50791, + "grad_norm": 1.5399573986657789, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 50791 + }, + { + "epoch": 0.50792, + "grad_norm": 1.7146260338221557, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 50792 + }, + { + "epoch": 0.50793, + "grad_norm": 1.5340127019268486, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 50793 + }, + { + "epoch": 0.50794, + "grad_norm": 1.564265137210424, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 50794 + }, + { + "epoch": 0.50795, + "grad_norm": 1.677996206043887, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 50795 + }, + { + "epoch": 0.50796, + "grad_norm": 1.2517614598211482, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 50796 + }, + { + "epoch": 0.50797, + "grad_norm": 1.8381915107934228, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50797 + }, + { + "epoch": 0.50798, + "grad_norm": 1.5305502466304453, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50798 + }, + { + "epoch": 0.50799, + "grad_norm": 1.7082487848069223, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 50799 + }, + { + "epoch": 0.508, + "grad_norm": 1.6195827468485886, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 50800 + }, + { + "epoch": 0.50801, + "grad_norm": 1.4234221255581518, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 50801 + }, + { + "epoch": 0.50802, + "grad_norm": 1.4761125135589623, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 50802 + }, + { + "epoch": 0.50803, + "grad_norm": 1.7092921132591699, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 50803 + }, + { + "epoch": 0.50804, + "grad_norm": 1.4301118330850822, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 50804 + }, + { + "epoch": 0.50805, + "grad_norm": 1.71235611491296, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 50805 + }, + { + "epoch": 0.50806, + "grad_norm": 1.2688711177958016, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 50806 + }, + { + "epoch": 0.50807, + "grad_norm": 1.7446826616854119, + "learning_rate": 0.003, + "loss": 4.048, + "step": 50807 + }, + { + "epoch": 0.50808, + "grad_norm": 1.2225675849009807, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 50808 + }, + { + "epoch": 0.50809, + "grad_norm": 1.784487098382097, + "learning_rate": 0.003, + "loss": 4.024, + "step": 50809 + }, + { + "epoch": 0.5081, + "grad_norm": 1.5333483289734893, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 50810 + }, + { + "epoch": 0.50811, + "grad_norm": 1.589671422860852, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 50811 + }, + { + "epoch": 0.50812, + "grad_norm": 1.8048045678729878, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 50812 + }, + { + "epoch": 0.50813, + "grad_norm": 1.3369916910882305, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 50813 + }, + { + "epoch": 0.50814, + "grad_norm": 1.5254199114978306, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 50814 + }, + { + "epoch": 0.50815, + "grad_norm": 1.637055438471523, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50815 + }, + { + "epoch": 0.50816, + "grad_norm": 1.516392369297383, + "learning_rate": 0.003, + "loss": 4.035, + "step": 50816 + }, + { + "epoch": 0.50817, + "grad_norm": 1.3941704365857388, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 50817 + }, + { + "epoch": 0.50818, + "grad_norm": 1.7748218066319628, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 50818 + }, + { + "epoch": 0.50819, + "grad_norm": 1.8663127525712584, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 50819 + }, + { + "epoch": 0.5082, + "grad_norm": 1.493745640475036, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 50820 + }, + { + "epoch": 0.50821, + "grad_norm": 1.7775367578870636, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50821 + }, + { + "epoch": 0.50822, + "grad_norm": 1.6800940361777041, + "learning_rate": 0.003, + "loss": 4.059, + "step": 50822 + }, + { + "epoch": 0.50823, + "grad_norm": 1.383749793259203, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 50823 + }, + { + "epoch": 0.50824, + "grad_norm": 1.815992318065767, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 50824 + }, + { + "epoch": 0.50825, + "grad_norm": 1.3825262948190027, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50825 + }, + { + "epoch": 0.50826, + "grad_norm": 1.6985928936051593, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 50826 + }, + { + "epoch": 0.50827, + "grad_norm": 1.5118875746684002, + "learning_rate": 0.003, + "loss": 4.082, + "step": 50827 + }, + { + "epoch": 0.50828, + "grad_norm": 1.7137837119489243, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 50828 + }, + { + "epoch": 0.50829, + "grad_norm": 1.3162420544557, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 50829 + }, + { + "epoch": 0.5083, + "grad_norm": 1.6011901784628622, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 50830 + }, + { + "epoch": 0.50831, + "grad_norm": 1.7483242688624787, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 50831 + }, + { + "epoch": 0.50832, + "grad_norm": 1.8311717906794043, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 50832 + }, + { + "epoch": 0.50833, + "grad_norm": 1.219378967837434, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50833 + }, + { + "epoch": 0.50834, + "grad_norm": 1.6183788568070783, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50834 + }, + { + "epoch": 0.50835, + "grad_norm": 1.9410160886041445, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 50835 + }, + { + "epoch": 0.50836, + "grad_norm": 1.2939583484123944, + "learning_rate": 0.003, + "loss": 4.062, + "step": 50836 + }, + { + "epoch": 0.50837, + "grad_norm": 1.545303324281158, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 50837 + }, + { + "epoch": 0.50838, + "grad_norm": 1.2675794189609322, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 50838 + }, + { + "epoch": 0.50839, + "grad_norm": 1.9500165112028063, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 50839 + }, + { + "epoch": 0.5084, + "grad_norm": 1.271570568800538, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 50840 + }, + { + "epoch": 0.50841, + "grad_norm": 1.75559671112365, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 50841 + }, + { + "epoch": 0.50842, + "grad_norm": 1.5224895000901995, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 50842 + }, + { + "epoch": 0.50843, + "grad_norm": 1.7697692766774005, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 50843 + }, + { + "epoch": 0.50844, + "grad_norm": 1.89880513560648, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 50844 + }, + { + "epoch": 0.50845, + "grad_norm": 1.3280033083362934, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 50845 + }, + { + "epoch": 0.50846, + "grad_norm": 1.6277400661627772, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 50846 + }, + { + "epoch": 0.50847, + "grad_norm": 1.8201831681130556, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 50847 + }, + { + "epoch": 0.50848, + "grad_norm": 1.4484126741035486, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 50848 + }, + { + "epoch": 0.50849, + "grad_norm": 1.6244171569233876, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 50849 + }, + { + "epoch": 0.5085, + "grad_norm": 1.3760577963393277, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 50850 + }, + { + "epoch": 0.50851, + "grad_norm": 1.8248771910837402, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 50851 + }, + { + "epoch": 0.50852, + "grad_norm": 1.4709817419449538, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50852 + }, + { + "epoch": 0.50853, + "grad_norm": 1.2915257677916723, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 50853 + }, + { + "epoch": 0.50854, + "grad_norm": 1.6777553696658143, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 50854 + }, + { + "epoch": 0.50855, + "grad_norm": 1.41061228567446, + "learning_rate": 0.003, + "loss": 4.044, + "step": 50855 + }, + { + "epoch": 0.50856, + "grad_norm": 1.8239698464801815, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 50856 + }, + { + "epoch": 0.50857, + "grad_norm": 1.7399502537165867, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50857 + }, + { + "epoch": 0.50858, + "grad_norm": 1.5389815163639138, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 50858 + }, + { + "epoch": 0.50859, + "grad_norm": 1.7340418076341368, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 50859 + }, + { + "epoch": 0.5086, + "grad_norm": 1.4549909330257704, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 50860 + }, + { + "epoch": 0.50861, + "grad_norm": 1.6163965133877625, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 50861 + }, + { + "epoch": 0.50862, + "grad_norm": 1.639574650360603, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50862 + }, + { + "epoch": 0.50863, + "grad_norm": 1.4792649378706717, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 50863 + }, + { + "epoch": 0.50864, + "grad_norm": 1.5607956602234003, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 50864 + }, + { + "epoch": 0.50865, + "grad_norm": 1.4949485068386636, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 50865 + }, + { + "epoch": 0.50866, + "grad_norm": 1.9068005321741723, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 50866 + }, + { + "epoch": 0.50867, + "grad_norm": 1.4567537927429492, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 50867 + }, + { + "epoch": 0.50868, + "grad_norm": 1.4815992502671016, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 50868 + }, + { + "epoch": 0.50869, + "grad_norm": 1.637147022737994, + "learning_rate": 0.003, + "loss": 4.027, + "step": 50869 + }, + { + "epoch": 0.5087, + "grad_norm": 1.869039203268362, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 50870 + }, + { + "epoch": 0.50871, + "grad_norm": 1.3460122949294648, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 50871 + }, + { + "epoch": 0.50872, + "grad_norm": 1.5186441847331031, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 50872 + }, + { + "epoch": 0.50873, + "grad_norm": 1.7728820804783954, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 50873 + }, + { + "epoch": 0.50874, + "grad_norm": 1.3532713347635208, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 50874 + }, + { + "epoch": 0.50875, + "grad_norm": 1.77173167917732, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 50875 + }, + { + "epoch": 0.50876, + "grad_norm": 1.4498992135550857, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50876 + }, + { + "epoch": 0.50877, + "grad_norm": 1.730659713491192, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 50877 + }, + { + "epoch": 0.50878, + "grad_norm": 1.501842344875415, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50878 + }, + { + "epoch": 0.50879, + "grad_norm": 1.3964898591430734, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50879 + }, + { + "epoch": 0.5088, + "grad_norm": 1.4839633511190573, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 50880 + }, + { + "epoch": 0.50881, + "grad_norm": 1.689182074455771, + "learning_rate": 0.003, + "loss": 4.051, + "step": 50881 + }, + { + "epoch": 0.50882, + "grad_norm": 1.330346095477005, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 50882 + }, + { + "epoch": 0.50883, + "grad_norm": 1.6559530808681608, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 50883 + }, + { + "epoch": 0.50884, + "grad_norm": 1.2979110762241135, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 50884 + }, + { + "epoch": 0.50885, + "grad_norm": 1.897750109269166, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 50885 + }, + { + "epoch": 0.50886, + "grad_norm": 1.6682736378062308, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 50886 + }, + { + "epoch": 0.50887, + "grad_norm": 1.8056923551982282, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50887 + }, + { + "epoch": 0.50888, + "grad_norm": 1.639661748101224, + "learning_rate": 0.003, + "loss": 4.0, + "step": 50888 + }, + { + "epoch": 0.50889, + "grad_norm": 1.6884625390812988, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50889 + }, + { + "epoch": 0.5089, + "grad_norm": 1.5778751165612697, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 50890 + }, + { + "epoch": 0.50891, + "grad_norm": 1.235941531741966, + "learning_rate": 0.003, + "loss": 4.053, + "step": 50891 + }, + { + "epoch": 0.50892, + "grad_norm": 1.725750673576096, + "learning_rate": 0.003, + "loss": 4.036, + "step": 50892 + }, + { + "epoch": 0.50893, + "grad_norm": 1.4151168340821216, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 50893 + }, + { + "epoch": 0.50894, + "grad_norm": 2.0342908315916115, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 50894 + }, + { + "epoch": 0.50895, + "grad_norm": 1.4992953594078644, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 50895 + }, + { + "epoch": 0.50896, + "grad_norm": 1.763652636276399, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 50896 + }, + { + "epoch": 0.50897, + "grad_norm": 1.7032942053068088, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 50897 + }, + { + "epoch": 0.50898, + "grad_norm": 1.326560829771289, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 50898 + }, + { + "epoch": 0.50899, + "grad_norm": 1.7657447136755215, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 50899 + }, + { + "epoch": 0.509, + "grad_norm": 1.572389156466625, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 50900 + }, + { + "epoch": 0.50901, + "grad_norm": 1.6968267730700364, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 50901 + }, + { + "epoch": 0.50902, + "grad_norm": 1.346824231215947, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 50902 + }, + { + "epoch": 0.50903, + "grad_norm": 1.915873940215023, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 50903 + }, + { + "epoch": 0.50904, + "grad_norm": 1.3389404884503497, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 50904 + }, + { + "epoch": 0.50905, + "grad_norm": 1.783975166117801, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 50905 + }, + { + "epoch": 0.50906, + "grad_norm": 1.2362561433860875, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 50906 + }, + { + "epoch": 0.50907, + "grad_norm": 1.890891815174427, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 50907 + }, + { + "epoch": 0.50908, + "grad_norm": 1.7625067422521359, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 50908 + }, + { + "epoch": 0.50909, + "grad_norm": 1.6106188234810261, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50909 + }, + { + "epoch": 0.5091, + "grad_norm": 1.6701959173547904, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50910 + }, + { + "epoch": 0.50911, + "grad_norm": 1.2479638027942646, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 50911 + }, + { + "epoch": 0.50912, + "grad_norm": 1.5529636181667827, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 50912 + }, + { + "epoch": 0.50913, + "grad_norm": 1.59329807632668, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 50913 + }, + { + "epoch": 0.50914, + "grad_norm": 1.5715268328498313, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 50914 + }, + { + "epoch": 0.50915, + "grad_norm": 1.483921901707074, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 50915 + }, + { + "epoch": 0.50916, + "grad_norm": 1.5681793919055902, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 50916 + }, + { + "epoch": 0.50917, + "grad_norm": 1.706079851989887, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 50917 + }, + { + "epoch": 0.50918, + "grad_norm": 1.5626380122898929, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50918 + }, + { + "epoch": 0.50919, + "grad_norm": 1.623069568849669, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 50919 + }, + { + "epoch": 0.5092, + "grad_norm": 1.7242203485449308, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 50920 + }, + { + "epoch": 0.50921, + "grad_norm": 1.140782034105534, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 50921 + }, + { + "epoch": 0.50922, + "grad_norm": 1.9501221568305287, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 50922 + }, + { + "epoch": 0.50923, + "grad_norm": 1.2249577008253276, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 50923 + }, + { + "epoch": 0.50924, + "grad_norm": 1.863283985969505, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 50924 + }, + { + "epoch": 0.50925, + "grad_norm": 1.324656630363756, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 50925 + }, + { + "epoch": 0.50926, + "grad_norm": 1.4497140889576894, + "learning_rate": 0.003, + "loss": 4.017, + "step": 50926 + }, + { + "epoch": 0.50927, + "grad_norm": 1.7157335482378104, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 50927 + }, + { + "epoch": 0.50928, + "grad_norm": 1.7318118550497785, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 50928 + }, + { + "epoch": 0.50929, + "grad_norm": 2.129162588699757, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 50929 + }, + { + "epoch": 0.5093, + "grad_norm": 1.2412172646425388, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 50930 + }, + { + "epoch": 0.50931, + "grad_norm": 1.795065280465416, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 50931 + }, + { + "epoch": 0.50932, + "grad_norm": 1.5810865910184249, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 50932 + }, + { + "epoch": 0.50933, + "grad_norm": 1.7378989219361451, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 50933 + }, + { + "epoch": 0.50934, + "grad_norm": 1.4515724936292334, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 50934 + }, + { + "epoch": 0.50935, + "grad_norm": 1.8548111377609673, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 50935 + }, + { + "epoch": 0.50936, + "grad_norm": 1.429766468055553, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 50936 + }, + { + "epoch": 0.50937, + "grad_norm": 1.3763530667534853, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 50937 + }, + { + "epoch": 0.50938, + "grad_norm": 1.6641996891694104, + "learning_rate": 0.003, + "loss": 4.022, + "step": 50938 + }, + { + "epoch": 0.50939, + "grad_norm": 1.564718032875471, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 50939 + }, + { + "epoch": 0.5094, + "grad_norm": 1.5836839965278933, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 50940 + }, + { + "epoch": 0.50941, + "grad_norm": 1.5276169878048436, + "learning_rate": 0.003, + "loss": 4.032, + "step": 50941 + }, + { + "epoch": 0.50942, + "grad_norm": 1.6968935065894841, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 50942 + }, + { + "epoch": 0.50943, + "grad_norm": 1.4888670712748289, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 50943 + }, + { + "epoch": 0.50944, + "grad_norm": 1.537279975735043, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 50944 + }, + { + "epoch": 0.50945, + "grad_norm": 1.5334446775381096, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 50945 + }, + { + "epoch": 0.50946, + "grad_norm": 1.647564752230263, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 50946 + }, + { + "epoch": 0.50947, + "grad_norm": 1.5641808862835684, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 50947 + }, + { + "epoch": 0.50948, + "grad_norm": 1.4270926946506626, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 50948 + }, + { + "epoch": 0.50949, + "grad_norm": 1.6311212815253415, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 50949 + }, + { + "epoch": 0.5095, + "grad_norm": 1.5864383315404667, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 50950 + }, + { + "epoch": 0.50951, + "grad_norm": 1.576221146420163, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 50951 + }, + { + "epoch": 0.50952, + "grad_norm": 1.7228679876433015, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50952 + }, + { + "epoch": 0.50953, + "grad_norm": 1.3779347945953675, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50953 + }, + { + "epoch": 0.50954, + "grad_norm": 1.589045066749319, + "learning_rate": 0.003, + "loss": 4.031, + "step": 50954 + }, + { + "epoch": 0.50955, + "grad_norm": 1.3333702238922118, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50955 + }, + { + "epoch": 0.50956, + "grad_norm": 1.6068101051582757, + "learning_rate": 0.003, + "loss": 4.034, + "step": 50956 + }, + { + "epoch": 0.50957, + "grad_norm": 1.3555403466951244, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 50957 + }, + { + "epoch": 0.50958, + "grad_norm": 1.5872899201394561, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 50958 + }, + { + "epoch": 0.50959, + "grad_norm": 1.4488786497293793, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 50959 + }, + { + "epoch": 0.5096, + "grad_norm": 1.7435582679700115, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 50960 + }, + { + "epoch": 0.50961, + "grad_norm": 1.6638523189601042, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 50961 + }, + { + "epoch": 0.50962, + "grad_norm": 1.6994156516663632, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 50962 + }, + { + "epoch": 0.50963, + "grad_norm": 1.519842972353351, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 50963 + }, + { + "epoch": 0.50964, + "grad_norm": 1.7528474708251436, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 50964 + }, + { + "epoch": 0.50965, + "grad_norm": 1.564993864600597, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 50965 + }, + { + "epoch": 0.50966, + "grad_norm": 1.4476637071045615, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 50966 + }, + { + "epoch": 0.50967, + "grad_norm": 1.5033526491480425, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 50967 + }, + { + "epoch": 0.50968, + "grad_norm": 1.5440059087295832, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 50968 + }, + { + "epoch": 0.50969, + "grad_norm": 1.4780255823550847, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50969 + }, + { + "epoch": 0.5097, + "grad_norm": 1.52681977554603, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 50970 + }, + { + "epoch": 0.50971, + "grad_norm": 1.6317851492753102, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 50971 + }, + { + "epoch": 0.50972, + "grad_norm": 1.7893610542360079, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 50972 + }, + { + "epoch": 0.50973, + "grad_norm": 1.6451717085326483, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 50973 + }, + { + "epoch": 0.50974, + "grad_norm": 1.4481024298815068, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 50974 + }, + { + "epoch": 0.50975, + "grad_norm": 1.691191157166342, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 50975 + }, + { + "epoch": 0.50976, + "grad_norm": 1.3982084899166987, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 50976 + }, + { + "epoch": 0.50977, + "grad_norm": 1.7231817907639821, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 50977 + }, + { + "epoch": 0.50978, + "grad_norm": 1.3187668625655982, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 50978 + }, + { + "epoch": 0.50979, + "grad_norm": 1.7696495446498346, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50979 + }, + { + "epoch": 0.5098, + "grad_norm": 1.491270900280761, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 50980 + }, + { + "epoch": 0.50981, + "grad_norm": 1.5070918562371483, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 50981 + }, + { + "epoch": 0.50982, + "grad_norm": 1.5382294563067203, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 50982 + }, + { + "epoch": 0.50983, + "grad_norm": 1.7424035349663047, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 50983 + }, + { + "epoch": 0.50984, + "grad_norm": 1.6651182175287422, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 50984 + }, + { + "epoch": 0.50985, + "grad_norm": 1.7757403339385447, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 50985 + }, + { + "epoch": 0.50986, + "grad_norm": 1.6969539575482246, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 50986 + }, + { + "epoch": 0.50987, + "grad_norm": 1.5231978127726502, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 50987 + }, + { + "epoch": 0.50988, + "grad_norm": 1.6442125046970868, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 50988 + }, + { + "epoch": 0.50989, + "grad_norm": 1.6989203234123058, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 50989 + }, + { + "epoch": 0.5099, + "grad_norm": 1.4624222015368376, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 50990 + }, + { + "epoch": 0.50991, + "grad_norm": 1.792408086346165, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 50991 + }, + { + "epoch": 0.50992, + "grad_norm": 1.5250235939264236, + "learning_rate": 0.003, + "loss": 4.033, + "step": 50992 + }, + { + "epoch": 0.50993, + "grad_norm": 1.9212260947842106, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 50993 + }, + { + "epoch": 0.50994, + "grad_norm": 1.6081604775914071, + "learning_rate": 0.003, + "loss": 4.039, + "step": 50994 + }, + { + "epoch": 0.50995, + "grad_norm": 1.33196445239802, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 50995 + }, + { + "epoch": 0.50996, + "grad_norm": 1.4784590200688175, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 50996 + }, + { + "epoch": 0.50997, + "grad_norm": 1.4219277018906278, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 50997 + }, + { + "epoch": 0.50998, + "grad_norm": 1.6941409644830234, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 50998 + }, + { + "epoch": 0.50999, + "grad_norm": 1.5500543008009817, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 50999 + }, + { + "epoch": 0.51, + "grad_norm": 1.5932758595157608, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 51000 + }, + { + "epoch": 0.51001, + "grad_norm": 1.639889580967102, + "learning_rate": 0.003, + "loss": 4.072, + "step": 51001 + }, + { + "epoch": 0.51002, + "grad_norm": 1.2725310334908244, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 51002 + }, + { + "epoch": 0.51003, + "grad_norm": 1.602141205089841, + "learning_rate": 0.003, + "loss": 4.061, + "step": 51003 + }, + { + "epoch": 0.51004, + "grad_norm": 1.5329779993337622, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 51004 + }, + { + "epoch": 0.51005, + "grad_norm": 1.7942747256313558, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 51005 + }, + { + "epoch": 0.51006, + "grad_norm": 1.7200132510926722, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 51006 + }, + { + "epoch": 0.51007, + "grad_norm": 1.3208672068639815, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51007 + }, + { + "epoch": 0.51008, + "grad_norm": 1.853003012693865, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 51008 + }, + { + "epoch": 0.51009, + "grad_norm": 1.1500524924338071, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51009 + }, + { + "epoch": 0.5101, + "grad_norm": 1.730594573845929, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51010 + }, + { + "epoch": 0.51011, + "grad_norm": 1.4085031890861461, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51011 + }, + { + "epoch": 0.51012, + "grad_norm": 1.4896814218464065, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51012 + }, + { + "epoch": 0.51013, + "grad_norm": 1.3976105873190705, + "learning_rate": 0.003, + "loss": 4.041, + "step": 51013 + }, + { + "epoch": 0.51014, + "grad_norm": 1.823465688056142, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51014 + }, + { + "epoch": 0.51015, + "grad_norm": 1.8955891620883718, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 51015 + }, + { + "epoch": 0.51016, + "grad_norm": 1.5204094017442824, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51016 + }, + { + "epoch": 0.51017, + "grad_norm": 1.4713942864731704, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 51017 + }, + { + "epoch": 0.51018, + "grad_norm": 1.5815994795434631, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 51018 + }, + { + "epoch": 0.51019, + "grad_norm": 1.3391958196180327, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 51019 + }, + { + "epoch": 0.5102, + "grad_norm": 1.6886848385111495, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 51020 + }, + { + "epoch": 0.51021, + "grad_norm": 1.557096924284872, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 51021 + }, + { + "epoch": 0.51022, + "grad_norm": 1.60965176306314, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 51022 + }, + { + "epoch": 0.51023, + "grad_norm": 1.5574341529561493, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 51023 + }, + { + "epoch": 0.51024, + "grad_norm": 1.630198209024033, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51024 + }, + { + "epoch": 0.51025, + "grad_norm": 1.457715273661437, + "learning_rate": 0.003, + "loss": 4.011, + "step": 51025 + }, + { + "epoch": 0.51026, + "grad_norm": 1.541015996468764, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 51026 + }, + { + "epoch": 0.51027, + "grad_norm": 1.565864278622047, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 51027 + }, + { + "epoch": 0.51028, + "grad_norm": 1.3690644391173827, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 51028 + }, + { + "epoch": 0.51029, + "grad_norm": 1.6929708645357484, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51029 + }, + { + "epoch": 0.5103, + "grad_norm": 1.612342650079766, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 51030 + }, + { + "epoch": 0.51031, + "grad_norm": 1.43264302207003, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 51031 + }, + { + "epoch": 0.51032, + "grad_norm": 1.4703918163190794, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 51032 + }, + { + "epoch": 0.51033, + "grad_norm": 1.5759096606357736, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 51033 + }, + { + "epoch": 0.51034, + "grad_norm": 1.309418359409977, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 51034 + }, + { + "epoch": 0.51035, + "grad_norm": 1.5208380589417114, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51035 + }, + { + "epoch": 0.51036, + "grad_norm": 1.5657735150801635, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51036 + }, + { + "epoch": 0.51037, + "grad_norm": 1.417755743696076, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 51037 + }, + { + "epoch": 0.51038, + "grad_norm": 2.3241103358303983, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 51038 + }, + { + "epoch": 0.51039, + "grad_norm": 1.521441175747266, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 51039 + }, + { + "epoch": 0.5104, + "grad_norm": 1.9321557361738222, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 51040 + }, + { + "epoch": 0.51041, + "grad_norm": 1.9677761231348234, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51041 + }, + { + "epoch": 0.51042, + "grad_norm": 1.3112319947735473, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 51042 + }, + { + "epoch": 0.51043, + "grad_norm": 1.7930539820785607, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51043 + }, + { + "epoch": 0.51044, + "grad_norm": 1.5679895503920815, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51044 + }, + { + "epoch": 0.51045, + "grad_norm": 1.6568202535568333, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51045 + }, + { + "epoch": 0.51046, + "grad_norm": 1.5288697841268486, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 51046 + }, + { + "epoch": 0.51047, + "grad_norm": 2.061614033656634, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 51047 + }, + { + "epoch": 0.51048, + "grad_norm": 1.4112355235005658, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 51048 + }, + { + "epoch": 0.51049, + "grad_norm": 1.612187524446505, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 51049 + }, + { + "epoch": 0.5105, + "grad_norm": 1.5248169952834731, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 51050 + }, + { + "epoch": 0.51051, + "grad_norm": 1.7723193755707307, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 51051 + }, + { + "epoch": 0.51052, + "grad_norm": 1.3542234079669486, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51052 + }, + { + "epoch": 0.51053, + "grad_norm": 1.5939443423510964, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 51053 + }, + { + "epoch": 0.51054, + "grad_norm": 1.2531722769461564, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 51054 + }, + { + "epoch": 0.51055, + "grad_norm": 1.7000762269885972, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 51055 + }, + { + "epoch": 0.51056, + "grad_norm": 1.4978684449101412, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51056 + }, + { + "epoch": 0.51057, + "grad_norm": 2.023549682965934, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 51057 + }, + { + "epoch": 0.51058, + "grad_norm": 1.7871542009973242, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 51058 + }, + { + "epoch": 0.51059, + "grad_norm": 1.9457551480546291, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51059 + }, + { + "epoch": 0.5106, + "grad_norm": 1.476759965376664, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 51060 + }, + { + "epoch": 0.51061, + "grad_norm": 1.5937182634604035, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 51061 + }, + { + "epoch": 0.51062, + "grad_norm": 1.5640159020893651, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 51062 + }, + { + "epoch": 0.51063, + "grad_norm": 1.6648573248864937, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 51063 + }, + { + "epoch": 0.51064, + "grad_norm": 1.2848214458385379, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 51064 + }, + { + "epoch": 0.51065, + "grad_norm": 1.634929751871227, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 51065 + }, + { + "epoch": 0.51066, + "grad_norm": 1.342842571486023, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 51066 + }, + { + "epoch": 0.51067, + "grad_norm": 1.6659275606688604, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 51067 + }, + { + "epoch": 0.51068, + "grad_norm": 1.370924315874315, + "learning_rate": 0.003, + "loss": 4.029, + "step": 51068 + }, + { + "epoch": 0.51069, + "grad_norm": 1.5834257357857409, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 51069 + }, + { + "epoch": 0.5107, + "grad_norm": 1.7554699772968576, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51070 + }, + { + "epoch": 0.51071, + "grad_norm": 1.9187963532414172, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 51071 + }, + { + "epoch": 0.51072, + "grad_norm": 1.5402329572794913, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 51072 + }, + { + "epoch": 0.51073, + "grad_norm": 1.7060455151245872, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 51073 + }, + { + "epoch": 0.51074, + "grad_norm": 1.445886251413068, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 51074 + }, + { + "epoch": 0.51075, + "grad_norm": 1.8061414777720017, + "learning_rate": 0.003, + "loss": 4.018, + "step": 51075 + }, + { + "epoch": 0.51076, + "grad_norm": 1.2700962174703543, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51076 + }, + { + "epoch": 0.51077, + "grad_norm": 1.6740547058746373, + "learning_rate": 0.003, + "loss": 4.04, + "step": 51077 + }, + { + "epoch": 0.51078, + "grad_norm": 1.5382360837329603, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 51078 + }, + { + "epoch": 0.51079, + "grad_norm": 1.5727782699048807, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 51079 + }, + { + "epoch": 0.5108, + "grad_norm": 1.6464195757084727, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51080 + }, + { + "epoch": 0.51081, + "grad_norm": 1.3223413251920106, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51081 + }, + { + "epoch": 0.51082, + "grad_norm": 1.5782346744490063, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 51082 + }, + { + "epoch": 0.51083, + "grad_norm": 1.698465372259566, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 51083 + }, + { + "epoch": 0.51084, + "grad_norm": 1.549490825287196, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 51084 + }, + { + "epoch": 0.51085, + "grad_norm": 1.396532094981883, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 51085 + }, + { + "epoch": 0.51086, + "grad_norm": 1.4937418095571182, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 51086 + }, + { + "epoch": 0.51087, + "grad_norm": 1.4901412334055644, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 51087 + }, + { + "epoch": 0.51088, + "grad_norm": 1.3706595327279127, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 51088 + }, + { + "epoch": 0.51089, + "grad_norm": 1.6064111035396205, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 51089 + }, + { + "epoch": 0.5109, + "grad_norm": 1.4110874774511444, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 51090 + }, + { + "epoch": 0.51091, + "grad_norm": 1.7488725702061687, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 51091 + }, + { + "epoch": 0.51092, + "grad_norm": 1.503542502251722, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 51092 + }, + { + "epoch": 0.51093, + "grad_norm": 1.4181662756548161, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 51093 + }, + { + "epoch": 0.51094, + "grad_norm": 1.6570708430422365, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 51094 + }, + { + "epoch": 0.51095, + "grad_norm": 1.535453171393185, + "learning_rate": 0.003, + "loss": 4.018, + "step": 51095 + }, + { + "epoch": 0.51096, + "grad_norm": 1.6200724869549081, + "learning_rate": 0.003, + "loss": 4.023, + "step": 51096 + }, + { + "epoch": 0.51097, + "grad_norm": 1.6669322466742806, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 51097 + }, + { + "epoch": 0.51098, + "grad_norm": 1.4747345357751709, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 51098 + }, + { + "epoch": 0.51099, + "grad_norm": 1.7314129030933136, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 51099 + }, + { + "epoch": 0.511, + "grad_norm": 1.7789080377957134, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51100 + }, + { + "epoch": 0.51101, + "grad_norm": 2.008245887554292, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 51101 + }, + { + "epoch": 0.51102, + "grad_norm": 1.5170546676564243, + "learning_rate": 0.003, + "loss": 4.041, + "step": 51102 + }, + { + "epoch": 0.51103, + "grad_norm": 1.6363934148263612, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 51103 + }, + { + "epoch": 0.51104, + "grad_norm": 1.5527936307454615, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 51104 + }, + { + "epoch": 0.51105, + "grad_norm": 1.3336094336064717, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 51105 + }, + { + "epoch": 0.51106, + "grad_norm": 1.7770874484899153, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 51106 + }, + { + "epoch": 0.51107, + "grad_norm": 1.351485013703423, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 51107 + }, + { + "epoch": 0.51108, + "grad_norm": 1.705148272334536, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 51108 + }, + { + "epoch": 0.51109, + "grad_norm": 1.2522058885410265, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 51109 + }, + { + "epoch": 0.5111, + "grad_norm": 1.7212943273613024, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51110 + }, + { + "epoch": 0.51111, + "grad_norm": 1.625123692003469, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51111 + }, + { + "epoch": 0.51112, + "grad_norm": 1.9723882131125154, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 51112 + }, + { + "epoch": 0.51113, + "grad_norm": 1.382780675625774, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 51113 + }, + { + "epoch": 0.51114, + "grad_norm": 1.7832957724848868, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 51114 + }, + { + "epoch": 0.51115, + "grad_norm": 1.6980517764305276, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 51115 + }, + { + "epoch": 0.51116, + "grad_norm": 1.3815725396520389, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51116 + }, + { + "epoch": 0.51117, + "grad_norm": 1.442988750488706, + "learning_rate": 0.003, + "loss": 4.023, + "step": 51117 + }, + { + "epoch": 0.51118, + "grad_norm": 1.230953764505247, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 51118 + }, + { + "epoch": 0.51119, + "grad_norm": 2.0569495830229716, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 51119 + }, + { + "epoch": 0.5112, + "grad_norm": 1.541430512672491, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 51120 + }, + { + "epoch": 0.51121, + "grad_norm": 2.047836791290045, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51121 + }, + { + "epoch": 0.51122, + "grad_norm": 1.354637034759406, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51122 + }, + { + "epoch": 0.51123, + "grad_norm": 1.8889793070001717, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51123 + }, + { + "epoch": 0.51124, + "grad_norm": 1.6653496280634597, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 51124 + }, + { + "epoch": 0.51125, + "grad_norm": 1.4119178118263647, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 51125 + }, + { + "epoch": 0.51126, + "grad_norm": 1.4715120908704014, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 51126 + }, + { + "epoch": 0.51127, + "grad_norm": 1.5833867026917114, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 51127 + }, + { + "epoch": 0.51128, + "grad_norm": 1.3608634245568147, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 51128 + }, + { + "epoch": 0.51129, + "grad_norm": 1.5791775692053611, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 51129 + }, + { + "epoch": 0.5113, + "grad_norm": 1.416806857190009, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51130 + }, + { + "epoch": 0.51131, + "grad_norm": 1.7082978499072186, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 51131 + }, + { + "epoch": 0.51132, + "grad_norm": 1.475093243871082, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 51132 + }, + { + "epoch": 0.51133, + "grad_norm": 1.5548232832828421, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 51133 + }, + { + "epoch": 0.51134, + "grad_norm": 1.662996936894898, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 51134 + }, + { + "epoch": 0.51135, + "grad_norm": 1.5111114402197985, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 51135 + }, + { + "epoch": 0.51136, + "grad_norm": 1.7130039864775832, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51136 + }, + { + "epoch": 0.51137, + "grad_norm": 1.302218957516452, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 51137 + }, + { + "epoch": 0.51138, + "grad_norm": 1.667741870046485, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51138 + }, + { + "epoch": 0.51139, + "grad_norm": 1.5358786412978818, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51139 + }, + { + "epoch": 0.5114, + "grad_norm": 1.230221983921038, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 51140 + }, + { + "epoch": 0.51141, + "grad_norm": 1.6863603814588028, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 51141 + }, + { + "epoch": 0.51142, + "grad_norm": 1.6348574638954048, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 51142 + }, + { + "epoch": 0.51143, + "grad_norm": 1.591436587963657, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 51143 + }, + { + "epoch": 0.51144, + "grad_norm": 1.7132653946724377, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 51144 + }, + { + "epoch": 0.51145, + "grad_norm": 1.3476758255523897, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 51145 + }, + { + "epoch": 0.51146, + "grad_norm": 1.3186024010882926, + "learning_rate": 0.003, + "loss": 4.037, + "step": 51146 + }, + { + "epoch": 0.51147, + "grad_norm": 1.7633186382521366, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 51147 + }, + { + "epoch": 0.51148, + "grad_norm": 1.5085988742628444, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 51148 + }, + { + "epoch": 0.51149, + "grad_norm": 1.4148242162590043, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 51149 + }, + { + "epoch": 0.5115, + "grad_norm": 1.671933443403937, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 51150 + }, + { + "epoch": 0.51151, + "grad_norm": 1.588468435443876, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 51151 + }, + { + "epoch": 0.51152, + "grad_norm": 1.2185466556105797, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 51152 + }, + { + "epoch": 0.51153, + "grad_norm": 1.8607073308896036, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 51153 + }, + { + "epoch": 0.51154, + "grad_norm": 1.324876386393465, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 51154 + }, + { + "epoch": 0.51155, + "grad_norm": 1.6718386679669779, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 51155 + }, + { + "epoch": 0.51156, + "grad_norm": 1.3614897714791963, + "learning_rate": 0.003, + "loss": 4.034, + "step": 51156 + }, + { + "epoch": 0.51157, + "grad_norm": 1.5789828127289454, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51157 + }, + { + "epoch": 0.51158, + "grad_norm": 1.5467097052102239, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51158 + }, + { + "epoch": 0.51159, + "grad_norm": 1.8141340749912578, + "learning_rate": 0.003, + "loss": 4.03, + "step": 51159 + }, + { + "epoch": 0.5116, + "grad_norm": 1.5499911538339115, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 51160 + }, + { + "epoch": 0.51161, + "grad_norm": 1.7350136922680628, + "learning_rate": 0.003, + "loss": 4.048, + "step": 51161 + }, + { + "epoch": 0.51162, + "grad_norm": 1.8420998458938924, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 51162 + }, + { + "epoch": 0.51163, + "grad_norm": 1.331151419141666, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 51163 + }, + { + "epoch": 0.51164, + "grad_norm": 1.5867974920723233, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 51164 + }, + { + "epoch": 0.51165, + "grad_norm": 1.4560676933291197, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 51165 + }, + { + "epoch": 0.51166, + "grad_norm": 1.870925498678557, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 51166 + }, + { + "epoch": 0.51167, + "grad_norm": 1.3654971127964795, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 51167 + }, + { + "epoch": 0.51168, + "grad_norm": 1.8522592149401127, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 51168 + }, + { + "epoch": 0.51169, + "grad_norm": 1.4291250950727012, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51169 + }, + { + "epoch": 0.5117, + "grad_norm": 1.4705943818668596, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 51170 + }, + { + "epoch": 0.51171, + "grad_norm": 1.626426404940036, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 51171 + }, + { + "epoch": 0.51172, + "grad_norm": 1.440279911207778, + "learning_rate": 0.003, + "loss": 4.026, + "step": 51172 + }, + { + "epoch": 0.51173, + "grad_norm": 1.6712926675893949, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51173 + }, + { + "epoch": 0.51174, + "grad_norm": 1.5873477258599338, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 51174 + }, + { + "epoch": 0.51175, + "grad_norm": 1.6994106503214446, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51175 + }, + { + "epoch": 0.51176, + "grad_norm": 1.3170173126086424, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51176 + }, + { + "epoch": 0.51177, + "grad_norm": 1.707805931682472, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 51177 + }, + { + "epoch": 0.51178, + "grad_norm": 1.35010249088562, + "learning_rate": 0.003, + "loss": 4.014, + "step": 51178 + }, + { + "epoch": 0.51179, + "grad_norm": 1.7183078266859253, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51179 + }, + { + "epoch": 0.5118, + "grad_norm": 1.3134641837560233, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 51180 + }, + { + "epoch": 0.51181, + "grad_norm": 1.691120686143733, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51181 + }, + { + "epoch": 0.51182, + "grad_norm": 1.2891072325144388, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51182 + }, + { + "epoch": 0.51183, + "grad_norm": 1.6861737065315825, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51183 + }, + { + "epoch": 0.51184, + "grad_norm": 1.4168693815228688, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 51184 + }, + { + "epoch": 0.51185, + "grad_norm": 2.017926150698442, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 51185 + }, + { + "epoch": 0.51186, + "grad_norm": 1.3951539964058826, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51186 + }, + { + "epoch": 0.51187, + "grad_norm": 2.091457936881222, + "learning_rate": 0.003, + "loss": 4.049, + "step": 51187 + }, + { + "epoch": 0.51188, + "grad_norm": 1.447371157048677, + "learning_rate": 0.003, + "loss": 4.073, + "step": 51188 + }, + { + "epoch": 0.51189, + "grad_norm": 1.5598489022766453, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51189 + }, + { + "epoch": 0.5119, + "grad_norm": 1.7462079224823635, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 51190 + }, + { + "epoch": 0.51191, + "grad_norm": 1.3521417141115053, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51191 + }, + { + "epoch": 0.51192, + "grad_norm": 1.7134209137317642, + "learning_rate": 0.003, + "loss": 4.056, + "step": 51192 + }, + { + "epoch": 0.51193, + "grad_norm": 1.468206415887436, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 51193 + }, + { + "epoch": 0.51194, + "grad_norm": 1.6172515770970035, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 51194 + }, + { + "epoch": 0.51195, + "grad_norm": 1.383843754409078, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 51195 + }, + { + "epoch": 0.51196, + "grad_norm": 1.4489863338180062, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 51196 + }, + { + "epoch": 0.51197, + "grad_norm": 1.699681857661678, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 51197 + }, + { + "epoch": 0.51198, + "grad_norm": 1.3520114557632021, + "learning_rate": 0.003, + "loss": 4.068, + "step": 51198 + }, + { + "epoch": 0.51199, + "grad_norm": 1.757285659274222, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 51199 + }, + { + "epoch": 0.512, + "grad_norm": 1.7540388204280122, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 51200 + }, + { + "epoch": 0.51201, + "grad_norm": 1.4505061696033048, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 51201 + }, + { + "epoch": 0.51202, + "grad_norm": 1.8613296525986875, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 51202 + }, + { + "epoch": 0.51203, + "grad_norm": 1.5310803762973892, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 51203 + }, + { + "epoch": 0.51204, + "grad_norm": 1.754561979034162, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51204 + }, + { + "epoch": 0.51205, + "grad_norm": 1.6147136247534475, + "learning_rate": 0.003, + "loss": 4.049, + "step": 51205 + }, + { + "epoch": 0.51206, + "grad_norm": 1.4896529916136925, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 51206 + }, + { + "epoch": 0.51207, + "grad_norm": 1.8037240824658876, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 51207 + }, + { + "epoch": 0.51208, + "grad_norm": 1.7837622628119594, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51208 + }, + { + "epoch": 0.51209, + "grad_norm": 1.2405653730639308, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 51209 + }, + { + "epoch": 0.5121, + "grad_norm": 1.6564906890287963, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51210 + }, + { + "epoch": 0.51211, + "grad_norm": 1.448168449270582, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 51211 + }, + { + "epoch": 0.51212, + "grad_norm": 1.8506915860577464, + "learning_rate": 0.003, + "loss": 4.038, + "step": 51212 + }, + { + "epoch": 0.51213, + "grad_norm": 1.5699420514867932, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 51213 + }, + { + "epoch": 0.51214, + "grad_norm": 1.4704116297301604, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51214 + }, + { + "epoch": 0.51215, + "grad_norm": 1.6625364269267149, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 51215 + }, + { + "epoch": 0.51216, + "grad_norm": 1.3722392327334856, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 51216 + }, + { + "epoch": 0.51217, + "grad_norm": 1.4838940735235342, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 51217 + }, + { + "epoch": 0.51218, + "grad_norm": 1.6514105548493094, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 51218 + }, + { + "epoch": 0.51219, + "grad_norm": 1.377254879957822, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51219 + }, + { + "epoch": 0.5122, + "grad_norm": 1.696695187798047, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 51220 + }, + { + "epoch": 0.51221, + "grad_norm": 1.500873193785215, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 51221 + }, + { + "epoch": 0.51222, + "grad_norm": 1.8285207434109834, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 51222 + }, + { + "epoch": 0.51223, + "grad_norm": 1.3677640085620866, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 51223 + }, + { + "epoch": 0.51224, + "grad_norm": 1.7312792567129995, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 51224 + }, + { + "epoch": 0.51225, + "grad_norm": 1.5762481451281212, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51225 + }, + { + "epoch": 0.51226, + "grad_norm": 1.2813161631887489, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 51226 + }, + { + "epoch": 0.51227, + "grad_norm": 1.4795013264787282, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 51227 + }, + { + "epoch": 0.51228, + "grad_norm": 1.5919034715674822, + "learning_rate": 0.003, + "loss": 4.048, + "step": 51228 + }, + { + "epoch": 0.51229, + "grad_norm": 1.522581875715783, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51229 + }, + { + "epoch": 0.5123, + "grad_norm": 1.4990319138737407, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 51230 + }, + { + "epoch": 0.51231, + "grad_norm": 1.6081629470707848, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 51231 + }, + { + "epoch": 0.51232, + "grad_norm": 1.4281639745765795, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51232 + }, + { + "epoch": 0.51233, + "grad_norm": 1.367282438335221, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 51233 + }, + { + "epoch": 0.51234, + "grad_norm": 1.5242746626948427, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51234 + }, + { + "epoch": 0.51235, + "grad_norm": 1.6053271994487959, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 51235 + }, + { + "epoch": 0.51236, + "grad_norm": 1.5874943082245465, + "learning_rate": 0.003, + "loss": 4.008, + "step": 51236 + }, + { + "epoch": 0.51237, + "grad_norm": 1.3638853468566665, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 51237 + }, + { + "epoch": 0.51238, + "grad_norm": 1.9793947730604502, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 51238 + }, + { + "epoch": 0.51239, + "grad_norm": 1.2759235778919886, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51239 + }, + { + "epoch": 0.5124, + "grad_norm": 1.77738503253274, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 51240 + }, + { + "epoch": 0.51241, + "grad_norm": 1.360856405018428, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51241 + }, + { + "epoch": 0.51242, + "grad_norm": 2.005574812673528, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 51242 + }, + { + "epoch": 0.51243, + "grad_norm": 1.408775045522796, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 51243 + }, + { + "epoch": 0.51244, + "grad_norm": 1.4466596278036632, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 51244 + }, + { + "epoch": 0.51245, + "grad_norm": 1.7727224714505996, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 51245 + }, + { + "epoch": 0.51246, + "grad_norm": 1.4774326312189803, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 51246 + }, + { + "epoch": 0.51247, + "grad_norm": 1.6349187617689693, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 51247 + }, + { + "epoch": 0.51248, + "grad_norm": 1.389217721523753, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 51248 + }, + { + "epoch": 0.51249, + "grad_norm": 1.6176850023696254, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 51249 + }, + { + "epoch": 0.5125, + "grad_norm": 1.8344270259974578, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51250 + }, + { + "epoch": 0.51251, + "grad_norm": 1.6204474037410272, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51251 + }, + { + "epoch": 0.51252, + "grad_norm": 1.4857982596896269, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 51252 + }, + { + "epoch": 0.51253, + "grad_norm": 1.4287916042262054, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 51253 + }, + { + "epoch": 0.51254, + "grad_norm": 1.7811418034404745, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 51254 + }, + { + "epoch": 0.51255, + "grad_norm": 1.4998023695776226, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 51255 + }, + { + "epoch": 0.51256, + "grad_norm": 1.556220732392439, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 51256 + }, + { + "epoch": 0.51257, + "grad_norm": 1.3061746494316147, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 51257 + }, + { + "epoch": 0.51258, + "grad_norm": 1.7117468160270264, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 51258 + }, + { + "epoch": 0.51259, + "grad_norm": 1.4596199736916085, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 51259 + }, + { + "epoch": 0.5126, + "grad_norm": 1.664977580242391, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 51260 + }, + { + "epoch": 0.51261, + "grad_norm": 1.5003414621432363, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 51261 + }, + { + "epoch": 0.51262, + "grad_norm": 1.6065509359397678, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 51262 + }, + { + "epoch": 0.51263, + "grad_norm": 1.8755129377445483, + "learning_rate": 0.003, + "loss": 4.064, + "step": 51263 + }, + { + "epoch": 0.51264, + "grad_norm": 1.5441056881958317, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 51264 + }, + { + "epoch": 0.51265, + "grad_norm": 1.575986522030534, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51265 + }, + { + "epoch": 0.51266, + "grad_norm": 1.4859826009886947, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51266 + }, + { + "epoch": 0.51267, + "grad_norm": 1.8534744534434073, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 51267 + }, + { + "epoch": 0.51268, + "grad_norm": 1.538822949613574, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 51268 + }, + { + "epoch": 0.51269, + "grad_norm": 1.3983207026689843, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 51269 + }, + { + "epoch": 0.5127, + "grad_norm": 1.5226693836606044, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51270 + }, + { + "epoch": 0.51271, + "grad_norm": 1.6235845032076195, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51271 + }, + { + "epoch": 0.51272, + "grad_norm": 1.418696416352583, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51272 + }, + { + "epoch": 0.51273, + "grad_norm": 1.3707676813481178, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 51273 + }, + { + "epoch": 0.51274, + "grad_norm": 1.5333365949718718, + "learning_rate": 0.003, + "loss": 4.038, + "step": 51274 + }, + { + "epoch": 0.51275, + "grad_norm": 1.4734857398140282, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 51275 + }, + { + "epoch": 0.51276, + "grad_norm": 1.751554094382518, + "learning_rate": 0.003, + "loss": 4.02, + "step": 51276 + }, + { + "epoch": 0.51277, + "grad_norm": 1.5415424763115915, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 51277 + }, + { + "epoch": 0.51278, + "grad_norm": 1.73952276401917, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 51278 + }, + { + "epoch": 0.51279, + "grad_norm": 1.225790334131462, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51279 + }, + { + "epoch": 0.5128, + "grad_norm": 1.8505197126100905, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51280 + }, + { + "epoch": 0.51281, + "grad_norm": 1.5689004803223883, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51281 + }, + { + "epoch": 0.51282, + "grad_norm": 1.6320828743837978, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51282 + }, + { + "epoch": 0.51283, + "grad_norm": 1.479534138148139, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 51283 + }, + { + "epoch": 0.51284, + "grad_norm": 1.9983332471680102, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 51284 + }, + { + "epoch": 0.51285, + "grad_norm": 1.2439182079849167, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 51285 + }, + { + "epoch": 0.51286, + "grad_norm": 1.611990742820919, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51286 + }, + { + "epoch": 0.51287, + "grad_norm": 1.359802908784393, + "learning_rate": 0.003, + "loss": 4.062, + "step": 51287 + }, + { + "epoch": 0.51288, + "grad_norm": 1.7809518252392345, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51288 + }, + { + "epoch": 0.51289, + "grad_norm": 1.708756674827299, + "learning_rate": 0.003, + "loss": 4.026, + "step": 51289 + }, + { + "epoch": 0.5129, + "grad_norm": 1.39936959565031, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51290 + }, + { + "epoch": 0.51291, + "grad_norm": 2.001514188941458, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51291 + }, + { + "epoch": 0.51292, + "grad_norm": 1.5363487694579465, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 51292 + }, + { + "epoch": 0.51293, + "grad_norm": 1.8046472930076758, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 51293 + }, + { + "epoch": 0.51294, + "grad_norm": 1.3022179138421226, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 51294 + }, + { + "epoch": 0.51295, + "grad_norm": 1.7366245547567551, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 51295 + }, + { + "epoch": 0.51296, + "grad_norm": 1.3601456041730364, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 51296 + }, + { + "epoch": 0.51297, + "grad_norm": 1.4877965065193262, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51297 + }, + { + "epoch": 0.51298, + "grad_norm": 1.5200028546755662, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51298 + }, + { + "epoch": 0.51299, + "grad_norm": 1.4858558638714183, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 51299 + }, + { + "epoch": 0.513, + "grad_norm": 1.3447845299237982, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 51300 + }, + { + "epoch": 0.51301, + "grad_norm": 1.411270157736964, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51301 + }, + { + "epoch": 0.51302, + "grad_norm": 1.8860614541104355, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 51302 + }, + { + "epoch": 0.51303, + "grad_norm": 1.7944740110645845, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 51303 + }, + { + "epoch": 0.51304, + "grad_norm": 1.4767364480302108, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51304 + }, + { + "epoch": 0.51305, + "grad_norm": 1.8128510781550946, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51305 + }, + { + "epoch": 0.51306, + "grad_norm": 1.3650927084397286, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 51306 + }, + { + "epoch": 0.51307, + "grad_norm": 1.7448494923165256, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 51307 + }, + { + "epoch": 0.51308, + "grad_norm": 1.565064401064547, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 51308 + }, + { + "epoch": 0.51309, + "grad_norm": 1.5899858625455103, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 51309 + }, + { + "epoch": 0.5131, + "grad_norm": 1.3273313534264557, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51310 + }, + { + "epoch": 0.51311, + "grad_norm": 1.7023709357394434, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51311 + }, + { + "epoch": 0.51312, + "grad_norm": 1.4463512171022679, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51312 + }, + { + "epoch": 0.51313, + "grad_norm": 2.0370287866913683, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 51313 + }, + { + "epoch": 0.51314, + "grad_norm": 1.2069900386474364, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51314 + }, + { + "epoch": 0.51315, + "grad_norm": 1.8742017328959117, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 51315 + }, + { + "epoch": 0.51316, + "grad_norm": 1.629677428256179, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 51316 + }, + { + "epoch": 0.51317, + "grad_norm": 1.683668416040639, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 51317 + }, + { + "epoch": 0.51318, + "grad_norm": 1.6139165548040662, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 51318 + }, + { + "epoch": 0.51319, + "grad_norm": 1.4672394842090506, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 51319 + }, + { + "epoch": 0.5132, + "grad_norm": 1.5961839332124348, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 51320 + }, + { + "epoch": 0.51321, + "grad_norm": 1.5636365059494672, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51321 + }, + { + "epoch": 0.51322, + "grad_norm": 1.4137615675182222, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 51322 + }, + { + "epoch": 0.51323, + "grad_norm": 1.5851547704809552, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 51323 + }, + { + "epoch": 0.51324, + "grad_norm": 1.3503017413619702, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 51324 + }, + { + "epoch": 0.51325, + "grad_norm": 1.7307623140830208, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 51325 + }, + { + "epoch": 0.51326, + "grad_norm": 1.4425162056899228, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 51326 + }, + { + "epoch": 0.51327, + "grad_norm": 1.6473952114034196, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 51327 + }, + { + "epoch": 0.51328, + "grad_norm": 1.3755926799657192, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 51328 + }, + { + "epoch": 0.51329, + "grad_norm": 1.697134423874978, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 51329 + }, + { + "epoch": 0.5133, + "grad_norm": 1.5454385654243876, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 51330 + }, + { + "epoch": 0.51331, + "grad_norm": 1.7439873162332031, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 51331 + }, + { + "epoch": 0.51332, + "grad_norm": 1.7688379953855144, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 51332 + }, + { + "epoch": 0.51333, + "grad_norm": 1.8244263436936992, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51333 + }, + { + "epoch": 0.51334, + "grad_norm": 1.3537494510885195, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 51334 + }, + { + "epoch": 0.51335, + "grad_norm": 1.9278411104176105, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51335 + }, + { + "epoch": 0.51336, + "grad_norm": 1.6703357827279632, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51336 + }, + { + "epoch": 0.51337, + "grad_norm": 1.5286301516561436, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51337 + }, + { + "epoch": 0.51338, + "grad_norm": 1.566373028802153, + "learning_rate": 0.003, + "loss": 4.049, + "step": 51338 + }, + { + "epoch": 0.51339, + "grad_norm": 1.4118172148732513, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 51339 + }, + { + "epoch": 0.5134, + "grad_norm": 1.4593041829082003, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 51340 + }, + { + "epoch": 0.51341, + "grad_norm": 1.767058932796948, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 51341 + }, + { + "epoch": 0.51342, + "grad_norm": 1.4981104776791372, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 51342 + }, + { + "epoch": 0.51343, + "grad_norm": 1.5620230400795851, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 51343 + }, + { + "epoch": 0.51344, + "grad_norm": 1.4024943519192659, + "learning_rate": 0.003, + "loss": 4.006, + "step": 51344 + }, + { + "epoch": 0.51345, + "grad_norm": 1.5083064741572003, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 51345 + }, + { + "epoch": 0.51346, + "grad_norm": 1.4171948237229801, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 51346 + }, + { + "epoch": 0.51347, + "grad_norm": 1.3480132920594865, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 51347 + }, + { + "epoch": 0.51348, + "grad_norm": 1.7347626100638704, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 51348 + }, + { + "epoch": 0.51349, + "grad_norm": 1.4251262650306002, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51349 + }, + { + "epoch": 0.5135, + "grad_norm": 1.5807311040049028, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51350 + }, + { + "epoch": 0.51351, + "grad_norm": 1.4313226780759538, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 51351 + }, + { + "epoch": 0.51352, + "grad_norm": 1.5744603006017364, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 51352 + }, + { + "epoch": 0.51353, + "grad_norm": 1.4509822960217391, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 51353 + }, + { + "epoch": 0.51354, + "grad_norm": 1.4842747345059797, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 51354 + }, + { + "epoch": 0.51355, + "grad_norm": 1.7938417025914168, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51355 + }, + { + "epoch": 0.51356, + "grad_norm": 1.8764785861366136, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 51356 + }, + { + "epoch": 0.51357, + "grad_norm": 1.7540279652277557, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 51357 + }, + { + "epoch": 0.51358, + "grad_norm": 1.6316390925783189, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 51358 + }, + { + "epoch": 0.51359, + "grad_norm": 1.2534486322322995, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 51359 + }, + { + "epoch": 0.5136, + "grad_norm": 2.00944790865361, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 51360 + }, + { + "epoch": 0.51361, + "grad_norm": 1.3369096674290253, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51361 + }, + { + "epoch": 0.51362, + "grad_norm": 1.6625520522437778, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51362 + }, + { + "epoch": 0.51363, + "grad_norm": 1.7670067369544966, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 51363 + }, + { + "epoch": 0.51364, + "grad_norm": 1.2968028259535542, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 51364 + }, + { + "epoch": 0.51365, + "grad_norm": 1.6785532205092757, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51365 + }, + { + "epoch": 0.51366, + "grad_norm": 1.617116571021586, + "learning_rate": 0.003, + "loss": 4.07, + "step": 51366 + }, + { + "epoch": 0.51367, + "grad_norm": 1.5955733288060694, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 51367 + }, + { + "epoch": 0.51368, + "grad_norm": 1.4682527642605034, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 51368 + }, + { + "epoch": 0.51369, + "grad_norm": 1.5179239124518673, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 51369 + }, + { + "epoch": 0.5137, + "grad_norm": 1.4443614433498417, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51370 + }, + { + "epoch": 0.51371, + "grad_norm": 1.4714786048375397, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51371 + }, + { + "epoch": 0.51372, + "grad_norm": 1.843233312431351, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 51372 + }, + { + "epoch": 0.51373, + "grad_norm": 1.4212425580573926, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 51373 + }, + { + "epoch": 0.51374, + "grad_norm": 1.607895048352583, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51374 + }, + { + "epoch": 0.51375, + "grad_norm": 1.5568129304434684, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 51375 + }, + { + "epoch": 0.51376, + "grad_norm": 1.486888390334085, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51376 + }, + { + "epoch": 0.51377, + "grad_norm": 1.339392556749323, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 51377 + }, + { + "epoch": 0.51378, + "grad_norm": 1.5989057410893126, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 51378 + }, + { + "epoch": 0.51379, + "grad_norm": 1.2929695864884345, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 51379 + }, + { + "epoch": 0.5138, + "grad_norm": 1.8308076408803262, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 51380 + }, + { + "epoch": 0.51381, + "grad_norm": 1.434963337190914, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 51381 + }, + { + "epoch": 0.51382, + "grad_norm": 1.6910261894720375, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51382 + }, + { + "epoch": 0.51383, + "grad_norm": 1.600700686586583, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 51383 + }, + { + "epoch": 0.51384, + "grad_norm": 1.4142416862249727, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 51384 + }, + { + "epoch": 0.51385, + "grad_norm": 1.7258715353020493, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 51385 + }, + { + "epoch": 0.51386, + "grad_norm": 1.4933176065393114, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 51386 + }, + { + "epoch": 0.51387, + "grad_norm": 1.6962757554605599, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51387 + }, + { + "epoch": 0.51388, + "grad_norm": 1.7019875842794228, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 51388 + }, + { + "epoch": 0.51389, + "grad_norm": 1.7064057729377207, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51389 + }, + { + "epoch": 0.5139, + "grad_norm": 1.5291282402503064, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 51390 + }, + { + "epoch": 0.51391, + "grad_norm": 1.8197652988877748, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 51391 + }, + { + "epoch": 0.51392, + "grad_norm": 1.4726201436565864, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51392 + }, + { + "epoch": 0.51393, + "grad_norm": 1.4780698215090862, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 51393 + }, + { + "epoch": 0.51394, + "grad_norm": 1.4633875493327453, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 51394 + }, + { + "epoch": 0.51395, + "grad_norm": 1.5347442247072616, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51395 + }, + { + "epoch": 0.51396, + "grad_norm": 1.6087520446551808, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51396 + }, + { + "epoch": 0.51397, + "grad_norm": 1.7069596421612068, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 51397 + }, + { + "epoch": 0.51398, + "grad_norm": 1.4428464345007486, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51398 + }, + { + "epoch": 0.51399, + "grad_norm": 1.5975438753551667, + "learning_rate": 0.003, + "loss": 4.016, + "step": 51399 + }, + { + "epoch": 0.514, + "grad_norm": 1.7418565139652515, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 51400 + }, + { + "epoch": 0.51401, + "grad_norm": 1.4308916670822605, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 51401 + }, + { + "epoch": 0.51402, + "grad_norm": 1.4670665136239052, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51402 + }, + { + "epoch": 0.51403, + "grad_norm": 1.7393715138866985, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 51403 + }, + { + "epoch": 0.51404, + "grad_norm": 1.3783167835470786, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 51404 + }, + { + "epoch": 0.51405, + "grad_norm": 1.75464162190282, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 51405 + }, + { + "epoch": 0.51406, + "grad_norm": 1.3613181114682096, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 51406 + }, + { + "epoch": 0.51407, + "grad_norm": 1.6628963093923517, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 51407 + }, + { + "epoch": 0.51408, + "grad_norm": 1.3858561719211824, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 51408 + }, + { + "epoch": 0.51409, + "grad_norm": 1.669347025735369, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 51409 + }, + { + "epoch": 0.5141, + "grad_norm": 1.4786615334369058, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 51410 + }, + { + "epoch": 0.51411, + "grad_norm": 1.8016416816107543, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51411 + }, + { + "epoch": 0.51412, + "grad_norm": 1.360798627816167, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 51412 + }, + { + "epoch": 0.51413, + "grad_norm": 1.7003765679807763, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 51413 + }, + { + "epoch": 0.51414, + "grad_norm": 1.6382755275282461, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 51414 + }, + { + "epoch": 0.51415, + "grad_norm": 1.4772297071479426, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51415 + }, + { + "epoch": 0.51416, + "grad_norm": 1.4989143252448405, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51416 + }, + { + "epoch": 0.51417, + "grad_norm": 1.6456644330077015, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 51417 + }, + { + "epoch": 0.51418, + "grad_norm": 1.4575515395347196, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 51418 + }, + { + "epoch": 0.51419, + "grad_norm": 1.758398022360136, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 51419 + }, + { + "epoch": 0.5142, + "grad_norm": 1.3990824556516361, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 51420 + }, + { + "epoch": 0.51421, + "grad_norm": 1.7154808079330635, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 51421 + }, + { + "epoch": 0.51422, + "grad_norm": 1.8055588437166477, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 51422 + }, + { + "epoch": 0.51423, + "grad_norm": 1.4436050058582357, + "learning_rate": 0.003, + "loss": 4.015, + "step": 51423 + }, + { + "epoch": 0.51424, + "grad_norm": 1.8028144210495425, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 51424 + }, + { + "epoch": 0.51425, + "grad_norm": 1.5011676614744804, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 51425 + }, + { + "epoch": 0.51426, + "grad_norm": 1.2755062396032844, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51426 + }, + { + "epoch": 0.51427, + "grad_norm": 1.5282539769221575, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51427 + }, + { + "epoch": 0.51428, + "grad_norm": 1.3787757951196251, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51428 + }, + { + "epoch": 0.51429, + "grad_norm": 1.8413040528330544, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 51429 + }, + { + "epoch": 0.5143, + "grad_norm": 1.53111471626421, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51430 + }, + { + "epoch": 0.51431, + "grad_norm": 1.6988104027440651, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 51431 + }, + { + "epoch": 0.51432, + "grad_norm": 1.4556854396755277, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 51432 + }, + { + "epoch": 0.51433, + "grad_norm": 1.5769930612336474, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 51433 + }, + { + "epoch": 0.51434, + "grad_norm": 1.7563610535457357, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51434 + }, + { + "epoch": 0.51435, + "grad_norm": 1.470073776525021, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 51435 + }, + { + "epoch": 0.51436, + "grad_norm": 1.7316331209700828, + "learning_rate": 0.003, + "loss": 4.013, + "step": 51436 + }, + { + "epoch": 0.51437, + "grad_norm": 1.6172732293253227, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51437 + }, + { + "epoch": 0.51438, + "grad_norm": 1.5132110321629364, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51438 + }, + { + "epoch": 0.51439, + "grad_norm": 1.308357208832105, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 51439 + }, + { + "epoch": 0.5144, + "grad_norm": 1.8640827029715479, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 51440 + }, + { + "epoch": 0.51441, + "grad_norm": 1.3460674949350777, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 51441 + }, + { + "epoch": 0.51442, + "grad_norm": 1.7651497830464853, + "learning_rate": 0.003, + "loss": 4.034, + "step": 51442 + }, + { + "epoch": 0.51443, + "grad_norm": 1.3571882925852936, + "learning_rate": 0.003, + "loss": 4.052, + "step": 51443 + }, + { + "epoch": 0.51444, + "grad_norm": 1.6016857189815807, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 51444 + }, + { + "epoch": 0.51445, + "grad_norm": 1.26635990229387, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 51445 + }, + { + "epoch": 0.51446, + "grad_norm": 1.541889001148449, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 51446 + }, + { + "epoch": 0.51447, + "grad_norm": 1.6313693900029196, + "learning_rate": 0.003, + "loss": 4.063, + "step": 51447 + }, + { + "epoch": 0.51448, + "grad_norm": 1.4585065798803898, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51448 + }, + { + "epoch": 0.51449, + "grad_norm": 1.5470321837741532, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 51449 + }, + { + "epoch": 0.5145, + "grad_norm": 1.459518122719423, + "learning_rate": 0.003, + "loss": 4.022, + "step": 51450 + }, + { + "epoch": 0.51451, + "grad_norm": 1.535205923578708, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 51451 + }, + { + "epoch": 0.51452, + "grad_norm": 1.5813118197522262, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51452 + }, + { + "epoch": 0.51453, + "grad_norm": 1.5884833652448704, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 51453 + }, + { + "epoch": 0.51454, + "grad_norm": 1.8115458455809952, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 51454 + }, + { + "epoch": 0.51455, + "grad_norm": 1.4239031713307928, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 51455 + }, + { + "epoch": 0.51456, + "grad_norm": 2.131867649558439, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51456 + }, + { + "epoch": 0.51457, + "grad_norm": 1.6522931743997227, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 51457 + }, + { + "epoch": 0.51458, + "grad_norm": 1.6849525787776702, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51458 + }, + { + "epoch": 0.51459, + "grad_norm": 1.4021709562395723, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 51459 + }, + { + "epoch": 0.5146, + "grad_norm": 1.5973447602522994, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 51460 + }, + { + "epoch": 0.51461, + "grad_norm": 1.6088084672430427, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 51461 + }, + { + "epoch": 0.51462, + "grad_norm": 1.6808161184070416, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 51462 + }, + { + "epoch": 0.51463, + "grad_norm": 1.5317692318068954, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 51463 + }, + { + "epoch": 0.51464, + "grad_norm": 1.6184125901245685, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51464 + }, + { + "epoch": 0.51465, + "grad_norm": 1.4590039385180844, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 51465 + }, + { + "epoch": 0.51466, + "grad_norm": 1.7777645463099332, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 51466 + }, + { + "epoch": 0.51467, + "grad_norm": 1.4062818486889737, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 51467 + }, + { + "epoch": 0.51468, + "grad_norm": 1.8212454435667647, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 51468 + }, + { + "epoch": 0.51469, + "grad_norm": 1.626395072826754, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 51469 + }, + { + "epoch": 0.5147, + "grad_norm": 1.71234539567337, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 51470 + }, + { + "epoch": 0.51471, + "grad_norm": 1.8674409010704456, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 51471 + }, + { + "epoch": 0.51472, + "grad_norm": 1.4379615350082484, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 51472 + }, + { + "epoch": 0.51473, + "grad_norm": 1.7113600871843155, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 51473 + }, + { + "epoch": 0.51474, + "grad_norm": 1.7575805996493932, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 51474 + }, + { + "epoch": 0.51475, + "grad_norm": 1.1728753530763674, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 51475 + }, + { + "epoch": 0.51476, + "grad_norm": 1.7563433536759636, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 51476 + }, + { + "epoch": 0.51477, + "grad_norm": 1.22863315839338, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 51477 + }, + { + "epoch": 0.51478, + "grad_norm": 1.8731860704528795, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 51478 + }, + { + "epoch": 0.51479, + "grad_norm": 1.3078072301069348, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 51479 + }, + { + "epoch": 0.5148, + "grad_norm": 1.6320376970786221, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 51480 + }, + { + "epoch": 0.51481, + "grad_norm": 1.7035372053984719, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 51481 + }, + { + "epoch": 0.51482, + "grad_norm": 1.7031270886263363, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 51482 + }, + { + "epoch": 0.51483, + "grad_norm": 1.4397570980224272, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51483 + }, + { + "epoch": 0.51484, + "grad_norm": 1.7400442201473518, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51484 + }, + { + "epoch": 0.51485, + "grad_norm": 1.4828112155503674, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 51485 + }, + { + "epoch": 0.51486, + "grad_norm": 1.4633966037335897, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 51486 + }, + { + "epoch": 0.51487, + "grad_norm": 1.406461402288847, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 51487 + }, + { + "epoch": 0.51488, + "grad_norm": 1.6907671583085802, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 51488 + }, + { + "epoch": 0.51489, + "grad_norm": 1.0900479753923273, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 51489 + }, + { + "epoch": 0.5149, + "grad_norm": 1.6154994068501987, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51490 + }, + { + "epoch": 0.51491, + "grad_norm": 1.3824704802573948, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 51491 + }, + { + "epoch": 0.51492, + "grad_norm": 1.7987897083574549, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 51492 + }, + { + "epoch": 0.51493, + "grad_norm": 1.4517368986156614, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51493 + }, + { + "epoch": 0.51494, + "grad_norm": 1.867097957869351, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 51494 + }, + { + "epoch": 0.51495, + "grad_norm": 1.9364211856527294, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 51495 + }, + { + "epoch": 0.51496, + "grad_norm": 1.11675304199185, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 51496 + }, + { + "epoch": 0.51497, + "grad_norm": 1.8851261255699256, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51497 + }, + { + "epoch": 0.51498, + "grad_norm": 1.6723513611126006, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 51498 + }, + { + "epoch": 0.51499, + "grad_norm": 1.5878459314081093, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 51499 + }, + { + "epoch": 0.515, + "grad_norm": 1.5975051907708362, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51500 + }, + { + "epoch": 0.51501, + "grad_norm": 1.7420014599084184, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 51501 + }, + { + "epoch": 0.51502, + "grad_norm": 1.1960434157681983, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51502 + }, + { + "epoch": 0.51503, + "grad_norm": 1.5353205905176945, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 51503 + }, + { + "epoch": 0.51504, + "grad_norm": 1.4243860757000402, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 51504 + }, + { + "epoch": 0.51505, + "grad_norm": 1.536366698411091, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 51505 + }, + { + "epoch": 0.51506, + "grad_norm": 1.5336605971974462, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 51506 + }, + { + "epoch": 0.51507, + "grad_norm": 1.6173432215369214, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51507 + }, + { + "epoch": 0.51508, + "grad_norm": 1.469413297461884, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 51508 + }, + { + "epoch": 0.51509, + "grad_norm": 1.7672911047286468, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 51509 + }, + { + "epoch": 0.5151, + "grad_norm": 1.4846359383570338, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51510 + }, + { + "epoch": 0.51511, + "grad_norm": 1.6886861608185526, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 51511 + }, + { + "epoch": 0.51512, + "grad_norm": 1.4876058768987377, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 51512 + }, + { + "epoch": 0.51513, + "grad_norm": 1.7238113049290853, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 51513 + }, + { + "epoch": 0.51514, + "grad_norm": 2.0232598507436914, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 51514 + }, + { + "epoch": 0.51515, + "grad_norm": 1.328675853168239, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51515 + }, + { + "epoch": 0.51516, + "grad_norm": 2.0112444779121748, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 51516 + }, + { + "epoch": 0.51517, + "grad_norm": 1.389954524102205, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 51517 + }, + { + "epoch": 0.51518, + "grad_norm": 1.7259393652679436, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 51518 + }, + { + "epoch": 0.51519, + "grad_norm": 1.1731593698959142, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 51519 + }, + { + "epoch": 0.5152, + "grad_norm": 1.7982475159222278, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 51520 + }, + { + "epoch": 0.51521, + "grad_norm": 1.6120157874774779, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 51521 + }, + { + "epoch": 0.51522, + "grad_norm": 1.7819431042937242, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 51522 + }, + { + "epoch": 0.51523, + "grad_norm": 1.473665305540452, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 51523 + }, + { + "epoch": 0.51524, + "grad_norm": 1.376800702493582, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51524 + }, + { + "epoch": 0.51525, + "grad_norm": 1.6028170999983147, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51525 + }, + { + "epoch": 0.51526, + "grad_norm": 1.6967655593886855, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 51526 + }, + { + "epoch": 0.51527, + "grad_norm": 1.3970256821044809, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 51527 + }, + { + "epoch": 0.51528, + "grad_norm": 1.5672246331049455, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 51528 + }, + { + "epoch": 0.51529, + "grad_norm": 1.4116117476511074, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 51529 + }, + { + "epoch": 0.5153, + "grad_norm": 1.4371364633073695, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 51530 + }, + { + "epoch": 0.51531, + "grad_norm": 1.5404448153749664, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51531 + }, + { + "epoch": 0.51532, + "grad_norm": 1.7781419940070808, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 51532 + }, + { + "epoch": 0.51533, + "grad_norm": 1.1910977549157804, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 51533 + }, + { + "epoch": 0.51534, + "grad_norm": 2.072908096788469, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 51534 + }, + { + "epoch": 0.51535, + "grad_norm": 1.1840787050055575, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 51535 + }, + { + "epoch": 0.51536, + "grad_norm": 2.018168196622249, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 51536 + }, + { + "epoch": 0.51537, + "grad_norm": 1.751415183171546, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 51537 + }, + { + "epoch": 0.51538, + "grad_norm": 1.5177569353704872, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 51538 + }, + { + "epoch": 0.51539, + "grad_norm": 1.3076109998421321, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 51539 + }, + { + "epoch": 0.5154, + "grad_norm": 1.7134119659425582, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 51540 + }, + { + "epoch": 0.51541, + "grad_norm": 1.760470127116552, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 51541 + }, + { + "epoch": 0.51542, + "grad_norm": 1.5662163873481025, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 51542 + }, + { + "epoch": 0.51543, + "grad_norm": 1.5688003825045105, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 51543 + }, + { + "epoch": 0.51544, + "grad_norm": 1.5797980803945928, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 51544 + }, + { + "epoch": 0.51545, + "grad_norm": 1.4665813336106872, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 51545 + }, + { + "epoch": 0.51546, + "grad_norm": 1.5009275882938633, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 51546 + }, + { + "epoch": 0.51547, + "grad_norm": 1.616809726120181, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 51547 + }, + { + "epoch": 0.51548, + "grad_norm": 1.5880855048080467, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 51548 + }, + { + "epoch": 0.51549, + "grad_norm": 1.98213578372564, + "learning_rate": 0.003, + "loss": 4.04, + "step": 51549 + }, + { + "epoch": 0.5155, + "grad_norm": 1.2705483911370061, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 51550 + }, + { + "epoch": 0.51551, + "grad_norm": 1.5879699011500559, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 51551 + }, + { + "epoch": 0.51552, + "grad_norm": 1.6312540403331768, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 51552 + }, + { + "epoch": 0.51553, + "grad_norm": 1.3743701320209891, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 51553 + }, + { + "epoch": 0.51554, + "grad_norm": 1.5890663808718017, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51554 + }, + { + "epoch": 0.51555, + "grad_norm": 1.4912138466223515, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51555 + }, + { + "epoch": 0.51556, + "grad_norm": 1.7200868094175272, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 51556 + }, + { + "epoch": 0.51557, + "grad_norm": 1.6154843693339882, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 51557 + }, + { + "epoch": 0.51558, + "grad_norm": 1.607486721881634, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 51558 + }, + { + "epoch": 0.51559, + "grad_norm": 1.2027998453385673, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 51559 + }, + { + "epoch": 0.5156, + "grad_norm": 1.6410656759473186, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51560 + }, + { + "epoch": 0.51561, + "grad_norm": 1.525424445362252, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 51561 + }, + { + "epoch": 0.51562, + "grad_norm": 1.5413103257619523, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 51562 + }, + { + "epoch": 0.51563, + "grad_norm": 1.4479599457440389, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 51563 + }, + { + "epoch": 0.51564, + "grad_norm": 1.6272998019352385, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 51564 + }, + { + "epoch": 0.51565, + "grad_norm": 1.4196534334599586, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 51565 + }, + { + "epoch": 0.51566, + "grad_norm": 1.7247577115050763, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51566 + }, + { + "epoch": 0.51567, + "grad_norm": 1.5529034657777876, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 51567 + }, + { + "epoch": 0.51568, + "grad_norm": 1.5454062098767691, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 51568 + }, + { + "epoch": 0.51569, + "grad_norm": 1.265599471471066, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 51569 + }, + { + "epoch": 0.5157, + "grad_norm": 1.9058340193060586, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 51570 + }, + { + "epoch": 0.51571, + "grad_norm": 1.2515245516366358, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 51571 + }, + { + "epoch": 0.51572, + "grad_norm": 1.5431522421412334, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 51572 + }, + { + "epoch": 0.51573, + "grad_norm": 1.5384412210876348, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 51573 + }, + { + "epoch": 0.51574, + "grad_norm": 1.7514061814170574, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51574 + }, + { + "epoch": 0.51575, + "grad_norm": 1.8431501171317675, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 51575 + }, + { + "epoch": 0.51576, + "grad_norm": 1.2937934040216976, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 51576 + }, + { + "epoch": 0.51577, + "grad_norm": 1.6378755761430055, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51577 + }, + { + "epoch": 0.51578, + "grad_norm": 1.7429472892090612, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 51578 + }, + { + "epoch": 0.51579, + "grad_norm": 1.4672645190550235, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 51579 + }, + { + "epoch": 0.5158, + "grad_norm": 1.556276925208221, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 51580 + }, + { + "epoch": 0.51581, + "grad_norm": 1.953266326101508, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51581 + }, + { + "epoch": 0.51582, + "grad_norm": 1.2208401785352547, + "learning_rate": 0.003, + "loss": 4.046, + "step": 51582 + }, + { + "epoch": 0.51583, + "grad_norm": 1.6966908824199498, + "learning_rate": 0.003, + "loss": 4.075, + "step": 51583 + }, + { + "epoch": 0.51584, + "grad_norm": 1.5865749294673492, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 51584 + }, + { + "epoch": 0.51585, + "grad_norm": 1.6138873716507478, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 51585 + }, + { + "epoch": 0.51586, + "grad_norm": 1.383764243078382, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 51586 + }, + { + "epoch": 0.51587, + "grad_norm": 1.544969460947507, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51587 + }, + { + "epoch": 0.51588, + "grad_norm": 1.7167713312654251, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 51588 + }, + { + "epoch": 0.51589, + "grad_norm": 1.4236383002411512, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 51589 + }, + { + "epoch": 0.5159, + "grad_norm": 1.7741571669916285, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51590 + }, + { + "epoch": 0.51591, + "grad_norm": 1.2921388783941832, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 51591 + }, + { + "epoch": 0.51592, + "grad_norm": 1.5990544691756685, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 51592 + }, + { + "epoch": 0.51593, + "grad_norm": 1.5208559486444908, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 51593 + }, + { + "epoch": 0.51594, + "grad_norm": 1.638470222991462, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 51594 + }, + { + "epoch": 0.51595, + "grad_norm": 1.5241196184646872, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 51595 + }, + { + "epoch": 0.51596, + "grad_norm": 1.8914601485784297, + "learning_rate": 0.003, + "loss": 4.022, + "step": 51596 + }, + { + "epoch": 0.51597, + "grad_norm": 1.4070202865701682, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 51597 + }, + { + "epoch": 0.51598, + "grad_norm": 1.6745317522039562, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 51598 + }, + { + "epoch": 0.51599, + "grad_norm": 1.570307077996802, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51599 + }, + { + "epoch": 0.516, + "grad_norm": 1.5756573358115802, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 51600 + }, + { + "epoch": 0.51601, + "grad_norm": 1.4558495455321423, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 51601 + }, + { + "epoch": 0.51602, + "grad_norm": 1.335159254101434, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 51602 + }, + { + "epoch": 0.51603, + "grad_norm": 1.5906095976944072, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 51603 + }, + { + "epoch": 0.51604, + "grad_norm": 1.9956390999085467, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51604 + }, + { + "epoch": 0.51605, + "grad_norm": 1.7137161693250014, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 51605 + }, + { + "epoch": 0.51606, + "grad_norm": 1.750733763484742, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 51606 + }, + { + "epoch": 0.51607, + "grad_norm": 1.3853773665454765, + "learning_rate": 0.003, + "loss": 4.066, + "step": 51607 + }, + { + "epoch": 0.51608, + "grad_norm": 1.6432616803233615, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 51608 + }, + { + "epoch": 0.51609, + "grad_norm": 1.4154278605019126, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 51609 + }, + { + "epoch": 0.5161, + "grad_norm": 1.7194906667964291, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51610 + }, + { + "epoch": 0.51611, + "grad_norm": 1.4740873618493955, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 51611 + }, + { + "epoch": 0.51612, + "grad_norm": 1.8758778235619595, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 51612 + }, + { + "epoch": 0.51613, + "grad_norm": 1.2062076221600417, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 51613 + }, + { + "epoch": 0.51614, + "grad_norm": 1.6763346283941623, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 51614 + }, + { + "epoch": 0.51615, + "grad_norm": 1.18275952833538, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 51615 + }, + { + "epoch": 0.51616, + "grad_norm": 1.8773164921983159, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 51616 + }, + { + "epoch": 0.51617, + "grad_norm": 1.1911230677738383, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 51617 + }, + { + "epoch": 0.51618, + "grad_norm": 1.8363567716331661, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 51618 + }, + { + "epoch": 0.51619, + "grad_norm": 1.5271081009359193, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 51619 + }, + { + "epoch": 0.5162, + "grad_norm": 2.413463033204334, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 51620 + }, + { + "epoch": 0.51621, + "grad_norm": 1.3359540252951074, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 51621 + }, + { + "epoch": 0.51622, + "grad_norm": 1.615952477096724, + "learning_rate": 0.003, + "loss": 4.047, + "step": 51622 + }, + { + "epoch": 0.51623, + "grad_norm": 1.4419774804756056, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 51623 + }, + { + "epoch": 0.51624, + "grad_norm": 1.6854857997691284, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 51624 + }, + { + "epoch": 0.51625, + "grad_norm": 1.746895514562888, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 51625 + }, + { + "epoch": 0.51626, + "grad_norm": 1.5616558272808538, + "learning_rate": 0.003, + "loss": 4.019, + "step": 51626 + }, + { + "epoch": 0.51627, + "grad_norm": 1.5243845741183697, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 51627 + }, + { + "epoch": 0.51628, + "grad_norm": 1.5282506352736278, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 51628 + }, + { + "epoch": 0.51629, + "grad_norm": 1.4580407037012804, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 51629 + }, + { + "epoch": 0.5163, + "grad_norm": 1.3224878090987218, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 51630 + }, + { + "epoch": 0.51631, + "grad_norm": 1.7243174498876421, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 51631 + }, + { + "epoch": 0.51632, + "grad_norm": 1.484810339204094, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 51632 + }, + { + "epoch": 0.51633, + "grad_norm": 1.458638813092642, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 51633 + }, + { + "epoch": 0.51634, + "grad_norm": 1.8887661658443444, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 51634 + }, + { + "epoch": 0.51635, + "grad_norm": 1.27547885685843, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 51635 + }, + { + "epoch": 0.51636, + "grad_norm": 1.7852211190816545, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 51636 + }, + { + "epoch": 0.51637, + "grad_norm": 1.527315487443218, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 51637 + }, + { + "epoch": 0.51638, + "grad_norm": 1.8422742225896633, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 51638 + }, + { + "epoch": 0.51639, + "grad_norm": 1.358707074916634, + "learning_rate": 0.003, + "loss": 4.03, + "step": 51639 + }, + { + "epoch": 0.5164, + "grad_norm": 1.6152099175855232, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 51640 + }, + { + "epoch": 0.51641, + "grad_norm": 1.483884635685569, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 51641 + }, + { + "epoch": 0.51642, + "grad_norm": 1.4478809551553398, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 51642 + }, + { + "epoch": 0.51643, + "grad_norm": 1.7051113982410242, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 51643 + }, + { + "epoch": 0.51644, + "grad_norm": 1.3356362933748642, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 51644 + }, + { + "epoch": 0.51645, + "grad_norm": 1.7218322860060251, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 51645 + }, + { + "epoch": 0.51646, + "grad_norm": 1.4876148090483523, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 51646 + }, + { + "epoch": 0.51647, + "grad_norm": 1.856540724003774, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 51647 + }, + { + "epoch": 0.51648, + "grad_norm": 1.5139024274960262, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 51648 + }, + { + "epoch": 0.51649, + "grad_norm": 1.670727820831582, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 51649 + }, + { + "epoch": 0.5165, + "grad_norm": 1.5553134229871484, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 51650 + }, + { + "epoch": 0.51651, + "grad_norm": 1.626314194087974, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 51651 + }, + { + "epoch": 0.51652, + "grad_norm": 1.418150250850149, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 51652 + }, + { + "epoch": 0.51653, + "grad_norm": 1.6920546183725722, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51653 + }, + { + "epoch": 0.51654, + "grad_norm": 1.3898700457733428, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 51654 + }, + { + "epoch": 0.51655, + "grad_norm": 1.6454859448385208, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51655 + }, + { + "epoch": 0.51656, + "grad_norm": 1.4643540980101917, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 51656 + }, + { + "epoch": 0.51657, + "grad_norm": 1.6983484498708792, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 51657 + }, + { + "epoch": 0.51658, + "grad_norm": 1.5831225472658745, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 51658 + }, + { + "epoch": 0.51659, + "grad_norm": 1.4115122117849859, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 51659 + }, + { + "epoch": 0.5166, + "grad_norm": 1.504776540345942, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 51660 + }, + { + "epoch": 0.51661, + "grad_norm": 1.499982767118776, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 51661 + }, + { + "epoch": 0.51662, + "grad_norm": 1.5550131836305587, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 51662 + }, + { + "epoch": 0.51663, + "grad_norm": 1.4278867589589113, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51663 + }, + { + "epoch": 0.51664, + "grad_norm": 1.7220761303603975, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51664 + }, + { + "epoch": 0.51665, + "grad_norm": 1.4660966908445219, + "learning_rate": 0.003, + "loss": 4.012, + "step": 51665 + }, + { + "epoch": 0.51666, + "grad_norm": 1.5407445961295605, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 51666 + }, + { + "epoch": 0.51667, + "grad_norm": 1.6431150821780336, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 51667 + }, + { + "epoch": 0.51668, + "grad_norm": 1.5925052190357813, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 51668 + }, + { + "epoch": 0.51669, + "grad_norm": 1.6933530156780414, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 51669 + }, + { + "epoch": 0.5167, + "grad_norm": 1.3359747343020587, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 51670 + }, + { + "epoch": 0.51671, + "grad_norm": 1.8002015235367064, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51671 + }, + { + "epoch": 0.51672, + "grad_norm": 1.1971487516355082, + "learning_rate": 0.003, + "loss": 4.012, + "step": 51672 + }, + { + "epoch": 0.51673, + "grad_norm": 1.8968278514758123, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 51673 + }, + { + "epoch": 0.51674, + "grad_norm": 1.2217475488163947, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 51674 + }, + { + "epoch": 0.51675, + "grad_norm": 1.5500648416121696, + "learning_rate": 0.003, + "loss": 3.994, + "step": 51675 + }, + { + "epoch": 0.51676, + "grad_norm": 1.451645437360482, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 51676 + }, + { + "epoch": 0.51677, + "grad_norm": 1.4934234457250954, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51677 + }, + { + "epoch": 0.51678, + "grad_norm": 1.5237127292866652, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 51678 + }, + { + "epoch": 0.51679, + "grad_norm": 1.8393983973962904, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 51679 + }, + { + "epoch": 0.5168, + "grad_norm": 2.0691022454308134, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 51680 + }, + { + "epoch": 0.51681, + "grad_norm": 1.2666501881504793, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 51681 + }, + { + "epoch": 0.51682, + "grad_norm": 1.6019021847592598, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 51682 + }, + { + "epoch": 0.51683, + "grad_norm": 1.5731222403745564, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51683 + }, + { + "epoch": 0.51684, + "grad_norm": 1.7013437453926579, + "learning_rate": 0.003, + "loss": 4.027, + "step": 51684 + }, + { + "epoch": 0.51685, + "grad_norm": 1.532883725052802, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51685 + }, + { + "epoch": 0.51686, + "grad_norm": 1.6810667354436417, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 51686 + }, + { + "epoch": 0.51687, + "grad_norm": 1.2628185958020177, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 51687 + }, + { + "epoch": 0.51688, + "grad_norm": 1.5835466197504464, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 51688 + }, + { + "epoch": 0.51689, + "grad_norm": 1.619743404404982, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 51689 + }, + { + "epoch": 0.5169, + "grad_norm": 1.3713632088755368, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 51690 + }, + { + "epoch": 0.51691, + "grad_norm": 1.6939221072933972, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 51691 + }, + { + "epoch": 0.51692, + "grad_norm": 1.403376653222366, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 51692 + }, + { + "epoch": 0.51693, + "grad_norm": 1.8668617339664535, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 51693 + }, + { + "epoch": 0.51694, + "grad_norm": 1.481383840119227, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 51694 + }, + { + "epoch": 0.51695, + "grad_norm": 1.861609240751685, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51695 + }, + { + "epoch": 0.51696, + "grad_norm": 1.4525625029836773, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 51696 + }, + { + "epoch": 0.51697, + "grad_norm": 1.8640966175963503, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 51697 + }, + { + "epoch": 0.51698, + "grad_norm": 1.6576700084222042, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 51698 + }, + { + "epoch": 0.51699, + "grad_norm": 1.5826871743138335, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 51699 + }, + { + "epoch": 0.517, + "grad_norm": 1.556215960882519, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 51700 + }, + { + "epoch": 0.51701, + "grad_norm": 1.5587846222438921, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 51701 + }, + { + "epoch": 0.51702, + "grad_norm": 1.5133237569780102, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 51702 + }, + { + "epoch": 0.51703, + "grad_norm": 1.678758132749893, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 51703 + }, + { + "epoch": 0.51704, + "grad_norm": 1.8019300857499434, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 51704 + }, + { + "epoch": 0.51705, + "grad_norm": 1.4306248880040466, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51705 + }, + { + "epoch": 0.51706, + "grad_norm": 1.366968164424459, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51706 + }, + { + "epoch": 0.51707, + "grad_norm": 1.6274000161680255, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51707 + }, + { + "epoch": 0.51708, + "grad_norm": 1.4895522870869033, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51708 + }, + { + "epoch": 0.51709, + "grad_norm": 1.3812809827042547, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 51709 + }, + { + "epoch": 0.5171, + "grad_norm": 1.8516143603829855, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 51710 + }, + { + "epoch": 0.51711, + "grad_norm": 1.3065835299787578, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 51711 + }, + { + "epoch": 0.51712, + "grad_norm": 1.6642368078999599, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 51712 + }, + { + "epoch": 0.51713, + "grad_norm": 1.3717494364457024, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 51713 + }, + { + "epoch": 0.51714, + "grad_norm": 1.4476812976255358, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51714 + }, + { + "epoch": 0.51715, + "grad_norm": 1.6285480691387033, + "learning_rate": 0.003, + "loss": 4.014, + "step": 51715 + }, + { + "epoch": 0.51716, + "grad_norm": 1.5509356746055014, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 51716 + }, + { + "epoch": 0.51717, + "grad_norm": 1.8314316303482079, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 51717 + }, + { + "epoch": 0.51718, + "grad_norm": 1.6215806689166126, + "learning_rate": 0.003, + "loss": 4.05, + "step": 51718 + }, + { + "epoch": 0.51719, + "grad_norm": 1.5446937164725254, + "learning_rate": 0.003, + "loss": 4.035, + "step": 51719 + }, + { + "epoch": 0.5172, + "grad_norm": 1.5746722237197086, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 51720 + }, + { + "epoch": 0.51721, + "grad_norm": 1.3773069578016695, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 51721 + }, + { + "epoch": 0.51722, + "grad_norm": 1.3918994247159775, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 51722 + }, + { + "epoch": 0.51723, + "grad_norm": 1.6703359509429092, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 51723 + }, + { + "epoch": 0.51724, + "grad_norm": 2.01380605033872, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51724 + }, + { + "epoch": 0.51725, + "grad_norm": 1.2371539984774975, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 51725 + }, + { + "epoch": 0.51726, + "grad_norm": 2.004248278303805, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51726 + }, + { + "epoch": 0.51727, + "grad_norm": 1.1884018589767635, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51727 + }, + { + "epoch": 0.51728, + "grad_norm": 1.9304720555008303, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 51728 + }, + { + "epoch": 0.51729, + "grad_norm": 1.5197738528821938, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 51729 + }, + { + "epoch": 0.5173, + "grad_norm": 1.2327727114498013, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 51730 + }, + { + "epoch": 0.51731, + "grad_norm": 2.055963872799128, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 51731 + }, + { + "epoch": 0.51732, + "grad_norm": 1.3242370465545354, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 51732 + }, + { + "epoch": 0.51733, + "grad_norm": 1.7845447014227915, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 51733 + }, + { + "epoch": 0.51734, + "grad_norm": 1.685840584306787, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 51734 + }, + { + "epoch": 0.51735, + "grad_norm": 1.5071041444516866, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 51735 + }, + { + "epoch": 0.51736, + "grad_norm": 1.6606526290892132, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 51736 + }, + { + "epoch": 0.51737, + "grad_norm": 1.388808216322583, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51737 + }, + { + "epoch": 0.51738, + "grad_norm": 1.615646441258868, + "learning_rate": 0.003, + "loss": 4.062, + "step": 51738 + }, + { + "epoch": 0.51739, + "grad_norm": 1.2734760861072765, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 51739 + }, + { + "epoch": 0.5174, + "grad_norm": 1.5795442366453099, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51740 + }, + { + "epoch": 0.51741, + "grad_norm": 2.0105045687174914, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 51741 + }, + { + "epoch": 0.51742, + "grad_norm": 1.4600894149901358, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 51742 + }, + { + "epoch": 0.51743, + "grad_norm": 1.7084312312721417, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 51743 + }, + { + "epoch": 0.51744, + "grad_norm": 1.488049636097678, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 51744 + }, + { + "epoch": 0.51745, + "grad_norm": 1.5606116270361343, + "learning_rate": 0.003, + "loss": 4.054, + "step": 51745 + }, + { + "epoch": 0.51746, + "grad_norm": 1.6378147849540554, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51746 + }, + { + "epoch": 0.51747, + "grad_norm": 1.2521703002211617, + "learning_rate": 0.003, + "loss": 4.011, + "step": 51747 + }, + { + "epoch": 0.51748, + "grad_norm": 1.3864280662928776, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 51748 + }, + { + "epoch": 0.51749, + "grad_norm": 1.458749415406428, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 51749 + }, + { + "epoch": 0.5175, + "grad_norm": 1.619812705051788, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 51750 + }, + { + "epoch": 0.51751, + "grad_norm": 1.5189536234347145, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 51751 + }, + { + "epoch": 0.51752, + "grad_norm": 1.76982746311325, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 51752 + }, + { + "epoch": 0.51753, + "grad_norm": 1.4285494935915835, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 51753 + }, + { + "epoch": 0.51754, + "grad_norm": 1.4426069337218612, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51754 + }, + { + "epoch": 0.51755, + "grad_norm": 1.4466955890609177, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 51755 + }, + { + "epoch": 0.51756, + "grad_norm": 1.9453230600851659, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 51756 + }, + { + "epoch": 0.51757, + "grad_norm": 1.254802245993286, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 51757 + }, + { + "epoch": 0.51758, + "grad_norm": 1.8910932590194875, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51758 + }, + { + "epoch": 0.51759, + "grad_norm": 1.425467478361341, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 51759 + }, + { + "epoch": 0.5176, + "grad_norm": 1.8303567894924195, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 51760 + }, + { + "epoch": 0.51761, + "grad_norm": 1.5504448414253165, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 51761 + }, + { + "epoch": 0.51762, + "grad_norm": 1.5920757682322701, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51762 + }, + { + "epoch": 0.51763, + "grad_norm": 1.4303734932828958, + "learning_rate": 0.003, + "loss": 4.032, + "step": 51763 + }, + { + "epoch": 0.51764, + "grad_norm": 1.5105353346074337, + "learning_rate": 0.003, + "loss": 4.037, + "step": 51764 + }, + { + "epoch": 0.51765, + "grad_norm": 1.3068987425495069, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 51765 + }, + { + "epoch": 0.51766, + "grad_norm": 1.8895799140759573, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 51766 + }, + { + "epoch": 0.51767, + "grad_norm": 1.2804561031785646, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 51767 + }, + { + "epoch": 0.51768, + "grad_norm": 1.6806957434750045, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 51768 + }, + { + "epoch": 0.51769, + "grad_norm": 1.406643252530562, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 51769 + }, + { + "epoch": 0.5177, + "grad_norm": 1.6167406523199364, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 51770 + }, + { + "epoch": 0.51771, + "grad_norm": 1.5936786275238464, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 51771 + }, + { + "epoch": 0.51772, + "grad_norm": 1.4017795138882252, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51772 + }, + { + "epoch": 0.51773, + "grad_norm": 1.9238684617567636, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 51773 + }, + { + "epoch": 0.51774, + "grad_norm": 1.6547730981371567, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 51774 + }, + { + "epoch": 0.51775, + "grad_norm": 1.5547690758550636, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 51775 + }, + { + "epoch": 0.51776, + "grad_norm": 1.4683665880753323, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 51776 + }, + { + "epoch": 0.51777, + "grad_norm": 1.8488763970650093, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 51777 + }, + { + "epoch": 0.51778, + "grad_norm": 1.7256524603357883, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 51778 + }, + { + "epoch": 0.51779, + "grad_norm": 1.5450715819524945, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51779 + }, + { + "epoch": 0.5178, + "grad_norm": 1.601660497110653, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 51780 + }, + { + "epoch": 0.51781, + "grad_norm": 1.49374667533609, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51781 + }, + { + "epoch": 0.51782, + "grad_norm": 1.3626574744217017, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51782 + }, + { + "epoch": 0.51783, + "grad_norm": 1.9234025815294848, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 51783 + }, + { + "epoch": 0.51784, + "grad_norm": 1.5568617195350232, + "learning_rate": 0.003, + "loss": 4.042, + "step": 51784 + }, + { + "epoch": 0.51785, + "grad_norm": 1.7210386535792486, + "learning_rate": 0.003, + "loss": 4.044, + "step": 51785 + }, + { + "epoch": 0.51786, + "grad_norm": 1.3341598086933677, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 51786 + }, + { + "epoch": 0.51787, + "grad_norm": 1.493232975444907, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 51787 + }, + { + "epoch": 0.51788, + "grad_norm": 1.737069366071011, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 51788 + }, + { + "epoch": 0.51789, + "grad_norm": 1.1520593044541543, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 51789 + }, + { + "epoch": 0.5179, + "grad_norm": 1.8609252376333294, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 51790 + }, + { + "epoch": 0.51791, + "grad_norm": 1.2614855533639175, + "learning_rate": 0.003, + "loss": 4.057, + "step": 51791 + }, + { + "epoch": 0.51792, + "grad_norm": 1.8499627850284461, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 51792 + }, + { + "epoch": 0.51793, + "grad_norm": 1.3753620409618441, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 51793 + }, + { + "epoch": 0.51794, + "grad_norm": 1.5760908886791967, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 51794 + }, + { + "epoch": 0.51795, + "grad_norm": 1.669646150019578, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 51795 + }, + { + "epoch": 0.51796, + "grad_norm": 1.4372777017844556, + "learning_rate": 0.003, + "loss": 4.021, + "step": 51796 + }, + { + "epoch": 0.51797, + "grad_norm": 1.7403292117089608, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 51797 + }, + { + "epoch": 0.51798, + "grad_norm": 1.3719796330120597, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 51798 + }, + { + "epoch": 0.51799, + "grad_norm": 1.752107975385293, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 51799 + }, + { + "epoch": 0.518, + "grad_norm": 1.5177445738839586, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 51800 + }, + { + "epoch": 0.51801, + "grad_norm": 1.5330191284644858, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 51801 + }, + { + "epoch": 0.51802, + "grad_norm": 1.706110368356187, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51802 + }, + { + "epoch": 0.51803, + "grad_norm": 1.478572248598704, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 51803 + }, + { + "epoch": 0.51804, + "grad_norm": 1.6973208779977114, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 51804 + }, + { + "epoch": 0.51805, + "grad_norm": 1.4027484947573006, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 51805 + }, + { + "epoch": 0.51806, + "grad_norm": 1.8118904315215425, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 51806 + }, + { + "epoch": 0.51807, + "grad_norm": 1.2561738663693292, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 51807 + }, + { + "epoch": 0.51808, + "grad_norm": 1.8713524475153211, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 51808 + }, + { + "epoch": 0.51809, + "grad_norm": 1.399392978564258, + "learning_rate": 0.003, + "loss": 4.057, + "step": 51809 + }, + { + "epoch": 0.5181, + "grad_norm": 1.7587592425171266, + "learning_rate": 0.003, + "loss": 4.041, + "step": 51810 + }, + { + "epoch": 0.51811, + "grad_norm": 1.2557721313879753, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 51811 + }, + { + "epoch": 0.51812, + "grad_norm": 1.427821694959845, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 51812 + }, + { + "epoch": 0.51813, + "grad_norm": 1.6989252374735235, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51813 + }, + { + "epoch": 0.51814, + "grad_norm": 1.481427317916997, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 51814 + }, + { + "epoch": 0.51815, + "grad_norm": 1.839906454322507, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 51815 + }, + { + "epoch": 0.51816, + "grad_norm": 1.354857358394804, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51816 + }, + { + "epoch": 0.51817, + "grad_norm": 1.8295001949099796, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 51817 + }, + { + "epoch": 0.51818, + "grad_norm": 1.3910484671611616, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 51818 + }, + { + "epoch": 0.51819, + "grad_norm": 1.4858797562412387, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 51819 + }, + { + "epoch": 0.5182, + "grad_norm": 1.4042109404434688, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 51820 + }, + { + "epoch": 0.51821, + "grad_norm": 1.7027224471681266, + "learning_rate": 0.003, + "loss": 4.081, + "step": 51821 + }, + { + "epoch": 0.51822, + "grad_norm": 1.7041241516783385, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 51822 + }, + { + "epoch": 0.51823, + "grad_norm": 1.3950044365686427, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 51823 + }, + { + "epoch": 0.51824, + "grad_norm": 1.724205775583452, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 51824 + }, + { + "epoch": 0.51825, + "grad_norm": 1.3897296217599275, + "learning_rate": 0.003, + "loss": 4.045, + "step": 51825 + }, + { + "epoch": 0.51826, + "grad_norm": 1.6059908406208225, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 51826 + }, + { + "epoch": 0.51827, + "grad_norm": 1.5561060426343574, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 51827 + }, + { + "epoch": 0.51828, + "grad_norm": 1.7409856714372094, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 51828 + }, + { + "epoch": 0.51829, + "grad_norm": 1.3526341791171164, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 51829 + }, + { + "epoch": 0.5183, + "grad_norm": 1.4599007505995885, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 51830 + }, + { + "epoch": 0.51831, + "grad_norm": 1.726365522588939, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 51831 + }, + { + "epoch": 0.51832, + "grad_norm": 1.4759838060826167, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 51832 + }, + { + "epoch": 0.51833, + "grad_norm": 1.901862046201052, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 51833 + }, + { + "epoch": 0.51834, + "grad_norm": 1.1332169313434302, + "learning_rate": 0.003, + "loss": 4.036, + "step": 51834 + }, + { + "epoch": 0.51835, + "grad_norm": 1.9114494530350055, + "learning_rate": 0.003, + "loss": 4.062, + "step": 51835 + }, + { + "epoch": 0.51836, + "grad_norm": 1.6157533933037493, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51836 + }, + { + "epoch": 0.51837, + "grad_norm": 1.7388206058682731, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 51837 + }, + { + "epoch": 0.51838, + "grad_norm": 1.4600984944330884, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 51838 + }, + { + "epoch": 0.51839, + "grad_norm": 1.4374978767635609, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 51839 + }, + { + "epoch": 0.5184, + "grad_norm": 1.8941621248777623, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51840 + }, + { + "epoch": 0.51841, + "grad_norm": 1.2007051952546233, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 51841 + }, + { + "epoch": 0.51842, + "grad_norm": 1.815704101641304, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 51842 + }, + { + "epoch": 0.51843, + "grad_norm": 1.1725779257138844, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51843 + }, + { + "epoch": 0.51844, + "grad_norm": 1.9000173537007905, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 51844 + }, + { + "epoch": 0.51845, + "grad_norm": 1.446854951451014, + "learning_rate": 0.003, + "loss": 4.053, + "step": 51845 + }, + { + "epoch": 0.51846, + "grad_norm": 1.356995559011848, + "learning_rate": 0.003, + "loss": 4.039, + "step": 51846 + }, + { + "epoch": 0.51847, + "grad_norm": 1.7569008670827109, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 51847 + }, + { + "epoch": 0.51848, + "grad_norm": 1.47395231211763, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 51848 + }, + { + "epoch": 0.51849, + "grad_norm": 1.6453855839984575, + "learning_rate": 0.003, + "loss": 4.052, + "step": 51849 + }, + { + "epoch": 0.5185, + "grad_norm": 1.4825519866205152, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 51850 + }, + { + "epoch": 0.51851, + "grad_norm": 1.7631669304773605, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 51851 + }, + { + "epoch": 0.51852, + "grad_norm": 1.5914121151939207, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 51852 + }, + { + "epoch": 0.51853, + "grad_norm": 1.5903185858209343, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 51853 + }, + { + "epoch": 0.51854, + "grad_norm": 1.4273936251088823, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 51854 + }, + { + "epoch": 0.51855, + "grad_norm": 1.4910404877454415, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 51855 + }, + { + "epoch": 0.51856, + "grad_norm": 1.6373188584239593, + "learning_rate": 0.003, + "loss": 4.045, + "step": 51856 + }, + { + "epoch": 0.51857, + "grad_norm": 1.5697784610050785, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 51857 + }, + { + "epoch": 0.51858, + "grad_norm": 1.8101437140565693, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 51858 + }, + { + "epoch": 0.51859, + "grad_norm": 1.3787320933601237, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51859 + }, + { + "epoch": 0.5186, + "grad_norm": 1.7024032837881413, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 51860 + }, + { + "epoch": 0.51861, + "grad_norm": 1.2999532622643983, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 51861 + }, + { + "epoch": 0.51862, + "grad_norm": 1.2167830509003779, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 51862 + }, + { + "epoch": 0.51863, + "grad_norm": 1.9724063637788776, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 51863 + }, + { + "epoch": 0.51864, + "grad_norm": 1.4917337538381825, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 51864 + }, + { + "epoch": 0.51865, + "grad_norm": 1.5590815163973462, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 51865 + }, + { + "epoch": 0.51866, + "grad_norm": 1.7636983008200182, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 51866 + }, + { + "epoch": 0.51867, + "grad_norm": 1.6331804218053756, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 51867 + }, + { + "epoch": 0.51868, + "grad_norm": 1.2728195350318143, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 51868 + }, + { + "epoch": 0.51869, + "grad_norm": 1.7916447196277574, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 51869 + }, + { + "epoch": 0.5187, + "grad_norm": 1.4812541280087312, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51870 + }, + { + "epoch": 0.51871, + "grad_norm": 1.6724361697949606, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 51871 + }, + { + "epoch": 0.51872, + "grad_norm": 1.4365179575333586, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 51872 + }, + { + "epoch": 0.51873, + "grad_norm": 1.607130189830785, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51873 + }, + { + "epoch": 0.51874, + "grad_norm": 1.4160854995123098, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 51874 + }, + { + "epoch": 0.51875, + "grad_norm": 1.6488454593448336, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 51875 + }, + { + "epoch": 0.51876, + "grad_norm": 1.3474443271254284, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 51876 + }, + { + "epoch": 0.51877, + "grad_norm": 1.4346981679964756, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 51877 + }, + { + "epoch": 0.51878, + "grad_norm": 1.8326948007499553, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 51878 + }, + { + "epoch": 0.51879, + "grad_norm": 1.4597557112544712, + "learning_rate": 0.003, + "loss": 3.996, + "step": 51879 + }, + { + "epoch": 0.5188, + "grad_norm": 1.5950168693966873, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 51880 + }, + { + "epoch": 0.51881, + "grad_norm": 1.2807119781004204, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 51881 + }, + { + "epoch": 0.51882, + "grad_norm": 1.4883258458081763, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 51882 + }, + { + "epoch": 0.51883, + "grad_norm": 1.663530090831979, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 51883 + }, + { + "epoch": 0.51884, + "grad_norm": 1.2294226791926097, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 51884 + }, + { + "epoch": 0.51885, + "grad_norm": 1.8055794750554048, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 51885 + }, + { + "epoch": 0.51886, + "grad_norm": 1.3700257810216832, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 51886 + }, + { + "epoch": 0.51887, + "grad_norm": 2.4444359897815846, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 51887 + }, + { + "epoch": 0.51888, + "grad_norm": 1.0061342710517607, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 51888 + }, + { + "epoch": 0.51889, + "grad_norm": 1.8403397602004319, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 51889 + }, + { + "epoch": 0.5189, + "grad_norm": 1.5515318417732293, + "learning_rate": 0.003, + "loss": 4.051, + "step": 51890 + }, + { + "epoch": 0.51891, + "grad_norm": 1.722635496047221, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 51891 + }, + { + "epoch": 0.51892, + "grad_norm": 1.4146939138138244, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 51892 + }, + { + "epoch": 0.51893, + "grad_norm": 1.6653212129525459, + "learning_rate": 0.003, + "loss": 4.094, + "step": 51893 + }, + { + "epoch": 0.51894, + "grad_norm": 1.469084810953321, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 51894 + }, + { + "epoch": 0.51895, + "grad_norm": 1.5928418415140875, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 51895 + }, + { + "epoch": 0.51896, + "grad_norm": 1.7490767068225355, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 51896 + }, + { + "epoch": 0.51897, + "grad_norm": 1.5710943534089419, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 51897 + }, + { + "epoch": 0.51898, + "grad_norm": 1.5283988294559105, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 51898 + }, + { + "epoch": 0.51899, + "grad_norm": 1.509229679729986, + "learning_rate": 0.003, + "loss": 4.047, + "step": 51899 + }, + { + "epoch": 0.519, + "grad_norm": 1.4825756384595705, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 51900 + }, + { + "epoch": 0.51901, + "grad_norm": 1.4991038289876069, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51901 + }, + { + "epoch": 0.51902, + "grad_norm": 1.7403853444779323, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 51902 + }, + { + "epoch": 0.51903, + "grad_norm": 1.31596528738451, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 51903 + }, + { + "epoch": 0.51904, + "grad_norm": 1.5576827285753794, + "learning_rate": 0.003, + "loss": 4.031, + "step": 51904 + }, + { + "epoch": 0.51905, + "grad_norm": 1.711582989920481, + "learning_rate": 0.003, + "loss": 4.058, + "step": 51905 + }, + { + "epoch": 0.51906, + "grad_norm": 1.3162431073847096, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 51906 + }, + { + "epoch": 0.51907, + "grad_norm": 1.5063784356417047, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 51907 + }, + { + "epoch": 0.51908, + "grad_norm": 1.534208163305875, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 51908 + }, + { + "epoch": 0.51909, + "grad_norm": 1.5502441886971838, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 51909 + }, + { + "epoch": 0.5191, + "grad_norm": 1.454185697438428, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 51910 + }, + { + "epoch": 0.51911, + "grad_norm": 1.6027083733760077, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 51911 + }, + { + "epoch": 0.51912, + "grad_norm": 1.3812703047019108, + "learning_rate": 0.003, + "loss": 4.008, + "step": 51912 + }, + { + "epoch": 0.51913, + "grad_norm": 1.6999019837128697, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 51913 + }, + { + "epoch": 0.51914, + "grad_norm": 1.5604103091851143, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 51914 + }, + { + "epoch": 0.51915, + "grad_norm": 1.813717888175761, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 51915 + }, + { + "epoch": 0.51916, + "grad_norm": 1.3722450266079596, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 51916 + }, + { + "epoch": 0.51917, + "grad_norm": 1.7164646400192158, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 51917 + }, + { + "epoch": 0.51918, + "grad_norm": 1.2476356461749232, + "learning_rate": 0.003, + "loss": 3.991, + "step": 51918 + }, + { + "epoch": 0.51919, + "grad_norm": 1.7283015329643543, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 51919 + }, + { + "epoch": 0.5192, + "grad_norm": 1.498519041691234, + "learning_rate": 0.003, + "loss": 4.026, + "step": 51920 + }, + { + "epoch": 0.51921, + "grad_norm": 1.5626565891804756, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 51921 + }, + { + "epoch": 0.51922, + "grad_norm": 1.6735691897813179, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 51922 + }, + { + "epoch": 0.51923, + "grad_norm": 1.4511862838981924, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 51923 + }, + { + "epoch": 0.51924, + "grad_norm": 1.5574796694734536, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 51924 + }, + { + "epoch": 0.51925, + "grad_norm": 1.5121425617068234, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 51925 + }, + { + "epoch": 0.51926, + "grad_norm": 1.6038534562487814, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 51926 + }, + { + "epoch": 0.51927, + "grad_norm": 1.8151238816597475, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 51927 + }, + { + "epoch": 0.51928, + "grad_norm": 1.360150062724095, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 51928 + }, + { + "epoch": 0.51929, + "grad_norm": 1.6631807279899176, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 51929 + }, + { + "epoch": 0.5193, + "grad_norm": 1.5700579588000985, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 51930 + }, + { + "epoch": 0.51931, + "grad_norm": 1.630987858987436, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 51931 + }, + { + "epoch": 0.51932, + "grad_norm": 1.4704545154272273, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51932 + }, + { + "epoch": 0.51933, + "grad_norm": 1.4743894599866318, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 51933 + }, + { + "epoch": 0.51934, + "grad_norm": 1.8202903735090277, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 51934 + }, + { + "epoch": 0.51935, + "grad_norm": 1.5438170770599704, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 51935 + }, + { + "epoch": 0.51936, + "grad_norm": 1.4260122807958207, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 51936 + }, + { + "epoch": 0.51937, + "grad_norm": 1.5652469047955695, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 51937 + }, + { + "epoch": 0.51938, + "grad_norm": 1.6435520699970187, + "learning_rate": 0.003, + "loss": 4.014, + "step": 51938 + }, + { + "epoch": 0.51939, + "grad_norm": 1.7161787260520402, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 51939 + }, + { + "epoch": 0.5194, + "grad_norm": 1.7483936046504684, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 51940 + }, + { + "epoch": 0.51941, + "grad_norm": 1.5231189300147066, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 51941 + }, + { + "epoch": 0.51942, + "grad_norm": 1.5132230766871628, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 51942 + }, + { + "epoch": 0.51943, + "grad_norm": 1.4654167324929812, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 51943 + }, + { + "epoch": 0.51944, + "grad_norm": 1.3916864586558166, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 51944 + }, + { + "epoch": 0.51945, + "grad_norm": 1.4914121724600182, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 51945 + }, + { + "epoch": 0.51946, + "grad_norm": 1.5985053058641112, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51946 + }, + { + "epoch": 0.51947, + "grad_norm": 1.3413010671217485, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 51947 + }, + { + "epoch": 0.51948, + "grad_norm": 1.7840154353580648, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 51948 + }, + { + "epoch": 0.51949, + "grad_norm": 1.3472829887512727, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51949 + }, + { + "epoch": 0.5195, + "grad_norm": 1.726654294892419, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 51950 + }, + { + "epoch": 0.51951, + "grad_norm": 1.133918453150249, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 51951 + }, + { + "epoch": 0.51952, + "grad_norm": 1.8553447811075996, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 51952 + }, + { + "epoch": 0.51953, + "grad_norm": 1.3080637417737968, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 51953 + }, + { + "epoch": 0.51954, + "grad_norm": 1.96798829100829, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 51954 + }, + { + "epoch": 0.51955, + "grad_norm": 1.27591223228795, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 51955 + }, + { + "epoch": 0.51956, + "grad_norm": 1.830741277749369, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 51956 + }, + { + "epoch": 0.51957, + "grad_norm": 1.3008205203766146, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 51957 + }, + { + "epoch": 0.51958, + "grad_norm": 1.4368903805839859, + "learning_rate": 0.003, + "loss": 4.028, + "step": 51958 + }, + { + "epoch": 0.51959, + "grad_norm": 1.4718498899663228, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 51959 + }, + { + "epoch": 0.5196, + "grad_norm": 1.5751873869562396, + "learning_rate": 0.003, + "loss": 4.044, + "step": 51960 + }, + { + "epoch": 0.51961, + "grad_norm": 1.465428902159411, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 51961 + }, + { + "epoch": 0.51962, + "grad_norm": 2.043647810777192, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51962 + }, + { + "epoch": 0.51963, + "grad_norm": 1.4384861755531908, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 51963 + }, + { + "epoch": 0.51964, + "grad_norm": 1.725252635638175, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 51964 + }, + { + "epoch": 0.51965, + "grad_norm": 1.7758841895929904, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 51965 + }, + { + "epoch": 0.51966, + "grad_norm": 1.184562897188455, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 51966 + }, + { + "epoch": 0.51967, + "grad_norm": 1.8260319234472933, + "learning_rate": 0.003, + "loss": 4.044, + "step": 51967 + }, + { + "epoch": 0.51968, + "grad_norm": 1.5065570275203615, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 51968 + }, + { + "epoch": 0.51969, + "grad_norm": 1.3477816950854025, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 51969 + }, + { + "epoch": 0.5197, + "grad_norm": 1.6953694500574805, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51970 + }, + { + "epoch": 0.51971, + "grad_norm": 1.685319774878692, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 51971 + }, + { + "epoch": 0.51972, + "grad_norm": 1.5598901077147083, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 51972 + }, + { + "epoch": 0.51973, + "grad_norm": 1.3559869858689655, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 51973 + }, + { + "epoch": 0.51974, + "grad_norm": 1.8427651812675876, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 51974 + }, + { + "epoch": 0.51975, + "grad_norm": 1.355230419045306, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 51975 + }, + { + "epoch": 0.51976, + "grad_norm": 1.8194377628930039, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 51976 + }, + { + "epoch": 0.51977, + "grad_norm": 1.208411814894467, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 51977 + }, + { + "epoch": 0.51978, + "grad_norm": 1.549356300276107, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 51978 + }, + { + "epoch": 0.51979, + "grad_norm": 1.6525951209103262, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 51979 + }, + { + "epoch": 0.5198, + "grad_norm": 1.5719641921812975, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 51980 + }, + { + "epoch": 0.51981, + "grad_norm": 1.4380036983115818, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 51981 + }, + { + "epoch": 0.51982, + "grad_norm": 1.7972218194120149, + "learning_rate": 0.003, + "loss": 4.045, + "step": 51982 + }, + { + "epoch": 0.51983, + "grad_norm": 1.3808435516892619, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51983 + }, + { + "epoch": 0.51984, + "grad_norm": 1.685296515718034, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 51984 + }, + { + "epoch": 0.51985, + "grad_norm": 1.3887066998926645, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 51985 + }, + { + "epoch": 0.51986, + "grad_norm": 1.774160968237483, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 51986 + }, + { + "epoch": 0.51987, + "grad_norm": 1.3617556851551758, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51987 + }, + { + "epoch": 0.51988, + "grad_norm": 1.441882481601765, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51988 + }, + { + "epoch": 0.51989, + "grad_norm": 1.3527322182466819, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51989 + }, + { + "epoch": 0.5199, + "grad_norm": 1.8408133364068928, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 51990 + }, + { + "epoch": 0.51991, + "grad_norm": 1.282469677533682, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 51991 + }, + { + "epoch": 0.51992, + "grad_norm": 1.6536652994132388, + "learning_rate": 0.003, + "loss": 4.056, + "step": 51992 + }, + { + "epoch": 0.51993, + "grad_norm": 1.4533552851309004, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 51993 + }, + { + "epoch": 0.51994, + "grad_norm": 1.867863076451088, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 51994 + }, + { + "epoch": 0.51995, + "grad_norm": 1.5666780546103158, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 51995 + }, + { + "epoch": 0.51996, + "grad_norm": 1.6356919811686554, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 51996 + }, + { + "epoch": 0.51997, + "grad_norm": 1.5900372881649072, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 51997 + }, + { + "epoch": 0.51998, + "grad_norm": 1.4945412966097884, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 51998 + }, + { + "epoch": 0.51999, + "grad_norm": 1.577191408939309, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 51999 + }, + { + "epoch": 0.52, + "grad_norm": 1.5300362789569983, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52000 + }, + { + "epoch": 0.52001, + "grad_norm": 1.691188249952909, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 52001 + }, + { + "epoch": 0.52002, + "grad_norm": 1.5880995338804273, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 52002 + }, + { + "epoch": 0.52003, + "grad_norm": 1.631303312589561, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 52003 + }, + { + "epoch": 0.52004, + "grad_norm": 1.3841149078335895, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52004 + }, + { + "epoch": 0.52005, + "grad_norm": 1.5891508557867053, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 52005 + }, + { + "epoch": 0.52006, + "grad_norm": 1.3617021361250636, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52006 + }, + { + "epoch": 0.52007, + "grad_norm": 1.6460119117337166, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 52007 + }, + { + "epoch": 0.52008, + "grad_norm": 1.5439674747489651, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 52008 + }, + { + "epoch": 0.52009, + "grad_norm": 1.7304964271028633, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 52009 + }, + { + "epoch": 0.5201, + "grad_norm": 1.3411759963245993, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 52010 + }, + { + "epoch": 0.52011, + "grad_norm": 1.9753664687403656, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 52011 + }, + { + "epoch": 0.52012, + "grad_norm": 1.4233052370739827, + "learning_rate": 0.003, + "loss": 4.028, + "step": 52012 + }, + { + "epoch": 0.52013, + "grad_norm": 1.8475660545391848, + "learning_rate": 0.003, + "loss": 4.054, + "step": 52013 + }, + { + "epoch": 0.52014, + "grad_norm": 1.3968550491860579, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 52014 + }, + { + "epoch": 0.52015, + "grad_norm": 1.420381124920659, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52015 + }, + { + "epoch": 0.52016, + "grad_norm": 1.453767833637682, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52016 + }, + { + "epoch": 0.52017, + "grad_norm": 1.7929899155258724, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 52017 + }, + { + "epoch": 0.52018, + "grad_norm": 1.45366314908029, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 52018 + }, + { + "epoch": 0.52019, + "grad_norm": 2.0990511548115856, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 52019 + }, + { + "epoch": 0.5202, + "grad_norm": 1.416389591317553, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52020 + }, + { + "epoch": 0.52021, + "grad_norm": 1.5115379593358882, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 52021 + }, + { + "epoch": 0.52022, + "grad_norm": 1.7655376317653781, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 52022 + }, + { + "epoch": 0.52023, + "grad_norm": 1.7112614510680775, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 52023 + }, + { + "epoch": 0.52024, + "grad_norm": 1.379490066233306, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52024 + }, + { + "epoch": 0.52025, + "grad_norm": 1.5524705900412108, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 52025 + }, + { + "epoch": 0.52026, + "grad_norm": 1.8239911952748686, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 52026 + }, + { + "epoch": 0.52027, + "grad_norm": 1.5019717165742408, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 52027 + }, + { + "epoch": 0.52028, + "grad_norm": 1.688163032943548, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52028 + }, + { + "epoch": 0.52029, + "grad_norm": 1.4204817542746424, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 52029 + }, + { + "epoch": 0.5203, + "grad_norm": 1.663695547820293, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52030 + }, + { + "epoch": 0.52031, + "grad_norm": 1.4358719062217282, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 52031 + }, + { + "epoch": 0.52032, + "grad_norm": 1.773632951937089, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 52032 + }, + { + "epoch": 0.52033, + "grad_norm": 1.361286754806877, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 52033 + }, + { + "epoch": 0.52034, + "grad_norm": 1.7508519688817972, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52034 + }, + { + "epoch": 0.52035, + "grad_norm": 1.4059884083778789, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 52035 + }, + { + "epoch": 0.52036, + "grad_norm": 1.6651838442517581, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 52036 + }, + { + "epoch": 0.52037, + "grad_norm": 1.521774446620009, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 52037 + }, + { + "epoch": 0.52038, + "grad_norm": 1.5544729332599132, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 52038 + }, + { + "epoch": 0.52039, + "grad_norm": 1.685296519582303, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 52039 + }, + { + "epoch": 0.5204, + "grad_norm": 1.459554985955361, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52040 + }, + { + "epoch": 0.52041, + "grad_norm": 1.3625514596710817, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 52041 + }, + { + "epoch": 0.52042, + "grad_norm": 1.5921182818865105, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52042 + }, + { + "epoch": 0.52043, + "grad_norm": 1.611902640518294, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 52043 + }, + { + "epoch": 0.52044, + "grad_norm": 1.7593161196286768, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52044 + }, + { + "epoch": 0.52045, + "grad_norm": 1.1485468712958309, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52045 + }, + { + "epoch": 0.52046, + "grad_norm": 2.062589871399686, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 52046 + }, + { + "epoch": 0.52047, + "grad_norm": 1.4332215952975866, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 52047 + }, + { + "epoch": 0.52048, + "grad_norm": 1.7426092038406662, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 52048 + }, + { + "epoch": 0.52049, + "grad_norm": 1.6210732689746485, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 52049 + }, + { + "epoch": 0.5205, + "grad_norm": 1.6234841287076818, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 52050 + }, + { + "epoch": 0.52051, + "grad_norm": 1.2738699029684988, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 52051 + }, + { + "epoch": 0.52052, + "grad_norm": 1.5580566711807984, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 52052 + }, + { + "epoch": 0.52053, + "grad_norm": 1.708300180253508, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 52053 + }, + { + "epoch": 0.52054, + "grad_norm": 1.6527985557026967, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 52054 + }, + { + "epoch": 0.52055, + "grad_norm": 1.7883805392767411, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 52055 + }, + { + "epoch": 0.52056, + "grad_norm": 1.5525729258992647, + "learning_rate": 0.003, + "loss": 4.036, + "step": 52056 + }, + { + "epoch": 0.52057, + "grad_norm": 1.5666947877558766, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 52057 + }, + { + "epoch": 0.52058, + "grad_norm": 1.1435513484106692, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 52058 + }, + { + "epoch": 0.52059, + "grad_norm": 1.7314606315846635, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52059 + }, + { + "epoch": 0.5206, + "grad_norm": 1.5170808540260352, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 52060 + }, + { + "epoch": 0.52061, + "grad_norm": 1.5648440178974399, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 52061 + }, + { + "epoch": 0.52062, + "grad_norm": 1.453963320965387, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 52062 + }, + { + "epoch": 0.52063, + "grad_norm": 1.525403689096192, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52063 + }, + { + "epoch": 0.52064, + "grad_norm": 1.7732156122174705, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 52064 + }, + { + "epoch": 0.52065, + "grad_norm": 1.451418706865689, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 52065 + }, + { + "epoch": 0.52066, + "grad_norm": 1.7215508443793421, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 52066 + }, + { + "epoch": 0.52067, + "grad_norm": 1.5988065870736061, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 52067 + }, + { + "epoch": 0.52068, + "grad_norm": 1.4548638415308721, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 52068 + }, + { + "epoch": 0.52069, + "grad_norm": 1.4336302418226217, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52069 + }, + { + "epoch": 0.5207, + "grad_norm": 1.5533476975058473, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 52070 + }, + { + "epoch": 0.52071, + "grad_norm": 1.6258799461456166, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52071 + }, + { + "epoch": 0.52072, + "grad_norm": 1.5359218518206903, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52072 + }, + { + "epoch": 0.52073, + "grad_norm": 1.7615742912380912, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 52073 + }, + { + "epoch": 0.52074, + "grad_norm": 1.7037735406481964, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 52074 + }, + { + "epoch": 0.52075, + "grad_norm": 1.7553972124551138, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 52075 + }, + { + "epoch": 0.52076, + "grad_norm": 1.5300911328684856, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 52076 + }, + { + "epoch": 0.52077, + "grad_norm": 1.5696986632524454, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52077 + }, + { + "epoch": 0.52078, + "grad_norm": 1.681564551285718, + "learning_rate": 0.003, + "loss": 4.058, + "step": 52078 + }, + { + "epoch": 0.52079, + "grad_norm": 1.2245135144284252, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 52079 + }, + { + "epoch": 0.5208, + "grad_norm": 1.4924615160134602, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52080 + }, + { + "epoch": 0.52081, + "grad_norm": 1.342335522189065, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 52081 + }, + { + "epoch": 0.52082, + "grad_norm": 1.696730409919574, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52082 + }, + { + "epoch": 0.52083, + "grad_norm": 1.5686820573017766, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 52083 + }, + { + "epoch": 0.52084, + "grad_norm": 2.0205822500685295, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52084 + }, + { + "epoch": 0.52085, + "grad_norm": 1.5424823832735775, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52085 + }, + { + "epoch": 0.52086, + "grad_norm": 1.5609339969596703, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52086 + }, + { + "epoch": 0.52087, + "grad_norm": 1.4965831234632025, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52087 + }, + { + "epoch": 0.52088, + "grad_norm": 1.864133695605441, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52088 + }, + { + "epoch": 0.52089, + "grad_norm": 1.1841311053465986, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 52089 + }, + { + "epoch": 0.5209, + "grad_norm": 1.7438621981639855, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 52090 + }, + { + "epoch": 0.52091, + "grad_norm": 1.3434991669561227, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 52091 + }, + { + "epoch": 0.52092, + "grad_norm": 1.7925276050452468, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 52092 + }, + { + "epoch": 0.52093, + "grad_norm": 1.4921172138198147, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52093 + }, + { + "epoch": 0.52094, + "grad_norm": 1.7008954703447905, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 52094 + }, + { + "epoch": 0.52095, + "grad_norm": 1.5252256669118813, + "learning_rate": 0.003, + "loss": 4.021, + "step": 52095 + }, + { + "epoch": 0.52096, + "grad_norm": 1.6214519145738875, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52096 + }, + { + "epoch": 0.52097, + "grad_norm": 1.60144276328119, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52097 + }, + { + "epoch": 0.52098, + "grad_norm": 1.4298627239844486, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52098 + }, + { + "epoch": 0.52099, + "grad_norm": 1.5002947410277219, + "learning_rate": 0.003, + "loss": 3.998, + "step": 52099 + }, + { + "epoch": 0.521, + "grad_norm": 1.7315969491942904, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 52100 + }, + { + "epoch": 0.52101, + "grad_norm": 1.369327140098979, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 52101 + }, + { + "epoch": 0.52102, + "grad_norm": 1.527882257197602, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52102 + }, + { + "epoch": 0.52103, + "grad_norm": 1.5611448613388454, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52103 + }, + { + "epoch": 0.52104, + "grad_norm": 1.5557341104155933, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 52104 + }, + { + "epoch": 0.52105, + "grad_norm": 1.5260282454605993, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 52105 + }, + { + "epoch": 0.52106, + "grad_norm": 1.5244664681265023, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52106 + }, + { + "epoch": 0.52107, + "grad_norm": 1.7479111403064493, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 52107 + }, + { + "epoch": 0.52108, + "grad_norm": 1.310327965516909, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 52108 + }, + { + "epoch": 0.52109, + "grad_norm": 1.858908100408623, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 52109 + }, + { + "epoch": 0.5211, + "grad_norm": 1.9888049122477762, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52110 + }, + { + "epoch": 0.52111, + "grad_norm": 1.599251996019687, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 52111 + }, + { + "epoch": 0.52112, + "grad_norm": 1.7454119202710145, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 52112 + }, + { + "epoch": 0.52113, + "grad_norm": 1.3823939379906838, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 52113 + }, + { + "epoch": 0.52114, + "grad_norm": 1.7782059381890587, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52114 + }, + { + "epoch": 0.52115, + "grad_norm": 1.4738838045130687, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 52115 + }, + { + "epoch": 0.52116, + "grad_norm": 1.5251519585010755, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 52116 + }, + { + "epoch": 0.52117, + "grad_norm": 1.4356552638930953, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 52117 + }, + { + "epoch": 0.52118, + "grad_norm": 1.749305518864153, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 52118 + }, + { + "epoch": 0.52119, + "grad_norm": 1.2673890355417283, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52119 + }, + { + "epoch": 0.5212, + "grad_norm": 1.5244199691294253, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 52120 + }, + { + "epoch": 0.52121, + "grad_norm": 1.5147276172565574, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 52121 + }, + { + "epoch": 0.52122, + "grad_norm": 1.6925197036564408, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52122 + }, + { + "epoch": 0.52123, + "grad_norm": 1.5606759544829307, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 52123 + }, + { + "epoch": 0.52124, + "grad_norm": 1.571477840456124, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52124 + }, + { + "epoch": 0.52125, + "grad_norm": 1.508722145727862, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 52125 + }, + { + "epoch": 0.52126, + "grad_norm": 1.6208636533966296, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 52126 + }, + { + "epoch": 0.52127, + "grad_norm": 1.4051720457156995, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 52127 + }, + { + "epoch": 0.52128, + "grad_norm": 1.6519550310959663, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 52128 + }, + { + "epoch": 0.52129, + "grad_norm": 1.4665942734208783, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 52129 + }, + { + "epoch": 0.5213, + "grad_norm": 1.7297176931330396, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 52130 + }, + { + "epoch": 0.52131, + "grad_norm": 1.4888836244052894, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 52131 + }, + { + "epoch": 0.52132, + "grad_norm": 1.6702212541564243, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 52132 + }, + { + "epoch": 0.52133, + "grad_norm": 1.6623827456010951, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 52133 + }, + { + "epoch": 0.52134, + "grad_norm": 1.8053772736135236, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 52134 + }, + { + "epoch": 0.52135, + "grad_norm": 1.5651425391984266, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52135 + }, + { + "epoch": 0.52136, + "grad_norm": 1.698378870031921, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52136 + }, + { + "epoch": 0.52137, + "grad_norm": 1.494239121498373, + "learning_rate": 0.003, + "loss": 4.005, + "step": 52137 + }, + { + "epoch": 0.52138, + "grad_norm": 1.5308705385260952, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 52138 + }, + { + "epoch": 0.52139, + "grad_norm": 1.2557800785094395, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 52139 + }, + { + "epoch": 0.5214, + "grad_norm": 1.591960134292105, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52140 + }, + { + "epoch": 0.52141, + "grad_norm": 1.3829128368892796, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 52141 + }, + { + "epoch": 0.52142, + "grad_norm": 1.9943023060538252, + "learning_rate": 0.003, + "loss": 4.044, + "step": 52142 + }, + { + "epoch": 0.52143, + "grad_norm": 2.022704875960234, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 52143 + }, + { + "epoch": 0.52144, + "grad_norm": 1.5358468201828523, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 52144 + }, + { + "epoch": 0.52145, + "grad_norm": 1.4875194014144681, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 52145 + }, + { + "epoch": 0.52146, + "grad_norm": 1.5438804982770218, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 52146 + }, + { + "epoch": 0.52147, + "grad_norm": 1.59504733412294, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52147 + }, + { + "epoch": 0.52148, + "grad_norm": 1.4994064728143646, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 52148 + }, + { + "epoch": 0.52149, + "grad_norm": 1.483740499678533, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52149 + }, + { + "epoch": 0.5215, + "grad_norm": 1.6519759144725445, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52150 + }, + { + "epoch": 0.52151, + "grad_norm": 1.2510499974762788, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52151 + }, + { + "epoch": 0.52152, + "grad_norm": 1.77620354699926, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52152 + }, + { + "epoch": 0.52153, + "grad_norm": 1.391739937128246, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52153 + }, + { + "epoch": 0.52154, + "grad_norm": 1.4939974940409873, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 52154 + }, + { + "epoch": 0.52155, + "grad_norm": 1.9534232455333393, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 52155 + }, + { + "epoch": 0.52156, + "grad_norm": 1.6553520774337696, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 52156 + }, + { + "epoch": 0.52157, + "grad_norm": 1.4209892618165847, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 52157 + }, + { + "epoch": 0.52158, + "grad_norm": 1.6462812859743219, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 52158 + }, + { + "epoch": 0.52159, + "grad_norm": 1.5259469479644674, + "learning_rate": 0.003, + "loss": 4.035, + "step": 52159 + }, + { + "epoch": 0.5216, + "grad_norm": 1.4413845143821173, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 52160 + }, + { + "epoch": 0.52161, + "grad_norm": 1.5136584806763433, + "learning_rate": 0.003, + "loss": 4.033, + "step": 52161 + }, + { + "epoch": 0.52162, + "grad_norm": 1.7199445491833543, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 52162 + }, + { + "epoch": 0.52163, + "grad_norm": 1.294073851675041, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52163 + }, + { + "epoch": 0.52164, + "grad_norm": 1.6437617502952404, + "learning_rate": 0.003, + "loss": 4.039, + "step": 52164 + }, + { + "epoch": 0.52165, + "grad_norm": 1.619938774377031, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 52165 + }, + { + "epoch": 0.52166, + "grad_norm": 1.2848853034957275, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 52166 + }, + { + "epoch": 0.52167, + "grad_norm": 1.6322448091947657, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 52167 + }, + { + "epoch": 0.52168, + "grad_norm": 1.3295702535521303, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 52168 + }, + { + "epoch": 0.52169, + "grad_norm": 1.383703767757821, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 52169 + }, + { + "epoch": 0.5217, + "grad_norm": 1.7568600728457553, + "learning_rate": 0.003, + "loss": 4.047, + "step": 52170 + }, + { + "epoch": 0.52171, + "grad_norm": 1.718520320729335, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 52171 + }, + { + "epoch": 0.52172, + "grad_norm": 1.7432992980616593, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52172 + }, + { + "epoch": 0.52173, + "grad_norm": 1.844852369723745, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 52173 + }, + { + "epoch": 0.52174, + "grad_norm": 1.5419448419488673, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 52174 + }, + { + "epoch": 0.52175, + "grad_norm": 1.5927455910253738, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 52175 + }, + { + "epoch": 0.52176, + "grad_norm": 1.5554566401312946, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52176 + }, + { + "epoch": 0.52177, + "grad_norm": 1.7479967584380292, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 52177 + }, + { + "epoch": 0.52178, + "grad_norm": 1.481426697804332, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 52178 + }, + { + "epoch": 0.52179, + "grad_norm": 1.4853790285593294, + "learning_rate": 0.003, + "loss": 4.006, + "step": 52179 + }, + { + "epoch": 0.5218, + "grad_norm": 1.4908037277844344, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52180 + }, + { + "epoch": 0.52181, + "grad_norm": 1.4509798457770433, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 52181 + }, + { + "epoch": 0.52182, + "grad_norm": 1.605437899943983, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 52182 + }, + { + "epoch": 0.52183, + "grad_norm": 1.737347916438756, + "learning_rate": 0.003, + "loss": 4.045, + "step": 52183 + }, + { + "epoch": 0.52184, + "grad_norm": 1.5878057415953155, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 52184 + }, + { + "epoch": 0.52185, + "grad_norm": 1.86014626086439, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 52185 + }, + { + "epoch": 0.52186, + "grad_norm": 1.71382069476353, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 52186 + }, + { + "epoch": 0.52187, + "grad_norm": 1.35678653448407, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 52187 + }, + { + "epoch": 0.52188, + "grad_norm": 1.8222998772096142, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 52188 + }, + { + "epoch": 0.52189, + "grad_norm": 1.752728768894567, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 52189 + }, + { + "epoch": 0.5219, + "grad_norm": 1.2639027331946184, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52190 + }, + { + "epoch": 0.52191, + "grad_norm": 1.6658351295241378, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52191 + }, + { + "epoch": 0.52192, + "grad_norm": 1.3211694385641308, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 52192 + }, + { + "epoch": 0.52193, + "grad_norm": 1.6022728973453402, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 52193 + }, + { + "epoch": 0.52194, + "grad_norm": 1.7733579427823571, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 52194 + }, + { + "epoch": 0.52195, + "grad_norm": 1.7375358899709281, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 52195 + }, + { + "epoch": 0.52196, + "grad_norm": 1.668879263481781, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 52196 + }, + { + "epoch": 0.52197, + "grad_norm": 1.331374609117241, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 52197 + }, + { + "epoch": 0.52198, + "grad_norm": 1.6504832208503082, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 52198 + }, + { + "epoch": 0.52199, + "grad_norm": 1.6186873733139164, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 52199 + }, + { + "epoch": 0.522, + "grad_norm": 1.4266740553081394, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52200 + }, + { + "epoch": 0.52201, + "grad_norm": 1.5438414282125303, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 52201 + }, + { + "epoch": 0.52202, + "grad_norm": 1.2814232454006789, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 52202 + }, + { + "epoch": 0.52203, + "grad_norm": 1.74008757722727, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 52203 + }, + { + "epoch": 0.52204, + "grad_norm": 1.2855708614580188, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52204 + }, + { + "epoch": 0.52205, + "grad_norm": 1.8727203444259992, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 52205 + }, + { + "epoch": 0.52206, + "grad_norm": 1.5023605305393049, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52206 + }, + { + "epoch": 0.52207, + "grad_norm": 1.9525322091987185, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 52207 + }, + { + "epoch": 0.52208, + "grad_norm": 2.082701617490719, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 52208 + }, + { + "epoch": 0.52209, + "grad_norm": 1.2576786890573923, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52209 + }, + { + "epoch": 0.5221, + "grad_norm": 1.5398481266554682, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52210 + }, + { + "epoch": 0.52211, + "grad_norm": 1.9323917940567104, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52211 + }, + { + "epoch": 0.52212, + "grad_norm": 1.6273682161880418, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 52212 + }, + { + "epoch": 0.52213, + "grad_norm": 1.2108406091623178, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52213 + }, + { + "epoch": 0.52214, + "grad_norm": 1.937757723750447, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 52214 + }, + { + "epoch": 0.52215, + "grad_norm": 1.6114844469200087, + "learning_rate": 0.003, + "loss": 4.041, + "step": 52215 + }, + { + "epoch": 0.52216, + "grad_norm": 1.2365594261430515, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52216 + }, + { + "epoch": 0.52217, + "grad_norm": 1.5223794878999433, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 52217 + }, + { + "epoch": 0.52218, + "grad_norm": 1.3650227132457033, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 52218 + }, + { + "epoch": 0.52219, + "grad_norm": 1.774658684575488, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 52219 + }, + { + "epoch": 0.5222, + "grad_norm": 1.459404762181143, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 52220 + }, + { + "epoch": 0.52221, + "grad_norm": 1.8775126363681032, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 52221 + }, + { + "epoch": 0.52222, + "grad_norm": 1.7058551460860956, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 52222 + }, + { + "epoch": 0.52223, + "grad_norm": 1.5233161060711882, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52223 + }, + { + "epoch": 0.52224, + "grad_norm": 1.4175259402703557, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 52224 + }, + { + "epoch": 0.52225, + "grad_norm": 1.79482936937732, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 52225 + }, + { + "epoch": 0.52226, + "grad_norm": 1.5088222776381877, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 52226 + }, + { + "epoch": 0.52227, + "grad_norm": 1.5349610279243466, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 52227 + }, + { + "epoch": 0.52228, + "grad_norm": 1.611076865678997, + "learning_rate": 0.003, + "loss": 4.041, + "step": 52228 + }, + { + "epoch": 0.52229, + "grad_norm": 1.616182467007457, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 52229 + }, + { + "epoch": 0.5223, + "grad_norm": 1.5634541085810985, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 52230 + }, + { + "epoch": 0.52231, + "grad_norm": 1.522836870269127, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 52231 + }, + { + "epoch": 0.52232, + "grad_norm": 1.3713078111444719, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52232 + }, + { + "epoch": 0.52233, + "grad_norm": 1.7002235176454712, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 52233 + }, + { + "epoch": 0.52234, + "grad_norm": 1.6164310755182913, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 52234 + }, + { + "epoch": 0.52235, + "grad_norm": 2.0400979714715244, + "learning_rate": 0.003, + "loss": 4.076, + "step": 52235 + }, + { + "epoch": 0.52236, + "grad_norm": 1.2031817022548439, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 52236 + }, + { + "epoch": 0.52237, + "grad_norm": 1.5730557008814712, + "learning_rate": 0.003, + "loss": 4.064, + "step": 52237 + }, + { + "epoch": 0.52238, + "grad_norm": 1.6163589619500978, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 52238 + }, + { + "epoch": 0.52239, + "grad_norm": 1.9420837054567985, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 52239 + }, + { + "epoch": 0.5224, + "grad_norm": 1.3262185186275064, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 52240 + }, + { + "epoch": 0.52241, + "grad_norm": 1.4100206207115014, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 52241 + }, + { + "epoch": 0.52242, + "grad_norm": 1.609901443270704, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 52242 + }, + { + "epoch": 0.52243, + "grad_norm": 1.3175815653584368, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 52243 + }, + { + "epoch": 0.52244, + "grad_norm": 1.7698070644717396, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 52244 + }, + { + "epoch": 0.52245, + "grad_norm": 1.6413972338665357, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 52245 + }, + { + "epoch": 0.52246, + "grad_norm": 1.5666417613019687, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 52246 + }, + { + "epoch": 0.52247, + "grad_norm": 1.6484141320986525, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 52247 + }, + { + "epoch": 0.52248, + "grad_norm": 1.6610141029804188, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 52248 + }, + { + "epoch": 0.52249, + "grad_norm": 1.5997437125709177, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52249 + }, + { + "epoch": 0.5225, + "grad_norm": 1.5010207146396362, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52250 + }, + { + "epoch": 0.52251, + "grad_norm": 1.5958409067097703, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52251 + }, + { + "epoch": 0.52252, + "grad_norm": 1.2980111572594566, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 52252 + }, + { + "epoch": 0.52253, + "grad_norm": 1.926389360859279, + "learning_rate": 0.003, + "loss": 4.011, + "step": 52253 + }, + { + "epoch": 0.52254, + "grad_norm": 1.00194149752034, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 52254 + }, + { + "epoch": 0.52255, + "grad_norm": 2.051284964434478, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 52255 + }, + { + "epoch": 0.52256, + "grad_norm": 1.4305407877664043, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 52256 + }, + { + "epoch": 0.52257, + "grad_norm": 1.6323697198057134, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52257 + }, + { + "epoch": 0.52258, + "grad_norm": 1.6774487791808534, + "learning_rate": 0.003, + "loss": 4.049, + "step": 52258 + }, + { + "epoch": 0.52259, + "grad_norm": 1.916926117168866, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 52259 + }, + { + "epoch": 0.5226, + "grad_norm": 1.2626632665377842, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 52260 + }, + { + "epoch": 0.52261, + "grad_norm": 1.7654996556295741, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 52261 + }, + { + "epoch": 0.52262, + "grad_norm": 1.5445627897919796, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 52262 + }, + { + "epoch": 0.52263, + "grad_norm": 1.5135354208897038, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 52263 + }, + { + "epoch": 0.52264, + "grad_norm": 1.432900866453553, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 52264 + }, + { + "epoch": 0.52265, + "grad_norm": 1.695948152555201, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 52265 + }, + { + "epoch": 0.52266, + "grad_norm": 1.3838295987570957, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52266 + }, + { + "epoch": 0.52267, + "grad_norm": 1.6622579708480172, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 52267 + }, + { + "epoch": 0.52268, + "grad_norm": 1.8320480141371969, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 52268 + }, + { + "epoch": 0.52269, + "grad_norm": 1.3443833726624683, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 52269 + }, + { + "epoch": 0.5227, + "grad_norm": 1.9887374303416046, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52270 + }, + { + "epoch": 0.52271, + "grad_norm": 1.4972255033721202, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 52271 + }, + { + "epoch": 0.52272, + "grad_norm": 1.577278986790464, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 52272 + }, + { + "epoch": 0.52273, + "grad_norm": 2.0002824273809288, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 52273 + }, + { + "epoch": 0.52274, + "grad_norm": 1.4060152276002218, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 52274 + }, + { + "epoch": 0.52275, + "grad_norm": 1.309499186042873, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 52275 + }, + { + "epoch": 0.52276, + "grad_norm": 1.6425679726203317, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52276 + }, + { + "epoch": 0.52277, + "grad_norm": 1.5109649220580195, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 52277 + }, + { + "epoch": 0.52278, + "grad_norm": 1.8748105833233017, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 52278 + }, + { + "epoch": 0.52279, + "grad_norm": 1.2137414600213892, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 52279 + }, + { + "epoch": 0.5228, + "grad_norm": 1.549423085797477, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52280 + }, + { + "epoch": 0.52281, + "grad_norm": 1.301445204185722, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 52281 + }, + { + "epoch": 0.52282, + "grad_norm": 1.5906830380041752, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 52282 + }, + { + "epoch": 0.52283, + "grad_norm": 1.4943627888778694, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 52283 + }, + { + "epoch": 0.52284, + "grad_norm": 1.752683023145845, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 52284 + }, + { + "epoch": 0.52285, + "grad_norm": 1.5918542013212749, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 52285 + }, + { + "epoch": 0.52286, + "grad_norm": 1.591966046892281, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52286 + }, + { + "epoch": 0.52287, + "grad_norm": 1.5314830669036432, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 52287 + }, + { + "epoch": 0.52288, + "grad_norm": 1.3518469119669823, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52288 + }, + { + "epoch": 0.52289, + "grad_norm": 1.6356750056283145, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52289 + }, + { + "epoch": 0.5229, + "grad_norm": 1.5850771100398322, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52290 + }, + { + "epoch": 0.52291, + "grad_norm": 1.6061967103413703, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52291 + }, + { + "epoch": 0.52292, + "grad_norm": 1.5428428322859409, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52292 + }, + { + "epoch": 0.52293, + "grad_norm": 1.7714306794363586, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 52293 + }, + { + "epoch": 0.52294, + "grad_norm": 1.5284366502479279, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 52294 + }, + { + "epoch": 0.52295, + "grad_norm": 1.611993771281484, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 52295 + }, + { + "epoch": 0.52296, + "grad_norm": 1.5167174123975888, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 52296 + }, + { + "epoch": 0.52297, + "grad_norm": 1.4478516830902306, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52297 + }, + { + "epoch": 0.52298, + "grad_norm": 1.4512656192762197, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52298 + }, + { + "epoch": 0.52299, + "grad_norm": 1.6687965052932348, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 52299 + }, + { + "epoch": 0.523, + "grad_norm": 1.5597560397904489, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52300 + }, + { + "epoch": 0.52301, + "grad_norm": 1.4615145927708522, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 52301 + }, + { + "epoch": 0.52302, + "grad_norm": 1.3211609356326008, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 52302 + }, + { + "epoch": 0.52303, + "grad_norm": 1.8906903554949837, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 52303 + }, + { + "epoch": 0.52304, + "grad_norm": 1.3120937721003112, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 52304 + }, + { + "epoch": 0.52305, + "grad_norm": 1.7257405915448236, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 52305 + }, + { + "epoch": 0.52306, + "grad_norm": 1.6629297981035, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 52306 + }, + { + "epoch": 0.52307, + "grad_norm": 1.6854424099915755, + "learning_rate": 0.003, + "loss": 4.019, + "step": 52307 + }, + { + "epoch": 0.52308, + "grad_norm": 1.3378146847215058, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 52308 + }, + { + "epoch": 0.52309, + "grad_norm": 1.7512468942941883, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52309 + }, + { + "epoch": 0.5231, + "grad_norm": 1.33098472568512, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 52310 + }, + { + "epoch": 0.52311, + "grad_norm": 1.5009746043322756, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 52311 + }, + { + "epoch": 0.52312, + "grad_norm": 1.6051745133495905, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 52312 + }, + { + "epoch": 0.52313, + "grad_norm": 1.9106941126751513, + "learning_rate": 0.003, + "loss": 4.048, + "step": 52313 + }, + { + "epoch": 0.52314, + "grad_norm": 1.3172106942494364, + "learning_rate": 0.003, + "loss": 4.047, + "step": 52314 + }, + { + "epoch": 0.52315, + "grad_norm": 1.827303099298261, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 52315 + }, + { + "epoch": 0.52316, + "grad_norm": 1.2874193353187013, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 52316 + }, + { + "epoch": 0.52317, + "grad_norm": 1.7680938831592599, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 52317 + }, + { + "epoch": 0.52318, + "grad_norm": 1.5375541489134517, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 52318 + }, + { + "epoch": 0.52319, + "grad_norm": 1.4166728207275021, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 52319 + }, + { + "epoch": 0.5232, + "grad_norm": 1.72912128237929, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 52320 + }, + { + "epoch": 0.52321, + "grad_norm": 1.3748358259177753, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 52321 + }, + { + "epoch": 0.52322, + "grad_norm": 1.7951755124890212, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 52322 + }, + { + "epoch": 0.52323, + "grad_norm": 1.5395242603661867, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 52323 + }, + { + "epoch": 0.52324, + "grad_norm": 1.506799444965746, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52324 + }, + { + "epoch": 0.52325, + "grad_norm": 1.54402006832546, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52325 + }, + { + "epoch": 0.52326, + "grad_norm": 1.5924443879328976, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 52326 + }, + { + "epoch": 0.52327, + "grad_norm": 1.5094532894945276, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 52327 + }, + { + "epoch": 0.52328, + "grad_norm": 1.5580050654005007, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52328 + }, + { + "epoch": 0.52329, + "grad_norm": 1.5378631433134813, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 52329 + }, + { + "epoch": 0.5233, + "grad_norm": 1.6277437403568382, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 52330 + }, + { + "epoch": 0.52331, + "grad_norm": 1.3518817541641013, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 52331 + }, + { + "epoch": 0.52332, + "grad_norm": 1.8738620118721552, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52332 + }, + { + "epoch": 0.52333, + "grad_norm": 1.5234856495500884, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 52333 + }, + { + "epoch": 0.52334, + "grad_norm": 1.6112745908172528, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 52334 + }, + { + "epoch": 0.52335, + "grad_norm": 1.653093246768048, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52335 + }, + { + "epoch": 0.52336, + "grad_norm": 1.6356021828645508, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 52336 + }, + { + "epoch": 0.52337, + "grad_norm": 1.4910764650834525, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 52337 + }, + { + "epoch": 0.52338, + "grad_norm": 1.7640711326692655, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 52338 + }, + { + "epoch": 0.52339, + "grad_norm": 1.2663471782972102, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 52339 + }, + { + "epoch": 0.5234, + "grad_norm": 1.680362038412724, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 52340 + }, + { + "epoch": 0.52341, + "grad_norm": 1.4481594427793787, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 52341 + }, + { + "epoch": 0.52342, + "grad_norm": 1.951341623546251, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 52342 + }, + { + "epoch": 0.52343, + "grad_norm": 1.313457808863974, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 52343 + }, + { + "epoch": 0.52344, + "grad_norm": 2.2731243001754677, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 52344 + }, + { + "epoch": 0.52345, + "grad_norm": 1.8023860866375916, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 52345 + }, + { + "epoch": 0.52346, + "grad_norm": 1.235120222265918, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52346 + }, + { + "epoch": 0.52347, + "grad_norm": 1.5816075391986566, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 52347 + }, + { + "epoch": 0.52348, + "grad_norm": 1.4794471542077674, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 52348 + }, + { + "epoch": 0.52349, + "grad_norm": 1.8857595168702843, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 52349 + }, + { + "epoch": 0.5235, + "grad_norm": 1.3934496830334266, + "learning_rate": 0.003, + "loss": 4.015, + "step": 52350 + }, + { + "epoch": 0.52351, + "grad_norm": 1.9152081572536956, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52351 + }, + { + "epoch": 0.52352, + "grad_norm": 1.3016567866836644, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52352 + }, + { + "epoch": 0.52353, + "grad_norm": 1.5102415330060202, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52353 + }, + { + "epoch": 0.52354, + "grad_norm": 1.4957653048718655, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 52354 + }, + { + "epoch": 0.52355, + "grad_norm": 1.6460041983282043, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 52355 + }, + { + "epoch": 0.52356, + "grad_norm": 1.626255596346924, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 52356 + }, + { + "epoch": 0.52357, + "grad_norm": 1.649298240611961, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 52357 + }, + { + "epoch": 0.52358, + "grad_norm": 1.363969830672953, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 52358 + }, + { + "epoch": 0.52359, + "grad_norm": 1.7657547741084951, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 52359 + }, + { + "epoch": 0.5236, + "grad_norm": 1.4873127618735573, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52360 + }, + { + "epoch": 0.52361, + "grad_norm": 1.7738539219791287, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 52361 + }, + { + "epoch": 0.52362, + "grad_norm": 1.155140312866341, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 52362 + }, + { + "epoch": 0.52363, + "grad_norm": 1.9743782153569234, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 52363 + }, + { + "epoch": 0.52364, + "grad_norm": 1.6924680684937379, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 52364 + }, + { + "epoch": 0.52365, + "grad_norm": 1.467620669636447, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 52365 + }, + { + "epoch": 0.52366, + "grad_norm": 1.375724826929561, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 52366 + }, + { + "epoch": 0.52367, + "grad_norm": 1.7892078480993625, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 52367 + }, + { + "epoch": 0.52368, + "grad_norm": 1.1282020686949308, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 52368 + }, + { + "epoch": 0.52369, + "grad_norm": 1.5761963173968336, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 52369 + }, + { + "epoch": 0.5237, + "grad_norm": 1.5052590957056264, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 52370 + }, + { + "epoch": 0.52371, + "grad_norm": 1.739738667851665, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52371 + }, + { + "epoch": 0.52372, + "grad_norm": 1.6302217495583677, + "learning_rate": 0.003, + "loss": 4.038, + "step": 52372 + }, + { + "epoch": 0.52373, + "grad_norm": 1.3536015548715226, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 52373 + }, + { + "epoch": 0.52374, + "grad_norm": 1.5852939220799107, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 52374 + }, + { + "epoch": 0.52375, + "grad_norm": 1.6565198580767548, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 52375 + }, + { + "epoch": 0.52376, + "grad_norm": 1.8169755923043227, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52376 + }, + { + "epoch": 0.52377, + "grad_norm": 1.371570686145544, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52377 + }, + { + "epoch": 0.52378, + "grad_norm": 1.6689584998443818, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 52378 + }, + { + "epoch": 0.52379, + "grad_norm": 1.721060924692459, + "learning_rate": 0.003, + "loss": 4.035, + "step": 52379 + }, + { + "epoch": 0.5238, + "grad_norm": 1.5379507366851273, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 52380 + }, + { + "epoch": 0.52381, + "grad_norm": 1.6342200691788498, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 52381 + }, + { + "epoch": 0.52382, + "grad_norm": 1.4255252535863734, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52382 + }, + { + "epoch": 0.52383, + "grad_norm": 1.64631394006485, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52383 + }, + { + "epoch": 0.52384, + "grad_norm": 1.2227196613059372, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52384 + }, + { + "epoch": 0.52385, + "grad_norm": 1.8322318839118885, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 52385 + }, + { + "epoch": 0.52386, + "grad_norm": 1.2768680151095686, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 52386 + }, + { + "epoch": 0.52387, + "grad_norm": 1.8283600902287365, + "learning_rate": 0.003, + "loss": 4.068, + "step": 52387 + }, + { + "epoch": 0.52388, + "grad_norm": 1.3008635241545918, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52388 + }, + { + "epoch": 0.52389, + "grad_norm": 1.8665356979078596, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52389 + }, + { + "epoch": 0.5239, + "grad_norm": 1.5833236280498326, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 52390 + }, + { + "epoch": 0.52391, + "grad_norm": 1.9000796744329944, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 52391 + }, + { + "epoch": 0.52392, + "grad_norm": 1.7136635762421524, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 52392 + }, + { + "epoch": 0.52393, + "grad_norm": 1.5240320252839403, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 52393 + }, + { + "epoch": 0.52394, + "grad_norm": 1.5514109926904216, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 52394 + }, + { + "epoch": 0.52395, + "grad_norm": 1.636346340837063, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 52395 + }, + { + "epoch": 0.52396, + "grad_norm": 1.5734000416751117, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 52396 + }, + { + "epoch": 0.52397, + "grad_norm": 1.4166850573821284, + "learning_rate": 0.003, + "loss": 4.048, + "step": 52397 + }, + { + "epoch": 0.52398, + "grad_norm": 1.5893671003621843, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 52398 + }, + { + "epoch": 0.52399, + "grad_norm": 1.4982361314470107, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 52399 + }, + { + "epoch": 0.524, + "grad_norm": 1.5144887723519695, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 52400 + }, + { + "epoch": 0.52401, + "grad_norm": 1.4760658718297879, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 52401 + }, + { + "epoch": 0.52402, + "grad_norm": 1.5322820462479552, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 52402 + }, + { + "epoch": 0.52403, + "grad_norm": 1.5559161744917744, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 52403 + }, + { + "epoch": 0.52404, + "grad_norm": 1.8864878258381605, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 52404 + }, + { + "epoch": 0.52405, + "grad_norm": 1.6611195328442725, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 52405 + }, + { + "epoch": 0.52406, + "grad_norm": 1.6947837894683333, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52406 + }, + { + "epoch": 0.52407, + "grad_norm": 1.657217833505021, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 52407 + }, + { + "epoch": 0.52408, + "grad_norm": 1.5979737752591747, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 52408 + }, + { + "epoch": 0.52409, + "grad_norm": 1.5500072863180434, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 52409 + }, + { + "epoch": 0.5241, + "grad_norm": 1.5562956773751495, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 52410 + }, + { + "epoch": 0.52411, + "grad_norm": 1.6416118200000538, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 52411 + }, + { + "epoch": 0.52412, + "grad_norm": 1.626644882669054, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 52412 + }, + { + "epoch": 0.52413, + "grad_norm": 1.4846212119128337, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 52413 + }, + { + "epoch": 0.52414, + "grad_norm": 1.5646965489616569, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52414 + }, + { + "epoch": 0.52415, + "grad_norm": 1.742706343258165, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52415 + }, + { + "epoch": 0.52416, + "grad_norm": 1.4645775194892283, + "learning_rate": 0.003, + "loss": 4.028, + "step": 52416 + }, + { + "epoch": 0.52417, + "grad_norm": 1.5337128546791838, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 52417 + }, + { + "epoch": 0.52418, + "grad_norm": 1.7991008890490459, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52418 + }, + { + "epoch": 0.52419, + "grad_norm": 1.3445196899725336, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 52419 + }, + { + "epoch": 0.5242, + "grad_norm": 1.6474419182536237, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 52420 + }, + { + "epoch": 0.52421, + "grad_norm": 1.2876038854013678, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 52421 + }, + { + "epoch": 0.52422, + "grad_norm": 1.5399059295808346, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 52422 + }, + { + "epoch": 0.52423, + "grad_norm": 1.4267066538910438, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52423 + }, + { + "epoch": 0.52424, + "grad_norm": 1.3686220320260403, + "learning_rate": 0.003, + "loss": 4.046, + "step": 52424 + }, + { + "epoch": 0.52425, + "grad_norm": 1.4640517752547229, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 52425 + }, + { + "epoch": 0.52426, + "grad_norm": 1.2270856983050356, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 52426 + }, + { + "epoch": 0.52427, + "grad_norm": 1.6590259168353516, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52427 + }, + { + "epoch": 0.52428, + "grad_norm": 1.4275084088455123, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 52428 + }, + { + "epoch": 0.52429, + "grad_norm": 1.979740558611239, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52429 + }, + { + "epoch": 0.5243, + "grad_norm": 1.9019983505975226, + "learning_rate": 0.003, + "loss": 4.062, + "step": 52430 + }, + { + "epoch": 0.52431, + "grad_norm": 1.809427835729836, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52431 + }, + { + "epoch": 0.52432, + "grad_norm": 1.6487226029093833, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 52432 + }, + { + "epoch": 0.52433, + "grad_norm": 1.5744838662817622, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52433 + }, + { + "epoch": 0.52434, + "grad_norm": 1.6470764447418893, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 52434 + }, + { + "epoch": 0.52435, + "grad_norm": 1.6189256346569039, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 52435 + }, + { + "epoch": 0.52436, + "grad_norm": 1.3713641770068206, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52436 + }, + { + "epoch": 0.52437, + "grad_norm": 1.7561970296864131, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 52437 + }, + { + "epoch": 0.52438, + "grad_norm": 1.7156539802660815, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 52438 + }, + { + "epoch": 0.52439, + "grad_norm": 1.3822855605830389, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 52439 + }, + { + "epoch": 0.5244, + "grad_norm": 1.6251984527966339, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 52440 + }, + { + "epoch": 0.52441, + "grad_norm": 1.5928845500166127, + "learning_rate": 0.003, + "loss": 4.033, + "step": 52441 + }, + { + "epoch": 0.52442, + "grad_norm": 1.749654551285293, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 52442 + }, + { + "epoch": 0.52443, + "grad_norm": 1.374217738492876, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52443 + }, + { + "epoch": 0.52444, + "grad_norm": 1.890660621489261, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 52444 + }, + { + "epoch": 0.52445, + "grad_norm": 1.3724197320509088, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52445 + }, + { + "epoch": 0.52446, + "grad_norm": 1.6847399565625814, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 52446 + }, + { + "epoch": 0.52447, + "grad_norm": 1.4680389410526522, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 52447 + }, + { + "epoch": 0.52448, + "grad_norm": 1.4576015056007365, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 52448 + }, + { + "epoch": 0.52449, + "grad_norm": 1.6950306877299446, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 52449 + }, + { + "epoch": 0.5245, + "grad_norm": 1.3686454530498453, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 52450 + }, + { + "epoch": 0.52451, + "grad_norm": 1.8408063407881832, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 52451 + }, + { + "epoch": 0.52452, + "grad_norm": 1.4996616419904663, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52452 + }, + { + "epoch": 0.52453, + "grad_norm": 1.5058310934954675, + "learning_rate": 0.003, + "loss": 4.024, + "step": 52453 + }, + { + "epoch": 0.52454, + "grad_norm": 1.4200362983800727, + "learning_rate": 0.003, + "loss": 4.06, + "step": 52454 + }, + { + "epoch": 0.52455, + "grad_norm": 1.7189944190285356, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 52455 + }, + { + "epoch": 0.52456, + "grad_norm": 1.515261629409625, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 52456 + }, + { + "epoch": 0.52457, + "grad_norm": 1.2706212331646463, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52457 + }, + { + "epoch": 0.52458, + "grad_norm": 1.7593997214843482, + "learning_rate": 0.003, + "loss": 4.057, + "step": 52458 + }, + { + "epoch": 0.52459, + "grad_norm": 1.4333243230478545, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 52459 + }, + { + "epoch": 0.5246, + "grad_norm": 1.696930774342953, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 52460 + }, + { + "epoch": 0.52461, + "grad_norm": 1.3324340118056908, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 52461 + }, + { + "epoch": 0.52462, + "grad_norm": 1.6250906564596816, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 52462 + }, + { + "epoch": 0.52463, + "grad_norm": 1.5674900230670612, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 52463 + }, + { + "epoch": 0.52464, + "grad_norm": 1.7952153288221002, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 52464 + }, + { + "epoch": 0.52465, + "grad_norm": 1.378583422219416, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 52465 + }, + { + "epoch": 0.52466, + "grad_norm": 1.8400076919139958, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 52466 + }, + { + "epoch": 0.52467, + "grad_norm": 1.2028086133973743, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 52467 + }, + { + "epoch": 0.52468, + "grad_norm": 1.913700294725184, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 52468 + }, + { + "epoch": 0.52469, + "grad_norm": 1.220591672149388, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 52469 + }, + { + "epoch": 0.5247, + "grad_norm": 1.5943130587967105, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 52470 + }, + { + "epoch": 0.52471, + "grad_norm": 1.495219658236524, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52471 + }, + { + "epoch": 0.52472, + "grad_norm": 2.0759688264141407, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 52472 + }, + { + "epoch": 0.52473, + "grad_norm": 1.2419931467586307, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52473 + }, + { + "epoch": 0.52474, + "grad_norm": 1.7885151360684302, + "learning_rate": 0.003, + "loss": 4.016, + "step": 52474 + }, + { + "epoch": 0.52475, + "grad_norm": 1.7915506629240572, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 52475 + }, + { + "epoch": 0.52476, + "grad_norm": 1.645962870506179, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 52476 + }, + { + "epoch": 0.52477, + "grad_norm": 1.3912290261860316, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52477 + }, + { + "epoch": 0.52478, + "grad_norm": 1.6231800850412201, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 52478 + }, + { + "epoch": 0.52479, + "grad_norm": 1.4535213059177394, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 52479 + }, + { + "epoch": 0.5248, + "grad_norm": 1.462274276828559, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52480 + }, + { + "epoch": 0.52481, + "grad_norm": 1.669112912361285, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 52481 + }, + { + "epoch": 0.52482, + "grad_norm": 1.4860847581384315, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 52482 + }, + { + "epoch": 0.52483, + "grad_norm": 1.724878428341215, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 52483 + }, + { + "epoch": 0.52484, + "grad_norm": 1.253000030215079, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 52484 + }, + { + "epoch": 0.52485, + "grad_norm": 1.9228322425031528, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 52485 + }, + { + "epoch": 0.52486, + "grad_norm": 1.3989572592827815, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52486 + }, + { + "epoch": 0.52487, + "grad_norm": 1.6343134049392705, + "learning_rate": 0.003, + "loss": 4.044, + "step": 52487 + }, + { + "epoch": 0.52488, + "grad_norm": 1.538506613528269, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 52488 + }, + { + "epoch": 0.52489, + "grad_norm": 1.573902413250057, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 52489 + }, + { + "epoch": 0.5249, + "grad_norm": 1.8980107798145311, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 52490 + }, + { + "epoch": 0.52491, + "grad_norm": 1.2832520227119035, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 52491 + }, + { + "epoch": 0.52492, + "grad_norm": 1.8619661787638013, + "learning_rate": 0.003, + "loss": 4.039, + "step": 52492 + }, + { + "epoch": 0.52493, + "grad_norm": 1.1051580787122652, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52493 + }, + { + "epoch": 0.52494, + "grad_norm": 1.6692035741729736, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 52494 + }, + { + "epoch": 0.52495, + "grad_norm": 1.2936869614832087, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 52495 + }, + { + "epoch": 0.52496, + "grad_norm": 1.7587577301614108, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 52496 + }, + { + "epoch": 0.52497, + "grad_norm": 1.3230554488560753, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 52497 + }, + { + "epoch": 0.52498, + "grad_norm": 2.057135991360038, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 52498 + }, + { + "epoch": 0.52499, + "grad_norm": 1.6794554663230998, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 52499 + }, + { + "epoch": 0.525, + "grad_norm": 1.6447312824383615, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 52500 + }, + { + "epoch": 0.52501, + "grad_norm": 1.4943138898770916, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52501 + }, + { + "epoch": 0.52502, + "grad_norm": 1.6604120933490611, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 52502 + }, + { + "epoch": 0.52503, + "grad_norm": 1.3249756633345284, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 52503 + }, + { + "epoch": 0.52504, + "grad_norm": 1.7936274351139598, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 52504 + }, + { + "epoch": 0.52505, + "grad_norm": 1.4967195540366633, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 52505 + }, + { + "epoch": 0.52506, + "grad_norm": 1.3337251238967212, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 52506 + }, + { + "epoch": 0.52507, + "grad_norm": 1.7926937974325092, + "learning_rate": 0.003, + "loss": 4.047, + "step": 52507 + }, + { + "epoch": 0.52508, + "grad_norm": 1.4541045431502293, + "learning_rate": 0.003, + "loss": 4.042, + "step": 52508 + }, + { + "epoch": 0.52509, + "grad_norm": 1.5002588369185426, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 52509 + }, + { + "epoch": 0.5251, + "grad_norm": 1.2914620341092422, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 52510 + }, + { + "epoch": 0.52511, + "grad_norm": 1.7851991489747463, + "learning_rate": 0.003, + "loss": 4.014, + "step": 52511 + }, + { + "epoch": 0.52512, + "grad_norm": 1.6776155614671064, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 52512 + }, + { + "epoch": 0.52513, + "grad_norm": 1.6824828362548603, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 52513 + }, + { + "epoch": 0.52514, + "grad_norm": 1.4707089608836932, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 52514 + }, + { + "epoch": 0.52515, + "grad_norm": 1.4696658901072548, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52515 + }, + { + "epoch": 0.52516, + "grad_norm": 1.6037542686665953, + "learning_rate": 0.003, + "loss": 4.011, + "step": 52516 + }, + { + "epoch": 0.52517, + "grad_norm": 1.4280592006600403, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 52517 + }, + { + "epoch": 0.52518, + "grad_norm": 1.5625994900212457, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52518 + }, + { + "epoch": 0.52519, + "grad_norm": 1.3055956431501368, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 52519 + }, + { + "epoch": 0.5252, + "grad_norm": 1.7551654133551606, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52520 + }, + { + "epoch": 0.52521, + "grad_norm": 1.3785429652357912, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 52521 + }, + { + "epoch": 0.52522, + "grad_norm": 1.7689352158101752, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 52522 + }, + { + "epoch": 0.52523, + "grad_norm": 1.625705472128512, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 52523 + }, + { + "epoch": 0.52524, + "grad_norm": 1.8369264631704159, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 52524 + }, + { + "epoch": 0.52525, + "grad_norm": 1.7297324323763286, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 52525 + }, + { + "epoch": 0.52526, + "grad_norm": 1.6153383917785877, + "learning_rate": 0.003, + "loss": 4.046, + "step": 52526 + }, + { + "epoch": 0.52527, + "grad_norm": 1.8905136903622781, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52527 + }, + { + "epoch": 0.52528, + "grad_norm": 1.3758076724539412, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 52528 + }, + { + "epoch": 0.52529, + "grad_norm": 1.6417632290619433, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52529 + }, + { + "epoch": 0.5253, + "grad_norm": 1.4544679048066218, + "learning_rate": 0.003, + "loss": 3.99, + "step": 52530 + }, + { + "epoch": 0.52531, + "grad_norm": 1.6932815355708817, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 52531 + }, + { + "epoch": 0.52532, + "grad_norm": 1.628142229272678, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52532 + }, + { + "epoch": 0.52533, + "grad_norm": 1.30293073929666, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52533 + }, + { + "epoch": 0.52534, + "grad_norm": 1.585810171328354, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 52534 + }, + { + "epoch": 0.52535, + "grad_norm": 1.462179019085909, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 52535 + }, + { + "epoch": 0.52536, + "grad_norm": 1.6294719188814866, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 52536 + }, + { + "epoch": 0.52537, + "grad_norm": 1.6243879865781086, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 52537 + }, + { + "epoch": 0.52538, + "grad_norm": 1.7113228878883617, + "learning_rate": 0.003, + "loss": 4.06, + "step": 52538 + }, + { + "epoch": 0.52539, + "grad_norm": 1.452261267029837, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52539 + }, + { + "epoch": 0.5254, + "grad_norm": 1.5743968342524748, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 52540 + }, + { + "epoch": 0.52541, + "grad_norm": 1.6898866397972994, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 52541 + }, + { + "epoch": 0.52542, + "grad_norm": 1.4266417645551777, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 52542 + }, + { + "epoch": 0.52543, + "grad_norm": 1.5796085584886903, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52543 + }, + { + "epoch": 0.52544, + "grad_norm": 1.3879440375243333, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 52544 + }, + { + "epoch": 0.52545, + "grad_norm": 1.8809253568261222, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52545 + }, + { + "epoch": 0.52546, + "grad_norm": 1.374319851143915, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 52546 + }, + { + "epoch": 0.52547, + "grad_norm": 1.6014732489068813, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 52547 + }, + { + "epoch": 0.52548, + "grad_norm": 1.420334118039134, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52548 + }, + { + "epoch": 0.52549, + "grad_norm": 1.5828279540622008, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 52549 + }, + { + "epoch": 0.5255, + "grad_norm": 1.5470434454744777, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 52550 + }, + { + "epoch": 0.52551, + "grad_norm": 1.5347783699544821, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 52551 + }, + { + "epoch": 0.52552, + "grad_norm": 1.5449379673167574, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 52552 + }, + { + "epoch": 0.52553, + "grad_norm": 1.9108640819037952, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 52553 + }, + { + "epoch": 0.52554, + "grad_norm": 1.4142054337908794, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52554 + }, + { + "epoch": 0.52555, + "grad_norm": 1.6122373768877385, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 52555 + }, + { + "epoch": 0.52556, + "grad_norm": 1.5597159901473931, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 52556 + }, + { + "epoch": 0.52557, + "grad_norm": 1.3718219204046098, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52557 + }, + { + "epoch": 0.52558, + "grad_norm": 1.507942455528019, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52558 + }, + { + "epoch": 0.52559, + "grad_norm": 1.4965508521715918, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52559 + }, + { + "epoch": 0.5256, + "grad_norm": 1.4405868450379424, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52560 + }, + { + "epoch": 0.52561, + "grad_norm": 1.8539024727421904, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 52561 + }, + { + "epoch": 0.52562, + "grad_norm": 1.6157935634338287, + "learning_rate": 0.003, + "loss": 4.011, + "step": 52562 + }, + { + "epoch": 0.52563, + "grad_norm": 1.759699275487731, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 52563 + }, + { + "epoch": 0.52564, + "grad_norm": 1.3521535819342618, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 52564 + }, + { + "epoch": 0.52565, + "grad_norm": 1.8389364569306816, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 52565 + }, + { + "epoch": 0.52566, + "grad_norm": 1.6531196142592797, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 52566 + }, + { + "epoch": 0.52567, + "grad_norm": 1.3720020963109483, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 52567 + }, + { + "epoch": 0.52568, + "grad_norm": 1.7377895365224498, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 52568 + }, + { + "epoch": 0.52569, + "grad_norm": 1.396872451968783, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52569 + }, + { + "epoch": 0.5257, + "grad_norm": 1.6926801455292197, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 52570 + }, + { + "epoch": 0.52571, + "grad_norm": 1.4190105995426465, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 52571 + }, + { + "epoch": 0.52572, + "grad_norm": 1.7346687656176392, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 52572 + }, + { + "epoch": 0.52573, + "grad_norm": 1.269367805118452, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 52573 + }, + { + "epoch": 0.52574, + "grad_norm": 1.463897782818267, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52574 + }, + { + "epoch": 0.52575, + "grad_norm": 1.5508355963829885, + "learning_rate": 0.003, + "loss": 4.044, + "step": 52575 + }, + { + "epoch": 0.52576, + "grad_norm": 1.2239923154008845, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52576 + }, + { + "epoch": 0.52577, + "grad_norm": 1.548685491852905, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 52577 + }, + { + "epoch": 0.52578, + "grad_norm": 1.3083090041985619, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 52578 + }, + { + "epoch": 0.52579, + "grad_norm": 1.6021122673553383, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 52579 + }, + { + "epoch": 0.5258, + "grad_norm": 2.0290458718412747, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 52580 + }, + { + "epoch": 0.52581, + "grad_norm": 1.9191563219364551, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 52581 + }, + { + "epoch": 0.52582, + "grad_norm": 1.4289696387493003, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 52582 + }, + { + "epoch": 0.52583, + "grad_norm": 1.6933938254444645, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 52583 + }, + { + "epoch": 0.52584, + "grad_norm": 1.8293037788784454, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 52584 + }, + { + "epoch": 0.52585, + "grad_norm": 1.4395625447371188, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 52585 + }, + { + "epoch": 0.52586, + "grad_norm": 1.780330096921893, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 52586 + }, + { + "epoch": 0.52587, + "grad_norm": 1.3655770414122042, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52587 + }, + { + "epoch": 0.52588, + "grad_norm": 1.8439298011543372, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52588 + }, + { + "epoch": 0.52589, + "grad_norm": 1.3561274960865408, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52589 + }, + { + "epoch": 0.5259, + "grad_norm": 1.7589331693152588, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 52590 + }, + { + "epoch": 0.52591, + "grad_norm": 1.2626898607107824, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 52591 + }, + { + "epoch": 0.52592, + "grad_norm": 1.887867060603084, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 52592 + }, + { + "epoch": 0.52593, + "grad_norm": 1.2501272149933411, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 52593 + }, + { + "epoch": 0.52594, + "grad_norm": 1.6318611169021837, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 52594 + }, + { + "epoch": 0.52595, + "grad_norm": 1.7641648367437333, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 52595 + }, + { + "epoch": 0.52596, + "grad_norm": 1.5276197768818693, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 52596 + }, + { + "epoch": 0.52597, + "grad_norm": 1.5522415927104876, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 52597 + }, + { + "epoch": 0.52598, + "grad_norm": 1.3794704614491906, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52598 + }, + { + "epoch": 0.52599, + "grad_norm": 1.6380059192240113, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 52599 + }, + { + "epoch": 0.526, + "grad_norm": 1.362701127076073, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 52600 + }, + { + "epoch": 0.52601, + "grad_norm": 1.7382306368742901, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52601 + }, + { + "epoch": 0.52602, + "grad_norm": 1.5552293601470253, + "learning_rate": 0.003, + "loss": 4.025, + "step": 52602 + }, + { + "epoch": 0.52603, + "grad_norm": 1.5544099766425432, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 52603 + }, + { + "epoch": 0.52604, + "grad_norm": 2.021205490180084, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 52604 + }, + { + "epoch": 0.52605, + "grad_norm": 1.7046322327348058, + "learning_rate": 0.003, + "loss": 4.039, + "step": 52605 + }, + { + "epoch": 0.52606, + "grad_norm": 1.5987648207499703, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52606 + }, + { + "epoch": 0.52607, + "grad_norm": 1.4625344833426484, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 52607 + }, + { + "epoch": 0.52608, + "grad_norm": 1.579426829591367, + "learning_rate": 0.003, + "loss": 4.031, + "step": 52608 + }, + { + "epoch": 0.52609, + "grad_norm": 1.549517806030528, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 52609 + }, + { + "epoch": 0.5261, + "grad_norm": 1.4930155588375473, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 52610 + }, + { + "epoch": 0.52611, + "grad_norm": 1.5148900698709893, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 52611 + }, + { + "epoch": 0.52612, + "grad_norm": 1.5191141946073503, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52612 + }, + { + "epoch": 0.52613, + "grad_norm": 1.3850644144625268, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 52613 + }, + { + "epoch": 0.52614, + "grad_norm": 1.3328412418667344, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 52614 + }, + { + "epoch": 0.52615, + "grad_norm": 1.7658871647020558, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 52615 + }, + { + "epoch": 0.52616, + "grad_norm": 1.5863302524960707, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 52616 + }, + { + "epoch": 0.52617, + "grad_norm": 1.6007842823241913, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 52617 + }, + { + "epoch": 0.52618, + "grad_norm": 1.5718095867361341, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 52618 + }, + { + "epoch": 0.52619, + "grad_norm": 1.3846612057433374, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 52619 + }, + { + "epoch": 0.5262, + "grad_norm": 1.53426363151694, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 52620 + }, + { + "epoch": 0.52621, + "grad_norm": 1.5444678729030077, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 52621 + }, + { + "epoch": 0.52622, + "grad_norm": 1.9184821343254643, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 52622 + }, + { + "epoch": 0.52623, + "grad_norm": 1.5676579157609243, + "learning_rate": 0.003, + "loss": 3.999, + "step": 52623 + }, + { + "epoch": 0.52624, + "grad_norm": 1.5519375244643436, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52624 + }, + { + "epoch": 0.52625, + "grad_norm": 1.9374320398265905, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 52625 + }, + { + "epoch": 0.52626, + "grad_norm": 1.0951086689134546, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 52626 + }, + { + "epoch": 0.52627, + "grad_norm": 1.5695422362299294, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 52627 + }, + { + "epoch": 0.52628, + "grad_norm": 1.455938866671437, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 52628 + }, + { + "epoch": 0.52629, + "grad_norm": 1.481692949312504, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 52629 + }, + { + "epoch": 0.5263, + "grad_norm": 1.5322151874286263, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 52630 + }, + { + "epoch": 0.52631, + "grad_norm": 1.6154029725965102, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 52631 + }, + { + "epoch": 0.52632, + "grad_norm": 2.1709348317103063, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 52632 + }, + { + "epoch": 0.52633, + "grad_norm": 1.291343917952208, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 52633 + }, + { + "epoch": 0.52634, + "grad_norm": 1.7772876352447577, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 52634 + }, + { + "epoch": 0.52635, + "grad_norm": 1.5682903651320874, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52635 + }, + { + "epoch": 0.52636, + "grad_norm": 1.8678070681589172, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52636 + }, + { + "epoch": 0.52637, + "grad_norm": 1.437042581144199, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 52637 + }, + { + "epoch": 0.52638, + "grad_norm": 1.3622636441283689, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 52638 + }, + { + "epoch": 0.52639, + "grad_norm": 1.6013231301318829, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 52639 + }, + { + "epoch": 0.5264, + "grad_norm": 1.329208423786449, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 52640 + }, + { + "epoch": 0.52641, + "grad_norm": 1.402703336817092, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 52641 + }, + { + "epoch": 0.52642, + "grad_norm": 1.5306974756405696, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 52642 + }, + { + "epoch": 0.52643, + "grad_norm": 1.7551687378169438, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 52643 + }, + { + "epoch": 0.52644, + "grad_norm": 1.5131488334734207, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 52644 + }, + { + "epoch": 0.52645, + "grad_norm": 1.5860234033410836, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 52645 + }, + { + "epoch": 0.52646, + "grad_norm": 1.8756216903652554, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 52646 + }, + { + "epoch": 0.52647, + "grad_norm": 1.3748567331373818, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 52647 + }, + { + "epoch": 0.52648, + "grad_norm": 1.85960318766828, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 52648 + }, + { + "epoch": 0.52649, + "grad_norm": 1.2873636889245565, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52649 + }, + { + "epoch": 0.5265, + "grad_norm": 1.7938647059460955, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52650 + }, + { + "epoch": 0.52651, + "grad_norm": 1.4476047717301195, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 52651 + }, + { + "epoch": 0.52652, + "grad_norm": 1.6836041511169388, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52652 + }, + { + "epoch": 0.52653, + "grad_norm": 1.8738144918482609, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 52653 + }, + { + "epoch": 0.52654, + "grad_norm": 1.3639975040935892, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 52654 + }, + { + "epoch": 0.52655, + "grad_norm": 1.7844762383869361, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 52655 + }, + { + "epoch": 0.52656, + "grad_norm": 1.2875457973332074, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52656 + }, + { + "epoch": 0.52657, + "grad_norm": 1.732672591305739, + "learning_rate": 0.003, + "loss": 4.048, + "step": 52657 + }, + { + "epoch": 0.52658, + "grad_norm": 1.4080731028519369, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 52658 + }, + { + "epoch": 0.52659, + "grad_norm": 1.7224563062875928, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 52659 + }, + { + "epoch": 0.5266, + "grad_norm": 1.5727717480923487, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52660 + }, + { + "epoch": 0.52661, + "grad_norm": 1.4015233205128972, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 52661 + }, + { + "epoch": 0.52662, + "grad_norm": 1.6288761293226588, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 52662 + }, + { + "epoch": 0.52663, + "grad_norm": 1.3087545175459991, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 52663 + }, + { + "epoch": 0.52664, + "grad_norm": 1.9723230663221933, + "learning_rate": 0.003, + "loss": 4.064, + "step": 52664 + }, + { + "epoch": 0.52665, + "grad_norm": 1.2240604959703076, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 52665 + }, + { + "epoch": 0.52666, + "grad_norm": 1.8466327422574025, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 52666 + }, + { + "epoch": 0.52667, + "grad_norm": 1.4808052570497443, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 52667 + }, + { + "epoch": 0.52668, + "grad_norm": 2.419675615933188, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 52668 + }, + { + "epoch": 0.52669, + "grad_norm": 1.8600392200993134, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 52669 + }, + { + "epoch": 0.5267, + "grad_norm": 1.3649100994778207, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 52670 + }, + { + "epoch": 0.52671, + "grad_norm": 1.622252536968426, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52671 + }, + { + "epoch": 0.52672, + "grad_norm": 1.8266724674135482, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 52672 + }, + { + "epoch": 0.52673, + "grad_norm": 1.385186040268793, + "learning_rate": 0.003, + "loss": 4.04, + "step": 52673 + }, + { + "epoch": 0.52674, + "grad_norm": 1.688055918876781, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 52674 + }, + { + "epoch": 0.52675, + "grad_norm": 1.6025301330960913, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52675 + }, + { + "epoch": 0.52676, + "grad_norm": 1.4427862987587352, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 52676 + }, + { + "epoch": 0.52677, + "grad_norm": 1.6689249533873025, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 52677 + }, + { + "epoch": 0.52678, + "grad_norm": 1.5609796320057767, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 52678 + }, + { + "epoch": 0.52679, + "grad_norm": 1.5009703175085547, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 52679 + }, + { + "epoch": 0.5268, + "grad_norm": 1.8333464795692411, + "learning_rate": 0.003, + "loss": 4.032, + "step": 52680 + }, + { + "epoch": 0.52681, + "grad_norm": 1.6414069251199184, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 52681 + }, + { + "epoch": 0.52682, + "grad_norm": 1.721886833490025, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52682 + }, + { + "epoch": 0.52683, + "grad_norm": 1.5894397754515879, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52683 + }, + { + "epoch": 0.52684, + "grad_norm": 1.5039249105031587, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52684 + }, + { + "epoch": 0.52685, + "grad_norm": 1.653609281607006, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 52685 + }, + { + "epoch": 0.52686, + "grad_norm": 1.2283265043260174, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52686 + }, + { + "epoch": 0.52687, + "grad_norm": 1.6473770196620086, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 52687 + }, + { + "epoch": 0.52688, + "grad_norm": 1.3742596848049036, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 52688 + }, + { + "epoch": 0.52689, + "grad_norm": 1.9145941404626745, + "learning_rate": 0.003, + "loss": 4.049, + "step": 52689 + }, + { + "epoch": 0.5269, + "grad_norm": 1.3973273244675986, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 52690 + }, + { + "epoch": 0.52691, + "grad_norm": 1.7114097547358174, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 52691 + }, + { + "epoch": 0.52692, + "grad_norm": 1.5960169550259604, + "learning_rate": 0.003, + "loss": 4.052, + "step": 52692 + }, + { + "epoch": 0.52693, + "grad_norm": 1.435221833351532, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 52693 + }, + { + "epoch": 0.52694, + "grad_norm": 1.5889233661613258, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 52694 + }, + { + "epoch": 0.52695, + "grad_norm": 1.3689302043819938, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 52695 + }, + { + "epoch": 0.52696, + "grad_norm": 1.7948847439383757, + "learning_rate": 0.003, + "loss": 4.035, + "step": 52696 + }, + { + "epoch": 0.52697, + "grad_norm": 1.512490398601947, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52697 + }, + { + "epoch": 0.52698, + "grad_norm": 1.8069991560691245, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 52698 + }, + { + "epoch": 0.52699, + "grad_norm": 1.5083843939848087, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52699 + }, + { + "epoch": 0.527, + "grad_norm": 1.5857856020640202, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 52700 + }, + { + "epoch": 0.52701, + "grad_norm": 1.3800301741597951, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 52701 + }, + { + "epoch": 0.52702, + "grad_norm": 1.7108846963143747, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52702 + }, + { + "epoch": 0.52703, + "grad_norm": 1.5768189229471417, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 52703 + }, + { + "epoch": 0.52704, + "grad_norm": 1.9400174543134332, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 52704 + }, + { + "epoch": 0.52705, + "grad_norm": 1.2594350767265992, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 52705 + }, + { + "epoch": 0.52706, + "grad_norm": 1.7009403322591918, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 52706 + }, + { + "epoch": 0.52707, + "grad_norm": 1.2520445138021707, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52707 + }, + { + "epoch": 0.52708, + "grad_norm": 1.612386982483596, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 52708 + }, + { + "epoch": 0.52709, + "grad_norm": 1.5577954333484327, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52709 + }, + { + "epoch": 0.5271, + "grad_norm": 1.688906246430214, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 52710 + }, + { + "epoch": 0.52711, + "grad_norm": 1.5821597618079155, + "learning_rate": 0.003, + "loss": 4.04, + "step": 52711 + }, + { + "epoch": 0.52712, + "grad_norm": 1.7524459217200377, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 52712 + }, + { + "epoch": 0.52713, + "grad_norm": 1.52630293976218, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 52713 + }, + { + "epoch": 0.52714, + "grad_norm": 1.4670748486935115, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 52714 + }, + { + "epoch": 0.52715, + "grad_norm": 1.8172312108840918, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 52715 + }, + { + "epoch": 0.52716, + "grad_norm": 1.5647079108863169, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52716 + }, + { + "epoch": 0.52717, + "grad_norm": 1.7233790230846069, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 52717 + }, + { + "epoch": 0.52718, + "grad_norm": 1.524764380951554, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 52718 + }, + { + "epoch": 0.52719, + "grad_norm": 1.7367048982154312, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52719 + }, + { + "epoch": 0.5272, + "grad_norm": 1.7237481235325696, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 52720 + }, + { + "epoch": 0.52721, + "grad_norm": 1.34308389254386, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 52721 + }, + { + "epoch": 0.52722, + "grad_norm": 1.3565980033035365, + "learning_rate": 0.003, + "loss": 4.034, + "step": 52722 + }, + { + "epoch": 0.52723, + "grad_norm": 1.5050461727398774, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 52723 + }, + { + "epoch": 0.52724, + "grad_norm": 1.27416514719393, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 52724 + }, + { + "epoch": 0.52725, + "grad_norm": 1.7465828801734364, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 52725 + }, + { + "epoch": 0.52726, + "grad_norm": 1.4080694312650417, + "learning_rate": 0.003, + "loss": 3.99, + "step": 52726 + }, + { + "epoch": 0.52727, + "grad_norm": 1.8598548390946388, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 52727 + }, + { + "epoch": 0.52728, + "grad_norm": 1.4551315251605945, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 52728 + }, + { + "epoch": 0.52729, + "grad_norm": 1.629258588762332, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52729 + }, + { + "epoch": 0.5273, + "grad_norm": 1.6451938877168861, + "learning_rate": 0.003, + "loss": 4.026, + "step": 52730 + }, + { + "epoch": 0.52731, + "grad_norm": 1.4463943200122795, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 52731 + }, + { + "epoch": 0.52732, + "grad_norm": 1.5810257148183116, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52732 + }, + { + "epoch": 0.52733, + "grad_norm": 1.4609453668073615, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52733 + }, + { + "epoch": 0.52734, + "grad_norm": 1.418326896879269, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 52734 + }, + { + "epoch": 0.52735, + "grad_norm": 1.7320625956942064, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 52735 + }, + { + "epoch": 0.52736, + "grad_norm": 1.3091662426665256, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 52736 + }, + { + "epoch": 0.52737, + "grad_norm": 1.555812179990933, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 52737 + }, + { + "epoch": 0.52738, + "grad_norm": 1.771469376353081, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 52738 + }, + { + "epoch": 0.52739, + "grad_norm": 1.832335025575245, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 52739 + }, + { + "epoch": 0.5274, + "grad_norm": 1.344625567819817, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 52740 + }, + { + "epoch": 0.52741, + "grad_norm": 1.8862140200804811, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 52741 + }, + { + "epoch": 0.52742, + "grad_norm": 1.7477590291938672, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 52742 + }, + { + "epoch": 0.52743, + "grad_norm": 1.4003949201938002, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52743 + }, + { + "epoch": 0.52744, + "grad_norm": 1.5930386751856, + "learning_rate": 0.003, + "loss": 4.037, + "step": 52744 + }, + { + "epoch": 0.52745, + "grad_norm": 1.5037912950489576, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 52745 + }, + { + "epoch": 0.52746, + "grad_norm": 1.6476964402501828, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 52746 + }, + { + "epoch": 0.52747, + "grad_norm": 1.2919588053880282, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 52747 + }, + { + "epoch": 0.52748, + "grad_norm": 1.6502738324476327, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 52748 + }, + { + "epoch": 0.52749, + "grad_norm": 1.6020792637134864, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 52749 + }, + { + "epoch": 0.5275, + "grad_norm": 1.6684772527384808, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 52750 + }, + { + "epoch": 0.52751, + "grad_norm": 1.510406204002577, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 52751 + }, + { + "epoch": 0.52752, + "grad_norm": 1.7235236994301928, + "learning_rate": 0.003, + "loss": 4.051, + "step": 52752 + }, + { + "epoch": 0.52753, + "grad_norm": 1.5634133560655152, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 52753 + }, + { + "epoch": 0.52754, + "grad_norm": 1.6476415417336867, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 52754 + }, + { + "epoch": 0.52755, + "grad_norm": 1.593238276456077, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 52755 + }, + { + "epoch": 0.52756, + "grad_norm": 1.5495321902451673, + "learning_rate": 0.003, + "loss": 4.034, + "step": 52756 + }, + { + "epoch": 0.52757, + "grad_norm": 1.7085778359733925, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 52757 + }, + { + "epoch": 0.52758, + "grad_norm": 1.6328394552340586, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 52758 + }, + { + "epoch": 0.52759, + "grad_norm": 1.342504476720062, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 52759 + }, + { + "epoch": 0.5276, + "grad_norm": 1.609325365239786, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 52760 + }, + { + "epoch": 0.52761, + "grad_norm": 1.739544595164986, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 52761 + }, + { + "epoch": 0.52762, + "grad_norm": 1.3712519647000936, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 52762 + }, + { + "epoch": 0.52763, + "grad_norm": 1.577499746319492, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 52763 + }, + { + "epoch": 0.52764, + "grad_norm": 1.447023320042865, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 52764 + }, + { + "epoch": 0.52765, + "grad_norm": 1.7299362680969854, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 52765 + }, + { + "epoch": 0.52766, + "grad_norm": 1.528099754203192, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 52766 + }, + { + "epoch": 0.52767, + "grad_norm": 1.5408444545582558, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 52767 + }, + { + "epoch": 0.52768, + "grad_norm": 1.8911002968404709, + "learning_rate": 0.003, + "loss": 4.015, + "step": 52768 + }, + { + "epoch": 0.52769, + "grad_norm": 1.4778714079807826, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 52769 + }, + { + "epoch": 0.5277, + "grad_norm": 1.3973263822355702, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 52770 + }, + { + "epoch": 0.52771, + "grad_norm": 1.6158720902837154, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 52771 + }, + { + "epoch": 0.52772, + "grad_norm": 1.4677101138540964, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52772 + }, + { + "epoch": 0.52773, + "grad_norm": 1.4771748285539403, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 52773 + }, + { + "epoch": 0.52774, + "grad_norm": 1.3635524568513677, + "learning_rate": 0.003, + "loss": 4.078, + "step": 52774 + }, + { + "epoch": 0.52775, + "grad_norm": 1.7800609528524132, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 52775 + }, + { + "epoch": 0.52776, + "grad_norm": 1.7387906174241943, + "learning_rate": 0.003, + "loss": 3.999, + "step": 52776 + }, + { + "epoch": 0.52777, + "grad_norm": 1.7230059710855798, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 52777 + }, + { + "epoch": 0.52778, + "grad_norm": 1.4036275203592692, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 52778 + }, + { + "epoch": 0.52779, + "grad_norm": 1.6027170510120916, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52779 + }, + { + "epoch": 0.5278, + "grad_norm": 1.7053358372733591, + "learning_rate": 0.003, + "loss": 4.047, + "step": 52780 + }, + { + "epoch": 0.52781, + "grad_norm": 1.5408400714659627, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 52781 + }, + { + "epoch": 0.52782, + "grad_norm": 1.3558103921258469, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 52782 + }, + { + "epoch": 0.52783, + "grad_norm": 1.6592982567891137, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 52783 + }, + { + "epoch": 0.52784, + "grad_norm": 1.347482054309172, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 52784 + }, + { + "epoch": 0.52785, + "grad_norm": 1.79370660881329, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 52785 + }, + { + "epoch": 0.52786, + "grad_norm": 1.3353357588366257, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52786 + }, + { + "epoch": 0.52787, + "grad_norm": 1.764139552228244, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 52787 + }, + { + "epoch": 0.52788, + "grad_norm": 1.4477226903103881, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 52788 + }, + { + "epoch": 0.52789, + "grad_norm": 1.7375865664774368, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 52789 + }, + { + "epoch": 0.5279, + "grad_norm": 1.4666363909183742, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52790 + }, + { + "epoch": 0.52791, + "grad_norm": 1.6096220890048394, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 52791 + }, + { + "epoch": 0.52792, + "grad_norm": 1.9047183413791067, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 52792 + }, + { + "epoch": 0.52793, + "grad_norm": 1.6213297222218177, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 52793 + }, + { + "epoch": 0.52794, + "grad_norm": 1.5547369167025424, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 52794 + }, + { + "epoch": 0.52795, + "grad_norm": 1.6147837245744503, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 52795 + }, + { + "epoch": 0.52796, + "grad_norm": 1.6168347081489989, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 52796 + }, + { + "epoch": 0.52797, + "grad_norm": 1.3273165787040384, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 52797 + }, + { + "epoch": 0.52798, + "grad_norm": 1.7127901154962586, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 52798 + }, + { + "epoch": 0.52799, + "grad_norm": 1.5638130315484537, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 52799 + }, + { + "epoch": 0.528, + "grad_norm": 1.3604389134583996, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52800 + }, + { + "epoch": 0.52801, + "grad_norm": 1.4307385525955405, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52801 + }, + { + "epoch": 0.52802, + "grad_norm": 1.4885226738947754, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 52802 + }, + { + "epoch": 0.52803, + "grad_norm": 1.629858923162539, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 52803 + }, + { + "epoch": 0.52804, + "grad_norm": 1.4059661299928152, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 52804 + }, + { + "epoch": 0.52805, + "grad_norm": 1.7987597818466057, + "learning_rate": 0.003, + "loss": 4.012, + "step": 52805 + }, + { + "epoch": 0.52806, + "grad_norm": 1.9006301323525308, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52806 + }, + { + "epoch": 0.52807, + "grad_norm": 1.3774528635529222, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52807 + }, + { + "epoch": 0.52808, + "grad_norm": 1.8971035577384925, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 52808 + }, + { + "epoch": 0.52809, + "grad_norm": 1.485422571182299, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 52809 + }, + { + "epoch": 0.5281, + "grad_norm": 1.5617943869221786, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 52810 + }, + { + "epoch": 0.52811, + "grad_norm": 1.4793121365857693, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 52811 + }, + { + "epoch": 0.52812, + "grad_norm": 1.7973803734902665, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 52812 + }, + { + "epoch": 0.52813, + "grad_norm": 1.2191941192732865, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 52813 + }, + { + "epoch": 0.52814, + "grad_norm": 1.798756682623382, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 52814 + }, + { + "epoch": 0.52815, + "grad_norm": 1.2179603772507683, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 52815 + }, + { + "epoch": 0.52816, + "grad_norm": 1.6120220153326175, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 52816 + }, + { + "epoch": 0.52817, + "grad_norm": 1.508764647805438, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 52817 + }, + { + "epoch": 0.52818, + "grad_norm": 1.4321039603531447, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 52818 + }, + { + "epoch": 0.52819, + "grad_norm": 1.6005087787599357, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52819 + }, + { + "epoch": 0.5282, + "grad_norm": 1.442216638000718, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 52820 + }, + { + "epoch": 0.52821, + "grad_norm": 2.263957743685405, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 52821 + }, + { + "epoch": 0.52822, + "grad_norm": 1.256541050514924, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 52822 + }, + { + "epoch": 0.52823, + "grad_norm": 1.5347496911552732, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 52823 + }, + { + "epoch": 0.52824, + "grad_norm": 1.8065176885705623, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 52824 + }, + { + "epoch": 0.52825, + "grad_norm": 1.6544354741945029, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 52825 + }, + { + "epoch": 0.52826, + "grad_norm": 1.4235368304960456, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 52826 + }, + { + "epoch": 0.52827, + "grad_norm": 1.5693594333435976, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 52827 + }, + { + "epoch": 0.52828, + "grad_norm": 1.6623542110998988, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 52828 + }, + { + "epoch": 0.52829, + "grad_norm": 1.4379390237171952, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 52829 + }, + { + "epoch": 0.5283, + "grad_norm": 1.9680521892880813, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 52830 + }, + { + "epoch": 0.52831, + "grad_norm": 1.3109093894717259, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 52831 + }, + { + "epoch": 0.52832, + "grad_norm": 1.8064125273884046, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 52832 + }, + { + "epoch": 0.52833, + "grad_norm": 1.6480208080148038, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 52833 + }, + { + "epoch": 0.52834, + "grad_norm": 1.2885528362542469, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 52834 + }, + { + "epoch": 0.52835, + "grad_norm": 1.7941526046608989, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 52835 + }, + { + "epoch": 0.52836, + "grad_norm": 1.6723934831238956, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 52836 + }, + { + "epoch": 0.52837, + "grad_norm": 1.1270104943258807, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52837 + }, + { + "epoch": 0.52838, + "grad_norm": 1.9311977653950299, + "learning_rate": 0.003, + "loss": 4.068, + "step": 52838 + }, + { + "epoch": 0.52839, + "grad_norm": 1.1373222279781396, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 52839 + }, + { + "epoch": 0.5284, + "grad_norm": 1.7559919340176686, + "learning_rate": 0.003, + "loss": 4.036, + "step": 52840 + }, + { + "epoch": 0.52841, + "grad_norm": 1.5375769642045574, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 52841 + }, + { + "epoch": 0.52842, + "grad_norm": 1.6782675540281828, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 52842 + }, + { + "epoch": 0.52843, + "grad_norm": 1.5131646570530344, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52843 + }, + { + "epoch": 0.52844, + "grad_norm": 1.9098672808703232, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 52844 + }, + { + "epoch": 0.52845, + "grad_norm": 1.563207244457398, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 52845 + }, + { + "epoch": 0.52846, + "grad_norm": 1.7175087802851632, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52846 + }, + { + "epoch": 0.52847, + "grad_norm": 1.611161326818101, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52847 + }, + { + "epoch": 0.52848, + "grad_norm": 1.336546741392157, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 52848 + }, + { + "epoch": 0.52849, + "grad_norm": 1.350601603445575, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52849 + }, + { + "epoch": 0.5285, + "grad_norm": 1.4714164597273676, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 52850 + }, + { + "epoch": 0.52851, + "grad_norm": 1.7223044648753676, + "learning_rate": 0.003, + "loss": 4.029, + "step": 52851 + }, + { + "epoch": 0.52852, + "grad_norm": 1.4938224153608288, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 52852 + }, + { + "epoch": 0.52853, + "grad_norm": 1.3183373158214566, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 52853 + }, + { + "epoch": 0.52854, + "grad_norm": 1.6136348742459956, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 52854 + }, + { + "epoch": 0.52855, + "grad_norm": 1.5241420711233813, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 52855 + }, + { + "epoch": 0.52856, + "grad_norm": 1.4290283568817506, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 52856 + }, + { + "epoch": 0.52857, + "grad_norm": 1.7710844322608996, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 52857 + }, + { + "epoch": 0.52858, + "grad_norm": 1.3531621790532244, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52858 + }, + { + "epoch": 0.52859, + "grad_norm": 1.9117169933591, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 52859 + }, + { + "epoch": 0.5286, + "grad_norm": 1.5638598346111017, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 52860 + }, + { + "epoch": 0.52861, + "grad_norm": 1.7226468156453867, + "learning_rate": 0.003, + "loss": 4.026, + "step": 52861 + }, + { + "epoch": 0.52862, + "grad_norm": 1.9333094916055746, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 52862 + }, + { + "epoch": 0.52863, + "grad_norm": 1.1249327493773593, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 52863 + }, + { + "epoch": 0.52864, + "grad_norm": 1.9415227999786067, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 52864 + }, + { + "epoch": 0.52865, + "grad_norm": 1.5899107882344568, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 52865 + }, + { + "epoch": 0.52866, + "grad_norm": 1.5182813362265384, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 52866 + }, + { + "epoch": 0.52867, + "grad_norm": 1.5630950771022236, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 52867 + }, + { + "epoch": 0.52868, + "grad_norm": 1.7458326282278784, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 52868 + }, + { + "epoch": 0.52869, + "grad_norm": 1.522573782575932, + "learning_rate": 0.003, + "loss": 4.011, + "step": 52869 + }, + { + "epoch": 0.5287, + "grad_norm": 1.2963600391127914, + "learning_rate": 0.003, + "loss": 4.01, + "step": 52870 + }, + { + "epoch": 0.52871, + "grad_norm": 1.690175105753996, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 52871 + }, + { + "epoch": 0.52872, + "grad_norm": 1.8675832933336651, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 52872 + }, + { + "epoch": 0.52873, + "grad_norm": 1.3686876176369025, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 52873 + }, + { + "epoch": 0.52874, + "grad_norm": 1.6376531441744544, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 52874 + }, + { + "epoch": 0.52875, + "grad_norm": 1.2730068254238225, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 52875 + }, + { + "epoch": 0.52876, + "grad_norm": 1.747446710118621, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 52876 + }, + { + "epoch": 0.52877, + "grad_norm": 1.204446053200769, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 52877 + }, + { + "epoch": 0.52878, + "grad_norm": 2.0577468866651447, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 52878 + }, + { + "epoch": 0.52879, + "grad_norm": 1.1602530384676943, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 52879 + }, + { + "epoch": 0.5288, + "grad_norm": 1.942046000537706, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52880 + }, + { + "epoch": 0.52881, + "grad_norm": 1.9304470045674638, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 52881 + }, + { + "epoch": 0.52882, + "grad_norm": 1.7333514454219612, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 52882 + }, + { + "epoch": 0.52883, + "grad_norm": 1.2524257844397693, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 52883 + }, + { + "epoch": 0.52884, + "grad_norm": 1.968620672685957, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 52884 + }, + { + "epoch": 0.52885, + "grad_norm": 1.757510582030265, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 52885 + }, + { + "epoch": 0.52886, + "grad_norm": 1.374075184095485, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 52886 + }, + { + "epoch": 0.52887, + "grad_norm": 1.832080625224176, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52887 + }, + { + "epoch": 0.52888, + "grad_norm": 1.6061774820026449, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 52888 + }, + { + "epoch": 0.52889, + "grad_norm": 1.141977859898051, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 52889 + }, + { + "epoch": 0.5289, + "grad_norm": 1.7853089837133147, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 52890 + }, + { + "epoch": 0.52891, + "grad_norm": 1.3431405810881818, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 52891 + }, + { + "epoch": 0.52892, + "grad_norm": 1.4881328367607842, + "learning_rate": 0.003, + "loss": 4.02, + "step": 52892 + }, + { + "epoch": 0.52893, + "grad_norm": 1.595711940416054, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 52893 + }, + { + "epoch": 0.52894, + "grad_norm": 1.6332672241094734, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 52894 + }, + { + "epoch": 0.52895, + "grad_norm": 1.4321680077203558, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 52895 + }, + { + "epoch": 0.52896, + "grad_norm": 1.6735541601421875, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 52896 + }, + { + "epoch": 0.52897, + "grad_norm": 1.7058615298045254, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 52897 + }, + { + "epoch": 0.52898, + "grad_norm": 1.4523219493774986, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 52898 + }, + { + "epoch": 0.52899, + "grad_norm": 1.6089174467693805, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 52899 + }, + { + "epoch": 0.529, + "grad_norm": 1.4387631225177715, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 52900 + }, + { + "epoch": 0.52901, + "grad_norm": 1.4780325584726584, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 52901 + }, + { + "epoch": 0.52902, + "grad_norm": 1.8719446511989377, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 52902 + }, + { + "epoch": 0.52903, + "grad_norm": 1.158682495718179, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 52903 + }, + { + "epoch": 0.52904, + "grad_norm": 1.7557420510660653, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 52904 + }, + { + "epoch": 0.52905, + "grad_norm": 1.5481579997041104, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 52905 + }, + { + "epoch": 0.52906, + "grad_norm": 1.5770736936164198, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52906 + }, + { + "epoch": 0.52907, + "grad_norm": 1.6533155197666323, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 52907 + }, + { + "epoch": 0.52908, + "grad_norm": 1.2922787457128495, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 52908 + }, + { + "epoch": 0.52909, + "grad_norm": 1.8396614523806358, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 52909 + }, + { + "epoch": 0.5291, + "grad_norm": 1.4749871861710697, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 52910 + }, + { + "epoch": 0.52911, + "grad_norm": 1.7611562217091667, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 52911 + }, + { + "epoch": 0.52912, + "grad_norm": 1.821386222083833, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 52912 + }, + { + "epoch": 0.52913, + "grad_norm": 1.3549365896954904, + "learning_rate": 0.003, + "loss": 4.033, + "step": 52913 + }, + { + "epoch": 0.52914, + "grad_norm": 1.857059118197674, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 52914 + }, + { + "epoch": 0.52915, + "grad_norm": 1.326494890886855, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 52915 + }, + { + "epoch": 0.52916, + "grad_norm": 1.5008722760509197, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 52916 + }, + { + "epoch": 0.52917, + "grad_norm": 1.4187869568562368, + "learning_rate": 0.003, + "loss": 4.045, + "step": 52917 + }, + { + "epoch": 0.52918, + "grad_norm": 1.671946171016072, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52918 + }, + { + "epoch": 0.52919, + "grad_norm": 1.5647248396429416, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 52919 + }, + { + "epoch": 0.5292, + "grad_norm": 1.5550101421526192, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 52920 + }, + { + "epoch": 0.52921, + "grad_norm": 1.3130884540478875, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 52921 + }, + { + "epoch": 0.52922, + "grad_norm": 1.5802235997467085, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52922 + }, + { + "epoch": 0.52923, + "grad_norm": 1.2638079602511132, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 52923 + }, + { + "epoch": 0.52924, + "grad_norm": 1.6900073155011714, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 52924 + }, + { + "epoch": 0.52925, + "grad_norm": 1.2803521848507589, + "learning_rate": 0.003, + "loss": 4.027, + "step": 52925 + }, + { + "epoch": 0.52926, + "grad_norm": 1.7022856735397927, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 52926 + }, + { + "epoch": 0.52927, + "grad_norm": 1.4363550288793498, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 52927 + }, + { + "epoch": 0.52928, + "grad_norm": 1.7535176245173336, + "learning_rate": 0.003, + "loss": 4.037, + "step": 52928 + }, + { + "epoch": 0.52929, + "grad_norm": 1.4490108129165946, + "learning_rate": 0.003, + "loss": 4.022, + "step": 52929 + }, + { + "epoch": 0.5293, + "grad_norm": 1.8774495307184664, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 52930 + }, + { + "epoch": 0.52931, + "grad_norm": 1.502121902326377, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 52931 + }, + { + "epoch": 0.52932, + "grad_norm": 1.6489012425187861, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 52932 + }, + { + "epoch": 0.52933, + "grad_norm": 1.7864214961654943, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52933 + }, + { + "epoch": 0.52934, + "grad_norm": 1.3280974634793243, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 52934 + }, + { + "epoch": 0.52935, + "grad_norm": 1.9139149049275224, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 52935 + }, + { + "epoch": 0.52936, + "grad_norm": 1.2023060770538312, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 52936 + }, + { + "epoch": 0.52937, + "grad_norm": 1.971038481406677, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 52937 + }, + { + "epoch": 0.52938, + "grad_norm": 1.1033337379373356, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 52938 + }, + { + "epoch": 0.52939, + "grad_norm": 1.75948093282032, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 52939 + }, + { + "epoch": 0.5294, + "grad_norm": 1.67916879652247, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 52940 + }, + { + "epoch": 0.52941, + "grad_norm": 1.905973242803676, + "learning_rate": 0.003, + "loss": 4.06, + "step": 52941 + }, + { + "epoch": 0.52942, + "grad_norm": 1.6639481786733576, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 52942 + }, + { + "epoch": 0.52943, + "grad_norm": 1.5479066537936044, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 52943 + }, + { + "epoch": 0.52944, + "grad_norm": 1.4437386882338121, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 52944 + }, + { + "epoch": 0.52945, + "grad_norm": 1.8040387157989557, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 52945 + }, + { + "epoch": 0.52946, + "grad_norm": 1.4362296640159176, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 52946 + }, + { + "epoch": 0.52947, + "grad_norm": 1.757859284860353, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 52947 + }, + { + "epoch": 0.52948, + "grad_norm": 1.6609097993367434, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 52948 + }, + { + "epoch": 0.52949, + "grad_norm": 1.2315414511490128, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52949 + }, + { + "epoch": 0.5295, + "grad_norm": 1.4074231034306444, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 52950 + }, + { + "epoch": 0.52951, + "grad_norm": 1.8504726344647198, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 52951 + }, + { + "epoch": 0.52952, + "grad_norm": 1.4649675677154232, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 52952 + }, + { + "epoch": 0.52953, + "grad_norm": 1.6472096544713448, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 52953 + }, + { + "epoch": 0.52954, + "grad_norm": 1.3582356157607087, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 52954 + }, + { + "epoch": 0.52955, + "grad_norm": 1.578432840593053, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 52955 + }, + { + "epoch": 0.52956, + "grad_norm": 1.5314705549392125, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 52956 + }, + { + "epoch": 0.52957, + "grad_norm": 1.563546286902707, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 52957 + }, + { + "epoch": 0.52958, + "grad_norm": 1.3181930117847194, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 52958 + }, + { + "epoch": 0.52959, + "grad_norm": 1.2986229745938025, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 52959 + }, + { + "epoch": 0.5296, + "grad_norm": 1.6273928658348658, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 52960 + }, + { + "epoch": 0.52961, + "grad_norm": 1.2280039262410518, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 52961 + }, + { + "epoch": 0.52962, + "grad_norm": 1.8343770582708225, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 52962 + }, + { + "epoch": 0.52963, + "grad_norm": 1.9222440443225233, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 52963 + }, + { + "epoch": 0.52964, + "grad_norm": 1.4821384954866208, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 52964 + }, + { + "epoch": 0.52965, + "grad_norm": 1.917971551209818, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 52965 + }, + { + "epoch": 0.52966, + "grad_norm": 1.735733070605013, + "learning_rate": 0.003, + "loss": 4.028, + "step": 52966 + }, + { + "epoch": 0.52967, + "grad_norm": 1.4266064322396348, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 52967 + }, + { + "epoch": 0.52968, + "grad_norm": 1.4746676185913032, + "learning_rate": 0.003, + "loss": 4.054, + "step": 52968 + }, + { + "epoch": 0.52969, + "grad_norm": 1.9125608725123222, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 52969 + }, + { + "epoch": 0.5297, + "grad_norm": 1.344288932959556, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 52970 + }, + { + "epoch": 0.52971, + "grad_norm": 1.567488391444039, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 52971 + }, + { + "epoch": 0.52972, + "grad_norm": 1.7676256335540728, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 52972 + }, + { + "epoch": 0.52973, + "grad_norm": 1.6561342549010496, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 52973 + }, + { + "epoch": 0.52974, + "grad_norm": 1.4743354105758095, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 52974 + }, + { + "epoch": 0.52975, + "grad_norm": 1.6219904532094587, + "learning_rate": 0.003, + "loss": 4.054, + "step": 52975 + }, + { + "epoch": 0.52976, + "grad_norm": 1.4128059871078433, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 52976 + }, + { + "epoch": 0.52977, + "grad_norm": 1.519612345923186, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 52977 + }, + { + "epoch": 0.52978, + "grad_norm": 1.473862060433202, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 52978 + }, + { + "epoch": 0.52979, + "grad_norm": 1.4434411834992296, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 52979 + }, + { + "epoch": 0.5298, + "grad_norm": 1.8591085465253114, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 52980 + }, + { + "epoch": 0.52981, + "grad_norm": 1.733957716092651, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 52981 + }, + { + "epoch": 0.52982, + "grad_norm": 1.4961634185631947, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 52982 + }, + { + "epoch": 0.52983, + "grad_norm": 1.7763906306204438, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 52983 + }, + { + "epoch": 0.52984, + "grad_norm": 1.5165616562997983, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 52984 + }, + { + "epoch": 0.52985, + "grad_norm": 1.507839202418934, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 52985 + }, + { + "epoch": 0.52986, + "grad_norm": 1.4770257462849217, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 52986 + }, + { + "epoch": 0.52987, + "grad_norm": 1.523560936403895, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 52987 + }, + { + "epoch": 0.52988, + "grad_norm": 1.549359073666075, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 52988 + }, + { + "epoch": 0.52989, + "grad_norm": 1.7319583093500586, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 52989 + }, + { + "epoch": 0.5299, + "grad_norm": 1.1570336371363266, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 52990 + }, + { + "epoch": 0.52991, + "grad_norm": 1.868007020972576, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 52991 + }, + { + "epoch": 0.52992, + "grad_norm": 1.848593892464678, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 52992 + }, + { + "epoch": 0.52993, + "grad_norm": 1.27794897254665, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 52993 + }, + { + "epoch": 0.52994, + "grad_norm": 1.636744676036931, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 52994 + }, + { + "epoch": 0.52995, + "grad_norm": 1.7450055539174112, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 52995 + }, + { + "epoch": 0.52996, + "grad_norm": 1.3384314488433415, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 52996 + }, + { + "epoch": 0.52997, + "grad_norm": 1.541925707109788, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 52997 + }, + { + "epoch": 0.52998, + "grad_norm": 1.4196961484082478, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 52998 + }, + { + "epoch": 0.52999, + "grad_norm": 1.2329145864428364, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 52999 + }, + { + "epoch": 0.53, + "grad_norm": 1.6271781307794078, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 53000 + }, + { + "epoch": 0.53001, + "grad_norm": 1.7709563790506275, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 53001 + }, + { + "epoch": 0.53002, + "grad_norm": 1.4742063572332893, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 53002 + }, + { + "epoch": 0.53003, + "grad_norm": 1.7079660663969438, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 53003 + }, + { + "epoch": 0.53004, + "grad_norm": 1.55437473729301, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 53004 + }, + { + "epoch": 0.53005, + "grad_norm": 1.6970519217562443, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 53005 + }, + { + "epoch": 0.53006, + "grad_norm": 1.4429568296493955, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 53006 + }, + { + "epoch": 0.53007, + "grad_norm": 1.6493356865361266, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 53007 + }, + { + "epoch": 0.53008, + "grad_norm": 1.455488375613173, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 53008 + }, + { + "epoch": 0.53009, + "grad_norm": 1.606463569894655, + "learning_rate": 0.003, + "loss": 3.993, + "step": 53009 + }, + { + "epoch": 0.5301, + "grad_norm": 1.401241447072866, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 53010 + }, + { + "epoch": 0.53011, + "grad_norm": 1.5907302707472926, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 53011 + }, + { + "epoch": 0.53012, + "grad_norm": 1.3153662872899987, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 53012 + }, + { + "epoch": 0.53013, + "grad_norm": 1.4700513959376234, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 53013 + }, + { + "epoch": 0.53014, + "grad_norm": 1.571756106470037, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 53014 + }, + { + "epoch": 0.53015, + "grad_norm": 1.5471505076454057, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 53015 + }, + { + "epoch": 0.53016, + "grad_norm": 1.6565959704408206, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 53016 + }, + { + "epoch": 0.53017, + "grad_norm": 1.6261987244974896, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 53017 + }, + { + "epoch": 0.53018, + "grad_norm": 1.4063841468674525, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 53018 + }, + { + "epoch": 0.53019, + "grad_norm": 1.7064054308905396, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 53019 + }, + { + "epoch": 0.5302, + "grad_norm": 1.9213859266936488, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 53020 + }, + { + "epoch": 0.53021, + "grad_norm": 1.7249479846988613, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 53021 + }, + { + "epoch": 0.53022, + "grad_norm": 1.5498684460579977, + "learning_rate": 0.003, + "loss": 4.041, + "step": 53022 + }, + { + "epoch": 0.53023, + "grad_norm": 1.5776264978255135, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 53023 + }, + { + "epoch": 0.53024, + "grad_norm": 1.4531980182375621, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 53024 + }, + { + "epoch": 0.53025, + "grad_norm": 1.4653451970566236, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 53025 + }, + { + "epoch": 0.53026, + "grad_norm": 1.9418396638203985, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 53026 + }, + { + "epoch": 0.53027, + "grad_norm": 1.1709459435355112, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 53027 + }, + { + "epoch": 0.53028, + "grad_norm": 1.8907669312681272, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 53028 + }, + { + "epoch": 0.53029, + "grad_norm": 1.3677920970285071, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 53029 + }, + { + "epoch": 0.5303, + "grad_norm": 1.7157801297591657, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 53030 + }, + { + "epoch": 0.53031, + "grad_norm": 1.5267841817409307, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 53031 + }, + { + "epoch": 0.53032, + "grad_norm": 2.0987420761377593, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 53032 + }, + { + "epoch": 0.53033, + "grad_norm": 1.635747828775945, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 53033 + }, + { + "epoch": 0.53034, + "grad_norm": 1.5427333821193365, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 53034 + }, + { + "epoch": 0.53035, + "grad_norm": 1.4554236724743674, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 53035 + }, + { + "epoch": 0.53036, + "grad_norm": 1.7920678998521473, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 53036 + }, + { + "epoch": 0.53037, + "grad_norm": 1.2667793053437324, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 53037 + }, + { + "epoch": 0.53038, + "grad_norm": 2.009506379608158, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 53038 + }, + { + "epoch": 0.53039, + "grad_norm": 1.3847481456309612, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 53039 + }, + { + "epoch": 0.5304, + "grad_norm": 1.4493182826460846, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 53040 + }, + { + "epoch": 0.53041, + "grad_norm": 1.6130897693692516, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 53041 + }, + { + "epoch": 0.53042, + "grad_norm": 1.3397205487486896, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 53042 + }, + { + "epoch": 0.53043, + "grad_norm": 2.0065783211165233, + "learning_rate": 0.003, + "loss": 4.037, + "step": 53043 + }, + { + "epoch": 0.53044, + "grad_norm": 1.7765991329613595, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 53044 + }, + { + "epoch": 0.53045, + "grad_norm": 1.7585325515305679, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 53045 + }, + { + "epoch": 0.53046, + "grad_norm": 1.3207963119001613, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 53046 + }, + { + "epoch": 0.53047, + "grad_norm": 1.835687495597899, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 53047 + }, + { + "epoch": 0.53048, + "grad_norm": 1.4349378667551886, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 53048 + }, + { + "epoch": 0.53049, + "grad_norm": 1.5413719138849369, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53049 + }, + { + "epoch": 0.5305, + "grad_norm": 1.4620687172535625, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 53050 + }, + { + "epoch": 0.53051, + "grad_norm": 1.9287988276380073, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 53051 + }, + { + "epoch": 0.53052, + "grad_norm": 1.0922895901614573, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 53052 + }, + { + "epoch": 0.53053, + "grad_norm": 1.7852373776607748, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 53053 + }, + { + "epoch": 0.53054, + "grad_norm": 1.4132046122333413, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 53054 + }, + { + "epoch": 0.53055, + "grad_norm": 1.6458791866229745, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 53055 + }, + { + "epoch": 0.53056, + "grad_norm": 1.5156511385943108, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 53056 + }, + { + "epoch": 0.53057, + "grad_norm": 2.114903597473215, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 53057 + }, + { + "epoch": 0.53058, + "grad_norm": 2.2661118721209865, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 53058 + }, + { + "epoch": 0.53059, + "grad_norm": 1.3798533156724286, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 53059 + }, + { + "epoch": 0.5306, + "grad_norm": 1.7998349230166588, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 53060 + }, + { + "epoch": 0.53061, + "grad_norm": 1.7149147906910702, + "learning_rate": 0.003, + "loss": 4.067, + "step": 53061 + }, + { + "epoch": 0.53062, + "grad_norm": 1.9124308913979844, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 53062 + }, + { + "epoch": 0.53063, + "grad_norm": 2.41621313739244, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 53063 + }, + { + "epoch": 0.53064, + "grad_norm": 1.8039497557671376, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 53064 + }, + { + "epoch": 0.53065, + "grad_norm": 1.4005455717983382, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 53065 + }, + { + "epoch": 0.53066, + "grad_norm": 1.942828029766617, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 53066 + }, + { + "epoch": 0.53067, + "grad_norm": 1.6201023213860553, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 53067 + }, + { + "epoch": 0.53068, + "grad_norm": 1.517551310741863, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 53068 + }, + { + "epoch": 0.53069, + "grad_norm": 2.05397488538995, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 53069 + }, + { + "epoch": 0.5307, + "grad_norm": 1.5561478771283364, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 53070 + }, + { + "epoch": 0.53071, + "grad_norm": 1.4136317807272691, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 53071 + }, + { + "epoch": 0.53072, + "grad_norm": 1.599524329556552, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 53072 + }, + { + "epoch": 0.53073, + "grad_norm": 1.7808836561715946, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 53073 + }, + { + "epoch": 0.53074, + "grad_norm": 1.5201541817343263, + "learning_rate": 0.003, + "loss": 4.062, + "step": 53074 + }, + { + "epoch": 0.53075, + "grad_norm": 1.7822057194645857, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 53075 + }, + { + "epoch": 0.53076, + "grad_norm": 1.6305991827162742, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 53076 + }, + { + "epoch": 0.53077, + "grad_norm": 1.7371083121072939, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 53077 + }, + { + "epoch": 0.53078, + "grad_norm": 1.6155925786942615, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 53078 + }, + { + "epoch": 0.53079, + "grad_norm": 1.9178798579505896, + "learning_rate": 0.003, + "loss": 4.087, + "step": 53079 + }, + { + "epoch": 0.5308, + "grad_norm": 1.9219969301296116, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 53080 + }, + { + "epoch": 0.53081, + "grad_norm": 1.7143965736413478, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 53081 + }, + { + "epoch": 0.53082, + "grad_norm": 1.7292693379865223, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 53082 + }, + { + "epoch": 0.53083, + "grad_norm": 1.9109131071613095, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 53083 + }, + { + "epoch": 0.53084, + "grad_norm": 1.6678738960545645, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 53084 + }, + { + "epoch": 0.53085, + "grad_norm": 1.2282485085321846, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 53085 + }, + { + "epoch": 0.53086, + "grad_norm": 2.0155645355146903, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 53086 + }, + { + "epoch": 0.53087, + "grad_norm": 1.7430976163456293, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 53087 + }, + { + "epoch": 0.53088, + "grad_norm": 1.4496245057968258, + "learning_rate": 0.003, + "loss": 4.07, + "step": 53088 + }, + { + "epoch": 0.53089, + "grad_norm": 1.6735847110931379, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 53089 + }, + { + "epoch": 0.5309, + "grad_norm": 1.4124851939926204, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 53090 + }, + { + "epoch": 0.53091, + "grad_norm": 1.633577156222487, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 53091 + }, + { + "epoch": 0.53092, + "grad_norm": 1.5439544459918215, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 53092 + }, + { + "epoch": 0.53093, + "grad_norm": 1.7409173784403853, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 53093 + }, + { + "epoch": 0.53094, + "grad_norm": 1.771143255641844, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 53094 + }, + { + "epoch": 0.53095, + "grad_norm": 1.9156635527614314, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 53095 + }, + { + "epoch": 0.53096, + "grad_norm": 1.7590903096032375, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 53096 + }, + { + "epoch": 0.53097, + "grad_norm": 1.2883090180778713, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 53097 + }, + { + "epoch": 0.53098, + "grad_norm": 1.8502611859015714, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 53098 + }, + { + "epoch": 0.53099, + "grad_norm": 1.3793039921206676, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 53099 + }, + { + "epoch": 0.531, + "grad_norm": 1.5836554216260046, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 53100 + }, + { + "epoch": 0.53101, + "grad_norm": 1.296507549123238, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 53101 + }, + { + "epoch": 0.53102, + "grad_norm": 2.02233361844666, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 53102 + }, + { + "epoch": 0.53103, + "grad_norm": 1.2683840631164356, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 53103 + }, + { + "epoch": 0.53104, + "grad_norm": 1.668888470345275, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 53104 + }, + { + "epoch": 0.53105, + "grad_norm": 1.4656992991655133, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 53105 + }, + { + "epoch": 0.53106, + "grad_norm": 1.8750760609043355, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 53106 + }, + { + "epoch": 0.53107, + "grad_norm": 1.8421101160743503, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 53107 + }, + { + "epoch": 0.53108, + "grad_norm": 1.5408091059964963, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 53108 + }, + { + "epoch": 0.53109, + "grad_norm": 1.7896611315582296, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 53109 + }, + { + "epoch": 0.5311, + "grad_norm": 1.5327638873785434, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53110 + }, + { + "epoch": 0.53111, + "grad_norm": 1.4567286538072544, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 53111 + }, + { + "epoch": 0.53112, + "grad_norm": 1.5521244388407605, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 53112 + }, + { + "epoch": 0.53113, + "grad_norm": 1.408133337742696, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 53113 + }, + { + "epoch": 0.53114, + "grad_norm": 1.6569848106209784, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 53114 + }, + { + "epoch": 0.53115, + "grad_norm": 1.4287962962186664, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 53115 + }, + { + "epoch": 0.53116, + "grad_norm": 1.6265663307114686, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 53116 + }, + { + "epoch": 0.53117, + "grad_norm": 1.5201718905155075, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 53117 + }, + { + "epoch": 0.53118, + "grad_norm": 1.2820070092705917, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 53118 + }, + { + "epoch": 0.53119, + "grad_norm": 1.5639390416696584, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 53119 + }, + { + "epoch": 0.5312, + "grad_norm": 1.3724648320146722, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 53120 + }, + { + "epoch": 0.53121, + "grad_norm": 1.453357179792477, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 53121 + }, + { + "epoch": 0.53122, + "grad_norm": 1.6872036940525499, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53122 + }, + { + "epoch": 0.53123, + "grad_norm": 1.9932381445280578, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 53123 + }, + { + "epoch": 0.53124, + "grad_norm": 1.7087272746339688, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 53124 + }, + { + "epoch": 0.53125, + "grad_norm": 1.376155809461124, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 53125 + }, + { + "epoch": 0.53126, + "grad_norm": 1.6780404263443882, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 53126 + }, + { + "epoch": 0.53127, + "grad_norm": 1.4040086193909744, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 53127 + }, + { + "epoch": 0.53128, + "grad_norm": 1.6854219889894289, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 53128 + }, + { + "epoch": 0.53129, + "grad_norm": 1.2241629645071133, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 53129 + }, + { + "epoch": 0.5313, + "grad_norm": 1.9277107610618827, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 53130 + }, + { + "epoch": 0.53131, + "grad_norm": 1.1748508353179252, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 53131 + }, + { + "epoch": 0.53132, + "grad_norm": 1.5484086807661315, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 53132 + }, + { + "epoch": 0.53133, + "grad_norm": 1.6165142133062809, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 53133 + }, + { + "epoch": 0.53134, + "grad_norm": 1.5080258587554474, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 53134 + }, + { + "epoch": 0.53135, + "grad_norm": 1.6354875480314606, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 53135 + }, + { + "epoch": 0.53136, + "grad_norm": 1.7868348039859259, + "learning_rate": 0.003, + "loss": 4.062, + "step": 53136 + }, + { + "epoch": 0.53137, + "grad_norm": 1.454910322419889, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 53137 + }, + { + "epoch": 0.53138, + "grad_norm": 1.6923628235641004, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 53138 + }, + { + "epoch": 0.53139, + "grad_norm": 1.4173052820069754, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 53139 + }, + { + "epoch": 0.5314, + "grad_norm": 1.5680676817765273, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 53140 + }, + { + "epoch": 0.53141, + "grad_norm": 1.2950555105166626, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 53141 + }, + { + "epoch": 0.53142, + "grad_norm": 1.7841658695653964, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 53142 + }, + { + "epoch": 0.53143, + "grad_norm": 1.168972837850347, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 53143 + }, + { + "epoch": 0.53144, + "grad_norm": 1.7988877787238193, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 53144 + }, + { + "epoch": 0.53145, + "grad_norm": 1.2780149677856407, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 53145 + }, + { + "epoch": 0.53146, + "grad_norm": 1.7981883444258087, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 53146 + }, + { + "epoch": 0.53147, + "grad_norm": 1.6373898106665783, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 53147 + }, + { + "epoch": 0.53148, + "grad_norm": 1.775022067454859, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 53148 + }, + { + "epoch": 0.53149, + "grad_norm": 1.4782522991932727, + "learning_rate": 0.003, + "loss": 4.045, + "step": 53149 + }, + { + "epoch": 0.5315, + "grad_norm": 1.4162386565988891, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 53150 + }, + { + "epoch": 0.53151, + "grad_norm": 1.8043142295043446, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53151 + }, + { + "epoch": 0.53152, + "grad_norm": 1.5997673894398419, + "learning_rate": 0.003, + "loss": 4.025, + "step": 53152 + }, + { + "epoch": 0.53153, + "grad_norm": 1.537150816823403, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 53153 + }, + { + "epoch": 0.53154, + "grad_norm": 1.3354501327861643, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 53154 + }, + { + "epoch": 0.53155, + "grad_norm": 1.6148109153222507, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 53155 + }, + { + "epoch": 0.53156, + "grad_norm": 1.4054192673406378, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 53156 + }, + { + "epoch": 0.53157, + "grad_norm": 1.5694886953114204, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 53157 + }, + { + "epoch": 0.53158, + "grad_norm": 1.6913493151314198, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 53158 + }, + { + "epoch": 0.53159, + "grad_norm": 1.357400224072602, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 53159 + }, + { + "epoch": 0.5316, + "grad_norm": 1.554888169734317, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 53160 + }, + { + "epoch": 0.53161, + "grad_norm": 1.4419782037712618, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 53161 + }, + { + "epoch": 0.53162, + "grad_norm": 1.5846940198299486, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 53162 + }, + { + "epoch": 0.53163, + "grad_norm": 1.2540963796727116, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 53163 + }, + { + "epoch": 0.53164, + "grad_norm": 1.8159819068078382, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 53164 + }, + { + "epoch": 0.53165, + "grad_norm": 1.2869996714591372, + "learning_rate": 0.003, + "loss": 4.049, + "step": 53165 + }, + { + "epoch": 0.53166, + "grad_norm": 1.7446926907360623, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 53166 + }, + { + "epoch": 0.53167, + "grad_norm": 1.198378629240135, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 53167 + }, + { + "epoch": 0.53168, + "grad_norm": 1.6154935887638402, + "learning_rate": 0.003, + "loss": 4.024, + "step": 53168 + }, + { + "epoch": 0.53169, + "grad_norm": 1.6116146412469237, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 53169 + }, + { + "epoch": 0.5317, + "grad_norm": 1.7346576963164655, + "learning_rate": 0.003, + "loss": 4.036, + "step": 53170 + }, + { + "epoch": 0.53171, + "grad_norm": 1.7478899318997427, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 53171 + }, + { + "epoch": 0.53172, + "grad_norm": 1.357212119630547, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 53172 + }, + { + "epoch": 0.53173, + "grad_norm": 2.192553938403577, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 53173 + }, + { + "epoch": 0.53174, + "grad_norm": 1.3394163857038996, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53174 + }, + { + "epoch": 0.53175, + "grad_norm": 1.6196339160498177, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 53175 + }, + { + "epoch": 0.53176, + "grad_norm": 1.5296620909620569, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 53176 + }, + { + "epoch": 0.53177, + "grad_norm": 1.4278576529699343, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 53177 + }, + { + "epoch": 0.53178, + "grad_norm": 1.5909030204865433, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 53178 + }, + { + "epoch": 0.53179, + "grad_norm": 1.4329086513605302, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 53179 + }, + { + "epoch": 0.5318, + "grad_norm": 1.7189740841048993, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 53180 + }, + { + "epoch": 0.53181, + "grad_norm": 1.4460437555648171, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 53181 + }, + { + "epoch": 0.53182, + "grad_norm": 1.9707266761453188, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 53182 + }, + { + "epoch": 0.53183, + "grad_norm": 1.6012714028978874, + "learning_rate": 0.003, + "loss": 4.054, + "step": 53183 + }, + { + "epoch": 0.53184, + "grad_norm": 1.228979256163889, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 53184 + }, + { + "epoch": 0.53185, + "grad_norm": 1.6432681386958818, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 53185 + }, + { + "epoch": 0.53186, + "grad_norm": 1.58580966099072, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 53186 + }, + { + "epoch": 0.53187, + "grad_norm": 1.5201780873389736, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 53187 + }, + { + "epoch": 0.53188, + "grad_norm": 1.6801147389548519, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 53188 + }, + { + "epoch": 0.53189, + "grad_norm": 1.3277121595591383, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 53189 + }, + { + "epoch": 0.5319, + "grad_norm": 1.5191747683108248, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 53190 + }, + { + "epoch": 0.53191, + "grad_norm": 1.3549683451144179, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 53191 + }, + { + "epoch": 0.53192, + "grad_norm": 1.6790386719567363, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 53192 + }, + { + "epoch": 0.53193, + "grad_norm": 1.385856217728964, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 53193 + }, + { + "epoch": 0.53194, + "grad_norm": 1.6252622837527957, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 53194 + }, + { + "epoch": 0.53195, + "grad_norm": 1.4872388165402655, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 53195 + }, + { + "epoch": 0.53196, + "grad_norm": 1.8048202763647447, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 53196 + }, + { + "epoch": 0.53197, + "grad_norm": 1.3325646594565548, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 53197 + }, + { + "epoch": 0.53198, + "grad_norm": 1.3566690986761367, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 53198 + }, + { + "epoch": 0.53199, + "grad_norm": 1.619578058948849, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 53199 + }, + { + "epoch": 0.532, + "grad_norm": 1.7015042920024412, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 53200 + }, + { + "epoch": 0.53201, + "grad_norm": 1.9006992310613011, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 53201 + }, + { + "epoch": 0.53202, + "grad_norm": 1.4900051357447375, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 53202 + }, + { + "epoch": 0.53203, + "grad_norm": 1.603011704738949, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 53203 + }, + { + "epoch": 0.53204, + "grad_norm": 1.542047831371847, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 53204 + }, + { + "epoch": 0.53205, + "grad_norm": 2.0109955750275055, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 53205 + }, + { + "epoch": 0.53206, + "grad_norm": 1.3218712150862046, + "learning_rate": 0.003, + "loss": 4.041, + "step": 53206 + }, + { + "epoch": 0.53207, + "grad_norm": 1.7512725901143822, + "learning_rate": 0.003, + "loss": 4.031, + "step": 53207 + }, + { + "epoch": 0.53208, + "grad_norm": 1.3360684984391533, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 53208 + }, + { + "epoch": 0.53209, + "grad_norm": 1.616299737160409, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 53209 + }, + { + "epoch": 0.5321, + "grad_norm": 1.216937278442808, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 53210 + }, + { + "epoch": 0.53211, + "grad_norm": 1.4577056334255938, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 53211 + }, + { + "epoch": 0.53212, + "grad_norm": 1.6190515008148059, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 53212 + }, + { + "epoch": 0.53213, + "grad_norm": 1.3703224768907931, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 53213 + }, + { + "epoch": 0.53214, + "grad_norm": 1.4616774737096623, + "learning_rate": 0.003, + "loss": 4.045, + "step": 53214 + }, + { + "epoch": 0.53215, + "grad_norm": 1.686832008001778, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 53215 + }, + { + "epoch": 0.53216, + "grad_norm": 1.3577245410100227, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 53216 + }, + { + "epoch": 0.53217, + "grad_norm": 1.8229113314906498, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 53217 + }, + { + "epoch": 0.53218, + "grad_norm": 2.0726742579350854, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 53218 + }, + { + "epoch": 0.53219, + "grad_norm": 1.6032540566311213, + "learning_rate": 0.003, + "loss": 4.042, + "step": 53219 + }, + { + "epoch": 0.5322, + "grad_norm": 1.6914377249678008, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 53220 + }, + { + "epoch": 0.53221, + "grad_norm": 1.350498729711726, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 53221 + }, + { + "epoch": 0.53222, + "grad_norm": 2.1034339828073954, + "learning_rate": 0.003, + "loss": 4.027, + "step": 53222 + }, + { + "epoch": 0.53223, + "grad_norm": 1.017849930840395, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 53223 + }, + { + "epoch": 0.53224, + "grad_norm": 1.8012677191115631, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 53224 + }, + { + "epoch": 0.53225, + "grad_norm": 1.6635740420237781, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 53225 + }, + { + "epoch": 0.53226, + "grad_norm": 1.4495433164878067, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 53226 + }, + { + "epoch": 0.53227, + "grad_norm": 1.8290027561996933, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 53227 + }, + { + "epoch": 0.53228, + "grad_norm": 1.3501872812074636, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 53228 + }, + { + "epoch": 0.53229, + "grad_norm": 1.7713399296046537, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 53229 + }, + { + "epoch": 0.5323, + "grad_norm": 1.269947329872597, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 53230 + }, + { + "epoch": 0.53231, + "grad_norm": 2.161494978618786, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 53231 + }, + { + "epoch": 0.53232, + "grad_norm": 1.496949830177156, + "learning_rate": 0.003, + "loss": 4.043, + "step": 53232 + }, + { + "epoch": 0.53233, + "grad_norm": 1.4828665697079506, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 53233 + }, + { + "epoch": 0.53234, + "grad_norm": 1.455897856706885, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 53234 + }, + { + "epoch": 0.53235, + "grad_norm": 1.5632387532734884, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 53235 + }, + { + "epoch": 0.53236, + "grad_norm": 1.6401941137279556, + "learning_rate": 0.003, + "loss": 4.041, + "step": 53236 + }, + { + "epoch": 0.53237, + "grad_norm": 1.150902704242544, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 53237 + }, + { + "epoch": 0.53238, + "grad_norm": 1.5531790756448023, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 53238 + }, + { + "epoch": 0.53239, + "grad_norm": 1.448596024667001, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 53239 + }, + { + "epoch": 0.5324, + "grad_norm": 1.6878916977579244, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 53240 + }, + { + "epoch": 0.53241, + "grad_norm": 1.3262742114326533, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 53241 + }, + { + "epoch": 0.53242, + "grad_norm": 1.979238400058443, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 53242 + }, + { + "epoch": 0.53243, + "grad_norm": 1.1564595217912252, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 53243 + }, + { + "epoch": 0.53244, + "grad_norm": 1.7574282479955325, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 53244 + }, + { + "epoch": 0.53245, + "grad_norm": 1.4981941493978492, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 53245 + }, + { + "epoch": 0.53246, + "grad_norm": 1.8118508692471267, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 53246 + }, + { + "epoch": 0.53247, + "grad_norm": 1.8324108958617318, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 53247 + }, + { + "epoch": 0.53248, + "grad_norm": 1.2119450288546574, + "learning_rate": 0.003, + "loss": 4.022, + "step": 53248 + }, + { + "epoch": 0.53249, + "grad_norm": 1.6479537576988381, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 53249 + }, + { + "epoch": 0.5325, + "grad_norm": 1.5408692367425205, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 53250 + }, + { + "epoch": 0.53251, + "grad_norm": 1.7075266007209577, + "learning_rate": 0.003, + "loss": 4.066, + "step": 53251 + }, + { + "epoch": 0.53252, + "grad_norm": 1.2698255959734464, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 53252 + }, + { + "epoch": 0.53253, + "grad_norm": 1.8960787656604599, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 53253 + }, + { + "epoch": 0.53254, + "grad_norm": 1.17442371218301, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 53254 + }, + { + "epoch": 0.53255, + "grad_norm": 1.50511842685864, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 53255 + }, + { + "epoch": 0.53256, + "grad_norm": 1.8564847318877697, + "learning_rate": 0.003, + "loss": 4.049, + "step": 53256 + }, + { + "epoch": 0.53257, + "grad_norm": 1.5018689780187502, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 53257 + }, + { + "epoch": 0.53258, + "grad_norm": 1.4698753224095173, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 53258 + }, + { + "epoch": 0.53259, + "grad_norm": 1.45889758433215, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53259 + }, + { + "epoch": 0.5326, + "grad_norm": 1.6166417841270673, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 53260 + }, + { + "epoch": 0.53261, + "grad_norm": 1.307695164235158, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 53261 + }, + { + "epoch": 0.53262, + "grad_norm": 2.0436600096333977, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 53262 + }, + { + "epoch": 0.53263, + "grad_norm": 1.2508321280239472, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 53263 + }, + { + "epoch": 0.53264, + "grad_norm": 1.6650749513941197, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 53264 + }, + { + "epoch": 0.53265, + "grad_norm": 1.5992656248871295, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 53265 + }, + { + "epoch": 0.53266, + "grad_norm": 1.8094055297156815, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 53266 + }, + { + "epoch": 0.53267, + "grad_norm": 1.349821764326583, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 53267 + }, + { + "epoch": 0.53268, + "grad_norm": 1.73305948913676, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 53268 + }, + { + "epoch": 0.53269, + "grad_norm": 1.3772179712224955, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 53269 + }, + { + "epoch": 0.5327, + "grad_norm": 1.5303863079563844, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 53270 + }, + { + "epoch": 0.53271, + "grad_norm": 1.6237305869317538, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 53271 + }, + { + "epoch": 0.53272, + "grad_norm": 1.507185675479042, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 53272 + }, + { + "epoch": 0.53273, + "grad_norm": 1.573021274197401, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 53273 + }, + { + "epoch": 0.53274, + "grad_norm": 1.6677760339521026, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 53274 + }, + { + "epoch": 0.53275, + "grad_norm": 1.5335446678685596, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 53275 + }, + { + "epoch": 0.53276, + "grad_norm": 1.1451524524652894, + "learning_rate": 0.003, + "loss": 4.052, + "step": 53276 + }, + { + "epoch": 0.53277, + "grad_norm": 1.8952894864428098, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53277 + }, + { + "epoch": 0.53278, + "grad_norm": 1.370294717277727, + "learning_rate": 0.003, + "loss": 4.026, + "step": 53278 + }, + { + "epoch": 0.53279, + "grad_norm": 1.479654799282248, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 53279 + }, + { + "epoch": 0.5328, + "grad_norm": 1.4652927331376917, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 53280 + }, + { + "epoch": 0.53281, + "grad_norm": 1.3224420323862875, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 53281 + }, + { + "epoch": 0.53282, + "grad_norm": 2.1092173528885536, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 53282 + }, + { + "epoch": 0.53283, + "grad_norm": 1.215935440375233, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 53283 + }, + { + "epoch": 0.53284, + "grad_norm": 1.892297887509125, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 53284 + }, + { + "epoch": 0.53285, + "grad_norm": 1.6000124990670552, + "learning_rate": 0.003, + "loss": 4.043, + "step": 53285 + }, + { + "epoch": 0.53286, + "grad_norm": 1.41973050745204, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 53286 + }, + { + "epoch": 0.53287, + "grad_norm": 1.9178377472789037, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 53287 + }, + { + "epoch": 0.53288, + "grad_norm": 1.6556173007542874, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53288 + }, + { + "epoch": 0.53289, + "grad_norm": 1.2962357949383956, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 53289 + }, + { + "epoch": 0.5329, + "grad_norm": 1.4155243208930732, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 53290 + }, + { + "epoch": 0.53291, + "grad_norm": 1.679745299625484, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 53291 + }, + { + "epoch": 0.53292, + "grad_norm": 1.145029294912297, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 53292 + }, + { + "epoch": 0.53293, + "grad_norm": 1.6721670693624453, + "learning_rate": 0.003, + "loss": 4.031, + "step": 53293 + }, + { + "epoch": 0.53294, + "grad_norm": 1.531314745526702, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 53294 + }, + { + "epoch": 0.53295, + "grad_norm": 1.546800692594864, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 53295 + }, + { + "epoch": 0.53296, + "grad_norm": 1.58834916082423, + "learning_rate": 0.003, + "loss": 4.014, + "step": 53296 + }, + { + "epoch": 0.53297, + "grad_norm": 1.4299518469540398, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 53297 + }, + { + "epoch": 0.53298, + "grad_norm": 1.6155382867558947, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 53298 + }, + { + "epoch": 0.53299, + "grad_norm": 1.4701536303598104, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 53299 + }, + { + "epoch": 0.533, + "grad_norm": 1.6938820064270526, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 53300 + }, + { + "epoch": 0.53301, + "grad_norm": 1.4994300882587004, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 53301 + }, + { + "epoch": 0.53302, + "grad_norm": 1.455868509020913, + "learning_rate": 0.003, + "loss": 3.996, + "step": 53302 + }, + { + "epoch": 0.53303, + "grad_norm": 1.6987626435159962, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 53303 + }, + { + "epoch": 0.53304, + "grad_norm": 1.4067448855472622, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 53304 + }, + { + "epoch": 0.53305, + "grad_norm": 1.7358052663277617, + "learning_rate": 0.003, + "loss": 4.025, + "step": 53305 + }, + { + "epoch": 0.53306, + "grad_norm": 1.6442478464105552, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 53306 + }, + { + "epoch": 0.53307, + "grad_norm": 1.6036089180520843, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 53307 + }, + { + "epoch": 0.53308, + "grad_norm": 1.6518271707402683, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 53308 + }, + { + "epoch": 0.53309, + "grad_norm": 1.475149817855065, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 53309 + }, + { + "epoch": 0.5331, + "grad_norm": 1.504316931545742, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 53310 + }, + { + "epoch": 0.53311, + "grad_norm": 1.4659655589850686, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 53311 + }, + { + "epoch": 0.53312, + "grad_norm": 1.6106544027101741, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 53312 + }, + { + "epoch": 0.53313, + "grad_norm": 1.3317681565293316, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 53313 + }, + { + "epoch": 0.53314, + "grad_norm": 1.8952110548356706, + "learning_rate": 0.003, + "loss": 4.029, + "step": 53314 + }, + { + "epoch": 0.53315, + "grad_norm": 1.093843367395429, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 53315 + }, + { + "epoch": 0.53316, + "grad_norm": 1.9129177146213376, + "learning_rate": 0.003, + "loss": 4.02, + "step": 53316 + }, + { + "epoch": 0.53317, + "grad_norm": 1.2041251855141089, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 53317 + }, + { + "epoch": 0.53318, + "grad_norm": 2.1027292828608704, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 53318 + }, + { + "epoch": 0.53319, + "grad_norm": 1.1548986307513804, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 53319 + }, + { + "epoch": 0.5332, + "grad_norm": 2.1272639347299083, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 53320 + }, + { + "epoch": 0.53321, + "grad_norm": 1.8765937515847833, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 53321 + }, + { + "epoch": 0.53322, + "grad_norm": 1.3896382667089588, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 53322 + }, + { + "epoch": 0.53323, + "grad_norm": 1.7313013060418738, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 53323 + }, + { + "epoch": 0.53324, + "grad_norm": 1.5566303415460823, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 53324 + }, + { + "epoch": 0.53325, + "grad_norm": 1.5255152202076958, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 53325 + }, + { + "epoch": 0.53326, + "grad_norm": 1.491667767755423, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 53326 + }, + { + "epoch": 0.53327, + "grad_norm": 1.6923898612889376, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 53327 + }, + { + "epoch": 0.53328, + "grad_norm": 1.3410645195071418, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53328 + }, + { + "epoch": 0.53329, + "grad_norm": 1.3959688812925626, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 53329 + }, + { + "epoch": 0.5333, + "grad_norm": 1.6131666524653943, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 53330 + }, + { + "epoch": 0.53331, + "grad_norm": 1.6048640245938577, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 53331 + }, + { + "epoch": 0.53332, + "grad_norm": 1.4340306449929612, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 53332 + }, + { + "epoch": 0.53333, + "grad_norm": 1.397766956141027, + "learning_rate": 0.003, + "loss": 4.038, + "step": 53333 + }, + { + "epoch": 0.53334, + "grad_norm": 1.4130303671514512, + "learning_rate": 0.003, + "loss": 4.043, + "step": 53334 + }, + { + "epoch": 0.53335, + "grad_norm": 1.472867842421546, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 53335 + }, + { + "epoch": 0.53336, + "grad_norm": 1.359377616544775, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 53336 + }, + { + "epoch": 0.53337, + "grad_norm": 1.3647295676656344, + "learning_rate": 0.003, + "loss": 4.047, + "step": 53337 + }, + { + "epoch": 0.53338, + "grad_norm": 1.34226731948281, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 53338 + }, + { + "epoch": 0.53339, + "grad_norm": 1.5430909710206309, + "learning_rate": 0.003, + "loss": 4.013, + "step": 53339 + }, + { + "epoch": 0.5334, + "grad_norm": 1.708116367257662, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 53340 + }, + { + "epoch": 0.53341, + "grad_norm": 1.8825326851493842, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 53341 + }, + { + "epoch": 0.53342, + "grad_norm": 1.6469830065301143, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 53342 + }, + { + "epoch": 0.53343, + "grad_norm": 1.5056983718682775, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 53343 + }, + { + "epoch": 0.53344, + "grad_norm": 1.3854936046152977, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 53344 + }, + { + "epoch": 0.53345, + "grad_norm": 1.4703196760337371, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 53345 + }, + { + "epoch": 0.53346, + "grad_norm": 1.5637999637422522, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 53346 + }, + { + "epoch": 0.53347, + "grad_norm": 1.6325701984319576, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 53347 + }, + { + "epoch": 0.53348, + "grad_norm": 1.8816682540999856, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 53348 + }, + { + "epoch": 0.53349, + "grad_norm": 1.2090330636025082, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 53349 + }, + { + "epoch": 0.5335, + "grad_norm": 1.7590345726392622, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 53350 + }, + { + "epoch": 0.53351, + "grad_norm": 1.5002913083243419, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 53351 + }, + { + "epoch": 0.53352, + "grad_norm": 1.84311257589545, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 53352 + }, + { + "epoch": 0.53353, + "grad_norm": 1.3150528814335865, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 53353 + }, + { + "epoch": 0.53354, + "grad_norm": 1.4700480174755972, + "learning_rate": 0.003, + "loss": 4.007, + "step": 53354 + }, + { + "epoch": 0.53355, + "grad_norm": 1.5542420208779122, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 53355 + }, + { + "epoch": 0.53356, + "grad_norm": 1.4276367371329146, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 53356 + }, + { + "epoch": 0.53357, + "grad_norm": 1.6695921351079528, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 53357 + }, + { + "epoch": 0.53358, + "grad_norm": 1.4470845044725595, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 53358 + }, + { + "epoch": 0.53359, + "grad_norm": 1.934613214650811, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 53359 + }, + { + "epoch": 0.5336, + "grad_norm": 1.434156499011399, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 53360 + }, + { + "epoch": 0.53361, + "grad_norm": 1.839648118541235, + "learning_rate": 0.003, + "loss": 4.049, + "step": 53361 + }, + { + "epoch": 0.53362, + "grad_norm": 1.264260697520475, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 53362 + }, + { + "epoch": 0.53363, + "grad_norm": 1.781194725936543, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 53363 + }, + { + "epoch": 0.53364, + "grad_norm": 1.4153578005909657, + "learning_rate": 0.003, + "loss": 4.023, + "step": 53364 + }, + { + "epoch": 0.53365, + "grad_norm": 1.6794428491102267, + "learning_rate": 0.003, + "loss": 4.04, + "step": 53365 + }, + { + "epoch": 0.53366, + "grad_norm": 1.5716108241770932, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 53366 + }, + { + "epoch": 0.53367, + "grad_norm": 1.6360025178175852, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 53367 + }, + { + "epoch": 0.53368, + "grad_norm": 1.4722891203458501, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 53368 + }, + { + "epoch": 0.53369, + "grad_norm": 1.3938347138622076, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 53369 + }, + { + "epoch": 0.5337, + "grad_norm": 1.383197802519479, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 53370 + }, + { + "epoch": 0.53371, + "grad_norm": 1.734551212732326, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 53371 + }, + { + "epoch": 0.53372, + "grad_norm": 1.4182524916929358, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53372 + }, + { + "epoch": 0.53373, + "grad_norm": 1.6399936100756152, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 53373 + }, + { + "epoch": 0.53374, + "grad_norm": 1.325878058346917, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 53374 + }, + { + "epoch": 0.53375, + "grad_norm": 2.1714771457810538, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 53375 + }, + { + "epoch": 0.53376, + "grad_norm": 1.3475407014752454, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 53376 + }, + { + "epoch": 0.53377, + "grad_norm": 1.7762136134506086, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 53377 + }, + { + "epoch": 0.53378, + "grad_norm": 1.6236551278879687, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 53378 + }, + { + "epoch": 0.53379, + "grad_norm": 1.5818752448863216, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 53379 + }, + { + "epoch": 0.5338, + "grad_norm": 1.6639219885590175, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 53380 + }, + { + "epoch": 0.53381, + "grad_norm": 1.4838317168758066, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 53381 + }, + { + "epoch": 0.53382, + "grad_norm": 1.7032294388615805, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 53382 + }, + { + "epoch": 0.53383, + "grad_norm": 1.6351558014892495, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 53383 + }, + { + "epoch": 0.53384, + "grad_norm": 1.5412217079743893, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 53384 + }, + { + "epoch": 0.53385, + "grad_norm": 1.3474870116584927, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 53385 + }, + { + "epoch": 0.53386, + "grad_norm": 1.6450830056975554, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 53386 + }, + { + "epoch": 0.53387, + "grad_norm": 1.2390829330091424, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53387 + }, + { + "epoch": 0.53388, + "grad_norm": 1.690716002183131, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 53388 + }, + { + "epoch": 0.53389, + "grad_norm": 1.605446486691748, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 53389 + }, + { + "epoch": 0.5339, + "grad_norm": 1.6946155267747742, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 53390 + }, + { + "epoch": 0.53391, + "grad_norm": 1.6241606809407094, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 53391 + }, + { + "epoch": 0.53392, + "grad_norm": 1.4024452931812832, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 53392 + }, + { + "epoch": 0.53393, + "grad_norm": 1.7330720583025534, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 53393 + }, + { + "epoch": 0.53394, + "grad_norm": 1.2750450258072976, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 53394 + }, + { + "epoch": 0.53395, + "grad_norm": 1.6420456730874462, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 53395 + }, + { + "epoch": 0.53396, + "grad_norm": 1.5062610735578497, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 53396 + }, + { + "epoch": 0.53397, + "grad_norm": 2.0558580300208704, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 53397 + }, + { + "epoch": 0.53398, + "grad_norm": 1.358533846413405, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 53398 + }, + { + "epoch": 0.53399, + "grad_norm": 1.8778974705497056, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 53399 + }, + { + "epoch": 0.534, + "grad_norm": 1.4945739423394508, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 53400 + }, + { + "epoch": 0.53401, + "grad_norm": 1.3922599090211105, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 53401 + }, + { + "epoch": 0.53402, + "grad_norm": 1.435811933718417, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 53402 + }, + { + "epoch": 0.53403, + "grad_norm": 1.3379246169865675, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 53403 + }, + { + "epoch": 0.53404, + "grad_norm": 1.8873974952696948, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 53404 + }, + { + "epoch": 0.53405, + "grad_norm": 1.4950476435842102, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 53405 + }, + { + "epoch": 0.53406, + "grad_norm": 1.6269963184117744, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 53406 + }, + { + "epoch": 0.53407, + "grad_norm": 1.579392094572197, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 53407 + }, + { + "epoch": 0.53408, + "grad_norm": 1.5540325038598894, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 53408 + }, + { + "epoch": 0.53409, + "grad_norm": 1.8969122131267928, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 53409 + }, + { + "epoch": 0.5341, + "grad_norm": 1.4629127247507656, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 53410 + }, + { + "epoch": 0.53411, + "grad_norm": 1.6719842672507783, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 53411 + }, + { + "epoch": 0.53412, + "grad_norm": 1.1672013844673876, + "learning_rate": 0.003, + "loss": 4.02, + "step": 53412 + }, + { + "epoch": 0.53413, + "grad_norm": 1.4832847997288126, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 53413 + }, + { + "epoch": 0.53414, + "grad_norm": 1.7281211854380798, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 53414 + }, + { + "epoch": 0.53415, + "grad_norm": 1.659309325460214, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 53415 + }, + { + "epoch": 0.53416, + "grad_norm": 1.5204965328309763, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 53416 + }, + { + "epoch": 0.53417, + "grad_norm": 1.6275991431234336, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 53417 + }, + { + "epoch": 0.53418, + "grad_norm": 1.5111552848067502, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 53418 + }, + { + "epoch": 0.53419, + "grad_norm": 1.6760840307737135, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 53419 + }, + { + "epoch": 0.5342, + "grad_norm": 1.3075467668363114, + "learning_rate": 0.003, + "loss": 4.016, + "step": 53420 + }, + { + "epoch": 0.53421, + "grad_norm": 1.7367522422305486, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 53421 + }, + { + "epoch": 0.53422, + "grad_norm": 1.345372443428069, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 53422 + }, + { + "epoch": 0.53423, + "grad_norm": 1.7570772722600174, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 53423 + }, + { + "epoch": 0.53424, + "grad_norm": 1.3680721159826816, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 53424 + }, + { + "epoch": 0.53425, + "grad_norm": 1.495248160445952, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 53425 + }, + { + "epoch": 0.53426, + "grad_norm": 1.5781035850992615, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 53426 + }, + { + "epoch": 0.53427, + "grad_norm": 1.6900460982102716, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 53427 + }, + { + "epoch": 0.53428, + "grad_norm": 1.5178213460966634, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 53428 + }, + { + "epoch": 0.53429, + "grad_norm": 1.991616883712534, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 53429 + }, + { + "epoch": 0.5343, + "grad_norm": 1.4680856803179019, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 53430 + }, + { + "epoch": 0.53431, + "grad_norm": 1.6410082115401456, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 53431 + }, + { + "epoch": 0.53432, + "grad_norm": 1.6397510792480345, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 53432 + }, + { + "epoch": 0.53433, + "grad_norm": 1.4322122996250406, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 53433 + }, + { + "epoch": 0.53434, + "grad_norm": 1.506755611845792, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 53434 + }, + { + "epoch": 0.53435, + "grad_norm": 1.7894151022601228, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 53435 + }, + { + "epoch": 0.53436, + "grad_norm": 1.4818923051053479, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 53436 + }, + { + "epoch": 0.53437, + "grad_norm": 1.6243364977372794, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 53437 + }, + { + "epoch": 0.53438, + "grad_norm": 1.6178519705105294, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 53438 + }, + { + "epoch": 0.53439, + "grad_norm": 1.360919909618899, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 53439 + }, + { + "epoch": 0.5344, + "grad_norm": 1.530031065288852, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 53440 + }, + { + "epoch": 0.53441, + "grad_norm": 1.6441038170988231, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 53441 + }, + { + "epoch": 0.53442, + "grad_norm": 1.4863989154534991, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 53442 + }, + { + "epoch": 0.53443, + "grad_norm": 1.8847864028089245, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 53443 + }, + { + "epoch": 0.53444, + "grad_norm": 1.415860415897316, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 53444 + }, + { + "epoch": 0.53445, + "grad_norm": 1.6879920648223743, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 53445 + }, + { + "epoch": 0.53446, + "grad_norm": 1.705806030336412, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 53446 + }, + { + "epoch": 0.53447, + "grad_norm": 1.1821139618750773, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 53447 + }, + { + "epoch": 0.53448, + "grad_norm": 1.6559915874381161, + "learning_rate": 0.003, + "loss": 4.016, + "step": 53448 + }, + { + "epoch": 0.53449, + "grad_norm": 1.3781546559098055, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 53449 + }, + { + "epoch": 0.5345, + "grad_norm": 1.8162600072303683, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 53450 + }, + { + "epoch": 0.53451, + "grad_norm": 1.3765209697835858, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 53451 + }, + { + "epoch": 0.53452, + "grad_norm": 1.887922978263478, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53452 + }, + { + "epoch": 0.53453, + "grad_norm": 1.61607106323695, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 53453 + }, + { + "epoch": 0.53454, + "grad_norm": 1.3915641361703113, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 53454 + }, + { + "epoch": 0.53455, + "grad_norm": 1.780907868969791, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 53455 + }, + { + "epoch": 0.53456, + "grad_norm": 1.0669908060792488, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 53456 + }, + { + "epoch": 0.53457, + "grad_norm": 1.852256259054923, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53457 + }, + { + "epoch": 0.53458, + "grad_norm": 1.4037452253289724, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 53458 + }, + { + "epoch": 0.53459, + "grad_norm": 1.6535381913623997, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 53459 + }, + { + "epoch": 0.5346, + "grad_norm": 1.7681486739222265, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 53460 + }, + { + "epoch": 0.53461, + "grad_norm": 1.524262576854258, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 53461 + }, + { + "epoch": 0.53462, + "grad_norm": 1.6355549195021526, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 53462 + }, + { + "epoch": 0.53463, + "grad_norm": 1.497514230425898, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 53463 + }, + { + "epoch": 0.53464, + "grad_norm": 1.5103844360232077, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 53464 + }, + { + "epoch": 0.53465, + "grad_norm": 1.454441210505984, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 53465 + }, + { + "epoch": 0.53466, + "grad_norm": 1.6921510698398141, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 53466 + }, + { + "epoch": 0.53467, + "grad_norm": 1.1839812123518796, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 53467 + }, + { + "epoch": 0.53468, + "grad_norm": 1.8786389154427101, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 53468 + }, + { + "epoch": 0.53469, + "grad_norm": 1.866929037315964, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 53469 + }, + { + "epoch": 0.5347, + "grad_norm": 1.5196420355872318, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 53470 + }, + { + "epoch": 0.53471, + "grad_norm": 1.4953792540033923, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53471 + }, + { + "epoch": 0.53472, + "grad_norm": 1.743358303877721, + "learning_rate": 0.003, + "loss": 4.046, + "step": 53472 + }, + { + "epoch": 0.53473, + "grad_norm": 1.4537345463237163, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 53473 + }, + { + "epoch": 0.53474, + "grad_norm": 1.458284187500015, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 53474 + }, + { + "epoch": 0.53475, + "grad_norm": 1.5715676006833617, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 53475 + }, + { + "epoch": 0.53476, + "grad_norm": 1.3804791555093703, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 53476 + }, + { + "epoch": 0.53477, + "grad_norm": 1.8537178742913267, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 53477 + }, + { + "epoch": 0.53478, + "grad_norm": 1.1640409073337745, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 53478 + }, + { + "epoch": 0.53479, + "grad_norm": 2.253700262562711, + "learning_rate": 0.003, + "loss": 4.014, + "step": 53479 + }, + { + "epoch": 0.5348, + "grad_norm": 1.5893623853022645, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 53480 + }, + { + "epoch": 0.53481, + "grad_norm": 1.494079669685287, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 53481 + }, + { + "epoch": 0.53482, + "grad_norm": 1.7933181932547924, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 53482 + }, + { + "epoch": 0.53483, + "grad_norm": 1.354030628728431, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 53483 + }, + { + "epoch": 0.53484, + "grad_norm": 1.3791860692905338, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 53484 + }, + { + "epoch": 0.53485, + "grad_norm": 1.5033908789376391, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 53485 + }, + { + "epoch": 0.53486, + "grad_norm": 1.5893391645870663, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 53486 + }, + { + "epoch": 0.53487, + "grad_norm": 1.2435880393983398, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53487 + }, + { + "epoch": 0.53488, + "grad_norm": 1.7323002724306134, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 53488 + }, + { + "epoch": 0.53489, + "grad_norm": 1.476505299777924, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 53489 + }, + { + "epoch": 0.5349, + "grad_norm": 1.4341679928218172, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 53490 + }, + { + "epoch": 0.53491, + "grad_norm": 1.7653778954190031, + "learning_rate": 0.003, + "loss": 4.066, + "step": 53491 + }, + { + "epoch": 0.53492, + "grad_norm": 1.5459895218125095, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 53492 + }, + { + "epoch": 0.53493, + "grad_norm": 2.0653454887605007, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 53493 + }, + { + "epoch": 0.53494, + "grad_norm": 1.8963429073705587, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 53494 + }, + { + "epoch": 0.53495, + "grad_norm": 1.2281562361131082, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 53495 + }, + { + "epoch": 0.53496, + "grad_norm": 1.7558898903264584, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 53496 + }, + { + "epoch": 0.53497, + "grad_norm": 1.5215677672441459, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 53497 + }, + { + "epoch": 0.53498, + "grad_norm": 1.5239920334192465, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 53498 + }, + { + "epoch": 0.53499, + "grad_norm": 1.4786626663823887, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 53499 + }, + { + "epoch": 0.535, + "grad_norm": 1.3789096054290444, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 53500 + }, + { + "epoch": 0.53501, + "grad_norm": 1.6382454706455312, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 53501 + }, + { + "epoch": 0.53502, + "grad_norm": 1.3356925723697015, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 53502 + }, + { + "epoch": 0.53503, + "grad_norm": 1.5016946416508719, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 53503 + }, + { + "epoch": 0.53504, + "grad_norm": 1.5153754832809, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 53504 + }, + { + "epoch": 0.53505, + "grad_norm": 1.464820214893811, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 53505 + }, + { + "epoch": 0.53506, + "grad_norm": 1.7872072385254396, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53506 + }, + { + "epoch": 0.53507, + "grad_norm": 2.0327014363332094, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 53507 + }, + { + "epoch": 0.53508, + "grad_norm": 1.4761721541783355, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 53508 + }, + { + "epoch": 0.53509, + "grad_norm": 1.3980871586484196, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 53509 + }, + { + "epoch": 0.5351, + "grad_norm": 2.0418869902399828, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 53510 + }, + { + "epoch": 0.53511, + "grad_norm": 1.6811546816773957, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 53511 + }, + { + "epoch": 0.53512, + "grad_norm": 1.3969689433815966, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 53512 + }, + { + "epoch": 0.53513, + "grad_norm": 1.477024039930233, + "learning_rate": 0.003, + "loss": 4.029, + "step": 53513 + }, + { + "epoch": 0.53514, + "grad_norm": 1.5216251272260068, + "learning_rate": 0.003, + "loss": 4.056, + "step": 53514 + }, + { + "epoch": 0.53515, + "grad_norm": 1.4909315281744266, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 53515 + }, + { + "epoch": 0.53516, + "grad_norm": 1.4149719036346369, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 53516 + }, + { + "epoch": 0.53517, + "grad_norm": 1.5483879229282311, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 53517 + }, + { + "epoch": 0.53518, + "grad_norm": 1.2546226016326028, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 53518 + }, + { + "epoch": 0.53519, + "grad_norm": 1.6579564670555433, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 53519 + }, + { + "epoch": 0.5352, + "grad_norm": 1.4983246869051419, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 53520 + }, + { + "epoch": 0.53521, + "grad_norm": 1.9024368912471532, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 53521 + }, + { + "epoch": 0.53522, + "grad_norm": 1.3193172607650987, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 53522 + }, + { + "epoch": 0.53523, + "grad_norm": 1.6936772059067307, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 53523 + }, + { + "epoch": 0.53524, + "grad_norm": 1.6957446411235821, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 53524 + }, + { + "epoch": 0.53525, + "grad_norm": 1.3284497810587315, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 53525 + }, + { + "epoch": 0.53526, + "grad_norm": 2.071256736978089, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 53526 + }, + { + "epoch": 0.53527, + "grad_norm": 1.2528270228814542, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 53527 + }, + { + "epoch": 0.53528, + "grad_norm": 1.7765117045449046, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 53528 + }, + { + "epoch": 0.53529, + "grad_norm": 1.368747714889234, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 53529 + }, + { + "epoch": 0.5353, + "grad_norm": 1.7775828648319165, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53530 + }, + { + "epoch": 0.53531, + "grad_norm": 1.5211635718088243, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 53531 + }, + { + "epoch": 0.53532, + "grad_norm": 1.4389417356416732, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 53532 + }, + { + "epoch": 0.53533, + "grad_norm": 1.250207225558927, + "learning_rate": 0.003, + "loss": 4.04, + "step": 53533 + }, + { + "epoch": 0.53534, + "grad_norm": 1.9192672462815392, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 53534 + }, + { + "epoch": 0.53535, + "grad_norm": 1.649779744845703, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 53535 + }, + { + "epoch": 0.53536, + "grad_norm": 1.4298470428867944, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 53536 + }, + { + "epoch": 0.53537, + "grad_norm": 1.9432239411589138, + "learning_rate": 0.003, + "loss": 4.059, + "step": 53537 + }, + { + "epoch": 0.53538, + "grad_norm": 1.3185396469316637, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 53538 + }, + { + "epoch": 0.53539, + "grad_norm": 1.4404078159235523, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 53539 + }, + { + "epoch": 0.5354, + "grad_norm": 1.5306137836478653, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 53540 + }, + { + "epoch": 0.53541, + "grad_norm": 1.5733419266132387, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 53541 + }, + { + "epoch": 0.53542, + "grad_norm": 1.4106668526757118, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 53542 + }, + { + "epoch": 0.53543, + "grad_norm": 1.262248323024816, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 53543 + }, + { + "epoch": 0.53544, + "grad_norm": 1.6053503457102114, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 53544 + }, + { + "epoch": 0.53545, + "grad_norm": 1.8298406155582427, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 53545 + }, + { + "epoch": 0.53546, + "grad_norm": 1.7470774897806571, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 53546 + }, + { + "epoch": 0.53547, + "grad_norm": 1.4458087094643146, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 53547 + }, + { + "epoch": 0.53548, + "grad_norm": 1.7085631117915203, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 53548 + }, + { + "epoch": 0.53549, + "grad_norm": 1.44282465590473, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 53549 + }, + { + "epoch": 0.5355, + "grad_norm": 1.5729982023439755, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 53550 + }, + { + "epoch": 0.53551, + "grad_norm": 1.75052026571513, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 53551 + }, + { + "epoch": 0.53552, + "grad_norm": 1.2392281125789144, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 53552 + }, + { + "epoch": 0.53553, + "grad_norm": 2.0388404398271205, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 53553 + }, + { + "epoch": 0.53554, + "grad_norm": 1.4387974252094013, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 53554 + }, + { + "epoch": 0.53555, + "grad_norm": 1.4776303941869475, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 53555 + }, + { + "epoch": 0.53556, + "grad_norm": 1.4363886565671373, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 53556 + }, + { + "epoch": 0.53557, + "grad_norm": 1.785446551474925, + "learning_rate": 0.003, + "loss": 4.059, + "step": 53557 + }, + { + "epoch": 0.53558, + "grad_norm": 1.3491468705500371, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 53558 + }, + { + "epoch": 0.53559, + "grad_norm": 1.5265622225393711, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 53559 + }, + { + "epoch": 0.5356, + "grad_norm": 1.7554185103761215, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 53560 + }, + { + "epoch": 0.53561, + "grad_norm": 1.229573975030319, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 53561 + }, + { + "epoch": 0.53562, + "grad_norm": 1.6560651426457769, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 53562 + }, + { + "epoch": 0.53563, + "grad_norm": 1.4816087225627843, + "learning_rate": 0.003, + "loss": 4.037, + "step": 53563 + }, + { + "epoch": 0.53564, + "grad_norm": 1.8128722176672676, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 53564 + }, + { + "epoch": 0.53565, + "grad_norm": 1.6060637835680014, + "learning_rate": 0.003, + "loss": 4.037, + "step": 53565 + }, + { + "epoch": 0.53566, + "grad_norm": 1.675238380639863, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 53566 + }, + { + "epoch": 0.53567, + "grad_norm": 1.4210253794510916, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 53567 + }, + { + "epoch": 0.53568, + "grad_norm": 1.554303813980998, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 53568 + }, + { + "epoch": 0.53569, + "grad_norm": 1.6740486599903355, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 53569 + }, + { + "epoch": 0.5357, + "grad_norm": 1.3086939667397712, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 53570 + }, + { + "epoch": 0.53571, + "grad_norm": 1.5929795838890408, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 53571 + }, + { + "epoch": 0.53572, + "grad_norm": 1.3560927896245774, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 53572 + }, + { + "epoch": 0.53573, + "grad_norm": 1.7781449214207345, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 53573 + }, + { + "epoch": 0.53574, + "grad_norm": 1.1705771417855961, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 53574 + }, + { + "epoch": 0.53575, + "grad_norm": 1.789828655062147, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 53575 + }, + { + "epoch": 0.53576, + "grad_norm": 1.3609308804478852, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 53576 + }, + { + "epoch": 0.53577, + "grad_norm": 1.7059471541792133, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 53577 + }, + { + "epoch": 0.53578, + "grad_norm": 1.7081232698296454, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 53578 + }, + { + "epoch": 0.53579, + "grad_norm": 1.8723196571965157, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 53579 + }, + { + "epoch": 0.5358, + "grad_norm": 1.4040121865621038, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 53580 + }, + { + "epoch": 0.53581, + "grad_norm": 1.694596449036588, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 53581 + }, + { + "epoch": 0.53582, + "grad_norm": 1.697649623210832, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 53582 + }, + { + "epoch": 0.53583, + "grad_norm": 1.4396912255228789, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 53583 + }, + { + "epoch": 0.53584, + "grad_norm": 1.682684862475723, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 53584 + }, + { + "epoch": 0.53585, + "grad_norm": 1.4559512067889118, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 53585 + }, + { + "epoch": 0.53586, + "grad_norm": 1.6605603916101863, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 53586 + }, + { + "epoch": 0.53587, + "grad_norm": 1.3919602512264455, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53587 + }, + { + "epoch": 0.53588, + "grad_norm": 1.576961863249442, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 53588 + }, + { + "epoch": 0.53589, + "grad_norm": 1.4989509544647361, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 53589 + }, + { + "epoch": 0.5359, + "grad_norm": 1.8524411189002499, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 53590 + }, + { + "epoch": 0.53591, + "grad_norm": 1.407969460098439, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 53591 + }, + { + "epoch": 0.53592, + "grad_norm": 1.5451545108089961, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 53592 + }, + { + "epoch": 0.53593, + "grad_norm": 1.8205496503523275, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 53593 + }, + { + "epoch": 0.53594, + "grad_norm": 1.3516173331975792, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 53594 + }, + { + "epoch": 0.53595, + "grad_norm": 1.9473657254498018, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 53595 + }, + { + "epoch": 0.53596, + "grad_norm": 1.2192844323949246, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 53596 + }, + { + "epoch": 0.53597, + "grad_norm": 1.5910755815797621, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 53597 + }, + { + "epoch": 0.53598, + "grad_norm": 1.4479317924955506, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 53598 + }, + { + "epoch": 0.53599, + "grad_norm": 1.473382322167697, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 53599 + }, + { + "epoch": 0.536, + "grad_norm": 1.4776299061125784, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 53600 + }, + { + "epoch": 0.53601, + "grad_norm": 1.093493583427605, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 53601 + }, + { + "epoch": 0.53602, + "grad_norm": 1.7878492629214349, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 53602 + }, + { + "epoch": 0.53603, + "grad_norm": 1.7947056597882824, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 53603 + }, + { + "epoch": 0.53604, + "grad_norm": 1.5560790679549, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 53604 + }, + { + "epoch": 0.53605, + "grad_norm": 1.615905523814953, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53605 + }, + { + "epoch": 0.53606, + "grad_norm": 1.5747383258614, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 53606 + }, + { + "epoch": 0.53607, + "grad_norm": 1.5555483311697011, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 53607 + }, + { + "epoch": 0.53608, + "grad_norm": 1.511582161478779, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 53608 + }, + { + "epoch": 0.53609, + "grad_norm": 1.6847894294191115, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 53609 + }, + { + "epoch": 0.5361, + "grad_norm": 1.454818495792913, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53610 + }, + { + "epoch": 0.53611, + "grad_norm": 1.6736709862572712, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 53611 + }, + { + "epoch": 0.53612, + "grad_norm": 1.5876025870413426, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 53612 + }, + { + "epoch": 0.53613, + "grad_norm": 1.5296267201778166, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 53613 + }, + { + "epoch": 0.53614, + "grad_norm": 1.742228247509381, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 53614 + }, + { + "epoch": 0.53615, + "grad_norm": 1.3404984778321214, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 53615 + }, + { + "epoch": 0.53616, + "grad_norm": 1.6163312562064298, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 53616 + }, + { + "epoch": 0.53617, + "grad_norm": 1.5647128714989773, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 53617 + }, + { + "epoch": 0.53618, + "grad_norm": 1.5452169043836763, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 53618 + }, + { + "epoch": 0.53619, + "grad_norm": 1.537531861964667, + "learning_rate": 0.003, + "loss": 4.042, + "step": 53619 + }, + { + "epoch": 0.5362, + "grad_norm": 1.3364395390963664, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 53620 + }, + { + "epoch": 0.53621, + "grad_norm": 1.6966616575359101, + "learning_rate": 0.003, + "loss": 4.026, + "step": 53621 + }, + { + "epoch": 0.53622, + "grad_norm": 1.5915208547221333, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 53622 + }, + { + "epoch": 0.53623, + "grad_norm": 1.722401064807915, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 53623 + }, + { + "epoch": 0.53624, + "grad_norm": 2.056047367870351, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 53624 + }, + { + "epoch": 0.53625, + "grad_norm": 1.278439570717565, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 53625 + }, + { + "epoch": 0.53626, + "grad_norm": 1.907532249276817, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 53626 + }, + { + "epoch": 0.53627, + "grad_norm": 1.5276057199649873, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 53627 + }, + { + "epoch": 0.53628, + "grad_norm": 1.7186963878491295, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 53628 + }, + { + "epoch": 0.53629, + "grad_norm": 1.5289261171312167, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 53629 + }, + { + "epoch": 0.5363, + "grad_norm": 1.2517983326381177, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 53630 + }, + { + "epoch": 0.53631, + "grad_norm": 2.0859020690005385, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 53631 + }, + { + "epoch": 0.53632, + "grad_norm": 1.3509722692307136, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 53632 + }, + { + "epoch": 0.53633, + "grad_norm": 1.6024966918187997, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 53633 + }, + { + "epoch": 0.53634, + "grad_norm": 1.6650771043769863, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 53634 + }, + { + "epoch": 0.53635, + "grad_norm": 1.4001486932475131, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 53635 + }, + { + "epoch": 0.53636, + "grad_norm": 1.4894593667374878, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 53636 + }, + { + "epoch": 0.53637, + "grad_norm": 1.2077886812052803, + "learning_rate": 0.003, + "loss": 4.041, + "step": 53637 + }, + { + "epoch": 0.53638, + "grad_norm": 1.8660371150653614, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 53638 + }, + { + "epoch": 0.53639, + "grad_norm": 1.3568803085735694, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 53639 + }, + { + "epoch": 0.5364, + "grad_norm": 1.5788998385359507, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 53640 + }, + { + "epoch": 0.53641, + "grad_norm": 1.4966508256860627, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 53641 + }, + { + "epoch": 0.53642, + "grad_norm": 1.645327035147196, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 53642 + }, + { + "epoch": 0.53643, + "grad_norm": 1.697002248776963, + "learning_rate": 0.003, + "loss": 4.04, + "step": 53643 + }, + { + "epoch": 0.53644, + "grad_norm": 1.4229099808548222, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 53644 + }, + { + "epoch": 0.53645, + "grad_norm": 1.6386618461898055, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 53645 + }, + { + "epoch": 0.53646, + "grad_norm": 1.3968742590011471, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 53646 + }, + { + "epoch": 0.53647, + "grad_norm": 1.8519443089452776, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 53647 + }, + { + "epoch": 0.53648, + "grad_norm": 1.5966003809932579, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 53648 + }, + { + "epoch": 0.53649, + "grad_norm": 1.4552034537100578, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 53649 + }, + { + "epoch": 0.5365, + "grad_norm": 1.7925773552098927, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 53650 + }, + { + "epoch": 0.53651, + "grad_norm": 1.2883488156480007, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 53651 + }, + { + "epoch": 0.53652, + "grad_norm": 1.621166142049974, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 53652 + }, + { + "epoch": 0.53653, + "grad_norm": 1.5693032119111459, + "learning_rate": 0.003, + "loss": 4.015, + "step": 53653 + }, + { + "epoch": 0.53654, + "grad_norm": 1.2766116175121984, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 53654 + }, + { + "epoch": 0.53655, + "grad_norm": 1.772259990760207, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 53655 + }, + { + "epoch": 0.53656, + "grad_norm": 1.5177958758643997, + "learning_rate": 0.003, + "loss": 4.035, + "step": 53656 + }, + { + "epoch": 0.53657, + "grad_norm": 1.5236339183882042, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 53657 + }, + { + "epoch": 0.53658, + "grad_norm": 1.5693044655739135, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 53658 + }, + { + "epoch": 0.53659, + "grad_norm": 1.5391309640044573, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 53659 + }, + { + "epoch": 0.5366, + "grad_norm": 1.7649588551460018, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 53660 + }, + { + "epoch": 0.53661, + "grad_norm": 1.402289844259078, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 53661 + }, + { + "epoch": 0.53662, + "grad_norm": 1.9078291214033356, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 53662 + }, + { + "epoch": 0.53663, + "grad_norm": 1.3009070601069062, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 53663 + }, + { + "epoch": 0.53664, + "grad_norm": 1.9690961544587993, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 53664 + }, + { + "epoch": 0.53665, + "grad_norm": 1.3220516728057488, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 53665 + }, + { + "epoch": 0.53666, + "grad_norm": 1.8055952244906424, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 53666 + }, + { + "epoch": 0.53667, + "grad_norm": 1.5303208194486573, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 53667 + }, + { + "epoch": 0.53668, + "grad_norm": 1.6189736749708397, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 53668 + }, + { + "epoch": 0.53669, + "grad_norm": 1.7651888926520052, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 53669 + }, + { + "epoch": 0.5367, + "grad_norm": 1.4054790071954022, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 53670 + }, + { + "epoch": 0.53671, + "grad_norm": 1.546798393286222, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 53671 + }, + { + "epoch": 0.53672, + "grad_norm": 1.6439440606245872, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 53672 + }, + { + "epoch": 0.53673, + "grad_norm": 1.4232707958685562, + "learning_rate": 0.003, + "loss": 4.053, + "step": 53673 + }, + { + "epoch": 0.53674, + "grad_norm": 1.4683401408531098, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 53674 + }, + { + "epoch": 0.53675, + "grad_norm": 1.8915340787835782, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 53675 + }, + { + "epoch": 0.53676, + "grad_norm": 1.35147999568109, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 53676 + }, + { + "epoch": 0.53677, + "grad_norm": 1.7253238041318186, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 53677 + }, + { + "epoch": 0.53678, + "grad_norm": 1.3598209105725736, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53678 + }, + { + "epoch": 0.53679, + "grad_norm": 1.489258091479862, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 53679 + }, + { + "epoch": 0.5368, + "grad_norm": 1.3561382160284743, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53680 + }, + { + "epoch": 0.53681, + "grad_norm": 1.612891383728352, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 53681 + }, + { + "epoch": 0.53682, + "grad_norm": 1.511345353001791, + "learning_rate": 0.003, + "loss": 4.001, + "step": 53682 + }, + { + "epoch": 0.53683, + "grad_norm": 2.2421718336677787, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 53683 + }, + { + "epoch": 0.53684, + "grad_norm": 1.2435749698906502, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 53684 + }, + { + "epoch": 0.53685, + "grad_norm": 2.138387547492581, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 53685 + }, + { + "epoch": 0.53686, + "grad_norm": 1.6874310185884935, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 53686 + }, + { + "epoch": 0.53687, + "grad_norm": 1.5255476591913912, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 53687 + }, + { + "epoch": 0.53688, + "grad_norm": 1.7603120155858023, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 53688 + }, + { + "epoch": 0.53689, + "grad_norm": 1.4257389618306315, + "learning_rate": 0.003, + "loss": 4.028, + "step": 53689 + }, + { + "epoch": 0.5369, + "grad_norm": 1.2461104737691548, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 53690 + }, + { + "epoch": 0.53691, + "grad_norm": 1.6237901397052041, + "learning_rate": 0.003, + "loss": 4.044, + "step": 53691 + }, + { + "epoch": 0.53692, + "grad_norm": 1.5041662260762882, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 53692 + }, + { + "epoch": 0.53693, + "grad_norm": 1.6635668110269795, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 53693 + }, + { + "epoch": 0.53694, + "grad_norm": 1.3335907575556816, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 53694 + }, + { + "epoch": 0.53695, + "grad_norm": 2.0661269027465123, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 53695 + }, + { + "epoch": 0.53696, + "grad_norm": 1.4529132089337156, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 53696 + }, + { + "epoch": 0.53697, + "grad_norm": 1.2797794158370004, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53697 + }, + { + "epoch": 0.53698, + "grad_norm": 1.5376403704538684, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 53698 + }, + { + "epoch": 0.53699, + "grad_norm": 1.9008560274193518, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 53699 + }, + { + "epoch": 0.537, + "grad_norm": 2.1882475976226434, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 53700 + }, + { + "epoch": 0.53701, + "grad_norm": 1.336057291177142, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 53701 + }, + { + "epoch": 0.53702, + "grad_norm": 1.6547282466251365, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 53702 + }, + { + "epoch": 0.53703, + "grad_norm": 1.4961537067997859, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 53703 + }, + { + "epoch": 0.53704, + "grad_norm": 1.8349403184044728, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 53704 + }, + { + "epoch": 0.53705, + "grad_norm": 1.4238324343578306, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 53705 + }, + { + "epoch": 0.53706, + "grad_norm": 1.427613963187147, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 53706 + }, + { + "epoch": 0.53707, + "grad_norm": 1.5137170091830439, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 53707 + }, + { + "epoch": 0.53708, + "grad_norm": 1.841819492112519, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 53708 + }, + { + "epoch": 0.53709, + "grad_norm": 1.4206855723437581, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 53709 + }, + { + "epoch": 0.5371, + "grad_norm": 1.9359135302212134, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 53710 + }, + { + "epoch": 0.53711, + "grad_norm": 1.3646358819651232, + "learning_rate": 0.003, + "loss": 4.048, + "step": 53711 + }, + { + "epoch": 0.53712, + "grad_norm": 1.3991775605967454, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 53712 + }, + { + "epoch": 0.53713, + "grad_norm": 1.8343178797434079, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 53713 + }, + { + "epoch": 0.53714, + "grad_norm": 1.6188997576692339, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 53714 + }, + { + "epoch": 0.53715, + "grad_norm": 1.4564226117257473, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 53715 + }, + { + "epoch": 0.53716, + "grad_norm": 1.5858686270023103, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 53716 + }, + { + "epoch": 0.53717, + "grad_norm": 1.1851323250452788, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 53717 + }, + { + "epoch": 0.53718, + "grad_norm": 1.8430949633477627, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 53718 + }, + { + "epoch": 0.53719, + "grad_norm": 1.2314638537640845, + "learning_rate": 0.003, + "loss": 4.037, + "step": 53719 + }, + { + "epoch": 0.5372, + "grad_norm": 1.924046027468344, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 53720 + }, + { + "epoch": 0.53721, + "grad_norm": 1.5236905603468447, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 53721 + }, + { + "epoch": 0.53722, + "grad_norm": 1.7227893384092914, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 53722 + }, + { + "epoch": 0.53723, + "grad_norm": 1.5515157122296193, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 53723 + }, + { + "epoch": 0.53724, + "grad_norm": 1.5297996399765932, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 53724 + }, + { + "epoch": 0.53725, + "grad_norm": 1.8005287564669312, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 53725 + }, + { + "epoch": 0.53726, + "grad_norm": 1.5637713249394238, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 53726 + }, + { + "epoch": 0.53727, + "grad_norm": 1.6561663108894775, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 53727 + }, + { + "epoch": 0.53728, + "grad_norm": 1.5751887149423913, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 53728 + }, + { + "epoch": 0.53729, + "grad_norm": 1.507673728915949, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 53729 + }, + { + "epoch": 0.5373, + "grad_norm": 1.716779405412689, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 53730 + }, + { + "epoch": 0.53731, + "grad_norm": 1.4615018416804073, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 53731 + }, + { + "epoch": 0.53732, + "grad_norm": 1.3887700716150413, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 53732 + }, + { + "epoch": 0.53733, + "grad_norm": 1.554734342445412, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 53733 + }, + { + "epoch": 0.53734, + "grad_norm": 1.5770691286011922, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 53734 + }, + { + "epoch": 0.53735, + "grad_norm": 1.2446871610209924, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 53735 + }, + { + "epoch": 0.53736, + "grad_norm": 1.7083005153834634, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 53736 + }, + { + "epoch": 0.53737, + "grad_norm": 1.2502096023604734, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 53737 + }, + { + "epoch": 0.53738, + "grad_norm": 2.043787744503767, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 53738 + }, + { + "epoch": 0.53739, + "grad_norm": 1.3243961032435734, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 53739 + }, + { + "epoch": 0.5374, + "grad_norm": 1.6445154250946297, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 53740 + }, + { + "epoch": 0.53741, + "grad_norm": 1.8064949153745904, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 53741 + }, + { + "epoch": 0.53742, + "grad_norm": 1.8152633660361355, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 53742 + }, + { + "epoch": 0.53743, + "grad_norm": 1.7089755412988816, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 53743 + }, + { + "epoch": 0.53744, + "grad_norm": 1.332297818448147, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 53744 + }, + { + "epoch": 0.53745, + "grad_norm": 1.6444649603278998, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 53745 + }, + { + "epoch": 0.53746, + "grad_norm": 1.4444029056295786, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 53746 + }, + { + "epoch": 0.53747, + "grad_norm": 1.6708485968589106, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 53747 + }, + { + "epoch": 0.53748, + "grad_norm": 1.4159057702926017, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 53748 + }, + { + "epoch": 0.53749, + "grad_norm": 1.698363721928134, + "learning_rate": 0.003, + "loss": 4.043, + "step": 53749 + }, + { + "epoch": 0.5375, + "grad_norm": 1.5972264301470214, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 53750 + }, + { + "epoch": 0.53751, + "grad_norm": 1.651355606593782, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 53751 + }, + { + "epoch": 0.53752, + "grad_norm": 1.7035706546082745, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 53752 + }, + { + "epoch": 0.53753, + "grad_norm": 1.351937246937344, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 53753 + }, + { + "epoch": 0.53754, + "grad_norm": 1.644116771336386, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 53754 + }, + { + "epoch": 0.53755, + "grad_norm": 1.5382925035964692, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 53755 + }, + { + "epoch": 0.53756, + "grad_norm": 1.8885865782677604, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 53756 + }, + { + "epoch": 0.53757, + "grad_norm": 1.273878800797696, + "learning_rate": 0.003, + "loss": 4.064, + "step": 53757 + }, + { + "epoch": 0.53758, + "grad_norm": 1.8190447421011702, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 53758 + }, + { + "epoch": 0.53759, + "grad_norm": 1.5550902258917059, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 53759 + }, + { + "epoch": 0.5376, + "grad_norm": 1.3766649734226952, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 53760 + }, + { + "epoch": 0.53761, + "grad_norm": 1.895861493983907, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 53761 + }, + { + "epoch": 0.53762, + "grad_norm": 1.1509740101371677, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 53762 + }, + { + "epoch": 0.53763, + "grad_norm": 1.8721115052272623, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 53763 + }, + { + "epoch": 0.53764, + "grad_norm": 1.237281784843624, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 53764 + }, + { + "epoch": 0.53765, + "grad_norm": 2.0181153539746925, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 53765 + }, + { + "epoch": 0.53766, + "grad_norm": 1.7218745932442117, + "learning_rate": 0.003, + "loss": 4.048, + "step": 53766 + }, + { + "epoch": 0.53767, + "grad_norm": 1.5749155181479377, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 53767 + }, + { + "epoch": 0.53768, + "grad_norm": 1.5038808842179778, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 53768 + }, + { + "epoch": 0.53769, + "grad_norm": 1.7300846710866016, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 53769 + }, + { + "epoch": 0.5377, + "grad_norm": 1.3989820439390008, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 53770 + }, + { + "epoch": 0.53771, + "grad_norm": 1.4606460575339604, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 53771 + }, + { + "epoch": 0.53772, + "grad_norm": 1.5319676244007365, + "learning_rate": 0.003, + "loss": 4.04, + "step": 53772 + }, + { + "epoch": 0.53773, + "grad_norm": 1.7003453285329084, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 53773 + }, + { + "epoch": 0.53774, + "grad_norm": 1.7945376387807923, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 53774 + }, + { + "epoch": 0.53775, + "grad_norm": 1.4684113213521657, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 53775 + }, + { + "epoch": 0.53776, + "grad_norm": 1.5906360065087077, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 53776 + }, + { + "epoch": 0.53777, + "grad_norm": 1.532321357868074, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 53777 + }, + { + "epoch": 0.53778, + "grad_norm": 1.8964229744669154, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 53778 + }, + { + "epoch": 0.53779, + "grad_norm": 1.3602891782530615, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 53779 + }, + { + "epoch": 0.5378, + "grad_norm": 1.6295871060001201, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 53780 + }, + { + "epoch": 0.53781, + "grad_norm": 1.3168969696577792, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 53781 + }, + { + "epoch": 0.53782, + "grad_norm": 1.6275356820969298, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 53782 + }, + { + "epoch": 0.53783, + "grad_norm": 1.5552205931831538, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 53783 + }, + { + "epoch": 0.53784, + "grad_norm": 1.6157968056097858, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 53784 + }, + { + "epoch": 0.53785, + "grad_norm": 1.6217742999115978, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 53785 + }, + { + "epoch": 0.53786, + "grad_norm": 1.4394218907325471, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 53786 + }, + { + "epoch": 0.53787, + "grad_norm": 1.39729645244773, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 53787 + }, + { + "epoch": 0.53788, + "grad_norm": 1.4740568701471841, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 53788 + }, + { + "epoch": 0.53789, + "grad_norm": 1.5211638409315849, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 53789 + }, + { + "epoch": 0.5379, + "grad_norm": 1.4640547643123303, + "learning_rate": 0.003, + "loss": 4.034, + "step": 53790 + }, + { + "epoch": 0.53791, + "grad_norm": 1.6996744898643157, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 53791 + }, + { + "epoch": 0.53792, + "grad_norm": 1.5298116187139654, + "learning_rate": 0.003, + "loss": 4.001, + "step": 53792 + }, + { + "epoch": 0.53793, + "grad_norm": 1.6016988042233093, + "learning_rate": 0.003, + "loss": 4.024, + "step": 53793 + }, + { + "epoch": 0.53794, + "grad_norm": 1.6242959473555112, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 53794 + }, + { + "epoch": 0.53795, + "grad_norm": 1.5605051076941496, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 53795 + }, + { + "epoch": 0.53796, + "grad_norm": 1.5996326342933187, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 53796 + }, + { + "epoch": 0.53797, + "grad_norm": 1.5887834863929304, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 53797 + }, + { + "epoch": 0.53798, + "grad_norm": 1.3464993027061425, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 53798 + }, + { + "epoch": 0.53799, + "grad_norm": 1.6481232285419736, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 53799 + }, + { + "epoch": 0.538, + "grad_norm": 1.484826810898476, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 53800 + }, + { + "epoch": 0.53801, + "grad_norm": 1.6953703767999284, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 53801 + }, + { + "epoch": 0.53802, + "grad_norm": 1.681896192184611, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 53802 + }, + { + "epoch": 0.53803, + "grad_norm": 1.4753174970649956, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 53803 + }, + { + "epoch": 0.53804, + "grad_norm": 2.0126441117438683, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 53804 + }, + { + "epoch": 0.53805, + "grad_norm": 1.462599930095322, + "learning_rate": 0.003, + "loss": 4.026, + "step": 53805 + }, + { + "epoch": 0.53806, + "grad_norm": 1.9801538691597425, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 53806 + }, + { + "epoch": 0.53807, + "grad_norm": 1.4233087986898254, + "learning_rate": 0.003, + "loss": 4.027, + "step": 53807 + }, + { + "epoch": 0.53808, + "grad_norm": 1.6880742840202376, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 53808 + }, + { + "epoch": 0.53809, + "grad_norm": 1.4445046463642617, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 53809 + }, + { + "epoch": 0.5381, + "grad_norm": 1.7013226223269795, + "learning_rate": 0.003, + "loss": 4.029, + "step": 53810 + }, + { + "epoch": 0.53811, + "grad_norm": 1.5043630377070876, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 53811 + }, + { + "epoch": 0.53812, + "grad_norm": 1.4975771579838106, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 53812 + }, + { + "epoch": 0.53813, + "grad_norm": 1.7651458753699192, + "learning_rate": 0.003, + "loss": 3.988, + "step": 53813 + }, + { + "epoch": 0.53814, + "grad_norm": 1.1727351780079347, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 53814 + }, + { + "epoch": 0.53815, + "grad_norm": 1.5923541069308857, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 53815 + }, + { + "epoch": 0.53816, + "grad_norm": 1.4396763870543068, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 53816 + }, + { + "epoch": 0.53817, + "grad_norm": 1.5793601249835107, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 53817 + }, + { + "epoch": 0.53818, + "grad_norm": 1.3030476926870345, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 53818 + }, + { + "epoch": 0.53819, + "grad_norm": 1.8610916678010778, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 53819 + }, + { + "epoch": 0.5382, + "grad_norm": 1.3340739614926387, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 53820 + }, + { + "epoch": 0.53821, + "grad_norm": 1.4478035234631348, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 53821 + }, + { + "epoch": 0.53822, + "grad_norm": 1.5565290454933374, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 53822 + }, + { + "epoch": 0.53823, + "grad_norm": 1.6560121770411427, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 53823 + }, + { + "epoch": 0.53824, + "grad_norm": 1.4369436010218974, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 53824 + }, + { + "epoch": 0.53825, + "grad_norm": 1.7835295240175, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 53825 + }, + { + "epoch": 0.53826, + "grad_norm": 1.8241340892353408, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 53826 + }, + { + "epoch": 0.53827, + "grad_norm": 1.412176672265658, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 53827 + }, + { + "epoch": 0.53828, + "grad_norm": 1.7244993303198057, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 53828 + }, + { + "epoch": 0.53829, + "grad_norm": 1.4544201536737018, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 53829 + }, + { + "epoch": 0.5383, + "grad_norm": 1.7750613959498265, + "learning_rate": 0.003, + "loss": 4.025, + "step": 53830 + }, + { + "epoch": 0.53831, + "grad_norm": 1.313826558969305, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 53831 + }, + { + "epoch": 0.53832, + "grad_norm": 1.6808878628886763, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 53832 + }, + { + "epoch": 0.53833, + "grad_norm": 1.3011543604100484, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 53833 + }, + { + "epoch": 0.53834, + "grad_norm": 1.9282929924099024, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 53834 + }, + { + "epoch": 0.53835, + "grad_norm": 1.531982487556883, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 53835 + }, + { + "epoch": 0.53836, + "grad_norm": 1.5889799811920688, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 53836 + }, + { + "epoch": 0.53837, + "grad_norm": 1.4976766325745996, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 53837 + }, + { + "epoch": 0.53838, + "grad_norm": 1.4885719700026647, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 53838 + }, + { + "epoch": 0.53839, + "grad_norm": 2.072610429003309, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 53839 + }, + { + "epoch": 0.5384, + "grad_norm": 1.3202345914520757, + "learning_rate": 0.003, + "loss": 4.006, + "step": 53840 + }, + { + "epoch": 0.53841, + "grad_norm": 1.4789863502177811, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 53841 + }, + { + "epoch": 0.53842, + "grad_norm": 1.8215502527332592, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 53842 + }, + { + "epoch": 0.53843, + "grad_norm": 1.5407614893132722, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 53843 + }, + { + "epoch": 0.53844, + "grad_norm": 1.504138062899459, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 53844 + }, + { + "epoch": 0.53845, + "grad_norm": 1.6059899338950283, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 53845 + }, + { + "epoch": 0.53846, + "grad_norm": 1.5006607655940642, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 53846 + }, + { + "epoch": 0.53847, + "grad_norm": 1.3411070364727746, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 53847 + }, + { + "epoch": 0.53848, + "grad_norm": 1.6144186035993762, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 53848 + }, + { + "epoch": 0.53849, + "grad_norm": 1.3666147778568767, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 53849 + }, + { + "epoch": 0.5385, + "grad_norm": 1.769991560401641, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 53850 + }, + { + "epoch": 0.53851, + "grad_norm": 1.3243345194726286, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 53851 + }, + { + "epoch": 0.53852, + "grad_norm": 1.5476340923274179, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 53852 + }, + { + "epoch": 0.53853, + "grad_norm": 1.6359155763838462, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 53853 + }, + { + "epoch": 0.53854, + "grad_norm": 1.6927931034313812, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 53854 + }, + { + "epoch": 0.53855, + "grad_norm": 1.972356847839806, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 53855 + }, + { + "epoch": 0.53856, + "grad_norm": 2.1474483555951713, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 53856 + }, + { + "epoch": 0.53857, + "grad_norm": 1.2386146783997327, + "learning_rate": 0.003, + "loss": 4.032, + "step": 53857 + }, + { + "epoch": 0.53858, + "grad_norm": 1.88316868380961, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 53858 + }, + { + "epoch": 0.53859, + "grad_norm": 1.6510478308863927, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 53859 + }, + { + "epoch": 0.5386, + "grad_norm": 1.5170240537456134, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 53860 + }, + { + "epoch": 0.53861, + "grad_norm": 1.6126910497169817, + "learning_rate": 0.003, + "loss": 4.05, + "step": 53861 + }, + { + "epoch": 0.53862, + "grad_norm": 1.8401246096857238, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 53862 + }, + { + "epoch": 0.53863, + "grad_norm": 1.3894435169144825, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 53863 + }, + { + "epoch": 0.53864, + "grad_norm": 1.603200083905397, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 53864 + }, + { + "epoch": 0.53865, + "grad_norm": 1.8045280643994615, + "learning_rate": 0.003, + "loss": 4.034, + "step": 53865 + }, + { + "epoch": 0.53866, + "grad_norm": 1.5722083138935945, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 53866 + }, + { + "epoch": 0.53867, + "grad_norm": 1.4537112203122522, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 53867 + }, + { + "epoch": 0.53868, + "grad_norm": 1.6253627681013243, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 53868 + }, + { + "epoch": 0.53869, + "grad_norm": 1.336372256127655, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 53869 + }, + { + "epoch": 0.5387, + "grad_norm": 1.7770531881382015, + "learning_rate": 0.003, + "loss": 4.05, + "step": 53870 + }, + { + "epoch": 0.53871, + "grad_norm": 1.249844179328197, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 53871 + }, + { + "epoch": 0.53872, + "grad_norm": 1.574572344483978, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 53872 + }, + { + "epoch": 0.53873, + "grad_norm": 1.5989884798422018, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 53873 + }, + { + "epoch": 0.53874, + "grad_norm": 1.53423451259967, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 53874 + }, + { + "epoch": 0.53875, + "grad_norm": 2.1047006572992606, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 53875 + }, + { + "epoch": 0.53876, + "grad_norm": 1.5811444104906305, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 53876 + }, + { + "epoch": 0.53877, + "grad_norm": 1.6464930610353798, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 53877 + }, + { + "epoch": 0.53878, + "grad_norm": 1.4596960084247592, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 53878 + }, + { + "epoch": 0.53879, + "grad_norm": 1.746475278950323, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 53879 + }, + { + "epoch": 0.5388, + "grad_norm": 1.2690846981372792, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 53880 + }, + { + "epoch": 0.53881, + "grad_norm": 1.7964789621487687, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53881 + }, + { + "epoch": 0.53882, + "grad_norm": 1.561228564312726, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 53882 + }, + { + "epoch": 0.53883, + "grad_norm": 1.4928734679078202, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53883 + }, + { + "epoch": 0.53884, + "grad_norm": 1.6530170165140228, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 53884 + }, + { + "epoch": 0.53885, + "grad_norm": 1.4112094492598712, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 53885 + }, + { + "epoch": 0.53886, + "grad_norm": 1.528721248147578, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 53886 + }, + { + "epoch": 0.53887, + "grad_norm": 1.3662294813285305, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 53887 + }, + { + "epoch": 0.53888, + "grad_norm": 1.633216719186148, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 53888 + }, + { + "epoch": 0.53889, + "grad_norm": 1.513573752113875, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 53889 + }, + { + "epoch": 0.5389, + "grad_norm": 1.8774156225321845, + "learning_rate": 0.003, + "loss": 4.039, + "step": 53890 + }, + { + "epoch": 0.53891, + "grad_norm": 1.3092399880404089, + "learning_rate": 0.003, + "loss": 4.008, + "step": 53891 + }, + { + "epoch": 0.53892, + "grad_norm": 1.675696530072826, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 53892 + }, + { + "epoch": 0.53893, + "grad_norm": 1.5533766659799764, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 53893 + }, + { + "epoch": 0.53894, + "grad_norm": 1.7000571230553536, + "learning_rate": 0.003, + "loss": 4.007, + "step": 53894 + }, + { + "epoch": 0.53895, + "grad_norm": 1.5386775788340998, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 53895 + }, + { + "epoch": 0.53896, + "grad_norm": 1.5018106234667497, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 53896 + }, + { + "epoch": 0.53897, + "grad_norm": 1.5822585588170053, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 53897 + }, + { + "epoch": 0.53898, + "grad_norm": 1.8161910311369196, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 53898 + }, + { + "epoch": 0.53899, + "grad_norm": 1.3160744452237008, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 53899 + }, + { + "epoch": 0.539, + "grad_norm": 1.5825115752940921, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 53900 + }, + { + "epoch": 0.53901, + "grad_norm": 1.392404832714326, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 53901 + }, + { + "epoch": 0.53902, + "grad_norm": 1.530299375005297, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 53902 + }, + { + "epoch": 0.53903, + "grad_norm": 1.5212825682500846, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 53903 + }, + { + "epoch": 0.53904, + "grad_norm": 1.5662314045604389, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 53904 + }, + { + "epoch": 0.53905, + "grad_norm": 1.7140470123313771, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 53905 + }, + { + "epoch": 0.53906, + "grad_norm": 1.4757014861161117, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 53906 + }, + { + "epoch": 0.53907, + "grad_norm": 1.60822800569548, + "learning_rate": 0.003, + "loss": 4.049, + "step": 53907 + }, + { + "epoch": 0.53908, + "grad_norm": 1.5329682958087012, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 53908 + }, + { + "epoch": 0.53909, + "grad_norm": 1.9031924668291402, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 53909 + }, + { + "epoch": 0.5391, + "grad_norm": 1.4316993383586851, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 53910 + }, + { + "epoch": 0.53911, + "grad_norm": 1.7502641823145366, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 53911 + }, + { + "epoch": 0.53912, + "grad_norm": 1.301895269790338, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 53912 + }, + { + "epoch": 0.53913, + "grad_norm": 1.8140506769702713, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 53913 + }, + { + "epoch": 0.53914, + "grad_norm": 1.6756103758276477, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 53914 + }, + { + "epoch": 0.53915, + "grad_norm": 1.7643896935352927, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 53915 + }, + { + "epoch": 0.53916, + "grad_norm": 1.51427782602509, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 53916 + }, + { + "epoch": 0.53917, + "grad_norm": 1.6773099182495825, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 53917 + }, + { + "epoch": 0.53918, + "grad_norm": 1.4944179128264112, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 53918 + }, + { + "epoch": 0.53919, + "grad_norm": 1.5536488209373873, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 53919 + }, + { + "epoch": 0.5392, + "grad_norm": 1.4673735829614252, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 53920 + }, + { + "epoch": 0.53921, + "grad_norm": 1.7063491995972084, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 53921 + }, + { + "epoch": 0.53922, + "grad_norm": 1.4954065983467895, + "learning_rate": 0.003, + "loss": 4.027, + "step": 53922 + }, + { + "epoch": 0.53923, + "grad_norm": 1.503283285428641, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 53923 + }, + { + "epoch": 0.53924, + "grad_norm": 1.41787108923306, + "learning_rate": 0.003, + "loss": 4.009, + "step": 53924 + }, + { + "epoch": 0.53925, + "grad_norm": 1.7214328476609597, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 53925 + }, + { + "epoch": 0.53926, + "grad_norm": 1.431177937331995, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 53926 + }, + { + "epoch": 0.53927, + "grad_norm": 1.899880055273053, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 53927 + }, + { + "epoch": 0.53928, + "grad_norm": 1.3469858503630532, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 53928 + }, + { + "epoch": 0.53929, + "grad_norm": 2.0581957052504913, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 53929 + }, + { + "epoch": 0.5393, + "grad_norm": 1.5466216153471901, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 53930 + }, + { + "epoch": 0.53931, + "grad_norm": 1.712195136114229, + "learning_rate": 0.003, + "loss": 4.037, + "step": 53931 + }, + { + "epoch": 0.53932, + "grad_norm": 1.2570460542494573, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 53932 + }, + { + "epoch": 0.53933, + "grad_norm": 1.789371844887893, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 53933 + }, + { + "epoch": 0.53934, + "grad_norm": 1.440591218786328, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 53934 + }, + { + "epoch": 0.53935, + "grad_norm": 1.5361569109707904, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 53935 + }, + { + "epoch": 0.53936, + "grad_norm": 1.3723944234294523, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 53936 + }, + { + "epoch": 0.53937, + "grad_norm": 1.667050192512472, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 53937 + }, + { + "epoch": 0.53938, + "grad_norm": 1.736068318474724, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 53938 + }, + { + "epoch": 0.53939, + "grad_norm": 1.6580685467325267, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 53939 + }, + { + "epoch": 0.5394, + "grad_norm": 1.7345565826925013, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 53940 + }, + { + "epoch": 0.53941, + "grad_norm": 1.3325074432988266, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 53941 + }, + { + "epoch": 0.53942, + "grad_norm": 1.6413128010112261, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 53942 + }, + { + "epoch": 0.53943, + "grad_norm": 1.571545987623805, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 53943 + }, + { + "epoch": 0.53944, + "grad_norm": 1.271120366078819, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 53944 + }, + { + "epoch": 0.53945, + "grad_norm": 1.658377807337409, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 53945 + }, + { + "epoch": 0.53946, + "grad_norm": 1.5069517820837155, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 53946 + }, + { + "epoch": 0.53947, + "grad_norm": 1.7664465614122533, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 53947 + }, + { + "epoch": 0.53948, + "grad_norm": 1.2656018514105443, + "learning_rate": 0.003, + "loss": 4.043, + "step": 53948 + }, + { + "epoch": 0.53949, + "grad_norm": 1.6516139617485364, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 53949 + }, + { + "epoch": 0.5395, + "grad_norm": 1.676157471285838, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 53950 + }, + { + "epoch": 0.53951, + "grad_norm": 1.9449561881275246, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 53951 + }, + { + "epoch": 0.53952, + "grad_norm": 1.4611890535762493, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 53952 + }, + { + "epoch": 0.53953, + "grad_norm": 1.39724321087621, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 53953 + }, + { + "epoch": 0.53954, + "grad_norm": 2.264057148222176, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 53954 + }, + { + "epoch": 0.53955, + "grad_norm": 1.3994346410564344, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 53955 + }, + { + "epoch": 0.53956, + "grad_norm": 1.4434270752972496, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 53956 + }, + { + "epoch": 0.53957, + "grad_norm": 1.8088046128654642, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 53957 + }, + { + "epoch": 0.53958, + "grad_norm": 1.5957812772019642, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 53958 + }, + { + "epoch": 0.53959, + "grad_norm": 1.507214508832078, + "learning_rate": 0.003, + "loss": 4.022, + "step": 53959 + }, + { + "epoch": 0.5396, + "grad_norm": 1.544226351170576, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 53960 + }, + { + "epoch": 0.53961, + "grad_norm": 1.499251108964462, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 53961 + }, + { + "epoch": 0.53962, + "grad_norm": 1.6003434643760142, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 53962 + }, + { + "epoch": 0.53963, + "grad_norm": 1.6204465738876412, + "learning_rate": 0.003, + "loss": 4.021, + "step": 53963 + }, + { + "epoch": 0.53964, + "grad_norm": 1.3575538746372973, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 53964 + }, + { + "epoch": 0.53965, + "grad_norm": 1.4238513441285963, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 53965 + }, + { + "epoch": 0.53966, + "grad_norm": 1.4755426962104834, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 53966 + }, + { + "epoch": 0.53967, + "grad_norm": 1.6816910733974704, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 53967 + }, + { + "epoch": 0.53968, + "grad_norm": 1.45268561275283, + "learning_rate": 0.003, + "loss": 4.023, + "step": 53968 + }, + { + "epoch": 0.53969, + "grad_norm": 1.5621598088030557, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 53969 + }, + { + "epoch": 0.5397, + "grad_norm": 1.4877016969153742, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 53970 + }, + { + "epoch": 0.53971, + "grad_norm": 1.5651972086884638, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 53971 + }, + { + "epoch": 0.53972, + "grad_norm": 1.4479618921982151, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 53972 + }, + { + "epoch": 0.53973, + "grad_norm": 1.568625807946879, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 53973 + }, + { + "epoch": 0.53974, + "grad_norm": 1.497444793728139, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 53974 + }, + { + "epoch": 0.53975, + "grad_norm": 1.3847630055186013, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 53975 + }, + { + "epoch": 0.53976, + "grad_norm": 1.6690668769690717, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 53976 + }, + { + "epoch": 0.53977, + "grad_norm": 1.5048560578984305, + "learning_rate": 0.003, + "loss": 4.015, + "step": 53977 + }, + { + "epoch": 0.53978, + "grad_norm": 1.8042763382017026, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 53978 + }, + { + "epoch": 0.53979, + "grad_norm": 2.0958332307603524, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 53979 + }, + { + "epoch": 0.5398, + "grad_norm": 1.4466838469845702, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 53980 + }, + { + "epoch": 0.53981, + "grad_norm": 1.6475607242097272, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 53981 + }, + { + "epoch": 0.53982, + "grad_norm": 1.7018914098487838, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 53982 + }, + { + "epoch": 0.53983, + "grad_norm": 1.9568590041810574, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 53983 + }, + { + "epoch": 0.53984, + "grad_norm": 1.2038717421294276, + "learning_rate": 0.003, + "loss": 4.009, + "step": 53984 + }, + { + "epoch": 0.53985, + "grad_norm": 1.6877428790658162, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 53985 + }, + { + "epoch": 0.53986, + "grad_norm": 1.448242321472981, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 53986 + }, + { + "epoch": 0.53987, + "grad_norm": 1.9764013577152046, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 53987 + }, + { + "epoch": 0.53988, + "grad_norm": 1.2711452896917625, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 53988 + }, + { + "epoch": 0.53989, + "grad_norm": 1.5873276862944627, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 53989 + }, + { + "epoch": 0.5399, + "grad_norm": 1.5463417565408122, + "learning_rate": 0.003, + "loss": 4.033, + "step": 53990 + }, + { + "epoch": 0.53991, + "grad_norm": 1.7801186039268402, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 53991 + }, + { + "epoch": 0.53992, + "grad_norm": 1.8225899358147193, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 53992 + }, + { + "epoch": 0.53993, + "grad_norm": 1.6997163662209251, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 53993 + }, + { + "epoch": 0.53994, + "grad_norm": 1.69437506052091, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 53994 + }, + { + "epoch": 0.53995, + "grad_norm": 1.2905437876945143, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 53995 + }, + { + "epoch": 0.53996, + "grad_norm": 1.9436379918197944, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 53996 + }, + { + "epoch": 0.53997, + "grad_norm": 1.2888355851773021, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 53997 + }, + { + "epoch": 0.53998, + "grad_norm": 1.4224782251882182, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 53998 + }, + { + "epoch": 0.53999, + "grad_norm": 1.683082923648302, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 53999 + }, + { + "epoch": 0.54, + "grad_norm": 1.5428109489126598, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 54000 + }, + { + "epoch": 0.54001, + "grad_norm": 1.8791818356612782, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 54001 + }, + { + "epoch": 0.54002, + "grad_norm": 1.622470667355651, + "learning_rate": 0.003, + "loss": 4.012, + "step": 54002 + }, + { + "epoch": 0.54003, + "grad_norm": 1.631037179718342, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 54003 + }, + { + "epoch": 0.54004, + "grad_norm": 1.327840972360089, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 54004 + }, + { + "epoch": 0.54005, + "grad_norm": 1.465070160737639, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 54005 + }, + { + "epoch": 0.54006, + "grad_norm": 1.4789181300034406, + "learning_rate": 0.003, + "loss": 4.04, + "step": 54006 + }, + { + "epoch": 0.54007, + "grad_norm": 1.3623738760609987, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 54007 + }, + { + "epoch": 0.54008, + "grad_norm": 1.8073825820337681, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 54008 + }, + { + "epoch": 0.54009, + "grad_norm": 1.4810218664284558, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 54009 + }, + { + "epoch": 0.5401, + "grad_norm": 1.9234458916919923, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 54010 + }, + { + "epoch": 0.54011, + "grad_norm": 1.5144413044565488, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 54011 + }, + { + "epoch": 0.54012, + "grad_norm": 1.618016820221912, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 54012 + }, + { + "epoch": 0.54013, + "grad_norm": 1.471091366914103, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 54013 + }, + { + "epoch": 0.54014, + "grad_norm": 1.4136408273160181, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 54014 + }, + { + "epoch": 0.54015, + "grad_norm": 1.664085273482806, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 54015 + }, + { + "epoch": 0.54016, + "grad_norm": 1.405704096108276, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 54016 + }, + { + "epoch": 0.54017, + "grad_norm": 1.7704247574543215, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 54017 + }, + { + "epoch": 0.54018, + "grad_norm": 1.514352033136462, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 54018 + }, + { + "epoch": 0.54019, + "grad_norm": 1.5579198047642657, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 54019 + }, + { + "epoch": 0.5402, + "grad_norm": 1.3566583183784906, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 54020 + }, + { + "epoch": 0.54021, + "grad_norm": 1.758176577938438, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 54021 + }, + { + "epoch": 0.54022, + "grad_norm": 1.3065558813897906, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 54022 + }, + { + "epoch": 0.54023, + "grad_norm": 1.3927916232655708, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 54023 + }, + { + "epoch": 0.54024, + "grad_norm": 1.5996286911969846, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 54024 + }, + { + "epoch": 0.54025, + "grad_norm": 1.519816393339942, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54025 + }, + { + "epoch": 0.54026, + "grad_norm": 1.589706961907297, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 54026 + }, + { + "epoch": 0.54027, + "grad_norm": 1.5439659699044999, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 54027 + }, + { + "epoch": 0.54028, + "grad_norm": 1.7511917762663123, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 54028 + }, + { + "epoch": 0.54029, + "grad_norm": 1.4245731981891825, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 54029 + }, + { + "epoch": 0.5403, + "grad_norm": 1.590766342712729, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 54030 + }, + { + "epoch": 0.54031, + "grad_norm": 1.6754100904951632, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 54031 + }, + { + "epoch": 0.54032, + "grad_norm": 1.4551519495873289, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 54032 + }, + { + "epoch": 0.54033, + "grad_norm": 2.044387369732751, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54033 + }, + { + "epoch": 0.54034, + "grad_norm": 1.2777789923355862, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 54034 + }, + { + "epoch": 0.54035, + "grad_norm": 1.6860995118374373, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 54035 + }, + { + "epoch": 0.54036, + "grad_norm": 1.9504302075434476, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 54036 + }, + { + "epoch": 0.54037, + "grad_norm": 1.60392285288199, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 54037 + }, + { + "epoch": 0.54038, + "grad_norm": 1.5267510918237972, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 54038 + }, + { + "epoch": 0.54039, + "grad_norm": 1.5529885527787848, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 54039 + }, + { + "epoch": 0.5404, + "grad_norm": 1.5056414694267428, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 54040 + }, + { + "epoch": 0.54041, + "grad_norm": 1.9226215950387453, + "learning_rate": 0.003, + "loss": 4.031, + "step": 54041 + }, + { + "epoch": 0.54042, + "grad_norm": 1.1729089983716332, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 54042 + }, + { + "epoch": 0.54043, + "grad_norm": 1.7493579101694494, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 54043 + }, + { + "epoch": 0.54044, + "grad_norm": 1.6656386765353866, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 54044 + }, + { + "epoch": 0.54045, + "grad_norm": 1.3343097040236087, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 54045 + }, + { + "epoch": 0.54046, + "grad_norm": 1.3928353881040207, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 54046 + }, + { + "epoch": 0.54047, + "grad_norm": 1.6041904711625286, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 54047 + }, + { + "epoch": 0.54048, + "grad_norm": 1.4927633372011684, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 54048 + }, + { + "epoch": 0.54049, + "grad_norm": 1.562897273075576, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 54049 + }, + { + "epoch": 0.5405, + "grad_norm": 1.4234029455675437, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 54050 + }, + { + "epoch": 0.54051, + "grad_norm": 1.5392121928304079, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 54051 + }, + { + "epoch": 0.54052, + "grad_norm": 2.072098835731799, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 54052 + }, + { + "epoch": 0.54053, + "grad_norm": 1.8153831785870953, + "learning_rate": 0.003, + "loss": 4.032, + "step": 54053 + }, + { + "epoch": 0.54054, + "grad_norm": 1.63554090604296, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 54054 + }, + { + "epoch": 0.54055, + "grad_norm": 1.6604687787146672, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 54055 + }, + { + "epoch": 0.54056, + "grad_norm": 1.3193062693458655, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 54056 + }, + { + "epoch": 0.54057, + "grad_norm": 1.661403178087486, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 54057 + }, + { + "epoch": 0.54058, + "grad_norm": 1.4653730919047652, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 54058 + }, + { + "epoch": 0.54059, + "grad_norm": 1.8276650063068804, + "learning_rate": 0.003, + "loss": 4.047, + "step": 54059 + }, + { + "epoch": 0.5406, + "grad_norm": 1.2987658965851805, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 54060 + }, + { + "epoch": 0.54061, + "grad_norm": 1.6053604820811345, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 54061 + }, + { + "epoch": 0.54062, + "grad_norm": 1.3669705368106655, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54062 + }, + { + "epoch": 0.54063, + "grad_norm": 1.7575307352306049, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54063 + }, + { + "epoch": 0.54064, + "grad_norm": 1.5808814870284211, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 54064 + }, + { + "epoch": 0.54065, + "grad_norm": 1.7666549261803475, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 54065 + }, + { + "epoch": 0.54066, + "grad_norm": 1.8101883175832831, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54066 + }, + { + "epoch": 0.54067, + "grad_norm": 1.348648747184343, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 54067 + }, + { + "epoch": 0.54068, + "grad_norm": 1.9269483604874975, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 54068 + }, + { + "epoch": 0.54069, + "grad_norm": 1.358513124689649, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 54069 + }, + { + "epoch": 0.5407, + "grad_norm": 1.4279379801592451, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 54070 + }, + { + "epoch": 0.54071, + "grad_norm": 1.5496603164649798, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 54071 + }, + { + "epoch": 0.54072, + "grad_norm": 1.7666433449676031, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 54072 + }, + { + "epoch": 0.54073, + "grad_norm": 1.4893063899366235, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 54073 + }, + { + "epoch": 0.54074, + "grad_norm": 1.5649730374608242, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 54074 + }, + { + "epoch": 0.54075, + "grad_norm": 1.6718067391160707, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 54075 + }, + { + "epoch": 0.54076, + "grad_norm": 1.635723654509342, + "learning_rate": 0.003, + "loss": 4.022, + "step": 54076 + }, + { + "epoch": 0.54077, + "grad_norm": 1.3600830008036766, + "learning_rate": 0.003, + "loss": 4.01, + "step": 54077 + }, + { + "epoch": 0.54078, + "grad_norm": 1.82176397751528, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 54078 + }, + { + "epoch": 0.54079, + "grad_norm": 1.3711678525813706, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 54079 + }, + { + "epoch": 0.5408, + "grad_norm": 1.698352836238497, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54080 + }, + { + "epoch": 0.54081, + "grad_norm": 1.3756078322204162, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 54081 + }, + { + "epoch": 0.54082, + "grad_norm": 1.721946846662828, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 54082 + }, + { + "epoch": 0.54083, + "grad_norm": 2.0290683668205056, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 54083 + }, + { + "epoch": 0.54084, + "grad_norm": 1.2261520729454414, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 54084 + }, + { + "epoch": 0.54085, + "grad_norm": 1.7751062466770953, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 54085 + }, + { + "epoch": 0.54086, + "grad_norm": 1.5840576505069877, + "learning_rate": 0.003, + "loss": 4.003, + "step": 54086 + }, + { + "epoch": 0.54087, + "grad_norm": 1.6717942317189927, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 54087 + }, + { + "epoch": 0.54088, + "grad_norm": 1.3929478822793209, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 54088 + }, + { + "epoch": 0.54089, + "grad_norm": 1.6597803443955337, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 54089 + }, + { + "epoch": 0.5409, + "grad_norm": 1.4753587450131438, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 54090 + }, + { + "epoch": 0.54091, + "grad_norm": 1.7101926491332264, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 54091 + }, + { + "epoch": 0.54092, + "grad_norm": 1.2997665244293648, + "learning_rate": 0.003, + "loss": 4.009, + "step": 54092 + }, + { + "epoch": 0.54093, + "grad_norm": 1.647547696121097, + "learning_rate": 0.003, + "loss": 4.038, + "step": 54093 + }, + { + "epoch": 0.54094, + "grad_norm": 1.439878918931944, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 54094 + }, + { + "epoch": 0.54095, + "grad_norm": 1.487117787065009, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 54095 + }, + { + "epoch": 0.54096, + "grad_norm": 1.369837902810512, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 54096 + }, + { + "epoch": 0.54097, + "grad_norm": 1.9322899568678904, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 54097 + }, + { + "epoch": 0.54098, + "grad_norm": 1.6503017910475182, + "learning_rate": 0.003, + "loss": 4.054, + "step": 54098 + }, + { + "epoch": 0.54099, + "grad_norm": 1.886861212727814, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 54099 + }, + { + "epoch": 0.541, + "grad_norm": 1.660126022614681, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54100 + }, + { + "epoch": 0.54101, + "grad_norm": 1.606296105431543, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 54101 + }, + { + "epoch": 0.54102, + "grad_norm": 1.4560011335112213, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54102 + }, + { + "epoch": 0.54103, + "grad_norm": 1.7626558108775505, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 54103 + }, + { + "epoch": 0.54104, + "grad_norm": 1.3166137086799847, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 54104 + }, + { + "epoch": 0.54105, + "grad_norm": 1.4871267805206376, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 54105 + }, + { + "epoch": 0.54106, + "grad_norm": 1.461631687766989, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 54106 + }, + { + "epoch": 0.54107, + "grad_norm": 1.5588376227542962, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 54107 + }, + { + "epoch": 0.54108, + "grad_norm": 1.4474418482846796, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 54108 + }, + { + "epoch": 0.54109, + "grad_norm": 1.752240881614876, + "learning_rate": 0.003, + "loss": 4.035, + "step": 54109 + }, + { + "epoch": 0.5411, + "grad_norm": 1.408005934772432, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 54110 + }, + { + "epoch": 0.54111, + "grad_norm": 1.6220018521407928, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 54111 + }, + { + "epoch": 0.54112, + "grad_norm": 1.211655426551483, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54112 + }, + { + "epoch": 0.54113, + "grad_norm": 1.830155364379755, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 54113 + }, + { + "epoch": 0.54114, + "grad_norm": 1.6084528311441317, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54114 + }, + { + "epoch": 0.54115, + "grad_norm": 1.5723664507605066, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54115 + }, + { + "epoch": 0.54116, + "grad_norm": 1.7928334797555492, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 54116 + }, + { + "epoch": 0.54117, + "grad_norm": 1.516403853975451, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 54117 + }, + { + "epoch": 0.54118, + "grad_norm": 1.7783404301179428, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 54118 + }, + { + "epoch": 0.54119, + "grad_norm": 1.606475970326138, + "learning_rate": 0.003, + "loss": 4.021, + "step": 54119 + }, + { + "epoch": 0.5412, + "grad_norm": 1.6885107159098927, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 54120 + }, + { + "epoch": 0.54121, + "grad_norm": 1.5278124229748995, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 54121 + }, + { + "epoch": 0.54122, + "grad_norm": 1.6608695453733706, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 54122 + }, + { + "epoch": 0.54123, + "grad_norm": 1.5557153840273252, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 54123 + }, + { + "epoch": 0.54124, + "grad_norm": 1.6776780248414527, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 54124 + }, + { + "epoch": 0.54125, + "grad_norm": 1.8484103110534493, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 54125 + }, + { + "epoch": 0.54126, + "grad_norm": 1.2947268125021887, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 54126 + }, + { + "epoch": 0.54127, + "grad_norm": 1.7840102207263842, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 54127 + }, + { + "epoch": 0.54128, + "grad_norm": 1.6188806999105658, + "learning_rate": 0.003, + "loss": 4.058, + "step": 54128 + }, + { + "epoch": 0.54129, + "grad_norm": 1.3886058955829352, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 54129 + }, + { + "epoch": 0.5413, + "grad_norm": 1.438947502858438, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 54130 + }, + { + "epoch": 0.54131, + "grad_norm": 1.5718134060192965, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 54131 + }, + { + "epoch": 0.54132, + "grad_norm": 1.5161052022034462, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 54132 + }, + { + "epoch": 0.54133, + "grad_norm": 1.5555996420455864, + "learning_rate": 0.003, + "loss": 4.011, + "step": 54133 + }, + { + "epoch": 0.54134, + "grad_norm": 1.5441858576557377, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54134 + }, + { + "epoch": 0.54135, + "grad_norm": 1.3981859957301392, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 54135 + }, + { + "epoch": 0.54136, + "grad_norm": 1.747216601267133, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 54136 + }, + { + "epoch": 0.54137, + "grad_norm": 1.234104557768065, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 54137 + }, + { + "epoch": 0.54138, + "grad_norm": 1.717779694490877, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 54138 + }, + { + "epoch": 0.54139, + "grad_norm": 2.0484472795722906, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 54139 + }, + { + "epoch": 0.5414, + "grad_norm": 1.6081655463445261, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 54140 + }, + { + "epoch": 0.54141, + "grad_norm": 1.4765088507394624, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 54141 + }, + { + "epoch": 0.54142, + "grad_norm": 1.7213193510464615, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 54142 + }, + { + "epoch": 0.54143, + "grad_norm": 1.8317729984424473, + "learning_rate": 0.003, + "loss": 4.027, + "step": 54143 + }, + { + "epoch": 0.54144, + "grad_norm": 1.2728971857117337, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 54144 + }, + { + "epoch": 0.54145, + "grad_norm": 1.7482529586777948, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 54145 + }, + { + "epoch": 0.54146, + "grad_norm": 1.3911012646804168, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 54146 + }, + { + "epoch": 0.54147, + "grad_norm": 1.5339199740903269, + "learning_rate": 0.003, + "loss": 4.078, + "step": 54147 + }, + { + "epoch": 0.54148, + "grad_norm": 1.7182836761658944, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 54148 + }, + { + "epoch": 0.54149, + "grad_norm": 1.499332465964261, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 54149 + }, + { + "epoch": 0.5415, + "grad_norm": 1.5326310783819932, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 54150 + }, + { + "epoch": 0.54151, + "grad_norm": 1.6586629703869649, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 54151 + }, + { + "epoch": 0.54152, + "grad_norm": 1.9733927223551642, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 54152 + }, + { + "epoch": 0.54153, + "grad_norm": 1.1143102539469478, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 54153 + }, + { + "epoch": 0.54154, + "grad_norm": 2.0774565271510332, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 54154 + }, + { + "epoch": 0.54155, + "grad_norm": 1.6549316748276368, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 54155 + }, + { + "epoch": 0.54156, + "grad_norm": 1.542142139363442, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 54156 + }, + { + "epoch": 0.54157, + "grad_norm": 1.5074144282230948, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 54157 + }, + { + "epoch": 0.54158, + "grad_norm": 1.479441188330229, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 54158 + }, + { + "epoch": 0.54159, + "grad_norm": 1.5998613446673766, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 54159 + }, + { + "epoch": 0.5416, + "grad_norm": 1.7243077559098907, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 54160 + }, + { + "epoch": 0.54161, + "grad_norm": 1.4843049649906792, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 54161 + }, + { + "epoch": 0.54162, + "grad_norm": 1.533862702968755, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54162 + }, + { + "epoch": 0.54163, + "grad_norm": 1.4990239046843443, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 54163 + }, + { + "epoch": 0.54164, + "grad_norm": 1.4430091321687641, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 54164 + }, + { + "epoch": 0.54165, + "grad_norm": 1.5444724645533208, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 54165 + }, + { + "epoch": 0.54166, + "grad_norm": 1.3400008651644086, + "learning_rate": 0.003, + "loss": 4.038, + "step": 54166 + }, + { + "epoch": 0.54167, + "grad_norm": 1.7115119227438267, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 54167 + }, + { + "epoch": 0.54168, + "grad_norm": 1.4076518052268046, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 54168 + }, + { + "epoch": 0.54169, + "grad_norm": 1.9420622473768312, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 54169 + }, + { + "epoch": 0.5417, + "grad_norm": 1.3287828714341878, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54170 + }, + { + "epoch": 0.54171, + "grad_norm": 2.043144531593811, + "learning_rate": 0.003, + "loss": 4.042, + "step": 54171 + }, + { + "epoch": 0.54172, + "grad_norm": 1.5718486732307957, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 54172 + }, + { + "epoch": 0.54173, + "grad_norm": 1.4930865953693997, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 54173 + }, + { + "epoch": 0.54174, + "grad_norm": 1.6555897219757971, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 54174 + }, + { + "epoch": 0.54175, + "grad_norm": 1.6639319800998569, + "learning_rate": 0.003, + "loss": 4.043, + "step": 54175 + }, + { + "epoch": 0.54176, + "grad_norm": 1.2995394023979736, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 54176 + }, + { + "epoch": 0.54177, + "grad_norm": 1.7143882328806124, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 54177 + }, + { + "epoch": 0.54178, + "grad_norm": 1.3569821648040088, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 54178 + }, + { + "epoch": 0.54179, + "grad_norm": 1.4719366261037665, + "learning_rate": 0.003, + "loss": 4.062, + "step": 54179 + }, + { + "epoch": 0.5418, + "grad_norm": 1.538543052527737, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 54180 + }, + { + "epoch": 0.54181, + "grad_norm": 1.9853561045213908, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 54181 + }, + { + "epoch": 0.54182, + "grad_norm": 1.5133208153913003, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 54182 + }, + { + "epoch": 0.54183, + "grad_norm": 1.3758197446289333, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54183 + }, + { + "epoch": 0.54184, + "grad_norm": 1.910628287694907, + "learning_rate": 0.003, + "loss": 4.037, + "step": 54184 + }, + { + "epoch": 0.54185, + "grad_norm": 1.2503654431197708, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 54185 + }, + { + "epoch": 0.54186, + "grad_norm": 1.9130138104307386, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 54186 + }, + { + "epoch": 0.54187, + "grad_norm": 1.3790025051322548, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 54187 + }, + { + "epoch": 0.54188, + "grad_norm": 1.716950508282994, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 54188 + }, + { + "epoch": 0.54189, + "grad_norm": 1.3863236943867234, + "learning_rate": 0.003, + "loss": 4.045, + "step": 54189 + }, + { + "epoch": 0.5419, + "grad_norm": 1.6135624260764252, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 54190 + }, + { + "epoch": 0.54191, + "grad_norm": 1.6727460632915199, + "learning_rate": 0.003, + "loss": 4.045, + "step": 54191 + }, + { + "epoch": 0.54192, + "grad_norm": 1.559637391409788, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 54192 + }, + { + "epoch": 0.54193, + "grad_norm": 1.5045890706977547, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 54193 + }, + { + "epoch": 0.54194, + "grad_norm": 1.2871461741778905, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 54194 + }, + { + "epoch": 0.54195, + "grad_norm": 1.695558905334632, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 54195 + }, + { + "epoch": 0.54196, + "grad_norm": 1.598517624141992, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 54196 + }, + { + "epoch": 0.54197, + "grad_norm": 1.8549718201075263, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 54197 + }, + { + "epoch": 0.54198, + "grad_norm": 1.6600281814786546, + "learning_rate": 0.003, + "loss": 4.043, + "step": 54198 + }, + { + "epoch": 0.54199, + "grad_norm": 1.6600052524704179, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 54199 + }, + { + "epoch": 0.542, + "grad_norm": 1.665946175512964, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 54200 + }, + { + "epoch": 0.54201, + "grad_norm": 1.2256717968739304, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 54201 + }, + { + "epoch": 0.54202, + "grad_norm": 1.5347876160295086, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 54202 + }, + { + "epoch": 0.54203, + "grad_norm": 1.443090325913287, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 54203 + }, + { + "epoch": 0.54204, + "grad_norm": 1.571439795022196, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 54204 + }, + { + "epoch": 0.54205, + "grad_norm": 1.4891856115712794, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 54205 + }, + { + "epoch": 0.54206, + "grad_norm": 1.5880139835438005, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 54206 + }, + { + "epoch": 0.54207, + "grad_norm": 1.385011215036405, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 54207 + }, + { + "epoch": 0.54208, + "grad_norm": 1.604368214543489, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 54208 + }, + { + "epoch": 0.54209, + "grad_norm": 1.4165788718890138, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 54209 + }, + { + "epoch": 0.5421, + "grad_norm": 1.7804212745852224, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 54210 + }, + { + "epoch": 0.54211, + "grad_norm": 1.3865012405886343, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 54211 + }, + { + "epoch": 0.54212, + "grad_norm": 1.7950915730507335, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 54212 + }, + { + "epoch": 0.54213, + "grad_norm": 1.6023797210293096, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 54213 + }, + { + "epoch": 0.54214, + "grad_norm": 1.6256363297976784, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 54214 + }, + { + "epoch": 0.54215, + "grad_norm": 1.8043447578217715, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 54215 + }, + { + "epoch": 0.54216, + "grad_norm": 1.664483652166758, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 54216 + }, + { + "epoch": 0.54217, + "grad_norm": 1.695725522203021, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 54217 + }, + { + "epoch": 0.54218, + "grad_norm": 1.5383602887922436, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 54218 + }, + { + "epoch": 0.54219, + "grad_norm": 1.7900666989436114, + "learning_rate": 0.003, + "loss": 4.027, + "step": 54219 + }, + { + "epoch": 0.5422, + "grad_norm": 1.3295655826796915, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 54220 + }, + { + "epoch": 0.54221, + "grad_norm": 1.608458568187073, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 54221 + }, + { + "epoch": 0.54222, + "grad_norm": 1.8356301431166255, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 54222 + }, + { + "epoch": 0.54223, + "grad_norm": 1.2112446487999693, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 54223 + }, + { + "epoch": 0.54224, + "grad_norm": 1.7824347723041851, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 54224 + }, + { + "epoch": 0.54225, + "grad_norm": 1.3938810800820238, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 54225 + }, + { + "epoch": 0.54226, + "grad_norm": 1.6868924131202212, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 54226 + }, + { + "epoch": 0.54227, + "grad_norm": 1.2992217647834072, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 54227 + }, + { + "epoch": 0.54228, + "grad_norm": 1.5626300451918496, + "learning_rate": 0.003, + "loss": 4.027, + "step": 54228 + }, + { + "epoch": 0.54229, + "grad_norm": 1.5477273774129798, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 54229 + }, + { + "epoch": 0.5423, + "grad_norm": 1.4884876239733649, + "learning_rate": 0.003, + "loss": 3.997, + "step": 54230 + }, + { + "epoch": 0.54231, + "grad_norm": 1.74895048440425, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 54231 + }, + { + "epoch": 0.54232, + "grad_norm": 1.4677175261436255, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 54232 + }, + { + "epoch": 0.54233, + "grad_norm": 1.7407140905694722, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 54233 + }, + { + "epoch": 0.54234, + "grad_norm": 1.4603724193783683, + "learning_rate": 0.003, + "loss": 4.027, + "step": 54234 + }, + { + "epoch": 0.54235, + "grad_norm": 1.8566254494626206, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 54235 + }, + { + "epoch": 0.54236, + "grad_norm": 1.5985859296100147, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 54236 + }, + { + "epoch": 0.54237, + "grad_norm": 1.5528073407984622, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 54237 + }, + { + "epoch": 0.54238, + "grad_norm": 1.2026539556429388, + "learning_rate": 0.003, + "loss": 4.013, + "step": 54238 + }, + { + "epoch": 0.54239, + "grad_norm": 1.8616264428817195, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 54239 + }, + { + "epoch": 0.5424, + "grad_norm": 1.4062349789320023, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 54240 + }, + { + "epoch": 0.54241, + "grad_norm": 1.8189866040704679, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 54241 + }, + { + "epoch": 0.54242, + "grad_norm": 1.5211774166639271, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 54242 + }, + { + "epoch": 0.54243, + "grad_norm": 1.5422639367426993, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 54243 + }, + { + "epoch": 0.54244, + "grad_norm": 1.3916211048585485, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 54244 + }, + { + "epoch": 0.54245, + "grad_norm": 1.756700971242531, + "learning_rate": 0.003, + "loss": 4.059, + "step": 54245 + }, + { + "epoch": 0.54246, + "grad_norm": 1.4297372687226695, + "learning_rate": 0.003, + "loss": 4.019, + "step": 54246 + }, + { + "epoch": 0.54247, + "grad_norm": 1.7239902974508223, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 54247 + }, + { + "epoch": 0.54248, + "grad_norm": 1.4247552756754032, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 54248 + }, + { + "epoch": 0.54249, + "grad_norm": 1.69086329656444, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 54249 + }, + { + "epoch": 0.5425, + "grad_norm": 1.3746186610085398, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 54250 + }, + { + "epoch": 0.54251, + "grad_norm": 1.5478771624937544, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 54251 + }, + { + "epoch": 0.54252, + "grad_norm": 1.6603703762875144, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 54252 + }, + { + "epoch": 0.54253, + "grad_norm": 1.5692540095712035, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 54253 + }, + { + "epoch": 0.54254, + "grad_norm": 1.6009073453415765, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 54254 + }, + { + "epoch": 0.54255, + "grad_norm": 1.544386023467326, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 54255 + }, + { + "epoch": 0.54256, + "grad_norm": 1.872142983152279, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 54256 + }, + { + "epoch": 0.54257, + "grad_norm": 1.461388564177192, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 54257 + }, + { + "epoch": 0.54258, + "grad_norm": 1.6820009832530056, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 54258 + }, + { + "epoch": 0.54259, + "grad_norm": 1.5049739140775173, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 54259 + }, + { + "epoch": 0.5426, + "grad_norm": 1.6434642808594644, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 54260 + }, + { + "epoch": 0.54261, + "grad_norm": 1.5281022306493577, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 54261 + }, + { + "epoch": 0.54262, + "grad_norm": 1.5352463780866492, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 54262 + }, + { + "epoch": 0.54263, + "grad_norm": 1.7990239131156422, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 54263 + }, + { + "epoch": 0.54264, + "grad_norm": 1.3772173743989382, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 54264 + }, + { + "epoch": 0.54265, + "grad_norm": 1.7446368825341607, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 54265 + }, + { + "epoch": 0.54266, + "grad_norm": 1.388738894390236, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 54266 + }, + { + "epoch": 0.54267, + "grad_norm": 1.7987022005380477, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54267 + }, + { + "epoch": 0.54268, + "grad_norm": 1.4744810682148732, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 54268 + }, + { + "epoch": 0.54269, + "grad_norm": 1.7770628313474701, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54269 + }, + { + "epoch": 0.5427, + "grad_norm": 1.5428228996329914, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54270 + }, + { + "epoch": 0.54271, + "grad_norm": 1.483361112446251, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 54271 + }, + { + "epoch": 0.54272, + "grad_norm": 1.5076540922627393, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 54272 + }, + { + "epoch": 0.54273, + "grad_norm": 1.419031078936631, + "learning_rate": 0.003, + "loss": 4.001, + "step": 54273 + }, + { + "epoch": 0.54274, + "grad_norm": 1.391389827198487, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 54274 + }, + { + "epoch": 0.54275, + "grad_norm": 1.4747077190312385, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 54275 + }, + { + "epoch": 0.54276, + "grad_norm": 1.8399471882379448, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 54276 + }, + { + "epoch": 0.54277, + "grad_norm": 1.2203253349444834, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 54277 + }, + { + "epoch": 0.54278, + "grad_norm": 1.8226147148581984, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 54278 + }, + { + "epoch": 0.54279, + "grad_norm": 1.5221049516501806, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54279 + }, + { + "epoch": 0.5428, + "grad_norm": 1.7596809985303956, + "learning_rate": 0.003, + "loss": 4.02, + "step": 54280 + }, + { + "epoch": 0.54281, + "grad_norm": 1.4975873212725257, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 54281 + }, + { + "epoch": 0.54282, + "grad_norm": 1.6991686058325504, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 54282 + }, + { + "epoch": 0.54283, + "grad_norm": 1.5149507785423504, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 54283 + }, + { + "epoch": 0.54284, + "grad_norm": 1.5626974990198428, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54284 + }, + { + "epoch": 0.54285, + "grad_norm": 1.5187603887361583, + "learning_rate": 0.003, + "loss": 4.061, + "step": 54285 + }, + { + "epoch": 0.54286, + "grad_norm": 1.525624040016642, + "learning_rate": 0.003, + "loss": 4.011, + "step": 54286 + }, + { + "epoch": 0.54287, + "grad_norm": 1.833085414130524, + "learning_rate": 0.003, + "loss": 4.012, + "step": 54287 + }, + { + "epoch": 0.54288, + "grad_norm": 1.3013863763959832, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 54288 + }, + { + "epoch": 0.54289, + "grad_norm": 2.037318643204515, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 54289 + }, + { + "epoch": 0.5429, + "grad_norm": 1.2781537381665977, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 54290 + }, + { + "epoch": 0.54291, + "grad_norm": 1.8164417118753957, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 54291 + }, + { + "epoch": 0.54292, + "grad_norm": 1.578810297332806, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 54292 + }, + { + "epoch": 0.54293, + "grad_norm": 1.709174642979793, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 54293 + }, + { + "epoch": 0.54294, + "grad_norm": 1.410018575649775, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 54294 + }, + { + "epoch": 0.54295, + "grad_norm": 1.5840171566100416, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 54295 + }, + { + "epoch": 0.54296, + "grad_norm": 1.5291963138105957, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 54296 + }, + { + "epoch": 0.54297, + "grad_norm": 1.4215274014454065, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 54297 + }, + { + "epoch": 0.54298, + "grad_norm": 1.4908155922635502, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 54298 + }, + { + "epoch": 0.54299, + "grad_norm": 1.358670596790898, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 54299 + }, + { + "epoch": 0.543, + "grad_norm": 1.7994667630648835, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 54300 + }, + { + "epoch": 0.54301, + "grad_norm": 1.2367934181709228, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 54301 + }, + { + "epoch": 0.54302, + "grad_norm": 1.8930412991996122, + "learning_rate": 0.003, + "loss": 4.06, + "step": 54302 + }, + { + "epoch": 0.54303, + "grad_norm": 1.6354348953952604, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 54303 + }, + { + "epoch": 0.54304, + "grad_norm": 1.296690216137188, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 54304 + }, + { + "epoch": 0.54305, + "grad_norm": 1.9747988898553972, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 54305 + }, + { + "epoch": 0.54306, + "grad_norm": 1.680013203427813, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 54306 + }, + { + "epoch": 0.54307, + "grad_norm": 2.1998437314298545, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 54307 + }, + { + "epoch": 0.54308, + "grad_norm": 1.6927638510958831, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 54308 + }, + { + "epoch": 0.54309, + "grad_norm": 1.248677377918697, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 54309 + }, + { + "epoch": 0.5431, + "grad_norm": 1.5941699464010963, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 54310 + }, + { + "epoch": 0.54311, + "grad_norm": 1.5697085078259443, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 54311 + }, + { + "epoch": 0.54312, + "grad_norm": 1.4895242463441254, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54312 + }, + { + "epoch": 0.54313, + "grad_norm": 1.6900613969724383, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54313 + }, + { + "epoch": 0.54314, + "grad_norm": 1.608807544708035, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 54314 + }, + { + "epoch": 0.54315, + "grad_norm": 1.7762979423079799, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 54315 + }, + { + "epoch": 0.54316, + "grad_norm": 1.3748513514210152, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 54316 + }, + { + "epoch": 0.54317, + "grad_norm": 1.6076577167617505, + "learning_rate": 0.003, + "loss": 4.049, + "step": 54317 + }, + { + "epoch": 0.54318, + "grad_norm": 1.5327147249944857, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 54318 + }, + { + "epoch": 0.54319, + "grad_norm": 2.0550707690313494, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 54319 + }, + { + "epoch": 0.5432, + "grad_norm": 1.731729324787878, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 54320 + }, + { + "epoch": 0.54321, + "grad_norm": 1.3458175563465355, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 54321 + }, + { + "epoch": 0.54322, + "grad_norm": 1.2978868407334148, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 54322 + }, + { + "epoch": 0.54323, + "grad_norm": 1.950248547972045, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 54323 + }, + { + "epoch": 0.54324, + "grad_norm": 1.386908364511677, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 54324 + }, + { + "epoch": 0.54325, + "grad_norm": 1.7259140608192773, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 54325 + }, + { + "epoch": 0.54326, + "grad_norm": 1.532924965090783, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 54326 + }, + { + "epoch": 0.54327, + "grad_norm": 1.6783849655194512, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 54327 + }, + { + "epoch": 0.54328, + "grad_norm": 1.5944732769330883, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 54328 + }, + { + "epoch": 0.54329, + "grad_norm": 1.6245241968382709, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54329 + }, + { + "epoch": 0.5433, + "grad_norm": 1.709295019966548, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 54330 + }, + { + "epoch": 0.54331, + "grad_norm": 1.5310879037685015, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 54331 + }, + { + "epoch": 0.54332, + "grad_norm": 1.3702486348858551, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 54332 + }, + { + "epoch": 0.54333, + "grad_norm": 1.7996595381159683, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 54333 + }, + { + "epoch": 0.54334, + "grad_norm": 1.6812808234271348, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 54334 + }, + { + "epoch": 0.54335, + "grad_norm": 1.6644494255281717, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 54335 + }, + { + "epoch": 0.54336, + "grad_norm": 1.7354077930102156, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 54336 + }, + { + "epoch": 0.54337, + "grad_norm": 1.3868910133944083, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 54337 + }, + { + "epoch": 0.54338, + "grad_norm": 1.572654222372756, + "learning_rate": 0.003, + "loss": 4.029, + "step": 54338 + }, + { + "epoch": 0.54339, + "grad_norm": 1.7698238831226825, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54339 + }, + { + "epoch": 0.5434, + "grad_norm": 1.5924414533962126, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 54340 + }, + { + "epoch": 0.54341, + "grad_norm": 1.5192789914812026, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 54341 + }, + { + "epoch": 0.54342, + "grad_norm": 1.5462158448717735, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 54342 + }, + { + "epoch": 0.54343, + "grad_norm": 1.722605732002241, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 54343 + }, + { + "epoch": 0.54344, + "grad_norm": 1.230471181546267, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 54344 + }, + { + "epoch": 0.54345, + "grad_norm": 1.842726520067874, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 54345 + }, + { + "epoch": 0.54346, + "grad_norm": 1.2472068481967318, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 54346 + }, + { + "epoch": 0.54347, + "grad_norm": 2.1647458354020594, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 54347 + }, + { + "epoch": 0.54348, + "grad_norm": 1.3044318384345566, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 54348 + }, + { + "epoch": 0.54349, + "grad_norm": 1.7188730646763377, + "learning_rate": 0.003, + "loss": 4.035, + "step": 54349 + }, + { + "epoch": 0.5435, + "grad_norm": 1.598649642915665, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 54350 + }, + { + "epoch": 0.54351, + "grad_norm": 1.501316251303862, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54351 + }, + { + "epoch": 0.54352, + "grad_norm": 1.7254233436154134, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 54352 + }, + { + "epoch": 0.54353, + "grad_norm": 1.313869330195355, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 54353 + }, + { + "epoch": 0.54354, + "grad_norm": 1.526344305335321, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 54354 + }, + { + "epoch": 0.54355, + "grad_norm": 1.5348097365241076, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 54355 + }, + { + "epoch": 0.54356, + "grad_norm": 1.6618335773503037, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 54356 + }, + { + "epoch": 0.54357, + "grad_norm": 1.4610191848607061, + "learning_rate": 0.003, + "loss": 4.031, + "step": 54357 + }, + { + "epoch": 0.54358, + "grad_norm": 1.651762407687016, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 54358 + }, + { + "epoch": 0.54359, + "grad_norm": 1.5344903064990376, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 54359 + }, + { + "epoch": 0.5436, + "grad_norm": 1.6362065118667855, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 54360 + }, + { + "epoch": 0.54361, + "grad_norm": 1.955725078847975, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54361 + }, + { + "epoch": 0.54362, + "grad_norm": 1.451400762390375, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 54362 + }, + { + "epoch": 0.54363, + "grad_norm": 1.6916061795612043, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 54363 + }, + { + "epoch": 0.54364, + "grad_norm": 1.7142431199476245, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 54364 + }, + { + "epoch": 0.54365, + "grad_norm": 1.3991389362693964, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 54365 + }, + { + "epoch": 0.54366, + "grad_norm": 1.6969952729235833, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 54366 + }, + { + "epoch": 0.54367, + "grad_norm": 1.4128638700625744, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54367 + }, + { + "epoch": 0.54368, + "grad_norm": 1.4029972887042854, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 54368 + }, + { + "epoch": 0.54369, + "grad_norm": 1.6958575946538776, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 54369 + }, + { + "epoch": 0.5437, + "grad_norm": 1.5262801403145898, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 54370 + }, + { + "epoch": 0.54371, + "grad_norm": 1.5635836783617378, + "learning_rate": 0.003, + "loss": 4.049, + "step": 54371 + }, + { + "epoch": 0.54372, + "grad_norm": 1.5599028152293581, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 54372 + }, + { + "epoch": 0.54373, + "grad_norm": 1.3814875540478095, + "learning_rate": 0.003, + "loss": 4.048, + "step": 54373 + }, + { + "epoch": 0.54374, + "grad_norm": 1.7509637157706448, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 54374 + }, + { + "epoch": 0.54375, + "grad_norm": 1.6066795524874429, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 54375 + }, + { + "epoch": 0.54376, + "grad_norm": 1.7162103049289124, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 54376 + }, + { + "epoch": 0.54377, + "grad_norm": 2.1587663534832027, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 54377 + }, + { + "epoch": 0.54378, + "grad_norm": 1.3933151548637484, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 54378 + }, + { + "epoch": 0.54379, + "grad_norm": 1.2764862301517765, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 54379 + }, + { + "epoch": 0.5438, + "grad_norm": 1.873073589192602, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 54380 + }, + { + "epoch": 0.54381, + "grad_norm": 1.4785229436236713, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54381 + }, + { + "epoch": 0.54382, + "grad_norm": 1.9727218594762794, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 54382 + }, + { + "epoch": 0.54383, + "grad_norm": 1.3252200573299846, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 54383 + }, + { + "epoch": 0.54384, + "grad_norm": 1.6510297838766856, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 54384 + }, + { + "epoch": 0.54385, + "grad_norm": 1.8728712684646704, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54385 + }, + { + "epoch": 0.54386, + "grad_norm": 1.525045063975927, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 54386 + }, + { + "epoch": 0.54387, + "grad_norm": 1.4704582202342336, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 54387 + }, + { + "epoch": 0.54388, + "grad_norm": 1.4903500669594125, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 54388 + }, + { + "epoch": 0.54389, + "grad_norm": 1.6533703981477235, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 54389 + }, + { + "epoch": 0.5439, + "grad_norm": 1.2423717050352256, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 54390 + }, + { + "epoch": 0.54391, + "grad_norm": 1.9489644704001952, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 54391 + }, + { + "epoch": 0.54392, + "grad_norm": 1.1557102033625932, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 54392 + }, + { + "epoch": 0.54393, + "grad_norm": 1.8480454262009074, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 54393 + }, + { + "epoch": 0.54394, + "grad_norm": 1.2551359995981584, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54394 + }, + { + "epoch": 0.54395, + "grad_norm": 1.5494072057974448, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 54395 + }, + { + "epoch": 0.54396, + "grad_norm": 1.2999960726478108, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 54396 + }, + { + "epoch": 0.54397, + "grad_norm": 1.5579300945613537, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 54397 + }, + { + "epoch": 0.54398, + "grad_norm": 1.5909705043965832, + "learning_rate": 0.003, + "loss": 4.027, + "step": 54398 + }, + { + "epoch": 0.54399, + "grad_norm": 1.323636119400058, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 54399 + }, + { + "epoch": 0.544, + "grad_norm": 1.9777861138412172, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 54400 + }, + { + "epoch": 0.54401, + "grad_norm": 2.109433267149971, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 54401 + }, + { + "epoch": 0.54402, + "grad_norm": 1.5510367084457957, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 54402 + }, + { + "epoch": 0.54403, + "grad_norm": 1.5707610578508486, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 54403 + }, + { + "epoch": 0.54404, + "grad_norm": 1.6011887438942851, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 54404 + }, + { + "epoch": 0.54405, + "grad_norm": 1.5441309257327829, + "learning_rate": 0.003, + "loss": 4.035, + "step": 54405 + }, + { + "epoch": 0.54406, + "grad_norm": 1.7306662192956301, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 54406 + }, + { + "epoch": 0.54407, + "grad_norm": 1.4811976197143881, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 54407 + }, + { + "epoch": 0.54408, + "grad_norm": 1.4405182500958418, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 54408 + }, + { + "epoch": 0.54409, + "grad_norm": 1.7667309243007592, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 54409 + }, + { + "epoch": 0.5441, + "grad_norm": 1.3825921993395798, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54410 + }, + { + "epoch": 0.54411, + "grad_norm": 1.610443290770545, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54411 + }, + { + "epoch": 0.54412, + "grad_norm": 1.430688172461036, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 54412 + }, + { + "epoch": 0.54413, + "grad_norm": 1.7173633604246843, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 54413 + }, + { + "epoch": 0.54414, + "grad_norm": 1.4499207543666812, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 54414 + }, + { + "epoch": 0.54415, + "grad_norm": 1.3335627703840411, + "learning_rate": 0.003, + "loss": 4.079, + "step": 54415 + }, + { + "epoch": 0.54416, + "grad_norm": 1.5569713761743251, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 54416 + }, + { + "epoch": 0.54417, + "grad_norm": 1.4632798211323954, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 54417 + }, + { + "epoch": 0.54418, + "grad_norm": 1.6335219398884426, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 54418 + }, + { + "epoch": 0.54419, + "grad_norm": 1.8122036539333055, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 54419 + }, + { + "epoch": 0.5442, + "grad_norm": 1.6582762157916915, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 54420 + }, + { + "epoch": 0.54421, + "grad_norm": 1.7596057040405844, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 54421 + }, + { + "epoch": 0.54422, + "grad_norm": 1.5551294800609297, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 54422 + }, + { + "epoch": 0.54423, + "grad_norm": 1.3707765020693508, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 54423 + }, + { + "epoch": 0.54424, + "grad_norm": 1.7703316588227456, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 54424 + }, + { + "epoch": 0.54425, + "grad_norm": 1.4334131775377574, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 54425 + }, + { + "epoch": 0.54426, + "grad_norm": 1.4616987782933428, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 54426 + }, + { + "epoch": 0.54427, + "grad_norm": 1.572172076306183, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 54427 + }, + { + "epoch": 0.54428, + "grad_norm": 1.544044894558827, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 54428 + }, + { + "epoch": 0.54429, + "grad_norm": 1.4110735697653536, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 54429 + }, + { + "epoch": 0.5443, + "grad_norm": 1.5923471528392992, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 54430 + }, + { + "epoch": 0.54431, + "grad_norm": 1.581057830017775, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 54431 + }, + { + "epoch": 0.54432, + "grad_norm": 1.4000558913515435, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 54432 + }, + { + "epoch": 0.54433, + "grad_norm": 1.814592357346721, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 54433 + }, + { + "epoch": 0.54434, + "grad_norm": 1.3719646582435683, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 54434 + }, + { + "epoch": 0.54435, + "grad_norm": 1.5339512181559136, + "learning_rate": 0.003, + "loss": 4.035, + "step": 54435 + }, + { + "epoch": 0.54436, + "grad_norm": 1.4919614055039934, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 54436 + }, + { + "epoch": 0.54437, + "grad_norm": 2.3692804888202526, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 54437 + }, + { + "epoch": 0.54438, + "grad_norm": 1.1213048018543668, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 54438 + }, + { + "epoch": 0.54439, + "grad_norm": 1.5944933563994668, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 54439 + }, + { + "epoch": 0.5444, + "grad_norm": 1.5956826487782967, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 54440 + }, + { + "epoch": 0.54441, + "grad_norm": 1.438734848185007, + "learning_rate": 0.003, + "loss": 4.012, + "step": 54441 + }, + { + "epoch": 0.54442, + "grad_norm": 1.631950760855928, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 54442 + }, + { + "epoch": 0.54443, + "grad_norm": 1.6831547296057714, + "learning_rate": 0.003, + "loss": 4.055, + "step": 54443 + }, + { + "epoch": 0.54444, + "grad_norm": 1.4674787289477877, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 54444 + }, + { + "epoch": 0.54445, + "grad_norm": 1.4534802873311654, + "learning_rate": 0.003, + "loss": 4.039, + "step": 54445 + }, + { + "epoch": 0.54446, + "grad_norm": 1.727233667684943, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 54446 + }, + { + "epoch": 0.54447, + "grad_norm": 1.5942292452965696, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 54447 + }, + { + "epoch": 0.54448, + "grad_norm": 1.6975027173775976, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 54448 + }, + { + "epoch": 0.54449, + "grad_norm": 1.3946194672985932, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 54449 + }, + { + "epoch": 0.5445, + "grad_norm": 1.9364969649409818, + "learning_rate": 0.003, + "loss": 4.034, + "step": 54450 + }, + { + "epoch": 0.54451, + "grad_norm": 1.3051623264088117, + "learning_rate": 0.003, + "loss": 4.053, + "step": 54451 + }, + { + "epoch": 0.54452, + "grad_norm": 1.560731756314872, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 54452 + }, + { + "epoch": 0.54453, + "grad_norm": 1.5172930464814105, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 54453 + }, + { + "epoch": 0.54454, + "grad_norm": 1.5604950838323695, + "learning_rate": 0.003, + "loss": 4.022, + "step": 54454 + }, + { + "epoch": 0.54455, + "grad_norm": 1.474316905795359, + "learning_rate": 0.003, + "loss": 4.037, + "step": 54455 + }, + { + "epoch": 0.54456, + "grad_norm": 1.484217990747013, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54456 + }, + { + "epoch": 0.54457, + "grad_norm": 1.781445892449564, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 54457 + }, + { + "epoch": 0.54458, + "grad_norm": 1.7558424585910495, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 54458 + }, + { + "epoch": 0.54459, + "grad_norm": 1.3792956618948262, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 54459 + }, + { + "epoch": 0.5446, + "grad_norm": 1.7922888411841083, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 54460 + }, + { + "epoch": 0.54461, + "grad_norm": 1.7254623171145402, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 54461 + }, + { + "epoch": 0.54462, + "grad_norm": 1.567652323826191, + "learning_rate": 0.003, + "loss": 4.0, + "step": 54462 + }, + { + "epoch": 0.54463, + "grad_norm": 1.802273677774292, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 54463 + }, + { + "epoch": 0.54464, + "grad_norm": 1.3373178361029192, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 54464 + }, + { + "epoch": 0.54465, + "grad_norm": 1.5417690908090116, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 54465 + }, + { + "epoch": 0.54466, + "grad_norm": 1.5552448772719123, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 54466 + }, + { + "epoch": 0.54467, + "grad_norm": 1.442276896674439, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 54467 + }, + { + "epoch": 0.54468, + "grad_norm": 1.3547018716305637, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 54468 + }, + { + "epoch": 0.54469, + "grad_norm": 1.7302165661729136, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 54469 + }, + { + "epoch": 0.5447, + "grad_norm": 1.4033459958810044, + "learning_rate": 0.003, + "loss": 4.005, + "step": 54470 + }, + { + "epoch": 0.54471, + "grad_norm": 1.9253300997321225, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 54471 + }, + { + "epoch": 0.54472, + "grad_norm": 1.5068459665674694, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 54472 + }, + { + "epoch": 0.54473, + "grad_norm": 1.7777030935985398, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 54473 + }, + { + "epoch": 0.54474, + "grad_norm": 1.667728440848711, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 54474 + }, + { + "epoch": 0.54475, + "grad_norm": 1.4841075882777193, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 54475 + }, + { + "epoch": 0.54476, + "grad_norm": 1.7518118045114974, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 54476 + }, + { + "epoch": 0.54477, + "grad_norm": 1.2556385552946623, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54477 + }, + { + "epoch": 0.54478, + "grad_norm": 1.8002058594176302, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 54478 + }, + { + "epoch": 0.54479, + "grad_norm": 1.5385387832232806, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 54479 + }, + { + "epoch": 0.5448, + "grad_norm": 1.7607109303852315, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 54480 + }, + { + "epoch": 0.54481, + "grad_norm": 1.3372181736829833, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 54481 + }, + { + "epoch": 0.54482, + "grad_norm": 1.4534530880191228, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 54482 + }, + { + "epoch": 0.54483, + "grad_norm": 1.4852969896587747, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 54483 + }, + { + "epoch": 0.54484, + "grad_norm": 2.03438146842482, + "learning_rate": 0.003, + "loss": 4.035, + "step": 54484 + }, + { + "epoch": 0.54485, + "grad_norm": 1.5753908820564257, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 54485 + }, + { + "epoch": 0.54486, + "grad_norm": 1.983041556357043, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 54486 + }, + { + "epoch": 0.54487, + "grad_norm": 1.5242581877164743, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 54487 + }, + { + "epoch": 0.54488, + "grad_norm": 1.6484113467433863, + "learning_rate": 0.003, + "loss": 4.024, + "step": 54488 + }, + { + "epoch": 0.54489, + "grad_norm": 1.4081525903369176, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 54489 + }, + { + "epoch": 0.5449, + "grad_norm": 1.8897943944094275, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 54490 + }, + { + "epoch": 0.54491, + "grad_norm": 1.400050554461867, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 54491 + }, + { + "epoch": 0.54492, + "grad_norm": 1.7156693980742026, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 54492 + }, + { + "epoch": 0.54493, + "grad_norm": 1.356964370973021, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 54493 + }, + { + "epoch": 0.54494, + "grad_norm": 1.6123394718710755, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 54494 + }, + { + "epoch": 0.54495, + "grad_norm": 1.4606811795404164, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54495 + }, + { + "epoch": 0.54496, + "grad_norm": 1.5840524576599089, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 54496 + }, + { + "epoch": 0.54497, + "grad_norm": 1.3110802714027416, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 54497 + }, + { + "epoch": 0.54498, + "grad_norm": 1.5131105092599073, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 54498 + }, + { + "epoch": 0.54499, + "grad_norm": 1.5559959756648771, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54499 + }, + { + "epoch": 0.545, + "grad_norm": 1.8574133783774287, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 54500 + }, + { + "epoch": 0.54501, + "grad_norm": 1.6018234827519011, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 54501 + }, + { + "epoch": 0.54502, + "grad_norm": 1.48521926223333, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 54502 + }, + { + "epoch": 0.54503, + "grad_norm": 1.524089157201405, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54503 + }, + { + "epoch": 0.54504, + "grad_norm": 1.437498080240485, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 54504 + }, + { + "epoch": 0.54505, + "grad_norm": 1.766531305919607, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 54505 + }, + { + "epoch": 0.54506, + "grad_norm": 1.484411922933223, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 54506 + }, + { + "epoch": 0.54507, + "grad_norm": 1.483622036160443, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 54507 + }, + { + "epoch": 0.54508, + "grad_norm": 1.2837265919849679, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 54508 + }, + { + "epoch": 0.54509, + "grad_norm": 1.7541938888624633, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 54509 + }, + { + "epoch": 0.5451, + "grad_norm": 1.2239352817352582, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 54510 + }, + { + "epoch": 0.54511, + "grad_norm": 2.1032291213033516, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 54511 + }, + { + "epoch": 0.54512, + "grad_norm": 1.4377323639506885, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 54512 + }, + { + "epoch": 0.54513, + "grad_norm": 1.66928290994253, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 54513 + }, + { + "epoch": 0.54514, + "grad_norm": 1.7719782704962266, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 54514 + }, + { + "epoch": 0.54515, + "grad_norm": 1.386407341936505, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 54515 + }, + { + "epoch": 0.54516, + "grad_norm": 1.5266220429445039, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 54516 + }, + { + "epoch": 0.54517, + "grad_norm": 1.6990562236313207, + "learning_rate": 0.003, + "loss": 4.05, + "step": 54517 + }, + { + "epoch": 0.54518, + "grad_norm": 1.6038075090234742, + "learning_rate": 0.003, + "loss": 4.031, + "step": 54518 + }, + { + "epoch": 0.54519, + "grad_norm": 1.620506583554574, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 54519 + }, + { + "epoch": 0.5452, + "grad_norm": 1.4904039506264974, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54520 + }, + { + "epoch": 0.54521, + "grad_norm": 1.4963657520679, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 54521 + }, + { + "epoch": 0.54522, + "grad_norm": 1.8212362279631225, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 54522 + }, + { + "epoch": 0.54523, + "grad_norm": 1.6276570799378647, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 54523 + }, + { + "epoch": 0.54524, + "grad_norm": 1.5465848391574766, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 54524 + }, + { + "epoch": 0.54525, + "grad_norm": 1.517759185953855, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 54525 + }, + { + "epoch": 0.54526, + "grad_norm": 1.517423586643446, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 54526 + }, + { + "epoch": 0.54527, + "grad_norm": 1.4391356638464576, + "learning_rate": 0.003, + "loss": 4.009, + "step": 54527 + }, + { + "epoch": 0.54528, + "grad_norm": 1.410796644956736, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 54528 + }, + { + "epoch": 0.54529, + "grad_norm": 1.4994801868106147, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 54529 + }, + { + "epoch": 0.5453, + "grad_norm": 1.3907925044186404, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 54530 + }, + { + "epoch": 0.54531, + "grad_norm": 1.632745940397701, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 54531 + }, + { + "epoch": 0.54532, + "grad_norm": 1.7526573800443344, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 54532 + }, + { + "epoch": 0.54533, + "grad_norm": 1.787378792280595, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 54533 + }, + { + "epoch": 0.54534, + "grad_norm": 1.6099507016915782, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 54534 + }, + { + "epoch": 0.54535, + "grad_norm": 1.550088478453092, + "learning_rate": 0.003, + "loss": 4.004, + "step": 54535 + }, + { + "epoch": 0.54536, + "grad_norm": 1.9450782608817803, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 54536 + }, + { + "epoch": 0.54537, + "grad_norm": 1.2663292850041628, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 54537 + }, + { + "epoch": 0.54538, + "grad_norm": 2.0621293764765785, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 54538 + }, + { + "epoch": 0.54539, + "grad_norm": 1.5327038908413253, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 54539 + }, + { + "epoch": 0.5454, + "grad_norm": 1.3266155301927116, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 54540 + }, + { + "epoch": 0.54541, + "grad_norm": 1.8527054468722588, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 54541 + }, + { + "epoch": 0.54542, + "grad_norm": 1.4614743368950691, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 54542 + }, + { + "epoch": 0.54543, + "grad_norm": 1.6041209579888733, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 54543 + }, + { + "epoch": 0.54544, + "grad_norm": 1.4091429133219975, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 54544 + }, + { + "epoch": 0.54545, + "grad_norm": 1.699107128230605, + "learning_rate": 0.003, + "loss": 4.039, + "step": 54545 + }, + { + "epoch": 0.54546, + "grad_norm": 1.5564019859540639, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 54546 + }, + { + "epoch": 0.54547, + "grad_norm": 1.410032728288163, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 54547 + }, + { + "epoch": 0.54548, + "grad_norm": 1.747620535169701, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 54548 + }, + { + "epoch": 0.54549, + "grad_norm": 1.3291968258462459, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 54549 + }, + { + "epoch": 0.5455, + "grad_norm": 2.191663322627776, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 54550 + }, + { + "epoch": 0.54551, + "grad_norm": 1.5674826403522517, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 54551 + }, + { + "epoch": 0.54552, + "grad_norm": 1.6773930399920876, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 54552 + }, + { + "epoch": 0.54553, + "grad_norm": 2.0251572274709253, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 54553 + }, + { + "epoch": 0.54554, + "grad_norm": 1.1488636113872228, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 54554 + }, + { + "epoch": 0.54555, + "grad_norm": 1.8447802264869846, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 54555 + }, + { + "epoch": 0.54556, + "grad_norm": 1.5719478518496228, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 54556 + }, + { + "epoch": 0.54557, + "grad_norm": 1.6362926248757528, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 54557 + }, + { + "epoch": 0.54558, + "grad_norm": 1.8156511737623504, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 54558 + }, + { + "epoch": 0.54559, + "grad_norm": 1.4446792628459408, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 54559 + }, + { + "epoch": 0.5456, + "grad_norm": 1.2257446181348217, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 54560 + }, + { + "epoch": 0.54561, + "grad_norm": 1.6605859281329687, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 54561 + }, + { + "epoch": 0.54562, + "grad_norm": 1.7306183296306576, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 54562 + }, + { + "epoch": 0.54563, + "grad_norm": 1.4729930597006171, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 54563 + }, + { + "epoch": 0.54564, + "grad_norm": 1.6096083497504108, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54564 + }, + { + "epoch": 0.54565, + "grad_norm": 1.5484763581457415, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 54565 + }, + { + "epoch": 0.54566, + "grad_norm": 1.3491001264536568, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 54566 + }, + { + "epoch": 0.54567, + "grad_norm": 1.7171545628322271, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 54567 + }, + { + "epoch": 0.54568, + "grad_norm": 1.4175308835423086, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 54568 + }, + { + "epoch": 0.54569, + "grad_norm": 1.5023533475331172, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 54569 + }, + { + "epoch": 0.5457, + "grad_norm": 1.6336505208969645, + "learning_rate": 0.003, + "loss": 4.025, + "step": 54570 + }, + { + "epoch": 0.54571, + "grad_norm": 1.652861078494024, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 54571 + }, + { + "epoch": 0.54572, + "grad_norm": 1.7100771333370237, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 54572 + }, + { + "epoch": 0.54573, + "grad_norm": 1.5124021881266596, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 54573 + }, + { + "epoch": 0.54574, + "grad_norm": 1.537675217944064, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 54574 + }, + { + "epoch": 0.54575, + "grad_norm": 1.4054284050549892, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 54575 + }, + { + "epoch": 0.54576, + "grad_norm": 1.622810392514186, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 54576 + }, + { + "epoch": 0.54577, + "grad_norm": 1.4643879123299255, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 54577 + }, + { + "epoch": 0.54578, + "grad_norm": 1.8546522246823536, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 54578 + }, + { + "epoch": 0.54579, + "grad_norm": 1.323136925846009, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 54579 + }, + { + "epoch": 0.5458, + "grad_norm": 1.7920058402743622, + "learning_rate": 0.003, + "loss": 4.01, + "step": 54580 + }, + { + "epoch": 0.54581, + "grad_norm": 1.6184556369570686, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 54581 + }, + { + "epoch": 0.54582, + "grad_norm": 1.3896580039363808, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 54582 + }, + { + "epoch": 0.54583, + "grad_norm": 1.6356398376423826, + "learning_rate": 0.003, + "loss": 4.001, + "step": 54583 + }, + { + "epoch": 0.54584, + "grad_norm": 1.3361622997967157, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 54584 + }, + { + "epoch": 0.54585, + "grad_norm": 1.7219610588047995, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 54585 + }, + { + "epoch": 0.54586, + "grad_norm": 1.4546195179009636, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 54586 + }, + { + "epoch": 0.54587, + "grad_norm": 1.653118404106418, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 54587 + }, + { + "epoch": 0.54588, + "grad_norm": 1.6277238832543297, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 54588 + }, + { + "epoch": 0.54589, + "grad_norm": 1.725833171135909, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 54589 + }, + { + "epoch": 0.5459, + "grad_norm": 1.5679077031535618, + "learning_rate": 0.003, + "loss": 4.024, + "step": 54590 + }, + { + "epoch": 0.54591, + "grad_norm": 1.3181457157029361, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 54591 + }, + { + "epoch": 0.54592, + "grad_norm": 1.5398078302553566, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54592 + }, + { + "epoch": 0.54593, + "grad_norm": 1.621098652394937, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 54593 + }, + { + "epoch": 0.54594, + "grad_norm": 1.6310969323510907, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 54594 + }, + { + "epoch": 0.54595, + "grad_norm": 1.7861873223355516, + "learning_rate": 0.003, + "loss": 4.032, + "step": 54595 + }, + { + "epoch": 0.54596, + "grad_norm": 1.687207513374486, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 54596 + }, + { + "epoch": 0.54597, + "grad_norm": 1.52879361260968, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 54597 + }, + { + "epoch": 0.54598, + "grad_norm": 1.8179207331712475, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 54598 + }, + { + "epoch": 0.54599, + "grad_norm": 1.2401983664449134, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 54599 + }, + { + "epoch": 0.546, + "grad_norm": 1.687269077095677, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 54600 + }, + { + "epoch": 0.54601, + "grad_norm": 1.490824095379841, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54601 + }, + { + "epoch": 0.54602, + "grad_norm": 1.5607143473048801, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 54602 + }, + { + "epoch": 0.54603, + "grad_norm": 1.5391536429259274, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 54603 + }, + { + "epoch": 0.54604, + "grad_norm": 1.4676421024654875, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 54604 + }, + { + "epoch": 0.54605, + "grad_norm": 1.8812140425383999, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 54605 + }, + { + "epoch": 0.54606, + "grad_norm": 1.487984782107856, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54606 + }, + { + "epoch": 0.54607, + "grad_norm": 1.745833617955257, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 54607 + }, + { + "epoch": 0.54608, + "grad_norm": 1.381762436589822, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 54608 + }, + { + "epoch": 0.54609, + "grad_norm": 1.8562257594137148, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 54609 + }, + { + "epoch": 0.5461, + "grad_norm": 1.4362678468556969, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 54610 + }, + { + "epoch": 0.54611, + "grad_norm": 1.7373195754256163, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 54611 + }, + { + "epoch": 0.54612, + "grad_norm": 1.3911899873870206, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54612 + }, + { + "epoch": 0.54613, + "grad_norm": 1.529546442516383, + "learning_rate": 0.003, + "loss": 4.007, + "step": 54613 + }, + { + "epoch": 0.54614, + "grad_norm": 1.6554254973860054, + "learning_rate": 0.003, + "loss": 4.063, + "step": 54614 + }, + { + "epoch": 0.54615, + "grad_norm": 1.7054782199653662, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 54615 + }, + { + "epoch": 0.54616, + "grad_norm": 1.465308928664045, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 54616 + }, + { + "epoch": 0.54617, + "grad_norm": 1.5795281465378104, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 54617 + }, + { + "epoch": 0.54618, + "grad_norm": 1.5402332673083978, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 54618 + }, + { + "epoch": 0.54619, + "grad_norm": 1.433945104033263, + "learning_rate": 0.003, + "loss": 3.997, + "step": 54619 + }, + { + "epoch": 0.5462, + "grad_norm": 1.74617811880436, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 54620 + }, + { + "epoch": 0.54621, + "grad_norm": 1.5071107165358137, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 54621 + }, + { + "epoch": 0.54622, + "grad_norm": 1.4379193614980903, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 54622 + }, + { + "epoch": 0.54623, + "grad_norm": 1.6991510781469736, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 54623 + }, + { + "epoch": 0.54624, + "grad_norm": 1.4608055428453586, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 54624 + }, + { + "epoch": 0.54625, + "grad_norm": 1.9046687105813969, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 54625 + }, + { + "epoch": 0.54626, + "grad_norm": 1.530730665564697, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 54626 + }, + { + "epoch": 0.54627, + "grad_norm": 1.8485421405198141, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 54627 + }, + { + "epoch": 0.54628, + "grad_norm": 1.3851828392620957, + "learning_rate": 0.003, + "loss": 4.034, + "step": 54628 + }, + { + "epoch": 0.54629, + "grad_norm": 1.553116854094002, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 54629 + }, + { + "epoch": 0.5463, + "grad_norm": 1.4396921099038886, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 54630 + }, + { + "epoch": 0.54631, + "grad_norm": 1.5715886125268899, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 54631 + }, + { + "epoch": 0.54632, + "grad_norm": 1.4910366529016115, + "learning_rate": 0.003, + "loss": 4.018, + "step": 54632 + }, + { + "epoch": 0.54633, + "grad_norm": 1.6056211011542016, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 54633 + }, + { + "epoch": 0.54634, + "grad_norm": 1.4496625535740095, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 54634 + }, + { + "epoch": 0.54635, + "grad_norm": 1.888849810944094, + "learning_rate": 0.003, + "loss": 4.024, + "step": 54635 + }, + { + "epoch": 0.54636, + "grad_norm": 1.17146632716028, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 54636 + }, + { + "epoch": 0.54637, + "grad_norm": 1.8187807457925993, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 54637 + }, + { + "epoch": 0.54638, + "grad_norm": 1.5643678241614816, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 54638 + }, + { + "epoch": 0.54639, + "grad_norm": 1.4036126637749393, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 54639 + }, + { + "epoch": 0.5464, + "grad_norm": 1.503184169335649, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 54640 + }, + { + "epoch": 0.54641, + "grad_norm": 1.5749075252049352, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54641 + }, + { + "epoch": 0.54642, + "grad_norm": 1.3965720217827051, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 54642 + }, + { + "epoch": 0.54643, + "grad_norm": 1.4899450746620613, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 54643 + }, + { + "epoch": 0.54644, + "grad_norm": 1.4860301712021178, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 54644 + }, + { + "epoch": 0.54645, + "grad_norm": 1.4332324965306533, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54645 + }, + { + "epoch": 0.54646, + "grad_norm": 1.721837916684611, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 54646 + }, + { + "epoch": 0.54647, + "grad_norm": 1.4154877311573446, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 54647 + }, + { + "epoch": 0.54648, + "grad_norm": 1.9176867257285524, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 54648 + }, + { + "epoch": 0.54649, + "grad_norm": 1.3365432832869217, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 54649 + }, + { + "epoch": 0.5465, + "grad_norm": 1.8777365643458033, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 54650 + }, + { + "epoch": 0.54651, + "grad_norm": 1.6130571861292906, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 54651 + }, + { + "epoch": 0.54652, + "grad_norm": 1.1847476914750963, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 54652 + }, + { + "epoch": 0.54653, + "grad_norm": 1.8442901045558566, + "learning_rate": 0.003, + "loss": 4.053, + "step": 54653 + }, + { + "epoch": 0.54654, + "grad_norm": 1.7870614128831526, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 54654 + }, + { + "epoch": 0.54655, + "grad_norm": 1.7125704954613412, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 54655 + }, + { + "epoch": 0.54656, + "grad_norm": 1.3741475923468325, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 54656 + }, + { + "epoch": 0.54657, + "grad_norm": 1.5746228999775957, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 54657 + }, + { + "epoch": 0.54658, + "grad_norm": 2.00673345640671, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 54658 + }, + { + "epoch": 0.54659, + "grad_norm": 1.3753602913611647, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 54659 + }, + { + "epoch": 0.5466, + "grad_norm": 1.831111224856383, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 54660 + }, + { + "epoch": 0.54661, + "grad_norm": 1.4227280799834539, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54661 + }, + { + "epoch": 0.54662, + "grad_norm": 1.4490677640900704, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 54662 + }, + { + "epoch": 0.54663, + "grad_norm": 1.433283799653362, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 54663 + }, + { + "epoch": 0.54664, + "grad_norm": 1.6477415962835207, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 54664 + }, + { + "epoch": 0.54665, + "grad_norm": 1.389668373851355, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 54665 + }, + { + "epoch": 0.54666, + "grad_norm": 1.419324895591199, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 54666 + }, + { + "epoch": 0.54667, + "grad_norm": 1.622316225875362, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 54667 + }, + { + "epoch": 0.54668, + "grad_norm": 1.5490566681908566, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 54668 + }, + { + "epoch": 0.54669, + "grad_norm": 1.3979189008886452, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 54669 + }, + { + "epoch": 0.5467, + "grad_norm": 1.7545826565457778, + "learning_rate": 0.003, + "loss": 4.044, + "step": 54670 + }, + { + "epoch": 0.54671, + "grad_norm": 1.8067772288593504, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 54671 + }, + { + "epoch": 0.54672, + "grad_norm": 1.3940512828741831, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 54672 + }, + { + "epoch": 0.54673, + "grad_norm": 1.8057444738499344, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 54673 + }, + { + "epoch": 0.54674, + "grad_norm": 1.3696332598257854, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 54674 + }, + { + "epoch": 0.54675, + "grad_norm": 2.2253755500226706, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 54675 + }, + { + "epoch": 0.54676, + "grad_norm": 1.8595280032045443, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 54676 + }, + { + "epoch": 0.54677, + "grad_norm": 1.3737555990965309, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 54677 + }, + { + "epoch": 0.54678, + "grad_norm": 1.8509156942904654, + "learning_rate": 0.003, + "loss": 4.07, + "step": 54678 + }, + { + "epoch": 0.54679, + "grad_norm": 1.439577045371358, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 54679 + }, + { + "epoch": 0.5468, + "grad_norm": 1.5869668199963791, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 54680 + }, + { + "epoch": 0.54681, + "grad_norm": 1.7236006569278526, + "learning_rate": 0.003, + "loss": 4.042, + "step": 54681 + }, + { + "epoch": 0.54682, + "grad_norm": 1.6788390901621395, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 54682 + }, + { + "epoch": 0.54683, + "grad_norm": 1.438541908318377, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 54683 + }, + { + "epoch": 0.54684, + "grad_norm": 1.4667476930970653, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 54684 + }, + { + "epoch": 0.54685, + "grad_norm": 1.6837520625024478, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 54685 + }, + { + "epoch": 0.54686, + "grad_norm": 1.3289326190759654, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 54686 + }, + { + "epoch": 0.54687, + "grad_norm": 1.753664217390644, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 54687 + }, + { + "epoch": 0.54688, + "grad_norm": 1.6102974268522348, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 54688 + }, + { + "epoch": 0.54689, + "grad_norm": 1.2718202377738015, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 54689 + }, + { + "epoch": 0.5469, + "grad_norm": 1.7011414196451744, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 54690 + }, + { + "epoch": 0.54691, + "grad_norm": 1.462057394524623, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 54691 + }, + { + "epoch": 0.54692, + "grad_norm": 1.488239873484059, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 54692 + }, + { + "epoch": 0.54693, + "grad_norm": 1.495502959340986, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 54693 + }, + { + "epoch": 0.54694, + "grad_norm": 1.4815386250414935, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 54694 + }, + { + "epoch": 0.54695, + "grad_norm": 1.7650707678611983, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 54695 + }, + { + "epoch": 0.54696, + "grad_norm": 1.6450011236651771, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 54696 + }, + { + "epoch": 0.54697, + "grad_norm": 1.72254625659057, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 54697 + }, + { + "epoch": 0.54698, + "grad_norm": 1.5650832717862846, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 54698 + }, + { + "epoch": 0.54699, + "grad_norm": 1.7203080836301503, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 54699 + }, + { + "epoch": 0.547, + "grad_norm": 1.1222375251483547, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54700 + }, + { + "epoch": 0.54701, + "grad_norm": 1.7541943187489222, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 54701 + }, + { + "epoch": 0.54702, + "grad_norm": 1.7331882461491155, + "learning_rate": 0.003, + "loss": 4.053, + "step": 54702 + }, + { + "epoch": 0.54703, + "grad_norm": 1.6839321951342243, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 54703 + }, + { + "epoch": 0.54704, + "grad_norm": 1.79742902361756, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 54704 + }, + { + "epoch": 0.54705, + "grad_norm": 1.3004893720400927, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 54705 + }, + { + "epoch": 0.54706, + "grad_norm": 1.531478667313301, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 54706 + }, + { + "epoch": 0.54707, + "grad_norm": 1.6664578258699103, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 54707 + }, + { + "epoch": 0.54708, + "grad_norm": 1.4700613779475358, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 54708 + }, + { + "epoch": 0.54709, + "grad_norm": 1.7247491792861225, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 54709 + }, + { + "epoch": 0.5471, + "grad_norm": 1.4787621618990299, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 54710 + }, + { + "epoch": 0.54711, + "grad_norm": 1.7106648240180535, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 54711 + }, + { + "epoch": 0.54712, + "grad_norm": 1.3888433443247337, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 54712 + }, + { + "epoch": 0.54713, + "grad_norm": 1.7921458611261234, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 54713 + }, + { + "epoch": 0.54714, + "grad_norm": 1.7513714149894053, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 54714 + }, + { + "epoch": 0.54715, + "grad_norm": 1.4796998327295838, + "learning_rate": 0.003, + "loss": 4.022, + "step": 54715 + }, + { + "epoch": 0.54716, + "grad_norm": 1.5482134762139241, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 54716 + }, + { + "epoch": 0.54717, + "grad_norm": 1.7846669198468892, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 54717 + }, + { + "epoch": 0.54718, + "grad_norm": 1.2856704969789008, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 54718 + }, + { + "epoch": 0.54719, + "grad_norm": 1.7336076332336348, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 54719 + }, + { + "epoch": 0.5472, + "grad_norm": 1.3956304618126505, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 54720 + }, + { + "epoch": 0.54721, + "grad_norm": 1.784007885238396, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 54721 + }, + { + "epoch": 0.54722, + "grad_norm": 1.1829237646447197, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 54722 + }, + { + "epoch": 0.54723, + "grad_norm": 1.837462894590451, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 54723 + }, + { + "epoch": 0.54724, + "grad_norm": 1.6476891478775146, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 54724 + }, + { + "epoch": 0.54725, + "grad_norm": 1.605326158947282, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 54725 + }, + { + "epoch": 0.54726, + "grad_norm": 1.5115446903517096, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 54726 + }, + { + "epoch": 0.54727, + "grad_norm": 1.7341295821405707, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 54727 + }, + { + "epoch": 0.54728, + "grad_norm": 1.3179882425655374, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 54728 + }, + { + "epoch": 0.54729, + "grad_norm": 1.805707793719584, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 54729 + }, + { + "epoch": 0.5473, + "grad_norm": 1.4141107203907777, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 54730 + }, + { + "epoch": 0.54731, + "grad_norm": 1.9861398647264517, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 54731 + }, + { + "epoch": 0.54732, + "grad_norm": 1.4327248367315835, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 54732 + }, + { + "epoch": 0.54733, + "grad_norm": 1.8813595938513366, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 54733 + }, + { + "epoch": 0.54734, + "grad_norm": 1.8136488882649788, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 54734 + }, + { + "epoch": 0.54735, + "grad_norm": 1.528688633337184, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 54735 + }, + { + "epoch": 0.54736, + "grad_norm": 1.2997322545035452, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 54736 + }, + { + "epoch": 0.54737, + "grad_norm": 1.7229972279719856, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 54737 + }, + { + "epoch": 0.54738, + "grad_norm": 1.5648556769535342, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 54738 + }, + { + "epoch": 0.54739, + "grad_norm": 1.4858117154036248, + "learning_rate": 0.003, + "loss": 4.048, + "step": 54739 + }, + { + "epoch": 0.5474, + "grad_norm": 1.7982316628483395, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54740 + }, + { + "epoch": 0.54741, + "grad_norm": 1.335124111945693, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54741 + }, + { + "epoch": 0.54742, + "grad_norm": 1.4839950730951534, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 54742 + }, + { + "epoch": 0.54743, + "grad_norm": 1.4811675171472318, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 54743 + }, + { + "epoch": 0.54744, + "grad_norm": 1.62077924863982, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 54744 + }, + { + "epoch": 0.54745, + "grad_norm": 1.5773000675699822, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 54745 + }, + { + "epoch": 0.54746, + "grad_norm": 1.476675670287325, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 54746 + }, + { + "epoch": 0.54747, + "grad_norm": 1.582577120952324, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 54747 + }, + { + "epoch": 0.54748, + "grad_norm": 1.41594765430279, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 54748 + }, + { + "epoch": 0.54749, + "grad_norm": 1.6381453441274294, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 54749 + }, + { + "epoch": 0.5475, + "grad_norm": 1.424560499814308, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 54750 + }, + { + "epoch": 0.54751, + "grad_norm": 1.91157649348387, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 54751 + }, + { + "epoch": 0.54752, + "grad_norm": 1.307872035516649, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 54752 + }, + { + "epoch": 0.54753, + "grad_norm": 1.8473323770155132, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 54753 + }, + { + "epoch": 0.54754, + "grad_norm": 1.6618612641486294, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 54754 + }, + { + "epoch": 0.54755, + "grad_norm": 1.366009268703105, + "learning_rate": 0.003, + "loss": 4.018, + "step": 54755 + }, + { + "epoch": 0.54756, + "grad_norm": 1.481603741443726, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 54756 + }, + { + "epoch": 0.54757, + "grad_norm": 1.627538061931491, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 54757 + }, + { + "epoch": 0.54758, + "grad_norm": 1.6527477670863981, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 54758 + }, + { + "epoch": 0.54759, + "grad_norm": 1.4957862343011208, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 54759 + }, + { + "epoch": 0.5476, + "grad_norm": 1.6399394533991156, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 54760 + }, + { + "epoch": 0.54761, + "grad_norm": 1.4508016121202678, + "learning_rate": 0.003, + "loss": 4.041, + "step": 54761 + }, + { + "epoch": 0.54762, + "grad_norm": 1.8359541335688547, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 54762 + }, + { + "epoch": 0.54763, + "grad_norm": 1.5607422894338692, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 54763 + }, + { + "epoch": 0.54764, + "grad_norm": 1.6090957600477012, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 54764 + }, + { + "epoch": 0.54765, + "grad_norm": 1.7793048252783128, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 54765 + }, + { + "epoch": 0.54766, + "grad_norm": 1.54849852716584, + "learning_rate": 0.003, + "loss": 4.015, + "step": 54766 + }, + { + "epoch": 0.54767, + "grad_norm": 1.6844574496226983, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 54767 + }, + { + "epoch": 0.54768, + "grad_norm": 1.5080116736612739, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 54768 + }, + { + "epoch": 0.54769, + "grad_norm": 1.9454577190583313, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 54769 + }, + { + "epoch": 0.5477, + "grad_norm": 1.2892462190698446, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 54770 + }, + { + "epoch": 0.54771, + "grad_norm": 1.6762481610076267, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 54771 + }, + { + "epoch": 0.54772, + "grad_norm": 1.2165084916366313, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54772 + }, + { + "epoch": 0.54773, + "grad_norm": 1.5459372073740405, + "learning_rate": 0.003, + "loss": 4.045, + "step": 54773 + }, + { + "epoch": 0.54774, + "grad_norm": 1.5533112243374638, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 54774 + }, + { + "epoch": 0.54775, + "grad_norm": 1.4535939248937644, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 54775 + }, + { + "epoch": 0.54776, + "grad_norm": 1.362734266613085, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 54776 + }, + { + "epoch": 0.54777, + "grad_norm": 1.6143169152325834, + "learning_rate": 0.003, + "loss": 4.015, + "step": 54777 + }, + { + "epoch": 0.54778, + "grad_norm": 1.6244376257749027, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 54778 + }, + { + "epoch": 0.54779, + "grad_norm": 1.5985138511936103, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 54779 + }, + { + "epoch": 0.5478, + "grad_norm": 1.5439021147459673, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 54780 + }, + { + "epoch": 0.54781, + "grad_norm": 1.8274993621091833, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 54781 + }, + { + "epoch": 0.54782, + "grad_norm": 1.8633408583519615, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 54782 + }, + { + "epoch": 0.54783, + "grad_norm": 1.3344638693761, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 54783 + }, + { + "epoch": 0.54784, + "grad_norm": 1.7222546758378576, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 54784 + }, + { + "epoch": 0.54785, + "grad_norm": 1.7897943935685945, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 54785 + }, + { + "epoch": 0.54786, + "grad_norm": 1.8338525913715529, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 54786 + }, + { + "epoch": 0.54787, + "grad_norm": 1.295806075828714, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 54787 + }, + { + "epoch": 0.54788, + "grad_norm": 1.5919975621047575, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 54788 + }, + { + "epoch": 0.54789, + "grad_norm": 1.5406626127525294, + "learning_rate": 0.003, + "loss": 4.011, + "step": 54789 + }, + { + "epoch": 0.5479, + "grad_norm": 1.3115800208844204, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 54790 + }, + { + "epoch": 0.54791, + "grad_norm": 1.9219665370700407, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 54791 + }, + { + "epoch": 0.54792, + "grad_norm": 1.1037674222151637, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 54792 + }, + { + "epoch": 0.54793, + "grad_norm": 1.8176798464288892, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 54793 + }, + { + "epoch": 0.54794, + "grad_norm": 1.5355488607948595, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 54794 + }, + { + "epoch": 0.54795, + "grad_norm": 1.7451311257418183, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 54795 + }, + { + "epoch": 0.54796, + "grad_norm": 1.3386089233180984, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 54796 + }, + { + "epoch": 0.54797, + "grad_norm": 1.4740190405391354, + "learning_rate": 0.003, + "loss": 4.041, + "step": 54797 + }, + { + "epoch": 0.54798, + "grad_norm": 1.6660099080587019, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 54798 + }, + { + "epoch": 0.54799, + "grad_norm": 1.4064199015453205, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 54799 + }, + { + "epoch": 0.548, + "grad_norm": 1.7206082670507306, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 54800 + }, + { + "epoch": 0.54801, + "grad_norm": 1.617599683531985, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 54801 + }, + { + "epoch": 0.54802, + "grad_norm": 1.4193258393167694, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 54802 + }, + { + "epoch": 0.54803, + "grad_norm": 1.5982913208093346, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 54803 + }, + { + "epoch": 0.54804, + "grad_norm": 1.5614904332701347, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 54804 + }, + { + "epoch": 0.54805, + "grad_norm": 1.295008360675931, + "learning_rate": 0.003, + "loss": 4.032, + "step": 54805 + }, + { + "epoch": 0.54806, + "grad_norm": 1.3813285207170443, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 54806 + }, + { + "epoch": 0.54807, + "grad_norm": 1.7726705817818802, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 54807 + }, + { + "epoch": 0.54808, + "grad_norm": 1.7648477678784844, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 54808 + }, + { + "epoch": 0.54809, + "grad_norm": 1.4843576804552194, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54809 + }, + { + "epoch": 0.5481, + "grad_norm": 1.8464664289892256, + "learning_rate": 0.003, + "loss": 4.043, + "step": 54810 + }, + { + "epoch": 0.54811, + "grad_norm": 1.799814858262002, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 54811 + }, + { + "epoch": 0.54812, + "grad_norm": 1.763815845924664, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 54812 + }, + { + "epoch": 0.54813, + "grad_norm": 1.288626675130138, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 54813 + }, + { + "epoch": 0.54814, + "grad_norm": 1.855141640341266, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54814 + }, + { + "epoch": 0.54815, + "grad_norm": 1.382465290866033, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 54815 + }, + { + "epoch": 0.54816, + "grad_norm": 1.9221520216975363, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 54816 + }, + { + "epoch": 0.54817, + "grad_norm": 1.3418940124880832, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 54817 + }, + { + "epoch": 0.54818, + "grad_norm": 1.3784090641516253, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 54818 + }, + { + "epoch": 0.54819, + "grad_norm": 1.569206373077985, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 54819 + }, + { + "epoch": 0.5482, + "grad_norm": 1.4107717160778932, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 54820 + }, + { + "epoch": 0.54821, + "grad_norm": 2.049822202151055, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 54821 + }, + { + "epoch": 0.54822, + "grad_norm": 1.3036155223449748, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 54822 + }, + { + "epoch": 0.54823, + "grad_norm": 1.6518973711632572, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 54823 + }, + { + "epoch": 0.54824, + "grad_norm": 1.8355002614271714, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 54824 + }, + { + "epoch": 0.54825, + "grad_norm": 1.3478810479180383, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 54825 + }, + { + "epoch": 0.54826, + "grad_norm": 1.9028115040303408, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 54826 + }, + { + "epoch": 0.54827, + "grad_norm": 1.3140785013730016, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 54827 + }, + { + "epoch": 0.54828, + "grad_norm": 1.6274697681439145, + "learning_rate": 0.003, + "loss": 4.029, + "step": 54828 + }, + { + "epoch": 0.54829, + "grad_norm": 1.7771096964615243, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 54829 + }, + { + "epoch": 0.5483, + "grad_norm": 1.5540112628536829, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 54830 + }, + { + "epoch": 0.54831, + "grad_norm": 1.6645891227100535, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 54831 + }, + { + "epoch": 0.54832, + "grad_norm": 1.4920476966810348, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 54832 + }, + { + "epoch": 0.54833, + "grad_norm": 1.4721561707763444, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 54833 + }, + { + "epoch": 0.54834, + "grad_norm": 1.682428030031367, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 54834 + }, + { + "epoch": 0.54835, + "grad_norm": 1.3344951802520686, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 54835 + }, + { + "epoch": 0.54836, + "grad_norm": 1.531504414353498, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 54836 + }, + { + "epoch": 0.54837, + "grad_norm": 1.322705195664857, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 54837 + }, + { + "epoch": 0.54838, + "grad_norm": 1.6288517360658261, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 54838 + }, + { + "epoch": 0.54839, + "grad_norm": 1.417585827875026, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 54839 + }, + { + "epoch": 0.5484, + "grad_norm": 2.2175180574016413, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 54840 + }, + { + "epoch": 0.54841, + "grad_norm": 1.3885073128137577, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 54841 + }, + { + "epoch": 0.54842, + "grad_norm": 1.8523924185450629, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 54842 + }, + { + "epoch": 0.54843, + "grad_norm": 1.4313848792027983, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 54843 + }, + { + "epoch": 0.54844, + "grad_norm": 1.6030993384711847, + "learning_rate": 0.003, + "loss": 4.056, + "step": 54844 + }, + { + "epoch": 0.54845, + "grad_norm": 1.3949873130308954, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 54845 + }, + { + "epoch": 0.54846, + "grad_norm": 1.6103226441105145, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 54846 + }, + { + "epoch": 0.54847, + "grad_norm": 1.637559852326798, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 54847 + }, + { + "epoch": 0.54848, + "grad_norm": 1.582077300456499, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 54848 + }, + { + "epoch": 0.54849, + "grad_norm": 1.7220110554413797, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 54849 + }, + { + "epoch": 0.5485, + "grad_norm": 1.526801537682072, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 54850 + }, + { + "epoch": 0.54851, + "grad_norm": 1.8112664533456586, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 54851 + }, + { + "epoch": 0.54852, + "grad_norm": 1.6118991804216851, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 54852 + }, + { + "epoch": 0.54853, + "grad_norm": 1.5530048650364985, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 54853 + }, + { + "epoch": 0.54854, + "grad_norm": 1.7144838684628638, + "learning_rate": 0.003, + "loss": 4.013, + "step": 54854 + }, + { + "epoch": 0.54855, + "grad_norm": 1.4165136767669309, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 54855 + }, + { + "epoch": 0.54856, + "grad_norm": 1.4889560157372317, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 54856 + }, + { + "epoch": 0.54857, + "grad_norm": 1.473268312543302, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 54857 + }, + { + "epoch": 0.54858, + "grad_norm": 1.7472892240883648, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 54858 + }, + { + "epoch": 0.54859, + "grad_norm": 1.2341697393622926, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 54859 + }, + { + "epoch": 0.5486, + "grad_norm": 1.8202591810080788, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 54860 + }, + { + "epoch": 0.54861, + "grad_norm": 1.4165111199837608, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 54861 + }, + { + "epoch": 0.54862, + "grad_norm": 2.304964365964043, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 54862 + }, + { + "epoch": 0.54863, + "grad_norm": 1.271135878227687, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 54863 + }, + { + "epoch": 0.54864, + "grad_norm": 1.6726926346145397, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 54864 + }, + { + "epoch": 0.54865, + "grad_norm": 1.6834197552431636, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 54865 + }, + { + "epoch": 0.54866, + "grad_norm": 1.5684556853137719, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 54866 + }, + { + "epoch": 0.54867, + "grad_norm": 1.3372364373448966, + "learning_rate": 0.003, + "loss": 4.055, + "step": 54867 + }, + { + "epoch": 0.54868, + "grad_norm": 1.5447325303065884, + "learning_rate": 0.003, + "loss": 4.01, + "step": 54868 + }, + { + "epoch": 0.54869, + "grad_norm": 1.5503490392521424, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 54869 + }, + { + "epoch": 0.5487, + "grad_norm": 1.3979826848888772, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 54870 + }, + { + "epoch": 0.54871, + "grad_norm": 1.4939865599714657, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 54871 + }, + { + "epoch": 0.54872, + "grad_norm": 1.727432211599931, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 54872 + }, + { + "epoch": 0.54873, + "grad_norm": 1.7069427995294288, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 54873 + }, + { + "epoch": 0.54874, + "grad_norm": 1.5346441226609269, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 54874 + }, + { + "epoch": 0.54875, + "grad_norm": 1.6191007673195281, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 54875 + }, + { + "epoch": 0.54876, + "grad_norm": 1.5812065179919792, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 54876 + }, + { + "epoch": 0.54877, + "grad_norm": 1.5415476658822584, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 54877 + }, + { + "epoch": 0.54878, + "grad_norm": 1.441603343755806, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 54878 + }, + { + "epoch": 0.54879, + "grad_norm": 1.7167918834547082, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 54879 + }, + { + "epoch": 0.5488, + "grad_norm": 2.2601703588514925, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 54880 + }, + { + "epoch": 0.54881, + "grad_norm": 1.0414871566402417, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 54881 + }, + { + "epoch": 0.54882, + "grad_norm": 1.8325519198498479, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 54882 + }, + { + "epoch": 0.54883, + "grad_norm": 1.665692002569338, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 54883 + }, + { + "epoch": 0.54884, + "grad_norm": 1.9063417977479808, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 54884 + }, + { + "epoch": 0.54885, + "grad_norm": 1.5228842046050817, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 54885 + }, + { + "epoch": 0.54886, + "grad_norm": 1.6973912598848702, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 54886 + }, + { + "epoch": 0.54887, + "grad_norm": 1.3838304533976151, + "learning_rate": 0.003, + "loss": 4.05, + "step": 54887 + }, + { + "epoch": 0.54888, + "grad_norm": 1.3445371084305493, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 54888 + }, + { + "epoch": 0.54889, + "grad_norm": 1.769362638481852, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 54889 + }, + { + "epoch": 0.5489, + "grad_norm": 1.4965743112801635, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 54890 + }, + { + "epoch": 0.54891, + "grad_norm": 2.012966093962274, + "learning_rate": 0.003, + "loss": 4.013, + "step": 54891 + }, + { + "epoch": 0.54892, + "grad_norm": 1.487097789869073, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 54892 + }, + { + "epoch": 0.54893, + "grad_norm": 1.5245675963744094, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 54893 + }, + { + "epoch": 0.54894, + "grad_norm": 1.53258904475176, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 54894 + }, + { + "epoch": 0.54895, + "grad_norm": 1.7347453627377543, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 54895 + }, + { + "epoch": 0.54896, + "grad_norm": 1.6202288366080795, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 54896 + }, + { + "epoch": 0.54897, + "grad_norm": 1.7127110002525467, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54897 + }, + { + "epoch": 0.54898, + "grad_norm": 1.2507945585762044, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 54898 + }, + { + "epoch": 0.54899, + "grad_norm": 1.6702347577839574, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 54899 + }, + { + "epoch": 0.549, + "grad_norm": 1.3616853902289507, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 54900 + }, + { + "epoch": 0.54901, + "grad_norm": 1.5629489456410668, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 54901 + }, + { + "epoch": 0.54902, + "grad_norm": 1.594055138110132, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 54902 + }, + { + "epoch": 0.54903, + "grad_norm": 1.7258735374102454, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 54903 + }, + { + "epoch": 0.54904, + "grad_norm": 1.7919438483263126, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 54904 + }, + { + "epoch": 0.54905, + "grad_norm": 1.2706646872817742, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 54905 + }, + { + "epoch": 0.54906, + "grad_norm": 1.9076237181429614, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 54906 + }, + { + "epoch": 0.54907, + "grad_norm": 1.4637682476630323, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 54907 + }, + { + "epoch": 0.54908, + "grad_norm": 1.5614406880259966, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54908 + }, + { + "epoch": 0.54909, + "grad_norm": 1.4316248064027866, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 54909 + }, + { + "epoch": 0.5491, + "grad_norm": 1.436282900298818, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 54910 + }, + { + "epoch": 0.54911, + "grad_norm": 1.4315963529638687, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 54911 + }, + { + "epoch": 0.54912, + "grad_norm": 1.4141050544054359, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 54912 + }, + { + "epoch": 0.54913, + "grad_norm": 1.6062219134730036, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 54913 + }, + { + "epoch": 0.54914, + "grad_norm": 1.807163534077414, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 54914 + }, + { + "epoch": 0.54915, + "grad_norm": 1.7609542024999743, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 54915 + }, + { + "epoch": 0.54916, + "grad_norm": 1.7931708793171164, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 54916 + }, + { + "epoch": 0.54917, + "grad_norm": 1.3732638951477696, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 54917 + }, + { + "epoch": 0.54918, + "grad_norm": 1.8037915400971118, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 54918 + }, + { + "epoch": 0.54919, + "grad_norm": 1.215253910926411, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 54919 + }, + { + "epoch": 0.5492, + "grad_norm": 1.9169872381317241, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 54920 + }, + { + "epoch": 0.54921, + "grad_norm": 1.4098837191341103, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 54921 + }, + { + "epoch": 0.54922, + "grad_norm": 1.7780374763735862, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 54922 + }, + { + "epoch": 0.54923, + "grad_norm": 1.6826966398394136, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 54923 + }, + { + "epoch": 0.54924, + "grad_norm": 1.3735330478080254, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 54924 + }, + { + "epoch": 0.54925, + "grad_norm": 1.6981591500368372, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 54925 + }, + { + "epoch": 0.54926, + "grad_norm": 1.4403035109056967, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 54926 + }, + { + "epoch": 0.54927, + "grad_norm": 1.5557662949850017, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 54927 + }, + { + "epoch": 0.54928, + "grad_norm": 1.7131674723630859, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 54928 + }, + { + "epoch": 0.54929, + "grad_norm": 1.4197592619362984, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 54929 + }, + { + "epoch": 0.5493, + "grad_norm": 1.5107906318668076, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 54930 + }, + { + "epoch": 0.54931, + "grad_norm": 1.461021367315407, + "learning_rate": 0.003, + "loss": 4.007, + "step": 54931 + }, + { + "epoch": 0.54932, + "grad_norm": 1.879142119851692, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 54932 + }, + { + "epoch": 0.54933, + "grad_norm": 1.6141477596966485, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 54933 + }, + { + "epoch": 0.54934, + "grad_norm": 1.789008424740402, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 54934 + }, + { + "epoch": 0.54935, + "grad_norm": 1.9367197390925621, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 54935 + }, + { + "epoch": 0.54936, + "grad_norm": 1.2627309471158161, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 54936 + }, + { + "epoch": 0.54937, + "grad_norm": 1.384341736080556, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 54937 + }, + { + "epoch": 0.54938, + "grad_norm": 1.801196423938777, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 54938 + }, + { + "epoch": 0.54939, + "grad_norm": 1.5274317779732718, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 54939 + }, + { + "epoch": 0.5494, + "grad_norm": 1.8147006306832962, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 54940 + }, + { + "epoch": 0.54941, + "grad_norm": 1.4515088811230858, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 54941 + }, + { + "epoch": 0.54942, + "grad_norm": 1.4287369779741095, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 54942 + }, + { + "epoch": 0.54943, + "grad_norm": 1.5709401946092112, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 54943 + }, + { + "epoch": 0.54944, + "grad_norm": 1.5317593716032105, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 54944 + }, + { + "epoch": 0.54945, + "grad_norm": 1.5272832820785078, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 54945 + }, + { + "epoch": 0.54946, + "grad_norm": 1.4679914987015272, + "learning_rate": 0.003, + "loss": 4.05, + "step": 54946 + }, + { + "epoch": 0.54947, + "grad_norm": 1.478192207016561, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 54947 + }, + { + "epoch": 0.54948, + "grad_norm": 1.4277912300173876, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 54948 + }, + { + "epoch": 0.54949, + "grad_norm": 1.7257329619662258, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 54949 + }, + { + "epoch": 0.5495, + "grad_norm": 1.7760602527915024, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 54950 + }, + { + "epoch": 0.54951, + "grad_norm": 1.4607016974206442, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 54951 + }, + { + "epoch": 0.54952, + "grad_norm": 1.6826457400234285, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 54952 + }, + { + "epoch": 0.54953, + "grad_norm": 1.7330307726621375, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 54953 + }, + { + "epoch": 0.54954, + "grad_norm": 1.5893755250533212, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 54954 + }, + { + "epoch": 0.54955, + "grad_norm": 1.1843992718864451, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 54955 + }, + { + "epoch": 0.54956, + "grad_norm": 1.8515504137098164, + "learning_rate": 0.003, + "loss": 4.033, + "step": 54956 + }, + { + "epoch": 0.54957, + "grad_norm": 1.27359285291328, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 54957 + }, + { + "epoch": 0.54958, + "grad_norm": 1.7429912395323819, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 54958 + }, + { + "epoch": 0.54959, + "grad_norm": 1.4458205878856054, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 54959 + }, + { + "epoch": 0.5496, + "grad_norm": 1.6545023306959221, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 54960 + }, + { + "epoch": 0.54961, + "grad_norm": 1.26689571641963, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 54961 + }, + { + "epoch": 0.54962, + "grad_norm": 1.435181317601764, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 54962 + }, + { + "epoch": 0.54963, + "grad_norm": 1.8168020757947665, + "learning_rate": 0.003, + "loss": 4.03, + "step": 54963 + }, + { + "epoch": 0.54964, + "grad_norm": 1.4875273088185772, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 54964 + }, + { + "epoch": 0.54965, + "grad_norm": 2.207291490601016, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 54965 + }, + { + "epoch": 0.54966, + "grad_norm": 1.2728275545037926, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 54966 + }, + { + "epoch": 0.54967, + "grad_norm": 1.4989281693305458, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 54967 + }, + { + "epoch": 0.54968, + "grad_norm": 1.5816831580787651, + "learning_rate": 0.003, + "loss": 4.004, + "step": 54968 + }, + { + "epoch": 0.54969, + "grad_norm": 1.5755220521348485, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 54969 + }, + { + "epoch": 0.5497, + "grad_norm": 1.6553885301983728, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 54970 + }, + { + "epoch": 0.54971, + "grad_norm": 1.7494926341036225, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 54971 + }, + { + "epoch": 0.54972, + "grad_norm": 1.5089271805937223, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 54972 + }, + { + "epoch": 0.54973, + "grad_norm": 1.5595612049408483, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 54973 + }, + { + "epoch": 0.54974, + "grad_norm": 1.6649562301441978, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 54974 + }, + { + "epoch": 0.54975, + "grad_norm": 1.4621079276319164, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 54975 + }, + { + "epoch": 0.54976, + "grad_norm": 1.3531982751195573, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 54976 + }, + { + "epoch": 0.54977, + "grad_norm": 1.560962182767277, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 54977 + }, + { + "epoch": 0.54978, + "grad_norm": 1.33173263720279, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 54978 + }, + { + "epoch": 0.54979, + "grad_norm": 1.7280554486568929, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 54979 + }, + { + "epoch": 0.5498, + "grad_norm": 1.2305093247898315, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 54980 + }, + { + "epoch": 0.54981, + "grad_norm": 1.6413260081759726, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 54981 + }, + { + "epoch": 0.54982, + "grad_norm": 1.558683890624818, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 54982 + }, + { + "epoch": 0.54983, + "grad_norm": 2.1604270655437743, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 54983 + }, + { + "epoch": 0.54984, + "grad_norm": 1.5675881307350858, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 54984 + }, + { + "epoch": 0.54985, + "grad_norm": 1.9603548903395853, + "learning_rate": 0.003, + "loss": 4.031, + "step": 54985 + }, + { + "epoch": 0.54986, + "grad_norm": 1.6332383478563126, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 54986 + }, + { + "epoch": 0.54987, + "grad_norm": 1.316617947324807, + "learning_rate": 0.003, + "loss": 4.068, + "step": 54987 + }, + { + "epoch": 0.54988, + "grad_norm": 1.6182909763295603, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 54988 + }, + { + "epoch": 0.54989, + "grad_norm": 1.4079063723807705, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 54989 + }, + { + "epoch": 0.5499, + "grad_norm": 1.7223707945256048, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 54990 + }, + { + "epoch": 0.54991, + "grad_norm": 1.3185569510193518, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 54991 + }, + { + "epoch": 0.54992, + "grad_norm": 2.0392000977654976, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 54992 + }, + { + "epoch": 0.54993, + "grad_norm": 1.4489356535783269, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 54993 + }, + { + "epoch": 0.54994, + "grad_norm": 1.5096666086205133, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 54994 + }, + { + "epoch": 0.54995, + "grad_norm": 1.665837028822768, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 54995 + }, + { + "epoch": 0.54996, + "grad_norm": 1.724622775716618, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 54996 + }, + { + "epoch": 0.54997, + "grad_norm": 1.3951752245569424, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 54997 + }, + { + "epoch": 0.54998, + "grad_norm": 1.7193555490995043, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 54998 + }, + { + "epoch": 0.54999, + "grad_norm": 1.3746695681747794, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 54999 + }, + { + "epoch": 0.55, + "grad_norm": 1.449814393890209, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 55000 + }, + { + "epoch": 0.55001, + "grad_norm": 1.7277405459523583, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 55001 + }, + { + "epoch": 0.55002, + "grad_norm": 1.563908099747275, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 55002 + }, + { + "epoch": 0.55003, + "grad_norm": 1.9872606192750841, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 55003 + }, + { + "epoch": 0.55004, + "grad_norm": 1.5174812563728637, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55004 + }, + { + "epoch": 0.55005, + "grad_norm": 1.3469016280356525, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 55005 + }, + { + "epoch": 0.55006, + "grad_norm": 1.7174149237811998, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 55006 + }, + { + "epoch": 0.55007, + "grad_norm": 1.386473693558225, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 55007 + }, + { + "epoch": 0.55008, + "grad_norm": 1.8677963494879009, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 55008 + }, + { + "epoch": 0.55009, + "grad_norm": 1.321781836442289, + "learning_rate": 0.003, + "loss": 4.02, + "step": 55009 + }, + { + "epoch": 0.5501, + "grad_norm": 1.6699670223860694, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 55010 + }, + { + "epoch": 0.55011, + "grad_norm": 1.2148864013957943, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55011 + }, + { + "epoch": 0.55012, + "grad_norm": 1.9043727041495242, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 55012 + }, + { + "epoch": 0.55013, + "grad_norm": 1.5328958150566385, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 55013 + }, + { + "epoch": 0.55014, + "grad_norm": 1.370855486098911, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 55014 + }, + { + "epoch": 0.55015, + "grad_norm": 1.6497230792384827, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 55015 + }, + { + "epoch": 0.55016, + "grad_norm": 1.6897895690844436, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 55016 + }, + { + "epoch": 0.55017, + "grad_norm": 1.5382052483270483, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 55017 + }, + { + "epoch": 0.55018, + "grad_norm": 1.805735015371691, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 55018 + }, + { + "epoch": 0.55019, + "grad_norm": 1.584354177278839, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 55019 + }, + { + "epoch": 0.5502, + "grad_norm": 1.5136888689830268, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 55020 + }, + { + "epoch": 0.55021, + "grad_norm": 1.475066869220537, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 55021 + }, + { + "epoch": 0.55022, + "grad_norm": 1.4488051006026714, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 55022 + }, + { + "epoch": 0.55023, + "grad_norm": 1.3489570139511973, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 55023 + }, + { + "epoch": 0.55024, + "grad_norm": 1.5631589732016957, + "learning_rate": 0.003, + "loss": 4.04, + "step": 55024 + }, + { + "epoch": 0.55025, + "grad_norm": 1.5155933509608408, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 55025 + }, + { + "epoch": 0.55026, + "grad_norm": 1.8756226204980309, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 55026 + }, + { + "epoch": 0.55027, + "grad_norm": 1.7446168117425918, + "learning_rate": 0.003, + "loss": 4.008, + "step": 55027 + }, + { + "epoch": 0.55028, + "grad_norm": 1.6180695608069495, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 55028 + }, + { + "epoch": 0.55029, + "grad_norm": 1.6552500674687827, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 55029 + }, + { + "epoch": 0.5503, + "grad_norm": 1.5624964005757185, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 55030 + }, + { + "epoch": 0.55031, + "grad_norm": 1.3719660112094834, + "learning_rate": 0.003, + "loss": 4.046, + "step": 55031 + }, + { + "epoch": 0.55032, + "grad_norm": 1.525543600999305, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 55032 + }, + { + "epoch": 0.55033, + "grad_norm": 1.2707230032514831, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 55033 + }, + { + "epoch": 0.55034, + "grad_norm": 1.6282807171420102, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 55034 + }, + { + "epoch": 0.55035, + "grad_norm": 1.6025013551329024, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 55035 + }, + { + "epoch": 0.55036, + "grad_norm": 1.3104661710656471, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 55036 + }, + { + "epoch": 0.55037, + "grad_norm": 2.0225802134564086, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 55037 + }, + { + "epoch": 0.55038, + "grad_norm": 1.7237757830687095, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 55038 + }, + { + "epoch": 0.55039, + "grad_norm": 1.9949090705689214, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 55039 + }, + { + "epoch": 0.5504, + "grad_norm": 1.495848123073626, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 55040 + }, + { + "epoch": 0.55041, + "grad_norm": 1.768222290480666, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 55041 + }, + { + "epoch": 0.55042, + "grad_norm": 1.705855623479194, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 55042 + }, + { + "epoch": 0.55043, + "grad_norm": 1.4180898842392866, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 55043 + }, + { + "epoch": 0.55044, + "grad_norm": 1.614876668009854, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 55044 + }, + { + "epoch": 0.55045, + "grad_norm": 1.501094601962657, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 55045 + }, + { + "epoch": 0.55046, + "grad_norm": 1.4534182672999685, + "learning_rate": 0.003, + "loss": 4.003, + "step": 55046 + }, + { + "epoch": 0.55047, + "grad_norm": 1.4634219632067036, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 55047 + }, + { + "epoch": 0.55048, + "grad_norm": 1.9731875262553533, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 55048 + }, + { + "epoch": 0.55049, + "grad_norm": 1.2574983665677049, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 55049 + }, + { + "epoch": 0.5505, + "grad_norm": 1.9279567087657443, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 55050 + }, + { + "epoch": 0.55051, + "grad_norm": 1.4117278915557983, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 55051 + }, + { + "epoch": 0.55052, + "grad_norm": 1.5965009837357613, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 55052 + }, + { + "epoch": 0.55053, + "grad_norm": 1.555369675734444, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 55053 + }, + { + "epoch": 0.55054, + "grad_norm": 1.4186109625525098, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 55054 + }, + { + "epoch": 0.55055, + "grad_norm": 1.728547312336767, + "learning_rate": 0.003, + "loss": 4.05, + "step": 55055 + }, + { + "epoch": 0.55056, + "grad_norm": 1.509188966578072, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 55056 + }, + { + "epoch": 0.55057, + "grad_norm": 1.7904177689854404, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 55057 + }, + { + "epoch": 0.55058, + "grad_norm": 1.5119545842966435, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 55058 + }, + { + "epoch": 0.55059, + "grad_norm": 1.543897911579199, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 55059 + }, + { + "epoch": 0.5506, + "grad_norm": 1.1635907164607024, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 55060 + }, + { + "epoch": 0.55061, + "grad_norm": 1.9062293956720178, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 55061 + }, + { + "epoch": 0.55062, + "grad_norm": 1.408053398246928, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 55062 + }, + { + "epoch": 0.55063, + "grad_norm": 1.5762993473160691, + "learning_rate": 0.003, + "loss": 4.066, + "step": 55063 + }, + { + "epoch": 0.55064, + "grad_norm": 1.948135311661131, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 55064 + }, + { + "epoch": 0.55065, + "grad_norm": 1.8225219980758143, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 55065 + }, + { + "epoch": 0.55066, + "grad_norm": 1.3410163059690217, + "learning_rate": 0.003, + "loss": 4.023, + "step": 55066 + }, + { + "epoch": 0.55067, + "grad_norm": 2.007644683882491, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 55067 + }, + { + "epoch": 0.55068, + "grad_norm": 1.2627555204196315, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 55068 + }, + { + "epoch": 0.55069, + "grad_norm": 1.8583751009947131, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 55069 + }, + { + "epoch": 0.5507, + "grad_norm": 1.4606508437309502, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 55070 + }, + { + "epoch": 0.55071, + "grad_norm": 1.8330533283802053, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 55071 + }, + { + "epoch": 0.55072, + "grad_norm": 1.4442716037111607, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 55072 + }, + { + "epoch": 0.55073, + "grad_norm": 1.7390494224971307, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 55073 + }, + { + "epoch": 0.55074, + "grad_norm": 1.4205729351809837, + "learning_rate": 0.003, + "loss": 4.025, + "step": 55074 + }, + { + "epoch": 0.55075, + "grad_norm": 1.569802885632464, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 55075 + }, + { + "epoch": 0.55076, + "grad_norm": 1.4239746693660282, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 55076 + }, + { + "epoch": 0.55077, + "grad_norm": 1.429533489773045, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 55077 + }, + { + "epoch": 0.55078, + "grad_norm": 1.4721485739723137, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 55078 + }, + { + "epoch": 0.55079, + "grad_norm": 1.5664264509499628, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 55079 + }, + { + "epoch": 0.5508, + "grad_norm": 1.5820555297118883, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 55080 + }, + { + "epoch": 0.55081, + "grad_norm": 1.5775287578556967, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 55081 + }, + { + "epoch": 0.55082, + "grad_norm": 1.738359004431675, + "learning_rate": 0.003, + "loss": 4.08, + "step": 55082 + }, + { + "epoch": 0.55083, + "grad_norm": 1.1697867156638038, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 55083 + }, + { + "epoch": 0.55084, + "grad_norm": 1.60193812617496, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 55084 + }, + { + "epoch": 0.55085, + "grad_norm": 1.386142358532812, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 55085 + }, + { + "epoch": 0.55086, + "grad_norm": 1.7025496437200636, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 55086 + }, + { + "epoch": 0.55087, + "grad_norm": 1.6605268869938827, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 55087 + }, + { + "epoch": 0.55088, + "grad_norm": 1.5023822024594335, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 55088 + }, + { + "epoch": 0.55089, + "grad_norm": 1.817872297859285, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 55089 + }, + { + "epoch": 0.5509, + "grad_norm": 1.4569517563166459, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 55090 + }, + { + "epoch": 0.55091, + "grad_norm": 1.639683137558779, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 55091 + }, + { + "epoch": 0.55092, + "grad_norm": 1.342605580814228, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 55092 + }, + { + "epoch": 0.55093, + "grad_norm": 1.8954075408993616, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 55093 + }, + { + "epoch": 0.55094, + "grad_norm": 1.386959789196942, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 55094 + }, + { + "epoch": 0.55095, + "grad_norm": 1.7223983032782402, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 55095 + }, + { + "epoch": 0.55096, + "grad_norm": 1.572481642292489, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 55096 + }, + { + "epoch": 0.55097, + "grad_norm": 1.5310085266941649, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 55097 + }, + { + "epoch": 0.55098, + "grad_norm": 1.5234084891945285, + "learning_rate": 0.003, + "loss": 4.039, + "step": 55098 + }, + { + "epoch": 0.55099, + "grad_norm": 1.5779699366870028, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 55099 + }, + { + "epoch": 0.551, + "grad_norm": 1.5077567341362288, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55100 + }, + { + "epoch": 0.55101, + "grad_norm": 1.4327152456586045, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 55101 + }, + { + "epoch": 0.55102, + "grad_norm": 1.523945422702843, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 55102 + }, + { + "epoch": 0.55103, + "grad_norm": 1.5153870378888767, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 55103 + }, + { + "epoch": 0.55104, + "grad_norm": 1.8482799578690572, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 55104 + }, + { + "epoch": 0.55105, + "grad_norm": 1.3230351664563083, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 55105 + }, + { + "epoch": 0.55106, + "grad_norm": 1.8978260212245732, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 55106 + }, + { + "epoch": 0.55107, + "grad_norm": 1.5475927951661665, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 55107 + }, + { + "epoch": 0.55108, + "grad_norm": 1.4299778225806796, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 55108 + }, + { + "epoch": 0.55109, + "grad_norm": 1.4443292855231744, + "learning_rate": 0.003, + "loss": 4.008, + "step": 55109 + }, + { + "epoch": 0.5511, + "grad_norm": 1.9330928079939091, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 55110 + }, + { + "epoch": 0.55111, + "grad_norm": 1.5104712040863637, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 55111 + }, + { + "epoch": 0.55112, + "grad_norm": 1.492707219938163, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 55112 + }, + { + "epoch": 0.55113, + "grad_norm": 1.5420114057648613, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 55113 + }, + { + "epoch": 0.55114, + "grad_norm": 1.6452686601108633, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 55114 + }, + { + "epoch": 0.55115, + "grad_norm": 1.4298120879524963, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55115 + }, + { + "epoch": 0.55116, + "grad_norm": 1.5818612230069031, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 55116 + }, + { + "epoch": 0.55117, + "grad_norm": 1.6645254058920396, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 55117 + }, + { + "epoch": 0.55118, + "grad_norm": 1.6157470270583647, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 55118 + }, + { + "epoch": 0.55119, + "grad_norm": 1.3571725517784516, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 55119 + }, + { + "epoch": 0.5512, + "grad_norm": 1.7637674864016744, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 55120 + }, + { + "epoch": 0.55121, + "grad_norm": 1.4763598055747569, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 55121 + }, + { + "epoch": 0.55122, + "grad_norm": 1.649853559467284, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 55122 + }, + { + "epoch": 0.55123, + "grad_norm": 1.7516172448053682, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 55123 + }, + { + "epoch": 0.55124, + "grad_norm": 1.3537809885706151, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 55124 + }, + { + "epoch": 0.55125, + "grad_norm": 1.8715945417268067, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 55125 + }, + { + "epoch": 0.55126, + "grad_norm": 1.324368220657926, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 55126 + }, + { + "epoch": 0.55127, + "grad_norm": 1.8393903293235332, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 55127 + }, + { + "epoch": 0.55128, + "grad_norm": 1.3817614536783094, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 55128 + }, + { + "epoch": 0.55129, + "grad_norm": 1.6550797180775136, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 55129 + }, + { + "epoch": 0.5513, + "grad_norm": 1.2534952890870998, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 55130 + }, + { + "epoch": 0.55131, + "grad_norm": 1.9222405480460456, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 55131 + }, + { + "epoch": 0.55132, + "grad_norm": 1.4114601050076592, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 55132 + }, + { + "epoch": 0.55133, + "grad_norm": 1.873987323811858, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 55133 + }, + { + "epoch": 0.55134, + "grad_norm": 1.5465448973975542, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 55134 + }, + { + "epoch": 0.55135, + "grad_norm": 1.4709155697101395, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 55135 + }, + { + "epoch": 0.55136, + "grad_norm": 1.392652408886761, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 55136 + }, + { + "epoch": 0.55137, + "grad_norm": 1.7142570916857571, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 55137 + }, + { + "epoch": 0.55138, + "grad_norm": 1.7790930853790359, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 55138 + }, + { + "epoch": 0.55139, + "grad_norm": 1.5134859103965235, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 55139 + }, + { + "epoch": 0.5514, + "grad_norm": 1.775410860159514, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 55140 + }, + { + "epoch": 0.55141, + "grad_norm": 1.5013148348317622, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 55141 + }, + { + "epoch": 0.55142, + "grad_norm": 1.435831501100193, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 55142 + }, + { + "epoch": 0.55143, + "grad_norm": 1.7980286289103107, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 55143 + }, + { + "epoch": 0.55144, + "grad_norm": 1.2813144543058386, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 55144 + }, + { + "epoch": 0.55145, + "grad_norm": 1.9309353489348116, + "learning_rate": 0.003, + "loss": 4.054, + "step": 55145 + }, + { + "epoch": 0.55146, + "grad_norm": 1.3042456946802694, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 55146 + }, + { + "epoch": 0.55147, + "grad_norm": 1.7945269417012597, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 55147 + }, + { + "epoch": 0.55148, + "grad_norm": 1.6109961473549268, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 55148 + }, + { + "epoch": 0.55149, + "grad_norm": 1.4569592181642728, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 55149 + }, + { + "epoch": 0.5515, + "grad_norm": 1.6844022522142192, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 55150 + }, + { + "epoch": 0.55151, + "grad_norm": 1.1927955386794986, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 55151 + }, + { + "epoch": 0.55152, + "grad_norm": 1.8205902373965464, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 55152 + }, + { + "epoch": 0.55153, + "grad_norm": 1.6760573297840915, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 55153 + }, + { + "epoch": 0.55154, + "grad_norm": 1.6421009969810991, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 55154 + }, + { + "epoch": 0.55155, + "grad_norm": 1.410464370774006, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 55155 + }, + { + "epoch": 0.55156, + "grad_norm": 1.4790304545963715, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 55156 + }, + { + "epoch": 0.55157, + "grad_norm": 1.7369790632928415, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 55157 + }, + { + "epoch": 0.55158, + "grad_norm": 1.3721138831122468, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 55158 + }, + { + "epoch": 0.55159, + "grad_norm": 1.7883213368010915, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 55159 + }, + { + "epoch": 0.5516, + "grad_norm": 1.2894946211662088, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 55160 + }, + { + "epoch": 0.55161, + "grad_norm": 1.6922880795567248, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 55161 + }, + { + "epoch": 0.55162, + "grad_norm": 1.3795455861006731, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 55162 + }, + { + "epoch": 0.55163, + "grad_norm": 2.051480415971494, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 55163 + }, + { + "epoch": 0.55164, + "grad_norm": 1.48818577521888, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 55164 + }, + { + "epoch": 0.55165, + "grad_norm": 1.6082531667815418, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 55165 + }, + { + "epoch": 0.55166, + "grad_norm": 1.4679616134779565, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 55166 + }, + { + "epoch": 0.55167, + "grad_norm": 1.6457048936576293, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 55167 + }, + { + "epoch": 0.55168, + "grad_norm": 1.3959245829822062, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 55168 + }, + { + "epoch": 0.55169, + "grad_norm": 2.038369394582539, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 55169 + }, + { + "epoch": 0.5517, + "grad_norm": 1.3563182352133256, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 55170 + }, + { + "epoch": 0.55171, + "grad_norm": 1.7770516333648494, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 55171 + }, + { + "epoch": 0.55172, + "grad_norm": 1.7188167247667625, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 55172 + }, + { + "epoch": 0.55173, + "grad_norm": 1.499569486117863, + "learning_rate": 0.003, + "loss": 4.034, + "step": 55173 + }, + { + "epoch": 0.55174, + "grad_norm": 1.6168596690353507, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 55174 + }, + { + "epoch": 0.55175, + "grad_norm": 1.3000028183715608, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 55175 + }, + { + "epoch": 0.55176, + "grad_norm": 1.822020432713553, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 55176 + }, + { + "epoch": 0.55177, + "grad_norm": 1.552230832193156, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 55177 + }, + { + "epoch": 0.55178, + "grad_norm": 1.9159588413015323, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 55178 + }, + { + "epoch": 0.55179, + "grad_norm": 1.427839095785095, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 55179 + }, + { + "epoch": 0.5518, + "grad_norm": 1.6711932921822583, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 55180 + }, + { + "epoch": 0.55181, + "grad_norm": 1.6176243415709124, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 55181 + }, + { + "epoch": 0.55182, + "grad_norm": 1.3496547926565081, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 55182 + }, + { + "epoch": 0.55183, + "grad_norm": 1.5450540050151773, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 55183 + }, + { + "epoch": 0.55184, + "grad_norm": 1.5443326685526484, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55184 + }, + { + "epoch": 0.55185, + "grad_norm": 1.256875154238345, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 55185 + }, + { + "epoch": 0.55186, + "grad_norm": 1.630758009581555, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 55186 + }, + { + "epoch": 0.55187, + "grad_norm": 1.398395270904639, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 55187 + }, + { + "epoch": 0.55188, + "grad_norm": 1.8133504013951445, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 55188 + }, + { + "epoch": 0.55189, + "grad_norm": 1.5229448758894446, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 55189 + }, + { + "epoch": 0.5519, + "grad_norm": 1.5489419577899544, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 55190 + }, + { + "epoch": 0.55191, + "grad_norm": 1.6887162372316535, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 55191 + }, + { + "epoch": 0.55192, + "grad_norm": 1.5333849753454512, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 55192 + }, + { + "epoch": 0.55193, + "grad_norm": 1.8570936634790876, + "learning_rate": 0.003, + "loss": 4.028, + "step": 55193 + }, + { + "epoch": 0.55194, + "grad_norm": 1.442366000847722, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55194 + }, + { + "epoch": 0.55195, + "grad_norm": 1.7631548797518504, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 55195 + }, + { + "epoch": 0.55196, + "grad_norm": 1.431185905875668, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55196 + }, + { + "epoch": 0.55197, + "grad_norm": 1.7238980308720444, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 55197 + }, + { + "epoch": 0.55198, + "grad_norm": 1.5172335553466192, + "learning_rate": 0.003, + "loss": 4.023, + "step": 55198 + }, + { + "epoch": 0.55199, + "grad_norm": 1.7975794923635648, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 55199 + }, + { + "epoch": 0.552, + "grad_norm": 1.473135695256772, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 55200 + }, + { + "epoch": 0.55201, + "grad_norm": 1.6875283494245197, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 55201 + }, + { + "epoch": 0.55202, + "grad_norm": 1.6163923846875705, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 55202 + }, + { + "epoch": 0.55203, + "grad_norm": 1.371595159186003, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 55203 + }, + { + "epoch": 0.55204, + "grad_norm": 1.440769851613525, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 55204 + }, + { + "epoch": 0.55205, + "grad_norm": 1.50261571971682, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 55205 + }, + { + "epoch": 0.55206, + "grad_norm": 1.4697957035227716, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 55206 + }, + { + "epoch": 0.55207, + "grad_norm": 1.5921813292854055, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 55207 + }, + { + "epoch": 0.55208, + "grad_norm": 1.4222524224909392, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 55208 + }, + { + "epoch": 0.55209, + "grad_norm": 1.8086057205097783, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55209 + }, + { + "epoch": 0.5521, + "grad_norm": 1.4134852204223702, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55210 + }, + { + "epoch": 0.55211, + "grad_norm": 1.6283021974804281, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 55211 + }, + { + "epoch": 0.55212, + "grad_norm": 1.5259339724948535, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 55212 + }, + { + "epoch": 0.55213, + "grad_norm": 1.6173262458742594, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 55213 + }, + { + "epoch": 0.55214, + "grad_norm": 1.8908720441798863, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 55214 + }, + { + "epoch": 0.55215, + "grad_norm": 1.7425603602183446, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 55215 + }, + { + "epoch": 0.55216, + "grad_norm": 1.456918127148663, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 55216 + }, + { + "epoch": 0.55217, + "grad_norm": 1.7546048558956735, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55217 + }, + { + "epoch": 0.55218, + "grad_norm": 1.7360372415208776, + "learning_rate": 0.003, + "loss": 4.011, + "step": 55218 + }, + { + "epoch": 0.55219, + "grad_norm": 1.4087145797122322, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55219 + }, + { + "epoch": 0.5522, + "grad_norm": 1.776997493916151, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 55220 + }, + { + "epoch": 0.55221, + "grad_norm": 1.41015463486106, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 55221 + }, + { + "epoch": 0.55222, + "grad_norm": 1.302886772746629, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 55222 + }, + { + "epoch": 0.55223, + "grad_norm": 1.4886678019986914, + "learning_rate": 0.003, + "loss": 4.047, + "step": 55223 + }, + { + "epoch": 0.55224, + "grad_norm": 1.5587960169756612, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 55224 + }, + { + "epoch": 0.55225, + "grad_norm": 1.5722043087267428, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 55225 + }, + { + "epoch": 0.55226, + "grad_norm": 1.5589754996824465, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 55226 + }, + { + "epoch": 0.55227, + "grad_norm": 1.789160979938708, + "learning_rate": 0.003, + "loss": 4.033, + "step": 55227 + }, + { + "epoch": 0.55228, + "grad_norm": 1.7041810079445932, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 55228 + }, + { + "epoch": 0.55229, + "grad_norm": 2.049351543863111, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 55229 + }, + { + "epoch": 0.5523, + "grad_norm": 1.6705794849829023, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 55230 + }, + { + "epoch": 0.55231, + "grad_norm": 1.1800315381242021, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 55231 + }, + { + "epoch": 0.55232, + "grad_norm": 1.6813484169934336, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 55232 + }, + { + "epoch": 0.55233, + "grad_norm": 1.5284633579901645, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 55233 + }, + { + "epoch": 0.55234, + "grad_norm": 1.5944001193090998, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 55234 + }, + { + "epoch": 0.55235, + "grad_norm": 1.4658059627391449, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 55235 + }, + { + "epoch": 0.55236, + "grad_norm": 1.5718422340027811, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 55236 + }, + { + "epoch": 0.55237, + "grad_norm": 1.6138992219434922, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 55237 + }, + { + "epoch": 0.55238, + "grad_norm": 1.793099269645201, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 55238 + }, + { + "epoch": 0.55239, + "grad_norm": 1.2988942353558555, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 55239 + }, + { + "epoch": 0.5524, + "grad_norm": 1.8630300824205568, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55240 + }, + { + "epoch": 0.55241, + "grad_norm": 1.2991141604405123, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 55241 + }, + { + "epoch": 0.55242, + "grad_norm": 1.5380836935332756, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 55242 + }, + { + "epoch": 0.55243, + "grad_norm": 1.6223523224649814, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 55243 + }, + { + "epoch": 0.55244, + "grad_norm": 1.5721004896235964, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 55244 + }, + { + "epoch": 0.55245, + "grad_norm": 1.5713848618381636, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 55245 + }, + { + "epoch": 0.55246, + "grad_norm": 1.7287991636373674, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 55246 + }, + { + "epoch": 0.55247, + "grad_norm": 1.663696674949889, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 55247 + }, + { + "epoch": 0.55248, + "grad_norm": 1.5977511287687052, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 55248 + }, + { + "epoch": 0.55249, + "grad_norm": 1.7698940523615219, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 55249 + }, + { + "epoch": 0.5525, + "grad_norm": 1.5884395171151733, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 55250 + }, + { + "epoch": 0.55251, + "grad_norm": 1.4601383043464145, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 55251 + }, + { + "epoch": 0.55252, + "grad_norm": 1.254087200205371, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 55252 + }, + { + "epoch": 0.55253, + "grad_norm": 1.5417453469515947, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 55253 + }, + { + "epoch": 0.55254, + "grad_norm": 1.3878820356426413, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 55254 + }, + { + "epoch": 0.55255, + "grad_norm": 1.7021803358107086, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 55255 + }, + { + "epoch": 0.55256, + "grad_norm": 1.7456827986758747, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 55256 + }, + { + "epoch": 0.55257, + "grad_norm": 1.531229066101312, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 55257 + }, + { + "epoch": 0.55258, + "grad_norm": 1.5133225991258845, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 55258 + }, + { + "epoch": 0.55259, + "grad_norm": 1.7573127501667651, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 55259 + }, + { + "epoch": 0.5526, + "grad_norm": 1.3570572950060988, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 55260 + }, + { + "epoch": 0.55261, + "grad_norm": 1.6375097387516304, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 55261 + }, + { + "epoch": 0.55262, + "grad_norm": 1.8759917356665563, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 55262 + }, + { + "epoch": 0.55263, + "grad_norm": 1.2257590612106504, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 55263 + }, + { + "epoch": 0.55264, + "grad_norm": 1.5605586097852515, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 55264 + }, + { + "epoch": 0.55265, + "grad_norm": 1.41306066220991, + "learning_rate": 0.003, + "loss": 4.041, + "step": 55265 + }, + { + "epoch": 0.55266, + "grad_norm": 1.691935001292243, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 55266 + }, + { + "epoch": 0.55267, + "grad_norm": 1.5840063304164078, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 55267 + }, + { + "epoch": 0.55268, + "grad_norm": 1.7956179446382938, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 55268 + }, + { + "epoch": 0.55269, + "grad_norm": 1.4685378514832756, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 55269 + }, + { + "epoch": 0.5527, + "grad_norm": 1.53297931979942, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 55270 + }, + { + "epoch": 0.55271, + "grad_norm": 1.6753760074256656, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 55271 + }, + { + "epoch": 0.55272, + "grad_norm": 1.3084231388341738, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 55272 + }, + { + "epoch": 0.55273, + "grad_norm": 1.9076804707774013, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 55273 + }, + { + "epoch": 0.55274, + "grad_norm": 1.5862847792628025, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55274 + }, + { + "epoch": 0.55275, + "grad_norm": 1.617965604879433, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 55275 + }, + { + "epoch": 0.55276, + "grad_norm": 1.608196276433938, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 55276 + }, + { + "epoch": 0.55277, + "grad_norm": 1.57015603108876, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 55277 + }, + { + "epoch": 0.55278, + "grad_norm": 1.6472204210441663, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 55278 + }, + { + "epoch": 0.55279, + "grad_norm": 1.8667387528708788, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 55279 + }, + { + "epoch": 0.5528, + "grad_norm": 1.3739418856883394, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 55280 + }, + { + "epoch": 0.55281, + "grad_norm": 1.9795089123627465, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 55281 + }, + { + "epoch": 0.55282, + "grad_norm": 1.8303348633360945, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 55282 + }, + { + "epoch": 0.55283, + "grad_norm": 1.615943271249922, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 55283 + }, + { + "epoch": 0.55284, + "grad_norm": 1.6898811670409988, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 55284 + }, + { + "epoch": 0.55285, + "grad_norm": 1.436471942824804, + "learning_rate": 0.003, + "loss": 4.025, + "step": 55285 + }, + { + "epoch": 0.55286, + "grad_norm": 1.5633942020647273, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 55286 + }, + { + "epoch": 0.55287, + "grad_norm": 1.2757377664954106, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 55287 + }, + { + "epoch": 0.55288, + "grad_norm": 1.6959737289397057, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 55288 + }, + { + "epoch": 0.55289, + "grad_norm": 1.1446225181495022, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 55289 + }, + { + "epoch": 0.5529, + "grad_norm": 1.8752561016073577, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 55290 + }, + { + "epoch": 0.55291, + "grad_norm": 1.2275226723386408, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 55291 + }, + { + "epoch": 0.55292, + "grad_norm": 1.8403573905608854, + "learning_rate": 0.003, + "loss": 4.07, + "step": 55292 + }, + { + "epoch": 0.55293, + "grad_norm": 1.65146255751541, + "learning_rate": 0.003, + "loss": 4.038, + "step": 55293 + }, + { + "epoch": 0.55294, + "grad_norm": 1.7911203282130592, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 55294 + }, + { + "epoch": 0.55295, + "grad_norm": 1.6477046459387648, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 55295 + }, + { + "epoch": 0.55296, + "grad_norm": 1.8457968013993842, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 55296 + }, + { + "epoch": 0.55297, + "grad_norm": 1.401472026025471, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 55297 + }, + { + "epoch": 0.55298, + "grad_norm": 1.6469670578748845, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 55298 + }, + { + "epoch": 0.55299, + "grad_norm": 1.3693496204085067, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 55299 + }, + { + "epoch": 0.553, + "grad_norm": 1.6551997594875805, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 55300 + }, + { + "epoch": 0.55301, + "grad_norm": 1.6135364564297596, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 55301 + }, + { + "epoch": 0.55302, + "grad_norm": 1.6671414163722749, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 55302 + }, + { + "epoch": 0.55303, + "grad_norm": 1.52909913462354, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 55303 + }, + { + "epoch": 0.55304, + "grad_norm": 1.3827177926653773, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 55304 + }, + { + "epoch": 0.55305, + "grad_norm": 1.5708183817249268, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55305 + }, + { + "epoch": 0.55306, + "grad_norm": 1.4483037890586754, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 55306 + }, + { + "epoch": 0.55307, + "grad_norm": 1.5480887883675776, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 55307 + }, + { + "epoch": 0.55308, + "grad_norm": 1.7407958707440025, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 55308 + }, + { + "epoch": 0.55309, + "grad_norm": 1.6344050290327174, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 55309 + }, + { + "epoch": 0.5531, + "grad_norm": 1.4167590372874, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 55310 + }, + { + "epoch": 0.55311, + "grad_norm": 1.6440304851529206, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 55311 + }, + { + "epoch": 0.55312, + "grad_norm": 1.6912105716054169, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 55312 + }, + { + "epoch": 0.55313, + "grad_norm": 1.445439311582907, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 55313 + }, + { + "epoch": 0.55314, + "grad_norm": 1.7995326429207679, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 55314 + }, + { + "epoch": 0.55315, + "grad_norm": 1.6541723137684163, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 55315 + }, + { + "epoch": 0.55316, + "grad_norm": 1.4915082371707749, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 55316 + }, + { + "epoch": 0.55317, + "grad_norm": 1.6515782061771174, + "learning_rate": 0.003, + "loss": 4.003, + "step": 55317 + }, + { + "epoch": 0.55318, + "grad_norm": 1.4972370425081378, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 55318 + }, + { + "epoch": 0.55319, + "grad_norm": 1.5327817289431238, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 55319 + }, + { + "epoch": 0.5532, + "grad_norm": 1.3742596815688257, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 55320 + }, + { + "epoch": 0.55321, + "grad_norm": 1.6741679624413224, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 55321 + }, + { + "epoch": 0.55322, + "grad_norm": 1.5979584544175114, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 55322 + }, + { + "epoch": 0.55323, + "grad_norm": 1.1985993138023836, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 55323 + }, + { + "epoch": 0.55324, + "grad_norm": 1.7218776065804464, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 55324 + }, + { + "epoch": 0.55325, + "grad_norm": 1.3890472044695517, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 55325 + }, + { + "epoch": 0.55326, + "grad_norm": 1.8098561858863036, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 55326 + }, + { + "epoch": 0.55327, + "grad_norm": 1.6661224819352207, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 55327 + }, + { + "epoch": 0.55328, + "grad_norm": 1.5862170025289581, + "learning_rate": 0.003, + "loss": 4.051, + "step": 55328 + }, + { + "epoch": 0.55329, + "grad_norm": 1.9995572296788642, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 55329 + }, + { + "epoch": 0.5533, + "grad_norm": 1.3484360818068246, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 55330 + }, + { + "epoch": 0.55331, + "grad_norm": 1.6090863918669502, + "learning_rate": 0.003, + "loss": 4.057, + "step": 55331 + }, + { + "epoch": 0.55332, + "grad_norm": 1.5417956140055735, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 55332 + }, + { + "epoch": 0.55333, + "grad_norm": 1.705613129987265, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 55333 + }, + { + "epoch": 0.55334, + "grad_norm": 1.7112741730557517, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 55334 + }, + { + "epoch": 0.55335, + "grad_norm": 1.5283634572829, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55335 + }, + { + "epoch": 0.55336, + "grad_norm": 1.543120789712945, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 55336 + }, + { + "epoch": 0.55337, + "grad_norm": 1.4899881053142965, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 55337 + }, + { + "epoch": 0.55338, + "grad_norm": 1.5314832189222156, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 55338 + }, + { + "epoch": 0.55339, + "grad_norm": 1.5702857584159542, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 55339 + }, + { + "epoch": 0.5534, + "grad_norm": 1.7335838760113593, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 55340 + }, + { + "epoch": 0.55341, + "grad_norm": 1.667980748045432, + "learning_rate": 0.003, + "loss": 4.029, + "step": 55341 + }, + { + "epoch": 0.55342, + "grad_norm": 1.6128899671599162, + "learning_rate": 0.003, + "loss": 4.034, + "step": 55342 + }, + { + "epoch": 0.55343, + "grad_norm": 1.4393278900048603, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 55343 + }, + { + "epoch": 0.55344, + "grad_norm": 1.7097953579139826, + "learning_rate": 0.003, + "loss": 4.043, + "step": 55344 + }, + { + "epoch": 0.55345, + "grad_norm": 1.5963586382295791, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 55345 + }, + { + "epoch": 0.55346, + "grad_norm": 1.6278909611216983, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 55346 + }, + { + "epoch": 0.55347, + "grad_norm": 1.6645298870914587, + "learning_rate": 0.003, + "loss": 4.023, + "step": 55347 + }, + { + "epoch": 0.55348, + "grad_norm": 1.4399740772205227, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 55348 + }, + { + "epoch": 0.55349, + "grad_norm": 1.3887018822578892, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 55349 + }, + { + "epoch": 0.5535, + "grad_norm": 1.4049489199138103, + "learning_rate": 0.003, + "loss": 4.014, + "step": 55350 + }, + { + "epoch": 0.55351, + "grad_norm": 1.3413666244482496, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 55351 + }, + { + "epoch": 0.55352, + "grad_norm": 1.5992395052243196, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 55352 + }, + { + "epoch": 0.55353, + "grad_norm": 1.3813403739636678, + "learning_rate": 0.003, + "loss": 4.024, + "step": 55353 + }, + { + "epoch": 0.55354, + "grad_norm": 2.234457040182082, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 55354 + }, + { + "epoch": 0.55355, + "grad_norm": 1.5081185470538838, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 55355 + }, + { + "epoch": 0.55356, + "grad_norm": 1.621011010840887, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 55356 + }, + { + "epoch": 0.55357, + "grad_norm": 1.9941652990344756, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 55357 + }, + { + "epoch": 0.55358, + "grad_norm": 1.3580878963820682, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 55358 + }, + { + "epoch": 0.55359, + "grad_norm": 1.6180729406297973, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 55359 + }, + { + "epoch": 0.5536, + "grad_norm": 1.4801108847758166, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 55360 + }, + { + "epoch": 0.55361, + "grad_norm": 1.5331569849823092, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 55361 + }, + { + "epoch": 0.55362, + "grad_norm": 1.6537197779302217, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 55362 + }, + { + "epoch": 0.55363, + "grad_norm": 1.4123257865099208, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 55363 + }, + { + "epoch": 0.55364, + "grad_norm": 1.7091067232138633, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 55364 + }, + { + "epoch": 0.55365, + "grad_norm": 1.5689521736886212, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 55365 + }, + { + "epoch": 0.55366, + "grad_norm": 1.7550368404050873, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 55366 + }, + { + "epoch": 0.55367, + "grad_norm": 1.505686105384609, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 55367 + }, + { + "epoch": 0.55368, + "grad_norm": 1.6264636653950446, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 55368 + }, + { + "epoch": 0.55369, + "grad_norm": 1.7257308520658876, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 55369 + }, + { + "epoch": 0.5537, + "grad_norm": 1.4599381854499693, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 55370 + }, + { + "epoch": 0.55371, + "grad_norm": 1.6738116211046683, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 55371 + }, + { + "epoch": 0.55372, + "grad_norm": 1.7561660579277771, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 55372 + }, + { + "epoch": 0.55373, + "grad_norm": 1.6185642999604186, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 55373 + }, + { + "epoch": 0.55374, + "grad_norm": 1.607603074062529, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 55374 + }, + { + "epoch": 0.55375, + "grad_norm": 1.6071941830893106, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 55375 + }, + { + "epoch": 0.55376, + "grad_norm": 1.7789734992954893, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 55376 + }, + { + "epoch": 0.55377, + "grad_norm": 1.8451147209504044, + "learning_rate": 0.003, + "loss": 4.04, + "step": 55377 + }, + { + "epoch": 0.55378, + "grad_norm": 1.389310472236995, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 55378 + }, + { + "epoch": 0.55379, + "grad_norm": 1.6390491403314926, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 55379 + }, + { + "epoch": 0.5538, + "grad_norm": 1.6289716042204356, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 55380 + }, + { + "epoch": 0.55381, + "grad_norm": 1.5601526288469683, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 55381 + }, + { + "epoch": 0.55382, + "grad_norm": 1.4685533924286776, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 55382 + }, + { + "epoch": 0.55383, + "grad_norm": 1.577162187798675, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 55383 + }, + { + "epoch": 0.55384, + "grad_norm": 1.369702919045224, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 55384 + }, + { + "epoch": 0.55385, + "grad_norm": 1.622674601050109, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 55385 + }, + { + "epoch": 0.55386, + "grad_norm": 1.5247960322942764, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55386 + }, + { + "epoch": 0.55387, + "grad_norm": 1.6821230340398265, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 55387 + }, + { + "epoch": 0.55388, + "grad_norm": 1.6680952490363248, + "learning_rate": 0.003, + "loss": 4.011, + "step": 55388 + }, + { + "epoch": 0.55389, + "grad_norm": 1.82588232823389, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 55389 + }, + { + "epoch": 0.5539, + "grad_norm": 1.6131773597265948, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 55390 + }, + { + "epoch": 0.55391, + "grad_norm": 1.5151561068846062, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 55391 + }, + { + "epoch": 0.55392, + "grad_norm": 1.739656937942846, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 55392 + }, + { + "epoch": 0.55393, + "grad_norm": 1.556118077636575, + "learning_rate": 0.003, + "loss": 3.991, + "step": 55393 + }, + { + "epoch": 0.55394, + "grad_norm": 1.724921000176868, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 55394 + }, + { + "epoch": 0.55395, + "grad_norm": 1.5377696063046096, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 55395 + }, + { + "epoch": 0.55396, + "grad_norm": 1.654994785610226, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 55396 + }, + { + "epoch": 0.55397, + "grad_norm": 1.2697497307352714, + "learning_rate": 0.003, + "loss": 4.064, + "step": 55397 + }, + { + "epoch": 0.55398, + "grad_norm": 1.5515501325502252, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 55398 + }, + { + "epoch": 0.55399, + "grad_norm": 1.619779535679676, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 55399 + }, + { + "epoch": 0.554, + "grad_norm": 1.6186954672471205, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 55400 + }, + { + "epoch": 0.55401, + "grad_norm": 1.4874291357238243, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 55401 + }, + { + "epoch": 0.55402, + "grad_norm": 1.4766439940532368, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 55402 + }, + { + "epoch": 0.55403, + "grad_norm": 1.7223845071473098, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 55403 + }, + { + "epoch": 0.55404, + "grad_norm": 1.5946551894881889, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 55404 + }, + { + "epoch": 0.55405, + "grad_norm": 1.951662346000338, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 55405 + }, + { + "epoch": 0.55406, + "grad_norm": 1.332071280180484, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 55406 + }, + { + "epoch": 0.55407, + "grad_norm": 1.6669837901776534, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 55407 + }, + { + "epoch": 0.55408, + "grad_norm": 2.082758462306685, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 55408 + }, + { + "epoch": 0.55409, + "grad_norm": 1.2223922295104557, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 55409 + }, + { + "epoch": 0.5541, + "grad_norm": 1.5828866953066967, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 55410 + }, + { + "epoch": 0.55411, + "grad_norm": 2.3585143045854826, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 55411 + }, + { + "epoch": 0.55412, + "grad_norm": 1.2200698663949918, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 55412 + }, + { + "epoch": 0.55413, + "grad_norm": 1.8680417808390104, + "learning_rate": 0.003, + "loss": 4.053, + "step": 55413 + }, + { + "epoch": 0.55414, + "grad_norm": 1.3195977003083335, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 55414 + }, + { + "epoch": 0.55415, + "grad_norm": 1.7987927171256284, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 55415 + }, + { + "epoch": 0.55416, + "grad_norm": 1.5205944696330187, + "learning_rate": 0.003, + "loss": 4.066, + "step": 55416 + }, + { + "epoch": 0.55417, + "grad_norm": 1.546553857997607, + "learning_rate": 0.003, + "loss": 4.023, + "step": 55417 + }, + { + "epoch": 0.55418, + "grad_norm": 1.5522123415816549, + "learning_rate": 0.003, + "loss": 4.015, + "step": 55418 + }, + { + "epoch": 0.55419, + "grad_norm": 1.6466841924071511, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 55419 + }, + { + "epoch": 0.5542, + "grad_norm": 1.5422259591478162, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 55420 + }, + { + "epoch": 0.55421, + "grad_norm": 1.523336122947932, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 55421 + }, + { + "epoch": 0.55422, + "grad_norm": 1.3250858617997952, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 55422 + }, + { + "epoch": 0.55423, + "grad_norm": 1.7921417815185514, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 55423 + }, + { + "epoch": 0.55424, + "grad_norm": 1.4803051042136646, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 55424 + }, + { + "epoch": 0.55425, + "grad_norm": 1.539827144502522, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 55425 + }, + { + "epoch": 0.55426, + "grad_norm": 1.4376659207347202, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 55426 + }, + { + "epoch": 0.55427, + "grad_norm": 2.0261093240936243, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 55427 + }, + { + "epoch": 0.55428, + "grad_norm": 1.5889088960019953, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 55428 + }, + { + "epoch": 0.55429, + "grad_norm": 1.5581616967205814, + "learning_rate": 0.003, + "loss": 4.039, + "step": 55429 + }, + { + "epoch": 0.5543, + "grad_norm": 1.7325612657030593, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 55430 + }, + { + "epoch": 0.55431, + "grad_norm": 1.7544331005450329, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 55431 + }, + { + "epoch": 0.55432, + "grad_norm": 1.312628209064902, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 55432 + }, + { + "epoch": 0.55433, + "grad_norm": 1.9755000770305968, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 55433 + }, + { + "epoch": 0.55434, + "grad_norm": 1.2936695003798835, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 55434 + }, + { + "epoch": 0.55435, + "grad_norm": 1.7023017529878524, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 55435 + }, + { + "epoch": 0.55436, + "grad_norm": 1.445260183313271, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55436 + }, + { + "epoch": 0.55437, + "grad_norm": 1.6019130936456345, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 55437 + }, + { + "epoch": 0.55438, + "grad_norm": 1.8106851408824505, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 55438 + }, + { + "epoch": 0.55439, + "grad_norm": 1.420967487841883, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 55439 + }, + { + "epoch": 0.5544, + "grad_norm": 1.6962366300494363, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 55440 + }, + { + "epoch": 0.55441, + "grad_norm": 1.7752774760619043, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 55441 + }, + { + "epoch": 0.55442, + "grad_norm": 1.2977423950278257, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 55442 + }, + { + "epoch": 0.55443, + "grad_norm": 1.7792504789381853, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 55443 + }, + { + "epoch": 0.55444, + "grad_norm": 1.5203673462952438, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 55444 + }, + { + "epoch": 0.55445, + "grad_norm": 1.7281424978979105, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 55445 + }, + { + "epoch": 0.55446, + "grad_norm": 1.4994023251558297, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 55446 + }, + { + "epoch": 0.55447, + "grad_norm": 1.5300422676134766, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 55447 + }, + { + "epoch": 0.55448, + "grad_norm": 1.416782122353656, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 55448 + }, + { + "epoch": 0.55449, + "grad_norm": 1.7843005549483537, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 55449 + }, + { + "epoch": 0.5545, + "grad_norm": 1.222547576608887, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55450 + }, + { + "epoch": 0.55451, + "grad_norm": 1.95838396331377, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 55451 + }, + { + "epoch": 0.55452, + "grad_norm": 1.4005780940933756, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 55452 + }, + { + "epoch": 0.55453, + "grad_norm": 1.7473848431658083, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 55453 + }, + { + "epoch": 0.55454, + "grad_norm": 1.363204767067982, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 55454 + }, + { + "epoch": 0.55455, + "grad_norm": 1.5193828479573328, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 55455 + }, + { + "epoch": 0.55456, + "grad_norm": 1.44839690095479, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 55456 + }, + { + "epoch": 0.55457, + "grad_norm": 1.6994181490829305, + "learning_rate": 0.003, + "loss": 4.04, + "step": 55457 + }, + { + "epoch": 0.55458, + "grad_norm": 1.6225236346850718, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 55458 + }, + { + "epoch": 0.55459, + "grad_norm": 1.9963794470150305, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 55459 + }, + { + "epoch": 0.5546, + "grad_norm": 1.2219987626365454, + "learning_rate": 0.003, + "loss": 4.047, + "step": 55460 + }, + { + "epoch": 0.55461, + "grad_norm": 1.708783307732266, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 55461 + }, + { + "epoch": 0.55462, + "grad_norm": 1.4389597549937467, + "learning_rate": 0.003, + "loss": 4.053, + "step": 55462 + }, + { + "epoch": 0.55463, + "grad_norm": 1.7183217104876747, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 55463 + }, + { + "epoch": 0.55464, + "grad_norm": 1.3410462084481418, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 55464 + }, + { + "epoch": 0.55465, + "grad_norm": 1.6340147180472104, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 55465 + }, + { + "epoch": 0.55466, + "grad_norm": 1.3933540454789826, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 55466 + }, + { + "epoch": 0.55467, + "grad_norm": 1.6078764215777925, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 55467 + }, + { + "epoch": 0.55468, + "grad_norm": 1.4855531842982372, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 55468 + }, + { + "epoch": 0.55469, + "grad_norm": 1.4898950571321887, + "learning_rate": 0.003, + "loss": 4.046, + "step": 55469 + }, + { + "epoch": 0.5547, + "grad_norm": 1.5267248600390013, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 55470 + }, + { + "epoch": 0.55471, + "grad_norm": 1.6950845859377075, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 55471 + }, + { + "epoch": 0.55472, + "grad_norm": 1.961143300201935, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 55472 + }, + { + "epoch": 0.55473, + "grad_norm": 1.943605657484478, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 55473 + }, + { + "epoch": 0.55474, + "grad_norm": 1.5146818356814393, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 55474 + }, + { + "epoch": 0.55475, + "grad_norm": 1.5370131859084015, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 55475 + }, + { + "epoch": 0.55476, + "grad_norm": 1.7705189606893115, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 55476 + }, + { + "epoch": 0.55477, + "grad_norm": 1.7640106668833513, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 55477 + }, + { + "epoch": 0.55478, + "grad_norm": 1.2427260801418736, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 55478 + }, + { + "epoch": 0.55479, + "grad_norm": 1.8106602016068574, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 55479 + }, + { + "epoch": 0.5548, + "grad_norm": 1.4097494578586456, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 55480 + }, + { + "epoch": 0.55481, + "grad_norm": 1.4693057624335053, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 55481 + }, + { + "epoch": 0.55482, + "grad_norm": 1.5657673108584353, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 55482 + }, + { + "epoch": 0.55483, + "grad_norm": 1.5026444129710408, + "learning_rate": 0.003, + "loss": 4.021, + "step": 55483 + }, + { + "epoch": 0.55484, + "grad_norm": 1.614513859627237, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 55484 + }, + { + "epoch": 0.55485, + "grad_norm": 1.5041966169929568, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 55485 + }, + { + "epoch": 0.55486, + "grad_norm": 2.097729518900467, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 55486 + }, + { + "epoch": 0.55487, + "grad_norm": 1.3197770450784254, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 55487 + }, + { + "epoch": 0.55488, + "grad_norm": 1.8046864708153196, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 55488 + }, + { + "epoch": 0.55489, + "grad_norm": 1.3938286390025156, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 55489 + }, + { + "epoch": 0.5549, + "grad_norm": 1.7492812102746476, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 55490 + }, + { + "epoch": 0.55491, + "grad_norm": 1.5152835252653643, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 55491 + }, + { + "epoch": 0.55492, + "grad_norm": 1.4041605136376383, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 55492 + }, + { + "epoch": 0.55493, + "grad_norm": 1.3470715990433277, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 55493 + }, + { + "epoch": 0.55494, + "grad_norm": 1.8688012852148264, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 55494 + }, + { + "epoch": 0.55495, + "grad_norm": 1.5609240746669255, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 55495 + }, + { + "epoch": 0.55496, + "grad_norm": 2.005408887185184, + "learning_rate": 0.003, + "loss": 4.048, + "step": 55496 + }, + { + "epoch": 0.55497, + "grad_norm": 1.4293895027236738, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 55497 + }, + { + "epoch": 0.55498, + "grad_norm": 1.6026169496443907, + "learning_rate": 0.003, + "loss": 4.024, + "step": 55498 + }, + { + "epoch": 0.55499, + "grad_norm": 1.8919839992442027, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 55499 + }, + { + "epoch": 0.555, + "grad_norm": 1.371906376527261, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 55500 + }, + { + "epoch": 0.55501, + "grad_norm": 1.747046001695213, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 55501 + }, + { + "epoch": 0.55502, + "grad_norm": 1.569665237515993, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 55502 + }, + { + "epoch": 0.55503, + "grad_norm": 1.6180863393188099, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 55503 + }, + { + "epoch": 0.55504, + "grad_norm": 1.5880940975864486, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 55504 + }, + { + "epoch": 0.55505, + "grad_norm": 1.5798007496286428, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 55505 + }, + { + "epoch": 0.55506, + "grad_norm": 1.6640918270990055, + "learning_rate": 0.003, + "loss": 4.033, + "step": 55506 + }, + { + "epoch": 0.55507, + "grad_norm": 1.4746889157082448, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 55507 + }, + { + "epoch": 0.55508, + "grad_norm": 1.6655912376873714, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 55508 + }, + { + "epoch": 0.55509, + "grad_norm": 1.2721161907309622, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 55509 + }, + { + "epoch": 0.5551, + "grad_norm": 1.7915278564289705, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 55510 + }, + { + "epoch": 0.55511, + "grad_norm": 1.458799113566644, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 55511 + }, + { + "epoch": 0.55512, + "grad_norm": 1.7238404117353543, + "learning_rate": 0.003, + "loss": 4.017, + "step": 55512 + }, + { + "epoch": 0.55513, + "grad_norm": 1.5040367788633853, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 55513 + }, + { + "epoch": 0.55514, + "grad_norm": 1.6561950234009135, + "learning_rate": 0.003, + "loss": 4.046, + "step": 55514 + }, + { + "epoch": 0.55515, + "grad_norm": 1.66807079462914, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 55515 + }, + { + "epoch": 0.55516, + "grad_norm": 1.3515702026372616, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 55516 + }, + { + "epoch": 0.55517, + "grad_norm": 1.825931502770136, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 55517 + }, + { + "epoch": 0.55518, + "grad_norm": 1.422681756098133, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 55518 + }, + { + "epoch": 0.55519, + "grad_norm": 1.4138842721331222, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 55519 + }, + { + "epoch": 0.5552, + "grad_norm": 1.6724941635878847, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 55520 + }, + { + "epoch": 0.55521, + "grad_norm": 1.4651347380431614, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 55521 + }, + { + "epoch": 0.55522, + "grad_norm": 1.9178435220772068, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 55522 + }, + { + "epoch": 0.55523, + "grad_norm": 1.414040077853743, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 55523 + }, + { + "epoch": 0.55524, + "grad_norm": 1.5286598049068851, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 55524 + }, + { + "epoch": 0.55525, + "grad_norm": 1.5034410556689688, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 55525 + }, + { + "epoch": 0.55526, + "grad_norm": 1.5229676147751552, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 55526 + }, + { + "epoch": 0.55527, + "grad_norm": 1.534521188790247, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 55527 + }, + { + "epoch": 0.55528, + "grad_norm": 1.3235844153192782, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 55528 + }, + { + "epoch": 0.55529, + "grad_norm": 1.530767519271969, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 55529 + }, + { + "epoch": 0.5553, + "grad_norm": 1.3990696496536972, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 55530 + }, + { + "epoch": 0.55531, + "grad_norm": 2.040667671075586, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 55531 + }, + { + "epoch": 0.55532, + "grad_norm": 1.2997889880222053, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 55532 + }, + { + "epoch": 0.55533, + "grad_norm": 1.5636776396106715, + "learning_rate": 0.003, + "loss": 4.018, + "step": 55533 + }, + { + "epoch": 0.55534, + "grad_norm": 1.7213576597420321, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 55534 + }, + { + "epoch": 0.55535, + "grad_norm": 1.6765948995558502, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 55535 + }, + { + "epoch": 0.55536, + "grad_norm": 1.692328890667857, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 55536 + }, + { + "epoch": 0.55537, + "grad_norm": 1.6543248769506096, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 55537 + }, + { + "epoch": 0.55538, + "grad_norm": 1.8005301079761566, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 55538 + }, + { + "epoch": 0.55539, + "grad_norm": 1.4897743552078222, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 55539 + }, + { + "epoch": 0.5554, + "grad_norm": 1.6365319567729544, + "learning_rate": 0.003, + "loss": 4.023, + "step": 55540 + }, + { + "epoch": 0.55541, + "grad_norm": 1.486638724515377, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 55541 + }, + { + "epoch": 0.55542, + "grad_norm": 1.6289027624241215, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 55542 + }, + { + "epoch": 0.55543, + "grad_norm": 1.476317110232534, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 55543 + }, + { + "epoch": 0.55544, + "grad_norm": 1.4504618049065643, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 55544 + }, + { + "epoch": 0.55545, + "grad_norm": 1.427610994640403, + "learning_rate": 0.003, + "loss": 4.034, + "step": 55545 + }, + { + "epoch": 0.55546, + "grad_norm": 1.5138744601293856, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 55546 + }, + { + "epoch": 0.55547, + "grad_norm": 1.6295225694272575, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 55547 + }, + { + "epoch": 0.55548, + "grad_norm": 1.5015275390983616, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55548 + }, + { + "epoch": 0.55549, + "grad_norm": 1.9308777785712794, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 55549 + }, + { + "epoch": 0.5555, + "grad_norm": 1.4450023409923702, + "learning_rate": 0.003, + "loss": 4.032, + "step": 55550 + }, + { + "epoch": 0.55551, + "grad_norm": 1.4571225626057827, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 55551 + }, + { + "epoch": 0.55552, + "grad_norm": 1.9641115026772407, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 55552 + }, + { + "epoch": 0.55553, + "grad_norm": 1.2184557956465751, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 55553 + }, + { + "epoch": 0.55554, + "grad_norm": 1.757068343037432, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 55554 + }, + { + "epoch": 0.55555, + "grad_norm": 1.7924018096981253, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 55555 + }, + { + "epoch": 0.55556, + "grad_norm": 1.6639551647038937, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 55556 + }, + { + "epoch": 0.55557, + "grad_norm": 1.5404980357084264, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 55557 + }, + { + "epoch": 0.55558, + "grad_norm": 1.682166282300696, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 55558 + }, + { + "epoch": 0.55559, + "grad_norm": 1.6168756264370503, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 55559 + }, + { + "epoch": 0.5556, + "grad_norm": 1.7217175144175585, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 55560 + }, + { + "epoch": 0.55561, + "grad_norm": 1.4038631725957138, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 55561 + }, + { + "epoch": 0.55562, + "grad_norm": 1.363849931801398, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 55562 + }, + { + "epoch": 0.55563, + "grad_norm": 1.5320251768911695, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 55563 + }, + { + "epoch": 0.55564, + "grad_norm": 1.5792993782867606, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55564 + }, + { + "epoch": 0.55565, + "grad_norm": 1.4989797128259064, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 55565 + }, + { + "epoch": 0.55566, + "grad_norm": 2.1358485975238763, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 55566 + }, + { + "epoch": 0.55567, + "grad_norm": 1.6689440013530932, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 55567 + }, + { + "epoch": 0.55568, + "grad_norm": 1.7253725884063464, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 55568 + }, + { + "epoch": 0.55569, + "grad_norm": 1.3888837264537166, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 55569 + }, + { + "epoch": 0.5557, + "grad_norm": 1.5707862864821514, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 55570 + }, + { + "epoch": 0.55571, + "grad_norm": 1.6241311992525458, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 55571 + }, + { + "epoch": 0.55572, + "grad_norm": 1.5798074508066215, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 55572 + }, + { + "epoch": 0.55573, + "grad_norm": 1.6003632357905813, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 55573 + }, + { + "epoch": 0.55574, + "grad_norm": 1.537164270750475, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 55574 + }, + { + "epoch": 0.55575, + "grad_norm": 1.7006314976080272, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 55575 + }, + { + "epoch": 0.55576, + "grad_norm": 1.3676910420461983, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 55576 + }, + { + "epoch": 0.55577, + "grad_norm": 1.5713219254734678, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 55577 + }, + { + "epoch": 0.55578, + "grad_norm": 1.3248724539982366, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 55578 + }, + { + "epoch": 0.55579, + "grad_norm": 1.5864444194905385, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 55579 + }, + { + "epoch": 0.5558, + "grad_norm": 1.5561388708995627, + "learning_rate": 0.003, + "loss": 4.043, + "step": 55580 + }, + { + "epoch": 0.55581, + "grad_norm": 1.7369700869650793, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 55581 + }, + { + "epoch": 0.55582, + "grad_norm": 1.1025114825668008, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 55582 + }, + { + "epoch": 0.55583, + "grad_norm": 1.8190281476905699, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 55583 + }, + { + "epoch": 0.55584, + "grad_norm": 1.3303726114091612, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 55584 + }, + { + "epoch": 0.55585, + "grad_norm": 1.496904435660712, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 55585 + }, + { + "epoch": 0.55586, + "grad_norm": 1.475617076815022, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 55586 + }, + { + "epoch": 0.55587, + "grad_norm": 1.6916523704843685, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 55587 + }, + { + "epoch": 0.55588, + "grad_norm": 1.7422672214381005, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 55588 + }, + { + "epoch": 0.55589, + "grad_norm": 1.6286474911491968, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 55589 + }, + { + "epoch": 0.5559, + "grad_norm": 1.7742007003064864, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 55590 + }, + { + "epoch": 0.55591, + "grad_norm": 1.9445890710730223, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 55591 + }, + { + "epoch": 0.55592, + "grad_norm": 1.568911065104612, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 55592 + }, + { + "epoch": 0.55593, + "grad_norm": 1.5769370937528453, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 55593 + }, + { + "epoch": 0.55594, + "grad_norm": 1.633619358349054, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55594 + }, + { + "epoch": 0.55595, + "grad_norm": 1.4464602916808635, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 55595 + }, + { + "epoch": 0.55596, + "grad_norm": 1.6490684082431428, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 55596 + }, + { + "epoch": 0.55597, + "grad_norm": 1.4747061350295378, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 55597 + }, + { + "epoch": 0.55598, + "grad_norm": 1.9416715503128206, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 55598 + }, + { + "epoch": 0.55599, + "grad_norm": 1.8061259048359986, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 55599 + }, + { + "epoch": 0.556, + "grad_norm": 1.4753495077554724, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 55600 + }, + { + "epoch": 0.55601, + "grad_norm": 1.4633074946908855, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 55601 + }, + { + "epoch": 0.55602, + "grad_norm": 1.5731899962921614, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 55602 + }, + { + "epoch": 0.55603, + "grad_norm": 1.5800723966132688, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 55603 + }, + { + "epoch": 0.55604, + "grad_norm": 1.5426558796365053, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 55604 + }, + { + "epoch": 0.55605, + "grad_norm": 1.331453906887947, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 55605 + }, + { + "epoch": 0.55606, + "grad_norm": 1.4006198016134068, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 55606 + }, + { + "epoch": 0.55607, + "grad_norm": 1.8225505815059793, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 55607 + }, + { + "epoch": 0.55608, + "grad_norm": 1.3234742160548958, + "learning_rate": 0.003, + "loss": 4.022, + "step": 55608 + }, + { + "epoch": 0.55609, + "grad_norm": 2.00783620926203, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 55609 + }, + { + "epoch": 0.5561, + "grad_norm": 1.6928527017805575, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 55610 + }, + { + "epoch": 0.55611, + "grad_norm": 1.7284556470234829, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 55611 + }, + { + "epoch": 0.55612, + "grad_norm": 1.5704560458411103, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 55612 + }, + { + "epoch": 0.55613, + "grad_norm": 1.9022958479140544, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 55613 + }, + { + "epoch": 0.55614, + "grad_norm": 1.3158689517027313, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 55614 + }, + { + "epoch": 0.55615, + "grad_norm": 1.6645019690623493, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 55615 + }, + { + "epoch": 0.55616, + "grad_norm": 1.5056238993788154, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 55616 + }, + { + "epoch": 0.55617, + "grad_norm": 1.717389160387043, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 55617 + }, + { + "epoch": 0.55618, + "grad_norm": 1.3441607151916146, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 55618 + }, + { + "epoch": 0.55619, + "grad_norm": 1.4755789131304675, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 55619 + }, + { + "epoch": 0.5562, + "grad_norm": 1.5283057010154133, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 55620 + }, + { + "epoch": 0.55621, + "grad_norm": 1.727853511378316, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 55621 + }, + { + "epoch": 0.55622, + "grad_norm": 1.486230680431549, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 55622 + }, + { + "epoch": 0.55623, + "grad_norm": 1.6175535459338148, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 55623 + }, + { + "epoch": 0.55624, + "grad_norm": 1.6117862837909538, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 55624 + }, + { + "epoch": 0.55625, + "grad_norm": 1.7542820749863188, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 55625 + }, + { + "epoch": 0.55626, + "grad_norm": 1.7524318489361717, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 55626 + }, + { + "epoch": 0.55627, + "grad_norm": 1.5309536636861638, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55627 + }, + { + "epoch": 0.55628, + "grad_norm": 1.7656984158960742, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 55628 + }, + { + "epoch": 0.55629, + "grad_norm": 1.691859468243143, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 55629 + }, + { + "epoch": 0.5563, + "grad_norm": 1.5441083745060904, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 55630 + }, + { + "epoch": 0.55631, + "grad_norm": 1.5122555447143748, + "learning_rate": 0.003, + "loss": 4.018, + "step": 55631 + }, + { + "epoch": 0.55632, + "grad_norm": 1.7862977916154332, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 55632 + }, + { + "epoch": 0.55633, + "grad_norm": 1.663185264518408, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 55633 + }, + { + "epoch": 0.55634, + "grad_norm": 1.4200270742427745, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 55634 + }, + { + "epoch": 0.55635, + "grad_norm": 1.6432906130610003, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 55635 + }, + { + "epoch": 0.55636, + "grad_norm": 1.4573195725396038, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 55636 + }, + { + "epoch": 0.55637, + "grad_norm": 1.621989534232004, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 55637 + }, + { + "epoch": 0.55638, + "grad_norm": 1.5944780215659664, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 55638 + }, + { + "epoch": 0.55639, + "grad_norm": 1.4432607623034404, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 55639 + }, + { + "epoch": 0.5564, + "grad_norm": 1.5422142044823632, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 55640 + }, + { + "epoch": 0.55641, + "grad_norm": 1.5499456387275228, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 55641 + }, + { + "epoch": 0.55642, + "grad_norm": 1.4221370154373125, + "learning_rate": 0.003, + "loss": 4.006, + "step": 55642 + }, + { + "epoch": 0.55643, + "grad_norm": 1.9602521745449548, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 55643 + }, + { + "epoch": 0.55644, + "grad_norm": 1.1540970422119003, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 55644 + }, + { + "epoch": 0.55645, + "grad_norm": 2.181535364352991, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 55645 + }, + { + "epoch": 0.55646, + "grad_norm": 1.4076965846025629, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 55646 + }, + { + "epoch": 0.55647, + "grad_norm": 1.58424509800184, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 55647 + }, + { + "epoch": 0.55648, + "grad_norm": 1.7636926716734274, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 55648 + }, + { + "epoch": 0.55649, + "grad_norm": 1.6258698237744342, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 55649 + }, + { + "epoch": 0.5565, + "grad_norm": 1.3244050521457769, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 55650 + }, + { + "epoch": 0.55651, + "grad_norm": 1.8391178355609294, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 55651 + }, + { + "epoch": 0.55652, + "grad_norm": 1.4573288004562908, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 55652 + }, + { + "epoch": 0.55653, + "grad_norm": 1.709956701108939, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 55653 + }, + { + "epoch": 0.55654, + "grad_norm": 1.3741976618191325, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 55654 + }, + { + "epoch": 0.55655, + "grad_norm": 1.6860700129531287, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 55655 + }, + { + "epoch": 0.55656, + "grad_norm": 1.6893664372926993, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 55656 + }, + { + "epoch": 0.55657, + "grad_norm": 1.673910744816844, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 55657 + }, + { + "epoch": 0.55658, + "grad_norm": 1.4763717266909433, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55658 + }, + { + "epoch": 0.55659, + "grad_norm": 1.3698857658401926, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 55659 + }, + { + "epoch": 0.5566, + "grad_norm": 1.5661447642863642, + "learning_rate": 0.003, + "loss": 4.033, + "step": 55660 + }, + { + "epoch": 0.55661, + "grad_norm": 1.2614406957369162, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55661 + }, + { + "epoch": 0.55662, + "grad_norm": 1.8853857900358344, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 55662 + }, + { + "epoch": 0.55663, + "grad_norm": 1.7562816019932008, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 55663 + }, + { + "epoch": 0.55664, + "grad_norm": 1.9762184342774476, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 55664 + }, + { + "epoch": 0.55665, + "grad_norm": 1.2799552391120221, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 55665 + }, + { + "epoch": 0.55666, + "grad_norm": 1.6632648704946844, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 55666 + }, + { + "epoch": 0.55667, + "grad_norm": 1.676029693692864, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 55667 + }, + { + "epoch": 0.55668, + "grad_norm": 1.6709621966197532, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 55668 + }, + { + "epoch": 0.55669, + "grad_norm": 1.7082153121131776, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 55669 + }, + { + "epoch": 0.5567, + "grad_norm": 1.546829469897668, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 55670 + }, + { + "epoch": 0.55671, + "grad_norm": 1.566320814692298, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 55671 + }, + { + "epoch": 0.55672, + "grad_norm": 1.5639616765093942, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 55672 + }, + { + "epoch": 0.55673, + "grad_norm": 1.6249599086915012, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 55673 + }, + { + "epoch": 0.55674, + "grad_norm": 1.352625103721266, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 55674 + }, + { + "epoch": 0.55675, + "grad_norm": 1.6122700581485228, + "learning_rate": 0.003, + "loss": 4.028, + "step": 55675 + }, + { + "epoch": 0.55676, + "grad_norm": 1.6959030720762256, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 55676 + }, + { + "epoch": 0.55677, + "grad_norm": 1.4545658242154182, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 55677 + }, + { + "epoch": 0.55678, + "grad_norm": 1.8750001484000298, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 55678 + }, + { + "epoch": 0.55679, + "grad_norm": 1.7371603838161238, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 55679 + }, + { + "epoch": 0.5568, + "grad_norm": 1.484652923689462, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 55680 + }, + { + "epoch": 0.55681, + "grad_norm": 1.656072135111523, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 55681 + }, + { + "epoch": 0.55682, + "grad_norm": 1.313239869824955, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55682 + }, + { + "epoch": 0.55683, + "grad_norm": 1.7486361137146758, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 55683 + }, + { + "epoch": 0.55684, + "grad_norm": 1.3464772567759327, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 55684 + }, + { + "epoch": 0.55685, + "grad_norm": 1.980589997908282, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 55685 + }, + { + "epoch": 0.55686, + "grad_norm": 1.3307102020448143, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 55686 + }, + { + "epoch": 0.55687, + "grad_norm": 1.984514594971292, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 55687 + }, + { + "epoch": 0.55688, + "grad_norm": 1.5553878133378283, + "learning_rate": 0.003, + "loss": 4.039, + "step": 55688 + }, + { + "epoch": 0.55689, + "grad_norm": 1.7939415353843562, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 55689 + }, + { + "epoch": 0.5569, + "grad_norm": 1.2926313763219839, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 55690 + }, + { + "epoch": 0.55691, + "grad_norm": 1.718500315209488, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 55691 + }, + { + "epoch": 0.55692, + "grad_norm": 1.8131644786963848, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 55692 + }, + { + "epoch": 0.55693, + "grad_norm": 1.5709162748760572, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 55693 + }, + { + "epoch": 0.55694, + "grad_norm": 1.5576194644364194, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 55694 + }, + { + "epoch": 0.55695, + "grad_norm": 1.4207514973156687, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 55695 + }, + { + "epoch": 0.55696, + "grad_norm": 1.2290126241692645, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 55696 + }, + { + "epoch": 0.55697, + "grad_norm": 1.85571096004682, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55697 + }, + { + "epoch": 0.55698, + "grad_norm": 1.2983950297218574, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 55698 + }, + { + "epoch": 0.55699, + "grad_norm": 1.7073604334136911, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 55699 + }, + { + "epoch": 0.557, + "grad_norm": 1.3168370622342382, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 55700 + }, + { + "epoch": 0.55701, + "grad_norm": 1.81068166790851, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 55701 + }, + { + "epoch": 0.55702, + "grad_norm": 1.3491869053936356, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 55702 + }, + { + "epoch": 0.55703, + "grad_norm": 1.655525019382765, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 55703 + }, + { + "epoch": 0.55704, + "grad_norm": 1.836936477111841, + "learning_rate": 0.003, + "loss": 4.037, + "step": 55704 + }, + { + "epoch": 0.55705, + "grad_norm": 1.6539302342662185, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 55705 + }, + { + "epoch": 0.55706, + "grad_norm": 1.6914357764344614, + "learning_rate": 0.003, + "loss": 4.032, + "step": 55706 + }, + { + "epoch": 0.55707, + "grad_norm": 1.5180468289398235, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 55707 + }, + { + "epoch": 0.55708, + "grad_norm": 1.4039557454660498, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 55708 + }, + { + "epoch": 0.55709, + "grad_norm": 1.7499697846062165, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 55709 + }, + { + "epoch": 0.5571, + "grad_norm": 1.5617467161615566, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 55710 + }, + { + "epoch": 0.55711, + "grad_norm": 1.7679926008243076, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 55711 + }, + { + "epoch": 0.55712, + "grad_norm": 1.5339970619925172, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 55712 + }, + { + "epoch": 0.55713, + "grad_norm": 1.843470852969606, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 55713 + }, + { + "epoch": 0.55714, + "grad_norm": 1.6602216024251115, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 55714 + }, + { + "epoch": 0.55715, + "grad_norm": 1.3184451431149418, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 55715 + }, + { + "epoch": 0.55716, + "grad_norm": 1.7572201099494662, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 55716 + }, + { + "epoch": 0.55717, + "grad_norm": 1.466891979477047, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 55717 + }, + { + "epoch": 0.55718, + "grad_norm": 1.6473682311654487, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 55718 + }, + { + "epoch": 0.55719, + "grad_norm": 1.644115621732127, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 55719 + }, + { + "epoch": 0.5572, + "grad_norm": 1.7284686072674784, + "learning_rate": 0.003, + "loss": 4.057, + "step": 55720 + }, + { + "epoch": 0.55721, + "grad_norm": 1.2888546870628728, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 55721 + }, + { + "epoch": 0.55722, + "grad_norm": 1.6152446893329278, + "learning_rate": 0.003, + "loss": 4.019, + "step": 55722 + }, + { + "epoch": 0.55723, + "grad_norm": 1.6006657409889975, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 55723 + }, + { + "epoch": 0.55724, + "grad_norm": 1.6784952535179665, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55724 + }, + { + "epoch": 0.55725, + "grad_norm": 1.6129834740194298, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 55725 + }, + { + "epoch": 0.55726, + "grad_norm": 1.5343472848866166, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 55726 + }, + { + "epoch": 0.55727, + "grad_norm": 1.7884062438212665, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 55727 + }, + { + "epoch": 0.55728, + "grad_norm": 1.2756375713827015, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 55728 + }, + { + "epoch": 0.55729, + "grad_norm": 1.838543933453625, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 55729 + }, + { + "epoch": 0.5573, + "grad_norm": 1.2683150673487897, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 55730 + }, + { + "epoch": 0.55731, + "grad_norm": 1.9274434735949812, + "learning_rate": 0.003, + "loss": 4.049, + "step": 55731 + }, + { + "epoch": 0.55732, + "grad_norm": 1.3522039767686473, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 55732 + }, + { + "epoch": 0.55733, + "grad_norm": 1.6696048321684644, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 55733 + }, + { + "epoch": 0.55734, + "grad_norm": 1.6414196285444567, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 55734 + }, + { + "epoch": 0.55735, + "grad_norm": 1.6761554178416533, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 55735 + }, + { + "epoch": 0.55736, + "grad_norm": 1.8029194942277698, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 55736 + }, + { + "epoch": 0.55737, + "grad_norm": 1.3933021786155197, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 55737 + }, + { + "epoch": 0.55738, + "grad_norm": 1.7440373672262164, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 55738 + }, + { + "epoch": 0.55739, + "grad_norm": 1.4395696632019992, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 55739 + }, + { + "epoch": 0.5574, + "grad_norm": 1.570333061196115, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 55740 + }, + { + "epoch": 0.55741, + "grad_norm": 1.5067297978210688, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 55741 + }, + { + "epoch": 0.55742, + "grad_norm": 1.5902323035999923, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 55742 + }, + { + "epoch": 0.55743, + "grad_norm": 1.456444617891264, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 55743 + }, + { + "epoch": 0.55744, + "grad_norm": 1.552944075869473, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 55744 + }, + { + "epoch": 0.55745, + "grad_norm": 1.4544100304932663, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 55745 + }, + { + "epoch": 0.55746, + "grad_norm": 1.5204053098245567, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 55746 + }, + { + "epoch": 0.55747, + "grad_norm": 1.4183261893348889, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 55747 + }, + { + "epoch": 0.55748, + "grad_norm": 1.6147563042585709, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 55748 + }, + { + "epoch": 0.55749, + "grad_norm": 1.9074076235737691, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 55749 + }, + { + "epoch": 0.5575, + "grad_norm": 1.434161456571585, + "learning_rate": 0.003, + "loss": 4.016, + "step": 55750 + }, + { + "epoch": 0.55751, + "grad_norm": 1.4234021747468462, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 55751 + }, + { + "epoch": 0.55752, + "grad_norm": 1.7252902181487286, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 55752 + }, + { + "epoch": 0.55753, + "grad_norm": 1.3957758452353308, + "learning_rate": 0.003, + "loss": 4.021, + "step": 55753 + }, + { + "epoch": 0.55754, + "grad_norm": 1.4070949542588111, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 55754 + }, + { + "epoch": 0.55755, + "grad_norm": 1.5591431309197459, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 55755 + }, + { + "epoch": 0.55756, + "grad_norm": 1.6156080344431922, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 55756 + }, + { + "epoch": 0.55757, + "grad_norm": 1.757268565510265, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 55757 + }, + { + "epoch": 0.55758, + "grad_norm": 1.7252663924834222, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 55758 + }, + { + "epoch": 0.55759, + "grad_norm": 1.8585411975449946, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 55759 + }, + { + "epoch": 0.5576, + "grad_norm": 1.464653965602625, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 55760 + }, + { + "epoch": 0.55761, + "grad_norm": 1.4608235878704587, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 55761 + }, + { + "epoch": 0.55762, + "grad_norm": 1.710565345670575, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 55762 + }, + { + "epoch": 0.55763, + "grad_norm": 1.4042968162787988, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 55763 + }, + { + "epoch": 0.55764, + "grad_norm": 1.6462947692931853, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 55764 + }, + { + "epoch": 0.55765, + "grad_norm": 1.4657511997233978, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 55765 + }, + { + "epoch": 0.55766, + "grad_norm": 1.8275357317953298, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 55766 + }, + { + "epoch": 0.55767, + "grad_norm": 1.7885156977315537, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 55767 + }, + { + "epoch": 0.55768, + "grad_norm": 1.690130631262929, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 55768 + }, + { + "epoch": 0.55769, + "grad_norm": 1.5358741380869212, + "learning_rate": 0.003, + "loss": 4.019, + "step": 55769 + }, + { + "epoch": 0.5577, + "grad_norm": 1.8023664117713623, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 55770 + }, + { + "epoch": 0.55771, + "grad_norm": 1.4780269429654318, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 55771 + }, + { + "epoch": 0.55772, + "grad_norm": 1.650733131201208, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 55772 + }, + { + "epoch": 0.55773, + "grad_norm": 1.391859378633336, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 55773 + }, + { + "epoch": 0.55774, + "grad_norm": 1.5578631616220253, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 55774 + }, + { + "epoch": 0.55775, + "grad_norm": 1.4781076903737524, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 55775 + }, + { + "epoch": 0.55776, + "grad_norm": 1.7549069971519033, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 55776 + }, + { + "epoch": 0.55777, + "grad_norm": 1.3064332551834796, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 55777 + }, + { + "epoch": 0.55778, + "grad_norm": 1.5472251807694923, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 55778 + }, + { + "epoch": 0.55779, + "grad_norm": 1.7357932581522681, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 55779 + }, + { + "epoch": 0.5578, + "grad_norm": 1.3838171944935167, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55780 + }, + { + "epoch": 0.55781, + "grad_norm": 1.637743649976164, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 55781 + }, + { + "epoch": 0.55782, + "grad_norm": 1.2566198952105614, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55782 + }, + { + "epoch": 0.55783, + "grad_norm": 1.9413823447953158, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 55783 + }, + { + "epoch": 0.55784, + "grad_norm": 1.7136473285043887, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 55784 + }, + { + "epoch": 0.55785, + "grad_norm": 1.7045480560162345, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55785 + }, + { + "epoch": 0.55786, + "grad_norm": 1.5418680260989654, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 55786 + }, + { + "epoch": 0.55787, + "grad_norm": 1.6374422704934382, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 55787 + }, + { + "epoch": 0.55788, + "grad_norm": 1.4912614109061304, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 55788 + }, + { + "epoch": 0.55789, + "grad_norm": 1.5888763807013047, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 55789 + }, + { + "epoch": 0.5579, + "grad_norm": 1.5368002736815578, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 55790 + }, + { + "epoch": 0.55791, + "grad_norm": 1.436351010156365, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 55791 + }, + { + "epoch": 0.55792, + "grad_norm": 1.6455744205491873, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 55792 + }, + { + "epoch": 0.55793, + "grad_norm": 1.3707657658232366, + "learning_rate": 0.003, + "loss": 4.027, + "step": 55793 + }, + { + "epoch": 0.55794, + "grad_norm": 1.8421168279660332, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55794 + }, + { + "epoch": 0.55795, + "grad_norm": 1.2972182026796386, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 55795 + }, + { + "epoch": 0.55796, + "grad_norm": 1.6015841642764606, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 55796 + }, + { + "epoch": 0.55797, + "grad_norm": 1.5487668554378466, + "learning_rate": 0.003, + "loss": 4.022, + "step": 55797 + }, + { + "epoch": 0.55798, + "grad_norm": 1.6275642741286926, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 55798 + }, + { + "epoch": 0.55799, + "grad_norm": 1.6286969690248956, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 55799 + }, + { + "epoch": 0.558, + "grad_norm": 1.4899967540978085, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 55800 + }, + { + "epoch": 0.55801, + "grad_norm": 1.5362938362084282, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 55801 + }, + { + "epoch": 0.55802, + "grad_norm": 1.5549342455607782, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 55802 + }, + { + "epoch": 0.55803, + "grad_norm": 1.693286323562827, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 55803 + }, + { + "epoch": 0.55804, + "grad_norm": 1.401202317399495, + "learning_rate": 0.003, + "loss": 3.989, + "step": 55804 + }, + { + "epoch": 0.55805, + "grad_norm": 1.700297842845936, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 55805 + }, + { + "epoch": 0.55806, + "grad_norm": 1.431751466181302, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 55806 + }, + { + "epoch": 0.55807, + "grad_norm": 1.545976404605659, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 55807 + }, + { + "epoch": 0.55808, + "grad_norm": 1.589231198043553, + "learning_rate": 0.003, + "loss": 4.018, + "step": 55808 + }, + { + "epoch": 0.55809, + "grad_norm": 1.5650019090456118, + "learning_rate": 0.003, + "loss": 4.027, + "step": 55809 + }, + { + "epoch": 0.5581, + "grad_norm": 1.5831668384297228, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 55810 + }, + { + "epoch": 0.55811, + "grad_norm": 1.4328012072206209, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 55811 + }, + { + "epoch": 0.55812, + "grad_norm": 1.793074778703684, + "learning_rate": 0.003, + "loss": 4.029, + "step": 55812 + }, + { + "epoch": 0.55813, + "grad_norm": 1.3634902465736323, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 55813 + }, + { + "epoch": 0.55814, + "grad_norm": 1.762463331381086, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 55814 + }, + { + "epoch": 0.55815, + "grad_norm": 1.6340408302866105, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 55815 + }, + { + "epoch": 0.55816, + "grad_norm": 2.0476112678654723, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 55816 + }, + { + "epoch": 0.55817, + "grad_norm": 1.6525350581758738, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 55817 + }, + { + "epoch": 0.55818, + "grad_norm": 1.4606675437819054, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 55818 + }, + { + "epoch": 0.55819, + "grad_norm": 1.988073011198186, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 55819 + }, + { + "epoch": 0.5582, + "grad_norm": 1.5965269477742663, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 55820 + }, + { + "epoch": 0.55821, + "grad_norm": 1.3853472798825863, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 55821 + }, + { + "epoch": 0.55822, + "grad_norm": 1.480216794667142, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 55822 + }, + { + "epoch": 0.55823, + "grad_norm": 1.5412186033008024, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 55823 + }, + { + "epoch": 0.55824, + "grad_norm": 1.6855441913006792, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 55824 + }, + { + "epoch": 0.55825, + "grad_norm": 1.6328384770688145, + "learning_rate": 0.003, + "loss": 4.042, + "step": 55825 + }, + { + "epoch": 0.55826, + "grad_norm": 1.6567821260721611, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 55826 + }, + { + "epoch": 0.55827, + "grad_norm": 1.646246109268027, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 55827 + }, + { + "epoch": 0.55828, + "grad_norm": 1.4422300944995214, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 55828 + }, + { + "epoch": 0.55829, + "grad_norm": 1.8231377569128537, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 55829 + }, + { + "epoch": 0.5583, + "grad_norm": 1.622849932134586, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 55830 + }, + { + "epoch": 0.55831, + "grad_norm": 1.5821745491991743, + "learning_rate": 0.003, + "loss": 4.057, + "step": 55831 + }, + { + "epoch": 0.55832, + "grad_norm": 1.586101354674864, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 55832 + }, + { + "epoch": 0.55833, + "grad_norm": 1.2800614674687847, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 55833 + }, + { + "epoch": 0.55834, + "grad_norm": 1.9516251341972024, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 55834 + }, + { + "epoch": 0.55835, + "grad_norm": 1.1849022410130032, + "learning_rate": 0.003, + "loss": 4.035, + "step": 55835 + }, + { + "epoch": 0.55836, + "grad_norm": 1.8437535311674675, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 55836 + }, + { + "epoch": 0.55837, + "grad_norm": 1.5190875745430596, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 55837 + }, + { + "epoch": 0.55838, + "grad_norm": 1.355214024025585, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 55838 + }, + { + "epoch": 0.55839, + "grad_norm": 1.8142520691961153, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 55839 + }, + { + "epoch": 0.5584, + "grad_norm": 1.5475353269797087, + "learning_rate": 0.003, + "loss": 4.034, + "step": 55840 + }, + { + "epoch": 0.55841, + "grad_norm": 1.7189480453468013, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 55841 + }, + { + "epoch": 0.55842, + "grad_norm": 1.386480618387093, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 55842 + }, + { + "epoch": 0.55843, + "grad_norm": 1.701812580802494, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 55843 + }, + { + "epoch": 0.55844, + "grad_norm": 1.3265385754023866, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 55844 + }, + { + "epoch": 0.55845, + "grad_norm": 1.6790380505378895, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 55845 + }, + { + "epoch": 0.55846, + "grad_norm": 1.5894647196011562, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 55846 + }, + { + "epoch": 0.55847, + "grad_norm": 1.9538462277403799, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 55847 + }, + { + "epoch": 0.55848, + "grad_norm": 1.3400134395440813, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 55848 + }, + { + "epoch": 0.55849, + "grad_norm": 1.8014505998617978, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 55849 + }, + { + "epoch": 0.5585, + "grad_norm": 1.9346155365426783, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 55850 + }, + { + "epoch": 0.55851, + "grad_norm": 1.3462483998370716, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 55851 + }, + { + "epoch": 0.55852, + "grad_norm": 1.5484388404856682, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 55852 + }, + { + "epoch": 0.55853, + "grad_norm": 1.6413430393291077, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 55853 + }, + { + "epoch": 0.55854, + "grad_norm": 1.355961306401036, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 55854 + }, + { + "epoch": 0.55855, + "grad_norm": 1.771909273522342, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 55855 + }, + { + "epoch": 0.55856, + "grad_norm": 1.3074929146516479, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 55856 + }, + { + "epoch": 0.55857, + "grad_norm": 1.9297701929548945, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 55857 + }, + { + "epoch": 0.55858, + "grad_norm": 1.3125809457712532, + "learning_rate": 0.003, + "loss": 4.069, + "step": 55858 + }, + { + "epoch": 0.55859, + "grad_norm": 2.0049451966192287, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 55859 + }, + { + "epoch": 0.5586, + "grad_norm": 1.5367025409591992, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 55860 + }, + { + "epoch": 0.55861, + "grad_norm": 1.6672869194409847, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 55861 + }, + { + "epoch": 0.55862, + "grad_norm": 1.4985071819707458, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 55862 + }, + { + "epoch": 0.55863, + "grad_norm": 1.6592993643683651, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 55863 + }, + { + "epoch": 0.55864, + "grad_norm": 1.3568580773645704, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 55864 + }, + { + "epoch": 0.55865, + "grad_norm": 1.672729752343554, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 55865 + }, + { + "epoch": 0.55866, + "grad_norm": 1.4690772799760323, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 55866 + }, + { + "epoch": 0.55867, + "grad_norm": 1.545609891015471, + "learning_rate": 0.003, + "loss": 4.045, + "step": 55867 + }, + { + "epoch": 0.55868, + "grad_norm": 1.5722000959555948, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 55868 + }, + { + "epoch": 0.55869, + "grad_norm": 1.674149816451418, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 55869 + }, + { + "epoch": 0.5587, + "grad_norm": 1.4877181638952184, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 55870 + }, + { + "epoch": 0.55871, + "grad_norm": 1.581177260632893, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 55871 + }, + { + "epoch": 0.55872, + "grad_norm": 1.8106778103292698, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 55872 + }, + { + "epoch": 0.55873, + "grad_norm": 1.4686128585772915, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 55873 + }, + { + "epoch": 0.55874, + "grad_norm": 1.5225264700834098, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 55874 + }, + { + "epoch": 0.55875, + "grad_norm": 1.6055612566237965, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 55875 + }, + { + "epoch": 0.55876, + "grad_norm": 1.7566320043927082, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 55876 + }, + { + "epoch": 0.55877, + "grad_norm": 1.379272885544812, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 55877 + }, + { + "epoch": 0.55878, + "grad_norm": 1.7892136623136894, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 55878 + }, + { + "epoch": 0.55879, + "grad_norm": 1.3465735953727025, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 55879 + }, + { + "epoch": 0.5588, + "grad_norm": 1.7055102819772034, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 55880 + }, + { + "epoch": 0.55881, + "grad_norm": 1.4980230049845265, + "learning_rate": 0.003, + "loss": 4.019, + "step": 55881 + }, + { + "epoch": 0.55882, + "grad_norm": 1.4493317312105876, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 55882 + }, + { + "epoch": 0.55883, + "grad_norm": 1.6545117396071412, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 55883 + }, + { + "epoch": 0.55884, + "grad_norm": 1.6464487082691852, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 55884 + }, + { + "epoch": 0.55885, + "grad_norm": 1.8006916676822613, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 55885 + }, + { + "epoch": 0.55886, + "grad_norm": 1.577761918627917, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 55886 + }, + { + "epoch": 0.55887, + "grad_norm": 1.43951270249822, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 55887 + }, + { + "epoch": 0.55888, + "grad_norm": 1.5281438160329097, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 55888 + }, + { + "epoch": 0.55889, + "grad_norm": 1.5379416425206098, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 55889 + }, + { + "epoch": 0.5589, + "grad_norm": 1.6793860081633114, + "learning_rate": 0.003, + "loss": 4.038, + "step": 55890 + }, + { + "epoch": 0.55891, + "grad_norm": 1.7524631150997474, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 55891 + }, + { + "epoch": 0.55892, + "grad_norm": 1.741250826280782, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 55892 + }, + { + "epoch": 0.55893, + "grad_norm": 1.5887885215347448, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 55893 + }, + { + "epoch": 0.55894, + "grad_norm": 1.9311971195152269, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 55894 + }, + { + "epoch": 0.55895, + "grad_norm": 1.1061352127875108, + "learning_rate": 0.003, + "loss": 4.01, + "step": 55895 + }, + { + "epoch": 0.55896, + "grad_norm": 1.7345641370142868, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 55896 + }, + { + "epoch": 0.55897, + "grad_norm": 1.5704198695041243, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 55897 + }, + { + "epoch": 0.55898, + "grad_norm": 1.677330029475357, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 55898 + }, + { + "epoch": 0.55899, + "grad_norm": 1.5232041885473997, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 55899 + }, + { + "epoch": 0.559, + "grad_norm": 1.5865743359159807, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 55900 + }, + { + "epoch": 0.55901, + "grad_norm": 1.3861185819645852, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 55901 + }, + { + "epoch": 0.55902, + "grad_norm": 1.5332482740894695, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 55902 + }, + { + "epoch": 0.55903, + "grad_norm": 1.5281060656832535, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 55903 + }, + { + "epoch": 0.55904, + "grad_norm": 1.4908843177635283, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 55904 + }, + { + "epoch": 0.55905, + "grad_norm": 1.5390624619617073, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 55905 + }, + { + "epoch": 0.55906, + "grad_norm": 1.8280822436485686, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 55906 + }, + { + "epoch": 0.55907, + "grad_norm": 1.5966494497916122, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 55907 + }, + { + "epoch": 0.55908, + "grad_norm": 1.6385292154694997, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 55908 + }, + { + "epoch": 0.55909, + "grad_norm": 1.6509075383487275, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 55909 + }, + { + "epoch": 0.5591, + "grad_norm": 1.6117516872624242, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 55910 + }, + { + "epoch": 0.55911, + "grad_norm": 1.4638328223114296, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 55911 + }, + { + "epoch": 0.55912, + "grad_norm": 1.5409094707442355, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 55912 + }, + { + "epoch": 0.55913, + "grad_norm": 1.5448595442157615, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 55913 + }, + { + "epoch": 0.55914, + "grad_norm": 1.3587146248824964, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 55914 + }, + { + "epoch": 0.55915, + "grad_norm": 1.5722373436259083, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 55915 + }, + { + "epoch": 0.55916, + "grad_norm": 1.5757825091367517, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 55916 + }, + { + "epoch": 0.55917, + "grad_norm": 1.944222864382521, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 55917 + }, + { + "epoch": 0.55918, + "grad_norm": 1.5068992778507309, + "learning_rate": 0.003, + "loss": 4.044, + "step": 55918 + }, + { + "epoch": 0.55919, + "grad_norm": 1.997348385461957, + "learning_rate": 0.003, + "loss": 4.018, + "step": 55919 + }, + { + "epoch": 0.5592, + "grad_norm": 1.494140820519232, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 55920 + }, + { + "epoch": 0.55921, + "grad_norm": 1.723882275960255, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 55921 + }, + { + "epoch": 0.55922, + "grad_norm": 1.7171209332906099, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 55922 + }, + { + "epoch": 0.55923, + "grad_norm": 1.5954447013560171, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 55923 + }, + { + "epoch": 0.55924, + "grad_norm": 1.8017387099942352, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 55924 + }, + { + "epoch": 0.55925, + "grad_norm": 1.4157002782586232, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 55925 + }, + { + "epoch": 0.55926, + "grad_norm": 1.4312156171773842, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 55926 + }, + { + "epoch": 0.55927, + "grad_norm": 1.7194520722691986, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 55927 + }, + { + "epoch": 0.55928, + "grad_norm": 1.6443818753923947, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 55928 + }, + { + "epoch": 0.55929, + "grad_norm": 1.850789570240993, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55929 + }, + { + "epoch": 0.5593, + "grad_norm": 1.2976331325912949, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 55930 + }, + { + "epoch": 0.55931, + "grad_norm": 1.8869226334416966, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 55931 + }, + { + "epoch": 0.55932, + "grad_norm": 1.3581031565894452, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 55932 + }, + { + "epoch": 0.55933, + "grad_norm": 1.921451562180607, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 55933 + }, + { + "epoch": 0.55934, + "grad_norm": 1.628032630065072, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 55934 + }, + { + "epoch": 0.55935, + "grad_norm": 1.5416210076381318, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 55935 + }, + { + "epoch": 0.55936, + "grad_norm": 1.2444966222606568, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 55936 + }, + { + "epoch": 0.55937, + "grad_norm": 1.7754129803094458, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 55937 + }, + { + "epoch": 0.55938, + "grad_norm": 1.6071974294783806, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 55938 + }, + { + "epoch": 0.55939, + "grad_norm": 1.4018257891713621, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 55939 + }, + { + "epoch": 0.5594, + "grad_norm": 1.465955632940637, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 55940 + }, + { + "epoch": 0.55941, + "grad_norm": 1.5714816341150377, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 55941 + }, + { + "epoch": 0.55942, + "grad_norm": 1.641023261222137, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 55942 + }, + { + "epoch": 0.55943, + "grad_norm": 1.7078550220490003, + "learning_rate": 0.003, + "loss": 4.011, + "step": 55943 + }, + { + "epoch": 0.55944, + "grad_norm": 1.5428316276236815, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 55944 + }, + { + "epoch": 0.55945, + "grad_norm": 1.780092562876783, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 55945 + }, + { + "epoch": 0.55946, + "grad_norm": 1.5282556325398868, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 55946 + }, + { + "epoch": 0.55947, + "grad_norm": 1.605937936228511, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 55947 + }, + { + "epoch": 0.55948, + "grad_norm": 1.5129625988847815, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 55948 + }, + { + "epoch": 0.55949, + "grad_norm": 1.4885710043881142, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 55949 + }, + { + "epoch": 0.5595, + "grad_norm": 1.299854405747269, + "learning_rate": 0.003, + "loss": 4.02, + "step": 55950 + }, + { + "epoch": 0.55951, + "grad_norm": 1.6254487200093999, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 55951 + }, + { + "epoch": 0.55952, + "grad_norm": 1.561179241492159, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 55952 + }, + { + "epoch": 0.55953, + "grad_norm": 1.4869299895676673, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 55953 + }, + { + "epoch": 0.55954, + "grad_norm": 1.6105501657179364, + "learning_rate": 0.003, + "loss": 4.037, + "step": 55954 + }, + { + "epoch": 0.55955, + "grad_norm": 1.5780922507628408, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 55955 + }, + { + "epoch": 0.55956, + "grad_norm": 1.4883304965366002, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 55956 + }, + { + "epoch": 0.55957, + "grad_norm": 1.8234682888216551, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 55957 + }, + { + "epoch": 0.55958, + "grad_norm": 1.5803650873004804, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 55958 + }, + { + "epoch": 0.55959, + "grad_norm": 1.8494602543259415, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 55959 + }, + { + "epoch": 0.5596, + "grad_norm": 1.5619504915630091, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 55960 + }, + { + "epoch": 0.55961, + "grad_norm": 1.6160640397498442, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 55961 + }, + { + "epoch": 0.55962, + "grad_norm": 1.767320831917898, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 55962 + }, + { + "epoch": 0.55963, + "grad_norm": 1.1983156546112903, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 55963 + }, + { + "epoch": 0.55964, + "grad_norm": 1.9345938869120214, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 55964 + }, + { + "epoch": 0.55965, + "grad_norm": 1.3556449805967916, + "learning_rate": 0.003, + "loss": 4.035, + "step": 55965 + }, + { + "epoch": 0.55966, + "grad_norm": 1.7610321410015468, + "learning_rate": 0.003, + "loss": 4.03, + "step": 55966 + }, + { + "epoch": 0.55967, + "grad_norm": 1.6352851855601809, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 55967 + }, + { + "epoch": 0.55968, + "grad_norm": 1.4398502413230814, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 55968 + }, + { + "epoch": 0.55969, + "grad_norm": 1.7703148000101725, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 55969 + }, + { + "epoch": 0.5597, + "grad_norm": 1.3707819598580748, + "learning_rate": 0.003, + "loss": 4.026, + "step": 55970 + }, + { + "epoch": 0.55971, + "grad_norm": 1.7893949417229764, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 55971 + }, + { + "epoch": 0.55972, + "grad_norm": 1.900065327333013, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 55972 + }, + { + "epoch": 0.55973, + "grad_norm": 1.4857559424929832, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 55973 + }, + { + "epoch": 0.55974, + "grad_norm": 1.482871646358572, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 55974 + }, + { + "epoch": 0.55975, + "grad_norm": 1.5559234158952742, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 55975 + }, + { + "epoch": 0.55976, + "grad_norm": 1.6300308825504288, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 55976 + }, + { + "epoch": 0.55977, + "grad_norm": 1.3510754315243139, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 55977 + }, + { + "epoch": 0.55978, + "grad_norm": 1.5234535503384004, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 55978 + }, + { + "epoch": 0.55979, + "grad_norm": 1.7106292466409574, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 55979 + }, + { + "epoch": 0.5598, + "grad_norm": 1.602964927052498, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 55980 + }, + { + "epoch": 0.55981, + "grad_norm": 1.639903553995427, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 55981 + }, + { + "epoch": 0.55982, + "grad_norm": 1.527655111991578, + "learning_rate": 0.003, + "loss": 4.04, + "step": 55982 + }, + { + "epoch": 0.55983, + "grad_norm": 1.5490959657794423, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 55983 + }, + { + "epoch": 0.55984, + "grad_norm": 1.5683979407670894, + "learning_rate": 0.003, + "loss": 4.044, + "step": 55984 + }, + { + "epoch": 0.55985, + "grad_norm": 1.4003032948203853, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 55985 + }, + { + "epoch": 0.55986, + "grad_norm": 1.7164198479867292, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 55986 + }, + { + "epoch": 0.55987, + "grad_norm": 1.3819864596618772, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 55987 + }, + { + "epoch": 0.55988, + "grad_norm": 1.6006968724712316, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 55988 + }, + { + "epoch": 0.55989, + "grad_norm": 1.4750697049279626, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 55989 + }, + { + "epoch": 0.5599, + "grad_norm": 1.7809590653190097, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 55990 + }, + { + "epoch": 0.55991, + "grad_norm": 1.808858213289703, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 55991 + }, + { + "epoch": 0.55992, + "grad_norm": 1.7591963194631328, + "learning_rate": 0.003, + "loss": 4.041, + "step": 55992 + }, + { + "epoch": 0.55993, + "grad_norm": 1.5372770604189754, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 55993 + }, + { + "epoch": 0.55994, + "grad_norm": 1.4747952135555453, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 55994 + }, + { + "epoch": 0.55995, + "grad_norm": 1.6535576478701903, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 55995 + }, + { + "epoch": 0.55996, + "grad_norm": 1.4772427740740217, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 55996 + }, + { + "epoch": 0.55997, + "grad_norm": 1.3877844566824271, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 55997 + }, + { + "epoch": 0.55998, + "grad_norm": 1.7447328192363, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 55998 + }, + { + "epoch": 0.55999, + "grad_norm": 1.5549098804272221, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 55999 + }, + { + "epoch": 0.56, + "grad_norm": 1.5790650580564567, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56000 + }, + { + "epoch": 0.56001, + "grad_norm": 1.5083241354845889, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 56001 + }, + { + "epoch": 0.56002, + "grad_norm": 1.5448070495592872, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 56002 + }, + { + "epoch": 0.56003, + "grad_norm": 1.6898169924964215, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 56003 + }, + { + "epoch": 0.56004, + "grad_norm": 1.616091913883818, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 56004 + }, + { + "epoch": 0.56005, + "grad_norm": 1.5107547574609528, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 56005 + }, + { + "epoch": 0.56006, + "grad_norm": 1.5733853244230247, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 56006 + }, + { + "epoch": 0.56007, + "grad_norm": 1.5075603311201813, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 56007 + }, + { + "epoch": 0.56008, + "grad_norm": 1.5258215474910104, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 56008 + }, + { + "epoch": 0.56009, + "grad_norm": 1.5088784434388585, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 56009 + }, + { + "epoch": 0.5601, + "grad_norm": 1.7556649913165567, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 56010 + }, + { + "epoch": 0.56011, + "grad_norm": 1.503221804387979, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 56011 + }, + { + "epoch": 0.56012, + "grad_norm": 1.9438745423599695, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 56012 + }, + { + "epoch": 0.56013, + "grad_norm": 1.4716591647455546, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 56013 + }, + { + "epoch": 0.56014, + "grad_norm": 1.7551191172139333, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 56014 + }, + { + "epoch": 0.56015, + "grad_norm": 1.7206693969324065, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 56015 + }, + { + "epoch": 0.56016, + "grad_norm": 1.6193362426097913, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 56016 + }, + { + "epoch": 0.56017, + "grad_norm": 1.466989448748819, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 56017 + }, + { + "epoch": 0.56018, + "grad_norm": 1.4761874536609865, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 56018 + }, + { + "epoch": 0.56019, + "grad_norm": 1.5802173657566396, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 56019 + }, + { + "epoch": 0.5602, + "grad_norm": 1.699858071105744, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 56020 + }, + { + "epoch": 0.56021, + "grad_norm": 1.5011789974453782, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 56021 + }, + { + "epoch": 0.56022, + "grad_norm": 1.6389485147172123, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 56022 + }, + { + "epoch": 0.56023, + "grad_norm": 1.4129991132985404, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56023 + }, + { + "epoch": 0.56024, + "grad_norm": 1.802625969055377, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 56024 + }, + { + "epoch": 0.56025, + "grad_norm": 1.4225183984416672, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 56025 + }, + { + "epoch": 0.56026, + "grad_norm": 1.9249323151722972, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 56026 + }, + { + "epoch": 0.56027, + "grad_norm": 1.5681923210161934, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 56027 + }, + { + "epoch": 0.56028, + "grad_norm": 1.7382957251040287, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 56028 + }, + { + "epoch": 0.56029, + "grad_norm": 1.601571764230415, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 56029 + }, + { + "epoch": 0.5603, + "grad_norm": 1.4802629431548235, + "learning_rate": 0.003, + "loss": 4.027, + "step": 56030 + }, + { + "epoch": 0.56031, + "grad_norm": 1.6280614874851496, + "learning_rate": 0.003, + "loss": 4.035, + "step": 56031 + }, + { + "epoch": 0.56032, + "grad_norm": 1.3986456924837583, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 56032 + }, + { + "epoch": 0.56033, + "grad_norm": 1.4217182397480486, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 56033 + }, + { + "epoch": 0.56034, + "grad_norm": 1.62425455534718, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 56034 + }, + { + "epoch": 0.56035, + "grad_norm": 1.4621119422310316, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 56035 + }, + { + "epoch": 0.56036, + "grad_norm": 1.6916995913907222, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 56036 + }, + { + "epoch": 0.56037, + "grad_norm": 1.6957609282580668, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 56037 + }, + { + "epoch": 0.56038, + "grad_norm": 1.8500811283267704, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 56038 + }, + { + "epoch": 0.56039, + "grad_norm": 1.2981567920201507, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 56039 + }, + { + "epoch": 0.5604, + "grad_norm": 1.7380607436737683, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 56040 + }, + { + "epoch": 0.56041, + "grad_norm": 1.644797700253207, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56041 + }, + { + "epoch": 0.56042, + "grad_norm": 1.7045218481893651, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 56042 + }, + { + "epoch": 0.56043, + "grad_norm": 1.137347389333279, + "learning_rate": 0.003, + "loss": 4.059, + "step": 56043 + }, + { + "epoch": 0.56044, + "grad_norm": 1.9529885213397848, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 56044 + }, + { + "epoch": 0.56045, + "grad_norm": 1.5303845706967643, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 56045 + }, + { + "epoch": 0.56046, + "grad_norm": 1.5145940394886415, + "learning_rate": 0.003, + "loss": 4.032, + "step": 56046 + }, + { + "epoch": 0.56047, + "grad_norm": 1.578642287344969, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56047 + }, + { + "epoch": 0.56048, + "grad_norm": 1.888559886412762, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 56048 + }, + { + "epoch": 0.56049, + "grad_norm": 1.4428788605946663, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56049 + }, + { + "epoch": 0.5605, + "grad_norm": 1.6443381863759599, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 56050 + }, + { + "epoch": 0.56051, + "grad_norm": 1.8743556284819733, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 56051 + }, + { + "epoch": 0.56052, + "grad_norm": 1.4049899452501557, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 56052 + }, + { + "epoch": 0.56053, + "grad_norm": 1.5917790375181569, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 56053 + }, + { + "epoch": 0.56054, + "grad_norm": 1.4591709747910033, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 56054 + }, + { + "epoch": 0.56055, + "grad_norm": 1.7382237203399895, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 56055 + }, + { + "epoch": 0.56056, + "grad_norm": 1.2012955359451367, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 56056 + }, + { + "epoch": 0.56057, + "grad_norm": 1.727388220425352, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56057 + }, + { + "epoch": 0.56058, + "grad_norm": 1.7521075918680638, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 56058 + }, + { + "epoch": 0.56059, + "grad_norm": 1.881748283131364, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 56059 + }, + { + "epoch": 0.5606, + "grad_norm": 1.702614745874381, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 56060 + }, + { + "epoch": 0.56061, + "grad_norm": 1.2129901627472683, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 56061 + }, + { + "epoch": 0.56062, + "grad_norm": 1.7126469696149833, + "learning_rate": 0.003, + "loss": 4.028, + "step": 56062 + }, + { + "epoch": 0.56063, + "grad_norm": 1.2838090972902123, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 56063 + }, + { + "epoch": 0.56064, + "grad_norm": 1.9985138990558093, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 56064 + }, + { + "epoch": 0.56065, + "grad_norm": 1.4537625690704283, + "learning_rate": 0.003, + "loss": 4.051, + "step": 56065 + }, + { + "epoch": 0.56066, + "grad_norm": 2.256802941894824, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 56066 + }, + { + "epoch": 0.56067, + "grad_norm": 1.2541752024447101, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 56067 + }, + { + "epoch": 0.56068, + "grad_norm": 1.5981532562265748, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 56068 + }, + { + "epoch": 0.56069, + "grad_norm": 1.8015139202517112, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 56069 + }, + { + "epoch": 0.5607, + "grad_norm": 1.7850541328688698, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 56070 + }, + { + "epoch": 0.56071, + "grad_norm": 1.6370562978842493, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 56071 + }, + { + "epoch": 0.56072, + "grad_norm": 1.80190546044449, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 56072 + }, + { + "epoch": 0.56073, + "grad_norm": 1.4685920694829997, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 56073 + }, + { + "epoch": 0.56074, + "grad_norm": 1.5946398509879167, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 56074 + }, + { + "epoch": 0.56075, + "grad_norm": 1.516023526656239, + "learning_rate": 0.003, + "loss": 4.053, + "step": 56075 + }, + { + "epoch": 0.56076, + "grad_norm": 1.624508034267559, + "learning_rate": 0.003, + "loss": 4.005, + "step": 56076 + }, + { + "epoch": 0.56077, + "grad_norm": 1.6711325101413876, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 56077 + }, + { + "epoch": 0.56078, + "grad_norm": 1.6423833174827487, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 56078 + }, + { + "epoch": 0.56079, + "grad_norm": 1.6875234139578883, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 56079 + }, + { + "epoch": 0.5608, + "grad_norm": 1.4985255275285818, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 56080 + }, + { + "epoch": 0.56081, + "grad_norm": 1.537588744100242, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 56081 + }, + { + "epoch": 0.56082, + "grad_norm": 1.381186741233017, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 56082 + }, + { + "epoch": 0.56083, + "grad_norm": 1.7969732374382208, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56083 + }, + { + "epoch": 0.56084, + "grad_norm": 1.6692997525486353, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 56084 + }, + { + "epoch": 0.56085, + "grad_norm": 1.726093478032872, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 56085 + }, + { + "epoch": 0.56086, + "grad_norm": 1.662034619207024, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 56086 + }, + { + "epoch": 0.56087, + "grad_norm": 1.543899314120322, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 56087 + }, + { + "epoch": 0.56088, + "grad_norm": 1.6994626955356107, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 56088 + }, + { + "epoch": 0.56089, + "grad_norm": 1.467088950893821, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 56089 + }, + { + "epoch": 0.5609, + "grad_norm": 1.5332100321226303, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 56090 + }, + { + "epoch": 0.56091, + "grad_norm": 1.4552757520988757, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 56091 + }, + { + "epoch": 0.56092, + "grad_norm": 1.2095302915137616, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 56092 + }, + { + "epoch": 0.56093, + "grad_norm": 1.7123964403684193, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56093 + }, + { + "epoch": 0.56094, + "grad_norm": 1.3658101446646147, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 56094 + }, + { + "epoch": 0.56095, + "grad_norm": 1.5700278387749902, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 56095 + }, + { + "epoch": 0.56096, + "grad_norm": 1.9194434968380873, + "learning_rate": 0.003, + "loss": 4.054, + "step": 56096 + }, + { + "epoch": 0.56097, + "grad_norm": 1.5306114721231965, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 56097 + }, + { + "epoch": 0.56098, + "grad_norm": 2.2037130328493704, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 56098 + }, + { + "epoch": 0.56099, + "grad_norm": 1.4759692985448571, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 56099 + }, + { + "epoch": 0.561, + "grad_norm": 1.6162287081888176, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 56100 + }, + { + "epoch": 0.56101, + "grad_norm": 1.8169168519598204, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 56101 + }, + { + "epoch": 0.56102, + "grad_norm": 1.42853950686868, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 56102 + }, + { + "epoch": 0.56103, + "grad_norm": 1.5624154017403478, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 56103 + }, + { + "epoch": 0.56104, + "grad_norm": 1.980450263530776, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 56104 + }, + { + "epoch": 0.56105, + "grad_norm": 1.0418610066848515, + "learning_rate": 0.003, + "loss": 4.057, + "step": 56105 + }, + { + "epoch": 0.56106, + "grad_norm": 1.9458556408265266, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56106 + }, + { + "epoch": 0.56107, + "grad_norm": 1.3449153242430663, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 56107 + }, + { + "epoch": 0.56108, + "grad_norm": 1.7210818203807408, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 56108 + }, + { + "epoch": 0.56109, + "grad_norm": 1.4522519899596729, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56109 + }, + { + "epoch": 0.5611, + "grad_norm": 1.6126240615958847, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 56110 + }, + { + "epoch": 0.56111, + "grad_norm": 1.7425552749342137, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 56111 + }, + { + "epoch": 0.56112, + "grad_norm": 1.4346623278243582, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 56112 + }, + { + "epoch": 0.56113, + "grad_norm": 1.6824177561143494, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 56113 + }, + { + "epoch": 0.56114, + "grad_norm": 1.7583254549369571, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 56114 + }, + { + "epoch": 0.56115, + "grad_norm": 1.63485272067493, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 56115 + }, + { + "epoch": 0.56116, + "grad_norm": 1.6187325959932148, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 56116 + }, + { + "epoch": 0.56117, + "grad_norm": 1.6983415477880706, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56117 + }, + { + "epoch": 0.56118, + "grad_norm": 1.3755152622197995, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 56118 + }, + { + "epoch": 0.56119, + "grad_norm": 1.7103084783900127, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 56119 + }, + { + "epoch": 0.5612, + "grad_norm": 1.17970993062412, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 56120 + }, + { + "epoch": 0.56121, + "grad_norm": 1.6644043127063324, + "learning_rate": 0.003, + "loss": 4.037, + "step": 56121 + }, + { + "epoch": 0.56122, + "grad_norm": 1.4789278018835086, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 56122 + }, + { + "epoch": 0.56123, + "grad_norm": 1.5756621645842839, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 56123 + }, + { + "epoch": 0.56124, + "grad_norm": 1.7002096827135, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56124 + }, + { + "epoch": 0.56125, + "grad_norm": 1.6298471829556975, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 56125 + }, + { + "epoch": 0.56126, + "grad_norm": 1.7035091761186534, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 56126 + }, + { + "epoch": 0.56127, + "grad_norm": 1.7182194247303286, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 56127 + }, + { + "epoch": 0.56128, + "grad_norm": 1.6734847814369633, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 56128 + }, + { + "epoch": 0.56129, + "grad_norm": 1.4774083456656593, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 56129 + }, + { + "epoch": 0.5613, + "grad_norm": 1.4622105895206838, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 56130 + }, + { + "epoch": 0.56131, + "grad_norm": 1.713398544649141, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 56131 + }, + { + "epoch": 0.56132, + "grad_norm": 1.3142003605154329, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 56132 + }, + { + "epoch": 0.56133, + "grad_norm": 1.7247555164095367, + "learning_rate": 0.003, + "loss": 4.043, + "step": 56133 + }, + { + "epoch": 0.56134, + "grad_norm": 1.2507757415339398, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 56134 + }, + { + "epoch": 0.56135, + "grad_norm": 1.7309616095389948, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 56135 + }, + { + "epoch": 0.56136, + "grad_norm": 1.1919977244374829, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 56136 + }, + { + "epoch": 0.56137, + "grad_norm": 1.7593471518454906, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 56137 + }, + { + "epoch": 0.56138, + "grad_norm": 1.5429943144181766, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 56138 + }, + { + "epoch": 0.56139, + "grad_norm": 1.6209050739454813, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 56139 + }, + { + "epoch": 0.5614, + "grad_norm": 1.6363067231752852, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 56140 + }, + { + "epoch": 0.56141, + "grad_norm": 1.8758876139505738, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 56141 + }, + { + "epoch": 0.56142, + "grad_norm": 1.6293984737112062, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 56142 + }, + { + "epoch": 0.56143, + "grad_norm": 1.643199815929257, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56143 + }, + { + "epoch": 0.56144, + "grad_norm": 1.5583186179841537, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 56144 + }, + { + "epoch": 0.56145, + "grad_norm": 1.3426036284273177, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 56145 + }, + { + "epoch": 0.56146, + "grad_norm": 2.1366073052462973, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 56146 + }, + { + "epoch": 0.56147, + "grad_norm": 1.6518988428694403, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 56147 + }, + { + "epoch": 0.56148, + "grad_norm": 1.6454897166016746, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 56148 + }, + { + "epoch": 0.56149, + "grad_norm": 1.5832666275688585, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 56149 + }, + { + "epoch": 0.5615, + "grad_norm": 1.5068475214057337, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 56150 + }, + { + "epoch": 0.56151, + "grad_norm": 1.3881205409854585, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 56151 + }, + { + "epoch": 0.56152, + "grad_norm": 1.5502619571817955, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 56152 + }, + { + "epoch": 0.56153, + "grad_norm": 1.6181314183745714, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 56153 + }, + { + "epoch": 0.56154, + "grad_norm": 1.456764243735726, + "learning_rate": 0.003, + "loss": 4.022, + "step": 56154 + }, + { + "epoch": 0.56155, + "grad_norm": 1.6176882071236764, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 56155 + }, + { + "epoch": 0.56156, + "grad_norm": 1.4639813464246765, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 56156 + }, + { + "epoch": 0.56157, + "grad_norm": 2.0466667061004813, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56157 + }, + { + "epoch": 0.56158, + "grad_norm": 1.2966320601577102, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 56158 + }, + { + "epoch": 0.56159, + "grad_norm": 1.9841029552737972, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 56159 + }, + { + "epoch": 0.5616, + "grad_norm": 1.6911937323614947, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 56160 + }, + { + "epoch": 0.56161, + "grad_norm": 1.4535749519491752, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 56161 + }, + { + "epoch": 0.56162, + "grad_norm": 1.5387798486369102, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 56162 + }, + { + "epoch": 0.56163, + "grad_norm": 1.7920357092690253, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 56163 + }, + { + "epoch": 0.56164, + "grad_norm": 1.353239620031759, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 56164 + }, + { + "epoch": 0.56165, + "grad_norm": 1.5497677325961439, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 56165 + }, + { + "epoch": 0.56166, + "grad_norm": 1.6893751549879132, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 56166 + }, + { + "epoch": 0.56167, + "grad_norm": 1.197757408046591, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 56167 + }, + { + "epoch": 0.56168, + "grad_norm": 1.6178095089194766, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 56168 + }, + { + "epoch": 0.56169, + "grad_norm": 1.3272477401372622, + "learning_rate": 0.003, + "loss": 4.026, + "step": 56169 + }, + { + "epoch": 0.5617, + "grad_norm": 1.6304273674764989, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 56170 + }, + { + "epoch": 0.56171, + "grad_norm": 1.5371082538524983, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 56171 + }, + { + "epoch": 0.56172, + "grad_norm": 1.8259220194104149, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 56172 + }, + { + "epoch": 0.56173, + "grad_norm": 1.7091988669793499, + "learning_rate": 0.003, + "loss": 4.026, + "step": 56173 + }, + { + "epoch": 0.56174, + "grad_norm": 1.5907715929979687, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 56174 + }, + { + "epoch": 0.56175, + "grad_norm": 1.6827951760520046, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 56175 + }, + { + "epoch": 0.56176, + "grad_norm": 1.4245098369563043, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 56176 + }, + { + "epoch": 0.56177, + "grad_norm": 1.4463674399253008, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 56177 + }, + { + "epoch": 0.56178, + "grad_norm": 1.5658907132000088, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 56178 + }, + { + "epoch": 0.56179, + "grad_norm": 1.630462487052508, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 56179 + }, + { + "epoch": 0.5618, + "grad_norm": 1.5448147147229436, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 56180 + }, + { + "epoch": 0.56181, + "grad_norm": 1.5334058566312943, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 56181 + }, + { + "epoch": 0.56182, + "grad_norm": 1.5373192042385913, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 56182 + }, + { + "epoch": 0.56183, + "grad_norm": 1.5119692807428375, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 56183 + }, + { + "epoch": 0.56184, + "grad_norm": 1.6890607212943842, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 56184 + }, + { + "epoch": 0.56185, + "grad_norm": 1.5013876082576407, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 56185 + }, + { + "epoch": 0.56186, + "grad_norm": 1.508566887978686, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 56186 + }, + { + "epoch": 0.56187, + "grad_norm": 1.56666477667358, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 56187 + }, + { + "epoch": 0.56188, + "grad_norm": 1.5319986057021704, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 56188 + }, + { + "epoch": 0.56189, + "grad_norm": 1.450763626890242, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 56189 + }, + { + "epoch": 0.5619, + "grad_norm": 1.5116531725783235, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 56190 + }, + { + "epoch": 0.56191, + "grad_norm": 1.4135097660188012, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 56191 + }, + { + "epoch": 0.56192, + "grad_norm": 1.563233522466735, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 56192 + }, + { + "epoch": 0.56193, + "grad_norm": 1.6613993523317407, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 56193 + }, + { + "epoch": 0.56194, + "grad_norm": 1.477584822300157, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 56194 + }, + { + "epoch": 0.56195, + "grad_norm": 1.8941127552251875, + "learning_rate": 0.003, + "loss": 4.019, + "step": 56195 + }, + { + "epoch": 0.56196, + "grad_norm": 1.469632030934154, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 56196 + }, + { + "epoch": 0.56197, + "grad_norm": 1.6173465790616313, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 56197 + }, + { + "epoch": 0.56198, + "grad_norm": 1.5822213463207317, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 56198 + }, + { + "epoch": 0.56199, + "grad_norm": 1.7431369738977467, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 56199 + }, + { + "epoch": 0.562, + "grad_norm": 1.5934794755492074, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 56200 + }, + { + "epoch": 0.56201, + "grad_norm": 1.6538406951372253, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 56201 + }, + { + "epoch": 0.56202, + "grad_norm": 2.059180816111038, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 56202 + }, + { + "epoch": 0.56203, + "grad_norm": 1.3429741901320411, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 56203 + }, + { + "epoch": 0.56204, + "grad_norm": 1.70689949392982, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 56204 + }, + { + "epoch": 0.56205, + "grad_norm": 1.3592162414029827, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 56205 + }, + { + "epoch": 0.56206, + "grad_norm": 1.9882653339845702, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 56206 + }, + { + "epoch": 0.56207, + "grad_norm": 1.6081217011223419, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 56207 + }, + { + "epoch": 0.56208, + "grad_norm": 1.570708075866886, + "learning_rate": 0.003, + "loss": 4.043, + "step": 56208 + }, + { + "epoch": 0.56209, + "grad_norm": 1.6867120663566149, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 56209 + }, + { + "epoch": 0.5621, + "grad_norm": 1.6015662156524728, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 56210 + }, + { + "epoch": 0.56211, + "grad_norm": 1.6505061951535218, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 56211 + }, + { + "epoch": 0.56212, + "grad_norm": 1.7859127890294015, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 56212 + }, + { + "epoch": 0.56213, + "grad_norm": 1.4241508466158512, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 56213 + }, + { + "epoch": 0.56214, + "grad_norm": 1.5218393399051118, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56214 + }, + { + "epoch": 0.56215, + "grad_norm": 1.5315380486870573, + "learning_rate": 0.003, + "loss": 4.033, + "step": 56215 + }, + { + "epoch": 0.56216, + "grad_norm": 1.382011122013449, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 56216 + }, + { + "epoch": 0.56217, + "grad_norm": 1.5298891581129732, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 56217 + }, + { + "epoch": 0.56218, + "grad_norm": 1.5124633315004952, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 56218 + }, + { + "epoch": 0.56219, + "grad_norm": 1.5502942219506195, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 56219 + }, + { + "epoch": 0.5622, + "grad_norm": 1.5881517178468367, + "learning_rate": 0.003, + "loss": 4.05, + "step": 56220 + }, + { + "epoch": 0.56221, + "grad_norm": 1.2999764270061778, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 56221 + }, + { + "epoch": 0.56222, + "grad_norm": 1.7552305688278471, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 56222 + }, + { + "epoch": 0.56223, + "grad_norm": 1.1160197401869716, + "learning_rate": 0.003, + "loss": 4.039, + "step": 56223 + }, + { + "epoch": 0.56224, + "grad_norm": 1.829288445204174, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 56224 + }, + { + "epoch": 0.56225, + "grad_norm": 1.3386120325623958, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 56225 + }, + { + "epoch": 0.56226, + "grad_norm": 1.7501651045702673, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 56226 + }, + { + "epoch": 0.56227, + "grad_norm": 1.7146654374990595, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 56227 + }, + { + "epoch": 0.56228, + "grad_norm": 1.4395407734510088, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 56228 + }, + { + "epoch": 0.56229, + "grad_norm": 1.6926955751882802, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 56229 + }, + { + "epoch": 0.5623, + "grad_norm": 1.5398033020986146, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 56230 + }, + { + "epoch": 0.56231, + "grad_norm": 1.9203972549910318, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 56231 + }, + { + "epoch": 0.56232, + "grad_norm": 1.4088291901069805, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 56232 + }, + { + "epoch": 0.56233, + "grad_norm": 1.7773934430119547, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 56233 + }, + { + "epoch": 0.56234, + "grad_norm": 1.333336917007322, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 56234 + }, + { + "epoch": 0.56235, + "grad_norm": 1.76396052555741, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 56235 + }, + { + "epoch": 0.56236, + "grad_norm": 1.4241250652237152, + "learning_rate": 0.003, + "loss": 4.019, + "step": 56236 + }, + { + "epoch": 0.56237, + "grad_norm": 1.9364764880891474, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 56237 + }, + { + "epoch": 0.56238, + "grad_norm": 1.5846403472033193, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 56238 + }, + { + "epoch": 0.56239, + "grad_norm": 1.450145490006639, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 56239 + }, + { + "epoch": 0.5624, + "grad_norm": 1.8065908526679066, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56240 + }, + { + "epoch": 0.56241, + "grad_norm": 1.4883369979685461, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 56241 + }, + { + "epoch": 0.56242, + "grad_norm": 1.8604006698658875, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 56242 + }, + { + "epoch": 0.56243, + "grad_norm": 1.2633910298625062, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 56243 + }, + { + "epoch": 0.56244, + "grad_norm": 1.841926706407234, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 56244 + }, + { + "epoch": 0.56245, + "grad_norm": 1.4304212524309512, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 56245 + }, + { + "epoch": 0.56246, + "grad_norm": 1.669702973555562, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 56246 + }, + { + "epoch": 0.56247, + "grad_norm": 1.7392877479913396, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 56247 + }, + { + "epoch": 0.56248, + "grad_norm": 1.6184338140911712, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 56248 + }, + { + "epoch": 0.56249, + "grad_norm": 1.4875942308590224, + "learning_rate": 0.003, + "loss": 4.014, + "step": 56249 + }, + { + "epoch": 0.5625, + "grad_norm": 1.529834064513803, + "learning_rate": 0.003, + "loss": 4.031, + "step": 56250 + }, + { + "epoch": 0.56251, + "grad_norm": 1.4777944595735093, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 56251 + }, + { + "epoch": 0.56252, + "grad_norm": 1.6115253862828418, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 56252 + }, + { + "epoch": 0.56253, + "grad_norm": 1.471192976184062, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 56253 + }, + { + "epoch": 0.56254, + "grad_norm": 1.4623626934393557, + "learning_rate": 0.003, + "loss": 4.01, + "step": 56254 + }, + { + "epoch": 0.56255, + "grad_norm": 1.845578559591472, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 56255 + }, + { + "epoch": 0.56256, + "grad_norm": 1.419918452978357, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 56256 + }, + { + "epoch": 0.56257, + "grad_norm": 1.58204083227221, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 56257 + }, + { + "epoch": 0.56258, + "grad_norm": 1.2391278323638388, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 56258 + }, + { + "epoch": 0.56259, + "grad_norm": 1.499104862817354, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 56259 + }, + { + "epoch": 0.5626, + "grad_norm": 1.4673580858463946, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 56260 + }, + { + "epoch": 0.56261, + "grad_norm": 1.907614517646585, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 56261 + }, + { + "epoch": 0.56262, + "grad_norm": 1.6632906704855206, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 56262 + }, + { + "epoch": 0.56263, + "grad_norm": 1.3237982763647196, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 56263 + }, + { + "epoch": 0.56264, + "grad_norm": 1.7564039699481864, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56264 + }, + { + "epoch": 0.56265, + "grad_norm": 1.3631619137212467, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 56265 + }, + { + "epoch": 0.56266, + "grad_norm": 1.906087578739117, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 56266 + }, + { + "epoch": 0.56267, + "grad_norm": 1.2620636136265326, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 56267 + }, + { + "epoch": 0.56268, + "grad_norm": 1.5765068440020573, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 56268 + }, + { + "epoch": 0.56269, + "grad_norm": 1.886043824004583, + "learning_rate": 0.003, + "loss": 4.071, + "step": 56269 + }, + { + "epoch": 0.5627, + "grad_norm": 1.4137023800873072, + "learning_rate": 0.003, + "loss": 4.042, + "step": 56270 + }, + { + "epoch": 0.56271, + "grad_norm": 1.554421434199855, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 56271 + }, + { + "epoch": 0.56272, + "grad_norm": 1.6014909993793645, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 56272 + }, + { + "epoch": 0.56273, + "grad_norm": 1.3935363004653611, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 56273 + }, + { + "epoch": 0.56274, + "grad_norm": 1.63592106130947, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 56274 + }, + { + "epoch": 0.56275, + "grad_norm": 1.6512262253755943, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 56275 + }, + { + "epoch": 0.56276, + "grad_norm": 1.5400499246107826, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 56276 + }, + { + "epoch": 0.56277, + "grad_norm": 1.5808175979888477, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 56277 + }, + { + "epoch": 0.56278, + "grad_norm": 1.695115763948575, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 56278 + }, + { + "epoch": 0.56279, + "grad_norm": 1.5034878244830807, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 56279 + }, + { + "epoch": 0.5628, + "grad_norm": 1.5907140690432449, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 56280 + }, + { + "epoch": 0.56281, + "grad_norm": 1.5256193504737596, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 56281 + }, + { + "epoch": 0.56282, + "grad_norm": 1.2285735534251905, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 56282 + }, + { + "epoch": 0.56283, + "grad_norm": 1.6172320242933973, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 56283 + }, + { + "epoch": 0.56284, + "grad_norm": 1.5028077484732136, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 56284 + }, + { + "epoch": 0.56285, + "grad_norm": 2.024291991581657, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 56285 + }, + { + "epoch": 0.56286, + "grad_norm": 2.030828396893329, + "learning_rate": 0.003, + "loss": 4.011, + "step": 56286 + }, + { + "epoch": 0.56287, + "grad_norm": 1.2361380817697838, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 56287 + }, + { + "epoch": 0.56288, + "grad_norm": 1.8959307862034278, + "learning_rate": 0.003, + "loss": 4.07, + "step": 56288 + }, + { + "epoch": 0.56289, + "grad_norm": 1.6212611351927688, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 56289 + }, + { + "epoch": 0.5629, + "grad_norm": 1.6423732357188332, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 56290 + }, + { + "epoch": 0.56291, + "grad_norm": 1.3687497829765907, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 56291 + }, + { + "epoch": 0.56292, + "grad_norm": 1.970419529405879, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 56292 + }, + { + "epoch": 0.56293, + "grad_norm": 1.1877506050565456, + "learning_rate": 0.003, + "loss": 4.043, + "step": 56293 + }, + { + "epoch": 0.56294, + "grad_norm": 1.448028733636759, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 56294 + }, + { + "epoch": 0.56295, + "grad_norm": 1.835895617441201, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 56295 + }, + { + "epoch": 0.56296, + "grad_norm": 1.8621470258772341, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 56296 + }, + { + "epoch": 0.56297, + "grad_norm": 1.577053097069343, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 56297 + }, + { + "epoch": 0.56298, + "grad_norm": 1.642566297198721, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 56298 + }, + { + "epoch": 0.56299, + "grad_norm": 1.564162237453397, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 56299 + }, + { + "epoch": 0.563, + "grad_norm": 1.6493005377114804, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 56300 + }, + { + "epoch": 0.56301, + "grad_norm": 1.497943264155287, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 56301 + }, + { + "epoch": 0.56302, + "grad_norm": 1.5747277288337944, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 56302 + }, + { + "epoch": 0.56303, + "grad_norm": 1.6855801836764854, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 56303 + }, + { + "epoch": 0.56304, + "grad_norm": 1.548316431387251, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 56304 + }, + { + "epoch": 0.56305, + "grad_norm": 1.6539855801132022, + "learning_rate": 0.003, + "loss": 4.02, + "step": 56305 + }, + { + "epoch": 0.56306, + "grad_norm": 1.5747959975359023, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 56306 + }, + { + "epoch": 0.56307, + "grad_norm": 1.7531713589620013, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 56307 + }, + { + "epoch": 0.56308, + "grad_norm": 1.781979072324047, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 56308 + }, + { + "epoch": 0.56309, + "grad_norm": 1.3023705516853714, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 56309 + }, + { + "epoch": 0.5631, + "grad_norm": 1.664633274740528, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 56310 + }, + { + "epoch": 0.56311, + "grad_norm": 1.5630504372559308, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 56311 + }, + { + "epoch": 0.56312, + "grad_norm": 1.4081000408576236, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 56312 + }, + { + "epoch": 0.56313, + "grad_norm": 1.8343291757587938, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 56313 + }, + { + "epoch": 0.56314, + "grad_norm": 1.569331690864705, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 56314 + }, + { + "epoch": 0.56315, + "grad_norm": 1.5220597673632341, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 56315 + }, + { + "epoch": 0.56316, + "grad_norm": 1.3943125234884572, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 56316 + }, + { + "epoch": 0.56317, + "grad_norm": 1.4843019565005016, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 56317 + }, + { + "epoch": 0.56318, + "grad_norm": 1.6505280621678942, + "learning_rate": 0.003, + "loss": 4.027, + "step": 56318 + }, + { + "epoch": 0.56319, + "grad_norm": 1.222830760933209, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 56319 + }, + { + "epoch": 0.5632, + "grad_norm": 1.7962674617586816, + "learning_rate": 0.003, + "loss": 4.021, + "step": 56320 + }, + { + "epoch": 0.56321, + "grad_norm": 1.4575934167659694, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 56321 + }, + { + "epoch": 0.56322, + "grad_norm": 1.6818659875169522, + "learning_rate": 0.003, + "loss": 4.013, + "step": 56322 + }, + { + "epoch": 0.56323, + "grad_norm": 1.4760250487617956, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 56323 + }, + { + "epoch": 0.56324, + "grad_norm": 1.5132219984408086, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 56324 + }, + { + "epoch": 0.56325, + "grad_norm": 1.8231419549803738, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56325 + }, + { + "epoch": 0.56326, + "grad_norm": 1.560913222295933, + "learning_rate": 0.003, + "loss": 4.053, + "step": 56326 + }, + { + "epoch": 0.56327, + "grad_norm": 1.6381521798717682, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 56327 + }, + { + "epoch": 0.56328, + "grad_norm": 1.8004830862744516, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 56328 + }, + { + "epoch": 0.56329, + "grad_norm": 1.4237492643846883, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 56329 + }, + { + "epoch": 0.5633, + "grad_norm": 1.5539076529241693, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 56330 + }, + { + "epoch": 0.56331, + "grad_norm": 1.5034837257877578, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 56331 + }, + { + "epoch": 0.56332, + "grad_norm": 1.7088012377681439, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 56332 + }, + { + "epoch": 0.56333, + "grad_norm": 1.4715006813193299, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 56333 + }, + { + "epoch": 0.56334, + "grad_norm": 1.6805049468776756, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 56334 + }, + { + "epoch": 0.56335, + "grad_norm": 1.8449643286471522, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 56335 + }, + { + "epoch": 0.56336, + "grad_norm": 1.659109895402205, + "learning_rate": 0.003, + "loss": 4.029, + "step": 56336 + }, + { + "epoch": 0.56337, + "grad_norm": 1.45255826219881, + "learning_rate": 0.003, + "loss": 4.006, + "step": 56337 + }, + { + "epoch": 0.56338, + "grad_norm": 1.559314462042449, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 56338 + }, + { + "epoch": 0.56339, + "grad_norm": 1.3514545379461222, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 56339 + }, + { + "epoch": 0.5634, + "grad_norm": 1.6887088451469994, + "learning_rate": 0.003, + "loss": 4.03, + "step": 56340 + }, + { + "epoch": 0.56341, + "grad_norm": 1.7726210970372476, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 56341 + }, + { + "epoch": 0.56342, + "grad_norm": 1.6885398553455138, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 56342 + }, + { + "epoch": 0.56343, + "grad_norm": 1.856189649144593, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 56343 + }, + { + "epoch": 0.56344, + "grad_norm": 1.4505807336478587, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 56344 + }, + { + "epoch": 0.56345, + "grad_norm": 1.503587724642768, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 56345 + }, + { + "epoch": 0.56346, + "grad_norm": 1.8507749413530254, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 56346 + }, + { + "epoch": 0.56347, + "grad_norm": 1.6076277415050768, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 56347 + }, + { + "epoch": 0.56348, + "grad_norm": 1.4699487445256223, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 56348 + }, + { + "epoch": 0.56349, + "grad_norm": 1.8283266956764521, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 56349 + }, + { + "epoch": 0.5635, + "grad_norm": 1.227768944646138, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 56350 + }, + { + "epoch": 0.56351, + "grad_norm": 1.5262031270077374, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 56351 + }, + { + "epoch": 0.56352, + "grad_norm": 1.8306264083950556, + "learning_rate": 0.003, + "loss": 4.037, + "step": 56352 + }, + { + "epoch": 0.56353, + "grad_norm": 1.7476987323316584, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 56353 + }, + { + "epoch": 0.56354, + "grad_norm": 1.4880290203572812, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 56354 + }, + { + "epoch": 0.56355, + "grad_norm": 1.4965011851621584, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 56355 + }, + { + "epoch": 0.56356, + "grad_norm": 1.7541146262079614, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 56356 + }, + { + "epoch": 0.56357, + "grad_norm": 1.5440436007919067, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 56357 + }, + { + "epoch": 0.56358, + "grad_norm": 1.6999531822506813, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 56358 + }, + { + "epoch": 0.56359, + "grad_norm": 1.5844799564711982, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 56359 + }, + { + "epoch": 0.5636, + "grad_norm": 1.5174798382442016, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56360 + }, + { + "epoch": 0.56361, + "grad_norm": 1.4861913987895248, + "learning_rate": 0.003, + "loss": 4.027, + "step": 56361 + }, + { + "epoch": 0.56362, + "grad_norm": 1.9069695358267553, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 56362 + }, + { + "epoch": 0.56363, + "grad_norm": 1.1579226514106584, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 56363 + }, + { + "epoch": 0.56364, + "grad_norm": 1.6518161897830965, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 56364 + }, + { + "epoch": 0.56365, + "grad_norm": 1.2034153005063286, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 56365 + }, + { + "epoch": 0.56366, + "grad_norm": 1.9981111701665717, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 56366 + }, + { + "epoch": 0.56367, + "grad_norm": 1.535852666458246, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 56367 + }, + { + "epoch": 0.56368, + "grad_norm": 1.9246785756495592, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 56368 + }, + { + "epoch": 0.56369, + "grad_norm": 1.6381787500419214, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 56369 + }, + { + "epoch": 0.5637, + "grad_norm": 1.7025339112525633, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 56370 + }, + { + "epoch": 0.56371, + "grad_norm": 1.368011399170478, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 56371 + }, + { + "epoch": 0.56372, + "grad_norm": 1.7033479802806704, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 56372 + }, + { + "epoch": 0.56373, + "grad_norm": 1.5306834179883022, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56373 + }, + { + "epoch": 0.56374, + "grad_norm": 1.9296328760733255, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56374 + }, + { + "epoch": 0.56375, + "grad_norm": 1.3040187277971187, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 56375 + }, + { + "epoch": 0.56376, + "grad_norm": 1.6353955641549462, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 56376 + }, + { + "epoch": 0.56377, + "grad_norm": 1.5443952559203407, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 56377 + }, + { + "epoch": 0.56378, + "grad_norm": 1.5661527090599099, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 56378 + }, + { + "epoch": 0.56379, + "grad_norm": 1.3134972755356504, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56379 + }, + { + "epoch": 0.5638, + "grad_norm": 1.5833424833883194, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 56380 + }, + { + "epoch": 0.56381, + "grad_norm": 1.8842177890714893, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 56381 + }, + { + "epoch": 0.56382, + "grad_norm": 1.6694550868751443, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 56382 + }, + { + "epoch": 0.56383, + "grad_norm": 1.6057864305379168, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 56383 + }, + { + "epoch": 0.56384, + "grad_norm": 1.4928156440162428, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56384 + }, + { + "epoch": 0.56385, + "grad_norm": 1.866904341454785, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 56385 + }, + { + "epoch": 0.56386, + "grad_norm": 1.5358855606445478, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 56386 + }, + { + "epoch": 0.56387, + "grad_norm": 1.4580215448308649, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 56387 + }, + { + "epoch": 0.56388, + "grad_norm": 1.760709725520533, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 56388 + }, + { + "epoch": 0.56389, + "grad_norm": 1.6408948920453836, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 56389 + }, + { + "epoch": 0.5639, + "grad_norm": 1.397776664233083, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 56390 + }, + { + "epoch": 0.56391, + "grad_norm": 1.8082069165069683, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 56391 + }, + { + "epoch": 0.56392, + "grad_norm": 1.282187786341918, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 56392 + }, + { + "epoch": 0.56393, + "grad_norm": 1.5952892119517537, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 56393 + }, + { + "epoch": 0.56394, + "grad_norm": 1.4767279983474357, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 56394 + }, + { + "epoch": 0.56395, + "grad_norm": 1.3598408765647785, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 56395 + }, + { + "epoch": 0.56396, + "grad_norm": 1.434018209166031, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 56396 + }, + { + "epoch": 0.56397, + "grad_norm": 1.6361711023023102, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 56397 + }, + { + "epoch": 0.56398, + "grad_norm": 1.6213307090084512, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 56398 + }, + { + "epoch": 0.56399, + "grad_norm": 1.3714560677444811, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 56399 + }, + { + "epoch": 0.564, + "grad_norm": 1.5594774548795234, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 56400 + }, + { + "epoch": 0.56401, + "grad_norm": 1.6589270247300234, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 56401 + }, + { + "epoch": 0.56402, + "grad_norm": 1.5114891683007152, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 56402 + }, + { + "epoch": 0.56403, + "grad_norm": 1.6002093203799657, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 56403 + }, + { + "epoch": 0.56404, + "grad_norm": 1.7846396360137946, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 56404 + }, + { + "epoch": 0.56405, + "grad_norm": 1.8239056927819768, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 56405 + }, + { + "epoch": 0.56406, + "grad_norm": 1.621404280569811, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 56406 + }, + { + "epoch": 0.56407, + "grad_norm": 1.9838911438206634, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 56407 + }, + { + "epoch": 0.56408, + "grad_norm": 1.542937121064198, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 56408 + }, + { + "epoch": 0.56409, + "grad_norm": 1.4145321248062968, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 56409 + }, + { + "epoch": 0.5641, + "grad_norm": 1.819150142523567, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 56410 + }, + { + "epoch": 0.56411, + "grad_norm": 1.4072898641011669, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 56411 + }, + { + "epoch": 0.56412, + "grad_norm": 1.4659728171233215, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 56412 + }, + { + "epoch": 0.56413, + "grad_norm": 1.3341213282101905, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 56413 + }, + { + "epoch": 0.56414, + "grad_norm": 1.6378603709653172, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 56414 + }, + { + "epoch": 0.56415, + "grad_norm": 1.3521741947713435, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 56415 + }, + { + "epoch": 0.56416, + "grad_norm": 1.5712307627215778, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 56416 + }, + { + "epoch": 0.56417, + "grad_norm": 1.6653104084433428, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 56417 + }, + { + "epoch": 0.56418, + "grad_norm": 1.9893267481374108, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 56418 + }, + { + "epoch": 0.56419, + "grad_norm": 1.7713975695853181, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 56419 + }, + { + "epoch": 0.5642, + "grad_norm": 1.7006832239863805, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 56420 + }, + { + "epoch": 0.56421, + "grad_norm": 1.538398220442289, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 56421 + }, + { + "epoch": 0.56422, + "grad_norm": 1.582836788567902, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 56422 + }, + { + "epoch": 0.56423, + "grad_norm": 1.4762788217301084, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 56423 + }, + { + "epoch": 0.56424, + "grad_norm": 1.7029718288481281, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 56424 + }, + { + "epoch": 0.56425, + "grad_norm": 1.4095432546115991, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 56425 + }, + { + "epoch": 0.56426, + "grad_norm": 1.6773284529237305, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 56426 + }, + { + "epoch": 0.56427, + "grad_norm": 1.5933022495473628, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 56427 + }, + { + "epoch": 0.56428, + "grad_norm": 1.429994945068548, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 56428 + }, + { + "epoch": 0.56429, + "grad_norm": 1.6870611435708065, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 56429 + }, + { + "epoch": 0.5643, + "grad_norm": 1.2896592522459585, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 56430 + }, + { + "epoch": 0.56431, + "grad_norm": 1.8766310519435914, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 56431 + }, + { + "epoch": 0.56432, + "grad_norm": 1.8099572787871, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 56432 + }, + { + "epoch": 0.56433, + "grad_norm": 1.4143811218399858, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 56433 + }, + { + "epoch": 0.56434, + "grad_norm": 1.6223562488679355, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 56434 + }, + { + "epoch": 0.56435, + "grad_norm": 1.6016604452332108, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 56435 + }, + { + "epoch": 0.56436, + "grad_norm": 1.5001322128761236, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 56436 + }, + { + "epoch": 0.56437, + "grad_norm": 1.474267683522091, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 56437 + }, + { + "epoch": 0.56438, + "grad_norm": 1.5307983775196583, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 56438 + }, + { + "epoch": 0.56439, + "grad_norm": 1.4548729571959707, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 56439 + }, + { + "epoch": 0.5644, + "grad_norm": 1.6754861928484732, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 56440 + }, + { + "epoch": 0.56441, + "grad_norm": 1.4651757231149214, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 56441 + }, + { + "epoch": 0.56442, + "grad_norm": 1.6729260410314235, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56442 + }, + { + "epoch": 0.56443, + "grad_norm": 1.6176244953362309, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 56443 + }, + { + "epoch": 0.56444, + "grad_norm": 1.7964751055881134, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56444 + }, + { + "epoch": 0.56445, + "grad_norm": 1.3486884390364728, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 56445 + }, + { + "epoch": 0.56446, + "grad_norm": 1.711405186109678, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 56446 + }, + { + "epoch": 0.56447, + "grad_norm": 1.6151620061207779, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 56447 + }, + { + "epoch": 0.56448, + "grad_norm": 1.5661407673407848, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 56448 + }, + { + "epoch": 0.56449, + "grad_norm": 1.6883042416672842, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 56449 + }, + { + "epoch": 0.5645, + "grad_norm": 1.5127981223676068, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 56450 + }, + { + "epoch": 0.56451, + "grad_norm": 1.4996979880401689, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 56451 + }, + { + "epoch": 0.56452, + "grad_norm": 1.6582605733331017, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 56452 + }, + { + "epoch": 0.56453, + "grad_norm": 1.5205633313838431, + "learning_rate": 0.003, + "loss": 4.039, + "step": 56453 + }, + { + "epoch": 0.56454, + "grad_norm": 1.5615725421725477, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 56454 + }, + { + "epoch": 0.56455, + "grad_norm": 1.6888459903228767, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 56455 + }, + { + "epoch": 0.56456, + "grad_norm": 1.9817661129049733, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 56456 + }, + { + "epoch": 0.56457, + "grad_norm": 1.3384331369071654, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 56457 + }, + { + "epoch": 0.56458, + "grad_norm": 1.762426604106959, + "learning_rate": 0.003, + "loss": 4.029, + "step": 56458 + }, + { + "epoch": 0.56459, + "grad_norm": 1.7893439426798068, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 56459 + }, + { + "epoch": 0.5646, + "grad_norm": 1.7131080578826545, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 56460 + }, + { + "epoch": 0.56461, + "grad_norm": 1.635224117281945, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 56461 + }, + { + "epoch": 0.56462, + "grad_norm": 1.422228347938256, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56462 + }, + { + "epoch": 0.56463, + "grad_norm": 1.8743211091244991, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 56463 + }, + { + "epoch": 0.56464, + "grad_norm": 1.0593515889449137, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 56464 + }, + { + "epoch": 0.56465, + "grad_norm": 1.7623658250351382, + "learning_rate": 0.003, + "loss": 4.013, + "step": 56465 + }, + { + "epoch": 0.56466, + "grad_norm": 1.8579992287825238, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 56466 + }, + { + "epoch": 0.56467, + "grad_norm": 1.650784424119748, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 56467 + }, + { + "epoch": 0.56468, + "grad_norm": 1.7916104036931713, + "learning_rate": 0.003, + "loss": 4.03, + "step": 56468 + }, + { + "epoch": 0.56469, + "grad_norm": 1.2588510181391164, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 56469 + }, + { + "epoch": 0.5647, + "grad_norm": 1.7470771553085722, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56470 + }, + { + "epoch": 0.56471, + "grad_norm": 1.45386474875532, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 56471 + }, + { + "epoch": 0.56472, + "grad_norm": 1.3631840202888554, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 56472 + }, + { + "epoch": 0.56473, + "grad_norm": 1.7451833694684116, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 56473 + }, + { + "epoch": 0.56474, + "grad_norm": 1.4154082944498543, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 56474 + }, + { + "epoch": 0.56475, + "grad_norm": 1.8283905735668553, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 56475 + }, + { + "epoch": 0.56476, + "grad_norm": 1.685446876092245, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 56476 + }, + { + "epoch": 0.56477, + "grad_norm": 1.940207854992061, + "learning_rate": 0.003, + "loss": 4.053, + "step": 56477 + }, + { + "epoch": 0.56478, + "grad_norm": 1.2968270517538234, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 56478 + }, + { + "epoch": 0.56479, + "grad_norm": 1.7883278606205166, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 56479 + }, + { + "epoch": 0.5648, + "grad_norm": 1.5667813681548026, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 56480 + }, + { + "epoch": 0.56481, + "grad_norm": 1.734932801550652, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 56481 + }, + { + "epoch": 0.56482, + "grad_norm": 1.4599576552961029, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56482 + }, + { + "epoch": 0.56483, + "grad_norm": 1.4354870009124678, + "learning_rate": 0.003, + "loss": 4.027, + "step": 56483 + }, + { + "epoch": 0.56484, + "grad_norm": 1.7964408349229626, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 56484 + }, + { + "epoch": 0.56485, + "grad_norm": 1.330711867793338, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56485 + }, + { + "epoch": 0.56486, + "grad_norm": 1.8089035035794276, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 56486 + }, + { + "epoch": 0.56487, + "grad_norm": 1.3785627905462419, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 56487 + }, + { + "epoch": 0.56488, + "grad_norm": 1.4367723464147244, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 56488 + }, + { + "epoch": 0.56489, + "grad_norm": 1.4973571063552507, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 56489 + }, + { + "epoch": 0.5649, + "grad_norm": 1.6312255340783304, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 56490 + }, + { + "epoch": 0.56491, + "grad_norm": 1.9629644841610674, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 56491 + }, + { + "epoch": 0.56492, + "grad_norm": 1.7419047728452026, + "learning_rate": 0.003, + "loss": 4.05, + "step": 56492 + }, + { + "epoch": 0.56493, + "grad_norm": 1.6398936888545668, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 56493 + }, + { + "epoch": 0.56494, + "grad_norm": 1.5787599148752356, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 56494 + }, + { + "epoch": 0.56495, + "grad_norm": 1.3704028609784453, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 56495 + }, + { + "epoch": 0.56496, + "grad_norm": 1.7560613234168772, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 56496 + }, + { + "epoch": 0.56497, + "grad_norm": 1.6457192100779534, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 56497 + }, + { + "epoch": 0.56498, + "grad_norm": 1.4075691875706147, + "learning_rate": 0.003, + "loss": 4.031, + "step": 56498 + }, + { + "epoch": 0.56499, + "grad_norm": 1.6960665950052134, + "learning_rate": 0.003, + "loss": 4.047, + "step": 56499 + }, + { + "epoch": 0.565, + "grad_norm": 1.7854472061000943, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 56500 + }, + { + "epoch": 0.56501, + "grad_norm": 1.5438718559859084, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 56501 + }, + { + "epoch": 0.56502, + "grad_norm": 1.3925162733039431, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 56502 + }, + { + "epoch": 0.56503, + "grad_norm": 1.398043080060356, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 56503 + }, + { + "epoch": 0.56504, + "grad_norm": 1.7205541459915337, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 56504 + }, + { + "epoch": 0.56505, + "grad_norm": 1.291591636330303, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 56505 + }, + { + "epoch": 0.56506, + "grad_norm": 2.142314090524997, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 56506 + }, + { + "epoch": 0.56507, + "grad_norm": 1.4189410041273454, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56507 + }, + { + "epoch": 0.56508, + "grad_norm": 1.5805535997351188, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 56508 + }, + { + "epoch": 0.56509, + "grad_norm": 1.6229995134387916, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 56509 + }, + { + "epoch": 0.5651, + "grad_norm": 1.399572077059356, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 56510 + }, + { + "epoch": 0.56511, + "grad_norm": 1.7813851436462533, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 56511 + }, + { + "epoch": 0.56512, + "grad_norm": 1.7354622107862163, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 56512 + }, + { + "epoch": 0.56513, + "grad_norm": 1.703046412484873, + "learning_rate": 0.003, + "loss": 4.007, + "step": 56513 + }, + { + "epoch": 0.56514, + "grad_norm": 1.525687560250005, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 56514 + }, + { + "epoch": 0.56515, + "grad_norm": 1.5682006939908744, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 56515 + }, + { + "epoch": 0.56516, + "grad_norm": 1.7934063930717703, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 56516 + }, + { + "epoch": 0.56517, + "grad_norm": 1.5821127543652787, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 56517 + }, + { + "epoch": 0.56518, + "grad_norm": 1.5595141938526442, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 56518 + }, + { + "epoch": 0.56519, + "grad_norm": 1.5160739876485885, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 56519 + }, + { + "epoch": 0.5652, + "grad_norm": 1.474736034535572, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 56520 + }, + { + "epoch": 0.56521, + "grad_norm": 1.813289566271452, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56521 + }, + { + "epoch": 0.56522, + "grad_norm": 1.2087888835184264, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56522 + }, + { + "epoch": 0.56523, + "grad_norm": 1.9723336229359218, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 56523 + }, + { + "epoch": 0.56524, + "grad_norm": 1.2170614066544785, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 56524 + }, + { + "epoch": 0.56525, + "grad_norm": 1.6778625260100157, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 56525 + }, + { + "epoch": 0.56526, + "grad_norm": 1.4410835764337975, + "learning_rate": 0.003, + "loss": 4.048, + "step": 56526 + }, + { + "epoch": 0.56527, + "grad_norm": 1.541161877112954, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 56527 + }, + { + "epoch": 0.56528, + "grad_norm": 1.5758580366773955, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56528 + }, + { + "epoch": 0.56529, + "grad_norm": 1.5758679765366606, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 56529 + }, + { + "epoch": 0.5653, + "grad_norm": 1.7629717764838462, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 56530 + }, + { + "epoch": 0.56531, + "grad_norm": 2.1541107478420654, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 56531 + }, + { + "epoch": 0.56532, + "grad_norm": 1.3099957495685812, + "learning_rate": 0.003, + "loss": 4.039, + "step": 56532 + }, + { + "epoch": 0.56533, + "grad_norm": 1.4925269196670816, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 56533 + }, + { + "epoch": 0.56534, + "grad_norm": 1.7940617506313588, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56534 + }, + { + "epoch": 0.56535, + "grad_norm": 1.5829385991079816, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 56535 + }, + { + "epoch": 0.56536, + "grad_norm": 2.0723072030718295, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 56536 + }, + { + "epoch": 0.56537, + "grad_norm": 1.3960441539723552, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 56537 + }, + { + "epoch": 0.56538, + "grad_norm": 1.8446685144276034, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 56538 + }, + { + "epoch": 0.56539, + "grad_norm": 1.421749958533072, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 56539 + }, + { + "epoch": 0.5654, + "grad_norm": 1.6714953095597005, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 56540 + }, + { + "epoch": 0.56541, + "grad_norm": 1.5081000232493738, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 56541 + }, + { + "epoch": 0.56542, + "grad_norm": 1.6784917713701621, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 56542 + }, + { + "epoch": 0.56543, + "grad_norm": 1.67410683960182, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 56543 + }, + { + "epoch": 0.56544, + "grad_norm": 1.518712429789068, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 56544 + }, + { + "epoch": 0.56545, + "grad_norm": 1.696501886532396, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 56545 + }, + { + "epoch": 0.56546, + "grad_norm": 1.439660585356201, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 56546 + }, + { + "epoch": 0.56547, + "grad_norm": 1.8977193632877933, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 56547 + }, + { + "epoch": 0.56548, + "grad_norm": 1.2574439483094626, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 56548 + }, + { + "epoch": 0.56549, + "grad_norm": 1.615950683639114, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 56549 + }, + { + "epoch": 0.5655, + "grad_norm": 1.5323740805855184, + "learning_rate": 0.003, + "loss": 4.031, + "step": 56550 + }, + { + "epoch": 0.56551, + "grad_norm": 1.3778628545275724, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 56551 + }, + { + "epoch": 0.56552, + "grad_norm": 1.9648957200121777, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 56552 + }, + { + "epoch": 0.56553, + "grad_norm": 1.489381017244567, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 56553 + }, + { + "epoch": 0.56554, + "grad_norm": 1.8450934429234027, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 56554 + }, + { + "epoch": 0.56555, + "grad_norm": 1.5174321199214484, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 56555 + }, + { + "epoch": 0.56556, + "grad_norm": 1.5666249276352682, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 56556 + }, + { + "epoch": 0.56557, + "grad_norm": 1.4457254638304504, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 56557 + }, + { + "epoch": 0.56558, + "grad_norm": 1.5919760426899319, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 56558 + }, + { + "epoch": 0.56559, + "grad_norm": 1.4704999124336668, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 56559 + }, + { + "epoch": 0.5656, + "grad_norm": 1.5642935809179097, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 56560 + }, + { + "epoch": 0.56561, + "grad_norm": 1.7075930875025498, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 56561 + }, + { + "epoch": 0.56562, + "grad_norm": 1.562911958756581, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 56562 + }, + { + "epoch": 0.56563, + "grad_norm": 1.5667409610919987, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 56563 + }, + { + "epoch": 0.56564, + "grad_norm": 1.764874921618993, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 56564 + }, + { + "epoch": 0.56565, + "grad_norm": 1.5623212476627177, + "learning_rate": 0.003, + "loss": 4.012, + "step": 56565 + }, + { + "epoch": 0.56566, + "grad_norm": 1.5509354937506616, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 56566 + }, + { + "epoch": 0.56567, + "grad_norm": 1.449180344882271, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 56567 + }, + { + "epoch": 0.56568, + "grad_norm": 1.7730005235076993, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 56568 + }, + { + "epoch": 0.56569, + "grad_norm": 1.5842971131018573, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 56569 + }, + { + "epoch": 0.5657, + "grad_norm": 1.6966377341364096, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56570 + }, + { + "epoch": 0.56571, + "grad_norm": 1.5119044419003107, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 56571 + }, + { + "epoch": 0.56572, + "grad_norm": 1.7438474826496562, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 56572 + }, + { + "epoch": 0.56573, + "grad_norm": 1.3894864974999879, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 56573 + }, + { + "epoch": 0.56574, + "grad_norm": 1.6446594681474742, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 56574 + }, + { + "epoch": 0.56575, + "grad_norm": 1.5099883428053722, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56575 + }, + { + "epoch": 0.56576, + "grad_norm": 1.745209562696036, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 56576 + }, + { + "epoch": 0.56577, + "grad_norm": 1.5215950911164273, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 56577 + }, + { + "epoch": 0.56578, + "grad_norm": 1.565985198776907, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 56578 + }, + { + "epoch": 0.56579, + "grad_norm": 1.3757579615577915, + "learning_rate": 0.003, + "loss": 4.037, + "step": 56579 + }, + { + "epoch": 0.5658, + "grad_norm": 1.643715556131211, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 56580 + }, + { + "epoch": 0.56581, + "grad_norm": 1.7187630683749393, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56581 + }, + { + "epoch": 0.56582, + "grad_norm": 1.8599963297667106, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 56582 + }, + { + "epoch": 0.56583, + "grad_norm": 1.9159846217236365, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 56583 + }, + { + "epoch": 0.56584, + "grad_norm": 1.4261892482960379, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 56584 + }, + { + "epoch": 0.56585, + "grad_norm": 1.5124201015680936, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 56585 + }, + { + "epoch": 0.56586, + "grad_norm": 2.060338119601697, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 56586 + }, + { + "epoch": 0.56587, + "grad_norm": 1.2559031378758199, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 56587 + }, + { + "epoch": 0.56588, + "grad_norm": 1.650915538882217, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 56588 + }, + { + "epoch": 0.56589, + "grad_norm": 1.5379525699398717, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 56589 + }, + { + "epoch": 0.5659, + "grad_norm": 1.745880672925872, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 56590 + }, + { + "epoch": 0.56591, + "grad_norm": 1.1621609712228216, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 56591 + }, + { + "epoch": 0.56592, + "grad_norm": 2.120294963557124, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 56592 + }, + { + "epoch": 0.56593, + "grad_norm": 1.570538736979387, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 56593 + }, + { + "epoch": 0.56594, + "grad_norm": 1.6628439657856793, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 56594 + }, + { + "epoch": 0.56595, + "grad_norm": 1.4846535643470655, + "learning_rate": 0.003, + "loss": 4.022, + "step": 56595 + }, + { + "epoch": 0.56596, + "grad_norm": 1.4840103661253365, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 56596 + }, + { + "epoch": 0.56597, + "grad_norm": 1.6743847639307825, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 56597 + }, + { + "epoch": 0.56598, + "grad_norm": 1.6522401326678617, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56598 + }, + { + "epoch": 0.56599, + "grad_norm": 1.1838686139746484, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 56599 + }, + { + "epoch": 0.566, + "grad_norm": 2.016595584461381, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 56600 + }, + { + "epoch": 0.56601, + "grad_norm": 1.4150635745481177, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 56601 + }, + { + "epoch": 0.56602, + "grad_norm": 1.8358698738898989, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 56602 + }, + { + "epoch": 0.56603, + "grad_norm": 1.1701811077010495, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 56603 + }, + { + "epoch": 0.56604, + "grad_norm": 1.994285620519307, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 56604 + }, + { + "epoch": 0.56605, + "grad_norm": 1.7041953389146154, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 56605 + }, + { + "epoch": 0.56606, + "grad_norm": 1.1462268659609207, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 56606 + }, + { + "epoch": 0.56607, + "grad_norm": 1.9735508847313044, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 56607 + }, + { + "epoch": 0.56608, + "grad_norm": 1.413850976949688, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 56608 + }, + { + "epoch": 0.56609, + "grad_norm": 1.7272752552318011, + "learning_rate": 0.003, + "loss": 4.04, + "step": 56609 + }, + { + "epoch": 0.5661, + "grad_norm": 1.5114945102022415, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 56610 + }, + { + "epoch": 0.56611, + "grad_norm": 1.5256092922816311, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 56611 + }, + { + "epoch": 0.56612, + "grad_norm": 1.3832439283542928, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 56612 + }, + { + "epoch": 0.56613, + "grad_norm": 1.6999057214039321, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 56613 + }, + { + "epoch": 0.56614, + "grad_norm": 1.4421045868923161, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 56614 + }, + { + "epoch": 0.56615, + "grad_norm": 1.392259825529093, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 56615 + }, + { + "epoch": 0.56616, + "grad_norm": 1.5591427894968855, + "learning_rate": 0.003, + "loss": 4.019, + "step": 56616 + }, + { + "epoch": 0.56617, + "grad_norm": 1.784934307480931, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 56617 + }, + { + "epoch": 0.56618, + "grad_norm": 1.7585706077188072, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 56618 + }, + { + "epoch": 0.56619, + "grad_norm": 1.466546481530902, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 56619 + }, + { + "epoch": 0.5662, + "grad_norm": 1.5935621311697035, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 56620 + }, + { + "epoch": 0.56621, + "grad_norm": 1.627840943022366, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 56621 + }, + { + "epoch": 0.56622, + "grad_norm": 1.4997978572387873, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 56622 + }, + { + "epoch": 0.56623, + "grad_norm": 1.6742431708199743, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 56623 + }, + { + "epoch": 0.56624, + "grad_norm": 1.2904227956606493, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 56624 + }, + { + "epoch": 0.56625, + "grad_norm": 1.4925071829519287, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 56625 + }, + { + "epoch": 0.56626, + "grad_norm": 1.2960450730183077, + "learning_rate": 0.003, + "loss": 4.051, + "step": 56626 + }, + { + "epoch": 0.56627, + "grad_norm": 1.5382179397959657, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 56627 + }, + { + "epoch": 0.56628, + "grad_norm": 1.4894818828824457, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 56628 + }, + { + "epoch": 0.56629, + "grad_norm": 1.7158276404586172, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 56629 + }, + { + "epoch": 0.5663, + "grad_norm": 1.2971799394130634, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 56630 + }, + { + "epoch": 0.56631, + "grad_norm": 1.6210023443460733, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 56631 + }, + { + "epoch": 0.56632, + "grad_norm": 1.5154841090743434, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 56632 + }, + { + "epoch": 0.56633, + "grad_norm": 1.909779713666353, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 56633 + }, + { + "epoch": 0.56634, + "grad_norm": 1.4586175712976632, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 56634 + }, + { + "epoch": 0.56635, + "grad_norm": 2.228722551267377, + "learning_rate": 0.003, + "loss": 4.059, + "step": 56635 + }, + { + "epoch": 0.56636, + "grad_norm": 1.711120613014037, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 56636 + }, + { + "epoch": 0.56637, + "grad_norm": 1.5854397395966224, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56637 + }, + { + "epoch": 0.56638, + "grad_norm": 1.6869782036916157, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 56638 + }, + { + "epoch": 0.56639, + "grad_norm": 2.1497473141376537, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 56639 + }, + { + "epoch": 0.5664, + "grad_norm": 0.977741279997171, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 56640 + }, + { + "epoch": 0.56641, + "grad_norm": 1.8199880150920875, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 56641 + }, + { + "epoch": 0.56642, + "grad_norm": 1.523619224144639, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 56642 + }, + { + "epoch": 0.56643, + "grad_norm": 1.7659713641888226, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 56643 + }, + { + "epoch": 0.56644, + "grad_norm": 1.5049895795330737, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 56644 + }, + { + "epoch": 0.56645, + "grad_norm": 2.0992738108685463, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 56645 + }, + { + "epoch": 0.56646, + "grad_norm": 1.330187096177046, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 56646 + }, + { + "epoch": 0.56647, + "grad_norm": 1.9017426775822963, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 56647 + }, + { + "epoch": 0.56648, + "grad_norm": 1.19496206104936, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 56648 + }, + { + "epoch": 0.56649, + "grad_norm": 1.768905708122054, + "learning_rate": 0.003, + "loss": 4.053, + "step": 56649 + }, + { + "epoch": 0.5665, + "grad_norm": 1.5511736432206649, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 56650 + }, + { + "epoch": 0.56651, + "grad_norm": 1.4987703982423528, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 56651 + }, + { + "epoch": 0.56652, + "grad_norm": 1.5419604766043733, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 56652 + }, + { + "epoch": 0.56653, + "grad_norm": 1.677057262994212, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 56653 + }, + { + "epoch": 0.56654, + "grad_norm": 1.4700215921840447, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 56654 + }, + { + "epoch": 0.56655, + "grad_norm": 1.5800948512170654, + "learning_rate": 0.003, + "loss": 4.04, + "step": 56655 + }, + { + "epoch": 0.56656, + "grad_norm": 1.4179690321437897, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56656 + }, + { + "epoch": 0.56657, + "grad_norm": 1.7094645886391884, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 56657 + }, + { + "epoch": 0.56658, + "grad_norm": 1.3954421832250514, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 56658 + }, + { + "epoch": 0.56659, + "grad_norm": 1.9101395736505025, + "learning_rate": 0.003, + "loss": 4.06, + "step": 56659 + }, + { + "epoch": 0.5666, + "grad_norm": 1.5776173652003285, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 56660 + }, + { + "epoch": 0.56661, + "grad_norm": 1.6066581227672994, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 56661 + }, + { + "epoch": 0.56662, + "grad_norm": 1.7466355711400763, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 56662 + }, + { + "epoch": 0.56663, + "grad_norm": 1.9928437010212212, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 56663 + }, + { + "epoch": 0.56664, + "grad_norm": 1.3535845821551615, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 56664 + }, + { + "epoch": 0.56665, + "grad_norm": 1.9678036168857511, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 56665 + }, + { + "epoch": 0.56666, + "grad_norm": 1.411166207064491, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 56666 + }, + { + "epoch": 0.56667, + "grad_norm": 1.437679795625751, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 56667 + }, + { + "epoch": 0.56668, + "grad_norm": 1.5347502761449199, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 56668 + }, + { + "epoch": 0.56669, + "grad_norm": 1.372126866059845, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 56669 + }, + { + "epoch": 0.5667, + "grad_norm": 1.7472973949354558, + "learning_rate": 0.003, + "loss": 4.066, + "step": 56670 + }, + { + "epoch": 0.56671, + "grad_norm": 1.2984077971530756, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 56671 + }, + { + "epoch": 0.56672, + "grad_norm": 1.882749194578949, + "learning_rate": 0.003, + "loss": 4.021, + "step": 56672 + }, + { + "epoch": 0.56673, + "grad_norm": 1.3375500173544321, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 56673 + }, + { + "epoch": 0.56674, + "grad_norm": 1.7730137149371308, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 56674 + }, + { + "epoch": 0.56675, + "grad_norm": 1.5606469653294066, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 56675 + }, + { + "epoch": 0.56676, + "grad_norm": 1.5692379381748898, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 56676 + }, + { + "epoch": 0.56677, + "grad_norm": 1.4346323743581653, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 56677 + }, + { + "epoch": 0.56678, + "grad_norm": 1.7246105812805865, + "learning_rate": 0.003, + "loss": 3.99, + "step": 56678 + }, + { + "epoch": 0.56679, + "grad_norm": 1.4367934493267633, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 56679 + }, + { + "epoch": 0.5668, + "grad_norm": 1.6577801459850834, + "learning_rate": 0.003, + "loss": 4.031, + "step": 56680 + }, + { + "epoch": 0.56681, + "grad_norm": 1.1494180435040116, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 56681 + }, + { + "epoch": 0.56682, + "grad_norm": 1.8429890230709225, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 56682 + }, + { + "epoch": 0.56683, + "grad_norm": 1.4453395206143114, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 56683 + }, + { + "epoch": 0.56684, + "grad_norm": 1.7021113633197205, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 56684 + }, + { + "epoch": 0.56685, + "grad_norm": 1.546888010899693, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 56685 + }, + { + "epoch": 0.56686, + "grad_norm": 1.429379259550975, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 56686 + }, + { + "epoch": 0.56687, + "grad_norm": 1.4160172630027783, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 56687 + }, + { + "epoch": 0.56688, + "grad_norm": 1.6251271753009335, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 56688 + }, + { + "epoch": 0.56689, + "grad_norm": 1.3583224994685288, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 56689 + }, + { + "epoch": 0.5669, + "grad_norm": 1.4040154736380446, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56690 + }, + { + "epoch": 0.56691, + "grad_norm": 1.7746908946624802, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 56691 + }, + { + "epoch": 0.56692, + "grad_norm": 1.2246661569998671, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 56692 + }, + { + "epoch": 0.56693, + "grad_norm": 1.9574561146392626, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 56693 + }, + { + "epoch": 0.56694, + "grad_norm": 1.355312356029919, + "learning_rate": 0.003, + "loss": 4.055, + "step": 56694 + }, + { + "epoch": 0.56695, + "grad_norm": 1.894843345686394, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 56695 + }, + { + "epoch": 0.56696, + "grad_norm": 1.4980445153274824, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 56696 + }, + { + "epoch": 0.56697, + "grad_norm": 1.6970572180775318, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 56697 + }, + { + "epoch": 0.56698, + "grad_norm": 1.8510413056990727, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 56698 + }, + { + "epoch": 0.56699, + "grad_norm": 1.4654021547818843, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 56699 + }, + { + "epoch": 0.567, + "grad_norm": 1.4481950457326964, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 56700 + }, + { + "epoch": 0.56701, + "grad_norm": 1.7097352033890199, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 56701 + }, + { + "epoch": 0.56702, + "grad_norm": 1.444860374118532, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 56702 + }, + { + "epoch": 0.56703, + "grad_norm": 1.7192038391268136, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 56703 + }, + { + "epoch": 0.56704, + "grad_norm": 1.4988205400724854, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 56704 + }, + { + "epoch": 0.56705, + "grad_norm": 1.677878857323478, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 56705 + }, + { + "epoch": 0.56706, + "grad_norm": 1.4883763263569887, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 56706 + }, + { + "epoch": 0.56707, + "grad_norm": 1.7996400094626002, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 56707 + }, + { + "epoch": 0.56708, + "grad_norm": 1.3470937656188429, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 56708 + }, + { + "epoch": 0.56709, + "grad_norm": 1.8259190283972189, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 56709 + }, + { + "epoch": 0.5671, + "grad_norm": 1.4910261992718963, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 56710 + }, + { + "epoch": 0.56711, + "grad_norm": 1.6179021572512096, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 56711 + }, + { + "epoch": 0.56712, + "grad_norm": 1.4937909437834307, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 56712 + }, + { + "epoch": 0.56713, + "grad_norm": 1.5320741058880964, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 56713 + }, + { + "epoch": 0.56714, + "grad_norm": 1.4423073044671535, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 56714 + }, + { + "epoch": 0.56715, + "grad_norm": 1.7124178233204765, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 56715 + }, + { + "epoch": 0.56716, + "grad_norm": 1.7164857231044195, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 56716 + }, + { + "epoch": 0.56717, + "grad_norm": 1.3391132630608737, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 56717 + }, + { + "epoch": 0.56718, + "grad_norm": 1.6651044740069247, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 56718 + }, + { + "epoch": 0.56719, + "grad_norm": 1.5178849855941963, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 56719 + }, + { + "epoch": 0.5672, + "grad_norm": 1.5661618001231017, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 56720 + }, + { + "epoch": 0.56721, + "grad_norm": 1.5351913968106075, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 56721 + }, + { + "epoch": 0.56722, + "grad_norm": 1.686423073336303, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 56722 + }, + { + "epoch": 0.56723, + "grad_norm": 1.514408916746434, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 56723 + }, + { + "epoch": 0.56724, + "grad_norm": 1.6308003889109262, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 56724 + }, + { + "epoch": 0.56725, + "grad_norm": 1.5938286974199944, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 56725 + }, + { + "epoch": 0.56726, + "grad_norm": 1.3638697199991383, + "learning_rate": 0.003, + "loss": 4.064, + "step": 56726 + }, + { + "epoch": 0.56727, + "grad_norm": 1.9184943547996527, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 56727 + }, + { + "epoch": 0.56728, + "grad_norm": 1.265338347187498, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 56728 + }, + { + "epoch": 0.56729, + "grad_norm": 1.7471172234609091, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 56729 + }, + { + "epoch": 0.5673, + "grad_norm": 1.4484418338602696, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 56730 + }, + { + "epoch": 0.56731, + "grad_norm": 1.4969945519097974, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 56731 + }, + { + "epoch": 0.56732, + "grad_norm": 1.5546776008177983, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56732 + }, + { + "epoch": 0.56733, + "grad_norm": 1.3037504694328208, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 56733 + }, + { + "epoch": 0.56734, + "grad_norm": 1.8192806583420396, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 56734 + }, + { + "epoch": 0.56735, + "grad_norm": 1.3878744175104096, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 56735 + }, + { + "epoch": 0.56736, + "grad_norm": 1.8472238313521923, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 56736 + }, + { + "epoch": 0.56737, + "grad_norm": 1.253511214970405, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 56737 + }, + { + "epoch": 0.56738, + "grad_norm": 1.9787112362817205, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 56738 + }, + { + "epoch": 0.56739, + "grad_norm": 1.2943120457390367, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 56739 + }, + { + "epoch": 0.5674, + "grad_norm": 1.6251952566612262, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 56740 + }, + { + "epoch": 0.56741, + "grad_norm": 1.658959621601446, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56741 + }, + { + "epoch": 0.56742, + "grad_norm": 1.864584869342838, + "learning_rate": 0.003, + "loss": 4.022, + "step": 56742 + }, + { + "epoch": 0.56743, + "grad_norm": 1.6151376376570605, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56743 + }, + { + "epoch": 0.56744, + "grad_norm": 1.64440621294848, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 56744 + }, + { + "epoch": 0.56745, + "grad_norm": 1.6276358901644346, + "learning_rate": 0.003, + "loss": 4.018, + "step": 56745 + }, + { + "epoch": 0.56746, + "grad_norm": 1.577226372327619, + "learning_rate": 0.003, + "loss": 4.023, + "step": 56746 + }, + { + "epoch": 0.56747, + "grad_norm": 1.58600777776376, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 56747 + }, + { + "epoch": 0.56748, + "grad_norm": 1.5007845172566743, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 56748 + }, + { + "epoch": 0.56749, + "grad_norm": 1.5373915324501555, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 56749 + }, + { + "epoch": 0.5675, + "grad_norm": 1.5657970860097048, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 56750 + }, + { + "epoch": 0.56751, + "grad_norm": 1.488475860886558, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 56751 + }, + { + "epoch": 0.56752, + "grad_norm": 2.059188771448791, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 56752 + }, + { + "epoch": 0.56753, + "grad_norm": 1.5570206685249062, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 56753 + }, + { + "epoch": 0.56754, + "grad_norm": 1.4540463397268917, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 56754 + }, + { + "epoch": 0.56755, + "grad_norm": 1.7020882463658922, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 56755 + }, + { + "epoch": 0.56756, + "grad_norm": 1.7168008944598594, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 56756 + }, + { + "epoch": 0.56757, + "grad_norm": 1.3499056483331622, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 56757 + }, + { + "epoch": 0.56758, + "grad_norm": 1.7363432615816718, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56758 + }, + { + "epoch": 0.56759, + "grad_norm": 1.2549321697993785, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 56759 + }, + { + "epoch": 0.5676, + "grad_norm": 1.9397683074607388, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 56760 + }, + { + "epoch": 0.56761, + "grad_norm": 1.6724773754403268, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 56761 + }, + { + "epoch": 0.56762, + "grad_norm": 1.5794080138374695, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 56762 + }, + { + "epoch": 0.56763, + "grad_norm": 1.8027509471704861, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 56763 + }, + { + "epoch": 0.56764, + "grad_norm": 1.3435722038848412, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 56764 + }, + { + "epoch": 0.56765, + "grad_norm": 1.9003048731886392, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 56765 + }, + { + "epoch": 0.56766, + "grad_norm": 1.413386202311785, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 56766 + }, + { + "epoch": 0.56767, + "grad_norm": 1.8715232246284748, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 56767 + }, + { + "epoch": 0.56768, + "grad_norm": 1.4941299191106232, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 56768 + }, + { + "epoch": 0.56769, + "grad_norm": 1.546431606845347, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 56769 + }, + { + "epoch": 0.5677, + "grad_norm": 1.3752806155707535, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 56770 + }, + { + "epoch": 0.56771, + "grad_norm": 1.532876291193005, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 56771 + }, + { + "epoch": 0.56772, + "grad_norm": 1.6663898082230189, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 56772 + }, + { + "epoch": 0.56773, + "grad_norm": 1.6514927327392537, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 56773 + }, + { + "epoch": 0.56774, + "grad_norm": 1.7041808449916154, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 56774 + }, + { + "epoch": 0.56775, + "grad_norm": 1.3968009423972871, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 56775 + }, + { + "epoch": 0.56776, + "grad_norm": 1.5253182679908375, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 56776 + }, + { + "epoch": 0.56777, + "grad_norm": 1.8317209867290132, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 56777 + }, + { + "epoch": 0.56778, + "grad_norm": 1.3127832091320228, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 56778 + }, + { + "epoch": 0.56779, + "grad_norm": 1.7853120291155584, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56779 + }, + { + "epoch": 0.5678, + "grad_norm": 1.5228494523396388, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 56780 + }, + { + "epoch": 0.56781, + "grad_norm": 1.8369396911484799, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 56781 + }, + { + "epoch": 0.56782, + "grad_norm": 1.528125333179926, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 56782 + }, + { + "epoch": 0.56783, + "grad_norm": 1.8898443545984318, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 56783 + }, + { + "epoch": 0.56784, + "grad_norm": 1.7516377843348898, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 56784 + }, + { + "epoch": 0.56785, + "grad_norm": 1.4045221103987149, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 56785 + }, + { + "epoch": 0.56786, + "grad_norm": 1.7645317210549742, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 56786 + }, + { + "epoch": 0.56787, + "grad_norm": 1.5217012539483423, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56787 + }, + { + "epoch": 0.56788, + "grad_norm": 1.7156577660831038, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 56788 + }, + { + "epoch": 0.56789, + "grad_norm": 1.5369386355541608, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 56789 + }, + { + "epoch": 0.5679, + "grad_norm": 1.4904878722433037, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 56790 + }, + { + "epoch": 0.56791, + "grad_norm": 1.628827363224678, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 56791 + }, + { + "epoch": 0.56792, + "grad_norm": 1.5626727399718354, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 56792 + }, + { + "epoch": 0.56793, + "grad_norm": 1.6174952860900234, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 56793 + }, + { + "epoch": 0.56794, + "grad_norm": 1.469026521830222, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 56794 + }, + { + "epoch": 0.56795, + "grad_norm": 1.644204947885293, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 56795 + }, + { + "epoch": 0.56796, + "grad_norm": 1.6657260562969338, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 56796 + }, + { + "epoch": 0.56797, + "grad_norm": 1.728923710354807, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 56797 + }, + { + "epoch": 0.56798, + "grad_norm": 1.6961615837934376, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56798 + }, + { + "epoch": 0.56799, + "grad_norm": 1.4063476731412266, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 56799 + }, + { + "epoch": 0.568, + "grad_norm": 1.932282674149228, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 56800 + }, + { + "epoch": 0.56801, + "grad_norm": 1.4914504911207456, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 56801 + }, + { + "epoch": 0.56802, + "grad_norm": 1.659867124940648, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 56802 + }, + { + "epoch": 0.56803, + "grad_norm": 1.750920581133936, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 56803 + }, + { + "epoch": 0.56804, + "grad_norm": 1.6769563793126592, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 56804 + }, + { + "epoch": 0.56805, + "grad_norm": 1.471554750799181, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 56805 + }, + { + "epoch": 0.56806, + "grad_norm": 1.591006019713322, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 56806 + }, + { + "epoch": 0.56807, + "grad_norm": 1.4686079644535477, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 56807 + }, + { + "epoch": 0.56808, + "grad_norm": 1.4617271327068073, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 56808 + }, + { + "epoch": 0.56809, + "grad_norm": 1.2107696678657807, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 56809 + }, + { + "epoch": 0.5681, + "grad_norm": 1.7031449786874995, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 56810 + }, + { + "epoch": 0.56811, + "grad_norm": 1.3298218619170101, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 56811 + }, + { + "epoch": 0.56812, + "grad_norm": 1.8835649787789144, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 56812 + }, + { + "epoch": 0.56813, + "grad_norm": 1.6856358625115797, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 56813 + }, + { + "epoch": 0.56814, + "grad_norm": 1.4055063069223013, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 56814 + }, + { + "epoch": 0.56815, + "grad_norm": 2.033714718245818, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 56815 + }, + { + "epoch": 0.56816, + "grad_norm": 1.5173689937892294, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 56816 + }, + { + "epoch": 0.56817, + "grad_norm": 1.5459632742262188, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56817 + }, + { + "epoch": 0.56818, + "grad_norm": 1.692886182565874, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 56818 + }, + { + "epoch": 0.56819, + "grad_norm": 1.5151135764073842, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56819 + }, + { + "epoch": 0.5682, + "grad_norm": 1.6117131458063154, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 56820 + }, + { + "epoch": 0.56821, + "grad_norm": 1.3498588602792214, + "learning_rate": 0.003, + "loss": 4.029, + "step": 56821 + }, + { + "epoch": 0.56822, + "grad_norm": 1.6444939971332215, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 56822 + }, + { + "epoch": 0.56823, + "grad_norm": 1.3403751348807036, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 56823 + }, + { + "epoch": 0.56824, + "grad_norm": 1.489921132397075, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 56824 + }, + { + "epoch": 0.56825, + "grad_norm": 1.5992090797908298, + "learning_rate": 0.003, + "loss": 4.032, + "step": 56825 + }, + { + "epoch": 0.56826, + "grad_norm": 1.5353088508484227, + "learning_rate": 0.003, + "loss": 4.018, + "step": 56826 + }, + { + "epoch": 0.56827, + "grad_norm": 1.5993810275122784, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 56827 + }, + { + "epoch": 0.56828, + "grad_norm": 1.2864637216063122, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 56828 + }, + { + "epoch": 0.56829, + "grad_norm": 1.8761773587423174, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 56829 + }, + { + "epoch": 0.5683, + "grad_norm": 1.4785842987455824, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 56830 + }, + { + "epoch": 0.56831, + "grad_norm": 1.4608804096719472, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 56831 + }, + { + "epoch": 0.56832, + "grad_norm": 1.8743998433076665, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 56832 + }, + { + "epoch": 0.56833, + "grad_norm": 1.4572671589418424, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 56833 + }, + { + "epoch": 0.56834, + "grad_norm": 2.011858048324306, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 56834 + }, + { + "epoch": 0.56835, + "grad_norm": 1.4593368946055587, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 56835 + }, + { + "epoch": 0.56836, + "grad_norm": 1.3294730504182708, + "learning_rate": 0.003, + "loss": 4.046, + "step": 56836 + }, + { + "epoch": 0.56837, + "grad_norm": 1.8323078183038108, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56837 + }, + { + "epoch": 0.56838, + "grad_norm": 1.8206968008221527, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 56838 + }, + { + "epoch": 0.56839, + "grad_norm": 1.530324673686084, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 56839 + }, + { + "epoch": 0.5684, + "grad_norm": 1.7266720256722745, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 56840 + }, + { + "epoch": 0.56841, + "grad_norm": 1.4954877297023277, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 56841 + }, + { + "epoch": 0.56842, + "grad_norm": 1.4918958689659678, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 56842 + }, + { + "epoch": 0.56843, + "grad_norm": 1.500966429799604, + "learning_rate": 0.003, + "loss": 4.039, + "step": 56843 + }, + { + "epoch": 0.56844, + "grad_norm": 1.3891211824872465, + "learning_rate": 0.003, + "loss": 3.999, + "step": 56844 + }, + { + "epoch": 0.56845, + "grad_norm": 1.85566029762477, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 56845 + }, + { + "epoch": 0.56846, + "grad_norm": 1.459472725332255, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 56846 + }, + { + "epoch": 0.56847, + "grad_norm": 1.3334477247211596, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 56847 + }, + { + "epoch": 0.56848, + "grad_norm": 1.5276111084059434, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 56848 + }, + { + "epoch": 0.56849, + "grad_norm": 1.3928950739239623, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 56849 + }, + { + "epoch": 0.5685, + "grad_norm": 1.6273574210113044, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 56850 + }, + { + "epoch": 0.56851, + "grad_norm": 1.4829061828521182, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 56851 + }, + { + "epoch": 0.56852, + "grad_norm": 1.2914057653717534, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56852 + }, + { + "epoch": 0.56853, + "grad_norm": 1.5262225274463257, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 56853 + }, + { + "epoch": 0.56854, + "grad_norm": 1.3608625129565932, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 56854 + }, + { + "epoch": 0.56855, + "grad_norm": 1.583235662962171, + "learning_rate": 0.003, + "loss": 4.05, + "step": 56855 + }, + { + "epoch": 0.56856, + "grad_norm": 1.8967491527705644, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 56856 + }, + { + "epoch": 0.56857, + "grad_norm": 2.042807111676155, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 56857 + }, + { + "epoch": 0.56858, + "grad_norm": 1.5789971533036902, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 56858 + }, + { + "epoch": 0.56859, + "grad_norm": 1.6357185016503009, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 56859 + }, + { + "epoch": 0.5686, + "grad_norm": 1.771232876370068, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 56860 + }, + { + "epoch": 0.56861, + "grad_norm": 1.4522285563250725, + "learning_rate": 0.003, + "loss": 4.046, + "step": 56861 + }, + { + "epoch": 0.56862, + "grad_norm": 1.667763105843688, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 56862 + }, + { + "epoch": 0.56863, + "grad_norm": 1.4673963456753278, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 56863 + }, + { + "epoch": 0.56864, + "grad_norm": 1.6596458023825882, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 56864 + }, + { + "epoch": 0.56865, + "grad_norm": 1.2211770553017172, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 56865 + }, + { + "epoch": 0.56866, + "grad_norm": 1.9766875701271558, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 56866 + }, + { + "epoch": 0.56867, + "grad_norm": 1.324856255354533, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 56867 + }, + { + "epoch": 0.56868, + "grad_norm": 1.8288578997015785, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 56868 + }, + { + "epoch": 0.56869, + "grad_norm": 1.315043075754402, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 56869 + }, + { + "epoch": 0.5687, + "grad_norm": 1.4796508267711814, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 56870 + }, + { + "epoch": 0.56871, + "grad_norm": 1.7875305891025264, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 56871 + }, + { + "epoch": 0.56872, + "grad_norm": 1.5385827414109237, + "learning_rate": 0.003, + "loss": 4.032, + "step": 56872 + }, + { + "epoch": 0.56873, + "grad_norm": 1.6289267895421635, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 56873 + }, + { + "epoch": 0.56874, + "grad_norm": 1.396717786072934, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 56874 + }, + { + "epoch": 0.56875, + "grad_norm": 1.4213506025835745, + "learning_rate": 0.003, + "loss": 4.029, + "step": 56875 + }, + { + "epoch": 0.56876, + "grad_norm": 1.4015415389931578, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 56876 + }, + { + "epoch": 0.56877, + "grad_norm": 1.8241619816849575, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 56877 + }, + { + "epoch": 0.56878, + "grad_norm": 1.8362755544852367, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 56878 + }, + { + "epoch": 0.56879, + "grad_norm": 1.576676380766892, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 56879 + }, + { + "epoch": 0.5688, + "grad_norm": 1.8541100117860516, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 56880 + }, + { + "epoch": 0.56881, + "grad_norm": 1.538575133532967, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 56881 + }, + { + "epoch": 0.56882, + "grad_norm": 1.3886137689263134, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 56882 + }, + { + "epoch": 0.56883, + "grad_norm": 1.8237220169287944, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 56883 + }, + { + "epoch": 0.56884, + "grad_norm": 1.5056239338317157, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 56884 + }, + { + "epoch": 0.56885, + "grad_norm": 1.7715300114700936, + "learning_rate": 0.003, + "loss": 4.037, + "step": 56885 + }, + { + "epoch": 0.56886, + "grad_norm": 1.2383300828202228, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 56886 + }, + { + "epoch": 0.56887, + "grad_norm": 2.069275847915428, + "learning_rate": 0.003, + "loss": 4.011, + "step": 56887 + }, + { + "epoch": 0.56888, + "grad_norm": 1.337107969802063, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 56888 + }, + { + "epoch": 0.56889, + "grad_norm": 1.8991820987248444, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 56889 + }, + { + "epoch": 0.5689, + "grad_norm": 1.5921256852253733, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 56890 + }, + { + "epoch": 0.56891, + "grad_norm": 1.3591549886115162, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 56891 + }, + { + "epoch": 0.56892, + "grad_norm": 1.5435408531149077, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 56892 + }, + { + "epoch": 0.56893, + "grad_norm": 1.4168133898213753, + "learning_rate": 0.003, + "loss": 4.049, + "step": 56893 + }, + { + "epoch": 0.56894, + "grad_norm": 1.5139775249348575, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 56894 + }, + { + "epoch": 0.56895, + "grad_norm": 1.7304194182263533, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 56895 + }, + { + "epoch": 0.56896, + "grad_norm": 1.5163884995832113, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 56896 + }, + { + "epoch": 0.56897, + "grad_norm": 1.6507606064651654, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 56897 + }, + { + "epoch": 0.56898, + "grad_norm": 1.3549098024321635, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 56898 + }, + { + "epoch": 0.56899, + "grad_norm": 1.6635176419352748, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 56899 + }, + { + "epoch": 0.569, + "grad_norm": 1.3746354800872398, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 56900 + }, + { + "epoch": 0.56901, + "grad_norm": 1.5726213011577999, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 56901 + }, + { + "epoch": 0.56902, + "grad_norm": 1.5053934736964258, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 56902 + }, + { + "epoch": 0.56903, + "grad_norm": 1.7505624945891536, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 56903 + }, + { + "epoch": 0.56904, + "grad_norm": 1.6360281682769768, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 56904 + }, + { + "epoch": 0.56905, + "grad_norm": 1.6063647418542906, + "learning_rate": 0.003, + "loss": 4.038, + "step": 56905 + }, + { + "epoch": 0.56906, + "grad_norm": 1.4844803965969502, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 56906 + }, + { + "epoch": 0.56907, + "grad_norm": 1.9011064163567672, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 56907 + }, + { + "epoch": 0.56908, + "grad_norm": 1.4562752936276362, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 56908 + }, + { + "epoch": 0.56909, + "grad_norm": 1.9170723873152693, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 56909 + }, + { + "epoch": 0.5691, + "grad_norm": 1.377817857352959, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 56910 + }, + { + "epoch": 0.56911, + "grad_norm": 1.7365150364072615, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 56911 + }, + { + "epoch": 0.56912, + "grad_norm": 1.2083267348478461, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 56912 + }, + { + "epoch": 0.56913, + "grad_norm": 1.590710360254768, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 56913 + }, + { + "epoch": 0.56914, + "grad_norm": 1.5403664479158528, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 56914 + }, + { + "epoch": 0.56915, + "grad_norm": 1.5828316186694307, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 56915 + }, + { + "epoch": 0.56916, + "grad_norm": 1.3906630283981767, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 56916 + }, + { + "epoch": 0.56917, + "grad_norm": 1.4583509676073265, + "learning_rate": 0.003, + "loss": 3.997, + "step": 56917 + }, + { + "epoch": 0.56918, + "grad_norm": 1.682701477544443, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 56918 + }, + { + "epoch": 0.56919, + "grad_norm": 1.5063585248096438, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 56919 + }, + { + "epoch": 0.5692, + "grad_norm": 1.7784706028446076, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 56920 + }, + { + "epoch": 0.56921, + "grad_norm": 1.3156478617730976, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 56921 + }, + { + "epoch": 0.56922, + "grad_norm": 1.8605931831102678, + "learning_rate": 0.003, + "loss": 4.036, + "step": 56922 + }, + { + "epoch": 0.56923, + "grad_norm": 1.6425963998748576, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 56923 + }, + { + "epoch": 0.56924, + "grad_norm": 1.5708180964787437, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 56924 + }, + { + "epoch": 0.56925, + "grad_norm": 1.5880993748628573, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 56925 + }, + { + "epoch": 0.56926, + "grad_norm": 1.484677191689772, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 56926 + }, + { + "epoch": 0.56927, + "grad_norm": 1.663046551228725, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 56927 + }, + { + "epoch": 0.56928, + "grad_norm": 1.313612045707248, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 56928 + }, + { + "epoch": 0.56929, + "grad_norm": 2.1063175917752184, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 56929 + }, + { + "epoch": 0.5693, + "grad_norm": 1.3966632973358324, + "learning_rate": 0.003, + "loss": 4.011, + "step": 56930 + }, + { + "epoch": 0.56931, + "grad_norm": 1.5916951500973062, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 56931 + }, + { + "epoch": 0.56932, + "grad_norm": 1.5104422476569173, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 56932 + }, + { + "epoch": 0.56933, + "grad_norm": 1.578668441579363, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 56933 + }, + { + "epoch": 0.56934, + "grad_norm": 1.4359133432664886, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 56934 + }, + { + "epoch": 0.56935, + "grad_norm": 1.5500140062048346, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 56935 + }, + { + "epoch": 0.56936, + "grad_norm": 1.692999855665226, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 56936 + }, + { + "epoch": 0.56937, + "grad_norm": 1.3784561735414311, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 56937 + }, + { + "epoch": 0.56938, + "grad_norm": 1.4598092519817254, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 56938 + }, + { + "epoch": 0.56939, + "grad_norm": 1.5972353376950958, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 56939 + }, + { + "epoch": 0.5694, + "grad_norm": 1.4227407241858803, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 56940 + }, + { + "epoch": 0.56941, + "grad_norm": 1.9913304494749695, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 56941 + }, + { + "epoch": 0.56942, + "grad_norm": 1.1509609820510471, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 56942 + }, + { + "epoch": 0.56943, + "grad_norm": 2.13941928268144, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 56943 + }, + { + "epoch": 0.56944, + "grad_norm": 1.4492116002869897, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 56944 + }, + { + "epoch": 0.56945, + "grad_norm": 1.5661324475415745, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 56945 + }, + { + "epoch": 0.56946, + "grad_norm": 1.779945948490474, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 56946 + }, + { + "epoch": 0.56947, + "grad_norm": 1.4909074473471673, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 56947 + }, + { + "epoch": 0.56948, + "grad_norm": 1.6003087355469652, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 56948 + }, + { + "epoch": 0.56949, + "grad_norm": 1.7928794559477446, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 56949 + }, + { + "epoch": 0.5695, + "grad_norm": 1.6714874464142722, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 56950 + }, + { + "epoch": 0.56951, + "grad_norm": 1.8432954114991589, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 56951 + }, + { + "epoch": 0.56952, + "grad_norm": 1.476223014162751, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 56952 + }, + { + "epoch": 0.56953, + "grad_norm": 1.6869970283592912, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 56953 + }, + { + "epoch": 0.56954, + "grad_norm": 1.4575821751764553, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 56954 + }, + { + "epoch": 0.56955, + "grad_norm": 1.7083792651312146, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 56955 + }, + { + "epoch": 0.56956, + "grad_norm": 1.320521598722143, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 56956 + }, + { + "epoch": 0.56957, + "grad_norm": 1.5464427427768639, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 56957 + }, + { + "epoch": 0.56958, + "grad_norm": 1.4105016944821398, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 56958 + }, + { + "epoch": 0.56959, + "grad_norm": 1.917968538542938, + "learning_rate": 0.003, + "loss": 4.069, + "step": 56959 + }, + { + "epoch": 0.5696, + "grad_norm": 1.4879666211107139, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 56960 + }, + { + "epoch": 0.56961, + "grad_norm": 1.598382061139946, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 56961 + }, + { + "epoch": 0.56962, + "grad_norm": 1.5686560968852268, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 56962 + }, + { + "epoch": 0.56963, + "grad_norm": 1.7647823973185723, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 56963 + }, + { + "epoch": 0.56964, + "grad_norm": 1.8137835249043424, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 56964 + }, + { + "epoch": 0.56965, + "grad_norm": 1.4597377587702076, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 56965 + }, + { + "epoch": 0.56966, + "grad_norm": 1.868595620816914, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 56966 + }, + { + "epoch": 0.56967, + "grad_norm": 1.4798735082549825, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 56967 + }, + { + "epoch": 0.56968, + "grad_norm": 1.5910299142756796, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 56968 + }, + { + "epoch": 0.56969, + "grad_norm": 1.6496316946903156, + "learning_rate": 0.003, + "loss": 4.014, + "step": 56969 + }, + { + "epoch": 0.5697, + "grad_norm": 1.5220572299471564, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 56970 + }, + { + "epoch": 0.56971, + "grad_norm": 1.5430191070362234, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 56971 + }, + { + "epoch": 0.56972, + "grad_norm": 1.4095245378178896, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 56972 + }, + { + "epoch": 0.56973, + "grad_norm": 1.6196230590457925, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 56973 + }, + { + "epoch": 0.56974, + "grad_norm": 1.6315930311389129, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 56974 + }, + { + "epoch": 0.56975, + "grad_norm": 1.9995058300895985, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 56975 + }, + { + "epoch": 0.56976, + "grad_norm": 1.515915515872944, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 56976 + }, + { + "epoch": 0.56977, + "grad_norm": 1.608814245555847, + "learning_rate": 0.003, + "loss": 4.046, + "step": 56977 + }, + { + "epoch": 0.56978, + "grad_norm": 1.5939797461268082, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 56978 + }, + { + "epoch": 0.56979, + "grad_norm": 1.5701925550544824, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 56979 + }, + { + "epoch": 0.5698, + "grad_norm": 1.6485049219987629, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 56980 + }, + { + "epoch": 0.56981, + "grad_norm": 1.4255093743905496, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 56981 + }, + { + "epoch": 0.56982, + "grad_norm": 1.478459452029668, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 56982 + }, + { + "epoch": 0.56983, + "grad_norm": 1.6358682564005322, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 56983 + }, + { + "epoch": 0.56984, + "grad_norm": 1.461570673106115, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 56984 + }, + { + "epoch": 0.56985, + "grad_norm": 2.0772331220540394, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 56985 + }, + { + "epoch": 0.56986, + "grad_norm": 1.2651121741675908, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 56986 + }, + { + "epoch": 0.56987, + "grad_norm": 2.068986282177144, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 56987 + }, + { + "epoch": 0.56988, + "grad_norm": 1.3557278764762943, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 56988 + }, + { + "epoch": 0.56989, + "grad_norm": 1.6577950207305598, + "learning_rate": 0.003, + "loss": 4.02, + "step": 56989 + }, + { + "epoch": 0.5699, + "grad_norm": 1.74707218927372, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 56990 + }, + { + "epoch": 0.56991, + "grad_norm": 1.4889524144575803, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 56991 + }, + { + "epoch": 0.56992, + "grad_norm": 1.7230828756809828, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 56992 + }, + { + "epoch": 0.56993, + "grad_norm": 1.5274298274723557, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 56993 + }, + { + "epoch": 0.56994, + "grad_norm": 1.7969419422099069, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 56994 + }, + { + "epoch": 0.56995, + "grad_norm": 1.6637723480376145, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 56995 + }, + { + "epoch": 0.56996, + "grad_norm": 1.4440571980737569, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 56996 + }, + { + "epoch": 0.56997, + "grad_norm": 1.5556424904410904, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 56997 + }, + { + "epoch": 0.56998, + "grad_norm": 1.3390358688841941, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 56998 + }, + { + "epoch": 0.56999, + "grad_norm": 1.9585270045954086, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 56999 + }, + { + "epoch": 0.57, + "grad_norm": 1.4766508456793286, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 57000 + }, + { + "epoch": 0.57001, + "grad_norm": 1.9241042481708046, + "learning_rate": 0.003, + "loss": 4.026, + "step": 57001 + }, + { + "epoch": 0.57002, + "grad_norm": 1.2854076312403824, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 57002 + }, + { + "epoch": 0.57003, + "grad_norm": 1.6520802565614792, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 57003 + }, + { + "epoch": 0.57004, + "grad_norm": 1.3544842916761899, + "learning_rate": 0.003, + "loss": 4.056, + "step": 57004 + }, + { + "epoch": 0.57005, + "grad_norm": 1.6287689148866775, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 57005 + }, + { + "epoch": 0.57006, + "grad_norm": 1.5986301512262586, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 57006 + }, + { + "epoch": 0.57007, + "grad_norm": 1.7115156241528857, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 57007 + }, + { + "epoch": 0.57008, + "grad_norm": 1.6348399795375668, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 57008 + }, + { + "epoch": 0.57009, + "grad_norm": 1.5019823332591737, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 57009 + }, + { + "epoch": 0.5701, + "grad_norm": 1.4228196676323834, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 57010 + }, + { + "epoch": 0.57011, + "grad_norm": 1.5763126446305402, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 57011 + }, + { + "epoch": 0.57012, + "grad_norm": 1.3432228399886117, + "learning_rate": 0.003, + "loss": 4.045, + "step": 57012 + }, + { + "epoch": 0.57013, + "grad_norm": 1.7682073292008056, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 57013 + }, + { + "epoch": 0.57014, + "grad_norm": 1.3767350479245062, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 57014 + }, + { + "epoch": 0.57015, + "grad_norm": 1.8900739178237727, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 57015 + }, + { + "epoch": 0.57016, + "grad_norm": 1.4782617099740225, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 57016 + }, + { + "epoch": 0.57017, + "grad_norm": 1.669892451595112, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 57017 + }, + { + "epoch": 0.57018, + "grad_norm": 1.551535112853622, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 57018 + }, + { + "epoch": 0.57019, + "grad_norm": 1.412903350242172, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 57019 + }, + { + "epoch": 0.5702, + "grad_norm": 1.5293124620144718, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 57020 + }, + { + "epoch": 0.57021, + "grad_norm": 1.640501953439825, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 57021 + }, + { + "epoch": 0.57022, + "grad_norm": 1.3298992868847859, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 57022 + }, + { + "epoch": 0.57023, + "grad_norm": 1.7754038064442974, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 57023 + }, + { + "epoch": 0.57024, + "grad_norm": 1.483674916364657, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 57024 + }, + { + "epoch": 0.57025, + "grad_norm": 1.9323024480184081, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 57025 + }, + { + "epoch": 0.57026, + "grad_norm": 1.555250935676317, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 57026 + }, + { + "epoch": 0.57027, + "grad_norm": 1.5835886754612996, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 57027 + }, + { + "epoch": 0.57028, + "grad_norm": 1.398691280143866, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 57028 + }, + { + "epoch": 0.57029, + "grad_norm": 2.127632997426885, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 57029 + }, + { + "epoch": 0.5703, + "grad_norm": 1.4094379913705053, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 57030 + }, + { + "epoch": 0.57031, + "grad_norm": 2.071380988222371, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 57031 + }, + { + "epoch": 0.57032, + "grad_norm": 1.3983239804071932, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57032 + }, + { + "epoch": 0.57033, + "grad_norm": 1.6360209335666231, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 57033 + }, + { + "epoch": 0.57034, + "grad_norm": 1.8332691791672857, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 57034 + }, + { + "epoch": 0.57035, + "grad_norm": 1.3754376392177128, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 57035 + }, + { + "epoch": 0.57036, + "grad_norm": 1.5221824811970526, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 57036 + }, + { + "epoch": 0.57037, + "grad_norm": 1.4119469736842507, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 57037 + }, + { + "epoch": 0.57038, + "grad_norm": 1.8424780955223992, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 57038 + }, + { + "epoch": 0.57039, + "grad_norm": 1.717314352596819, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 57039 + }, + { + "epoch": 0.5704, + "grad_norm": 1.8214315887033667, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 57040 + }, + { + "epoch": 0.57041, + "grad_norm": 1.5754129697006156, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 57041 + }, + { + "epoch": 0.57042, + "grad_norm": 1.6218999611137235, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 57042 + }, + { + "epoch": 0.57043, + "grad_norm": 1.4114895143855115, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 57043 + }, + { + "epoch": 0.57044, + "grad_norm": 1.5611119633708852, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 57044 + }, + { + "epoch": 0.57045, + "grad_norm": 1.1668967826924914, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 57045 + }, + { + "epoch": 0.57046, + "grad_norm": 1.8111962912231228, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 57046 + }, + { + "epoch": 0.57047, + "grad_norm": 1.2162355451080482, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 57047 + }, + { + "epoch": 0.57048, + "grad_norm": 1.6760667649436258, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 57048 + }, + { + "epoch": 0.57049, + "grad_norm": 1.405156161249089, + "learning_rate": 0.003, + "loss": 4.065, + "step": 57049 + }, + { + "epoch": 0.5705, + "grad_norm": 1.65094689790832, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 57050 + }, + { + "epoch": 0.57051, + "grad_norm": 1.730308119471349, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 57051 + }, + { + "epoch": 0.57052, + "grad_norm": 1.9682843303561168, + "learning_rate": 0.003, + "loss": 4.051, + "step": 57052 + }, + { + "epoch": 0.57053, + "grad_norm": 1.5667705749061327, + "learning_rate": 0.003, + "loss": 4.036, + "step": 57053 + }, + { + "epoch": 0.57054, + "grad_norm": 1.6587909846244981, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57054 + }, + { + "epoch": 0.57055, + "grad_norm": 1.3876769603666423, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 57055 + }, + { + "epoch": 0.57056, + "grad_norm": 1.6905977320397292, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 57056 + }, + { + "epoch": 0.57057, + "grad_norm": 1.4112280765866203, + "learning_rate": 0.003, + "loss": 4.029, + "step": 57057 + }, + { + "epoch": 0.57058, + "grad_norm": 1.7732199917653209, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 57058 + }, + { + "epoch": 0.57059, + "grad_norm": 1.4378333556783058, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 57059 + }, + { + "epoch": 0.5706, + "grad_norm": 2.033775740310553, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 57060 + }, + { + "epoch": 0.57061, + "grad_norm": 1.4058498538854654, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 57061 + }, + { + "epoch": 0.57062, + "grad_norm": 2.100371988820275, + "learning_rate": 0.003, + "loss": 4.017, + "step": 57062 + }, + { + "epoch": 0.57063, + "grad_norm": 1.4497939296109104, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57063 + }, + { + "epoch": 0.57064, + "grad_norm": 1.5342541694330014, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 57064 + }, + { + "epoch": 0.57065, + "grad_norm": 1.5906759294725954, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 57065 + }, + { + "epoch": 0.57066, + "grad_norm": 1.441447544768296, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57066 + }, + { + "epoch": 0.57067, + "grad_norm": 1.4863374615614398, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 57067 + }, + { + "epoch": 0.57068, + "grad_norm": 1.6994625520873161, + "learning_rate": 0.003, + "loss": 4.033, + "step": 57068 + }, + { + "epoch": 0.57069, + "grad_norm": 1.6128484996478387, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 57069 + }, + { + "epoch": 0.5707, + "grad_norm": 1.582293481714903, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57070 + }, + { + "epoch": 0.57071, + "grad_norm": 1.6175863450510521, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 57071 + }, + { + "epoch": 0.57072, + "grad_norm": 1.6746037988242883, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 57072 + }, + { + "epoch": 0.57073, + "grad_norm": 1.6508527104059338, + "learning_rate": 0.003, + "loss": 4.016, + "step": 57073 + }, + { + "epoch": 0.57074, + "grad_norm": 1.3550700240990257, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 57074 + }, + { + "epoch": 0.57075, + "grad_norm": 1.6251408505844211, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 57075 + }, + { + "epoch": 0.57076, + "grad_norm": 1.3906122738903106, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 57076 + }, + { + "epoch": 0.57077, + "grad_norm": 2.022679665011895, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 57077 + }, + { + "epoch": 0.57078, + "grad_norm": 1.5270612176840122, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 57078 + }, + { + "epoch": 0.57079, + "grad_norm": 1.9080307255194886, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 57079 + }, + { + "epoch": 0.5708, + "grad_norm": 1.8464990874009537, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 57080 + }, + { + "epoch": 0.57081, + "grad_norm": 1.5069715333343694, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 57081 + }, + { + "epoch": 0.57082, + "grad_norm": 1.6934996623842122, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 57082 + }, + { + "epoch": 0.57083, + "grad_norm": 1.504010047657534, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 57083 + }, + { + "epoch": 0.57084, + "grad_norm": 1.5336390185770892, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 57084 + }, + { + "epoch": 0.57085, + "grad_norm": 1.9545320936037596, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 57085 + }, + { + "epoch": 0.57086, + "grad_norm": 1.7569402248813577, + "learning_rate": 0.003, + "loss": 4.043, + "step": 57086 + }, + { + "epoch": 0.57087, + "grad_norm": 1.5009087368121077, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 57087 + }, + { + "epoch": 0.57088, + "grad_norm": 1.4972387215306382, + "learning_rate": 0.003, + "loss": 4.036, + "step": 57088 + }, + { + "epoch": 0.57089, + "grad_norm": 1.7936826125957466, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 57089 + }, + { + "epoch": 0.5709, + "grad_norm": 1.578411530371694, + "learning_rate": 0.003, + "loss": 4.016, + "step": 57090 + }, + { + "epoch": 0.57091, + "grad_norm": 1.5319019166946923, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 57091 + }, + { + "epoch": 0.57092, + "grad_norm": 1.526694403848885, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 57092 + }, + { + "epoch": 0.57093, + "grad_norm": 1.6099441692200072, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 57093 + }, + { + "epoch": 0.57094, + "grad_norm": 1.426900701689898, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 57094 + }, + { + "epoch": 0.57095, + "grad_norm": 1.638805719301913, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 57095 + }, + { + "epoch": 0.57096, + "grad_norm": 1.4158715918689844, + "learning_rate": 0.003, + "loss": 4.003, + "step": 57096 + }, + { + "epoch": 0.57097, + "grad_norm": 1.7088246998044183, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 57097 + }, + { + "epoch": 0.57098, + "grad_norm": 1.2586210976397523, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 57098 + }, + { + "epoch": 0.57099, + "grad_norm": 1.7392422076758782, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 57099 + }, + { + "epoch": 0.571, + "grad_norm": 1.309057574633611, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 57100 + }, + { + "epoch": 0.57101, + "grad_norm": 1.8370064026924375, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 57101 + }, + { + "epoch": 0.57102, + "grad_norm": 1.5199478195066183, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 57102 + }, + { + "epoch": 0.57103, + "grad_norm": 1.780912641569713, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 57103 + }, + { + "epoch": 0.57104, + "grad_norm": 1.4217532956027759, + "learning_rate": 0.003, + "loss": 4.013, + "step": 57104 + }, + { + "epoch": 0.57105, + "grad_norm": 1.5772331764558993, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 57105 + }, + { + "epoch": 0.57106, + "grad_norm": 1.6947586689960832, + "learning_rate": 0.003, + "loss": 4.034, + "step": 57106 + }, + { + "epoch": 0.57107, + "grad_norm": 1.3306720540166188, + "learning_rate": 0.003, + "loss": 4.02, + "step": 57107 + }, + { + "epoch": 0.57108, + "grad_norm": 1.6424830643258799, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 57108 + }, + { + "epoch": 0.57109, + "grad_norm": 1.4210054727503496, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 57109 + }, + { + "epoch": 0.5711, + "grad_norm": 1.5632232641194705, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 57110 + }, + { + "epoch": 0.57111, + "grad_norm": 1.740279373675073, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57111 + }, + { + "epoch": 0.57112, + "grad_norm": 1.3038766346327142, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 57112 + }, + { + "epoch": 0.57113, + "grad_norm": 1.8767470362645207, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 57113 + }, + { + "epoch": 0.57114, + "grad_norm": 1.6611205111261083, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 57114 + }, + { + "epoch": 0.57115, + "grad_norm": 1.6990484098948102, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 57115 + }, + { + "epoch": 0.57116, + "grad_norm": 1.5603228806482143, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 57116 + }, + { + "epoch": 0.57117, + "grad_norm": 1.3524552676166977, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 57117 + }, + { + "epoch": 0.57118, + "grad_norm": 2.119498293401525, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 57118 + }, + { + "epoch": 0.57119, + "grad_norm": 1.70453242067724, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 57119 + }, + { + "epoch": 0.5712, + "grad_norm": 1.717595066410197, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 57120 + }, + { + "epoch": 0.57121, + "grad_norm": 1.463202048490307, + "learning_rate": 0.003, + "loss": 4.01, + "step": 57121 + }, + { + "epoch": 0.57122, + "grad_norm": 1.5906426977629478, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 57122 + }, + { + "epoch": 0.57123, + "grad_norm": 1.6452939190665605, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 57123 + }, + { + "epoch": 0.57124, + "grad_norm": 1.4265224860186674, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 57124 + }, + { + "epoch": 0.57125, + "grad_norm": 1.5683057718375621, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 57125 + }, + { + "epoch": 0.57126, + "grad_norm": 1.4238649756816673, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 57126 + }, + { + "epoch": 0.57127, + "grad_norm": 1.623607033375514, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 57127 + }, + { + "epoch": 0.57128, + "grad_norm": 1.3756044723602996, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 57128 + }, + { + "epoch": 0.57129, + "grad_norm": 1.6158756675679284, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 57129 + }, + { + "epoch": 0.5713, + "grad_norm": 1.4817543022393407, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 57130 + }, + { + "epoch": 0.57131, + "grad_norm": 1.825777928927952, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 57131 + }, + { + "epoch": 0.57132, + "grad_norm": 1.6158868593125066, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 57132 + }, + { + "epoch": 0.57133, + "grad_norm": 1.283880620406461, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57133 + }, + { + "epoch": 0.57134, + "grad_norm": 1.6014476693656454, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 57134 + }, + { + "epoch": 0.57135, + "grad_norm": 1.6047671929306024, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 57135 + }, + { + "epoch": 0.57136, + "grad_norm": 1.9401430119966814, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 57136 + }, + { + "epoch": 0.57137, + "grad_norm": 1.451273552817913, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 57137 + }, + { + "epoch": 0.57138, + "grad_norm": 1.748795488631934, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 57138 + }, + { + "epoch": 0.57139, + "grad_norm": 1.703139796328794, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 57139 + }, + { + "epoch": 0.5714, + "grad_norm": 1.4996312292243048, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 57140 + }, + { + "epoch": 0.57141, + "grad_norm": 1.4846347427540054, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57141 + }, + { + "epoch": 0.57142, + "grad_norm": 1.6304775823641686, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 57142 + }, + { + "epoch": 0.57143, + "grad_norm": 1.5742413059887752, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 57143 + }, + { + "epoch": 0.57144, + "grad_norm": 1.6199159709899074, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 57144 + }, + { + "epoch": 0.57145, + "grad_norm": 1.4427026273840537, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 57145 + }, + { + "epoch": 0.57146, + "grad_norm": 1.5601323135123735, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 57146 + }, + { + "epoch": 0.57147, + "grad_norm": 1.599895170187323, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 57147 + }, + { + "epoch": 0.57148, + "grad_norm": 1.4772005077729389, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 57148 + }, + { + "epoch": 0.57149, + "grad_norm": 1.637144031019793, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 57149 + }, + { + "epoch": 0.5715, + "grad_norm": 1.8014408490021596, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 57150 + }, + { + "epoch": 0.57151, + "grad_norm": 1.6647516765864498, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 57151 + }, + { + "epoch": 0.57152, + "grad_norm": 1.685817293486327, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 57152 + }, + { + "epoch": 0.57153, + "grad_norm": 1.6324303111748775, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 57153 + }, + { + "epoch": 0.57154, + "grad_norm": 1.5663793151669811, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 57154 + }, + { + "epoch": 0.57155, + "grad_norm": 1.7453564361321527, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 57155 + }, + { + "epoch": 0.57156, + "grad_norm": 1.7294600459812168, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 57156 + }, + { + "epoch": 0.57157, + "grad_norm": 1.3063255719234115, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 57157 + }, + { + "epoch": 0.57158, + "grad_norm": 1.6750284668357514, + "learning_rate": 0.003, + "loss": 4.024, + "step": 57158 + }, + { + "epoch": 0.57159, + "grad_norm": 1.547066364363741, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 57159 + }, + { + "epoch": 0.5716, + "grad_norm": 1.5076704415805975, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 57160 + }, + { + "epoch": 0.57161, + "grad_norm": 1.4459957128361534, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 57161 + }, + { + "epoch": 0.57162, + "grad_norm": 1.5198413332424794, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 57162 + }, + { + "epoch": 0.57163, + "grad_norm": 1.9038423781823082, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 57163 + }, + { + "epoch": 0.57164, + "grad_norm": 1.191538926336113, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 57164 + }, + { + "epoch": 0.57165, + "grad_norm": 1.9094641634866543, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 57165 + }, + { + "epoch": 0.57166, + "grad_norm": 1.404720574935563, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 57166 + }, + { + "epoch": 0.57167, + "grad_norm": 1.635382769837572, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 57167 + }, + { + "epoch": 0.57168, + "grad_norm": 1.3916282794854655, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 57168 + }, + { + "epoch": 0.57169, + "grad_norm": 1.574020105965009, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 57169 + }, + { + "epoch": 0.5717, + "grad_norm": 1.6812851091534988, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 57170 + }, + { + "epoch": 0.57171, + "grad_norm": 1.3801340055081057, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 57171 + }, + { + "epoch": 0.57172, + "grad_norm": 1.7157217508561033, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 57172 + }, + { + "epoch": 0.57173, + "grad_norm": 1.7274773058944735, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 57173 + }, + { + "epoch": 0.57174, + "grad_norm": 1.8192710135310353, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 57174 + }, + { + "epoch": 0.57175, + "grad_norm": 1.573442498725844, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 57175 + }, + { + "epoch": 0.57176, + "grad_norm": 1.7845524350298618, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 57176 + }, + { + "epoch": 0.57177, + "grad_norm": 1.6006265664243857, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 57177 + }, + { + "epoch": 0.57178, + "grad_norm": 1.2975810082912944, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 57178 + }, + { + "epoch": 0.57179, + "grad_norm": 1.8464917320845606, + "learning_rate": 0.003, + "loss": 4.037, + "step": 57179 + }, + { + "epoch": 0.5718, + "grad_norm": 1.2115405206724024, + "learning_rate": 0.003, + "loss": 4.035, + "step": 57180 + }, + { + "epoch": 0.57181, + "grad_norm": 1.6897272267694377, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 57181 + }, + { + "epoch": 0.57182, + "grad_norm": 1.6047173541134427, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 57182 + }, + { + "epoch": 0.57183, + "grad_norm": 1.4110026342444026, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 57183 + }, + { + "epoch": 0.57184, + "grad_norm": 1.7708207464158792, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 57184 + }, + { + "epoch": 0.57185, + "grad_norm": 1.6631835407484359, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 57185 + }, + { + "epoch": 0.57186, + "grad_norm": 1.6916769687387585, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 57186 + }, + { + "epoch": 0.57187, + "grad_norm": 1.5818697924007197, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 57187 + }, + { + "epoch": 0.57188, + "grad_norm": 1.757208417630211, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 57188 + }, + { + "epoch": 0.57189, + "grad_norm": 1.5202046368153495, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 57189 + }, + { + "epoch": 0.5719, + "grad_norm": 1.5177863099082205, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 57190 + }, + { + "epoch": 0.57191, + "grad_norm": 1.7258221116997239, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57191 + }, + { + "epoch": 0.57192, + "grad_norm": 1.6624259219589281, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 57192 + }, + { + "epoch": 0.57193, + "grad_norm": 1.3580701729897127, + "learning_rate": 0.003, + "loss": 4.025, + "step": 57193 + }, + { + "epoch": 0.57194, + "grad_norm": 1.665822329412945, + "learning_rate": 0.003, + "loss": 4.024, + "step": 57194 + }, + { + "epoch": 0.57195, + "grad_norm": 1.412103973058782, + "learning_rate": 0.003, + "loss": 4.032, + "step": 57195 + }, + { + "epoch": 0.57196, + "grad_norm": 1.8016284478927707, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 57196 + }, + { + "epoch": 0.57197, + "grad_norm": 1.6141495816203826, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 57197 + }, + { + "epoch": 0.57198, + "grad_norm": 1.7876816658099148, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 57198 + }, + { + "epoch": 0.57199, + "grad_norm": 1.5265923804223802, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 57199 + }, + { + "epoch": 0.572, + "grad_norm": 1.723650264822254, + "learning_rate": 0.003, + "loss": 4.014, + "step": 57200 + }, + { + "epoch": 0.57201, + "grad_norm": 1.4736493062008513, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 57201 + }, + { + "epoch": 0.57202, + "grad_norm": 1.388500221344463, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 57202 + }, + { + "epoch": 0.57203, + "grad_norm": 1.6639272475916875, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 57203 + }, + { + "epoch": 0.57204, + "grad_norm": 1.6839749654089562, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 57204 + }, + { + "epoch": 0.57205, + "grad_norm": 1.6478691055715857, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 57205 + }, + { + "epoch": 0.57206, + "grad_norm": 1.6264904071659372, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 57206 + }, + { + "epoch": 0.57207, + "grad_norm": 1.7398317577252613, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 57207 + }, + { + "epoch": 0.57208, + "grad_norm": 1.1885491198794182, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 57208 + }, + { + "epoch": 0.57209, + "grad_norm": 1.7891644479738598, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 57209 + }, + { + "epoch": 0.5721, + "grad_norm": 1.294544327910033, + "learning_rate": 0.003, + "loss": 4.028, + "step": 57210 + }, + { + "epoch": 0.57211, + "grad_norm": 1.4969714696186716, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 57211 + }, + { + "epoch": 0.57212, + "grad_norm": 1.5884263405295942, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 57212 + }, + { + "epoch": 0.57213, + "grad_norm": 1.4708142391584487, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 57213 + }, + { + "epoch": 0.57214, + "grad_norm": 2.1660587899556027, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 57214 + }, + { + "epoch": 0.57215, + "grad_norm": 1.6177130185526982, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 57215 + }, + { + "epoch": 0.57216, + "grad_norm": 1.4562402917031507, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 57216 + }, + { + "epoch": 0.57217, + "grad_norm": 1.7893119536888038, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 57217 + }, + { + "epoch": 0.57218, + "grad_norm": 1.3839325305816308, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 57218 + }, + { + "epoch": 0.57219, + "grad_norm": 1.5458791673558923, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 57219 + }, + { + "epoch": 0.5722, + "grad_norm": 1.7288716364298549, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 57220 + }, + { + "epoch": 0.57221, + "grad_norm": 1.3848468464445347, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 57221 + }, + { + "epoch": 0.57222, + "grad_norm": 1.5545457951265684, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 57222 + }, + { + "epoch": 0.57223, + "grad_norm": 1.329076926035692, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57223 + }, + { + "epoch": 0.57224, + "grad_norm": 1.7174995832481548, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57224 + }, + { + "epoch": 0.57225, + "grad_norm": 1.3101006614519426, + "learning_rate": 0.003, + "loss": 4.058, + "step": 57225 + }, + { + "epoch": 0.57226, + "grad_norm": 1.5917205397424794, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 57226 + }, + { + "epoch": 0.57227, + "grad_norm": 1.6387807263531078, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 57227 + }, + { + "epoch": 0.57228, + "grad_norm": 1.4910446068252439, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 57228 + }, + { + "epoch": 0.57229, + "grad_norm": 1.5161445953751078, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 57229 + }, + { + "epoch": 0.5723, + "grad_norm": 1.9134429762659644, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 57230 + }, + { + "epoch": 0.57231, + "grad_norm": 1.8886471958058784, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 57231 + }, + { + "epoch": 0.57232, + "grad_norm": 1.4206350850793947, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 57232 + }, + { + "epoch": 0.57233, + "grad_norm": 1.9139754332677212, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 57233 + }, + { + "epoch": 0.57234, + "grad_norm": 1.587724548478216, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 57234 + }, + { + "epoch": 0.57235, + "grad_norm": 1.5560843038067842, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 57235 + }, + { + "epoch": 0.57236, + "grad_norm": 1.6691161034458284, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 57236 + }, + { + "epoch": 0.57237, + "grad_norm": 1.6930321580087833, + "learning_rate": 0.003, + "loss": 4.041, + "step": 57237 + }, + { + "epoch": 0.57238, + "grad_norm": 1.3402084383143642, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 57238 + }, + { + "epoch": 0.57239, + "grad_norm": 1.9289528118815784, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 57239 + }, + { + "epoch": 0.5724, + "grad_norm": 1.2673967292611572, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 57240 + }, + { + "epoch": 0.57241, + "grad_norm": 1.8521258468542299, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 57241 + }, + { + "epoch": 0.57242, + "grad_norm": 1.3897251539841111, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 57242 + }, + { + "epoch": 0.57243, + "grad_norm": 1.449935307144272, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 57243 + }, + { + "epoch": 0.57244, + "grad_norm": 1.6953127203550822, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 57244 + }, + { + "epoch": 0.57245, + "grad_norm": 1.3932213675575809, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 57245 + }, + { + "epoch": 0.57246, + "grad_norm": 1.5119248847968778, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 57246 + }, + { + "epoch": 0.57247, + "grad_norm": 1.644277237243183, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 57247 + }, + { + "epoch": 0.57248, + "grad_norm": 2.1696318371935974, + "learning_rate": 0.003, + "loss": 4.052, + "step": 57248 + }, + { + "epoch": 0.57249, + "grad_norm": 1.444032136441053, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 57249 + }, + { + "epoch": 0.5725, + "grad_norm": 1.5230000489834385, + "learning_rate": 0.003, + "loss": 4.009, + "step": 57250 + }, + { + "epoch": 0.57251, + "grad_norm": 1.799557663001347, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 57251 + }, + { + "epoch": 0.57252, + "grad_norm": 1.6224567591179075, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 57252 + }, + { + "epoch": 0.57253, + "grad_norm": 1.7407548082976045, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 57253 + }, + { + "epoch": 0.57254, + "grad_norm": 1.7262010645816588, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 57254 + }, + { + "epoch": 0.57255, + "grad_norm": 1.370026050141586, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 57255 + }, + { + "epoch": 0.57256, + "grad_norm": 1.5502692825070958, + "learning_rate": 0.003, + "loss": 4.042, + "step": 57256 + }, + { + "epoch": 0.57257, + "grad_norm": 1.3589245413939273, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 57257 + }, + { + "epoch": 0.57258, + "grad_norm": 1.7381585195398916, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 57258 + }, + { + "epoch": 0.57259, + "grad_norm": 1.2863471417849632, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 57259 + }, + { + "epoch": 0.5726, + "grad_norm": 1.7250498461340917, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 57260 + }, + { + "epoch": 0.57261, + "grad_norm": 1.2981685192043688, + "learning_rate": 0.003, + "loss": 4.027, + "step": 57261 + }, + { + "epoch": 0.57262, + "grad_norm": 1.7205156119836738, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 57262 + }, + { + "epoch": 0.57263, + "grad_norm": 1.4518887713380213, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 57263 + }, + { + "epoch": 0.57264, + "grad_norm": 1.5224742046516282, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57264 + }, + { + "epoch": 0.57265, + "grad_norm": 1.8530837098740465, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 57265 + }, + { + "epoch": 0.57266, + "grad_norm": 1.654867653157045, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 57266 + }, + { + "epoch": 0.57267, + "grad_norm": 1.6019105627948282, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 57267 + }, + { + "epoch": 0.57268, + "grad_norm": 1.551027442453529, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 57268 + }, + { + "epoch": 0.57269, + "grad_norm": 1.8469999424871841, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 57269 + }, + { + "epoch": 0.5727, + "grad_norm": 1.342511238140077, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 57270 + }, + { + "epoch": 0.57271, + "grad_norm": 1.9117407272152316, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 57271 + }, + { + "epoch": 0.57272, + "grad_norm": 1.5215436969658247, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 57272 + }, + { + "epoch": 0.57273, + "grad_norm": 1.6639442673005054, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 57273 + }, + { + "epoch": 0.57274, + "grad_norm": 1.2811295053824536, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 57274 + }, + { + "epoch": 0.57275, + "grad_norm": 1.7361472895950785, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 57275 + }, + { + "epoch": 0.57276, + "grad_norm": 1.5293339556940326, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 57276 + }, + { + "epoch": 0.57277, + "grad_norm": 1.5293324555318104, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 57277 + }, + { + "epoch": 0.57278, + "grad_norm": 1.6605793063526946, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 57278 + }, + { + "epoch": 0.57279, + "grad_norm": 1.8123832902762078, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 57279 + }, + { + "epoch": 0.5728, + "grad_norm": 1.5879818264761216, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 57280 + }, + { + "epoch": 0.57281, + "grad_norm": 1.5568373869257055, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 57281 + }, + { + "epoch": 0.57282, + "grad_norm": 1.7281747495616882, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 57282 + }, + { + "epoch": 0.57283, + "grad_norm": 1.5865242773862869, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 57283 + }, + { + "epoch": 0.57284, + "grad_norm": 1.38446148920207, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 57284 + }, + { + "epoch": 0.57285, + "grad_norm": 1.4403842074096813, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 57285 + }, + { + "epoch": 0.57286, + "grad_norm": 1.624946069246399, + "learning_rate": 0.003, + "loss": 4.027, + "step": 57286 + }, + { + "epoch": 0.57287, + "grad_norm": 1.4244530105067745, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 57287 + }, + { + "epoch": 0.57288, + "grad_norm": 1.593985022604783, + "learning_rate": 0.003, + "loss": 4.024, + "step": 57288 + }, + { + "epoch": 0.57289, + "grad_norm": 1.4878436307622607, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57289 + }, + { + "epoch": 0.5729, + "grad_norm": 1.675652839777338, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 57290 + }, + { + "epoch": 0.57291, + "grad_norm": 1.7103168687108015, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 57291 + }, + { + "epoch": 0.57292, + "grad_norm": 1.6780226416444455, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 57292 + }, + { + "epoch": 0.57293, + "grad_norm": 1.5618741048806422, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 57293 + }, + { + "epoch": 0.57294, + "grad_norm": 1.6407722192237957, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 57294 + }, + { + "epoch": 0.57295, + "grad_norm": 1.643470407334096, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 57295 + }, + { + "epoch": 0.57296, + "grad_norm": 1.494360677509016, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 57296 + }, + { + "epoch": 0.57297, + "grad_norm": 1.650374925332077, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 57297 + }, + { + "epoch": 0.57298, + "grad_norm": 1.4651287955510353, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 57298 + }, + { + "epoch": 0.57299, + "grad_norm": 1.5570732286863547, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 57299 + }, + { + "epoch": 0.573, + "grad_norm": 1.6253508016854894, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 57300 + }, + { + "epoch": 0.57301, + "grad_norm": 1.510264801160267, + "learning_rate": 0.003, + "loss": 4.058, + "step": 57301 + }, + { + "epoch": 0.57302, + "grad_norm": 1.4947607795549165, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 57302 + }, + { + "epoch": 0.57303, + "grad_norm": 1.5423688060425913, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 57303 + }, + { + "epoch": 0.57304, + "grad_norm": 1.5714140322470354, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 57304 + }, + { + "epoch": 0.57305, + "grad_norm": 1.613782463687738, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 57305 + }, + { + "epoch": 0.57306, + "grad_norm": 1.6309076543367707, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 57306 + }, + { + "epoch": 0.57307, + "grad_norm": 1.4779153837186687, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 57307 + }, + { + "epoch": 0.57308, + "grad_norm": 1.5475185075454232, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 57308 + }, + { + "epoch": 0.57309, + "grad_norm": 1.7877853251821654, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 57309 + }, + { + "epoch": 0.5731, + "grad_norm": 1.6095931716625516, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 57310 + }, + { + "epoch": 0.57311, + "grad_norm": 1.4743346388911536, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 57311 + }, + { + "epoch": 0.57312, + "grad_norm": 1.780033160650536, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 57312 + }, + { + "epoch": 0.57313, + "grad_norm": 1.324590662607226, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 57313 + }, + { + "epoch": 0.57314, + "grad_norm": 1.8695106330213012, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 57314 + }, + { + "epoch": 0.57315, + "grad_norm": 1.836172546732326, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 57315 + }, + { + "epoch": 0.57316, + "grad_norm": 1.8191528979954188, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 57316 + }, + { + "epoch": 0.57317, + "grad_norm": 1.4608121156190315, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 57317 + }, + { + "epoch": 0.57318, + "grad_norm": 1.4820690214867271, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 57318 + }, + { + "epoch": 0.57319, + "grad_norm": 1.854486120719843, + "learning_rate": 0.003, + "loss": 4.02, + "step": 57319 + }, + { + "epoch": 0.5732, + "grad_norm": 1.4398439133936887, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 57320 + }, + { + "epoch": 0.57321, + "grad_norm": 1.5321657428089297, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 57321 + }, + { + "epoch": 0.57322, + "grad_norm": 1.5867128147659857, + "learning_rate": 0.003, + "loss": 4.041, + "step": 57322 + }, + { + "epoch": 0.57323, + "grad_norm": 1.3604563914624705, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 57323 + }, + { + "epoch": 0.57324, + "grad_norm": 1.5212112793460035, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 57324 + }, + { + "epoch": 0.57325, + "grad_norm": 1.409621713594189, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 57325 + }, + { + "epoch": 0.57326, + "grad_norm": 1.569163810794593, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 57326 + }, + { + "epoch": 0.57327, + "grad_norm": 1.6163478937085394, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 57327 + }, + { + "epoch": 0.57328, + "grad_norm": 1.7432877372066677, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 57328 + }, + { + "epoch": 0.57329, + "grad_norm": 1.8773648271426717, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 57329 + }, + { + "epoch": 0.5733, + "grad_norm": 1.8182075168965808, + "learning_rate": 0.003, + "loss": 4.029, + "step": 57330 + }, + { + "epoch": 0.57331, + "grad_norm": 1.4808598717221768, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 57331 + }, + { + "epoch": 0.57332, + "grad_norm": 1.5642419452093599, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 57332 + }, + { + "epoch": 0.57333, + "grad_norm": 1.716161315282092, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 57333 + }, + { + "epoch": 0.57334, + "grad_norm": 1.5555056592804792, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 57334 + }, + { + "epoch": 0.57335, + "grad_norm": 1.7107938460165144, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 57335 + }, + { + "epoch": 0.57336, + "grad_norm": 1.4138003364681624, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 57336 + }, + { + "epoch": 0.57337, + "grad_norm": 1.6078407694589716, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 57337 + }, + { + "epoch": 0.57338, + "grad_norm": 1.574602984275391, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 57338 + }, + { + "epoch": 0.57339, + "grad_norm": 1.4958374482955183, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 57339 + }, + { + "epoch": 0.5734, + "grad_norm": 1.7538307547995922, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 57340 + }, + { + "epoch": 0.57341, + "grad_norm": 1.4845824170226591, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 57341 + }, + { + "epoch": 0.57342, + "grad_norm": 1.6209688640764508, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 57342 + }, + { + "epoch": 0.57343, + "grad_norm": 1.4101745363116702, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 57343 + }, + { + "epoch": 0.57344, + "grad_norm": 1.5804096064372777, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 57344 + }, + { + "epoch": 0.57345, + "grad_norm": 1.4831254112255572, + "learning_rate": 0.003, + "loss": 4.014, + "step": 57345 + }, + { + "epoch": 0.57346, + "grad_norm": 1.9423661476536607, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 57346 + }, + { + "epoch": 0.57347, + "grad_norm": 1.4062047973799812, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 57347 + }, + { + "epoch": 0.57348, + "grad_norm": 1.6207240340798652, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 57348 + }, + { + "epoch": 0.57349, + "grad_norm": 2.070973685641317, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 57349 + }, + { + "epoch": 0.5735, + "grad_norm": 1.7593331866723965, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 57350 + }, + { + "epoch": 0.57351, + "grad_norm": 1.2978437687973319, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 57351 + }, + { + "epoch": 0.57352, + "grad_norm": 1.5967635583411166, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 57352 + }, + { + "epoch": 0.57353, + "grad_norm": 1.636360112018483, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 57353 + }, + { + "epoch": 0.57354, + "grad_norm": 1.6200899450885136, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 57354 + }, + { + "epoch": 0.57355, + "grad_norm": 1.5432719126177359, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 57355 + }, + { + "epoch": 0.57356, + "grad_norm": 1.6575702716319562, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 57356 + }, + { + "epoch": 0.57357, + "grad_norm": 1.4973078039347225, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 57357 + }, + { + "epoch": 0.57358, + "grad_norm": 1.7585944220436878, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 57358 + }, + { + "epoch": 0.57359, + "grad_norm": 1.2048286805242978, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 57359 + }, + { + "epoch": 0.5736, + "grad_norm": 1.8589961388961638, + "learning_rate": 0.003, + "loss": 4.051, + "step": 57360 + }, + { + "epoch": 0.57361, + "grad_norm": 1.770942785536783, + "learning_rate": 0.003, + "loss": 4.001, + "step": 57361 + }, + { + "epoch": 0.57362, + "grad_norm": 1.3664941777521638, + "learning_rate": 0.003, + "loss": 4.04, + "step": 57362 + }, + { + "epoch": 0.57363, + "grad_norm": 1.7426751227973345, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 57363 + }, + { + "epoch": 0.57364, + "grad_norm": 1.4525954605954403, + "learning_rate": 0.003, + "loss": 4.049, + "step": 57364 + }, + { + "epoch": 0.57365, + "grad_norm": 1.913036335924713, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 57365 + }, + { + "epoch": 0.57366, + "grad_norm": 1.1828068036466333, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 57366 + }, + { + "epoch": 0.57367, + "grad_norm": 1.8255911353535725, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57367 + }, + { + "epoch": 0.57368, + "grad_norm": 1.4834606550059295, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 57368 + }, + { + "epoch": 0.57369, + "grad_norm": 1.4691823789835332, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 57369 + }, + { + "epoch": 0.5737, + "grad_norm": 1.7160705091310153, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 57370 + }, + { + "epoch": 0.57371, + "grad_norm": 1.6030509845098615, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 57371 + }, + { + "epoch": 0.57372, + "grad_norm": 2.185947154209666, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 57372 + }, + { + "epoch": 0.57373, + "grad_norm": 1.4683610277761974, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 57373 + }, + { + "epoch": 0.57374, + "grad_norm": 1.8402937332635982, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57374 + }, + { + "epoch": 0.57375, + "grad_norm": 1.6080398116771946, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 57375 + }, + { + "epoch": 0.57376, + "grad_norm": 1.7145174397741219, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 57376 + }, + { + "epoch": 0.57377, + "grad_norm": 1.5270657568152939, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 57377 + }, + { + "epoch": 0.57378, + "grad_norm": 1.3344142894467423, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 57378 + }, + { + "epoch": 0.57379, + "grad_norm": 1.7667991686812379, + "learning_rate": 0.003, + "loss": 4.016, + "step": 57379 + }, + { + "epoch": 0.5738, + "grad_norm": 1.4686429885897387, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 57380 + }, + { + "epoch": 0.57381, + "grad_norm": 1.6852454109560004, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 57381 + }, + { + "epoch": 0.57382, + "grad_norm": 1.4419455394790945, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 57382 + }, + { + "epoch": 0.57383, + "grad_norm": 1.8353052213072063, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 57383 + }, + { + "epoch": 0.57384, + "grad_norm": 1.4603157070343715, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 57384 + }, + { + "epoch": 0.57385, + "grad_norm": 1.8944215048449615, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 57385 + }, + { + "epoch": 0.57386, + "grad_norm": 1.413632726760585, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 57386 + }, + { + "epoch": 0.57387, + "grad_norm": 1.6807868559337877, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 57387 + }, + { + "epoch": 0.57388, + "grad_norm": 1.5227593981752763, + "learning_rate": 0.003, + "loss": 4.045, + "step": 57388 + }, + { + "epoch": 0.57389, + "grad_norm": 1.7023945744434088, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 57389 + }, + { + "epoch": 0.5739, + "grad_norm": 1.6129610821424227, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 57390 + }, + { + "epoch": 0.57391, + "grad_norm": 1.5376519415632826, + "learning_rate": 0.003, + "loss": 4.005, + "step": 57391 + }, + { + "epoch": 0.57392, + "grad_norm": 1.5805086358024756, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 57392 + }, + { + "epoch": 0.57393, + "grad_norm": 1.5361679294990367, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 57393 + }, + { + "epoch": 0.57394, + "grad_norm": 1.4405395942447816, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 57394 + }, + { + "epoch": 0.57395, + "grad_norm": 1.3465764582109674, + "learning_rate": 0.003, + "loss": 4.009, + "step": 57395 + }, + { + "epoch": 0.57396, + "grad_norm": 1.9513319876918924, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 57396 + }, + { + "epoch": 0.57397, + "grad_norm": 1.3776551268849242, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 57397 + }, + { + "epoch": 0.57398, + "grad_norm": 1.8078843087571541, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 57398 + }, + { + "epoch": 0.57399, + "grad_norm": 1.426919927782735, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 57399 + }, + { + "epoch": 0.574, + "grad_norm": 1.5396551831215928, + "learning_rate": 0.003, + "loss": 4.061, + "step": 57400 + }, + { + "epoch": 0.57401, + "grad_norm": 1.6646374220475957, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 57401 + }, + { + "epoch": 0.57402, + "grad_norm": 1.4255253144208824, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57402 + }, + { + "epoch": 0.57403, + "grad_norm": 1.5958466598712808, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 57403 + }, + { + "epoch": 0.57404, + "grad_norm": 1.441078495714436, + "learning_rate": 0.003, + "loss": 4.024, + "step": 57404 + }, + { + "epoch": 0.57405, + "grad_norm": 1.9142649742719933, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 57405 + }, + { + "epoch": 0.57406, + "grad_norm": 1.4530374648716087, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 57406 + }, + { + "epoch": 0.57407, + "grad_norm": 1.6050618699276529, + "learning_rate": 0.003, + "loss": 4.04, + "step": 57407 + }, + { + "epoch": 0.57408, + "grad_norm": 1.6573347261694988, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 57408 + }, + { + "epoch": 0.57409, + "grad_norm": 1.87638827926471, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 57409 + }, + { + "epoch": 0.5741, + "grad_norm": 1.5526707508937647, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 57410 + }, + { + "epoch": 0.57411, + "grad_norm": 1.7173808975656486, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 57411 + }, + { + "epoch": 0.57412, + "grad_norm": 1.4874387130861302, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 57412 + }, + { + "epoch": 0.57413, + "grad_norm": 1.5272002792413966, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 57413 + }, + { + "epoch": 0.57414, + "grad_norm": 1.8977719742481312, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 57414 + }, + { + "epoch": 0.57415, + "grad_norm": 1.4231500955151661, + "learning_rate": 0.003, + "loss": 4.03, + "step": 57415 + }, + { + "epoch": 0.57416, + "grad_norm": 1.7145830959151638, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 57416 + }, + { + "epoch": 0.57417, + "grad_norm": 1.7897085386417961, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 57417 + }, + { + "epoch": 0.57418, + "grad_norm": 1.3453200783332795, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 57418 + }, + { + "epoch": 0.57419, + "grad_norm": 1.8360030786391452, + "learning_rate": 0.003, + "loss": 4.037, + "step": 57419 + }, + { + "epoch": 0.5742, + "grad_norm": 1.4985876998665035, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 57420 + }, + { + "epoch": 0.57421, + "grad_norm": 1.7526984498552396, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 57421 + }, + { + "epoch": 0.57422, + "grad_norm": 1.5349403827830148, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 57422 + }, + { + "epoch": 0.57423, + "grad_norm": 1.4672921303800135, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 57423 + }, + { + "epoch": 0.57424, + "grad_norm": 1.334020617809309, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 57424 + }, + { + "epoch": 0.57425, + "grad_norm": 1.745153468030044, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 57425 + }, + { + "epoch": 0.57426, + "grad_norm": 1.420786250932367, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 57426 + }, + { + "epoch": 0.57427, + "grad_norm": 1.7946239680421656, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 57427 + }, + { + "epoch": 0.57428, + "grad_norm": 1.7891188105444729, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 57428 + }, + { + "epoch": 0.57429, + "grad_norm": 1.5500280167410811, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 57429 + }, + { + "epoch": 0.5743, + "grad_norm": 1.7030922118254987, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 57430 + }, + { + "epoch": 0.57431, + "grad_norm": 1.9468710817411914, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 57431 + }, + { + "epoch": 0.57432, + "grad_norm": 1.4153673897022827, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 57432 + }, + { + "epoch": 0.57433, + "grad_norm": 1.6685781372129158, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 57433 + }, + { + "epoch": 0.57434, + "grad_norm": 1.248549636188904, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 57434 + }, + { + "epoch": 0.57435, + "grad_norm": 1.9471095435284322, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 57435 + }, + { + "epoch": 0.57436, + "grad_norm": 1.3827325085674247, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 57436 + }, + { + "epoch": 0.57437, + "grad_norm": 1.6943006722479268, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 57437 + }, + { + "epoch": 0.57438, + "grad_norm": 1.6686671374530668, + "learning_rate": 0.003, + "loss": 4.032, + "step": 57438 + }, + { + "epoch": 0.57439, + "grad_norm": 1.5054902005170732, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 57439 + }, + { + "epoch": 0.5744, + "grad_norm": 1.7840434765752575, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 57440 + }, + { + "epoch": 0.57441, + "grad_norm": 1.3811086325260638, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 57441 + }, + { + "epoch": 0.57442, + "grad_norm": 1.7119702581684393, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 57442 + }, + { + "epoch": 0.57443, + "grad_norm": 1.7012749107666778, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57443 + }, + { + "epoch": 0.57444, + "grad_norm": 1.3011144439276212, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 57444 + }, + { + "epoch": 0.57445, + "grad_norm": 2.1740453046661923, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 57445 + }, + { + "epoch": 0.57446, + "grad_norm": 1.5491991858348961, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 57446 + }, + { + "epoch": 0.57447, + "grad_norm": 1.6268369680543047, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 57447 + }, + { + "epoch": 0.57448, + "grad_norm": 1.6346338208854831, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 57448 + }, + { + "epoch": 0.57449, + "grad_norm": 1.5560098645226543, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 57449 + }, + { + "epoch": 0.5745, + "grad_norm": 1.4978528876704797, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 57450 + }, + { + "epoch": 0.57451, + "grad_norm": 1.473623348780826, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 57451 + }, + { + "epoch": 0.57452, + "grad_norm": 1.6892902019439655, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 57452 + }, + { + "epoch": 0.57453, + "grad_norm": 1.276821514778993, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 57453 + }, + { + "epoch": 0.57454, + "grad_norm": 1.7224477403572682, + "learning_rate": 0.003, + "loss": 4.05, + "step": 57454 + }, + { + "epoch": 0.57455, + "grad_norm": 1.3523643541337942, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 57455 + }, + { + "epoch": 0.57456, + "grad_norm": 1.8160569895558327, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 57456 + }, + { + "epoch": 0.57457, + "grad_norm": 1.3742110146608768, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 57457 + }, + { + "epoch": 0.57458, + "grad_norm": 2.147892611551509, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 57458 + }, + { + "epoch": 0.57459, + "grad_norm": 1.3890246345612252, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 57459 + }, + { + "epoch": 0.5746, + "grad_norm": 1.6505249715501558, + "learning_rate": 0.003, + "loss": 4.022, + "step": 57460 + }, + { + "epoch": 0.57461, + "grad_norm": 1.9392118044498106, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 57461 + }, + { + "epoch": 0.57462, + "grad_norm": 1.3048099184703448, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 57462 + }, + { + "epoch": 0.57463, + "grad_norm": 1.6806215266322329, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 57463 + }, + { + "epoch": 0.57464, + "grad_norm": 1.6209657282960264, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57464 + }, + { + "epoch": 0.57465, + "grad_norm": 1.6420434869144485, + "learning_rate": 0.003, + "loss": 4.036, + "step": 57465 + }, + { + "epoch": 0.57466, + "grad_norm": 1.4109458012418443, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 57466 + }, + { + "epoch": 0.57467, + "grad_norm": 1.5575439582934196, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 57467 + }, + { + "epoch": 0.57468, + "grad_norm": 1.406698507324895, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 57468 + }, + { + "epoch": 0.57469, + "grad_norm": 1.4791453126780463, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 57469 + }, + { + "epoch": 0.5747, + "grad_norm": 1.4664703333840357, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 57470 + }, + { + "epoch": 0.57471, + "grad_norm": 1.5830731345356963, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 57471 + }, + { + "epoch": 0.57472, + "grad_norm": 1.4668742399960872, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 57472 + }, + { + "epoch": 0.57473, + "grad_norm": 1.675003205087458, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 57473 + }, + { + "epoch": 0.57474, + "grad_norm": 1.6683018784879293, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 57474 + }, + { + "epoch": 0.57475, + "grad_norm": 1.7611723578470524, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 57475 + }, + { + "epoch": 0.57476, + "grad_norm": 1.6428010553900054, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 57476 + }, + { + "epoch": 0.57477, + "grad_norm": 1.3411663448102196, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 57477 + }, + { + "epoch": 0.57478, + "grad_norm": 1.7391952988600685, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 57478 + }, + { + "epoch": 0.57479, + "grad_norm": 2.1606127487816402, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 57479 + }, + { + "epoch": 0.5748, + "grad_norm": 1.2075214257013622, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 57480 + }, + { + "epoch": 0.57481, + "grad_norm": 1.7785851690046992, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 57481 + }, + { + "epoch": 0.57482, + "grad_norm": 1.5302406490115197, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 57482 + }, + { + "epoch": 0.57483, + "grad_norm": 1.575756378265995, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 57483 + }, + { + "epoch": 0.57484, + "grad_norm": 1.6883995533689324, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57484 + }, + { + "epoch": 0.57485, + "grad_norm": 1.3946844541190833, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 57485 + }, + { + "epoch": 0.57486, + "grad_norm": 1.6058941748307318, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 57486 + }, + { + "epoch": 0.57487, + "grad_norm": 1.6655116496184168, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 57487 + }, + { + "epoch": 0.57488, + "grad_norm": 1.2769209355883264, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 57488 + }, + { + "epoch": 0.57489, + "grad_norm": 1.7623049464548033, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 57489 + }, + { + "epoch": 0.5749, + "grad_norm": 1.3034079240842766, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 57490 + }, + { + "epoch": 0.57491, + "grad_norm": 1.6549737282475385, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 57491 + }, + { + "epoch": 0.57492, + "grad_norm": 1.429699745520039, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 57492 + }, + { + "epoch": 0.57493, + "grad_norm": 1.649280612259243, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 57493 + }, + { + "epoch": 0.57494, + "grad_norm": 1.4133536049030835, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 57494 + }, + { + "epoch": 0.57495, + "grad_norm": 1.6909322407942429, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 57495 + }, + { + "epoch": 0.57496, + "grad_norm": 1.4614099911667977, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57496 + }, + { + "epoch": 0.57497, + "grad_norm": 1.7429202646737143, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 57497 + }, + { + "epoch": 0.57498, + "grad_norm": 1.6166780271032035, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 57498 + }, + { + "epoch": 0.57499, + "grad_norm": 1.5269805869100714, + "learning_rate": 0.003, + "loss": 4.037, + "step": 57499 + }, + { + "epoch": 0.575, + "grad_norm": 1.7782940196148074, + "learning_rate": 0.003, + "loss": 4.019, + "step": 57500 + }, + { + "epoch": 0.57501, + "grad_norm": 1.6343288912415568, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 57501 + }, + { + "epoch": 0.57502, + "grad_norm": 1.8964505907223559, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57502 + }, + { + "epoch": 0.57503, + "grad_norm": 1.4203957785048653, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 57503 + }, + { + "epoch": 0.57504, + "grad_norm": 1.7812666621605548, + "learning_rate": 0.003, + "loss": 4.053, + "step": 57504 + }, + { + "epoch": 0.57505, + "grad_norm": 1.4438149152246111, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 57505 + }, + { + "epoch": 0.57506, + "grad_norm": 1.7802402021712362, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 57506 + }, + { + "epoch": 0.57507, + "grad_norm": 1.3600414627751949, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 57507 + }, + { + "epoch": 0.57508, + "grad_norm": 1.673030364820609, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 57508 + }, + { + "epoch": 0.57509, + "grad_norm": 1.7920726130429432, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 57509 + }, + { + "epoch": 0.5751, + "grad_norm": 1.8382403943360845, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 57510 + }, + { + "epoch": 0.57511, + "grad_norm": 1.5778192989009838, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 57511 + }, + { + "epoch": 0.57512, + "grad_norm": 1.7323713924187847, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 57512 + }, + { + "epoch": 0.57513, + "grad_norm": 1.3765155488421554, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 57513 + }, + { + "epoch": 0.57514, + "grad_norm": 1.5527192180602543, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 57514 + }, + { + "epoch": 0.57515, + "grad_norm": 1.6125401768829735, + "learning_rate": 0.003, + "loss": 4.045, + "step": 57515 + }, + { + "epoch": 0.57516, + "grad_norm": 1.4651815219352164, + "learning_rate": 0.003, + "loss": 3.998, + "step": 57516 + }, + { + "epoch": 0.57517, + "grad_norm": 1.5273200749734526, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 57517 + }, + { + "epoch": 0.57518, + "grad_norm": 1.4611770603080838, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 57518 + }, + { + "epoch": 0.57519, + "grad_norm": 1.5403489336443434, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 57519 + }, + { + "epoch": 0.5752, + "grad_norm": 1.3825606693447103, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 57520 + }, + { + "epoch": 0.57521, + "grad_norm": 1.6293198653227359, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 57521 + }, + { + "epoch": 0.57522, + "grad_norm": 1.4333281893004728, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 57522 + }, + { + "epoch": 0.57523, + "grad_norm": 1.7359999311386562, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 57523 + }, + { + "epoch": 0.57524, + "grad_norm": 1.3395831100312063, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 57524 + }, + { + "epoch": 0.57525, + "grad_norm": 1.7171343375061268, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 57525 + }, + { + "epoch": 0.57526, + "grad_norm": 1.551333956616372, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 57526 + }, + { + "epoch": 0.57527, + "grad_norm": 1.6474799643892668, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 57527 + }, + { + "epoch": 0.57528, + "grad_norm": 1.630775023065861, + "learning_rate": 0.003, + "loss": 4.031, + "step": 57528 + }, + { + "epoch": 0.57529, + "grad_norm": 1.8418356877147883, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 57529 + }, + { + "epoch": 0.5753, + "grad_norm": 1.8864502579763305, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57530 + }, + { + "epoch": 0.57531, + "grad_norm": 1.2790134071309376, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 57531 + }, + { + "epoch": 0.57532, + "grad_norm": 2.0114483380505943, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 57532 + }, + { + "epoch": 0.57533, + "grad_norm": 1.4205330312760407, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57533 + }, + { + "epoch": 0.57534, + "grad_norm": 1.6268370606793052, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 57534 + }, + { + "epoch": 0.57535, + "grad_norm": 1.4397071007784632, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 57535 + }, + { + "epoch": 0.57536, + "grad_norm": 1.4543289394613643, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 57536 + }, + { + "epoch": 0.57537, + "grad_norm": 1.492540246634578, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 57537 + }, + { + "epoch": 0.57538, + "grad_norm": 1.5226882149364176, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 57538 + }, + { + "epoch": 0.57539, + "grad_norm": 1.656220152370945, + "learning_rate": 0.003, + "loss": 4.045, + "step": 57539 + }, + { + "epoch": 0.5754, + "grad_norm": 1.4038757362604188, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 57540 + }, + { + "epoch": 0.57541, + "grad_norm": 1.4857297448186781, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 57541 + }, + { + "epoch": 0.57542, + "grad_norm": 1.6860028418175688, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 57542 + }, + { + "epoch": 0.57543, + "grad_norm": 1.9643649251843383, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 57543 + }, + { + "epoch": 0.57544, + "grad_norm": 1.809422072726159, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 57544 + }, + { + "epoch": 0.57545, + "grad_norm": 1.2282931078151722, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 57545 + }, + { + "epoch": 0.57546, + "grad_norm": 1.8606478145250347, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 57546 + }, + { + "epoch": 0.57547, + "grad_norm": 1.458912096407571, + "learning_rate": 0.003, + "loss": 4.013, + "step": 57547 + }, + { + "epoch": 0.57548, + "grad_norm": 1.4796903102988512, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 57548 + }, + { + "epoch": 0.57549, + "grad_norm": 1.765849346220587, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 57549 + }, + { + "epoch": 0.5755, + "grad_norm": 1.6107011987210598, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 57550 + }, + { + "epoch": 0.57551, + "grad_norm": 1.3645546361723286, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 57551 + }, + { + "epoch": 0.57552, + "grad_norm": 1.743517106794816, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 57552 + }, + { + "epoch": 0.57553, + "grad_norm": 1.4362972102231581, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57553 + }, + { + "epoch": 0.57554, + "grad_norm": 1.636272201005904, + "learning_rate": 0.003, + "loss": 4.037, + "step": 57554 + }, + { + "epoch": 0.57555, + "grad_norm": 1.5879435931019585, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 57555 + }, + { + "epoch": 0.57556, + "grad_norm": 1.3979727776942708, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 57556 + }, + { + "epoch": 0.57557, + "grad_norm": 1.7989859725388693, + "learning_rate": 0.003, + "loss": 4.033, + "step": 57557 + }, + { + "epoch": 0.57558, + "grad_norm": 1.4921869378768962, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 57558 + }, + { + "epoch": 0.57559, + "grad_norm": 1.4617436452037573, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 57559 + }, + { + "epoch": 0.5756, + "grad_norm": 1.5704944970120727, + "learning_rate": 0.003, + "loss": 4.037, + "step": 57560 + }, + { + "epoch": 0.57561, + "grad_norm": 1.903856574212757, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 57561 + }, + { + "epoch": 0.57562, + "grad_norm": 1.7235557315948793, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 57562 + }, + { + "epoch": 0.57563, + "grad_norm": 1.5826912910580613, + "learning_rate": 0.003, + "loss": 4.021, + "step": 57563 + }, + { + "epoch": 0.57564, + "grad_norm": 1.5753020700526583, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 57564 + }, + { + "epoch": 0.57565, + "grad_norm": 1.6969502509316616, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 57565 + }, + { + "epoch": 0.57566, + "grad_norm": 1.5821694096897354, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 57566 + }, + { + "epoch": 0.57567, + "grad_norm": 1.3911998031777209, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 57567 + }, + { + "epoch": 0.57568, + "grad_norm": 1.6889435482547277, + "learning_rate": 0.003, + "loss": 4.015, + "step": 57568 + }, + { + "epoch": 0.57569, + "grad_norm": 1.4294540797620112, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 57569 + }, + { + "epoch": 0.5757, + "grad_norm": 1.8331914857190974, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 57570 + }, + { + "epoch": 0.57571, + "grad_norm": 1.1630480048614127, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 57571 + }, + { + "epoch": 0.57572, + "grad_norm": 1.846605694261479, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 57572 + }, + { + "epoch": 0.57573, + "grad_norm": 1.4200891552318526, + "learning_rate": 0.003, + "loss": 4.036, + "step": 57573 + }, + { + "epoch": 0.57574, + "grad_norm": 1.870884342341614, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 57574 + }, + { + "epoch": 0.57575, + "grad_norm": 1.4434594448951628, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 57575 + }, + { + "epoch": 0.57576, + "grad_norm": 1.6347365882402938, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 57576 + }, + { + "epoch": 0.57577, + "grad_norm": 1.5271272934110565, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 57577 + }, + { + "epoch": 0.57578, + "grad_norm": 1.6419583600464862, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 57578 + }, + { + "epoch": 0.57579, + "grad_norm": 1.5464819983100417, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57579 + }, + { + "epoch": 0.5758, + "grad_norm": 1.5394522535891815, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 57580 + }, + { + "epoch": 0.57581, + "grad_norm": 1.496720758093598, + "learning_rate": 0.003, + "loss": 4.024, + "step": 57581 + }, + { + "epoch": 0.57582, + "grad_norm": 1.7606721410578186, + "learning_rate": 0.003, + "loss": 4.036, + "step": 57582 + }, + { + "epoch": 0.57583, + "grad_norm": 1.5735950289614549, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 57583 + }, + { + "epoch": 0.57584, + "grad_norm": 1.7707465591294205, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 57584 + }, + { + "epoch": 0.57585, + "grad_norm": 1.7671162014632058, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 57585 + }, + { + "epoch": 0.57586, + "grad_norm": 1.4730222339323018, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 57586 + }, + { + "epoch": 0.57587, + "grad_norm": 1.8768745724341962, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 57587 + }, + { + "epoch": 0.57588, + "grad_norm": 1.438164897033053, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 57588 + }, + { + "epoch": 0.57589, + "grad_norm": 1.6480642628273379, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 57589 + }, + { + "epoch": 0.5759, + "grad_norm": 1.6743026248153794, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 57590 + }, + { + "epoch": 0.57591, + "grad_norm": 1.547279406783904, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 57591 + }, + { + "epoch": 0.57592, + "grad_norm": 1.4611931083256964, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 57592 + }, + { + "epoch": 0.57593, + "grad_norm": 1.7189374741308994, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 57593 + }, + { + "epoch": 0.57594, + "grad_norm": 1.3739405347216154, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 57594 + }, + { + "epoch": 0.57595, + "grad_norm": 1.523518119180164, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 57595 + }, + { + "epoch": 0.57596, + "grad_norm": 1.5938113474009161, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 57596 + }, + { + "epoch": 0.57597, + "grad_norm": 1.5421622279743623, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 57597 + }, + { + "epoch": 0.57598, + "grad_norm": 1.444271131223193, + "learning_rate": 0.003, + "loss": 4.05, + "step": 57598 + }, + { + "epoch": 0.57599, + "grad_norm": 1.682193387272921, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 57599 + }, + { + "epoch": 0.576, + "grad_norm": 1.397231465197799, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 57600 + }, + { + "epoch": 0.57601, + "grad_norm": 1.8882122140941218, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 57601 + }, + { + "epoch": 0.57602, + "grad_norm": 1.3981504202360704, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 57602 + }, + { + "epoch": 0.57603, + "grad_norm": 1.778828983818438, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57603 + }, + { + "epoch": 0.57604, + "grad_norm": 1.6184387331452341, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 57604 + }, + { + "epoch": 0.57605, + "grad_norm": 1.4428634983876376, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 57605 + }, + { + "epoch": 0.57606, + "grad_norm": 1.7878525702117793, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 57606 + }, + { + "epoch": 0.57607, + "grad_norm": 1.5184703785465816, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 57607 + }, + { + "epoch": 0.57608, + "grad_norm": 1.8115011555332825, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 57608 + }, + { + "epoch": 0.57609, + "grad_norm": 1.5836004282761993, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 57609 + }, + { + "epoch": 0.5761, + "grad_norm": 1.8975256672591516, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 57610 + }, + { + "epoch": 0.57611, + "grad_norm": 1.547591332743924, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 57611 + }, + { + "epoch": 0.57612, + "grad_norm": 1.5949317810210957, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 57612 + }, + { + "epoch": 0.57613, + "grad_norm": 1.3989050604482218, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 57613 + }, + { + "epoch": 0.57614, + "grad_norm": 1.4847716043607702, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 57614 + }, + { + "epoch": 0.57615, + "grad_norm": 1.6296410093502698, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 57615 + }, + { + "epoch": 0.57616, + "grad_norm": 1.4269042848423392, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 57616 + }, + { + "epoch": 0.57617, + "grad_norm": 1.4241030937398382, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 57617 + }, + { + "epoch": 0.57618, + "grad_norm": 1.6964603292964264, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 57618 + }, + { + "epoch": 0.57619, + "grad_norm": 1.9041797029735843, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 57619 + }, + { + "epoch": 0.5762, + "grad_norm": 1.4739905130834554, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57620 + }, + { + "epoch": 0.57621, + "grad_norm": 1.8500437697337184, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57621 + }, + { + "epoch": 0.57622, + "grad_norm": 1.37643298425587, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 57622 + }, + { + "epoch": 0.57623, + "grad_norm": 1.7213858197805791, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 57623 + }, + { + "epoch": 0.57624, + "grad_norm": 1.8027445324988791, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 57624 + }, + { + "epoch": 0.57625, + "grad_norm": 1.2053106847976387, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57625 + }, + { + "epoch": 0.57626, + "grad_norm": 1.859764450332714, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 57626 + }, + { + "epoch": 0.57627, + "grad_norm": 1.5389491317439463, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 57627 + }, + { + "epoch": 0.57628, + "grad_norm": 1.5474651993789204, + "learning_rate": 0.003, + "loss": 4.032, + "step": 57628 + }, + { + "epoch": 0.57629, + "grad_norm": 1.9383543954624756, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 57629 + }, + { + "epoch": 0.5763, + "grad_norm": 1.3591123339207063, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57630 + }, + { + "epoch": 0.57631, + "grad_norm": 1.848548226467854, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 57631 + }, + { + "epoch": 0.57632, + "grad_norm": 1.561355732714864, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 57632 + }, + { + "epoch": 0.57633, + "grad_norm": 1.6276396164042766, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57633 + }, + { + "epoch": 0.57634, + "grad_norm": 1.5373115674825508, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 57634 + }, + { + "epoch": 0.57635, + "grad_norm": 1.531640831353728, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 57635 + }, + { + "epoch": 0.57636, + "grad_norm": 1.4070725220454108, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 57636 + }, + { + "epoch": 0.57637, + "grad_norm": 1.609827943081552, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 57637 + }, + { + "epoch": 0.57638, + "grad_norm": 1.5698102340781557, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 57638 + }, + { + "epoch": 0.57639, + "grad_norm": 1.8849996055784717, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 57639 + }, + { + "epoch": 0.5764, + "grad_norm": 1.27996510431065, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 57640 + }, + { + "epoch": 0.57641, + "grad_norm": 1.7116086075209376, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 57641 + }, + { + "epoch": 0.57642, + "grad_norm": 1.5431337857374012, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 57642 + }, + { + "epoch": 0.57643, + "grad_norm": 1.7705169817051398, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 57643 + }, + { + "epoch": 0.57644, + "grad_norm": 1.2105922060205219, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 57644 + }, + { + "epoch": 0.57645, + "grad_norm": 1.803024176524946, + "learning_rate": 0.003, + "loss": 4.04, + "step": 57645 + }, + { + "epoch": 0.57646, + "grad_norm": 1.4601835543954784, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 57646 + }, + { + "epoch": 0.57647, + "grad_norm": 1.5587487527066612, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 57647 + }, + { + "epoch": 0.57648, + "grad_norm": 1.7344521926013574, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 57648 + }, + { + "epoch": 0.57649, + "grad_norm": 1.4561774938864689, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 57649 + }, + { + "epoch": 0.5765, + "grad_norm": 1.8921559904647884, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 57650 + }, + { + "epoch": 0.57651, + "grad_norm": 1.6958932601022536, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 57651 + }, + { + "epoch": 0.57652, + "grad_norm": 1.8364903183124244, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57652 + }, + { + "epoch": 0.57653, + "grad_norm": 1.2714608396101534, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 57653 + }, + { + "epoch": 0.57654, + "grad_norm": 1.8552469182340454, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57654 + }, + { + "epoch": 0.57655, + "grad_norm": 1.4024953359264258, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 57655 + }, + { + "epoch": 0.57656, + "grad_norm": 1.7496995340262456, + "learning_rate": 0.003, + "loss": 4.049, + "step": 57656 + }, + { + "epoch": 0.57657, + "grad_norm": 1.3505850452087786, + "learning_rate": 0.003, + "loss": 4.019, + "step": 57657 + }, + { + "epoch": 0.57658, + "grad_norm": 1.6617530421016045, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 57658 + }, + { + "epoch": 0.57659, + "grad_norm": 1.4150206520902553, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 57659 + }, + { + "epoch": 0.5766, + "grad_norm": 1.769216327703399, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57660 + }, + { + "epoch": 0.57661, + "grad_norm": 1.4203567999457154, + "learning_rate": 0.003, + "loss": 4.046, + "step": 57661 + }, + { + "epoch": 0.57662, + "grad_norm": 1.95306503162505, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 57662 + }, + { + "epoch": 0.57663, + "grad_norm": 1.2405307220979278, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 57663 + }, + { + "epoch": 0.57664, + "grad_norm": 1.7224015250696747, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 57664 + }, + { + "epoch": 0.57665, + "grad_norm": 1.6761453946333682, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 57665 + }, + { + "epoch": 0.57666, + "grad_norm": 1.9846612819874705, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 57666 + }, + { + "epoch": 0.57667, + "grad_norm": 1.2962358111828869, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 57667 + }, + { + "epoch": 0.57668, + "grad_norm": 1.745089834020313, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 57668 + }, + { + "epoch": 0.57669, + "grad_norm": 1.6405989243822312, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 57669 + }, + { + "epoch": 0.5767, + "grad_norm": 1.460480861723082, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 57670 + }, + { + "epoch": 0.57671, + "grad_norm": 1.770055205315284, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 57671 + }, + { + "epoch": 0.57672, + "grad_norm": 1.4034931896921197, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 57672 + }, + { + "epoch": 0.57673, + "grad_norm": 1.5746123394067857, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 57673 + }, + { + "epoch": 0.57674, + "grad_norm": 1.500640384614829, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 57674 + }, + { + "epoch": 0.57675, + "grad_norm": 1.4905737100049616, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 57675 + }, + { + "epoch": 0.57676, + "grad_norm": 1.6158058503202315, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 57676 + }, + { + "epoch": 0.57677, + "grad_norm": 1.4801720921610575, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 57677 + }, + { + "epoch": 0.57678, + "grad_norm": 1.5646751501884133, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 57678 + }, + { + "epoch": 0.57679, + "grad_norm": 1.583993027059523, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 57679 + }, + { + "epoch": 0.5768, + "grad_norm": 1.6126265132662947, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 57680 + }, + { + "epoch": 0.57681, + "grad_norm": 1.8249756347362256, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 57681 + }, + { + "epoch": 0.57682, + "grad_norm": 1.6359364047700182, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 57682 + }, + { + "epoch": 0.57683, + "grad_norm": 1.6813911500081362, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 57683 + }, + { + "epoch": 0.57684, + "grad_norm": 1.677708520962484, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 57684 + }, + { + "epoch": 0.57685, + "grad_norm": 1.6014368437497906, + "learning_rate": 0.003, + "loss": 4.025, + "step": 57685 + }, + { + "epoch": 0.57686, + "grad_norm": 1.111639139379655, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 57686 + }, + { + "epoch": 0.57687, + "grad_norm": 1.7763038886520757, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 57687 + }, + { + "epoch": 0.57688, + "grad_norm": 1.251629242680474, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 57688 + }, + { + "epoch": 0.57689, + "grad_norm": 1.7022056266385985, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 57689 + }, + { + "epoch": 0.5769, + "grad_norm": 1.592354887198759, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 57690 + }, + { + "epoch": 0.57691, + "grad_norm": 1.614223251136056, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 57691 + }, + { + "epoch": 0.57692, + "grad_norm": 1.8147739968002852, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 57692 + }, + { + "epoch": 0.57693, + "grad_norm": 1.5254993381089639, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 57693 + }, + { + "epoch": 0.57694, + "grad_norm": 1.388353417227346, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 57694 + }, + { + "epoch": 0.57695, + "grad_norm": 1.5665530915944337, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57695 + }, + { + "epoch": 0.57696, + "grad_norm": 1.4091349759533889, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 57696 + }, + { + "epoch": 0.57697, + "grad_norm": 1.5732153774943416, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 57697 + }, + { + "epoch": 0.57698, + "grad_norm": 1.7320462878227045, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 57698 + }, + { + "epoch": 0.57699, + "grad_norm": 1.7042002687661881, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 57699 + }, + { + "epoch": 0.577, + "grad_norm": 1.9748899422914397, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 57700 + }, + { + "epoch": 0.57701, + "grad_norm": 1.4977285258681632, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 57701 + }, + { + "epoch": 0.57702, + "grad_norm": 1.6456923084992183, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 57702 + }, + { + "epoch": 0.57703, + "grad_norm": 1.6252963142997443, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 57703 + }, + { + "epoch": 0.57704, + "grad_norm": 1.3376594772176935, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 57704 + }, + { + "epoch": 0.57705, + "grad_norm": 1.6577853569761516, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 57705 + }, + { + "epoch": 0.57706, + "grad_norm": 1.468892198718526, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 57706 + }, + { + "epoch": 0.57707, + "grad_norm": 1.5753553509664582, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 57707 + }, + { + "epoch": 0.57708, + "grad_norm": 1.534459332279108, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 57708 + }, + { + "epoch": 0.57709, + "grad_norm": 1.6608858253618406, + "learning_rate": 0.003, + "loss": 4.011, + "step": 57709 + }, + { + "epoch": 0.5771, + "grad_norm": 1.5124947525528805, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 57710 + }, + { + "epoch": 0.57711, + "grad_norm": 1.6795542415443423, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57711 + }, + { + "epoch": 0.57712, + "grad_norm": 1.3917773859015417, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 57712 + }, + { + "epoch": 0.57713, + "grad_norm": 1.6600762223916983, + "learning_rate": 0.003, + "loss": 4.012, + "step": 57713 + }, + { + "epoch": 0.57714, + "grad_norm": 1.5849611132617558, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 57714 + }, + { + "epoch": 0.57715, + "grad_norm": 1.5664852408839989, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 57715 + }, + { + "epoch": 0.57716, + "grad_norm": 1.6731361885237939, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 57716 + }, + { + "epoch": 0.57717, + "grad_norm": 1.3530423262158762, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 57717 + }, + { + "epoch": 0.57718, + "grad_norm": 1.7720511784515052, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 57718 + }, + { + "epoch": 0.57719, + "grad_norm": 1.1090592288771721, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 57719 + }, + { + "epoch": 0.5772, + "grad_norm": 1.7668095576102731, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57720 + }, + { + "epoch": 0.57721, + "grad_norm": 1.290684059033582, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 57721 + }, + { + "epoch": 0.57722, + "grad_norm": 1.912987553052005, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 57722 + }, + { + "epoch": 0.57723, + "grad_norm": 1.664031728360565, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 57723 + }, + { + "epoch": 0.57724, + "grad_norm": 1.5250173720199056, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 57724 + }, + { + "epoch": 0.57725, + "grad_norm": 1.6560037598111652, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 57725 + }, + { + "epoch": 0.57726, + "grad_norm": 2.1722787918111197, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 57726 + }, + { + "epoch": 0.57727, + "grad_norm": 1.3995201868817668, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 57727 + }, + { + "epoch": 0.57728, + "grad_norm": 2.0144889436742077, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 57728 + }, + { + "epoch": 0.57729, + "grad_norm": 1.5407973375108899, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 57729 + }, + { + "epoch": 0.5773, + "grad_norm": 1.6546627709895965, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57730 + }, + { + "epoch": 0.57731, + "grad_norm": 1.4059084346441992, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 57731 + }, + { + "epoch": 0.57732, + "grad_norm": 1.9370080561114227, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57732 + }, + { + "epoch": 0.57733, + "grad_norm": 1.2532735497058716, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 57733 + }, + { + "epoch": 0.57734, + "grad_norm": 1.9404857665166524, + "learning_rate": 0.003, + "loss": 4.052, + "step": 57734 + }, + { + "epoch": 0.57735, + "grad_norm": 1.5341520117565144, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 57735 + }, + { + "epoch": 0.57736, + "grad_norm": 1.3589442613051572, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 57736 + }, + { + "epoch": 0.57737, + "grad_norm": 1.6729358570466564, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57737 + }, + { + "epoch": 0.57738, + "grad_norm": 2.000233817909503, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 57738 + }, + { + "epoch": 0.57739, + "grad_norm": 1.436016040462806, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 57739 + }, + { + "epoch": 0.5774, + "grad_norm": 1.4759368200693683, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 57740 + }, + { + "epoch": 0.57741, + "grad_norm": 1.469122789450211, + "learning_rate": 0.003, + "loss": 4.041, + "step": 57741 + }, + { + "epoch": 0.57742, + "grad_norm": 1.442104329215446, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 57742 + }, + { + "epoch": 0.57743, + "grad_norm": 1.5379346206401021, + "learning_rate": 0.003, + "loss": 4.029, + "step": 57743 + }, + { + "epoch": 0.57744, + "grad_norm": 1.4121428034556556, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 57744 + }, + { + "epoch": 0.57745, + "grad_norm": 1.6799908312710619, + "learning_rate": 0.003, + "loss": 4.031, + "step": 57745 + }, + { + "epoch": 0.57746, + "grad_norm": 1.1751801074857091, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 57746 + }, + { + "epoch": 0.57747, + "grad_norm": 1.8365442547197655, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 57747 + }, + { + "epoch": 0.57748, + "grad_norm": 1.3874225627793406, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 57748 + }, + { + "epoch": 0.57749, + "grad_norm": 1.8276958194903918, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 57749 + }, + { + "epoch": 0.5775, + "grad_norm": 1.4751454448919017, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 57750 + }, + { + "epoch": 0.57751, + "grad_norm": 1.678964339533992, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 57751 + }, + { + "epoch": 0.57752, + "grad_norm": 1.78095859312813, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 57752 + }, + { + "epoch": 0.57753, + "grad_norm": 1.6625638048341398, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 57753 + }, + { + "epoch": 0.57754, + "grad_norm": 1.6278566956726082, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 57754 + }, + { + "epoch": 0.57755, + "grad_norm": 1.5992894898577736, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 57755 + }, + { + "epoch": 0.57756, + "grad_norm": 1.5520220103697253, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57756 + }, + { + "epoch": 0.57757, + "grad_norm": 1.467310279206527, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 57757 + }, + { + "epoch": 0.57758, + "grad_norm": 1.6775193075293178, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 57758 + }, + { + "epoch": 0.57759, + "grad_norm": 1.213746841836933, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 57759 + }, + { + "epoch": 0.5776, + "grad_norm": 1.7365728663403104, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 57760 + }, + { + "epoch": 0.57761, + "grad_norm": 1.4695991175377816, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 57761 + }, + { + "epoch": 0.57762, + "grad_norm": 1.7740361503120166, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 57762 + }, + { + "epoch": 0.57763, + "grad_norm": 1.9660214559591054, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 57763 + }, + { + "epoch": 0.57764, + "grad_norm": 1.5135705276269376, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 57764 + }, + { + "epoch": 0.57765, + "grad_norm": 1.6056863679828277, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 57765 + }, + { + "epoch": 0.57766, + "grad_norm": 1.9132275771501857, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 57766 + }, + { + "epoch": 0.57767, + "grad_norm": 1.3712969764825866, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 57767 + }, + { + "epoch": 0.57768, + "grad_norm": 1.557258601191343, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 57768 + }, + { + "epoch": 0.57769, + "grad_norm": 1.7797299811721312, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 57769 + }, + { + "epoch": 0.5777, + "grad_norm": 1.488578536925719, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 57770 + }, + { + "epoch": 0.57771, + "grad_norm": 1.552535599352127, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 57771 + }, + { + "epoch": 0.57772, + "grad_norm": 1.4542479477832657, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 57772 + }, + { + "epoch": 0.57773, + "grad_norm": 1.5454995785002954, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 57773 + }, + { + "epoch": 0.57774, + "grad_norm": 1.3523256541391788, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 57774 + }, + { + "epoch": 0.57775, + "grad_norm": 1.7031350922499222, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 57775 + }, + { + "epoch": 0.57776, + "grad_norm": 1.5734720907610698, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 57776 + }, + { + "epoch": 0.57777, + "grad_norm": 1.4788441052428511, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 57777 + }, + { + "epoch": 0.57778, + "grad_norm": 1.8249569349336328, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 57778 + }, + { + "epoch": 0.57779, + "grad_norm": 1.4051332547756854, + "learning_rate": 0.003, + "loss": 4.032, + "step": 57779 + }, + { + "epoch": 0.5778, + "grad_norm": 1.5487008045136723, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 57780 + }, + { + "epoch": 0.57781, + "grad_norm": 1.7911855112071913, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 57781 + }, + { + "epoch": 0.57782, + "grad_norm": 1.5378404761801394, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 57782 + }, + { + "epoch": 0.57783, + "grad_norm": 1.603752844831316, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57783 + }, + { + "epoch": 0.57784, + "grad_norm": 1.6693861385547442, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 57784 + }, + { + "epoch": 0.57785, + "grad_norm": 1.7793338705106947, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 57785 + }, + { + "epoch": 0.57786, + "grad_norm": 1.325758627778898, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 57786 + }, + { + "epoch": 0.57787, + "grad_norm": 1.673670618071037, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 57787 + }, + { + "epoch": 0.57788, + "grad_norm": 1.4884253891360877, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 57788 + }, + { + "epoch": 0.57789, + "grad_norm": 1.7129986155175887, + "learning_rate": 0.003, + "loss": 4.059, + "step": 57789 + }, + { + "epoch": 0.5779, + "grad_norm": 1.4609310432868796, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 57790 + }, + { + "epoch": 0.57791, + "grad_norm": 1.1726316090842295, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 57791 + }, + { + "epoch": 0.57792, + "grad_norm": 2.024680542242276, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 57792 + }, + { + "epoch": 0.57793, + "grad_norm": 1.7548389687113533, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 57793 + }, + { + "epoch": 0.57794, + "grad_norm": 1.849062917089833, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 57794 + }, + { + "epoch": 0.57795, + "grad_norm": 1.393854252661433, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57795 + }, + { + "epoch": 0.57796, + "grad_norm": 1.642762871910959, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 57796 + }, + { + "epoch": 0.57797, + "grad_norm": 1.7108437301584647, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 57797 + }, + { + "epoch": 0.57798, + "grad_norm": 1.5091039456197548, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 57798 + }, + { + "epoch": 0.57799, + "grad_norm": 1.8487852226750674, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 57799 + }, + { + "epoch": 0.578, + "grad_norm": 1.6761862991644243, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 57800 + }, + { + "epoch": 0.57801, + "grad_norm": 1.4789388464738005, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 57801 + }, + { + "epoch": 0.57802, + "grad_norm": 1.525883972509906, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 57802 + }, + { + "epoch": 0.57803, + "grad_norm": 1.586008570812673, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 57803 + }, + { + "epoch": 0.57804, + "grad_norm": 1.5765125137885114, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 57804 + }, + { + "epoch": 0.57805, + "grad_norm": 1.5917929608773185, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 57805 + }, + { + "epoch": 0.57806, + "grad_norm": 1.69478344839484, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 57806 + }, + { + "epoch": 0.57807, + "grad_norm": 1.4160784891185485, + "learning_rate": 0.003, + "loss": 4.031, + "step": 57807 + }, + { + "epoch": 0.57808, + "grad_norm": 1.8369526162999388, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 57808 + }, + { + "epoch": 0.57809, + "grad_norm": 1.5642013982350282, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 57809 + }, + { + "epoch": 0.5781, + "grad_norm": 1.4866635157165438, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 57810 + }, + { + "epoch": 0.57811, + "grad_norm": 1.2558144219242446, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 57811 + }, + { + "epoch": 0.57812, + "grad_norm": 1.770662143145302, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 57812 + }, + { + "epoch": 0.57813, + "grad_norm": 1.4216813556667915, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 57813 + }, + { + "epoch": 0.57814, + "grad_norm": 1.685249433452263, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 57814 + }, + { + "epoch": 0.57815, + "grad_norm": 1.8987541168597017, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 57815 + }, + { + "epoch": 0.57816, + "grad_norm": 1.4458969968646964, + "learning_rate": 0.003, + "loss": 4.022, + "step": 57816 + }, + { + "epoch": 0.57817, + "grad_norm": 1.9975301295495387, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 57817 + }, + { + "epoch": 0.57818, + "grad_norm": 1.3148687301912767, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 57818 + }, + { + "epoch": 0.57819, + "grad_norm": 1.540686142012626, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 57819 + }, + { + "epoch": 0.5782, + "grad_norm": 1.823927029745938, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 57820 + }, + { + "epoch": 0.57821, + "grad_norm": 1.3432382082063796, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 57821 + }, + { + "epoch": 0.57822, + "grad_norm": 1.5716601307704416, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 57822 + }, + { + "epoch": 0.57823, + "grad_norm": 1.3547827896201685, + "learning_rate": 0.003, + "loss": 4.045, + "step": 57823 + }, + { + "epoch": 0.57824, + "grad_norm": 1.6847948930136518, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 57824 + }, + { + "epoch": 0.57825, + "grad_norm": 1.6655825425344801, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 57825 + }, + { + "epoch": 0.57826, + "grad_norm": 1.4059719736060179, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 57826 + }, + { + "epoch": 0.57827, + "grad_norm": 1.6103213160090972, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 57827 + }, + { + "epoch": 0.57828, + "grad_norm": 1.6086886753843463, + "learning_rate": 0.003, + "loss": 4.07, + "step": 57828 + }, + { + "epoch": 0.57829, + "grad_norm": 1.85302276054908, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 57829 + }, + { + "epoch": 0.5783, + "grad_norm": 1.468821601047182, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 57830 + }, + { + "epoch": 0.57831, + "grad_norm": 1.578955462498401, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 57831 + }, + { + "epoch": 0.57832, + "grad_norm": 2.1200354265318437, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 57832 + }, + { + "epoch": 0.57833, + "grad_norm": 1.2859294565846715, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 57833 + }, + { + "epoch": 0.57834, + "grad_norm": 1.8672580857076568, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 57834 + }, + { + "epoch": 0.57835, + "grad_norm": 1.3541139123307624, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57835 + }, + { + "epoch": 0.57836, + "grad_norm": 1.5586809716418228, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57836 + }, + { + "epoch": 0.57837, + "grad_norm": 1.1905178105305365, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 57837 + }, + { + "epoch": 0.57838, + "grad_norm": 1.5735360656905408, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 57838 + }, + { + "epoch": 0.57839, + "grad_norm": 1.6740125606176866, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 57839 + }, + { + "epoch": 0.5784, + "grad_norm": 1.9903383724187331, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 57840 + }, + { + "epoch": 0.57841, + "grad_norm": 1.3727889153408326, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 57841 + }, + { + "epoch": 0.57842, + "grad_norm": 1.4808012111163655, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 57842 + }, + { + "epoch": 0.57843, + "grad_norm": 1.7659444526341435, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 57843 + }, + { + "epoch": 0.57844, + "grad_norm": 1.6411175066801587, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 57844 + }, + { + "epoch": 0.57845, + "grad_norm": 1.290535332030528, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 57845 + }, + { + "epoch": 0.57846, + "grad_norm": 1.9482116531209095, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 57846 + }, + { + "epoch": 0.57847, + "grad_norm": 1.1317519952877326, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 57847 + }, + { + "epoch": 0.57848, + "grad_norm": 1.7545461229029322, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 57848 + }, + { + "epoch": 0.57849, + "grad_norm": 1.5683663108601047, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 57849 + }, + { + "epoch": 0.5785, + "grad_norm": 1.8055320685057754, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 57850 + }, + { + "epoch": 0.57851, + "grad_norm": 1.6879838702303944, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 57851 + }, + { + "epoch": 0.57852, + "grad_norm": 1.5340532793605588, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 57852 + }, + { + "epoch": 0.57853, + "grad_norm": 1.5183200353415915, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 57853 + }, + { + "epoch": 0.57854, + "grad_norm": 1.4035418867472502, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 57854 + }, + { + "epoch": 0.57855, + "grad_norm": 1.8156748371470055, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 57855 + }, + { + "epoch": 0.57856, + "grad_norm": 1.4249308357811603, + "learning_rate": 0.003, + "loss": 4.022, + "step": 57856 + }, + { + "epoch": 0.57857, + "grad_norm": 1.7197359683589033, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 57857 + }, + { + "epoch": 0.57858, + "grad_norm": 1.3715698543435417, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 57858 + }, + { + "epoch": 0.57859, + "grad_norm": 1.6690266183024702, + "learning_rate": 0.003, + "loss": 4.011, + "step": 57859 + }, + { + "epoch": 0.5786, + "grad_norm": 1.5636561583407689, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 57860 + }, + { + "epoch": 0.57861, + "grad_norm": 1.4593304323761036, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 57861 + }, + { + "epoch": 0.57862, + "grad_norm": 1.8874977911757573, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 57862 + }, + { + "epoch": 0.57863, + "grad_norm": 1.4496499281287203, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 57863 + }, + { + "epoch": 0.57864, + "grad_norm": 1.4077785443986865, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 57864 + }, + { + "epoch": 0.57865, + "grad_norm": 1.734933114974535, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 57865 + }, + { + "epoch": 0.57866, + "grad_norm": 1.576140869140575, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 57866 + }, + { + "epoch": 0.57867, + "grad_norm": 1.372278722655877, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 57867 + }, + { + "epoch": 0.57868, + "grad_norm": 1.6517161352846237, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 57868 + }, + { + "epoch": 0.57869, + "grad_norm": 1.5593125462314263, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 57869 + }, + { + "epoch": 0.5787, + "grad_norm": 1.4587010082925551, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 57870 + }, + { + "epoch": 0.57871, + "grad_norm": 1.6224079180664828, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 57871 + }, + { + "epoch": 0.57872, + "grad_norm": 1.6284139965127595, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 57872 + }, + { + "epoch": 0.57873, + "grad_norm": 1.4084904481810683, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 57873 + }, + { + "epoch": 0.57874, + "grad_norm": 1.9210553612460137, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 57874 + }, + { + "epoch": 0.57875, + "grad_norm": 1.2594552742628708, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 57875 + }, + { + "epoch": 0.57876, + "grad_norm": 1.9494031224245343, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 57876 + }, + { + "epoch": 0.57877, + "grad_norm": 1.8655606743125797, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 57877 + }, + { + "epoch": 0.57878, + "grad_norm": 1.3885828683523926, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57878 + }, + { + "epoch": 0.57879, + "grad_norm": 2.0184630557743977, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 57879 + }, + { + "epoch": 0.5788, + "grad_norm": 1.2668412853258895, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 57880 + }, + { + "epoch": 0.57881, + "grad_norm": 1.588323531161569, + "learning_rate": 0.003, + "loss": 4.033, + "step": 57881 + }, + { + "epoch": 0.57882, + "grad_norm": 1.5905294932972718, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 57882 + }, + { + "epoch": 0.57883, + "grad_norm": 1.5683086775264323, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 57883 + }, + { + "epoch": 0.57884, + "grad_norm": 1.5726648289670249, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 57884 + }, + { + "epoch": 0.57885, + "grad_norm": 1.7198199419502795, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 57885 + }, + { + "epoch": 0.57886, + "grad_norm": 1.611574219311478, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 57886 + }, + { + "epoch": 0.57887, + "grad_norm": 1.4482633752441194, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 57887 + }, + { + "epoch": 0.57888, + "grad_norm": 1.7645849362869959, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 57888 + }, + { + "epoch": 0.57889, + "grad_norm": 1.3610903342959257, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 57889 + }, + { + "epoch": 0.5789, + "grad_norm": 1.8913010554755894, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 57890 + }, + { + "epoch": 0.57891, + "grad_norm": 1.2648352792642863, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 57891 + }, + { + "epoch": 0.57892, + "grad_norm": 2.009267225659109, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 57892 + }, + { + "epoch": 0.57893, + "grad_norm": 1.417945756677504, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 57893 + }, + { + "epoch": 0.57894, + "grad_norm": 1.8141636194144282, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 57894 + }, + { + "epoch": 0.57895, + "grad_norm": 1.616857287392823, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 57895 + }, + { + "epoch": 0.57896, + "grad_norm": 1.2177415273769243, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 57896 + }, + { + "epoch": 0.57897, + "grad_norm": 1.5194573286981325, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 57897 + }, + { + "epoch": 0.57898, + "grad_norm": 1.5510474207669203, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 57898 + }, + { + "epoch": 0.57899, + "grad_norm": 1.6038874370494194, + "learning_rate": 0.003, + "loss": 4.063, + "step": 57899 + }, + { + "epoch": 0.579, + "grad_norm": 1.382997114744454, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 57900 + }, + { + "epoch": 0.57901, + "grad_norm": 1.7376773679173931, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 57901 + }, + { + "epoch": 0.57902, + "grad_norm": 1.478307434534057, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 57902 + }, + { + "epoch": 0.57903, + "grad_norm": 1.60889017380926, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 57903 + }, + { + "epoch": 0.57904, + "grad_norm": 2.315731869305274, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 57904 + }, + { + "epoch": 0.57905, + "grad_norm": 1.1861959866811476, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 57905 + }, + { + "epoch": 0.57906, + "grad_norm": 1.6354279710187403, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 57906 + }, + { + "epoch": 0.57907, + "grad_norm": 1.433341380721545, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 57907 + }, + { + "epoch": 0.57908, + "grad_norm": 1.8672137390326713, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 57908 + }, + { + "epoch": 0.57909, + "grad_norm": 1.4780415259897874, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 57909 + }, + { + "epoch": 0.5791, + "grad_norm": 1.3314977389541982, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 57910 + }, + { + "epoch": 0.57911, + "grad_norm": 2.188265268493814, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 57911 + }, + { + "epoch": 0.57912, + "grad_norm": 1.407688080244486, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 57912 + }, + { + "epoch": 0.57913, + "grad_norm": 1.7542215565389796, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 57913 + }, + { + "epoch": 0.57914, + "grad_norm": 1.4980590528385183, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 57914 + }, + { + "epoch": 0.57915, + "grad_norm": 1.4928437667565924, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 57915 + }, + { + "epoch": 0.57916, + "grad_norm": 1.3921194168626516, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 57916 + }, + { + "epoch": 0.57917, + "grad_norm": 1.7978639813724782, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 57917 + }, + { + "epoch": 0.57918, + "grad_norm": 1.5161112161587689, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 57918 + }, + { + "epoch": 0.57919, + "grad_norm": 1.4671809968848368, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 57919 + }, + { + "epoch": 0.5792, + "grad_norm": 1.8369989447857282, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 57920 + }, + { + "epoch": 0.57921, + "grad_norm": 1.2820285452065794, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57921 + }, + { + "epoch": 0.57922, + "grad_norm": 1.2600519745744423, + "learning_rate": 0.003, + "loss": 4.027, + "step": 57922 + }, + { + "epoch": 0.57923, + "grad_norm": 1.77222829601755, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 57923 + }, + { + "epoch": 0.57924, + "grad_norm": 1.7375435640979293, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 57924 + }, + { + "epoch": 0.57925, + "grad_norm": 1.2259573432429978, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 57925 + }, + { + "epoch": 0.57926, + "grad_norm": 1.64868785166965, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 57926 + }, + { + "epoch": 0.57927, + "grad_norm": 1.7952348570701098, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57927 + }, + { + "epoch": 0.57928, + "grad_norm": 1.23615906921498, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 57928 + }, + { + "epoch": 0.57929, + "grad_norm": 2.013556858311337, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57929 + }, + { + "epoch": 0.5793, + "grad_norm": 1.1790093597809863, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 57930 + }, + { + "epoch": 0.57931, + "grad_norm": 2.0252762178200614, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 57931 + }, + { + "epoch": 0.57932, + "grad_norm": 1.475105232284004, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 57932 + }, + { + "epoch": 0.57933, + "grad_norm": 1.729334823258087, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 57933 + }, + { + "epoch": 0.57934, + "grad_norm": 1.6331613940028886, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 57934 + }, + { + "epoch": 0.57935, + "grad_norm": 1.611877484040477, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 57935 + }, + { + "epoch": 0.57936, + "grad_norm": 1.5935153557141482, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 57936 + }, + { + "epoch": 0.57937, + "grad_norm": 1.518137288987087, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 57937 + }, + { + "epoch": 0.57938, + "grad_norm": 1.9832267591561012, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 57938 + }, + { + "epoch": 0.57939, + "grad_norm": 1.2807407260607901, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 57939 + }, + { + "epoch": 0.5794, + "grad_norm": 1.7015853906048422, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 57940 + }, + { + "epoch": 0.57941, + "grad_norm": 1.5881218640088879, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 57941 + }, + { + "epoch": 0.57942, + "grad_norm": 1.7026849276129377, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 57942 + }, + { + "epoch": 0.57943, + "grad_norm": 1.4880223748175796, + "learning_rate": 0.003, + "loss": 3.998, + "step": 57943 + }, + { + "epoch": 0.57944, + "grad_norm": 1.4714462943685425, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 57944 + }, + { + "epoch": 0.57945, + "grad_norm": 1.4892261503018127, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 57945 + }, + { + "epoch": 0.57946, + "grad_norm": 1.455052411394938, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 57946 + }, + { + "epoch": 0.57947, + "grad_norm": 1.8742051753298539, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 57947 + }, + { + "epoch": 0.57948, + "grad_norm": 1.7694063897311287, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 57948 + }, + { + "epoch": 0.57949, + "grad_norm": 1.4586149935690718, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 57949 + }, + { + "epoch": 0.5795, + "grad_norm": 1.763895020313622, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 57950 + }, + { + "epoch": 0.57951, + "grad_norm": 1.5509286462931295, + "learning_rate": 0.003, + "loss": 4.047, + "step": 57951 + }, + { + "epoch": 0.57952, + "grad_norm": 1.3444520735060639, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 57952 + }, + { + "epoch": 0.57953, + "grad_norm": 1.7039122516272365, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 57953 + }, + { + "epoch": 0.57954, + "grad_norm": 1.6525686269033542, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 57954 + }, + { + "epoch": 0.57955, + "grad_norm": 1.520889863773035, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 57955 + }, + { + "epoch": 0.57956, + "grad_norm": 1.6302190627148823, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 57956 + }, + { + "epoch": 0.57957, + "grad_norm": 1.6543719306906821, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 57957 + }, + { + "epoch": 0.57958, + "grad_norm": 1.3396206021354915, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 57958 + }, + { + "epoch": 0.57959, + "grad_norm": 1.7061198919060192, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 57959 + }, + { + "epoch": 0.5796, + "grad_norm": 1.3155012061812594, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 57960 + }, + { + "epoch": 0.57961, + "grad_norm": 1.8982510669516306, + "learning_rate": 0.003, + "loss": 4.034, + "step": 57961 + }, + { + "epoch": 0.57962, + "grad_norm": 1.6080045732189365, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 57962 + }, + { + "epoch": 0.57963, + "grad_norm": 1.2866374348937017, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 57963 + }, + { + "epoch": 0.57964, + "grad_norm": 1.7533784690397092, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 57964 + }, + { + "epoch": 0.57965, + "grad_norm": 1.5174329602784193, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 57965 + }, + { + "epoch": 0.57966, + "grad_norm": 1.6054838576766604, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 57966 + }, + { + "epoch": 0.57967, + "grad_norm": 1.5874483832255644, + "learning_rate": 0.003, + "loss": 4.031, + "step": 57967 + }, + { + "epoch": 0.57968, + "grad_norm": 1.395788903468246, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 57968 + }, + { + "epoch": 0.57969, + "grad_norm": 1.539863132545249, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 57969 + }, + { + "epoch": 0.5797, + "grad_norm": 1.4950164732777513, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 57970 + }, + { + "epoch": 0.57971, + "grad_norm": 1.597300691747538, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 57971 + }, + { + "epoch": 0.57972, + "grad_norm": 1.4402999744024458, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 57972 + }, + { + "epoch": 0.57973, + "grad_norm": 1.6771714712868784, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 57973 + }, + { + "epoch": 0.57974, + "grad_norm": 1.2897298366457925, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 57974 + }, + { + "epoch": 0.57975, + "grad_norm": 1.7093893125914104, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 57975 + }, + { + "epoch": 0.57976, + "grad_norm": 1.3900731684810255, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 57976 + }, + { + "epoch": 0.57977, + "grad_norm": 2.1020548987910432, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 57977 + }, + { + "epoch": 0.57978, + "grad_norm": 1.3698697893363554, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 57978 + }, + { + "epoch": 0.57979, + "grad_norm": 1.694502110183097, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 57979 + }, + { + "epoch": 0.5798, + "grad_norm": 1.7139489190963464, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 57980 + }, + { + "epoch": 0.57981, + "grad_norm": 1.3493223491555253, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 57981 + }, + { + "epoch": 0.57982, + "grad_norm": 1.6862210277954555, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 57982 + }, + { + "epoch": 0.57983, + "grad_norm": 1.6234900550665863, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 57983 + }, + { + "epoch": 0.57984, + "grad_norm": 1.338225984129839, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 57984 + }, + { + "epoch": 0.57985, + "grad_norm": 1.5909318045838776, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 57985 + }, + { + "epoch": 0.57986, + "grad_norm": 1.5802125307977177, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 57986 + }, + { + "epoch": 0.57987, + "grad_norm": 1.4778517543872465, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 57987 + }, + { + "epoch": 0.57988, + "grad_norm": 1.4864586142227092, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 57988 + }, + { + "epoch": 0.57989, + "grad_norm": 1.72558480236629, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 57989 + }, + { + "epoch": 0.5799, + "grad_norm": 1.7559002205196381, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 57990 + }, + { + "epoch": 0.57991, + "grad_norm": 1.8586828186589, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 57991 + }, + { + "epoch": 0.57992, + "grad_norm": 1.7552926233637571, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 57992 + }, + { + "epoch": 0.57993, + "grad_norm": 1.4841851491363311, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 57993 + }, + { + "epoch": 0.57994, + "grad_norm": 1.7800720178367089, + "learning_rate": 0.003, + "loss": 4.021, + "step": 57994 + }, + { + "epoch": 0.57995, + "grad_norm": 1.3372882781567366, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 57995 + }, + { + "epoch": 0.57996, + "grad_norm": 1.6377122386413236, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 57996 + }, + { + "epoch": 0.57997, + "grad_norm": 1.527477342803634, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 57997 + }, + { + "epoch": 0.57998, + "grad_norm": 2.1846983327470317, + "learning_rate": 0.003, + "loss": 4.039, + "step": 57998 + }, + { + "epoch": 0.57999, + "grad_norm": 1.1411603904571366, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 57999 + }, + { + "epoch": 0.58, + "grad_norm": 1.815586569924389, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 58000 + }, + { + "epoch": 0.58001, + "grad_norm": 1.5563180639122063, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 58001 + }, + { + "epoch": 0.58002, + "grad_norm": 1.460470439835991, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 58002 + }, + { + "epoch": 0.58003, + "grad_norm": 1.672736301441314, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 58003 + }, + { + "epoch": 0.58004, + "grad_norm": 1.403481933160379, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 58004 + }, + { + "epoch": 0.58005, + "grad_norm": 1.576393064040786, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 58005 + }, + { + "epoch": 0.58006, + "grad_norm": 1.4781379687198282, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 58006 + }, + { + "epoch": 0.58007, + "grad_norm": 1.5429281571582667, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58007 + }, + { + "epoch": 0.58008, + "grad_norm": 1.6042733948075896, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 58008 + }, + { + "epoch": 0.58009, + "grad_norm": 1.51797449862489, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 58009 + }, + { + "epoch": 0.5801, + "grad_norm": 1.920315388915656, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 58010 + }, + { + "epoch": 0.58011, + "grad_norm": 1.6610186581421191, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 58011 + }, + { + "epoch": 0.58012, + "grad_norm": 1.3883109814358112, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 58012 + }, + { + "epoch": 0.58013, + "grad_norm": 1.845669737282771, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58013 + }, + { + "epoch": 0.58014, + "grad_norm": 1.6918778927206946, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 58014 + }, + { + "epoch": 0.58015, + "grad_norm": 1.3733646770615746, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 58015 + }, + { + "epoch": 0.58016, + "grad_norm": 1.6877535909988544, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 58016 + }, + { + "epoch": 0.58017, + "grad_norm": 1.298758252968307, + "learning_rate": 0.003, + "loss": 4.027, + "step": 58017 + }, + { + "epoch": 0.58018, + "grad_norm": 1.722609818210488, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 58018 + }, + { + "epoch": 0.58019, + "grad_norm": 1.2218181016280785, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 58019 + }, + { + "epoch": 0.5802, + "grad_norm": 1.589541953473173, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 58020 + }, + { + "epoch": 0.58021, + "grad_norm": 1.5376495626373623, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 58021 + }, + { + "epoch": 0.58022, + "grad_norm": 1.9109742048726326, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 58022 + }, + { + "epoch": 0.58023, + "grad_norm": 1.4024485523575252, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 58023 + }, + { + "epoch": 0.58024, + "grad_norm": 1.5974284358508413, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 58024 + }, + { + "epoch": 0.58025, + "grad_norm": 1.7261412994231382, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 58025 + }, + { + "epoch": 0.58026, + "grad_norm": 1.769985118327955, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 58026 + }, + { + "epoch": 0.58027, + "grad_norm": 1.3682918954892151, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 58027 + }, + { + "epoch": 0.58028, + "grad_norm": 1.644450973484525, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 58028 + }, + { + "epoch": 0.58029, + "grad_norm": 1.5420055003036766, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 58029 + }, + { + "epoch": 0.5803, + "grad_norm": 1.7243802833572794, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 58030 + }, + { + "epoch": 0.58031, + "grad_norm": 1.4484551277237276, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 58031 + }, + { + "epoch": 0.58032, + "grad_norm": 1.4887205327798494, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 58032 + }, + { + "epoch": 0.58033, + "grad_norm": 1.8208697029662555, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 58033 + }, + { + "epoch": 0.58034, + "grad_norm": 1.5322749858554514, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 58034 + }, + { + "epoch": 0.58035, + "grad_norm": 1.6506995260326116, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 58035 + }, + { + "epoch": 0.58036, + "grad_norm": 1.3729692846812325, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 58036 + }, + { + "epoch": 0.58037, + "grad_norm": 1.7073107827964158, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 58037 + }, + { + "epoch": 0.58038, + "grad_norm": 1.4050470030872815, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 58038 + }, + { + "epoch": 0.58039, + "grad_norm": 1.5835071281674298, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 58039 + }, + { + "epoch": 0.5804, + "grad_norm": 1.8497342790097242, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 58040 + }, + { + "epoch": 0.58041, + "grad_norm": 1.5838824285962774, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 58041 + }, + { + "epoch": 0.58042, + "grad_norm": 2.1086231204924224, + "learning_rate": 0.003, + "loss": 4.054, + "step": 58042 + }, + { + "epoch": 0.58043, + "grad_norm": 1.4824191015506343, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 58043 + }, + { + "epoch": 0.58044, + "grad_norm": 1.447378396242814, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 58044 + }, + { + "epoch": 0.58045, + "grad_norm": 1.9403604989142993, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 58045 + }, + { + "epoch": 0.58046, + "grad_norm": 1.2906900042291922, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 58046 + }, + { + "epoch": 0.58047, + "grad_norm": 1.9178732964548633, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 58047 + }, + { + "epoch": 0.58048, + "grad_norm": 1.667705403556363, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 58048 + }, + { + "epoch": 0.58049, + "grad_norm": 1.3536516998166284, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 58049 + }, + { + "epoch": 0.5805, + "grad_norm": 1.7750841693722401, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 58050 + }, + { + "epoch": 0.58051, + "grad_norm": 1.3618060756288843, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 58051 + }, + { + "epoch": 0.58052, + "grad_norm": 1.639346560478027, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 58052 + }, + { + "epoch": 0.58053, + "grad_norm": 1.3814270526595362, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 58053 + }, + { + "epoch": 0.58054, + "grad_norm": 1.6772436741997034, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 58054 + }, + { + "epoch": 0.58055, + "grad_norm": 1.5179251035511383, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 58055 + }, + { + "epoch": 0.58056, + "grad_norm": 1.9079680624193924, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 58056 + }, + { + "epoch": 0.58057, + "grad_norm": 1.6997313065381467, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 58057 + }, + { + "epoch": 0.58058, + "grad_norm": 1.6442690234539838, + "learning_rate": 0.003, + "loss": 4.062, + "step": 58058 + }, + { + "epoch": 0.58059, + "grad_norm": 1.6495983957049343, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 58059 + }, + { + "epoch": 0.5806, + "grad_norm": 1.2013021504169503, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 58060 + }, + { + "epoch": 0.58061, + "grad_norm": 1.7802147669222421, + "learning_rate": 0.003, + "loss": 4.032, + "step": 58061 + }, + { + "epoch": 0.58062, + "grad_norm": 1.3894944986606315, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 58062 + }, + { + "epoch": 0.58063, + "grad_norm": 1.7986465374160439, + "learning_rate": 0.003, + "loss": 4.005, + "step": 58063 + }, + { + "epoch": 0.58064, + "grad_norm": 1.5192747453934894, + "learning_rate": 0.003, + "loss": 4.05, + "step": 58064 + }, + { + "epoch": 0.58065, + "grad_norm": 1.4828883205724772, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 58065 + }, + { + "epoch": 0.58066, + "grad_norm": 1.4967888975055212, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 58066 + }, + { + "epoch": 0.58067, + "grad_norm": 1.8197181862223515, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 58067 + }, + { + "epoch": 0.58068, + "grad_norm": 1.5367384346737765, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 58068 + }, + { + "epoch": 0.58069, + "grad_norm": 1.5224917517233358, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 58069 + }, + { + "epoch": 0.5807, + "grad_norm": 1.7549508484050003, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 58070 + }, + { + "epoch": 0.58071, + "grad_norm": 1.4264472978774596, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 58071 + }, + { + "epoch": 0.58072, + "grad_norm": 1.6993740054505535, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 58072 + }, + { + "epoch": 0.58073, + "grad_norm": 1.3302095580869628, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 58073 + }, + { + "epoch": 0.58074, + "grad_norm": 1.5866294662404874, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 58074 + }, + { + "epoch": 0.58075, + "grad_norm": 1.4612864052247672, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 58075 + }, + { + "epoch": 0.58076, + "grad_norm": 1.7186074025486495, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 58076 + }, + { + "epoch": 0.58077, + "grad_norm": 1.5275392525127058, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 58077 + }, + { + "epoch": 0.58078, + "grad_norm": 1.6314276556785596, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 58078 + }, + { + "epoch": 0.58079, + "grad_norm": 1.6103928972397628, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 58079 + }, + { + "epoch": 0.5808, + "grad_norm": 1.6837045591558346, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 58080 + }, + { + "epoch": 0.58081, + "grad_norm": 1.7986585732871172, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 58081 + }, + { + "epoch": 0.58082, + "grad_norm": 1.2891396882421358, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 58082 + }, + { + "epoch": 0.58083, + "grad_norm": 1.807375971891515, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 58083 + }, + { + "epoch": 0.58084, + "grad_norm": 1.455920992386902, + "learning_rate": 0.003, + "loss": 4.045, + "step": 58084 + }, + { + "epoch": 0.58085, + "grad_norm": 1.5873608392820546, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 58085 + }, + { + "epoch": 0.58086, + "grad_norm": 1.7718523539205662, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 58086 + }, + { + "epoch": 0.58087, + "grad_norm": 1.4138786348431038, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 58087 + }, + { + "epoch": 0.58088, + "grad_norm": 1.9620072046397767, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 58088 + }, + { + "epoch": 0.58089, + "grad_norm": 1.4826570742226373, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 58089 + }, + { + "epoch": 0.5809, + "grad_norm": 1.5127035202619157, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 58090 + }, + { + "epoch": 0.58091, + "grad_norm": 1.5666923224065925, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 58091 + }, + { + "epoch": 0.58092, + "grad_norm": 1.6719036456548975, + "learning_rate": 0.003, + "loss": 4.019, + "step": 58092 + }, + { + "epoch": 0.58093, + "grad_norm": 1.5727999018431131, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 58093 + }, + { + "epoch": 0.58094, + "grad_norm": 1.5261150154485554, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 58094 + }, + { + "epoch": 0.58095, + "grad_norm": 1.7743736900241918, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 58095 + }, + { + "epoch": 0.58096, + "grad_norm": 1.2128132148443842, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 58096 + }, + { + "epoch": 0.58097, + "grad_norm": 1.6830282381212431, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 58097 + }, + { + "epoch": 0.58098, + "grad_norm": 1.1907944924859208, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 58098 + }, + { + "epoch": 0.58099, + "grad_norm": 1.5230100589328137, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 58099 + }, + { + "epoch": 0.581, + "grad_norm": 1.4267889204841213, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 58100 + }, + { + "epoch": 0.58101, + "grad_norm": 1.9505714310929243, + "learning_rate": 0.003, + "loss": 4.044, + "step": 58101 + }, + { + "epoch": 0.58102, + "grad_norm": 1.3286906633375246, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 58102 + }, + { + "epoch": 0.58103, + "grad_norm": 1.9197073300478038, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 58103 + }, + { + "epoch": 0.58104, + "grad_norm": 1.4645748820109008, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 58104 + }, + { + "epoch": 0.58105, + "grad_norm": 1.7152019824236249, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 58105 + }, + { + "epoch": 0.58106, + "grad_norm": 1.608578613605806, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 58106 + }, + { + "epoch": 0.58107, + "grad_norm": 1.5506691061001374, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 58107 + }, + { + "epoch": 0.58108, + "grad_norm": 1.8160864767861338, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 58108 + }, + { + "epoch": 0.58109, + "grad_norm": 1.6007337938929187, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 58109 + }, + { + "epoch": 0.5811, + "grad_norm": 1.3965753486638997, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 58110 + }, + { + "epoch": 0.58111, + "grad_norm": 1.6138471829501972, + "learning_rate": 0.003, + "loss": 4.017, + "step": 58111 + }, + { + "epoch": 0.58112, + "grad_norm": 1.73612368169382, + "learning_rate": 0.003, + "loss": 4.025, + "step": 58112 + }, + { + "epoch": 0.58113, + "grad_norm": 1.264575392306464, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 58113 + }, + { + "epoch": 0.58114, + "grad_norm": 1.7194336926803468, + "learning_rate": 0.003, + "loss": 4.056, + "step": 58114 + }, + { + "epoch": 0.58115, + "grad_norm": 1.3412066097586075, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 58115 + }, + { + "epoch": 0.58116, + "grad_norm": 1.595282360030179, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 58116 + }, + { + "epoch": 0.58117, + "grad_norm": 1.3854046236172897, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 58117 + }, + { + "epoch": 0.58118, + "grad_norm": 1.8116458640416706, + "learning_rate": 0.003, + "loss": 4.064, + "step": 58118 + }, + { + "epoch": 0.58119, + "grad_norm": 1.429924510714172, + "learning_rate": 0.003, + "loss": 4.06, + "step": 58119 + }, + { + "epoch": 0.5812, + "grad_norm": 1.4809207887208056, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 58120 + }, + { + "epoch": 0.58121, + "grad_norm": 1.8385965944860039, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 58121 + }, + { + "epoch": 0.58122, + "grad_norm": 1.7403858512963464, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 58122 + }, + { + "epoch": 0.58123, + "grad_norm": 2.3260413166009695, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 58123 + }, + { + "epoch": 0.58124, + "grad_norm": 1.4180763334259492, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 58124 + }, + { + "epoch": 0.58125, + "grad_norm": 1.6172484548320296, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 58125 + }, + { + "epoch": 0.58126, + "grad_norm": 1.5712629254366886, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 58126 + }, + { + "epoch": 0.58127, + "grad_norm": 1.709464968184306, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 58127 + }, + { + "epoch": 0.58128, + "grad_norm": 1.6039726547678221, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 58128 + }, + { + "epoch": 0.58129, + "grad_norm": 1.3679519685060195, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 58129 + }, + { + "epoch": 0.5813, + "grad_norm": 1.6204092553458274, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 58130 + }, + { + "epoch": 0.58131, + "grad_norm": 1.4680872598028973, + "learning_rate": 0.003, + "loss": 4.047, + "step": 58131 + }, + { + "epoch": 0.58132, + "grad_norm": 1.5899516644125713, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 58132 + }, + { + "epoch": 0.58133, + "grad_norm": 1.8042795207369458, + "learning_rate": 0.003, + "loss": 3.996, + "step": 58133 + }, + { + "epoch": 0.58134, + "grad_norm": 1.4147325952026168, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 58134 + }, + { + "epoch": 0.58135, + "grad_norm": 1.8515494974489197, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 58135 + }, + { + "epoch": 0.58136, + "grad_norm": 1.2264273589563055, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 58136 + }, + { + "epoch": 0.58137, + "grad_norm": 1.5262683125539795, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 58137 + }, + { + "epoch": 0.58138, + "grad_norm": 1.5990811989753535, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 58138 + }, + { + "epoch": 0.58139, + "grad_norm": 1.4877414104515092, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 58139 + }, + { + "epoch": 0.5814, + "grad_norm": 1.9321004809302684, + "learning_rate": 0.003, + "loss": 4.042, + "step": 58140 + }, + { + "epoch": 0.58141, + "grad_norm": 1.7038720855680791, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 58141 + }, + { + "epoch": 0.58142, + "grad_norm": 1.3278521916975914, + "learning_rate": 0.003, + "loss": 4.041, + "step": 58142 + }, + { + "epoch": 0.58143, + "grad_norm": 2.2678439112282454, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 58143 + }, + { + "epoch": 0.58144, + "grad_norm": 1.354137012852232, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 58144 + }, + { + "epoch": 0.58145, + "grad_norm": 1.4393829920506307, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 58145 + }, + { + "epoch": 0.58146, + "grad_norm": 1.998733818412058, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 58146 + }, + { + "epoch": 0.58147, + "grad_norm": 1.3696336563569134, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 58147 + }, + { + "epoch": 0.58148, + "grad_norm": 1.4109636397843228, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 58148 + }, + { + "epoch": 0.58149, + "grad_norm": 1.6792728792059188, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 58149 + }, + { + "epoch": 0.5815, + "grad_norm": 1.3464988827575763, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 58150 + }, + { + "epoch": 0.58151, + "grad_norm": 1.6176763313470062, + "learning_rate": 0.003, + "loss": 4.028, + "step": 58151 + }, + { + "epoch": 0.58152, + "grad_norm": 1.621499714282393, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 58152 + }, + { + "epoch": 0.58153, + "grad_norm": 1.4766345135748098, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 58153 + }, + { + "epoch": 0.58154, + "grad_norm": 1.682000620483449, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 58154 + }, + { + "epoch": 0.58155, + "grad_norm": 1.5883445888193495, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 58155 + }, + { + "epoch": 0.58156, + "grad_norm": 2.06483636480425, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 58156 + }, + { + "epoch": 0.58157, + "grad_norm": 1.3751151920302864, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 58157 + }, + { + "epoch": 0.58158, + "grad_norm": 1.4740563649009544, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 58158 + }, + { + "epoch": 0.58159, + "grad_norm": 2.021000825066195, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 58159 + }, + { + "epoch": 0.5816, + "grad_norm": 1.4364095602533704, + "learning_rate": 0.003, + "loss": 4.053, + "step": 58160 + }, + { + "epoch": 0.58161, + "grad_norm": 1.2721326432248012, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 58161 + }, + { + "epoch": 0.58162, + "grad_norm": 1.7228206432462172, + "learning_rate": 0.003, + "loss": 4.038, + "step": 58162 + }, + { + "epoch": 0.58163, + "grad_norm": 1.4773468577728526, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 58163 + }, + { + "epoch": 0.58164, + "grad_norm": 1.7452694022589137, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 58164 + }, + { + "epoch": 0.58165, + "grad_norm": 1.3436792783789646, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 58165 + }, + { + "epoch": 0.58166, + "grad_norm": 1.7076581837123053, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 58166 + }, + { + "epoch": 0.58167, + "grad_norm": 1.605312332613443, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 58167 + }, + { + "epoch": 0.58168, + "grad_norm": 1.5741190953551392, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 58168 + }, + { + "epoch": 0.58169, + "grad_norm": 1.7612690952311914, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 58169 + }, + { + "epoch": 0.5817, + "grad_norm": 1.5671933861032632, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 58170 + }, + { + "epoch": 0.58171, + "grad_norm": 1.5400328463741948, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 58171 + }, + { + "epoch": 0.58172, + "grad_norm": 1.558749583334564, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 58172 + }, + { + "epoch": 0.58173, + "grad_norm": 1.5516877642702556, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 58173 + }, + { + "epoch": 0.58174, + "grad_norm": 1.5443858505253096, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 58174 + }, + { + "epoch": 0.58175, + "grad_norm": 1.4186960013290375, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58175 + }, + { + "epoch": 0.58176, + "grad_norm": 1.5888641863462933, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 58176 + }, + { + "epoch": 0.58177, + "grad_norm": 1.6292551017157348, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 58177 + }, + { + "epoch": 0.58178, + "grad_norm": 1.4970683048105349, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 58178 + }, + { + "epoch": 0.58179, + "grad_norm": 1.670797927468529, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 58179 + }, + { + "epoch": 0.5818, + "grad_norm": 1.1847789523463341, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 58180 + }, + { + "epoch": 0.58181, + "grad_norm": 1.5898582615354842, + "learning_rate": 0.003, + "loss": 4.075, + "step": 58181 + }, + { + "epoch": 0.58182, + "grad_norm": 1.644065543134641, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58182 + }, + { + "epoch": 0.58183, + "grad_norm": 1.7620934526079441, + "learning_rate": 0.003, + "loss": 4.024, + "step": 58183 + }, + { + "epoch": 0.58184, + "grad_norm": 1.8543938380107265, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 58184 + }, + { + "epoch": 0.58185, + "grad_norm": 1.4113867579923334, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 58185 + }, + { + "epoch": 0.58186, + "grad_norm": 1.684781502349199, + "learning_rate": 0.003, + "loss": 4.039, + "step": 58186 + }, + { + "epoch": 0.58187, + "grad_norm": 1.7126667453477016, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 58187 + }, + { + "epoch": 0.58188, + "grad_norm": 1.5553519837042287, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 58188 + }, + { + "epoch": 0.58189, + "grad_norm": 1.5832303981437408, + "learning_rate": 0.003, + "loss": 4.052, + "step": 58189 + }, + { + "epoch": 0.5819, + "grad_norm": 1.8221289324617427, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 58190 + }, + { + "epoch": 0.58191, + "grad_norm": 1.478915128912362, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 58191 + }, + { + "epoch": 0.58192, + "grad_norm": 1.5848217911203408, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 58192 + }, + { + "epoch": 0.58193, + "grad_norm": 1.4701823835058299, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 58193 + }, + { + "epoch": 0.58194, + "grad_norm": 1.7394709289762722, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 58194 + }, + { + "epoch": 0.58195, + "grad_norm": 1.3465847106718047, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 58195 + }, + { + "epoch": 0.58196, + "grad_norm": 1.9619409796864495, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 58196 + }, + { + "epoch": 0.58197, + "grad_norm": 1.3023929735032163, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 58197 + }, + { + "epoch": 0.58198, + "grad_norm": 1.6058867220423363, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58198 + }, + { + "epoch": 0.58199, + "grad_norm": 1.8089638815078228, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 58199 + }, + { + "epoch": 0.582, + "grad_norm": 1.542495291319619, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 58200 + }, + { + "epoch": 0.58201, + "grad_norm": 1.645662822268379, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 58201 + }, + { + "epoch": 0.58202, + "grad_norm": 1.6159330651420676, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 58202 + }, + { + "epoch": 0.58203, + "grad_norm": 1.2949699584954197, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 58203 + }, + { + "epoch": 0.58204, + "grad_norm": 1.6879107664433095, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 58204 + }, + { + "epoch": 0.58205, + "grad_norm": 1.474539811650388, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 58205 + }, + { + "epoch": 0.58206, + "grad_norm": 1.8873775952883158, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 58206 + }, + { + "epoch": 0.58207, + "grad_norm": 1.3881956952390282, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 58207 + }, + { + "epoch": 0.58208, + "grad_norm": 1.5242444000434376, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 58208 + }, + { + "epoch": 0.58209, + "grad_norm": 1.9862831822652396, + "learning_rate": 0.003, + "loss": 4.042, + "step": 58209 + }, + { + "epoch": 0.5821, + "grad_norm": 1.1258084349277988, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 58210 + }, + { + "epoch": 0.58211, + "grad_norm": 2.1692464753035927, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 58211 + }, + { + "epoch": 0.58212, + "grad_norm": 1.3910132410381253, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 58212 + }, + { + "epoch": 0.58213, + "grad_norm": 1.6631930536730264, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 58213 + }, + { + "epoch": 0.58214, + "grad_norm": 1.8252516249684325, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 58214 + }, + { + "epoch": 0.58215, + "grad_norm": 1.589985116912279, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 58215 + }, + { + "epoch": 0.58216, + "grad_norm": 1.5266523897049136, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 58216 + }, + { + "epoch": 0.58217, + "grad_norm": 1.3909938699744666, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 58217 + }, + { + "epoch": 0.58218, + "grad_norm": 1.6404555318055394, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 58218 + }, + { + "epoch": 0.58219, + "grad_norm": 1.4584285149087124, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 58219 + }, + { + "epoch": 0.5822, + "grad_norm": 1.7721107500227309, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 58220 + }, + { + "epoch": 0.58221, + "grad_norm": 1.7108527370178612, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 58221 + }, + { + "epoch": 0.58222, + "grad_norm": 1.5340612706295174, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 58222 + }, + { + "epoch": 0.58223, + "grad_norm": 1.8007192191686954, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 58223 + }, + { + "epoch": 0.58224, + "grad_norm": 1.566299634652597, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 58224 + }, + { + "epoch": 0.58225, + "grad_norm": 1.5902530837901623, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 58225 + }, + { + "epoch": 0.58226, + "grad_norm": 1.5321071523855043, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 58226 + }, + { + "epoch": 0.58227, + "grad_norm": 1.7149297335703335, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 58227 + }, + { + "epoch": 0.58228, + "grad_norm": 1.4361735370839466, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 58228 + }, + { + "epoch": 0.58229, + "grad_norm": 1.5513464996221067, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 58229 + }, + { + "epoch": 0.5823, + "grad_norm": 1.6048474618489936, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58230 + }, + { + "epoch": 0.58231, + "grad_norm": 1.343605980031219, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 58231 + }, + { + "epoch": 0.58232, + "grad_norm": 1.6262973202385897, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 58232 + }, + { + "epoch": 0.58233, + "grad_norm": 1.3370264085465393, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 58233 + }, + { + "epoch": 0.58234, + "grad_norm": 1.740886563943456, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 58234 + }, + { + "epoch": 0.58235, + "grad_norm": 1.2533358147448876, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 58235 + }, + { + "epoch": 0.58236, + "grad_norm": 1.8751465002953402, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 58236 + }, + { + "epoch": 0.58237, + "grad_norm": 1.745271984044356, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 58237 + }, + { + "epoch": 0.58238, + "grad_norm": 1.2908059254591775, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 58238 + }, + { + "epoch": 0.58239, + "grad_norm": 1.6888252219141568, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 58239 + }, + { + "epoch": 0.5824, + "grad_norm": 1.5451118425064811, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 58240 + }, + { + "epoch": 0.58241, + "grad_norm": 2.1465346620296946, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 58241 + }, + { + "epoch": 0.58242, + "grad_norm": 1.669753248638396, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 58242 + }, + { + "epoch": 0.58243, + "grad_norm": 1.4541478497231584, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 58243 + }, + { + "epoch": 0.58244, + "grad_norm": 1.6057774495684958, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 58244 + }, + { + "epoch": 0.58245, + "grad_norm": 1.4783978722006004, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 58245 + }, + { + "epoch": 0.58246, + "grad_norm": 1.5320777616235552, + "learning_rate": 0.003, + "loss": 4.02, + "step": 58246 + }, + { + "epoch": 0.58247, + "grad_norm": 1.7213560477951644, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 58247 + }, + { + "epoch": 0.58248, + "grad_norm": 1.5184292145723077, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 58248 + }, + { + "epoch": 0.58249, + "grad_norm": 1.3917003136161998, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 58249 + }, + { + "epoch": 0.5825, + "grad_norm": 1.7243934796639442, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 58250 + }, + { + "epoch": 0.58251, + "grad_norm": 1.3472736793139062, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 58251 + }, + { + "epoch": 0.58252, + "grad_norm": 1.636449852996359, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 58252 + }, + { + "epoch": 0.58253, + "grad_norm": 1.5272835952464412, + "learning_rate": 0.003, + "loss": 4.048, + "step": 58253 + }, + { + "epoch": 0.58254, + "grad_norm": 1.738585095297876, + "learning_rate": 0.003, + "loss": 4.036, + "step": 58254 + }, + { + "epoch": 0.58255, + "grad_norm": 1.3025547130872683, + "learning_rate": 0.003, + "loss": 4.023, + "step": 58255 + }, + { + "epoch": 0.58256, + "grad_norm": 1.6419683298617884, + "learning_rate": 0.003, + "loss": 4.027, + "step": 58256 + }, + { + "epoch": 0.58257, + "grad_norm": 1.642640957231382, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 58257 + }, + { + "epoch": 0.58258, + "grad_norm": 1.5030897264685368, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 58258 + }, + { + "epoch": 0.58259, + "grad_norm": 1.7302123438583916, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 58259 + }, + { + "epoch": 0.5826, + "grad_norm": 1.7521194935302307, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 58260 + }, + { + "epoch": 0.58261, + "grad_norm": 1.4687594972757168, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 58261 + }, + { + "epoch": 0.58262, + "grad_norm": 1.4648278598601225, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 58262 + }, + { + "epoch": 0.58263, + "grad_norm": 1.6081763927730899, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 58263 + }, + { + "epoch": 0.58264, + "grad_norm": 1.6882310540935581, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 58264 + }, + { + "epoch": 0.58265, + "grad_norm": 1.835616770082309, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 58265 + }, + { + "epoch": 0.58266, + "grad_norm": 1.5430896299537824, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 58266 + }, + { + "epoch": 0.58267, + "grad_norm": 2.0085572850196596, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 58267 + }, + { + "epoch": 0.58268, + "grad_norm": 1.4500344339517637, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 58268 + }, + { + "epoch": 0.58269, + "grad_norm": 1.3375186414679399, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 58269 + }, + { + "epoch": 0.5827, + "grad_norm": 1.687781566739034, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 58270 + }, + { + "epoch": 0.58271, + "grad_norm": 1.8786607347980857, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 58271 + }, + { + "epoch": 0.58272, + "grad_norm": 1.2237657147148273, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58272 + }, + { + "epoch": 0.58273, + "grad_norm": 1.791137237216147, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 58273 + }, + { + "epoch": 0.58274, + "grad_norm": 1.4778361461345004, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 58274 + }, + { + "epoch": 0.58275, + "grad_norm": 1.8631820680403735, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 58275 + }, + { + "epoch": 0.58276, + "grad_norm": 1.4497619478621233, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 58276 + }, + { + "epoch": 0.58277, + "grad_norm": 1.5672834967137401, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 58277 + }, + { + "epoch": 0.58278, + "grad_norm": 1.7079176169849901, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 58278 + }, + { + "epoch": 0.58279, + "grad_norm": 1.6017623509608312, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 58279 + }, + { + "epoch": 0.5828, + "grad_norm": 1.9670522174016098, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 58280 + }, + { + "epoch": 0.58281, + "grad_norm": 1.4647475911871277, + "learning_rate": 0.003, + "loss": 4.025, + "step": 58281 + }, + { + "epoch": 0.58282, + "grad_norm": 2.0131536369735112, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 58282 + }, + { + "epoch": 0.58283, + "grad_norm": 1.6229552097416056, + "learning_rate": 0.003, + "loss": 4.069, + "step": 58283 + }, + { + "epoch": 0.58284, + "grad_norm": 1.3698978690981412, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 58284 + }, + { + "epoch": 0.58285, + "grad_norm": 1.7997426127455496, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 58285 + }, + { + "epoch": 0.58286, + "grad_norm": 1.416605202925685, + "learning_rate": 0.003, + "loss": 4.012, + "step": 58286 + }, + { + "epoch": 0.58287, + "grad_norm": 1.288157184350039, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 58287 + }, + { + "epoch": 0.58288, + "grad_norm": 1.8914488046512683, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 58288 + }, + { + "epoch": 0.58289, + "grad_norm": 1.5016950589762357, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 58289 + }, + { + "epoch": 0.5829, + "grad_norm": 1.4666209262822163, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 58290 + }, + { + "epoch": 0.58291, + "grad_norm": 1.8087026250661478, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 58291 + }, + { + "epoch": 0.58292, + "grad_norm": 1.5898548045215295, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 58292 + }, + { + "epoch": 0.58293, + "grad_norm": 1.3136127120173853, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 58293 + }, + { + "epoch": 0.58294, + "grad_norm": 1.755440581801388, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58294 + }, + { + "epoch": 0.58295, + "grad_norm": 1.4346780726574375, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 58295 + }, + { + "epoch": 0.58296, + "grad_norm": 1.6044755479086599, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 58296 + }, + { + "epoch": 0.58297, + "grad_norm": 1.3913123006072496, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 58297 + }, + { + "epoch": 0.58298, + "grad_norm": 2.0970848262409203, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 58298 + }, + { + "epoch": 0.58299, + "grad_norm": 1.1937669938440951, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 58299 + }, + { + "epoch": 0.583, + "grad_norm": 1.9388323850653222, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 58300 + }, + { + "epoch": 0.58301, + "grad_norm": 1.4815470758002256, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 58301 + }, + { + "epoch": 0.58302, + "grad_norm": 1.7460922383608417, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 58302 + }, + { + "epoch": 0.58303, + "grad_norm": 1.5108502698176405, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 58303 + }, + { + "epoch": 0.58304, + "grad_norm": 1.7537240666174339, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 58304 + }, + { + "epoch": 0.58305, + "grad_norm": 1.4373652667395636, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 58305 + }, + { + "epoch": 0.58306, + "grad_norm": 1.7766037958437277, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 58306 + }, + { + "epoch": 0.58307, + "grad_norm": 1.4834662013295539, + "learning_rate": 0.003, + "loss": 4.028, + "step": 58307 + }, + { + "epoch": 0.58308, + "grad_norm": 1.6117838653113126, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 58308 + }, + { + "epoch": 0.58309, + "grad_norm": 1.5551775724158752, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 58309 + }, + { + "epoch": 0.5831, + "grad_norm": 1.6556981335006564, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 58310 + }, + { + "epoch": 0.58311, + "grad_norm": 1.4930283094679715, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 58311 + }, + { + "epoch": 0.58312, + "grad_norm": 1.6486724437965166, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 58312 + }, + { + "epoch": 0.58313, + "grad_norm": 2.0087348429473986, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 58313 + }, + { + "epoch": 0.58314, + "grad_norm": 1.1633266946001932, + "learning_rate": 0.003, + "loss": 3.994, + "step": 58314 + }, + { + "epoch": 0.58315, + "grad_norm": 2.137899541892654, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 58315 + }, + { + "epoch": 0.58316, + "grad_norm": 1.4276270977662386, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 58316 + }, + { + "epoch": 0.58317, + "grad_norm": 1.6313105056126473, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 58317 + }, + { + "epoch": 0.58318, + "grad_norm": 1.881855672639429, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 58318 + }, + { + "epoch": 0.58319, + "grad_norm": 1.4469993094588278, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 58319 + }, + { + "epoch": 0.5832, + "grad_norm": 1.5449782011921849, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58320 + }, + { + "epoch": 0.58321, + "grad_norm": 1.462097092171496, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 58321 + }, + { + "epoch": 0.58322, + "grad_norm": 1.7705925669712752, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 58322 + }, + { + "epoch": 0.58323, + "grad_norm": 1.6614882819714454, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 58323 + }, + { + "epoch": 0.58324, + "grad_norm": 1.5774038749528199, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 58324 + }, + { + "epoch": 0.58325, + "grad_norm": 1.4307035301638087, + "learning_rate": 0.003, + "loss": 4.032, + "step": 58325 + }, + { + "epoch": 0.58326, + "grad_norm": 1.8521410185724445, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 58326 + }, + { + "epoch": 0.58327, + "grad_norm": 1.491588668374662, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 58327 + }, + { + "epoch": 0.58328, + "grad_norm": 1.6368267645145407, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 58328 + }, + { + "epoch": 0.58329, + "grad_norm": 1.5352584404464376, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 58329 + }, + { + "epoch": 0.5833, + "grad_norm": 1.889257229843557, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 58330 + }, + { + "epoch": 0.58331, + "grad_norm": 0.9989328223399979, + "learning_rate": 0.003, + "loss": 4.02, + "step": 58331 + }, + { + "epoch": 0.58332, + "grad_norm": 2.120754431258931, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 58332 + }, + { + "epoch": 0.58333, + "grad_norm": 1.7894626746582147, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 58333 + }, + { + "epoch": 0.58334, + "grad_norm": 1.8403129951249086, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 58334 + }, + { + "epoch": 0.58335, + "grad_norm": 1.6406452712483302, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 58335 + }, + { + "epoch": 0.58336, + "grad_norm": 1.7371862731095558, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 58336 + }, + { + "epoch": 0.58337, + "grad_norm": 1.2315007759204124, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 58337 + }, + { + "epoch": 0.58338, + "grad_norm": 1.8801895702853093, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 58338 + }, + { + "epoch": 0.58339, + "grad_norm": 1.2914481174395784, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 58339 + }, + { + "epoch": 0.5834, + "grad_norm": 1.6494489570572832, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 58340 + }, + { + "epoch": 0.58341, + "grad_norm": 1.4304547108338463, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 58341 + }, + { + "epoch": 0.58342, + "grad_norm": 1.6831893474679014, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 58342 + }, + { + "epoch": 0.58343, + "grad_norm": 1.3722763624279521, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 58343 + }, + { + "epoch": 0.58344, + "grad_norm": 1.6227785841831603, + "learning_rate": 0.003, + "loss": 4.064, + "step": 58344 + }, + { + "epoch": 0.58345, + "grad_norm": 1.5222827192305721, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 58345 + }, + { + "epoch": 0.58346, + "grad_norm": 1.8726460779064256, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 58346 + }, + { + "epoch": 0.58347, + "grad_norm": 1.788817783697714, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 58347 + }, + { + "epoch": 0.58348, + "grad_norm": 1.4689592476663704, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 58348 + }, + { + "epoch": 0.58349, + "grad_norm": 1.5000604334518235, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 58349 + }, + { + "epoch": 0.5835, + "grad_norm": 1.778755224528821, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 58350 + }, + { + "epoch": 0.58351, + "grad_norm": 1.3771924844296122, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 58351 + }, + { + "epoch": 0.58352, + "grad_norm": 1.7861180863311472, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 58352 + }, + { + "epoch": 0.58353, + "grad_norm": 1.3052209002805075, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 58353 + }, + { + "epoch": 0.58354, + "grad_norm": 1.4079315754709465, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 58354 + }, + { + "epoch": 0.58355, + "grad_norm": 1.570022934895564, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 58355 + }, + { + "epoch": 0.58356, + "grad_norm": 1.5502435551814109, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 58356 + }, + { + "epoch": 0.58357, + "grad_norm": 1.7504501725634685, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 58357 + }, + { + "epoch": 0.58358, + "grad_norm": 1.5885012569330077, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 58358 + }, + { + "epoch": 0.58359, + "grad_norm": 1.8821816985012252, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 58359 + }, + { + "epoch": 0.5836, + "grad_norm": 1.3638520648308232, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 58360 + }, + { + "epoch": 0.58361, + "grad_norm": 1.398050172605564, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 58361 + }, + { + "epoch": 0.58362, + "grad_norm": 1.6987785820969492, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 58362 + }, + { + "epoch": 0.58363, + "grad_norm": 1.727320973000462, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 58363 + }, + { + "epoch": 0.58364, + "grad_norm": 1.6005568804141226, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 58364 + }, + { + "epoch": 0.58365, + "grad_norm": 1.8319603883803206, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 58365 + }, + { + "epoch": 0.58366, + "grad_norm": 1.203261219821833, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 58366 + }, + { + "epoch": 0.58367, + "grad_norm": 1.587329036161193, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 58367 + }, + { + "epoch": 0.58368, + "grad_norm": 1.5556995626371841, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 58368 + }, + { + "epoch": 0.58369, + "grad_norm": 1.6321182772901213, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 58369 + }, + { + "epoch": 0.5837, + "grad_norm": 1.6392158979887432, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 58370 + }, + { + "epoch": 0.58371, + "grad_norm": 1.8122656010008515, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 58371 + }, + { + "epoch": 0.58372, + "grad_norm": 1.4081103782026592, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 58372 + }, + { + "epoch": 0.58373, + "grad_norm": 1.4759074363577303, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 58373 + }, + { + "epoch": 0.58374, + "grad_norm": 1.7858804520389948, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 58374 + }, + { + "epoch": 0.58375, + "grad_norm": 1.4824567644949742, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 58375 + }, + { + "epoch": 0.58376, + "grad_norm": 1.3149633183207898, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 58376 + }, + { + "epoch": 0.58377, + "grad_norm": 1.8167758686753888, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 58377 + }, + { + "epoch": 0.58378, + "grad_norm": 1.6126985127833764, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 58378 + }, + { + "epoch": 0.58379, + "grad_norm": 1.5651875117365008, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 58379 + }, + { + "epoch": 0.5838, + "grad_norm": 1.442653055997994, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 58380 + }, + { + "epoch": 0.58381, + "grad_norm": 1.5421668405153999, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 58381 + }, + { + "epoch": 0.58382, + "grad_norm": 1.4177349501993577, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 58382 + }, + { + "epoch": 0.58383, + "grad_norm": 1.5644419253756132, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 58383 + }, + { + "epoch": 0.58384, + "grad_norm": 1.3886493276306764, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 58384 + }, + { + "epoch": 0.58385, + "grad_norm": 1.6661359881873692, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 58385 + }, + { + "epoch": 0.58386, + "grad_norm": 1.4108754117107405, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 58386 + }, + { + "epoch": 0.58387, + "grad_norm": 2.219250399398817, + "learning_rate": 0.003, + "loss": 4.042, + "step": 58387 + }, + { + "epoch": 0.58388, + "grad_norm": 1.805977867219577, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 58388 + }, + { + "epoch": 0.58389, + "grad_norm": 1.2021539561960795, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 58389 + }, + { + "epoch": 0.5839, + "grad_norm": 2.032915551654226, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 58390 + }, + { + "epoch": 0.58391, + "grad_norm": 1.9173677102364657, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 58391 + }, + { + "epoch": 0.58392, + "grad_norm": 1.379031169665001, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 58392 + }, + { + "epoch": 0.58393, + "grad_norm": 1.6626268576131142, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 58393 + }, + { + "epoch": 0.58394, + "grad_norm": 1.5218061846968531, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 58394 + }, + { + "epoch": 0.58395, + "grad_norm": 1.5676966872131584, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 58395 + }, + { + "epoch": 0.58396, + "grad_norm": 1.3030380798081114, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 58396 + }, + { + "epoch": 0.58397, + "grad_norm": 1.7556986222330242, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 58397 + }, + { + "epoch": 0.58398, + "grad_norm": 2.060861091752672, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 58398 + }, + { + "epoch": 0.58399, + "grad_norm": 1.4126979841667573, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 58399 + }, + { + "epoch": 0.584, + "grad_norm": 1.6705130589831672, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 58400 + }, + { + "epoch": 0.58401, + "grad_norm": 1.6398905220501232, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 58401 + }, + { + "epoch": 0.58402, + "grad_norm": 1.4088033333392873, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 58402 + }, + { + "epoch": 0.58403, + "grad_norm": 1.4975746246584147, + "learning_rate": 0.003, + "loss": 4.026, + "step": 58403 + }, + { + "epoch": 0.58404, + "grad_norm": 1.5375192641999045, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 58404 + }, + { + "epoch": 0.58405, + "grad_norm": 1.5803471770090145, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 58405 + }, + { + "epoch": 0.58406, + "grad_norm": 1.5627359190247694, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 58406 + }, + { + "epoch": 0.58407, + "grad_norm": 1.5335563558430592, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 58407 + }, + { + "epoch": 0.58408, + "grad_norm": 1.6739794361698368, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 58408 + }, + { + "epoch": 0.58409, + "grad_norm": 1.8586506779158827, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 58409 + }, + { + "epoch": 0.5841, + "grad_norm": 1.1592218776288208, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 58410 + }, + { + "epoch": 0.58411, + "grad_norm": 1.6866879199239704, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 58411 + }, + { + "epoch": 0.58412, + "grad_norm": 1.5212439530356077, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 58412 + }, + { + "epoch": 0.58413, + "grad_norm": 2.2239706500304606, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 58413 + }, + { + "epoch": 0.58414, + "grad_norm": 1.1684498001237191, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58414 + }, + { + "epoch": 0.58415, + "grad_norm": 1.9935838027691033, + "learning_rate": 0.003, + "loss": 4.044, + "step": 58415 + }, + { + "epoch": 0.58416, + "grad_norm": 1.5149253709068435, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 58416 + }, + { + "epoch": 0.58417, + "grad_norm": 1.5458702977630205, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 58417 + }, + { + "epoch": 0.58418, + "grad_norm": 1.6851464507935916, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 58418 + }, + { + "epoch": 0.58419, + "grad_norm": 1.4066549679746343, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 58419 + }, + { + "epoch": 0.5842, + "grad_norm": 1.775084773303615, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 58420 + }, + { + "epoch": 0.58421, + "grad_norm": 1.477404663038583, + "learning_rate": 0.003, + "loss": 4.039, + "step": 58421 + }, + { + "epoch": 0.58422, + "grad_norm": 1.7165681306384375, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 58422 + }, + { + "epoch": 0.58423, + "grad_norm": 1.6527420346747426, + "learning_rate": 0.003, + "loss": 4.03, + "step": 58423 + }, + { + "epoch": 0.58424, + "grad_norm": 1.604307347716316, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 58424 + }, + { + "epoch": 0.58425, + "grad_norm": 1.41593544908276, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58425 + }, + { + "epoch": 0.58426, + "grad_norm": 1.5345559400687534, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 58426 + }, + { + "epoch": 0.58427, + "grad_norm": 1.6247218113339346, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 58427 + }, + { + "epoch": 0.58428, + "grad_norm": 1.4577377590673715, + "learning_rate": 0.003, + "loss": 4.077, + "step": 58428 + }, + { + "epoch": 0.58429, + "grad_norm": 1.727360754425368, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 58429 + }, + { + "epoch": 0.5843, + "grad_norm": 1.7129319373082634, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 58430 + }, + { + "epoch": 0.58431, + "grad_norm": 1.616412261751361, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 58431 + }, + { + "epoch": 0.58432, + "grad_norm": 1.4206614566234148, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 58432 + }, + { + "epoch": 0.58433, + "grad_norm": 1.6027579593531145, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 58433 + }, + { + "epoch": 0.58434, + "grad_norm": 1.6885721603459987, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 58434 + }, + { + "epoch": 0.58435, + "grad_norm": 1.511841605165213, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 58435 + }, + { + "epoch": 0.58436, + "grad_norm": 1.8847396725048562, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 58436 + }, + { + "epoch": 0.58437, + "grad_norm": 1.4055021659801774, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 58437 + }, + { + "epoch": 0.58438, + "grad_norm": 1.437235357962391, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 58438 + }, + { + "epoch": 0.58439, + "grad_norm": 1.4882750977456671, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 58439 + }, + { + "epoch": 0.5844, + "grad_norm": 1.6455450605072683, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 58440 + }, + { + "epoch": 0.58441, + "grad_norm": 1.8304555110985525, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 58441 + }, + { + "epoch": 0.58442, + "grad_norm": 1.7591705976646432, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 58442 + }, + { + "epoch": 0.58443, + "grad_norm": 1.251094382981331, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 58443 + }, + { + "epoch": 0.58444, + "grad_norm": 1.8501266029509638, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 58444 + }, + { + "epoch": 0.58445, + "grad_norm": 1.2916981462523087, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 58445 + }, + { + "epoch": 0.58446, + "grad_norm": 1.5754816313806823, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 58446 + }, + { + "epoch": 0.58447, + "grad_norm": 1.4260342025467154, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 58447 + }, + { + "epoch": 0.58448, + "grad_norm": 1.790729022003073, + "learning_rate": 0.003, + "loss": 4.031, + "step": 58448 + }, + { + "epoch": 0.58449, + "grad_norm": 1.7336590824450944, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 58449 + }, + { + "epoch": 0.5845, + "grad_norm": 1.623127934117915, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 58450 + }, + { + "epoch": 0.58451, + "grad_norm": 1.7485372983071454, + "learning_rate": 0.003, + "loss": 4.029, + "step": 58451 + }, + { + "epoch": 0.58452, + "grad_norm": 1.8301140437062384, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 58452 + }, + { + "epoch": 0.58453, + "grad_norm": 1.3418255877879357, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 58453 + }, + { + "epoch": 0.58454, + "grad_norm": 1.622781503861527, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 58454 + }, + { + "epoch": 0.58455, + "grad_norm": 1.4715855607867236, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 58455 + }, + { + "epoch": 0.58456, + "grad_norm": 1.5455745782003671, + "learning_rate": 0.003, + "loss": 4.04, + "step": 58456 + }, + { + "epoch": 0.58457, + "grad_norm": 1.9099369065789125, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 58457 + }, + { + "epoch": 0.58458, + "grad_norm": 1.4435733500897785, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 58458 + }, + { + "epoch": 0.58459, + "grad_norm": 1.6151955421299928, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 58459 + }, + { + "epoch": 0.5846, + "grad_norm": 1.4205603117810623, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 58460 + }, + { + "epoch": 0.58461, + "grad_norm": 1.5305238715536296, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 58461 + }, + { + "epoch": 0.58462, + "grad_norm": 1.6832980202729695, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 58462 + }, + { + "epoch": 0.58463, + "grad_norm": 2.114615711447475, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 58463 + }, + { + "epoch": 0.58464, + "grad_norm": 1.257474438026636, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 58464 + }, + { + "epoch": 0.58465, + "grad_norm": 1.6870044903670047, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 58465 + }, + { + "epoch": 0.58466, + "grad_norm": 1.767246072158195, + "learning_rate": 0.003, + "loss": 4.035, + "step": 58466 + }, + { + "epoch": 0.58467, + "grad_norm": 1.5930316350054288, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 58467 + }, + { + "epoch": 0.58468, + "grad_norm": 1.7020224708310983, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 58468 + }, + { + "epoch": 0.58469, + "grad_norm": 1.3505604831329239, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 58469 + }, + { + "epoch": 0.5847, + "grad_norm": 1.536067113207254, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 58470 + }, + { + "epoch": 0.58471, + "grad_norm": 1.3319485402093976, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 58471 + }, + { + "epoch": 0.58472, + "grad_norm": 1.9246944184687194, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 58472 + }, + { + "epoch": 0.58473, + "grad_norm": 1.5681064918575307, + "learning_rate": 0.003, + "loss": 4.04, + "step": 58473 + }, + { + "epoch": 0.58474, + "grad_norm": 1.4492712166293473, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 58474 + }, + { + "epoch": 0.58475, + "grad_norm": 1.530757411166538, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 58475 + }, + { + "epoch": 0.58476, + "grad_norm": 1.5192070390755552, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 58476 + }, + { + "epoch": 0.58477, + "grad_norm": 1.5233984129602713, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 58477 + }, + { + "epoch": 0.58478, + "grad_norm": 1.4581568825284716, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 58478 + }, + { + "epoch": 0.58479, + "grad_norm": 1.4416085409940629, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 58479 + }, + { + "epoch": 0.5848, + "grad_norm": 1.8867021730057918, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 58480 + }, + { + "epoch": 0.58481, + "grad_norm": 1.6578036707272465, + "learning_rate": 0.003, + "loss": 4.057, + "step": 58481 + }, + { + "epoch": 0.58482, + "grad_norm": 2.075149843260214, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 58482 + }, + { + "epoch": 0.58483, + "grad_norm": 1.4171994717684242, + "learning_rate": 0.003, + "loss": 4.023, + "step": 58483 + }, + { + "epoch": 0.58484, + "grad_norm": 1.3394354683321241, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 58484 + }, + { + "epoch": 0.58485, + "grad_norm": 1.8065769522773274, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 58485 + }, + { + "epoch": 0.58486, + "grad_norm": 1.7554482897701218, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 58486 + }, + { + "epoch": 0.58487, + "grad_norm": 1.3988216268320823, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 58487 + }, + { + "epoch": 0.58488, + "grad_norm": 2.016198186113905, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 58488 + }, + { + "epoch": 0.58489, + "grad_norm": 1.2418638287787402, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 58489 + }, + { + "epoch": 0.5849, + "grad_norm": 1.6899120125924443, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 58490 + }, + { + "epoch": 0.58491, + "grad_norm": 1.4662587672256842, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 58491 + }, + { + "epoch": 0.58492, + "grad_norm": 1.6250549850821088, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 58492 + }, + { + "epoch": 0.58493, + "grad_norm": 1.5912002566445538, + "learning_rate": 0.003, + "loss": 4.03, + "step": 58493 + }, + { + "epoch": 0.58494, + "grad_norm": 1.740865130571301, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 58494 + }, + { + "epoch": 0.58495, + "grad_norm": 1.49501175008497, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 58495 + }, + { + "epoch": 0.58496, + "grad_norm": 1.259030376643934, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 58496 + }, + { + "epoch": 0.58497, + "grad_norm": 1.9608074738592158, + "learning_rate": 0.003, + "loss": 4.025, + "step": 58497 + }, + { + "epoch": 0.58498, + "grad_norm": 1.626265718030213, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 58498 + }, + { + "epoch": 0.58499, + "grad_norm": 1.6933018191067615, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 58499 + }, + { + "epoch": 0.585, + "grad_norm": 1.41497363321234, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 58500 + }, + { + "epoch": 0.58501, + "grad_norm": 1.9062215553738995, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 58501 + }, + { + "epoch": 0.58502, + "grad_norm": 1.1031127427570635, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 58502 + }, + { + "epoch": 0.58503, + "grad_norm": 1.9263800291960367, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 58503 + }, + { + "epoch": 0.58504, + "grad_norm": 1.597103616394031, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 58504 + }, + { + "epoch": 0.58505, + "grad_norm": 1.577013622257441, + "learning_rate": 0.003, + "loss": 4.03, + "step": 58505 + }, + { + "epoch": 0.58506, + "grad_norm": 1.667608966264589, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 58506 + }, + { + "epoch": 0.58507, + "grad_norm": 1.557506315634068, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 58507 + }, + { + "epoch": 0.58508, + "grad_norm": 1.4249118370658487, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 58508 + }, + { + "epoch": 0.58509, + "grad_norm": 1.6006488980577354, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 58509 + }, + { + "epoch": 0.5851, + "grad_norm": 1.5801426636411082, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 58510 + }, + { + "epoch": 0.58511, + "grad_norm": 1.2685831038655442, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 58511 + }, + { + "epoch": 0.58512, + "grad_norm": 1.587033988773238, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 58512 + }, + { + "epoch": 0.58513, + "grad_norm": 1.4552670893954152, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 58513 + }, + { + "epoch": 0.58514, + "grad_norm": 1.4816276686786407, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 58514 + }, + { + "epoch": 0.58515, + "grad_norm": 1.7440673500816388, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 58515 + }, + { + "epoch": 0.58516, + "grad_norm": 1.5449432974321957, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 58516 + }, + { + "epoch": 0.58517, + "grad_norm": 1.8472549306695412, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 58517 + }, + { + "epoch": 0.58518, + "grad_norm": 1.5272383966564187, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 58518 + }, + { + "epoch": 0.58519, + "grad_norm": 1.9665020952246202, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 58519 + }, + { + "epoch": 0.5852, + "grad_norm": 1.212978088575291, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 58520 + }, + { + "epoch": 0.58521, + "grad_norm": 1.7605137599077003, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 58521 + }, + { + "epoch": 0.58522, + "grad_norm": 1.5332716695672488, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 58522 + }, + { + "epoch": 0.58523, + "grad_norm": 1.6299200119231503, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 58523 + }, + { + "epoch": 0.58524, + "grad_norm": 1.882754052980226, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 58524 + }, + { + "epoch": 0.58525, + "grad_norm": 1.411634788512324, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 58525 + }, + { + "epoch": 0.58526, + "grad_norm": 1.621096505511404, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 58526 + }, + { + "epoch": 0.58527, + "grad_norm": 1.5713494739597214, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 58527 + }, + { + "epoch": 0.58528, + "grad_norm": 1.2553311908616422, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 58528 + }, + { + "epoch": 0.58529, + "grad_norm": 1.8289163614847792, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 58529 + }, + { + "epoch": 0.5853, + "grad_norm": 1.485023876594529, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58530 + }, + { + "epoch": 0.58531, + "grad_norm": 1.6191372767954555, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 58531 + }, + { + "epoch": 0.58532, + "grad_norm": 1.7286696113900775, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 58532 + }, + { + "epoch": 0.58533, + "grad_norm": 1.941462141484889, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 58533 + }, + { + "epoch": 0.58534, + "grad_norm": 1.4347542409380076, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 58534 + }, + { + "epoch": 0.58535, + "grad_norm": 1.6984623548002775, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 58535 + }, + { + "epoch": 0.58536, + "grad_norm": 1.3471499411576713, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 58536 + }, + { + "epoch": 0.58537, + "grad_norm": 1.7130449413516102, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 58537 + }, + { + "epoch": 0.58538, + "grad_norm": 1.4994759289046924, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 58538 + }, + { + "epoch": 0.58539, + "grad_norm": 1.7299414786146636, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 58539 + }, + { + "epoch": 0.5854, + "grad_norm": 1.3758764674398125, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 58540 + }, + { + "epoch": 0.58541, + "grad_norm": 1.744310007656531, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 58541 + }, + { + "epoch": 0.58542, + "grad_norm": 1.4688109191348802, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 58542 + }, + { + "epoch": 0.58543, + "grad_norm": 1.5394966954110807, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 58543 + }, + { + "epoch": 0.58544, + "grad_norm": 1.5187460911551927, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 58544 + }, + { + "epoch": 0.58545, + "grad_norm": 1.4916151363870696, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 58545 + }, + { + "epoch": 0.58546, + "grad_norm": 1.4042269315692366, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 58546 + }, + { + "epoch": 0.58547, + "grad_norm": 1.7783901871991104, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 58547 + }, + { + "epoch": 0.58548, + "grad_norm": 1.7099545318160077, + "learning_rate": 0.003, + "loss": 4.055, + "step": 58548 + }, + { + "epoch": 0.58549, + "grad_norm": 1.3486163776527986, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 58549 + }, + { + "epoch": 0.5855, + "grad_norm": 1.850906915092449, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 58550 + }, + { + "epoch": 0.58551, + "grad_norm": 1.2211307874894712, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 58551 + }, + { + "epoch": 0.58552, + "grad_norm": 1.842483737617699, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 58552 + }, + { + "epoch": 0.58553, + "grad_norm": 1.3341639238404415, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 58553 + }, + { + "epoch": 0.58554, + "grad_norm": 1.8113043665324644, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 58554 + }, + { + "epoch": 0.58555, + "grad_norm": 1.4271652785556448, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 58555 + }, + { + "epoch": 0.58556, + "grad_norm": 1.7499165528631244, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 58556 + }, + { + "epoch": 0.58557, + "grad_norm": 1.5360037857875812, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 58557 + }, + { + "epoch": 0.58558, + "grad_norm": 1.825619099051333, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 58558 + }, + { + "epoch": 0.58559, + "grad_norm": 1.450979090085284, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 58559 + }, + { + "epoch": 0.5856, + "grad_norm": 1.6391716324539638, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 58560 + }, + { + "epoch": 0.58561, + "grad_norm": 1.7056998041638158, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 58561 + }, + { + "epoch": 0.58562, + "grad_norm": 1.3808415564703356, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 58562 + }, + { + "epoch": 0.58563, + "grad_norm": 1.8764525011740216, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 58563 + }, + { + "epoch": 0.58564, + "grad_norm": 1.187620042767567, + "learning_rate": 0.003, + "loss": 4.04, + "step": 58564 + }, + { + "epoch": 0.58565, + "grad_norm": 1.7567822020499928, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 58565 + }, + { + "epoch": 0.58566, + "grad_norm": 1.6025499390014257, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 58566 + }, + { + "epoch": 0.58567, + "grad_norm": 1.48601476144334, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 58567 + }, + { + "epoch": 0.58568, + "grad_norm": 1.7939675031350757, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 58568 + }, + { + "epoch": 0.58569, + "grad_norm": 1.5429553946582597, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58569 + }, + { + "epoch": 0.5857, + "grad_norm": 1.4474200947734135, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 58570 + }, + { + "epoch": 0.58571, + "grad_norm": 1.9249508216940483, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 58571 + }, + { + "epoch": 0.58572, + "grad_norm": 1.7580984913686115, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 58572 + }, + { + "epoch": 0.58573, + "grad_norm": 1.5770393482080516, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 58573 + }, + { + "epoch": 0.58574, + "grad_norm": 1.6955166528981176, + "learning_rate": 0.003, + "loss": 4.01, + "step": 58574 + }, + { + "epoch": 0.58575, + "grad_norm": 1.38861431022621, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 58575 + }, + { + "epoch": 0.58576, + "grad_norm": 1.6708327785580774, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 58576 + }, + { + "epoch": 0.58577, + "grad_norm": 1.5178467097352621, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 58577 + }, + { + "epoch": 0.58578, + "grad_norm": 1.7217148996269889, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 58578 + }, + { + "epoch": 0.58579, + "grad_norm": 1.6436135787646473, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 58579 + }, + { + "epoch": 0.5858, + "grad_norm": 1.5963707643619258, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 58580 + }, + { + "epoch": 0.58581, + "grad_norm": 1.8639449559934145, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 58581 + }, + { + "epoch": 0.58582, + "grad_norm": 1.323764141366979, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 58582 + }, + { + "epoch": 0.58583, + "grad_norm": 1.9678008939262113, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 58583 + }, + { + "epoch": 0.58584, + "grad_norm": 1.5751436625582445, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 58584 + }, + { + "epoch": 0.58585, + "grad_norm": 1.6207382866186484, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 58585 + }, + { + "epoch": 0.58586, + "grad_norm": 1.8980988015154805, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 58586 + }, + { + "epoch": 0.58587, + "grad_norm": 1.2132135720910118, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 58587 + }, + { + "epoch": 0.58588, + "grad_norm": 1.6978550896734308, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 58588 + }, + { + "epoch": 0.58589, + "grad_norm": 1.439238715164372, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 58589 + }, + { + "epoch": 0.5859, + "grad_norm": 1.7186106652772701, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 58590 + }, + { + "epoch": 0.58591, + "grad_norm": 1.6450118163877272, + "learning_rate": 0.003, + "loss": 4.041, + "step": 58591 + }, + { + "epoch": 0.58592, + "grad_norm": 1.5578216034707797, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 58592 + }, + { + "epoch": 0.58593, + "grad_norm": 1.5476631395089357, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 58593 + }, + { + "epoch": 0.58594, + "grad_norm": 1.5519951636382665, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 58594 + }, + { + "epoch": 0.58595, + "grad_norm": 1.3724183374607366, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 58595 + }, + { + "epoch": 0.58596, + "grad_norm": 1.4238691751155037, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 58596 + }, + { + "epoch": 0.58597, + "grad_norm": 1.3323085069015979, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 58597 + }, + { + "epoch": 0.58598, + "grad_norm": 1.7400524554464787, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 58598 + }, + { + "epoch": 0.58599, + "grad_norm": 1.586675184745535, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 58599 + }, + { + "epoch": 0.586, + "grad_norm": 1.6018998859736762, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 58600 + }, + { + "epoch": 0.58601, + "grad_norm": 1.4984394092322884, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 58601 + }, + { + "epoch": 0.58602, + "grad_norm": 1.560705343694411, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 58602 + }, + { + "epoch": 0.58603, + "grad_norm": 1.812410087814129, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 58603 + }, + { + "epoch": 0.58604, + "grad_norm": 1.7207960491646972, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 58604 + }, + { + "epoch": 0.58605, + "grad_norm": 1.5995864960200834, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 58605 + }, + { + "epoch": 0.58606, + "grad_norm": 1.546539624209543, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 58606 + }, + { + "epoch": 0.58607, + "grad_norm": 1.4917953047835844, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 58607 + }, + { + "epoch": 0.58608, + "grad_norm": 1.8003163384428422, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 58608 + }, + { + "epoch": 0.58609, + "grad_norm": 1.319027710431727, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 58609 + }, + { + "epoch": 0.5861, + "grad_norm": 2.186782225981946, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 58610 + }, + { + "epoch": 0.58611, + "grad_norm": 1.1494373930727735, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 58611 + }, + { + "epoch": 0.58612, + "grad_norm": 1.8544194033034767, + "learning_rate": 0.003, + "loss": 4.043, + "step": 58612 + }, + { + "epoch": 0.58613, + "grad_norm": 1.5250822309987007, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 58613 + }, + { + "epoch": 0.58614, + "grad_norm": 1.6296970257871555, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 58614 + }, + { + "epoch": 0.58615, + "grad_norm": 1.5035242118644556, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 58615 + }, + { + "epoch": 0.58616, + "grad_norm": 1.5903611950242027, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 58616 + }, + { + "epoch": 0.58617, + "grad_norm": 1.7855285872294906, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 58617 + }, + { + "epoch": 0.58618, + "grad_norm": 1.3544816798597723, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 58618 + }, + { + "epoch": 0.58619, + "grad_norm": 1.636628936340665, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 58619 + }, + { + "epoch": 0.5862, + "grad_norm": 1.5748944942871999, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 58620 + }, + { + "epoch": 0.58621, + "grad_norm": 1.6897201029485827, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 58621 + }, + { + "epoch": 0.58622, + "grad_norm": 1.571295065593124, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 58622 + }, + { + "epoch": 0.58623, + "grad_norm": 1.5492732915807843, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 58623 + }, + { + "epoch": 0.58624, + "grad_norm": 1.4763324476914026, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 58624 + }, + { + "epoch": 0.58625, + "grad_norm": 1.6139253681623467, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 58625 + }, + { + "epoch": 0.58626, + "grad_norm": 1.4955180810741264, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 58626 + }, + { + "epoch": 0.58627, + "grad_norm": 1.5643797843761627, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 58627 + }, + { + "epoch": 0.58628, + "grad_norm": 1.3150243166757853, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 58628 + }, + { + "epoch": 0.58629, + "grad_norm": 2.095351398910304, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 58629 + }, + { + "epoch": 0.5863, + "grad_norm": 1.3197274090739248, + "learning_rate": 0.003, + "loss": 4.049, + "step": 58630 + }, + { + "epoch": 0.58631, + "grad_norm": 1.7569817515186308, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 58631 + }, + { + "epoch": 0.58632, + "grad_norm": 1.429157043441385, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 58632 + }, + { + "epoch": 0.58633, + "grad_norm": 1.710077030257955, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58633 + }, + { + "epoch": 0.58634, + "grad_norm": 1.6267676608667987, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 58634 + }, + { + "epoch": 0.58635, + "grad_norm": 1.8645436166343352, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 58635 + }, + { + "epoch": 0.58636, + "grad_norm": 1.3722468417435065, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 58636 + }, + { + "epoch": 0.58637, + "grad_norm": 1.9099731147197814, + "learning_rate": 0.003, + "loss": 4.028, + "step": 58637 + }, + { + "epoch": 0.58638, + "grad_norm": 1.1934784920463501, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 58638 + }, + { + "epoch": 0.58639, + "grad_norm": 1.5556156161118222, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 58639 + }, + { + "epoch": 0.5864, + "grad_norm": 1.5412443724113665, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 58640 + }, + { + "epoch": 0.58641, + "grad_norm": 1.571798354772132, + "learning_rate": 0.003, + "loss": 4.043, + "step": 58641 + }, + { + "epoch": 0.58642, + "grad_norm": 1.4381210518654832, + "learning_rate": 0.003, + "loss": 4.026, + "step": 58642 + }, + { + "epoch": 0.58643, + "grad_norm": 1.5562011314890212, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 58643 + }, + { + "epoch": 0.58644, + "grad_norm": 1.5910421751652175, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 58644 + }, + { + "epoch": 0.58645, + "grad_norm": 1.5853103560414048, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 58645 + }, + { + "epoch": 0.58646, + "grad_norm": 1.297790983903055, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 58646 + }, + { + "epoch": 0.58647, + "grad_norm": 1.7621656925728273, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 58647 + }, + { + "epoch": 0.58648, + "grad_norm": 1.6907106932893685, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 58648 + }, + { + "epoch": 0.58649, + "grad_norm": 1.7609867114699918, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 58649 + }, + { + "epoch": 0.5865, + "grad_norm": 1.6216511193243388, + "learning_rate": 0.003, + "loss": 4.019, + "step": 58650 + }, + { + "epoch": 0.58651, + "grad_norm": 1.6198436867018797, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 58651 + }, + { + "epoch": 0.58652, + "grad_norm": 1.5021275299870422, + "learning_rate": 0.003, + "loss": 4.053, + "step": 58652 + }, + { + "epoch": 0.58653, + "grad_norm": 1.7800119297354242, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 58653 + }, + { + "epoch": 0.58654, + "grad_norm": 1.719919593714651, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 58654 + }, + { + "epoch": 0.58655, + "grad_norm": 1.4950262181997482, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 58655 + }, + { + "epoch": 0.58656, + "grad_norm": 1.514600990142436, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 58656 + }, + { + "epoch": 0.58657, + "grad_norm": 3.0856029715230466, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 58657 + }, + { + "epoch": 0.58658, + "grad_norm": 3.07892675113016, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 58658 + }, + { + "epoch": 0.58659, + "grad_norm": 5.285019308913642, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 58659 + }, + { + "epoch": 0.5866, + "grad_norm": 9.537751498964308, + "learning_rate": 0.003, + "loss": 4.3115, + "step": 58660 + }, + { + "epoch": 0.58661, + "grad_norm": 3.3435708541787688, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 58661 + }, + { + "epoch": 0.58662, + "grad_norm": 5.581153029994913, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 58662 + }, + { + "epoch": 0.58663, + "grad_norm": 3.69825294322414, + "learning_rate": 0.003, + "loss": 4.2213, + "step": 58663 + }, + { + "epoch": 0.58664, + "grad_norm": 7.3011423918312905, + "learning_rate": 0.003, + "loss": 4.3322, + "step": 58664 + }, + { + "epoch": 0.58665, + "grad_norm": 8.359380299316888, + "learning_rate": 0.003, + "loss": 4.3994, + "step": 58665 + }, + { + "epoch": 0.58666, + "grad_norm": 10.568274755379907, + "learning_rate": 0.003, + "loss": 4.3863, + "step": 58666 + }, + { + "epoch": 0.58667, + "grad_norm": 7.211478943146665, + "learning_rate": 0.003, + "loss": 4.4527, + "step": 58667 + }, + { + "epoch": 0.58668, + "grad_norm": 3.932508491858496, + "learning_rate": 0.003, + "loss": 4.3957, + "step": 58668 + }, + { + "epoch": 0.58669, + "grad_norm": 3.0179243354835537, + "learning_rate": 0.003, + "loss": 4.332, + "step": 58669 + }, + { + "epoch": 0.5867, + "grad_norm": 2.2248122399061168, + "learning_rate": 0.003, + "loss": 4.2975, + "step": 58670 + }, + { + "epoch": 0.58671, + "grad_norm": 2.9314373947690413, + "learning_rate": 0.003, + "loss": 4.3088, + "step": 58671 + }, + { + "epoch": 0.58672, + "grad_norm": 3.2948164108045277, + "learning_rate": 0.003, + "loss": 4.29, + "step": 58672 + }, + { + "epoch": 0.58673, + "grad_norm": 2.3043464548492882, + "learning_rate": 0.003, + "loss": 4.2994, + "step": 58673 + }, + { + "epoch": 0.58674, + "grad_norm": 3.078497481620938, + "learning_rate": 0.003, + "loss": 4.2767, + "step": 58674 + }, + { + "epoch": 0.58675, + "grad_norm": 3.0853946049705194, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 58675 + }, + { + "epoch": 0.58676, + "grad_norm": 2.747730976508321, + "learning_rate": 0.003, + "loss": 4.2823, + "step": 58676 + }, + { + "epoch": 0.58677, + "grad_norm": 2.3470682313776994, + "learning_rate": 0.003, + "loss": 4.232, + "step": 58677 + }, + { + "epoch": 0.58678, + "grad_norm": 3.0118046339566695, + "learning_rate": 0.003, + "loss": 4.2595, + "step": 58678 + }, + { + "epoch": 0.58679, + "grad_norm": 2.9706922203359167, + "learning_rate": 0.003, + "loss": 4.2686, + "step": 58679 + }, + { + "epoch": 0.5868, + "grad_norm": 2.3129939055007287, + "learning_rate": 0.003, + "loss": 4.2514, + "step": 58680 + }, + { + "epoch": 0.58681, + "grad_norm": 1.853778599934034, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 58681 + }, + { + "epoch": 0.58682, + "grad_norm": 1.8082773988507337, + "learning_rate": 0.003, + "loss": 4.2136, + "step": 58682 + }, + { + "epoch": 0.58683, + "grad_norm": 1.7622361640551627, + "learning_rate": 0.003, + "loss": 4.1972, + "step": 58683 + }, + { + "epoch": 0.58684, + "grad_norm": 2.3695032330768964, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 58684 + }, + { + "epoch": 0.58685, + "grad_norm": 2.3169416104803258, + "learning_rate": 0.003, + "loss": 4.1888, + "step": 58685 + }, + { + "epoch": 0.58686, + "grad_norm": 2.032342086263856, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 58686 + }, + { + "epoch": 0.58687, + "grad_norm": 2.408434718391147, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 58687 + }, + { + "epoch": 0.58688, + "grad_norm": 1.8173062647557086, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 58688 + }, + { + "epoch": 0.58689, + "grad_norm": 1.9632672742656159, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 58689 + }, + { + "epoch": 0.5869, + "grad_norm": 1.8908933031331157, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 58690 + }, + { + "epoch": 0.58691, + "grad_norm": 1.7580922122473317, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 58691 + }, + { + "epoch": 0.58692, + "grad_norm": 1.8781327555323504, + "learning_rate": 0.003, + "loss": 4.119, + "step": 58692 + }, + { + "epoch": 0.58693, + "grad_norm": 1.5934578586950912, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 58693 + }, + { + "epoch": 0.58694, + "grad_norm": 1.4605170571031412, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 58694 + }, + { + "epoch": 0.58695, + "grad_norm": 1.7643550388169256, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 58695 + }, + { + "epoch": 0.58696, + "grad_norm": 1.682966830939745, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 58696 + }, + { + "epoch": 0.58697, + "grad_norm": 1.928309584494891, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 58697 + }, + { + "epoch": 0.58698, + "grad_norm": 1.6920207225404982, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 58698 + }, + { + "epoch": 0.58699, + "grad_norm": 1.7084469980681425, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 58699 + }, + { + "epoch": 0.587, + "grad_norm": 1.7246760834431565, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 58700 + }, + { + "epoch": 0.58701, + "grad_norm": 1.6393099535325208, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 58701 + }, + { + "epoch": 0.58702, + "grad_norm": 1.7609680770220886, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 58702 + }, + { + "epoch": 0.58703, + "grad_norm": 1.6889177511533209, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 58703 + }, + { + "epoch": 0.58704, + "grad_norm": 1.7810672011236932, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 58704 + }, + { + "epoch": 0.58705, + "grad_norm": 1.3485533539118513, + "learning_rate": 0.003, + "loss": 4.069, + "step": 58705 + }, + { + "epoch": 0.58706, + "grad_norm": 1.9978106420519355, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 58706 + }, + { + "epoch": 0.58707, + "grad_norm": 1.4953791502157947, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 58707 + }, + { + "epoch": 0.58708, + "grad_norm": 1.74043054146384, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 58708 + }, + { + "epoch": 0.58709, + "grad_norm": 1.096513087372576, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 58709 + }, + { + "epoch": 0.5871, + "grad_norm": 2.023125883853139, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 58710 + }, + { + "epoch": 0.58711, + "grad_norm": 1.2590129113723902, + "learning_rate": 0.003, + "loss": 4.059, + "step": 58711 + }, + { + "epoch": 0.58712, + "grad_norm": 1.3802388896500664, + "learning_rate": 0.003, + "loss": 4.079, + "step": 58712 + }, + { + "epoch": 0.58713, + "grad_norm": 1.468564542952309, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 58713 + }, + { + "epoch": 0.58714, + "grad_norm": 1.6269761817239992, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 58714 + }, + { + "epoch": 0.58715, + "grad_norm": 1.424771341836781, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 58715 + }, + { + "epoch": 0.58716, + "grad_norm": 1.7975612514828243, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 58716 + }, + { + "epoch": 0.58717, + "grad_norm": 1.8334613623880833, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 58717 + }, + { + "epoch": 0.58718, + "grad_norm": 1.709925292476721, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 58718 + }, + { + "epoch": 0.58719, + "grad_norm": 1.291189463228111, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 58719 + }, + { + "epoch": 0.5872, + "grad_norm": 1.547786237127118, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 58720 + }, + { + "epoch": 0.58721, + "grad_norm": 1.7217311476525696, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 58721 + }, + { + "epoch": 0.58722, + "grad_norm": 1.9383755981833464, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 58722 + }, + { + "epoch": 0.58723, + "grad_norm": 1.1602921036310492, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 58723 + }, + { + "epoch": 0.58724, + "grad_norm": 1.6588638657225467, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 58724 + }, + { + "epoch": 0.58725, + "grad_norm": 1.2465548411788503, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 58725 + }, + { + "epoch": 0.58726, + "grad_norm": 1.9277601903244466, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 58726 + }, + { + "epoch": 0.58727, + "grad_norm": 1.1274745903293233, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 58727 + }, + { + "epoch": 0.58728, + "grad_norm": 1.6623344628937273, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 58728 + }, + { + "epoch": 0.58729, + "grad_norm": 1.5997336792563464, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 58729 + }, + { + "epoch": 0.5873, + "grad_norm": 1.6919545933491946, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 58730 + }, + { + "epoch": 0.58731, + "grad_norm": 1.547225855011808, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 58731 + }, + { + "epoch": 0.58732, + "grad_norm": 1.4308172625268534, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 58732 + }, + { + "epoch": 0.58733, + "grad_norm": 1.6034951592067912, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 58733 + }, + { + "epoch": 0.58734, + "grad_norm": 1.4309436234501383, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 58734 + }, + { + "epoch": 0.58735, + "grad_norm": 1.4303969054165024, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 58735 + }, + { + "epoch": 0.58736, + "grad_norm": 1.6730189404986209, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 58736 + }, + { + "epoch": 0.58737, + "grad_norm": 1.166767756453091, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 58737 + }, + { + "epoch": 0.58738, + "grad_norm": 2.0544916552443517, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 58738 + }, + { + "epoch": 0.58739, + "grad_norm": 1.4494179335984947, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 58739 + }, + { + "epoch": 0.5874, + "grad_norm": 1.4238136731832542, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 58740 + }, + { + "epoch": 0.58741, + "grad_norm": 1.3844331779342607, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 58741 + }, + { + "epoch": 0.58742, + "grad_norm": 2.3848474269534874, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58742 + }, + { + "epoch": 0.58743, + "grad_norm": 1.3603254667473232, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 58743 + }, + { + "epoch": 0.58744, + "grad_norm": 1.6131120299588337, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 58744 + }, + { + "epoch": 0.58745, + "grad_norm": 1.6106887905881635, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 58745 + }, + { + "epoch": 0.58746, + "grad_norm": 1.430394207478818, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 58746 + }, + { + "epoch": 0.58747, + "grad_norm": 1.9234734136751614, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 58747 + }, + { + "epoch": 0.58748, + "grad_norm": 1.5388273864862159, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 58748 + }, + { + "epoch": 0.58749, + "grad_norm": 1.5406874317803791, + "learning_rate": 0.003, + "loss": 4.058, + "step": 58749 + }, + { + "epoch": 0.5875, + "grad_norm": 1.4113293013225605, + "learning_rate": 0.003, + "loss": 4.062, + "step": 58750 + }, + { + "epoch": 0.58751, + "grad_norm": 1.5151501629423887, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 58751 + }, + { + "epoch": 0.58752, + "grad_norm": 1.801600832969483, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 58752 + }, + { + "epoch": 0.58753, + "grad_norm": 1.688041098411029, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 58753 + }, + { + "epoch": 0.58754, + "grad_norm": 1.5129136107441443, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 58754 + }, + { + "epoch": 0.58755, + "grad_norm": 1.5953520475134924, + "learning_rate": 0.003, + "loss": 4.055, + "step": 58755 + }, + { + "epoch": 0.58756, + "grad_norm": 1.4121874144394357, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 58756 + }, + { + "epoch": 0.58757, + "grad_norm": 1.4125958693821736, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 58757 + }, + { + "epoch": 0.58758, + "grad_norm": 1.3467856220801915, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 58758 + }, + { + "epoch": 0.58759, + "grad_norm": 1.5162501256152965, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 58759 + }, + { + "epoch": 0.5876, + "grad_norm": 1.3117788581894116, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 58760 + }, + { + "epoch": 0.58761, + "grad_norm": 1.9636849738301165, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 58761 + }, + { + "epoch": 0.58762, + "grad_norm": 1.3438434242976391, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 58762 + }, + { + "epoch": 0.58763, + "grad_norm": 1.6846556107553452, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 58763 + }, + { + "epoch": 0.58764, + "grad_norm": 1.55304741004106, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 58764 + }, + { + "epoch": 0.58765, + "grad_norm": 1.7591916875171476, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 58765 + }, + { + "epoch": 0.58766, + "grad_norm": 1.3787649414084389, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 58766 + }, + { + "epoch": 0.58767, + "grad_norm": 1.501547162088007, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 58767 + }, + { + "epoch": 0.58768, + "grad_norm": 1.6349641712021104, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 58768 + }, + { + "epoch": 0.58769, + "grad_norm": 1.490907873214788, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 58769 + }, + { + "epoch": 0.5877, + "grad_norm": 1.3506463722402273, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 58770 + }, + { + "epoch": 0.58771, + "grad_norm": 1.4851505814616883, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 58771 + }, + { + "epoch": 0.58772, + "grad_norm": 1.4367182850928903, + "learning_rate": 0.003, + "loss": 4.006, + "step": 58772 + }, + { + "epoch": 0.58773, + "grad_norm": 1.5678316066882945, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 58773 + }, + { + "epoch": 0.58774, + "grad_norm": 1.5517703406565497, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 58774 + }, + { + "epoch": 0.58775, + "grad_norm": 1.3745320829027692, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 58775 + }, + { + "epoch": 0.58776, + "grad_norm": 1.3858092845673702, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 58776 + }, + { + "epoch": 0.58777, + "grad_norm": 1.808130810578947, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 58777 + }, + { + "epoch": 0.58778, + "grad_norm": 1.7907132943989787, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 58778 + }, + { + "epoch": 0.58779, + "grad_norm": 1.4913991664127437, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 58779 + }, + { + "epoch": 0.5878, + "grad_norm": 1.7668858769919737, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 58780 + }, + { + "epoch": 0.58781, + "grad_norm": 1.276915323143477, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 58781 + }, + { + "epoch": 0.58782, + "grad_norm": 1.5991487229076924, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 58782 + }, + { + "epoch": 0.58783, + "grad_norm": 1.3508082509675199, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 58783 + }, + { + "epoch": 0.58784, + "grad_norm": 1.5269983501444835, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 58784 + }, + { + "epoch": 0.58785, + "grad_norm": 1.6286648853288366, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 58785 + }, + { + "epoch": 0.58786, + "grad_norm": 1.5482429811390832, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 58786 + }, + { + "epoch": 0.58787, + "grad_norm": 1.7348325279872874, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 58787 + }, + { + "epoch": 0.58788, + "grad_norm": 1.3707587862426323, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 58788 + }, + { + "epoch": 0.58789, + "grad_norm": 1.5143761592489031, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 58789 + }, + { + "epoch": 0.5879, + "grad_norm": 1.6091005374562837, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 58790 + }, + { + "epoch": 0.58791, + "grad_norm": 1.33771971541072, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 58791 + }, + { + "epoch": 0.58792, + "grad_norm": 1.8687578042312745, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 58792 + }, + { + "epoch": 0.58793, + "grad_norm": 1.5956250426120178, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 58793 + }, + { + "epoch": 0.58794, + "grad_norm": 1.5169591361235637, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 58794 + }, + { + "epoch": 0.58795, + "grad_norm": 1.6125819681436502, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 58795 + }, + { + "epoch": 0.58796, + "grad_norm": 1.454206713865769, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 58796 + }, + { + "epoch": 0.58797, + "grad_norm": 1.6722921678741274, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 58797 + }, + { + "epoch": 0.58798, + "grad_norm": 1.4838183220410217, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 58798 + }, + { + "epoch": 0.58799, + "grad_norm": 1.5694905899650582, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 58799 + }, + { + "epoch": 0.588, + "grad_norm": 1.6745015339154476, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 58800 + }, + { + "epoch": 0.58801, + "grad_norm": 1.6119576704140481, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 58801 + }, + { + "epoch": 0.58802, + "grad_norm": 1.5744532308111874, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 58802 + }, + { + "epoch": 0.58803, + "grad_norm": 1.31162019346104, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 58803 + }, + { + "epoch": 0.58804, + "grad_norm": 1.7672341175843602, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 58804 + }, + { + "epoch": 0.58805, + "grad_norm": 1.2324968299842456, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 58805 + }, + { + "epoch": 0.58806, + "grad_norm": 1.9371254061701437, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 58806 + }, + { + "epoch": 0.58807, + "grad_norm": 1.5197360016019539, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 58807 + }, + { + "epoch": 0.58808, + "grad_norm": 1.5179767351607278, + "learning_rate": 0.003, + "loss": 4.001, + "step": 58808 + }, + { + "epoch": 0.58809, + "grad_norm": 1.5134750789642601, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 58809 + }, + { + "epoch": 0.5881, + "grad_norm": 1.4693251528116553, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 58810 + }, + { + "epoch": 0.58811, + "grad_norm": 1.6819770479428895, + "learning_rate": 0.003, + "loss": 4.06, + "step": 58811 + }, + { + "epoch": 0.58812, + "grad_norm": 1.171761585703774, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 58812 + }, + { + "epoch": 0.58813, + "grad_norm": 1.7220293522304346, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 58813 + }, + { + "epoch": 0.58814, + "grad_norm": 1.4275539023695445, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 58814 + }, + { + "epoch": 0.58815, + "grad_norm": 1.2874071266333682, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 58815 + }, + { + "epoch": 0.58816, + "grad_norm": 1.7088622834038174, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 58816 + }, + { + "epoch": 0.58817, + "grad_norm": 1.1237732932074156, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 58817 + }, + { + "epoch": 0.58818, + "grad_norm": 1.7266081822882622, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 58818 + }, + { + "epoch": 0.58819, + "grad_norm": 1.1007292900092354, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 58819 + }, + { + "epoch": 0.5882, + "grad_norm": 1.7430203603070713, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 58820 + }, + { + "epoch": 0.58821, + "grad_norm": 1.6122023322107213, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 58821 + }, + { + "epoch": 0.58822, + "grad_norm": 1.8521370356213027, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 58822 + }, + { + "epoch": 0.58823, + "grad_norm": 1.4758632995699046, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 58823 + }, + { + "epoch": 0.58824, + "grad_norm": 1.5828687365685956, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 58824 + }, + { + "epoch": 0.58825, + "grad_norm": 1.5819859993482404, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 58825 + }, + { + "epoch": 0.58826, + "grad_norm": 1.4906762486185434, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 58826 + }, + { + "epoch": 0.58827, + "grad_norm": 1.4304446723803723, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 58827 + }, + { + "epoch": 0.58828, + "grad_norm": 1.670710236634712, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 58828 + }, + { + "epoch": 0.58829, + "grad_norm": 1.4649092577260518, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 58829 + }, + { + "epoch": 0.5883, + "grad_norm": 1.682445222716014, + "learning_rate": 0.003, + "loss": 4.015, + "step": 58830 + }, + { + "epoch": 0.58831, + "grad_norm": 1.4743298216183762, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 58831 + }, + { + "epoch": 0.58832, + "grad_norm": 1.6355078521554447, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 58832 + }, + { + "epoch": 0.58833, + "grad_norm": 1.5447118550102859, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 58833 + }, + { + "epoch": 0.58834, + "grad_norm": 1.2807178137065278, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 58834 + }, + { + "epoch": 0.58835, + "grad_norm": 1.4496607408285298, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 58835 + }, + { + "epoch": 0.58836, + "grad_norm": 1.4014122228908403, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 58836 + }, + { + "epoch": 0.58837, + "grad_norm": 1.1706026328620096, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 58837 + }, + { + "epoch": 0.58838, + "grad_norm": 1.7315423739119806, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 58838 + }, + { + "epoch": 0.58839, + "grad_norm": 1.295084994455735, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 58839 + }, + { + "epoch": 0.5884, + "grad_norm": 1.8702965061184906, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 58840 + }, + { + "epoch": 0.58841, + "grad_norm": 1.5573620139663251, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 58841 + }, + { + "epoch": 0.58842, + "grad_norm": 2.2229301863343682, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 58842 + }, + { + "epoch": 0.58843, + "grad_norm": 1.0647410784234337, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 58843 + }, + { + "epoch": 0.58844, + "grad_norm": 1.4744950792548714, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 58844 + }, + { + "epoch": 0.58845, + "grad_norm": 1.73401927905811, + "learning_rate": 0.003, + "loss": 4.048, + "step": 58845 + }, + { + "epoch": 0.58846, + "grad_norm": 2.0741988111611103, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 58846 + }, + { + "epoch": 0.58847, + "grad_norm": 1.4691688019790656, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 58847 + }, + { + "epoch": 0.58848, + "grad_norm": 1.739204861907959, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 58848 + }, + { + "epoch": 0.58849, + "grad_norm": 1.3050742748827115, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 58849 + }, + { + "epoch": 0.5885, + "grad_norm": 1.7135299242025237, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 58850 + }, + { + "epoch": 0.58851, + "grad_norm": 1.5562130791058262, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 58851 + }, + { + "epoch": 0.58852, + "grad_norm": 1.5215501909671176, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 58852 + }, + { + "epoch": 0.58853, + "grad_norm": 1.574810561924391, + "learning_rate": 0.003, + "loss": 4.035, + "step": 58853 + }, + { + "epoch": 0.58854, + "grad_norm": 1.3463092011870403, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 58854 + }, + { + "epoch": 0.58855, + "grad_norm": 1.6765573038151822, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 58855 + }, + { + "epoch": 0.58856, + "grad_norm": 1.2540106620672542, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 58856 + }, + { + "epoch": 0.58857, + "grad_norm": 1.6742552640968795, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 58857 + }, + { + "epoch": 0.58858, + "grad_norm": 1.3853454366650604, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 58858 + }, + { + "epoch": 0.58859, + "grad_norm": 1.647772674436609, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 58859 + }, + { + "epoch": 0.5886, + "grad_norm": 1.1371640382434967, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 58860 + }, + { + "epoch": 0.58861, + "grad_norm": 1.778429712653619, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 58861 + }, + { + "epoch": 0.58862, + "grad_norm": 1.519402085797893, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 58862 + }, + { + "epoch": 0.58863, + "grad_norm": 1.7128280443043629, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 58863 + }, + { + "epoch": 0.58864, + "grad_norm": 1.2321440691649086, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 58864 + }, + { + "epoch": 0.58865, + "grad_norm": 2.0622188687031535, + "learning_rate": 0.003, + "loss": 4.055, + "step": 58865 + }, + { + "epoch": 0.58866, + "grad_norm": 1.4700504738925662, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 58866 + }, + { + "epoch": 0.58867, + "grad_norm": 1.6347438368691058, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 58867 + }, + { + "epoch": 0.58868, + "grad_norm": 1.4335583554518923, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 58868 + }, + { + "epoch": 0.58869, + "grad_norm": 1.489965849447438, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 58869 + }, + { + "epoch": 0.5887, + "grad_norm": 1.8932533220124783, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 58870 + }, + { + "epoch": 0.58871, + "grad_norm": 1.3726186341334734, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 58871 + }, + { + "epoch": 0.58872, + "grad_norm": 1.8308869888155372, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 58872 + }, + { + "epoch": 0.58873, + "grad_norm": 1.373061531818949, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 58873 + }, + { + "epoch": 0.58874, + "grad_norm": 1.3659177136748737, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 58874 + }, + { + "epoch": 0.58875, + "grad_norm": 1.519138573924867, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 58875 + }, + { + "epoch": 0.58876, + "grad_norm": 1.5581865358525488, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 58876 + }, + { + "epoch": 0.58877, + "grad_norm": 1.5241635221702894, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 58877 + }, + { + "epoch": 0.58878, + "grad_norm": 1.5107416571427708, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 58878 + }, + { + "epoch": 0.58879, + "grad_norm": 1.438971769012427, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 58879 + }, + { + "epoch": 0.5888, + "grad_norm": 1.5800764141058088, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 58880 + }, + { + "epoch": 0.58881, + "grad_norm": 1.5352494944555573, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 58881 + }, + { + "epoch": 0.58882, + "grad_norm": 1.5900361383383055, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 58882 + }, + { + "epoch": 0.58883, + "grad_norm": 1.530232980312981, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 58883 + }, + { + "epoch": 0.58884, + "grad_norm": 1.672549268210794, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 58884 + }, + { + "epoch": 0.58885, + "grad_norm": 1.2616670626875408, + "learning_rate": 0.003, + "loss": 4.006, + "step": 58885 + }, + { + "epoch": 0.58886, + "grad_norm": 1.3743520645990528, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 58886 + }, + { + "epoch": 0.58887, + "grad_norm": 1.4901699404316702, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 58887 + }, + { + "epoch": 0.58888, + "grad_norm": 1.4906702468013884, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 58888 + }, + { + "epoch": 0.58889, + "grad_norm": 1.5396160872906532, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 58889 + }, + { + "epoch": 0.5889, + "grad_norm": 1.6288007870828398, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 58890 + }, + { + "epoch": 0.58891, + "grad_norm": 1.8179083351169, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 58891 + }, + { + "epoch": 0.58892, + "grad_norm": 1.3300009159964186, + "learning_rate": 0.003, + "loss": 4.056, + "step": 58892 + }, + { + "epoch": 0.58893, + "grad_norm": 1.6865473388552947, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 58893 + }, + { + "epoch": 0.58894, + "grad_norm": 1.7772685753444575, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 58894 + }, + { + "epoch": 0.58895, + "grad_norm": 1.5430747196072114, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 58895 + }, + { + "epoch": 0.58896, + "grad_norm": 1.3110973391453868, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 58896 + }, + { + "epoch": 0.58897, + "grad_norm": 1.7933464503639565, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 58897 + }, + { + "epoch": 0.58898, + "grad_norm": 1.3764873245092915, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 58898 + }, + { + "epoch": 0.58899, + "grad_norm": 1.697844063705507, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 58899 + }, + { + "epoch": 0.589, + "grad_norm": 1.1042303013081738, + "learning_rate": 0.003, + "loss": 4.01, + "step": 58900 + }, + { + "epoch": 0.58901, + "grad_norm": 1.7848794674646988, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 58901 + }, + { + "epoch": 0.58902, + "grad_norm": 1.502286381237986, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 58902 + }, + { + "epoch": 0.58903, + "grad_norm": 1.487415522739254, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 58903 + }, + { + "epoch": 0.58904, + "grad_norm": 1.310687069268779, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 58904 + }, + { + "epoch": 0.58905, + "grad_norm": 1.7928649553429914, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 58905 + }, + { + "epoch": 0.58906, + "grad_norm": 1.3522040536032107, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 58906 + }, + { + "epoch": 0.58907, + "grad_norm": 1.438884860655647, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 58907 + }, + { + "epoch": 0.58908, + "grad_norm": 1.912171544406681, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 58908 + }, + { + "epoch": 0.58909, + "grad_norm": 1.2885291380814592, + "learning_rate": 0.003, + "loss": 4.013, + "step": 58909 + }, + { + "epoch": 0.5891, + "grad_norm": 1.79174071901228, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 58910 + }, + { + "epoch": 0.58911, + "grad_norm": 1.2781160789973014, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 58911 + }, + { + "epoch": 0.58912, + "grad_norm": 1.8549114131251663, + "learning_rate": 0.003, + "loss": 4.034, + "step": 58912 + }, + { + "epoch": 0.58913, + "grad_norm": 1.4358520987166679, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 58913 + }, + { + "epoch": 0.58914, + "grad_norm": 1.8846970604299813, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 58914 + }, + { + "epoch": 0.58915, + "grad_norm": 1.668942644882983, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 58915 + }, + { + "epoch": 0.58916, + "grad_norm": 1.5323547677907627, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 58916 + }, + { + "epoch": 0.58917, + "grad_norm": 1.6420070607265778, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 58917 + }, + { + "epoch": 0.58918, + "grad_norm": 1.4910127479979374, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 58918 + }, + { + "epoch": 0.58919, + "grad_norm": 1.4531351026820503, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 58919 + }, + { + "epoch": 0.5892, + "grad_norm": 1.4325050444016847, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 58920 + }, + { + "epoch": 0.58921, + "grad_norm": 1.8101063575877843, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 58921 + }, + { + "epoch": 0.58922, + "grad_norm": 1.622462356286733, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 58922 + }, + { + "epoch": 0.58923, + "grad_norm": 1.7493151035834582, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 58923 + }, + { + "epoch": 0.58924, + "grad_norm": 1.5531220464925828, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 58924 + }, + { + "epoch": 0.58925, + "grad_norm": 1.5538411492160709, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 58925 + }, + { + "epoch": 0.58926, + "grad_norm": 1.4389671517297873, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 58926 + }, + { + "epoch": 0.58927, + "grad_norm": 1.554943453451828, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 58927 + }, + { + "epoch": 0.58928, + "grad_norm": 1.3921740822498079, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 58928 + }, + { + "epoch": 0.58929, + "grad_norm": 1.6459055951755335, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 58929 + }, + { + "epoch": 0.5893, + "grad_norm": 1.4416353736448961, + "learning_rate": 0.003, + "loss": 4.035, + "step": 58930 + }, + { + "epoch": 0.58931, + "grad_norm": 1.43441186260021, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 58931 + }, + { + "epoch": 0.58932, + "grad_norm": 1.6377761909733592, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 58932 + }, + { + "epoch": 0.58933, + "grad_norm": 1.5811151996762856, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 58933 + }, + { + "epoch": 0.58934, + "grad_norm": 1.569038722678743, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 58934 + }, + { + "epoch": 0.58935, + "grad_norm": 1.5361259442203041, + "learning_rate": 0.003, + "loss": 4.042, + "step": 58935 + }, + { + "epoch": 0.58936, + "grad_norm": 1.5744520949891985, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 58936 + }, + { + "epoch": 0.58937, + "grad_norm": 1.205197365857123, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 58937 + }, + { + "epoch": 0.58938, + "grad_norm": 1.7136963979502058, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 58938 + }, + { + "epoch": 0.58939, + "grad_norm": 1.1793640274964674, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 58939 + }, + { + "epoch": 0.5894, + "grad_norm": 1.820527714033749, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 58940 + }, + { + "epoch": 0.58941, + "grad_norm": 1.2152192989122752, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 58941 + }, + { + "epoch": 0.58942, + "grad_norm": 1.9355207704902264, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 58942 + }, + { + "epoch": 0.58943, + "grad_norm": 1.414136205182997, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 58943 + }, + { + "epoch": 0.58944, + "grad_norm": 1.5304594301209071, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 58944 + }, + { + "epoch": 0.58945, + "grad_norm": 1.530979396835447, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 58945 + }, + { + "epoch": 0.58946, + "grad_norm": 1.3540200908743196, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 58946 + }, + { + "epoch": 0.58947, + "grad_norm": 2.0602202262541836, + "learning_rate": 0.003, + "loss": 4.04, + "step": 58947 + }, + { + "epoch": 0.58948, + "grad_norm": 1.708039823468759, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 58948 + }, + { + "epoch": 0.58949, + "grad_norm": 1.3893197762932101, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 58949 + }, + { + "epoch": 0.5895, + "grad_norm": 1.552358117180007, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 58950 + }, + { + "epoch": 0.58951, + "grad_norm": 1.8513558305860134, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 58951 + }, + { + "epoch": 0.58952, + "grad_norm": 1.562040007312297, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 58952 + }, + { + "epoch": 0.58953, + "grad_norm": 1.73963276320866, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 58953 + }, + { + "epoch": 0.58954, + "grad_norm": 1.4546159726511791, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 58954 + }, + { + "epoch": 0.58955, + "grad_norm": 1.4388148922147768, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 58955 + }, + { + "epoch": 0.58956, + "grad_norm": 1.2481747854454757, + "learning_rate": 0.003, + "loss": 4.039, + "step": 58956 + }, + { + "epoch": 0.58957, + "grad_norm": 2.0587166510393247, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 58957 + }, + { + "epoch": 0.58958, + "grad_norm": 1.2448717611583242, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 58958 + }, + { + "epoch": 0.58959, + "grad_norm": 1.7544075394805407, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 58959 + }, + { + "epoch": 0.5896, + "grad_norm": 1.4409603262738575, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 58960 + }, + { + "epoch": 0.58961, + "grad_norm": 1.5504218176610887, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 58961 + }, + { + "epoch": 0.58962, + "grad_norm": 1.4581330222748548, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 58962 + }, + { + "epoch": 0.58963, + "grad_norm": 1.7541615747191044, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 58963 + }, + { + "epoch": 0.58964, + "grad_norm": 1.8475318755386494, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 58964 + }, + { + "epoch": 0.58965, + "grad_norm": 1.3402148589817557, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 58965 + }, + { + "epoch": 0.58966, + "grad_norm": 1.7769901747347097, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 58966 + }, + { + "epoch": 0.58967, + "grad_norm": 1.3920654207368814, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 58967 + }, + { + "epoch": 0.58968, + "grad_norm": 1.624172403799367, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 58968 + }, + { + "epoch": 0.58969, + "grad_norm": 1.5163728482304897, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 58969 + }, + { + "epoch": 0.5897, + "grad_norm": 1.2417995271144524, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 58970 + }, + { + "epoch": 0.58971, + "grad_norm": 1.6584853930970842, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 58971 + }, + { + "epoch": 0.58972, + "grad_norm": 1.233740268402035, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 58972 + }, + { + "epoch": 0.58973, + "grad_norm": 1.5671893117038997, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 58973 + }, + { + "epoch": 0.58974, + "grad_norm": 1.3199026148468316, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 58974 + }, + { + "epoch": 0.58975, + "grad_norm": 2.155071623584324, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 58975 + }, + { + "epoch": 0.58976, + "grad_norm": 1.2284644362626544, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 58976 + }, + { + "epoch": 0.58977, + "grad_norm": 1.623570415891869, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 58977 + }, + { + "epoch": 0.58978, + "grad_norm": 1.756897896245757, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 58978 + }, + { + "epoch": 0.58979, + "grad_norm": 1.5737468403762767, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 58979 + }, + { + "epoch": 0.5898, + "grad_norm": 1.3093929541767906, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 58980 + }, + { + "epoch": 0.58981, + "grad_norm": 1.5339238487190765, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 58981 + }, + { + "epoch": 0.58982, + "grad_norm": 1.2441579468815758, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 58982 + }, + { + "epoch": 0.58983, + "grad_norm": 1.6706406784668553, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 58983 + }, + { + "epoch": 0.58984, + "grad_norm": 2.185047934062051, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 58984 + }, + { + "epoch": 0.58985, + "grad_norm": 1.7184263530317547, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 58985 + }, + { + "epoch": 0.58986, + "grad_norm": 1.7648248824248103, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 58986 + }, + { + "epoch": 0.58987, + "grad_norm": 1.6775418952094587, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 58987 + }, + { + "epoch": 0.58988, + "grad_norm": 1.3689221478162548, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 58988 + }, + { + "epoch": 0.58989, + "grad_norm": 2.2951135444666475, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 58989 + }, + { + "epoch": 0.5899, + "grad_norm": 1.0635088990652828, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 58990 + }, + { + "epoch": 0.58991, + "grad_norm": 2.5713314950304214, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 58991 + }, + { + "epoch": 0.58992, + "grad_norm": 2.39631356905838, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 58992 + }, + { + "epoch": 0.58993, + "grad_norm": 3.6742522876702295, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 58993 + }, + { + "epoch": 0.58994, + "grad_norm": 2.929781111389481, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 58994 + }, + { + "epoch": 0.58995, + "grad_norm": 2.0894594287941195, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 58995 + }, + { + "epoch": 0.58996, + "grad_norm": 2.384969130826386, + "learning_rate": 0.003, + "loss": 4.1373, + "step": 58996 + }, + { + "epoch": 0.58997, + "grad_norm": 2.339208732127606, + "learning_rate": 0.003, + "loss": 4.101, + "step": 58997 + }, + { + "epoch": 0.58998, + "grad_norm": 2.5521657814127923, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 58998 + }, + { + "epoch": 0.58999, + "grad_norm": 1.5647975697812138, + "learning_rate": 0.003, + "loss": 4.066, + "step": 58999 + }, + { + "epoch": 0.59, + "grad_norm": 2.0541076055104712, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 59000 + }, + { + "epoch": 0.59001, + "grad_norm": 1.9057554390101759, + "learning_rate": 0.003, + "loss": 4.107, + "step": 59001 + }, + { + "epoch": 0.59002, + "grad_norm": 1.5824956285316005, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 59002 + }, + { + "epoch": 0.59003, + "grad_norm": 1.5151092934392043, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 59003 + }, + { + "epoch": 0.59004, + "grad_norm": 1.617616766442337, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 59004 + }, + { + "epoch": 0.59005, + "grad_norm": 1.6402114793608946, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 59005 + }, + { + "epoch": 0.59006, + "grad_norm": 1.6992377584583274, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 59006 + }, + { + "epoch": 0.59007, + "grad_norm": 1.5839255773953413, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 59007 + }, + { + "epoch": 0.59008, + "grad_norm": 1.5910319282804821, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 59008 + }, + { + "epoch": 0.59009, + "grad_norm": 1.5332468027349586, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 59009 + }, + { + "epoch": 0.5901, + "grad_norm": 1.752172125677888, + "learning_rate": 0.003, + "loss": 4.06, + "step": 59010 + }, + { + "epoch": 0.59011, + "grad_norm": 1.7805775703094497, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 59011 + }, + { + "epoch": 0.59012, + "grad_norm": 1.6263494747260543, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 59012 + }, + { + "epoch": 0.59013, + "grad_norm": 1.6972542843378995, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 59013 + }, + { + "epoch": 0.59014, + "grad_norm": 1.3859391985883187, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 59014 + }, + { + "epoch": 0.59015, + "grad_norm": 1.5536387354504892, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 59015 + }, + { + "epoch": 0.59016, + "grad_norm": 1.481917018096141, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 59016 + }, + { + "epoch": 0.59017, + "grad_norm": 1.6183776963236969, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 59017 + }, + { + "epoch": 0.59018, + "grad_norm": 1.7215252143888253, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 59018 + }, + { + "epoch": 0.59019, + "grad_norm": 1.3775165465974197, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 59019 + }, + { + "epoch": 0.5902, + "grad_norm": 1.2914629894927416, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 59020 + }, + { + "epoch": 0.59021, + "grad_norm": 1.8342160458997467, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 59021 + }, + { + "epoch": 0.59022, + "grad_norm": 1.21950080880731, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 59022 + }, + { + "epoch": 0.59023, + "grad_norm": 1.592466377983445, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 59023 + }, + { + "epoch": 0.59024, + "grad_norm": 1.2704151590277328, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 59024 + }, + { + "epoch": 0.59025, + "grad_norm": 1.7281547332966727, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 59025 + }, + { + "epoch": 0.59026, + "grad_norm": 1.3547098023370554, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 59026 + }, + { + "epoch": 0.59027, + "grad_norm": 1.4772695722697968, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 59027 + }, + { + "epoch": 0.59028, + "grad_norm": 1.875440476226843, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 59028 + }, + { + "epoch": 0.59029, + "grad_norm": 1.8914850169696107, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 59029 + }, + { + "epoch": 0.5903, + "grad_norm": 1.5014107488761346, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 59030 + }, + { + "epoch": 0.59031, + "grad_norm": 1.3766640615281607, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 59031 + }, + { + "epoch": 0.59032, + "grad_norm": 1.3941460426366055, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 59032 + }, + { + "epoch": 0.59033, + "grad_norm": 1.5022595619089079, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 59033 + }, + { + "epoch": 0.59034, + "grad_norm": 1.4657430078066511, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 59034 + }, + { + "epoch": 0.59035, + "grad_norm": 1.982467770190388, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 59035 + }, + { + "epoch": 0.59036, + "grad_norm": 1.425776330427221, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 59036 + }, + { + "epoch": 0.59037, + "grad_norm": 1.5912900508068941, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 59037 + }, + { + "epoch": 0.59038, + "grad_norm": 1.2780352923765825, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 59038 + }, + { + "epoch": 0.59039, + "grad_norm": 1.485915414653723, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 59039 + }, + { + "epoch": 0.5904, + "grad_norm": 1.3488924310686874, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 59040 + }, + { + "epoch": 0.59041, + "grad_norm": 1.8995471160142932, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 59041 + }, + { + "epoch": 0.59042, + "grad_norm": 1.245350635320777, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 59042 + }, + { + "epoch": 0.59043, + "grad_norm": 1.4500646069732372, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 59043 + }, + { + "epoch": 0.59044, + "grad_norm": 1.2932808021896334, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 59044 + }, + { + "epoch": 0.59045, + "grad_norm": 1.620165787476564, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 59045 + }, + { + "epoch": 0.59046, + "grad_norm": 1.1431785497080074, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 59046 + }, + { + "epoch": 0.59047, + "grad_norm": 1.5022361297449671, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 59047 + }, + { + "epoch": 0.59048, + "grad_norm": 1.5631871610519525, + "learning_rate": 0.003, + "loss": 4.052, + "step": 59048 + }, + { + "epoch": 0.59049, + "grad_norm": 1.3435264669855824, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 59049 + }, + { + "epoch": 0.5905, + "grad_norm": 1.855868299180346, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 59050 + }, + { + "epoch": 0.59051, + "grad_norm": 2.0348716452384283, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 59051 + }, + { + "epoch": 0.59052, + "grad_norm": 1.6234216252960907, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 59052 + }, + { + "epoch": 0.59053, + "grad_norm": 1.8123694233161014, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 59053 + }, + { + "epoch": 0.59054, + "grad_norm": 1.1977344675931043, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 59054 + }, + { + "epoch": 0.59055, + "grad_norm": 1.896896575487569, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 59055 + }, + { + "epoch": 0.59056, + "grad_norm": 1.4154435678172341, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 59056 + }, + { + "epoch": 0.59057, + "grad_norm": 1.617066410419371, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 59057 + }, + { + "epoch": 0.59058, + "grad_norm": 1.6255634209554382, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 59058 + }, + { + "epoch": 0.59059, + "grad_norm": 1.5497322964081968, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 59059 + }, + { + "epoch": 0.5906, + "grad_norm": 1.3743884922687106, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 59060 + }, + { + "epoch": 0.59061, + "grad_norm": 1.623908440074944, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 59061 + }, + { + "epoch": 0.59062, + "grad_norm": 1.7346421725436618, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 59062 + }, + { + "epoch": 0.59063, + "grad_norm": 1.1833947514367766, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 59063 + }, + { + "epoch": 0.59064, + "grad_norm": 1.721685993732229, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 59064 + }, + { + "epoch": 0.59065, + "grad_norm": 1.602267239773284, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 59065 + }, + { + "epoch": 0.59066, + "grad_norm": 1.4469254527842081, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 59066 + }, + { + "epoch": 0.59067, + "grad_norm": 1.5871605744713615, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 59067 + }, + { + "epoch": 0.59068, + "grad_norm": 1.2718267339097362, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 59068 + }, + { + "epoch": 0.59069, + "grad_norm": 1.5789143936751804, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 59069 + }, + { + "epoch": 0.5907, + "grad_norm": 1.5219219360667708, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 59070 + }, + { + "epoch": 0.59071, + "grad_norm": 1.4545537178173749, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 59071 + }, + { + "epoch": 0.59072, + "grad_norm": 1.7649470154675577, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 59072 + }, + { + "epoch": 0.59073, + "grad_norm": 1.4607801197519303, + "learning_rate": 0.003, + "loss": 4.045, + "step": 59073 + }, + { + "epoch": 0.59074, + "grad_norm": 1.395552949425407, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 59074 + }, + { + "epoch": 0.59075, + "grad_norm": 1.5908398553014869, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 59075 + }, + { + "epoch": 0.59076, + "grad_norm": 1.6383464674562662, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 59076 + }, + { + "epoch": 0.59077, + "grad_norm": 1.6079917909211139, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 59077 + }, + { + "epoch": 0.59078, + "grad_norm": 1.8693706627095688, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 59078 + }, + { + "epoch": 0.59079, + "grad_norm": 1.438675087274075, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 59079 + }, + { + "epoch": 0.5908, + "grad_norm": 1.7057613204187683, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 59080 + }, + { + "epoch": 0.59081, + "grad_norm": 1.5062767090197464, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 59081 + }, + { + "epoch": 0.59082, + "grad_norm": 1.9693640514475934, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 59082 + }, + { + "epoch": 0.59083, + "grad_norm": 1.4585781610363506, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 59083 + }, + { + "epoch": 0.59084, + "grad_norm": 1.5036412905105858, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 59084 + }, + { + "epoch": 0.59085, + "grad_norm": 1.391736604270403, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 59085 + }, + { + "epoch": 0.59086, + "grad_norm": 1.6359224976120008, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 59086 + }, + { + "epoch": 0.59087, + "grad_norm": 1.5160419938741088, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 59087 + }, + { + "epoch": 0.59088, + "grad_norm": 1.7755475237244043, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 59088 + }, + { + "epoch": 0.59089, + "grad_norm": 1.3975584192180313, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 59089 + }, + { + "epoch": 0.5909, + "grad_norm": 1.5777301051639376, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 59090 + }, + { + "epoch": 0.59091, + "grad_norm": 1.57867552393226, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 59091 + }, + { + "epoch": 0.59092, + "grad_norm": 1.5362730600209382, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 59092 + }, + { + "epoch": 0.59093, + "grad_norm": 1.4961954595937512, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 59093 + }, + { + "epoch": 0.59094, + "grad_norm": 1.299173177570075, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 59094 + }, + { + "epoch": 0.59095, + "grad_norm": 2.0600719633774167, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 59095 + }, + { + "epoch": 0.59096, + "grad_norm": 1.533769876080057, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 59096 + }, + { + "epoch": 0.59097, + "grad_norm": 1.6521133960279204, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 59097 + }, + { + "epoch": 0.59098, + "grad_norm": 1.346579068604454, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 59098 + }, + { + "epoch": 0.59099, + "grad_norm": 1.8360382485355806, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 59099 + }, + { + "epoch": 0.591, + "grad_norm": 1.4734039961119676, + "learning_rate": 0.003, + "loss": 4.043, + "step": 59100 + }, + { + "epoch": 0.59101, + "grad_norm": 1.4141607476968856, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 59101 + }, + { + "epoch": 0.59102, + "grad_norm": 1.622413694718851, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 59102 + }, + { + "epoch": 0.59103, + "grad_norm": 1.4429037905866695, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 59103 + }, + { + "epoch": 0.59104, + "grad_norm": 1.5929341808459225, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 59104 + }, + { + "epoch": 0.59105, + "grad_norm": 1.4818739069659532, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 59105 + }, + { + "epoch": 0.59106, + "grad_norm": 1.740819213604534, + "learning_rate": 0.003, + "loss": 4.05, + "step": 59106 + }, + { + "epoch": 0.59107, + "grad_norm": 1.683912617209508, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 59107 + }, + { + "epoch": 0.59108, + "grad_norm": 1.2396830574224031, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 59108 + }, + { + "epoch": 0.59109, + "grad_norm": 1.9139630915144068, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 59109 + }, + { + "epoch": 0.5911, + "grad_norm": 1.4223873452465947, + "learning_rate": 0.003, + "loss": 4.046, + "step": 59110 + }, + { + "epoch": 0.59111, + "grad_norm": 1.7200848008499108, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 59111 + }, + { + "epoch": 0.59112, + "grad_norm": 1.4522868733409864, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 59112 + }, + { + "epoch": 0.59113, + "grad_norm": 1.5879340750173103, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 59113 + }, + { + "epoch": 0.59114, + "grad_norm": 1.355845789329188, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 59114 + }, + { + "epoch": 0.59115, + "grad_norm": 1.4928321504855362, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 59115 + }, + { + "epoch": 0.59116, + "grad_norm": 1.7047163862552126, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 59116 + }, + { + "epoch": 0.59117, + "grad_norm": 1.3498637853029278, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 59117 + }, + { + "epoch": 0.59118, + "grad_norm": 1.5810426937004627, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 59118 + }, + { + "epoch": 0.59119, + "grad_norm": 1.4569655626575795, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 59119 + }, + { + "epoch": 0.5912, + "grad_norm": 1.5663871718681843, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 59120 + }, + { + "epoch": 0.59121, + "grad_norm": 1.61593914494703, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59121 + }, + { + "epoch": 0.59122, + "grad_norm": 1.5309989846564427, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 59122 + }, + { + "epoch": 0.59123, + "grad_norm": 1.6674254961545076, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 59123 + }, + { + "epoch": 0.59124, + "grad_norm": 1.6031343440806813, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 59124 + }, + { + "epoch": 0.59125, + "grad_norm": 1.7145939556401923, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 59125 + }, + { + "epoch": 0.59126, + "grad_norm": 1.856841339017944, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 59126 + }, + { + "epoch": 0.59127, + "grad_norm": 1.3935165556080766, + "learning_rate": 0.003, + "loss": 4.011, + "step": 59127 + }, + { + "epoch": 0.59128, + "grad_norm": 1.530638838822324, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 59128 + }, + { + "epoch": 0.59129, + "grad_norm": 1.9031157857813263, + "learning_rate": 0.003, + "loss": 4.018, + "step": 59129 + }, + { + "epoch": 0.5913, + "grad_norm": 1.496498404160746, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 59130 + }, + { + "epoch": 0.59131, + "grad_norm": 1.5150681776838646, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 59131 + }, + { + "epoch": 0.59132, + "grad_norm": 1.6392743074203082, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 59132 + }, + { + "epoch": 0.59133, + "grad_norm": 1.2692419542343363, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 59133 + }, + { + "epoch": 0.59134, + "grad_norm": 1.3932503705059882, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 59134 + }, + { + "epoch": 0.59135, + "grad_norm": 1.4587378952203158, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 59135 + }, + { + "epoch": 0.59136, + "grad_norm": 1.7294377261683378, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 59136 + }, + { + "epoch": 0.59137, + "grad_norm": 1.7666556308102233, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 59137 + }, + { + "epoch": 0.59138, + "grad_norm": 1.7105678148108963, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 59138 + }, + { + "epoch": 0.59139, + "grad_norm": 1.4468657544228511, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 59139 + }, + { + "epoch": 0.5914, + "grad_norm": 1.4533586767285769, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 59140 + }, + { + "epoch": 0.59141, + "grad_norm": 1.5713718237614671, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59141 + }, + { + "epoch": 0.59142, + "grad_norm": 1.4552721437781997, + "learning_rate": 0.003, + "loss": 4.027, + "step": 59142 + }, + { + "epoch": 0.59143, + "grad_norm": 1.5204863976356893, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 59143 + }, + { + "epoch": 0.59144, + "grad_norm": 1.7026729789981936, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 59144 + }, + { + "epoch": 0.59145, + "grad_norm": 1.455811173729769, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 59145 + }, + { + "epoch": 0.59146, + "grad_norm": 1.846077389183227, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 59146 + }, + { + "epoch": 0.59147, + "grad_norm": 1.2739157402997021, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 59147 + }, + { + "epoch": 0.59148, + "grad_norm": 1.9295866083915891, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 59148 + }, + { + "epoch": 0.59149, + "grad_norm": 1.3783576982609083, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 59149 + }, + { + "epoch": 0.5915, + "grad_norm": 1.4836758796974692, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 59150 + }, + { + "epoch": 0.59151, + "grad_norm": 1.8045363131437158, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 59151 + }, + { + "epoch": 0.59152, + "grad_norm": 2.237093089611085, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 59152 + }, + { + "epoch": 0.59153, + "grad_norm": 1.495582923997783, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 59153 + }, + { + "epoch": 0.59154, + "grad_norm": 1.573622279160979, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 59154 + }, + { + "epoch": 0.59155, + "grad_norm": 1.5770969469759197, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 59155 + }, + { + "epoch": 0.59156, + "grad_norm": 1.622457250653433, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 59156 + }, + { + "epoch": 0.59157, + "grad_norm": 1.5788521648123246, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 59157 + }, + { + "epoch": 0.59158, + "grad_norm": 1.4514431602017464, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 59158 + }, + { + "epoch": 0.59159, + "grad_norm": 1.7202287234088096, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 59159 + }, + { + "epoch": 0.5916, + "grad_norm": 1.3246446843798902, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 59160 + }, + { + "epoch": 0.59161, + "grad_norm": 1.6480798983767606, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 59161 + }, + { + "epoch": 0.59162, + "grad_norm": 1.5716163990021568, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 59162 + }, + { + "epoch": 0.59163, + "grad_norm": 1.570289298834415, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 59163 + }, + { + "epoch": 0.59164, + "grad_norm": 1.307825242997183, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 59164 + }, + { + "epoch": 0.59165, + "grad_norm": 1.891845643575484, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 59165 + }, + { + "epoch": 0.59166, + "grad_norm": 1.1519379784123587, + "learning_rate": 0.003, + "loss": 4.051, + "step": 59166 + }, + { + "epoch": 0.59167, + "grad_norm": 2.221113665090166, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 59167 + }, + { + "epoch": 0.59168, + "grad_norm": 1.1141857893617997, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 59168 + }, + { + "epoch": 0.59169, + "grad_norm": 2.316798813991551, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 59169 + }, + { + "epoch": 0.5917, + "grad_norm": 1.4446738171342843, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 59170 + }, + { + "epoch": 0.59171, + "grad_norm": 1.4407440246104246, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 59171 + }, + { + "epoch": 0.59172, + "grad_norm": 1.6748500513262952, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 59172 + }, + { + "epoch": 0.59173, + "grad_norm": 1.5550051358071975, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 59173 + }, + { + "epoch": 0.59174, + "grad_norm": 1.5773443298169705, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 59174 + }, + { + "epoch": 0.59175, + "grad_norm": 1.6384054294435546, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 59175 + }, + { + "epoch": 0.59176, + "grad_norm": 1.383250320700246, + "learning_rate": 0.003, + "loss": 4.044, + "step": 59176 + }, + { + "epoch": 0.59177, + "grad_norm": 1.7730851625152837, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 59177 + }, + { + "epoch": 0.59178, + "grad_norm": 1.3859822194471354, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 59178 + }, + { + "epoch": 0.59179, + "grad_norm": 1.5649556226334767, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 59179 + }, + { + "epoch": 0.5918, + "grad_norm": 1.6979903010555701, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 59180 + }, + { + "epoch": 0.59181, + "grad_norm": 1.2961338738312633, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 59181 + }, + { + "epoch": 0.59182, + "grad_norm": 1.9400132250444013, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 59182 + }, + { + "epoch": 0.59183, + "grad_norm": 1.4159550556124252, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 59183 + }, + { + "epoch": 0.59184, + "grad_norm": 1.430199620600649, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 59184 + }, + { + "epoch": 0.59185, + "grad_norm": 1.3499025551705186, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 59185 + }, + { + "epoch": 0.59186, + "grad_norm": 1.8807216574121037, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 59186 + }, + { + "epoch": 0.59187, + "grad_norm": 1.627976303449672, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 59187 + }, + { + "epoch": 0.59188, + "grad_norm": 1.4196648721412388, + "learning_rate": 0.003, + "loss": 4.055, + "step": 59188 + }, + { + "epoch": 0.59189, + "grad_norm": 1.9099009075516917, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 59189 + }, + { + "epoch": 0.5919, + "grad_norm": 1.290397132507347, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 59190 + }, + { + "epoch": 0.59191, + "grad_norm": 1.3806440496856869, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 59191 + }, + { + "epoch": 0.59192, + "grad_norm": 1.8307622053110544, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 59192 + }, + { + "epoch": 0.59193, + "grad_norm": 1.6019277847914466, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 59193 + }, + { + "epoch": 0.59194, + "grad_norm": 1.2256861847257494, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 59194 + }, + { + "epoch": 0.59195, + "grad_norm": 1.4513522500795943, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 59195 + }, + { + "epoch": 0.59196, + "grad_norm": 1.3458039021741341, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 59196 + }, + { + "epoch": 0.59197, + "grad_norm": 1.6681426089128486, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 59197 + }, + { + "epoch": 0.59198, + "grad_norm": 1.2150105168064396, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 59198 + }, + { + "epoch": 0.59199, + "grad_norm": 1.6950226215305637, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 59199 + }, + { + "epoch": 0.592, + "grad_norm": 1.3367287411174367, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 59200 + }, + { + "epoch": 0.59201, + "grad_norm": 1.6876398484295128, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 59201 + }, + { + "epoch": 0.59202, + "grad_norm": 1.4474976197422167, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 59202 + }, + { + "epoch": 0.59203, + "grad_norm": 1.6146756035175482, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 59203 + }, + { + "epoch": 0.59204, + "grad_norm": 1.2990617441730439, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 59204 + }, + { + "epoch": 0.59205, + "grad_norm": 1.543093894699601, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 59205 + }, + { + "epoch": 0.59206, + "grad_norm": 1.282402386747916, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 59206 + }, + { + "epoch": 0.59207, + "grad_norm": 1.4192863531145468, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 59207 + }, + { + "epoch": 0.59208, + "grad_norm": 1.4096302209359872, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 59208 + }, + { + "epoch": 0.59209, + "grad_norm": 1.792245957417751, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 59209 + }, + { + "epoch": 0.5921, + "grad_norm": 1.8235273185374203, + "learning_rate": 0.003, + "loss": 4.055, + "step": 59210 + }, + { + "epoch": 0.59211, + "grad_norm": 1.8629006013515437, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 59211 + }, + { + "epoch": 0.59212, + "grad_norm": 1.4963626099704848, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 59212 + }, + { + "epoch": 0.59213, + "grad_norm": 1.602756471454037, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 59213 + }, + { + "epoch": 0.59214, + "grad_norm": 1.7811131534904447, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 59214 + }, + { + "epoch": 0.59215, + "grad_norm": 1.6924305184688202, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 59215 + }, + { + "epoch": 0.59216, + "grad_norm": 1.4718494448002488, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 59216 + }, + { + "epoch": 0.59217, + "grad_norm": 1.5568166125701428, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 59217 + }, + { + "epoch": 0.59218, + "grad_norm": 1.624166651973885, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 59218 + }, + { + "epoch": 0.59219, + "grad_norm": 1.5153688824254732, + "learning_rate": 0.003, + "loss": 3.998, + "step": 59219 + }, + { + "epoch": 0.5922, + "grad_norm": 1.408737258456834, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 59220 + }, + { + "epoch": 0.59221, + "grad_norm": 1.6078698524752029, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 59221 + }, + { + "epoch": 0.59222, + "grad_norm": 1.729287796928545, + "learning_rate": 0.003, + "loss": 4.05, + "step": 59222 + }, + { + "epoch": 0.59223, + "grad_norm": 1.3994146310328097, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 59223 + }, + { + "epoch": 0.59224, + "grad_norm": 1.6498018822159068, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 59224 + }, + { + "epoch": 0.59225, + "grad_norm": 1.3377668145731307, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 59225 + }, + { + "epoch": 0.59226, + "grad_norm": 1.8009505541343165, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 59226 + }, + { + "epoch": 0.59227, + "grad_norm": 1.287682287402027, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 59227 + }, + { + "epoch": 0.59228, + "grad_norm": 1.7683211712152151, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 59228 + }, + { + "epoch": 0.59229, + "grad_norm": 1.5904871438729595, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59229 + }, + { + "epoch": 0.5923, + "grad_norm": 1.4857266561123748, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 59230 + }, + { + "epoch": 0.59231, + "grad_norm": 1.4646508920298182, + "learning_rate": 0.003, + "loss": 4.035, + "step": 59231 + }, + { + "epoch": 0.59232, + "grad_norm": 1.433316082681474, + "learning_rate": 0.003, + "loss": 4.045, + "step": 59232 + }, + { + "epoch": 0.59233, + "grad_norm": 1.4860819542209749, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 59233 + }, + { + "epoch": 0.59234, + "grad_norm": 1.8418891363829706, + "learning_rate": 0.003, + "loss": 4.044, + "step": 59234 + }, + { + "epoch": 0.59235, + "grad_norm": 1.5653002657469308, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 59235 + }, + { + "epoch": 0.59236, + "grad_norm": 1.7364643166159088, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 59236 + }, + { + "epoch": 0.59237, + "grad_norm": 1.4953212281336752, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 59237 + }, + { + "epoch": 0.59238, + "grad_norm": 1.2198180479023746, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 59238 + }, + { + "epoch": 0.59239, + "grad_norm": 1.9553573129441153, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 59239 + }, + { + "epoch": 0.5924, + "grad_norm": 1.2751809702638595, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 59240 + }, + { + "epoch": 0.59241, + "grad_norm": 1.671528798177877, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 59241 + }, + { + "epoch": 0.59242, + "grad_norm": 1.5124548528758432, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 59242 + }, + { + "epoch": 0.59243, + "grad_norm": 1.5758200285914061, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 59243 + }, + { + "epoch": 0.59244, + "grad_norm": 1.7109727894455011, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 59244 + }, + { + "epoch": 0.59245, + "grad_norm": 1.4878160198512174, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 59245 + }, + { + "epoch": 0.59246, + "grad_norm": 2.0255559673277648, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 59246 + }, + { + "epoch": 0.59247, + "grad_norm": 1.1488001870092999, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 59247 + }, + { + "epoch": 0.59248, + "grad_norm": 1.6287790454651263, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59248 + }, + { + "epoch": 0.59249, + "grad_norm": 1.8159297627572335, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 59249 + }, + { + "epoch": 0.5925, + "grad_norm": 1.8987908348358271, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 59250 + }, + { + "epoch": 0.59251, + "grad_norm": 1.1483395640955651, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 59251 + }, + { + "epoch": 0.59252, + "grad_norm": 1.4762427635311084, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 59252 + }, + { + "epoch": 0.59253, + "grad_norm": 1.3108791979997474, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 59253 + }, + { + "epoch": 0.59254, + "grad_norm": 1.6591919415013996, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 59254 + }, + { + "epoch": 0.59255, + "grad_norm": 1.5074437632504962, + "learning_rate": 0.003, + "loss": 4.049, + "step": 59255 + }, + { + "epoch": 0.59256, + "grad_norm": 1.9654439241242143, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 59256 + }, + { + "epoch": 0.59257, + "grad_norm": 1.4382225497434002, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 59257 + }, + { + "epoch": 0.59258, + "grad_norm": 1.4458253259611458, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59258 + }, + { + "epoch": 0.59259, + "grad_norm": 1.610807465592953, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 59259 + }, + { + "epoch": 0.5926, + "grad_norm": 1.6657493406088288, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 59260 + }, + { + "epoch": 0.59261, + "grad_norm": 1.5499518747874614, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 59261 + }, + { + "epoch": 0.59262, + "grad_norm": 1.3127893243506539, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 59262 + }, + { + "epoch": 0.59263, + "grad_norm": 1.8490724461073156, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 59263 + }, + { + "epoch": 0.59264, + "grad_norm": 1.6657760054436601, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 59264 + }, + { + "epoch": 0.59265, + "grad_norm": 1.8611948907870484, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 59265 + }, + { + "epoch": 0.59266, + "grad_norm": 1.3228678164347443, + "learning_rate": 0.003, + "loss": 4.019, + "step": 59266 + }, + { + "epoch": 0.59267, + "grad_norm": 1.340726408302849, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 59267 + }, + { + "epoch": 0.59268, + "grad_norm": 1.6341923974752453, + "learning_rate": 0.003, + "loss": 4.03, + "step": 59268 + }, + { + "epoch": 0.59269, + "grad_norm": 1.7614740727165366, + "learning_rate": 0.003, + "loss": 4.023, + "step": 59269 + }, + { + "epoch": 0.5927, + "grad_norm": 1.5918143487394611, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 59270 + }, + { + "epoch": 0.59271, + "grad_norm": 1.6694009487432646, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 59271 + }, + { + "epoch": 0.59272, + "grad_norm": 1.3968568322161967, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 59272 + }, + { + "epoch": 0.59273, + "grad_norm": 1.520233784182034, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 59273 + }, + { + "epoch": 0.59274, + "grad_norm": 1.4007588546661522, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 59274 + }, + { + "epoch": 0.59275, + "grad_norm": 1.4394636388811817, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 59275 + }, + { + "epoch": 0.59276, + "grad_norm": 1.8783916362543003, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 59276 + }, + { + "epoch": 0.59277, + "grad_norm": 1.3434355697211489, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 59277 + }, + { + "epoch": 0.59278, + "grad_norm": 1.5663285875660713, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 59278 + }, + { + "epoch": 0.59279, + "grad_norm": 1.5680402779823321, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 59279 + }, + { + "epoch": 0.5928, + "grad_norm": 1.5306055491015607, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 59280 + }, + { + "epoch": 0.59281, + "grad_norm": 1.4654070763366207, + "learning_rate": 0.003, + "loss": 4.039, + "step": 59281 + }, + { + "epoch": 0.59282, + "grad_norm": 1.5684322972482467, + "learning_rate": 0.003, + "loss": 4.045, + "step": 59282 + }, + { + "epoch": 0.59283, + "grad_norm": 1.8817574558068386, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 59283 + }, + { + "epoch": 0.59284, + "grad_norm": 1.7230107146125528, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 59284 + }, + { + "epoch": 0.59285, + "grad_norm": 1.6748934792742203, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 59285 + }, + { + "epoch": 0.59286, + "grad_norm": 1.570850286296761, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 59286 + }, + { + "epoch": 0.59287, + "grad_norm": 1.4734655576049578, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59287 + }, + { + "epoch": 0.59288, + "grad_norm": 1.8268668174995912, + "learning_rate": 0.003, + "loss": 4.008, + "step": 59288 + }, + { + "epoch": 0.59289, + "grad_norm": 1.406424948153058, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 59289 + }, + { + "epoch": 0.5929, + "grad_norm": 1.7045727763408072, + "learning_rate": 0.003, + "loss": 4.044, + "step": 59290 + }, + { + "epoch": 0.59291, + "grad_norm": 1.26298849644297, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 59291 + }, + { + "epoch": 0.59292, + "grad_norm": 1.549311960154481, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 59292 + }, + { + "epoch": 0.59293, + "grad_norm": 1.5337439534745765, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 59293 + }, + { + "epoch": 0.59294, + "grad_norm": 1.6399721700072283, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 59294 + }, + { + "epoch": 0.59295, + "grad_norm": 1.9266859440183983, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 59295 + }, + { + "epoch": 0.59296, + "grad_norm": 1.3968783547352974, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 59296 + }, + { + "epoch": 0.59297, + "grad_norm": 1.5649763127306089, + "learning_rate": 0.003, + "loss": 4.04, + "step": 59297 + }, + { + "epoch": 0.59298, + "grad_norm": 1.4555890852938216, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 59298 + }, + { + "epoch": 0.59299, + "grad_norm": 1.9497681099637678, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 59299 + }, + { + "epoch": 0.593, + "grad_norm": 1.2146705233700625, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 59300 + }, + { + "epoch": 0.59301, + "grad_norm": 1.777361635385311, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 59301 + }, + { + "epoch": 0.59302, + "grad_norm": 1.680347680221595, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 59302 + }, + { + "epoch": 0.59303, + "grad_norm": 1.5230667067502193, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 59303 + }, + { + "epoch": 0.59304, + "grad_norm": 1.4580782888275317, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 59304 + }, + { + "epoch": 0.59305, + "grad_norm": 1.6325930764390992, + "learning_rate": 0.003, + "loss": 4.014, + "step": 59305 + }, + { + "epoch": 0.59306, + "grad_norm": 1.47156714388376, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 59306 + }, + { + "epoch": 0.59307, + "grad_norm": 1.26947723937911, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 59307 + }, + { + "epoch": 0.59308, + "grad_norm": 1.5975098906856335, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 59308 + }, + { + "epoch": 0.59309, + "grad_norm": 1.477335817896129, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 59309 + }, + { + "epoch": 0.5931, + "grad_norm": 1.4653563587998424, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 59310 + }, + { + "epoch": 0.59311, + "grad_norm": 1.5057522380107846, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 59311 + }, + { + "epoch": 0.59312, + "grad_norm": 1.4106937282754668, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 59312 + }, + { + "epoch": 0.59313, + "grad_norm": 1.576360414847425, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 59313 + }, + { + "epoch": 0.59314, + "grad_norm": 1.318693888178584, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 59314 + }, + { + "epoch": 0.59315, + "grad_norm": 1.8200356954693029, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 59315 + }, + { + "epoch": 0.59316, + "grad_norm": 1.2120243119682295, + "learning_rate": 0.003, + "loss": 4.022, + "step": 59316 + }, + { + "epoch": 0.59317, + "grad_norm": 1.670913339155919, + "learning_rate": 0.003, + "loss": 4.013, + "step": 59317 + }, + { + "epoch": 0.59318, + "grad_norm": 1.670046063711645, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 59318 + }, + { + "epoch": 0.59319, + "grad_norm": 1.6988822597619369, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 59319 + }, + { + "epoch": 0.5932, + "grad_norm": 1.9229231057042278, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 59320 + }, + { + "epoch": 0.59321, + "grad_norm": 1.431583034192371, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 59321 + }, + { + "epoch": 0.59322, + "grad_norm": 1.7036681566246068, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 59322 + }, + { + "epoch": 0.59323, + "grad_norm": 1.3014544624018203, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 59323 + }, + { + "epoch": 0.59324, + "grad_norm": 1.9779621731637693, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 59324 + }, + { + "epoch": 0.59325, + "grad_norm": 1.2382232296953644, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 59325 + }, + { + "epoch": 0.59326, + "grad_norm": 1.6333737176126977, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 59326 + }, + { + "epoch": 0.59327, + "grad_norm": 1.3787407440952606, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 59327 + }, + { + "epoch": 0.59328, + "grad_norm": 1.6784154853305584, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 59328 + }, + { + "epoch": 0.59329, + "grad_norm": 1.7642524906723704, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 59329 + }, + { + "epoch": 0.5933, + "grad_norm": 1.459711108595886, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 59330 + }, + { + "epoch": 0.59331, + "grad_norm": 1.709715652897939, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 59331 + }, + { + "epoch": 0.59332, + "grad_norm": 1.4236039694673794, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 59332 + }, + { + "epoch": 0.59333, + "grad_norm": 1.2164269988917282, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 59333 + }, + { + "epoch": 0.59334, + "grad_norm": 1.8125683324567923, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 59334 + }, + { + "epoch": 0.59335, + "grad_norm": 1.2999227232543826, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 59335 + }, + { + "epoch": 0.59336, + "grad_norm": 1.6991920781760663, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 59336 + }, + { + "epoch": 0.59337, + "grad_norm": 1.4526243579002416, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 59337 + }, + { + "epoch": 0.59338, + "grad_norm": 1.740373100941739, + "learning_rate": 0.003, + "loss": 3.989, + "step": 59338 + }, + { + "epoch": 0.59339, + "grad_norm": 1.3555491299191142, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 59339 + }, + { + "epoch": 0.5934, + "grad_norm": 1.6201069763274516, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 59340 + }, + { + "epoch": 0.59341, + "grad_norm": 1.4225543677228936, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 59341 + }, + { + "epoch": 0.59342, + "grad_norm": 1.508266724555532, + "learning_rate": 0.003, + "loss": 4.015, + "step": 59342 + }, + { + "epoch": 0.59343, + "grad_norm": 1.5197781360274356, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 59343 + }, + { + "epoch": 0.59344, + "grad_norm": 1.3118786886719287, + "learning_rate": 0.003, + "loss": 4.032, + "step": 59344 + }, + { + "epoch": 0.59345, + "grad_norm": 1.6464970765187945, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 59345 + }, + { + "epoch": 0.59346, + "grad_norm": 1.1681616715221645, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 59346 + }, + { + "epoch": 0.59347, + "grad_norm": 1.821892185149168, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 59347 + }, + { + "epoch": 0.59348, + "grad_norm": 1.4199217289652628, + "learning_rate": 0.003, + "loss": 4.059, + "step": 59348 + }, + { + "epoch": 0.59349, + "grad_norm": 2.193108947190313, + "learning_rate": 0.003, + "loss": 4.011, + "step": 59349 + }, + { + "epoch": 0.5935, + "grad_norm": 1.4788977123103235, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 59350 + }, + { + "epoch": 0.59351, + "grad_norm": 1.7018923617368724, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 59351 + }, + { + "epoch": 0.59352, + "grad_norm": 1.6394501779211241, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 59352 + }, + { + "epoch": 0.59353, + "grad_norm": 1.4753386302083737, + "learning_rate": 0.003, + "loss": 4.016, + "step": 59353 + }, + { + "epoch": 0.59354, + "grad_norm": 1.8256258593240682, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 59354 + }, + { + "epoch": 0.59355, + "grad_norm": 1.594113780920626, + "learning_rate": 0.003, + "loss": 4.017, + "step": 59355 + }, + { + "epoch": 0.59356, + "grad_norm": 1.3676792620766705, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 59356 + }, + { + "epoch": 0.59357, + "grad_norm": 1.5007222473023072, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 59357 + }, + { + "epoch": 0.59358, + "grad_norm": 1.4164997906256946, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 59358 + }, + { + "epoch": 0.59359, + "grad_norm": 1.5451537836319273, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 59359 + }, + { + "epoch": 0.5936, + "grad_norm": 1.4723653796884468, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 59360 + }, + { + "epoch": 0.59361, + "grad_norm": 1.5888666597134207, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 59361 + }, + { + "epoch": 0.59362, + "grad_norm": 1.411363796837713, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 59362 + }, + { + "epoch": 0.59363, + "grad_norm": 1.5803632800263774, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59363 + }, + { + "epoch": 0.59364, + "grad_norm": 1.5009490008494846, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 59364 + }, + { + "epoch": 0.59365, + "grad_norm": 1.7815792109903361, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 59365 + }, + { + "epoch": 0.59366, + "grad_norm": 1.3294139512681429, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59366 + }, + { + "epoch": 0.59367, + "grad_norm": 1.5897501011816084, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 59367 + }, + { + "epoch": 0.59368, + "grad_norm": 1.6002064675806984, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 59368 + }, + { + "epoch": 0.59369, + "grad_norm": 1.4633724230529521, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 59369 + }, + { + "epoch": 0.5937, + "grad_norm": 1.5953039103877424, + "learning_rate": 0.003, + "loss": 4.043, + "step": 59370 + }, + { + "epoch": 0.59371, + "grad_norm": 1.4474811670057335, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 59371 + }, + { + "epoch": 0.59372, + "grad_norm": 1.803296842890718, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 59372 + }, + { + "epoch": 0.59373, + "grad_norm": 1.288126996635073, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 59373 + }, + { + "epoch": 0.59374, + "grad_norm": 1.7436112553584218, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 59374 + }, + { + "epoch": 0.59375, + "grad_norm": 1.5678681489654043, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 59375 + }, + { + "epoch": 0.59376, + "grad_norm": 1.532734205848502, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 59376 + }, + { + "epoch": 0.59377, + "grad_norm": 1.5384899433785726, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 59377 + }, + { + "epoch": 0.59378, + "grad_norm": 1.4955858490041187, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 59378 + }, + { + "epoch": 0.59379, + "grad_norm": 1.6844843199001773, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 59379 + }, + { + "epoch": 0.5938, + "grad_norm": 1.489624243127864, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 59380 + }, + { + "epoch": 0.59381, + "grad_norm": 1.7191304844256594, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 59381 + }, + { + "epoch": 0.59382, + "grad_norm": 1.8047994082542445, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 59382 + }, + { + "epoch": 0.59383, + "grad_norm": 1.7345171699265303, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 59383 + }, + { + "epoch": 0.59384, + "grad_norm": 1.4242211808029688, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 59384 + }, + { + "epoch": 0.59385, + "grad_norm": 1.7690883596983518, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 59385 + }, + { + "epoch": 0.59386, + "grad_norm": 1.3277502015613305, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 59386 + }, + { + "epoch": 0.59387, + "grad_norm": 1.6206422338459918, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 59387 + }, + { + "epoch": 0.59388, + "grad_norm": 1.5693134623180394, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 59388 + }, + { + "epoch": 0.59389, + "grad_norm": 1.6359717631360096, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 59389 + }, + { + "epoch": 0.5939, + "grad_norm": 1.474448807075897, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 59390 + }, + { + "epoch": 0.59391, + "grad_norm": 1.7194146454584442, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 59391 + }, + { + "epoch": 0.59392, + "grad_norm": 1.421993970901223, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 59392 + }, + { + "epoch": 0.59393, + "grad_norm": 1.5109935511969195, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 59393 + }, + { + "epoch": 0.59394, + "grad_norm": 1.8184049068747847, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 59394 + }, + { + "epoch": 0.59395, + "grad_norm": 1.5248611165084363, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 59395 + }, + { + "epoch": 0.59396, + "grad_norm": 1.5365770920071833, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 59396 + }, + { + "epoch": 0.59397, + "grad_norm": 1.585786320151067, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 59397 + }, + { + "epoch": 0.59398, + "grad_norm": 1.5995674106981244, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 59398 + }, + { + "epoch": 0.59399, + "grad_norm": 1.2983037285452494, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 59399 + }, + { + "epoch": 0.594, + "grad_norm": 1.7691128042415971, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 59400 + }, + { + "epoch": 0.59401, + "grad_norm": 1.3914553277660335, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 59401 + }, + { + "epoch": 0.59402, + "grad_norm": 1.735238042606836, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 59402 + }, + { + "epoch": 0.59403, + "grad_norm": 1.573318163924446, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 59403 + }, + { + "epoch": 0.59404, + "grad_norm": 1.3120725619293117, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 59404 + }, + { + "epoch": 0.59405, + "grad_norm": 1.6486499883300456, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 59405 + }, + { + "epoch": 0.59406, + "grad_norm": 1.4080352799807738, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 59406 + }, + { + "epoch": 0.59407, + "grad_norm": 1.6734143598708555, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 59407 + }, + { + "epoch": 0.59408, + "grad_norm": 1.2245261688284306, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 59408 + }, + { + "epoch": 0.59409, + "grad_norm": 1.6867109410760595, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 59409 + }, + { + "epoch": 0.5941, + "grad_norm": 1.749330418473594, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59410 + }, + { + "epoch": 0.59411, + "grad_norm": 1.6256148079244133, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 59411 + }, + { + "epoch": 0.59412, + "grad_norm": 1.689163284632774, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 59412 + }, + { + "epoch": 0.59413, + "grad_norm": 1.572816712788662, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 59413 + }, + { + "epoch": 0.59414, + "grad_norm": 1.5190274408609965, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 59414 + }, + { + "epoch": 0.59415, + "grad_norm": 1.5529955388259344, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 59415 + }, + { + "epoch": 0.59416, + "grad_norm": 1.5502973812237268, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 59416 + }, + { + "epoch": 0.59417, + "grad_norm": 1.3486783235841695, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 59417 + }, + { + "epoch": 0.59418, + "grad_norm": 1.5194830372375672, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 59418 + }, + { + "epoch": 0.59419, + "grad_norm": 1.6262292904537858, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 59419 + }, + { + "epoch": 0.5942, + "grad_norm": 1.522467334655696, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 59420 + }, + { + "epoch": 0.59421, + "grad_norm": 1.604492949595339, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 59421 + }, + { + "epoch": 0.59422, + "grad_norm": 1.536802154303846, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 59422 + }, + { + "epoch": 0.59423, + "grad_norm": 1.4878046166843626, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 59423 + }, + { + "epoch": 0.59424, + "grad_norm": 1.6328475478656637, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 59424 + }, + { + "epoch": 0.59425, + "grad_norm": 1.6291004998438876, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 59425 + }, + { + "epoch": 0.59426, + "grad_norm": 1.6142002280547423, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 59426 + }, + { + "epoch": 0.59427, + "grad_norm": 1.3796245031270724, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 59427 + }, + { + "epoch": 0.59428, + "grad_norm": 1.8855567620056142, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 59428 + }, + { + "epoch": 0.59429, + "grad_norm": 1.1824718965411816, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 59429 + }, + { + "epoch": 0.5943, + "grad_norm": 1.8631150244458936, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 59430 + }, + { + "epoch": 0.59431, + "grad_norm": 1.2246120247880707, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 59431 + }, + { + "epoch": 0.59432, + "grad_norm": 1.8553194432545637, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 59432 + }, + { + "epoch": 0.59433, + "grad_norm": 1.2514932917145465, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 59433 + }, + { + "epoch": 0.59434, + "grad_norm": 1.8615388127348955, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 59434 + }, + { + "epoch": 0.59435, + "grad_norm": 1.2977864258745284, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 59435 + }, + { + "epoch": 0.59436, + "grad_norm": 1.9233206131067453, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 59436 + }, + { + "epoch": 0.59437, + "grad_norm": 1.8683955927150553, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 59437 + }, + { + "epoch": 0.59438, + "grad_norm": 1.172734234249606, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 59438 + }, + { + "epoch": 0.59439, + "grad_norm": 1.7621212966932054, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 59439 + }, + { + "epoch": 0.5944, + "grad_norm": 1.4871057767113425, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 59440 + }, + { + "epoch": 0.59441, + "grad_norm": 1.5799177142522556, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 59441 + }, + { + "epoch": 0.59442, + "grad_norm": 1.6623806971269848, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 59442 + }, + { + "epoch": 0.59443, + "grad_norm": 1.4902540068014392, + "learning_rate": 0.003, + "loss": 4.048, + "step": 59443 + }, + { + "epoch": 0.59444, + "grad_norm": 1.581388776508093, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 59444 + }, + { + "epoch": 0.59445, + "grad_norm": 1.7774423557257635, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 59445 + }, + { + "epoch": 0.59446, + "grad_norm": 1.2099643985663777, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 59446 + }, + { + "epoch": 0.59447, + "grad_norm": 1.6939483443745231, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 59447 + }, + { + "epoch": 0.59448, + "grad_norm": 1.748259769402133, + "learning_rate": 0.003, + "loss": 4.037, + "step": 59448 + }, + { + "epoch": 0.59449, + "grad_norm": 1.6563992535189693, + "learning_rate": 0.003, + "loss": 4.036, + "step": 59449 + }, + { + "epoch": 0.5945, + "grad_norm": 1.5738851856478733, + "learning_rate": 0.003, + "loss": 4.058, + "step": 59450 + }, + { + "epoch": 0.59451, + "grad_norm": 1.67871549050545, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 59451 + }, + { + "epoch": 0.59452, + "grad_norm": 1.5957904648757704, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 59452 + }, + { + "epoch": 0.59453, + "grad_norm": 1.2578492395950949, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 59453 + }, + { + "epoch": 0.59454, + "grad_norm": 1.8660826793565974, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 59454 + }, + { + "epoch": 0.59455, + "grad_norm": 1.5481936471084226, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 59455 + }, + { + "epoch": 0.59456, + "grad_norm": 1.5457583552134864, + "learning_rate": 0.003, + "loss": 4.005, + "step": 59456 + }, + { + "epoch": 0.59457, + "grad_norm": 1.4875643421366034, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 59457 + }, + { + "epoch": 0.59458, + "grad_norm": 1.4611190212698606, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 59458 + }, + { + "epoch": 0.59459, + "grad_norm": 1.7851297153813832, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 59459 + }, + { + "epoch": 0.5946, + "grad_norm": 1.6857689593567318, + "learning_rate": 0.003, + "loss": 4.064, + "step": 59460 + }, + { + "epoch": 0.59461, + "grad_norm": 1.518259753970408, + "learning_rate": 0.003, + "loss": 4.05, + "step": 59461 + }, + { + "epoch": 0.59462, + "grad_norm": 1.6280499302589795, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 59462 + }, + { + "epoch": 0.59463, + "grad_norm": 1.5128934756575367, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 59463 + }, + { + "epoch": 0.59464, + "grad_norm": 1.7174223346973434, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 59464 + }, + { + "epoch": 0.59465, + "grad_norm": 1.3258774297504756, + "learning_rate": 0.003, + "loss": 4.041, + "step": 59465 + }, + { + "epoch": 0.59466, + "grad_norm": 1.8811606120343014, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 59466 + }, + { + "epoch": 0.59467, + "grad_norm": 1.6647813625092978, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 59467 + }, + { + "epoch": 0.59468, + "grad_norm": 1.5231861942645237, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 59468 + }, + { + "epoch": 0.59469, + "grad_norm": 1.6068300364053627, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 59469 + }, + { + "epoch": 0.5947, + "grad_norm": 1.742917176178978, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 59470 + }, + { + "epoch": 0.59471, + "grad_norm": 1.3800820888036127, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 59471 + }, + { + "epoch": 0.59472, + "grad_norm": 1.5061311036936507, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 59472 + }, + { + "epoch": 0.59473, + "grad_norm": 1.3123875180507114, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59473 + }, + { + "epoch": 0.59474, + "grad_norm": 1.4422864440830268, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 59474 + }, + { + "epoch": 0.59475, + "grad_norm": 1.6238346702480293, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 59475 + }, + { + "epoch": 0.59476, + "grad_norm": 1.516741265624519, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 59476 + }, + { + "epoch": 0.59477, + "grad_norm": 1.5711998731414347, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 59477 + }, + { + "epoch": 0.59478, + "grad_norm": 1.467329610031062, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 59478 + }, + { + "epoch": 0.59479, + "grad_norm": 1.7769199244493623, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 59479 + }, + { + "epoch": 0.5948, + "grad_norm": 1.2868035707916252, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 59480 + }, + { + "epoch": 0.59481, + "grad_norm": 1.5101580257012954, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 59481 + }, + { + "epoch": 0.59482, + "grad_norm": 1.6508448478590556, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 59482 + }, + { + "epoch": 0.59483, + "grad_norm": 1.5133016788670342, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 59483 + }, + { + "epoch": 0.59484, + "grad_norm": 1.3768682559893866, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 59484 + }, + { + "epoch": 0.59485, + "grad_norm": 1.639366574681348, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 59485 + }, + { + "epoch": 0.59486, + "grad_norm": 1.5021276222488549, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 59486 + }, + { + "epoch": 0.59487, + "grad_norm": 1.3745393762260498, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 59487 + }, + { + "epoch": 0.59488, + "grad_norm": 1.61455098956669, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 59488 + }, + { + "epoch": 0.59489, + "grad_norm": 1.485102307289063, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 59489 + }, + { + "epoch": 0.5949, + "grad_norm": 1.6736729188668176, + "learning_rate": 0.003, + "loss": 4.047, + "step": 59490 + }, + { + "epoch": 0.59491, + "grad_norm": 1.3920382474249977, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 59491 + }, + { + "epoch": 0.59492, + "grad_norm": 1.737070095053538, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 59492 + }, + { + "epoch": 0.59493, + "grad_norm": 1.5760481622589755, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 59493 + }, + { + "epoch": 0.59494, + "grad_norm": 2.0256966121950897, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 59494 + }, + { + "epoch": 0.59495, + "grad_norm": 1.73151418611671, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 59495 + }, + { + "epoch": 0.59496, + "grad_norm": 1.351157051036419, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 59496 + }, + { + "epoch": 0.59497, + "grad_norm": 2.0304368114284173, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 59497 + }, + { + "epoch": 0.59498, + "grad_norm": 1.5859627324999594, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 59498 + }, + { + "epoch": 0.59499, + "grad_norm": 1.4708260243333915, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59499 + }, + { + "epoch": 0.595, + "grad_norm": 1.672328569304013, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 59500 + }, + { + "epoch": 0.59501, + "grad_norm": 1.4576088901044808, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 59501 + }, + { + "epoch": 0.59502, + "grad_norm": 1.5839353858420229, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 59502 + }, + { + "epoch": 0.59503, + "grad_norm": 1.7900867502836846, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 59503 + }, + { + "epoch": 0.59504, + "grad_norm": 1.5453941390711394, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 59504 + }, + { + "epoch": 0.59505, + "grad_norm": 1.62829965047171, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 59505 + }, + { + "epoch": 0.59506, + "grad_norm": 1.6528595784539746, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 59506 + }, + { + "epoch": 0.59507, + "grad_norm": 1.310203298902498, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 59507 + }, + { + "epoch": 0.59508, + "grad_norm": 1.6750011028022558, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 59508 + }, + { + "epoch": 0.59509, + "grad_norm": 1.4753039210262793, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 59509 + }, + { + "epoch": 0.5951, + "grad_norm": 1.5522005196271786, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 59510 + }, + { + "epoch": 0.59511, + "grad_norm": 1.7804820253640385, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 59511 + }, + { + "epoch": 0.59512, + "grad_norm": 1.371512842940546, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 59512 + }, + { + "epoch": 0.59513, + "grad_norm": 1.5752919923920865, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59513 + }, + { + "epoch": 0.59514, + "grad_norm": 1.3499097682387806, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 59514 + }, + { + "epoch": 0.59515, + "grad_norm": 1.5296313851718424, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 59515 + }, + { + "epoch": 0.59516, + "grad_norm": 1.4012278793331439, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 59516 + }, + { + "epoch": 0.59517, + "grad_norm": 1.4662193739394926, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 59517 + }, + { + "epoch": 0.59518, + "grad_norm": 1.7816454617289568, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 59518 + }, + { + "epoch": 0.59519, + "grad_norm": 1.7201536652898766, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 59519 + }, + { + "epoch": 0.5952, + "grad_norm": 1.686759319072645, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59520 + }, + { + "epoch": 0.59521, + "grad_norm": 1.571953931805279, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 59521 + }, + { + "epoch": 0.59522, + "grad_norm": 1.4848542549987334, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 59522 + }, + { + "epoch": 0.59523, + "grad_norm": 1.5674862211640432, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 59523 + }, + { + "epoch": 0.59524, + "grad_norm": 1.7524205917703826, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 59524 + }, + { + "epoch": 0.59525, + "grad_norm": 1.5008422490419573, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 59525 + }, + { + "epoch": 0.59526, + "grad_norm": 1.4030359294760755, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 59526 + }, + { + "epoch": 0.59527, + "grad_norm": 1.695388451378732, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 59527 + }, + { + "epoch": 0.59528, + "grad_norm": 1.424411555829217, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 59528 + }, + { + "epoch": 0.59529, + "grad_norm": 1.754723184368798, + "learning_rate": 0.003, + "loss": 4.077, + "step": 59529 + }, + { + "epoch": 0.5953, + "grad_norm": 1.4505702163364251, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 59530 + }, + { + "epoch": 0.59531, + "grad_norm": 1.884954795542271, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 59531 + }, + { + "epoch": 0.59532, + "grad_norm": 1.5171785624411078, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 59532 + }, + { + "epoch": 0.59533, + "grad_norm": 1.542099298172011, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 59533 + }, + { + "epoch": 0.59534, + "grad_norm": 1.5488732127291547, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 59534 + }, + { + "epoch": 0.59535, + "grad_norm": 1.7146311884250032, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 59535 + }, + { + "epoch": 0.59536, + "grad_norm": 1.7500062984253915, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 59536 + }, + { + "epoch": 0.59537, + "grad_norm": 1.3469041284534244, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 59537 + }, + { + "epoch": 0.59538, + "grad_norm": 1.475832901221686, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 59538 + }, + { + "epoch": 0.59539, + "grad_norm": 1.549222803850725, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 59539 + }, + { + "epoch": 0.5954, + "grad_norm": 1.5416237428714399, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 59540 + }, + { + "epoch": 0.59541, + "grad_norm": 1.764827392862096, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 59541 + }, + { + "epoch": 0.59542, + "grad_norm": 1.3100061116497927, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 59542 + }, + { + "epoch": 0.59543, + "grad_norm": 1.7981929393116267, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 59543 + }, + { + "epoch": 0.59544, + "grad_norm": 1.3543351815840958, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 59544 + }, + { + "epoch": 0.59545, + "grad_norm": 1.5375745406136987, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 59545 + }, + { + "epoch": 0.59546, + "grad_norm": 1.379129286470833, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 59546 + }, + { + "epoch": 0.59547, + "grad_norm": 1.6802244145631213, + "learning_rate": 0.003, + "loss": 4.007, + "step": 59547 + }, + { + "epoch": 0.59548, + "grad_norm": 1.287664348313447, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 59548 + }, + { + "epoch": 0.59549, + "grad_norm": 1.8520885858908236, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 59549 + }, + { + "epoch": 0.5955, + "grad_norm": 1.5890707790408014, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 59550 + }, + { + "epoch": 0.59551, + "grad_norm": 1.6894472268177283, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 59551 + }, + { + "epoch": 0.59552, + "grad_norm": 1.7200295556834544, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59552 + }, + { + "epoch": 0.59553, + "grad_norm": 1.4993521955629199, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 59553 + }, + { + "epoch": 0.59554, + "grad_norm": 1.4096390073183154, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 59554 + }, + { + "epoch": 0.59555, + "grad_norm": 1.7393085962504937, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 59555 + }, + { + "epoch": 0.59556, + "grad_norm": 1.9355936380279926, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 59556 + }, + { + "epoch": 0.59557, + "grad_norm": 1.3391123361186894, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 59557 + }, + { + "epoch": 0.59558, + "grad_norm": 1.6574551549443208, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 59558 + }, + { + "epoch": 0.59559, + "grad_norm": 1.3966001611268772, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 59559 + }, + { + "epoch": 0.5956, + "grad_norm": 1.7381126747700066, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59560 + }, + { + "epoch": 0.59561, + "grad_norm": 1.5125065837181186, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 59561 + }, + { + "epoch": 0.59562, + "grad_norm": 1.7333452901960182, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 59562 + }, + { + "epoch": 0.59563, + "grad_norm": 1.1896511317417693, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 59563 + }, + { + "epoch": 0.59564, + "grad_norm": 1.5412325860867093, + "learning_rate": 0.003, + "loss": 4.048, + "step": 59564 + }, + { + "epoch": 0.59565, + "grad_norm": 1.6654625678133474, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 59565 + }, + { + "epoch": 0.59566, + "grad_norm": 1.2552747629678658, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 59566 + }, + { + "epoch": 0.59567, + "grad_norm": 1.695275567639224, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 59567 + }, + { + "epoch": 0.59568, + "grad_norm": 1.2050207951017933, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 59568 + }, + { + "epoch": 0.59569, + "grad_norm": 1.7889523595856516, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 59569 + }, + { + "epoch": 0.5957, + "grad_norm": 1.273718275690038, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 59570 + }, + { + "epoch": 0.59571, + "grad_norm": 1.8755006816443898, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 59571 + }, + { + "epoch": 0.59572, + "grad_norm": 2.1291203507686878, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 59572 + }, + { + "epoch": 0.59573, + "grad_norm": 1.364050596046236, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 59573 + }, + { + "epoch": 0.59574, + "grad_norm": 1.84318658958962, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 59574 + }, + { + "epoch": 0.59575, + "grad_norm": 1.9660286018975912, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 59575 + }, + { + "epoch": 0.59576, + "grad_norm": 1.2101030397847725, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 59576 + }, + { + "epoch": 0.59577, + "grad_norm": 1.8384557032866364, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 59577 + }, + { + "epoch": 0.59578, + "grad_norm": 1.3253379967394798, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 59578 + }, + { + "epoch": 0.59579, + "grad_norm": 1.6058591571686769, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59579 + }, + { + "epoch": 0.5958, + "grad_norm": 1.6247782825906811, + "learning_rate": 0.003, + "loss": 4.053, + "step": 59580 + }, + { + "epoch": 0.59581, + "grad_norm": 1.6437243804361528, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 59581 + }, + { + "epoch": 0.59582, + "grad_norm": 1.496152316302895, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 59582 + }, + { + "epoch": 0.59583, + "grad_norm": 1.7887220102249322, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 59583 + }, + { + "epoch": 0.59584, + "grad_norm": 1.356085611868516, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 59584 + }, + { + "epoch": 0.59585, + "grad_norm": 1.6485963512738684, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 59585 + }, + { + "epoch": 0.59586, + "grad_norm": 1.5308642210806311, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 59586 + }, + { + "epoch": 0.59587, + "grad_norm": 1.479386403106623, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 59587 + }, + { + "epoch": 0.59588, + "grad_norm": 1.6352562783108675, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 59588 + }, + { + "epoch": 0.59589, + "grad_norm": 1.5410614715183761, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 59589 + }, + { + "epoch": 0.5959, + "grad_norm": 1.6459905922093452, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 59590 + }, + { + "epoch": 0.59591, + "grad_norm": 1.3890491557082156, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 59591 + }, + { + "epoch": 0.59592, + "grad_norm": 1.6415370111032432, + "learning_rate": 0.003, + "loss": 4.048, + "step": 59592 + }, + { + "epoch": 0.59593, + "grad_norm": 1.3645800472968908, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 59593 + }, + { + "epoch": 0.59594, + "grad_norm": 1.8452108195295598, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 59594 + }, + { + "epoch": 0.59595, + "grad_norm": 1.2610586359950098, + "learning_rate": 0.003, + "loss": 4.053, + "step": 59595 + }, + { + "epoch": 0.59596, + "grad_norm": 1.6073321009361834, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 59596 + }, + { + "epoch": 0.59597, + "grad_norm": 1.2423808076818872, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 59597 + }, + { + "epoch": 0.59598, + "grad_norm": 1.607178085462123, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 59598 + }, + { + "epoch": 0.59599, + "grad_norm": 1.488891766077755, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 59599 + }, + { + "epoch": 0.596, + "grad_norm": 1.9964217581089923, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 59600 + }, + { + "epoch": 0.59601, + "grad_norm": 1.7343237160544984, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 59601 + }, + { + "epoch": 0.59602, + "grad_norm": 1.5959934984779713, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 59602 + }, + { + "epoch": 0.59603, + "grad_norm": 1.771922283949591, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 59603 + }, + { + "epoch": 0.59604, + "grad_norm": 1.24858647451435, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 59604 + }, + { + "epoch": 0.59605, + "grad_norm": 1.816995192091677, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 59605 + }, + { + "epoch": 0.59606, + "grad_norm": 1.4904248572160221, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 59606 + }, + { + "epoch": 0.59607, + "grad_norm": 1.6880543512591113, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 59607 + }, + { + "epoch": 0.59608, + "grad_norm": 1.6271275252078758, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 59608 + }, + { + "epoch": 0.59609, + "grad_norm": 1.4369258311031943, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 59609 + }, + { + "epoch": 0.5961, + "grad_norm": 1.5692903795623587, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 59610 + }, + { + "epoch": 0.59611, + "grad_norm": 1.1718561661594875, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 59611 + }, + { + "epoch": 0.59612, + "grad_norm": 1.5938086767458681, + "learning_rate": 0.003, + "loss": 4.012, + "step": 59612 + }, + { + "epoch": 0.59613, + "grad_norm": 1.3113581643689245, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 59613 + }, + { + "epoch": 0.59614, + "grad_norm": 1.7949132980678526, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 59614 + }, + { + "epoch": 0.59615, + "grad_norm": 1.805567803923286, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 59615 + }, + { + "epoch": 0.59616, + "grad_norm": 1.4485770084014458, + "learning_rate": 0.003, + "loss": 4.011, + "step": 59616 + }, + { + "epoch": 0.59617, + "grad_norm": 2.0352760014744478, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 59617 + }, + { + "epoch": 0.59618, + "grad_norm": 1.3468845951337305, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 59618 + }, + { + "epoch": 0.59619, + "grad_norm": 1.84303653048175, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 59619 + }, + { + "epoch": 0.5962, + "grad_norm": 1.3481819224546392, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 59620 + }, + { + "epoch": 0.59621, + "grad_norm": 1.6904756744628162, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 59621 + }, + { + "epoch": 0.59622, + "grad_norm": 1.4573523150127488, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 59622 + }, + { + "epoch": 0.59623, + "grad_norm": 1.4739123837122046, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 59623 + }, + { + "epoch": 0.59624, + "grad_norm": 1.7957621881334367, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 59624 + }, + { + "epoch": 0.59625, + "grad_norm": 1.418061350031498, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 59625 + }, + { + "epoch": 0.59626, + "grad_norm": 1.7410706311161919, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 59626 + }, + { + "epoch": 0.59627, + "grad_norm": 1.4808558445961426, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 59627 + }, + { + "epoch": 0.59628, + "grad_norm": 1.669540192723345, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 59628 + }, + { + "epoch": 0.59629, + "grad_norm": 1.5223899798706475, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59629 + }, + { + "epoch": 0.5963, + "grad_norm": 1.3889442781563313, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 59630 + }, + { + "epoch": 0.59631, + "grad_norm": 1.8158881619694582, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 59631 + }, + { + "epoch": 0.59632, + "grad_norm": 1.3435694858107334, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 59632 + }, + { + "epoch": 0.59633, + "grad_norm": 1.8511826864027747, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 59633 + }, + { + "epoch": 0.59634, + "grad_norm": 1.3021060602569232, + "learning_rate": 0.003, + "loss": 4.005, + "step": 59634 + }, + { + "epoch": 0.59635, + "grad_norm": 1.7211439599211802, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 59635 + }, + { + "epoch": 0.59636, + "grad_norm": 1.3174204954305835, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 59636 + }, + { + "epoch": 0.59637, + "grad_norm": 1.7420138656266357, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 59637 + }, + { + "epoch": 0.59638, + "grad_norm": 1.480481804432416, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 59638 + }, + { + "epoch": 0.59639, + "grad_norm": 1.7468897509276382, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 59639 + }, + { + "epoch": 0.5964, + "grad_norm": 1.6839198684104142, + "learning_rate": 0.003, + "loss": 4.032, + "step": 59640 + }, + { + "epoch": 0.59641, + "grad_norm": 1.5524385359114066, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 59641 + }, + { + "epoch": 0.59642, + "grad_norm": 1.593680305911817, + "learning_rate": 0.003, + "loss": 4.04, + "step": 59642 + }, + { + "epoch": 0.59643, + "grad_norm": 1.562278221500459, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 59643 + }, + { + "epoch": 0.59644, + "grad_norm": 1.3920481463913386, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 59644 + }, + { + "epoch": 0.59645, + "grad_norm": 1.474020852819873, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 59645 + }, + { + "epoch": 0.59646, + "grad_norm": 1.652640234673486, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 59646 + }, + { + "epoch": 0.59647, + "grad_norm": 1.7258929552886582, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 59647 + }, + { + "epoch": 0.59648, + "grad_norm": 1.342953907928157, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 59648 + }, + { + "epoch": 0.59649, + "grad_norm": 1.6931701864980158, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 59649 + }, + { + "epoch": 0.5965, + "grad_norm": 1.4877095782557967, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 59650 + }, + { + "epoch": 0.59651, + "grad_norm": 1.6124137695224816, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 59651 + }, + { + "epoch": 0.59652, + "grad_norm": 1.6046616946400947, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 59652 + }, + { + "epoch": 0.59653, + "grad_norm": 1.4417169411451962, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 59653 + }, + { + "epoch": 0.59654, + "grad_norm": 1.7496250533657018, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 59654 + }, + { + "epoch": 0.59655, + "grad_norm": 1.4351753419868538, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 59655 + }, + { + "epoch": 0.59656, + "grad_norm": 1.5977734351311041, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 59656 + }, + { + "epoch": 0.59657, + "grad_norm": 1.652091595796089, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 59657 + }, + { + "epoch": 0.59658, + "grad_norm": 1.6434239487486288, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 59658 + }, + { + "epoch": 0.59659, + "grad_norm": 1.228592976152935, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 59659 + }, + { + "epoch": 0.5966, + "grad_norm": 1.7800957248758953, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 59660 + }, + { + "epoch": 0.59661, + "grad_norm": 1.6816297698081226, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 59661 + }, + { + "epoch": 0.59662, + "grad_norm": 1.5319464022212785, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 59662 + }, + { + "epoch": 0.59663, + "grad_norm": 1.4407881395337088, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 59663 + }, + { + "epoch": 0.59664, + "grad_norm": 1.6953443364429515, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 59664 + }, + { + "epoch": 0.59665, + "grad_norm": 1.6072097114751447, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 59665 + }, + { + "epoch": 0.59666, + "grad_norm": 1.6695430979306947, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 59666 + }, + { + "epoch": 0.59667, + "grad_norm": 1.38826822474217, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 59667 + }, + { + "epoch": 0.59668, + "grad_norm": 1.6508337218833307, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 59668 + }, + { + "epoch": 0.59669, + "grad_norm": 1.3759656141446235, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 59669 + }, + { + "epoch": 0.5967, + "grad_norm": 1.3905983873082046, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 59670 + }, + { + "epoch": 0.59671, + "grad_norm": 1.6568927055785394, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 59671 + }, + { + "epoch": 0.59672, + "grad_norm": 1.528317136097671, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 59672 + }, + { + "epoch": 0.59673, + "grad_norm": 1.7043474038966648, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 59673 + }, + { + "epoch": 0.59674, + "grad_norm": 1.460443481173466, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 59674 + }, + { + "epoch": 0.59675, + "grad_norm": 1.7690460557963696, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 59675 + }, + { + "epoch": 0.59676, + "grad_norm": 1.2099596307420002, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 59676 + }, + { + "epoch": 0.59677, + "grad_norm": 1.8434200028343033, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 59677 + }, + { + "epoch": 0.59678, + "grad_norm": 1.3406160219046614, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59678 + }, + { + "epoch": 0.59679, + "grad_norm": 1.7861808236354793, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59679 + }, + { + "epoch": 0.5968, + "grad_norm": 1.6923636842535108, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 59680 + }, + { + "epoch": 0.59681, + "grad_norm": 1.4916753085544752, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 59681 + }, + { + "epoch": 0.59682, + "grad_norm": 1.9469427094571798, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59682 + }, + { + "epoch": 0.59683, + "grad_norm": 1.5156978587431105, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 59683 + }, + { + "epoch": 0.59684, + "grad_norm": 1.5534213409705404, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 59684 + }, + { + "epoch": 0.59685, + "grad_norm": 1.6882903513693535, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 59685 + }, + { + "epoch": 0.59686, + "grad_norm": 1.5150263243184052, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 59686 + }, + { + "epoch": 0.59687, + "grad_norm": 1.7271670895357147, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 59687 + }, + { + "epoch": 0.59688, + "grad_norm": 1.4429500182443953, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 59688 + }, + { + "epoch": 0.59689, + "grad_norm": 1.8318760084568013, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 59689 + }, + { + "epoch": 0.5969, + "grad_norm": 1.2265688126400986, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 59690 + }, + { + "epoch": 0.59691, + "grad_norm": 1.7399359426476813, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 59691 + }, + { + "epoch": 0.59692, + "grad_norm": 1.7696069772084357, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 59692 + }, + { + "epoch": 0.59693, + "grad_norm": 1.3811776495303352, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 59693 + }, + { + "epoch": 0.59694, + "grad_norm": 1.6229271029683459, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 59694 + }, + { + "epoch": 0.59695, + "grad_norm": 1.3994871885247344, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 59695 + }, + { + "epoch": 0.59696, + "grad_norm": 1.9224217073949244, + "learning_rate": 0.003, + "loss": 4.041, + "step": 59696 + }, + { + "epoch": 0.59697, + "grad_norm": 1.349966727836609, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 59697 + }, + { + "epoch": 0.59698, + "grad_norm": 1.5603883553588074, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 59698 + }, + { + "epoch": 0.59699, + "grad_norm": 1.4458299438590516, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 59699 + }, + { + "epoch": 0.597, + "grad_norm": 1.5882047837891877, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 59700 + }, + { + "epoch": 0.59701, + "grad_norm": 1.569607904853881, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 59701 + }, + { + "epoch": 0.59702, + "grad_norm": 1.8695672593564792, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 59702 + }, + { + "epoch": 0.59703, + "grad_norm": 1.7522802245583142, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59703 + }, + { + "epoch": 0.59704, + "grad_norm": 1.4404163909615046, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 59704 + }, + { + "epoch": 0.59705, + "grad_norm": 1.6822981150858418, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 59705 + }, + { + "epoch": 0.59706, + "grad_norm": 1.53532403763983, + "learning_rate": 0.003, + "loss": 4.047, + "step": 59706 + }, + { + "epoch": 0.59707, + "grad_norm": 1.3978219641197547, + "learning_rate": 0.003, + "loss": 4.016, + "step": 59707 + }, + { + "epoch": 0.59708, + "grad_norm": 1.7272076826771865, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 59708 + }, + { + "epoch": 0.59709, + "grad_norm": 1.2910426210039916, + "learning_rate": 0.003, + "loss": 4.025, + "step": 59709 + }, + { + "epoch": 0.5971, + "grad_norm": 1.9223410680619757, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 59710 + }, + { + "epoch": 0.59711, + "grad_norm": 1.3472134731632572, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 59711 + }, + { + "epoch": 0.59712, + "grad_norm": 1.7171980498013715, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 59712 + }, + { + "epoch": 0.59713, + "grad_norm": 1.67910713826353, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 59713 + }, + { + "epoch": 0.59714, + "grad_norm": 1.8167126692881321, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 59714 + }, + { + "epoch": 0.59715, + "grad_norm": 1.5089744018246216, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 59715 + }, + { + "epoch": 0.59716, + "grad_norm": 1.5709689899352748, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 59716 + }, + { + "epoch": 0.59717, + "grad_norm": 1.8134590697159985, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 59717 + }, + { + "epoch": 0.59718, + "grad_norm": 1.3391446058285876, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 59718 + }, + { + "epoch": 0.59719, + "grad_norm": 1.5479954746653, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 59719 + }, + { + "epoch": 0.5972, + "grad_norm": 1.4532816644146183, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 59720 + }, + { + "epoch": 0.59721, + "grad_norm": 1.6032511220094297, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 59721 + }, + { + "epoch": 0.59722, + "grad_norm": 1.617091983814945, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 59722 + }, + { + "epoch": 0.59723, + "grad_norm": 1.5294783220097354, + "learning_rate": 0.003, + "loss": 4.05, + "step": 59723 + }, + { + "epoch": 0.59724, + "grad_norm": 1.6215834409233065, + "learning_rate": 0.003, + "loss": 4.042, + "step": 59724 + }, + { + "epoch": 0.59725, + "grad_norm": 1.32666408465799, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 59725 + }, + { + "epoch": 0.59726, + "grad_norm": 1.652450977869261, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 59726 + }, + { + "epoch": 0.59727, + "grad_norm": 1.5500100124616691, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 59727 + }, + { + "epoch": 0.59728, + "grad_norm": 1.6278585360456639, + "learning_rate": 0.003, + "loss": 4.011, + "step": 59728 + }, + { + "epoch": 0.59729, + "grad_norm": 1.507555102477027, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 59729 + }, + { + "epoch": 0.5973, + "grad_norm": 1.4746880845476504, + "learning_rate": 0.003, + "loss": 4.022, + "step": 59730 + }, + { + "epoch": 0.59731, + "grad_norm": 2.081179650608291, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 59731 + }, + { + "epoch": 0.59732, + "grad_norm": 1.6870881571930294, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 59732 + }, + { + "epoch": 0.59733, + "grad_norm": 1.3318805845607276, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 59733 + }, + { + "epoch": 0.59734, + "grad_norm": 1.4057616727544928, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 59734 + }, + { + "epoch": 0.59735, + "grad_norm": 1.6226117733905847, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 59735 + }, + { + "epoch": 0.59736, + "grad_norm": 1.6852514278120225, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 59736 + }, + { + "epoch": 0.59737, + "grad_norm": 1.721960016224673, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 59737 + }, + { + "epoch": 0.59738, + "grad_norm": 1.4438721712530942, + "learning_rate": 0.003, + "loss": 4.006, + "step": 59738 + }, + { + "epoch": 0.59739, + "grad_norm": 1.6402120018330348, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 59739 + }, + { + "epoch": 0.5974, + "grad_norm": 1.4942783873784418, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 59740 + }, + { + "epoch": 0.59741, + "grad_norm": 1.4868664866530965, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 59741 + }, + { + "epoch": 0.59742, + "grad_norm": 1.647204604068297, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 59742 + }, + { + "epoch": 0.59743, + "grad_norm": 1.4505022274906543, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 59743 + }, + { + "epoch": 0.59744, + "grad_norm": 1.709107140179008, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 59744 + }, + { + "epoch": 0.59745, + "grad_norm": 1.2009312520596997, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 59745 + }, + { + "epoch": 0.59746, + "grad_norm": 1.7580251210174693, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 59746 + }, + { + "epoch": 0.59747, + "grad_norm": 1.2965490938740372, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 59747 + }, + { + "epoch": 0.59748, + "grad_norm": 1.7585568713447544, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 59748 + }, + { + "epoch": 0.59749, + "grad_norm": 1.3569641361022133, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 59749 + }, + { + "epoch": 0.5975, + "grad_norm": 1.5099781247838753, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 59750 + }, + { + "epoch": 0.59751, + "grad_norm": 1.4214303999859477, + "learning_rate": 0.003, + "loss": 4.037, + "step": 59751 + }, + { + "epoch": 0.59752, + "grad_norm": 1.7007720030897273, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 59752 + }, + { + "epoch": 0.59753, + "grad_norm": 1.7517795007560095, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 59753 + }, + { + "epoch": 0.59754, + "grad_norm": 1.8981418605844116, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 59754 + }, + { + "epoch": 0.59755, + "grad_norm": 1.4731248903807999, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 59755 + }, + { + "epoch": 0.59756, + "grad_norm": 1.679764116785176, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 59756 + }, + { + "epoch": 0.59757, + "grad_norm": 1.6410585042097143, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 59757 + }, + { + "epoch": 0.59758, + "grad_norm": 1.3761564579487426, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 59758 + }, + { + "epoch": 0.59759, + "grad_norm": 1.5930117824652452, + "learning_rate": 0.003, + "loss": 4.023, + "step": 59759 + }, + { + "epoch": 0.5976, + "grad_norm": 1.4237240402899156, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 59760 + }, + { + "epoch": 0.59761, + "grad_norm": 1.4712840322449563, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 59761 + }, + { + "epoch": 0.59762, + "grad_norm": 1.461453927656004, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 59762 + }, + { + "epoch": 0.59763, + "grad_norm": 1.8894197389158107, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 59763 + }, + { + "epoch": 0.59764, + "grad_norm": 1.4976132617105287, + "learning_rate": 0.003, + "loss": 4.034, + "step": 59764 + }, + { + "epoch": 0.59765, + "grad_norm": 1.7188957026816394, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 59765 + }, + { + "epoch": 0.59766, + "grad_norm": 1.4627194694306995, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 59766 + }, + { + "epoch": 0.59767, + "grad_norm": 1.547586112374024, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 59767 + }, + { + "epoch": 0.59768, + "grad_norm": 1.391651756479671, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 59768 + }, + { + "epoch": 0.59769, + "grad_norm": 1.2862613978595459, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 59769 + }, + { + "epoch": 0.5977, + "grad_norm": 1.418027809206172, + "learning_rate": 0.003, + "loss": 4.018, + "step": 59770 + }, + { + "epoch": 0.59771, + "grad_norm": 1.7161589762607905, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 59771 + }, + { + "epoch": 0.59772, + "grad_norm": 1.7618546297152085, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 59772 + }, + { + "epoch": 0.59773, + "grad_norm": 1.8783805901130046, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 59773 + }, + { + "epoch": 0.59774, + "grad_norm": 1.573628660691904, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 59774 + }, + { + "epoch": 0.59775, + "grad_norm": 1.7940503350363024, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 59775 + }, + { + "epoch": 0.59776, + "grad_norm": 1.2905394850539056, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 59776 + }, + { + "epoch": 0.59777, + "grad_norm": 1.6696153388711186, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 59777 + }, + { + "epoch": 0.59778, + "grad_norm": 1.3913690277937063, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 59778 + }, + { + "epoch": 0.59779, + "grad_norm": 1.6607648038785032, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 59779 + }, + { + "epoch": 0.5978, + "grad_norm": 1.4283091132013077, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 59780 + }, + { + "epoch": 0.59781, + "grad_norm": 1.5944871178942412, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 59781 + }, + { + "epoch": 0.59782, + "grad_norm": 1.5968915030632154, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 59782 + }, + { + "epoch": 0.59783, + "grad_norm": 1.4545265161916665, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 59783 + }, + { + "epoch": 0.59784, + "grad_norm": 1.7849295085683794, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 59784 + }, + { + "epoch": 0.59785, + "grad_norm": 1.4551736909899549, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 59785 + }, + { + "epoch": 0.59786, + "grad_norm": 1.6432773704606425, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 59786 + }, + { + "epoch": 0.59787, + "grad_norm": 1.440337853519741, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 59787 + }, + { + "epoch": 0.59788, + "grad_norm": 1.532848077621244, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 59788 + }, + { + "epoch": 0.59789, + "grad_norm": 1.7329420021699102, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 59789 + }, + { + "epoch": 0.5979, + "grad_norm": 1.5621330563613431, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 59790 + }, + { + "epoch": 0.59791, + "grad_norm": 1.8555287908312232, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 59791 + }, + { + "epoch": 0.59792, + "grad_norm": 1.5622678271799473, + "learning_rate": 0.003, + "loss": 4.008, + "step": 59792 + }, + { + "epoch": 0.59793, + "grad_norm": 1.8882274508216632, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 59793 + }, + { + "epoch": 0.59794, + "grad_norm": 1.8157421156643028, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 59794 + }, + { + "epoch": 0.59795, + "grad_norm": 1.2587878555480767, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 59795 + }, + { + "epoch": 0.59796, + "grad_norm": 1.8596193224007351, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 59796 + }, + { + "epoch": 0.59797, + "grad_norm": 1.737694261284724, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 59797 + }, + { + "epoch": 0.59798, + "grad_norm": 1.4773400536036625, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 59798 + }, + { + "epoch": 0.59799, + "grad_norm": 1.5601301387590005, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 59799 + }, + { + "epoch": 0.598, + "grad_norm": 1.4793734438881116, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 59800 + }, + { + "epoch": 0.59801, + "grad_norm": 1.8424019602276207, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 59801 + }, + { + "epoch": 0.59802, + "grad_norm": 1.2884718519949192, + "learning_rate": 0.003, + "loss": 4.02, + "step": 59802 + }, + { + "epoch": 0.59803, + "grad_norm": 1.8287686923967044, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 59803 + }, + { + "epoch": 0.59804, + "grad_norm": 1.332973756707441, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 59804 + }, + { + "epoch": 0.59805, + "grad_norm": 1.701147658853194, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 59805 + }, + { + "epoch": 0.59806, + "grad_norm": 1.5557741908682758, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 59806 + }, + { + "epoch": 0.59807, + "grad_norm": 1.7997164874838016, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 59807 + }, + { + "epoch": 0.59808, + "grad_norm": 1.4116366797107596, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 59808 + }, + { + "epoch": 0.59809, + "grad_norm": 1.4997739328025912, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 59809 + }, + { + "epoch": 0.5981, + "grad_norm": 1.398421044302444, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 59810 + }, + { + "epoch": 0.59811, + "grad_norm": 1.4435015894026244, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 59811 + }, + { + "epoch": 0.59812, + "grad_norm": 1.6793423434260448, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 59812 + }, + { + "epoch": 0.59813, + "grad_norm": 1.753154748226516, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 59813 + }, + { + "epoch": 0.59814, + "grad_norm": 2.0587261960043324, + "learning_rate": 0.003, + "loss": 4.036, + "step": 59814 + }, + { + "epoch": 0.59815, + "grad_norm": 1.5200237342612781, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 59815 + }, + { + "epoch": 0.59816, + "grad_norm": 1.385963022921016, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 59816 + }, + { + "epoch": 0.59817, + "grad_norm": 1.820414892699136, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 59817 + }, + { + "epoch": 0.59818, + "grad_norm": 1.509448035023985, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 59818 + }, + { + "epoch": 0.59819, + "grad_norm": 1.8516621265441175, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 59819 + }, + { + "epoch": 0.5982, + "grad_norm": 1.4383963149696757, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 59820 + }, + { + "epoch": 0.59821, + "grad_norm": 1.7450824255594604, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 59821 + }, + { + "epoch": 0.59822, + "grad_norm": 1.5993016245355791, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 59822 + }, + { + "epoch": 0.59823, + "grad_norm": 1.5974931331540698, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 59823 + }, + { + "epoch": 0.59824, + "grad_norm": 1.4825139176986748, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 59824 + }, + { + "epoch": 0.59825, + "grad_norm": 1.6215103701568052, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 59825 + }, + { + "epoch": 0.59826, + "grad_norm": 1.861202095575566, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 59826 + }, + { + "epoch": 0.59827, + "grad_norm": 1.5354714621657346, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 59827 + }, + { + "epoch": 0.59828, + "grad_norm": 1.5776357449492664, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 59828 + }, + { + "epoch": 0.59829, + "grad_norm": 1.4614734957156932, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 59829 + }, + { + "epoch": 0.5983, + "grad_norm": 1.3247185337276406, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 59830 + }, + { + "epoch": 0.59831, + "grad_norm": 1.5959603966420752, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 59831 + }, + { + "epoch": 0.59832, + "grad_norm": 1.4736904883985231, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 59832 + }, + { + "epoch": 0.59833, + "grad_norm": 1.6151122332019534, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 59833 + }, + { + "epoch": 0.59834, + "grad_norm": 1.484511464309268, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 59834 + }, + { + "epoch": 0.59835, + "grad_norm": 1.3587439397330523, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 59835 + }, + { + "epoch": 0.59836, + "grad_norm": 1.8400136804291611, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 59836 + }, + { + "epoch": 0.59837, + "grad_norm": 1.4542189123492735, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 59837 + }, + { + "epoch": 0.59838, + "grad_norm": 1.8564239167195296, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 59838 + }, + { + "epoch": 0.59839, + "grad_norm": 1.479206470138555, + "learning_rate": 0.003, + "loss": 4.048, + "step": 59839 + }, + { + "epoch": 0.5984, + "grad_norm": 1.7508898136695106, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 59840 + }, + { + "epoch": 0.59841, + "grad_norm": 1.914594995140847, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 59841 + }, + { + "epoch": 0.59842, + "grad_norm": 1.1282089935780628, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 59842 + }, + { + "epoch": 0.59843, + "grad_norm": 2.1541912782456523, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 59843 + }, + { + "epoch": 0.59844, + "grad_norm": 1.3051018162974364, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 59844 + }, + { + "epoch": 0.59845, + "grad_norm": 2.122118568047144, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 59845 + }, + { + "epoch": 0.59846, + "grad_norm": 1.3578741706568909, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 59846 + }, + { + "epoch": 0.59847, + "grad_norm": 1.4972223230145985, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 59847 + }, + { + "epoch": 0.59848, + "grad_norm": 1.728792120480365, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 59848 + }, + { + "epoch": 0.59849, + "grad_norm": 1.7194563944066084, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 59849 + }, + { + "epoch": 0.5985, + "grad_norm": 1.2531832761569857, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 59850 + }, + { + "epoch": 0.59851, + "grad_norm": 1.9123053975622597, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 59851 + }, + { + "epoch": 0.59852, + "grad_norm": 1.386798821424351, + "learning_rate": 0.003, + "loss": 4.049, + "step": 59852 + }, + { + "epoch": 0.59853, + "grad_norm": 1.5932943307850242, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 59853 + }, + { + "epoch": 0.59854, + "grad_norm": 1.538495705170951, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 59854 + }, + { + "epoch": 0.59855, + "grad_norm": 1.7873296091615876, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 59855 + }, + { + "epoch": 0.59856, + "grad_norm": 1.5241750367440703, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 59856 + }, + { + "epoch": 0.59857, + "grad_norm": 1.4704701487408736, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 59857 + }, + { + "epoch": 0.59858, + "grad_norm": 1.6291915445225005, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 59858 + }, + { + "epoch": 0.59859, + "grad_norm": 1.2831271036948786, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 59859 + }, + { + "epoch": 0.5986, + "grad_norm": 1.389923219062249, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 59860 + }, + { + "epoch": 0.59861, + "grad_norm": 1.5653963897282672, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 59861 + }, + { + "epoch": 0.59862, + "grad_norm": 1.5961164877267662, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 59862 + }, + { + "epoch": 0.59863, + "grad_norm": 1.3659440549509472, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 59863 + }, + { + "epoch": 0.59864, + "grad_norm": 1.6673120632091925, + "learning_rate": 0.003, + "loss": 4.041, + "step": 59864 + }, + { + "epoch": 0.59865, + "grad_norm": 1.3787844975512702, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 59865 + }, + { + "epoch": 0.59866, + "grad_norm": 1.5925185780385027, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 59866 + }, + { + "epoch": 0.59867, + "grad_norm": 1.6626671624399354, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 59867 + }, + { + "epoch": 0.59868, + "grad_norm": 1.8489231537337159, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 59868 + }, + { + "epoch": 0.59869, + "grad_norm": 2.258354176909436, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 59869 + }, + { + "epoch": 0.5987, + "grad_norm": 1.078558103104986, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 59870 + }, + { + "epoch": 0.59871, + "grad_norm": 1.670110071139575, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 59871 + }, + { + "epoch": 0.59872, + "grad_norm": 1.5419745873432948, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 59872 + }, + { + "epoch": 0.59873, + "grad_norm": 1.6075802712429392, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 59873 + }, + { + "epoch": 0.59874, + "grad_norm": 1.5358806523465924, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 59874 + }, + { + "epoch": 0.59875, + "grad_norm": 1.9992449676037973, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 59875 + }, + { + "epoch": 0.59876, + "grad_norm": 1.35552069587056, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 59876 + }, + { + "epoch": 0.59877, + "grad_norm": 1.5685508813334033, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 59877 + }, + { + "epoch": 0.59878, + "grad_norm": 1.6156848853451642, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 59878 + }, + { + "epoch": 0.59879, + "grad_norm": 1.7242602235862694, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 59879 + }, + { + "epoch": 0.5988, + "grad_norm": 1.3741704110397914, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 59880 + }, + { + "epoch": 0.59881, + "grad_norm": 1.5675563806954655, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 59881 + }, + { + "epoch": 0.59882, + "grad_norm": 1.7445314714381326, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 59882 + }, + { + "epoch": 0.59883, + "grad_norm": 1.284359172821299, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 59883 + }, + { + "epoch": 0.59884, + "grad_norm": 1.5509914448073454, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 59884 + }, + { + "epoch": 0.59885, + "grad_norm": 1.5602444513295888, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 59885 + }, + { + "epoch": 0.59886, + "grad_norm": 1.5433901103164207, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 59886 + }, + { + "epoch": 0.59887, + "grad_norm": 1.6873670044680087, + "learning_rate": 0.003, + "loss": 4.041, + "step": 59887 + }, + { + "epoch": 0.59888, + "grad_norm": 1.3037429550830648, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 59888 + }, + { + "epoch": 0.59889, + "grad_norm": 1.5664804644869177, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 59889 + }, + { + "epoch": 0.5989, + "grad_norm": 1.426589888900929, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 59890 + }, + { + "epoch": 0.59891, + "grad_norm": 1.5338595244587223, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 59891 + }, + { + "epoch": 0.59892, + "grad_norm": 1.514802682617255, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 59892 + }, + { + "epoch": 0.59893, + "grad_norm": 1.4095914448914173, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 59893 + }, + { + "epoch": 0.59894, + "grad_norm": 1.5834760909245322, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59894 + }, + { + "epoch": 0.59895, + "grad_norm": 1.699570397447232, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 59895 + }, + { + "epoch": 0.59896, + "grad_norm": 1.4646775584649794, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 59896 + }, + { + "epoch": 0.59897, + "grad_norm": 1.5674861416222299, + "learning_rate": 0.003, + "loss": 4.0, + "step": 59897 + }, + { + "epoch": 0.59898, + "grad_norm": 2.1610262001131524, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 59898 + }, + { + "epoch": 0.59899, + "grad_norm": 1.162523783671237, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 59899 + }, + { + "epoch": 0.599, + "grad_norm": 2.105495601029195, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 59900 + }, + { + "epoch": 0.59901, + "grad_norm": 1.562834296031644, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 59901 + }, + { + "epoch": 0.59902, + "grad_norm": 1.5864526634488003, + "learning_rate": 0.003, + "loss": 4.041, + "step": 59902 + }, + { + "epoch": 0.59903, + "grad_norm": 1.3610435206978522, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 59903 + }, + { + "epoch": 0.59904, + "grad_norm": 1.6454306083206791, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 59904 + }, + { + "epoch": 0.59905, + "grad_norm": 1.4378735320598435, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 59905 + }, + { + "epoch": 0.59906, + "grad_norm": 1.4037688088822755, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 59906 + }, + { + "epoch": 0.59907, + "grad_norm": 1.6039930573969394, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 59907 + }, + { + "epoch": 0.59908, + "grad_norm": 1.3753102417640872, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 59908 + }, + { + "epoch": 0.59909, + "grad_norm": 1.5734424244259042, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 59909 + }, + { + "epoch": 0.5991, + "grad_norm": 1.4323679207593538, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 59910 + }, + { + "epoch": 0.59911, + "grad_norm": 1.4748097389711472, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 59911 + }, + { + "epoch": 0.59912, + "grad_norm": 1.2929629104706388, + "learning_rate": 0.003, + "loss": 4.06, + "step": 59912 + }, + { + "epoch": 0.59913, + "grad_norm": 2.0648030941619435, + "learning_rate": 0.003, + "loss": 4.037, + "step": 59913 + }, + { + "epoch": 0.59914, + "grad_norm": 1.9805462325356746, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 59914 + }, + { + "epoch": 0.59915, + "grad_norm": 1.7744015837207383, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 59915 + }, + { + "epoch": 0.59916, + "grad_norm": 1.6121427279085017, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 59916 + }, + { + "epoch": 0.59917, + "grad_norm": 1.749477281701498, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 59917 + }, + { + "epoch": 0.59918, + "grad_norm": 1.4166154640858362, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 59918 + }, + { + "epoch": 0.59919, + "grad_norm": 1.8617966784051019, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 59919 + }, + { + "epoch": 0.5992, + "grad_norm": 1.0960434536723327, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 59920 + }, + { + "epoch": 0.59921, + "grad_norm": 1.9771308298541619, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 59921 + }, + { + "epoch": 0.59922, + "grad_norm": 2.044257684765474, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 59922 + }, + { + "epoch": 0.59923, + "grad_norm": 1.370799770294274, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 59923 + }, + { + "epoch": 0.59924, + "grad_norm": 1.7757017655133838, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 59924 + }, + { + "epoch": 0.59925, + "grad_norm": 1.7123293673996776, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 59925 + }, + { + "epoch": 0.59926, + "grad_norm": 1.666838979556726, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 59926 + }, + { + "epoch": 0.59927, + "grad_norm": 1.3859463303412487, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 59927 + }, + { + "epoch": 0.59928, + "grad_norm": 1.7041022043083967, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 59928 + }, + { + "epoch": 0.59929, + "grad_norm": 1.2790221681079679, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 59929 + }, + { + "epoch": 0.5993, + "grad_norm": 1.7881238083305686, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 59930 + }, + { + "epoch": 0.59931, + "grad_norm": 1.2393828411276666, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 59931 + }, + { + "epoch": 0.59932, + "grad_norm": 1.9174468868080454, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 59932 + }, + { + "epoch": 0.59933, + "grad_norm": 1.6526544469051065, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 59933 + }, + { + "epoch": 0.59934, + "grad_norm": 1.6159673824303646, + "learning_rate": 0.003, + "loss": 4.03, + "step": 59934 + }, + { + "epoch": 0.59935, + "grad_norm": 1.4158865133164855, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 59935 + }, + { + "epoch": 0.59936, + "grad_norm": 1.5444951039191825, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 59936 + }, + { + "epoch": 0.59937, + "grad_norm": 1.7093637657542495, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 59937 + }, + { + "epoch": 0.59938, + "grad_norm": 1.5144376715345016, + "learning_rate": 0.003, + "loss": 4.037, + "step": 59938 + }, + { + "epoch": 0.59939, + "grad_norm": 1.696217448269004, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 59939 + }, + { + "epoch": 0.5994, + "grad_norm": 1.4516065860475467, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 59940 + }, + { + "epoch": 0.59941, + "grad_norm": 1.559971350020425, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 59941 + }, + { + "epoch": 0.59942, + "grad_norm": 1.5993301326650873, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 59942 + }, + { + "epoch": 0.59943, + "grad_norm": 1.5493594649051234, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 59943 + }, + { + "epoch": 0.59944, + "grad_norm": 1.7225656063933728, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 59944 + }, + { + "epoch": 0.59945, + "grad_norm": 1.5997744095456476, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 59945 + }, + { + "epoch": 0.59946, + "grad_norm": 1.5549016278885148, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 59946 + }, + { + "epoch": 0.59947, + "grad_norm": 1.617839264146737, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 59947 + }, + { + "epoch": 0.59948, + "grad_norm": 1.48167901031751, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 59948 + }, + { + "epoch": 0.59949, + "grad_norm": 1.6855626896404028, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 59949 + }, + { + "epoch": 0.5995, + "grad_norm": 1.3602143536328946, + "learning_rate": 0.003, + "loss": 4.02, + "step": 59950 + }, + { + "epoch": 0.59951, + "grad_norm": 1.8047800545972275, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 59951 + }, + { + "epoch": 0.59952, + "grad_norm": 1.3668015847550077, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 59952 + }, + { + "epoch": 0.59953, + "grad_norm": 2.045655181341772, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 59953 + }, + { + "epoch": 0.59954, + "grad_norm": 1.3563161402208674, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 59954 + }, + { + "epoch": 0.59955, + "grad_norm": 1.7741427506034262, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 59955 + }, + { + "epoch": 0.59956, + "grad_norm": 1.4757769565946266, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 59956 + }, + { + "epoch": 0.59957, + "grad_norm": 1.7153046318098362, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 59957 + }, + { + "epoch": 0.59958, + "grad_norm": 1.2720004491866534, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 59958 + }, + { + "epoch": 0.59959, + "grad_norm": 2.024914832021339, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 59959 + }, + { + "epoch": 0.5996, + "grad_norm": 1.6722217996107849, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 59960 + }, + { + "epoch": 0.59961, + "grad_norm": 1.69324578776699, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 59961 + }, + { + "epoch": 0.59962, + "grad_norm": 1.4256439453672265, + "learning_rate": 0.003, + "loss": 4.039, + "step": 59962 + }, + { + "epoch": 0.59963, + "grad_norm": 1.9593689992600414, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 59963 + }, + { + "epoch": 0.59964, + "grad_norm": 4.805849052173863, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 59964 + }, + { + "epoch": 0.59965, + "grad_norm": 2.0835222512773766, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 59965 + }, + { + "epoch": 0.59966, + "grad_norm": 1.6439964661753341, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 59966 + }, + { + "epoch": 0.59967, + "grad_norm": 1.3442882724785918, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 59967 + }, + { + "epoch": 0.59968, + "grad_norm": 2.0438628567231767, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 59968 + }, + { + "epoch": 0.59969, + "grad_norm": 1.8707838298883541, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 59969 + }, + { + "epoch": 0.5997, + "grad_norm": 1.9079320048027137, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 59970 + }, + { + "epoch": 0.59971, + "grad_norm": 1.8776888661232474, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 59971 + }, + { + "epoch": 0.59972, + "grad_norm": 2.368545020998219, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 59972 + }, + { + "epoch": 0.59973, + "grad_norm": 2.1006201918523772, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 59973 + }, + { + "epoch": 0.59974, + "grad_norm": 2.133389664875934, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 59974 + }, + { + "epoch": 0.59975, + "grad_norm": 2.3808728758892603, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 59975 + }, + { + "epoch": 0.59976, + "grad_norm": 1.5991864462281287, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 59976 + }, + { + "epoch": 0.59977, + "grad_norm": 1.65449005796705, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 59977 + }, + { + "epoch": 0.59978, + "grad_norm": 2.1869474756262837, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 59978 + }, + { + "epoch": 0.59979, + "grad_norm": 1.796752259847935, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 59979 + }, + { + "epoch": 0.5998, + "grad_norm": 1.7448395729756359, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 59980 + }, + { + "epoch": 0.59981, + "grad_norm": 1.7158356641931347, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 59981 + }, + { + "epoch": 0.59982, + "grad_norm": 1.5028891124026065, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 59982 + }, + { + "epoch": 0.59983, + "grad_norm": 2.067228192534617, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 59983 + }, + { + "epoch": 0.59984, + "grad_norm": 1.710486060918132, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 59984 + }, + { + "epoch": 0.59985, + "grad_norm": 1.5049793541569922, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 59985 + }, + { + "epoch": 0.59986, + "grad_norm": 1.8401329321680329, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 59986 + }, + { + "epoch": 0.59987, + "grad_norm": 1.2207185805395668, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 59987 + }, + { + "epoch": 0.59988, + "grad_norm": 1.8667986121348914, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 59988 + }, + { + "epoch": 0.59989, + "grad_norm": 1.245029958165769, + "learning_rate": 0.003, + "loss": 4.075, + "step": 59989 + }, + { + "epoch": 0.5999, + "grad_norm": 1.8221371530104113, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 59990 + }, + { + "epoch": 0.59991, + "grad_norm": 1.3152097396829205, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 59991 + }, + { + "epoch": 0.59992, + "grad_norm": 1.6438237369587847, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 59992 + }, + { + "epoch": 0.59993, + "grad_norm": 1.7609839199410617, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 59993 + }, + { + "epoch": 0.59994, + "grad_norm": 2.0421233617388608, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 59994 + }, + { + "epoch": 0.59995, + "grad_norm": 1.4458704635843387, + "learning_rate": 0.003, + "loss": 4.024, + "step": 59995 + }, + { + "epoch": 0.59996, + "grad_norm": 1.8479977828370147, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 59996 + }, + { + "epoch": 0.59997, + "grad_norm": 1.6235469240995248, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 59997 + }, + { + "epoch": 0.59998, + "grad_norm": 1.692231806691251, + "learning_rate": 0.003, + "loss": 4.059, + "step": 59998 + }, + { + "epoch": 0.59999, + "grad_norm": 1.4345792905575367, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 59999 + }, + { + "epoch": 0.6, + "grad_norm": 1.5732999030891943, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 60000 + }, + { + "epoch": 0.60001, + "grad_norm": 1.4220087729662028, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 60001 + }, + { + "epoch": 0.60002, + "grad_norm": 1.680395319451223, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 60002 + }, + { + "epoch": 0.60003, + "grad_norm": 1.4595106393808472, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 60003 + }, + { + "epoch": 0.60004, + "grad_norm": 1.8852267902375754, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 60004 + }, + { + "epoch": 0.60005, + "grad_norm": 1.943076706873448, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 60005 + }, + { + "epoch": 0.60006, + "grad_norm": 2.2529546987498095, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 60006 + }, + { + "epoch": 0.60007, + "grad_norm": 1.6152694514565298, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 60007 + }, + { + "epoch": 0.60008, + "grad_norm": 1.7048200498680297, + "learning_rate": 0.003, + "loss": 4.09, + "step": 60008 + }, + { + "epoch": 0.60009, + "grad_norm": 2.1229186321234597, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 60009 + }, + { + "epoch": 0.6001, + "grad_norm": 2.5062951481328097, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 60010 + }, + { + "epoch": 0.60011, + "grad_norm": 1.5845591946781856, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 60011 + }, + { + "epoch": 0.60012, + "grad_norm": 1.8029009644224079, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 60012 + }, + { + "epoch": 0.60013, + "grad_norm": 1.5372512553641653, + "learning_rate": 0.003, + "loss": 4.079, + "step": 60013 + }, + { + "epoch": 0.60014, + "grad_norm": 1.996377332059518, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 60014 + }, + { + "epoch": 0.60015, + "grad_norm": 1.5372799898523137, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 60015 + }, + { + "epoch": 0.60016, + "grad_norm": 2.003145647408536, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 60016 + }, + { + "epoch": 0.60017, + "grad_norm": 1.8549394683051341, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 60017 + }, + { + "epoch": 0.60018, + "grad_norm": 1.901212615574601, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 60018 + }, + { + "epoch": 0.60019, + "grad_norm": 1.7585370004131273, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 60019 + }, + { + "epoch": 0.6002, + "grad_norm": 2.0339005721896615, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 60020 + }, + { + "epoch": 0.60021, + "grad_norm": 1.58925088647434, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 60021 + }, + { + "epoch": 0.60022, + "grad_norm": 1.7441545698392205, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 60022 + }, + { + "epoch": 0.60023, + "grad_norm": 2.603010955890891, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 60023 + }, + { + "epoch": 0.60024, + "grad_norm": 1.5402575163528534, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 60024 + }, + { + "epoch": 0.60025, + "grad_norm": 1.5888001652952137, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 60025 + }, + { + "epoch": 0.60026, + "grad_norm": 1.9368593651281363, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 60026 + }, + { + "epoch": 0.60027, + "grad_norm": 1.5033756709410784, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 60027 + }, + { + "epoch": 0.60028, + "grad_norm": 1.9500206213738018, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 60028 + }, + { + "epoch": 0.60029, + "grad_norm": 1.3186502029944096, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 60029 + }, + { + "epoch": 0.6003, + "grad_norm": 1.8012302293262874, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 60030 + }, + { + "epoch": 0.60031, + "grad_norm": 1.6635211077136716, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 60031 + }, + { + "epoch": 0.60032, + "grad_norm": 1.8902508248027816, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 60032 + }, + { + "epoch": 0.60033, + "grad_norm": 1.6601000365391643, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 60033 + }, + { + "epoch": 0.60034, + "grad_norm": 1.7847768931936543, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 60034 + }, + { + "epoch": 0.60035, + "grad_norm": 1.6518804642457254, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 60035 + }, + { + "epoch": 0.60036, + "grad_norm": 1.72370873974794, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 60036 + }, + { + "epoch": 0.60037, + "grad_norm": 1.3412066641908498, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 60037 + }, + { + "epoch": 0.60038, + "grad_norm": 1.7897922809877351, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 60038 + }, + { + "epoch": 0.60039, + "grad_norm": 1.3951159686486938, + "learning_rate": 0.003, + "loss": 4.038, + "step": 60039 + }, + { + "epoch": 0.6004, + "grad_norm": 1.9657571382991355, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 60040 + }, + { + "epoch": 0.60041, + "grad_norm": 1.6129080977748909, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 60041 + }, + { + "epoch": 0.60042, + "grad_norm": 1.6767043983517818, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 60042 + }, + { + "epoch": 0.60043, + "grad_norm": 1.627351540766552, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 60043 + }, + { + "epoch": 0.60044, + "grad_norm": 1.4711805897029768, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 60044 + }, + { + "epoch": 0.60045, + "grad_norm": 1.666192495541977, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 60045 + }, + { + "epoch": 0.60046, + "grad_norm": 1.5473900677817118, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 60046 + }, + { + "epoch": 0.60047, + "grad_norm": 1.5826596436552611, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 60047 + }, + { + "epoch": 0.60048, + "grad_norm": 1.238831706250076, + "learning_rate": 0.003, + "loss": 4.054, + "step": 60048 + }, + { + "epoch": 0.60049, + "grad_norm": 1.7899502216052778, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 60049 + }, + { + "epoch": 0.6005, + "grad_norm": 1.4577172289201898, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 60050 + }, + { + "epoch": 0.60051, + "grad_norm": 1.6163835680859933, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 60051 + }, + { + "epoch": 0.60052, + "grad_norm": 1.8429983539813364, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 60052 + }, + { + "epoch": 0.60053, + "grad_norm": 1.5638177513911482, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 60053 + }, + { + "epoch": 0.60054, + "grad_norm": 1.5913715099361634, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 60054 + }, + { + "epoch": 0.60055, + "grad_norm": 1.3865937267948336, + "learning_rate": 0.003, + "loss": 4.079, + "step": 60055 + }, + { + "epoch": 0.60056, + "grad_norm": 1.6935196031629147, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 60056 + }, + { + "epoch": 0.60057, + "grad_norm": 1.1763069470365102, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 60057 + }, + { + "epoch": 0.60058, + "grad_norm": 1.6975572560356993, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 60058 + }, + { + "epoch": 0.60059, + "grad_norm": 1.3697114698014117, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 60059 + }, + { + "epoch": 0.6006, + "grad_norm": 1.6812724208385634, + "learning_rate": 0.003, + "loss": 4.038, + "step": 60060 + }, + { + "epoch": 0.60061, + "grad_norm": 2.0296507931538184, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 60061 + }, + { + "epoch": 0.60062, + "grad_norm": 1.3608798750557909, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 60062 + }, + { + "epoch": 0.60063, + "grad_norm": 1.2914972102157145, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 60063 + }, + { + "epoch": 0.60064, + "grad_norm": 1.7783711629483296, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 60064 + }, + { + "epoch": 0.60065, + "grad_norm": 1.8070460158262325, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 60065 + }, + { + "epoch": 0.60066, + "grad_norm": 1.5152440074791516, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 60066 + }, + { + "epoch": 0.60067, + "grad_norm": 1.530530850936317, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 60067 + }, + { + "epoch": 0.60068, + "grad_norm": 1.4356695484322675, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 60068 + }, + { + "epoch": 0.60069, + "grad_norm": 1.426964533436205, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 60069 + }, + { + "epoch": 0.6007, + "grad_norm": 1.4863923769468086, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 60070 + }, + { + "epoch": 0.60071, + "grad_norm": 1.405467302346576, + "learning_rate": 0.003, + "loss": 4.025, + "step": 60071 + }, + { + "epoch": 0.60072, + "grad_norm": 1.7227710554518925, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 60072 + }, + { + "epoch": 0.60073, + "grad_norm": 1.5554594406882414, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 60073 + }, + { + "epoch": 0.60074, + "grad_norm": 1.682607076763209, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 60074 + }, + { + "epoch": 0.60075, + "grad_norm": 1.2994682145587586, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 60075 + }, + { + "epoch": 0.60076, + "grad_norm": 2.0078581470950354, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60076 + }, + { + "epoch": 0.60077, + "grad_norm": 1.1453724275917831, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 60077 + }, + { + "epoch": 0.60078, + "grad_norm": 2.030926775757416, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60078 + }, + { + "epoch": 0.60079, + "grad_norm": 1.5741750625316608, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 60079 + }, + { + "epoch": 0.6008, + "grad_norm": 1.9084794055504763, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 60080 + }, + { + "epoch": 0.60081, + "grad_norm": 1.4166876324157622, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 60081 + }, + { + "epoch": 0.60082, + "grad_norm": 1.4951584463960332, + "learning_rate": 0.003, + "loss": 4.029, + "step": 60082 + }, + { + "epoch": 0.60083, + "grad_norm": 1.7294910566470114, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 60083 + }, + { + "epoch": 0.60084, + "grad_norm": 1.5557437091681847, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60084 + }, + { + "epoch": 0.60085, + "grad_norm": 1.5009044913606806, + "learning_rate": 0.003, + "loss": 4.018, + "step": 60085 + }, + { + "epoch": 0.60086, + "grad_norm": 1.6120503269350646, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 60086 + }, + { + "epoch": 0.60087, + "grad_norm": 1.6347773031640143, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 60087 + }, + { + "epoch": 0.60088, + "grad_norm": 1.3413673312340038, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 60088 + }, + { + "epoch": 0.60089, + "grad_norm": 1.711962712686804, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 60089 + }, + { + "epoch": 0.6009, + "grad_norm": 1.6208893759295482, + "learning_rate": 0.003, + "loss": 4.038, + "step": 60090 + }, + { + "epoch": 0.60091, + "grad_norm": 1.5981331293768326, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60091 + }, + { + "epoch": 0.60092, + "grad_norm": 1.5562903747081993, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 60092 + }, + { + "epoch": 0.60093, + "grad_norm": 1.4249782039660086, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 60093 + }, + { + "epoch": 0.60094, + "grad_norm": 1.40002942250853, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 60094 + }, + { + "epoch": 0.60095, + "grad_norm": 1.4411527306121348, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 60095 + }, + { + "epoch": 0.60096, + "grad_norm": 1.648618228485054, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 60096 + }, + { + "epoch": 0.60097, + "grad_norm": 1.5104044319218906, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 60097 + }, + { + "epoch": 0.60098, + "grad_norm": 1.6600396685556646, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 60098 + }, + { + "epoch": 0.60099, + "grad_norm": 1.471519159286577, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 60099 + }, + { + "epoch": 0.601, + "grad_norm": 1.6173011449955854, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 60100 + }, + { + "epoch": 0.60101, + "grad_norm": 1.590615711805899, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 60101 + }, + { + "epoch": 0.60102, + "grad_norm": 1.5137022964222322, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 60102 + }, + { + "epoch": 0.60103, + "grad_norm": 1.34563550578989, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 60103 + }, + { + "epoch": 0.60104, + "grad_norm": 1.9612576174043774, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 60104 + }, + { + "epoch": 0.60105, + "grad_norm": 1.1346642506907707, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 60105 + }, + { + "epoch": 0.60106, + "grad_norm": 1.6472334561380684, + "learning_rate": 0.003, + "loss": 4.006, + "step": 60106 + }, + { + "epoch": 0.60107, + "grad_norm": 1.6260397910829838, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 60107 + }, + { + "epoch": 0.60108, + "grad_norm": 1.9808545602746739, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 60108 + }, + { + "epoch": 0.60109, + "grad_norm": 1.537111018446546, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 60109 + }, + { + "epoch": 0.6011, + "grad_norm": 1.5109967876298276, + "learning_rate": 0.003, + "loss": 4.067, + "step": 60110 + }, + { + "epoch": 0.60111, + "grad_norm": 1.5099084460050054, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 60111 + }, + { + "epoch": 0.60112, + "grad_norm": 1.2761092014708812, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 60112 + }, + { + "epoch": 0.60113, + "grad_norm": 1.711158743250311, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 60113 + }, + { + "epoch": 0.60114, + "grad_norm": 1.1302164362427933, + "learning_rate": 0.003, + "loss": 4.044, + "step": 60114 + }, + { + "epoch": 0.60115, + "grad_norm": 1.891134962230167, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 60115 + }, + { + "epoch": 0.60116, + "grad_norm": 1.5931617603337287, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 60116 + }, + { + "epoch": 0.60117, + "grad_norm": 2.3042013511175425, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 60117 + }, + { + "epoch": 0.60118, + "grad_norm": 1.2842043778700338, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 60118 + }, + { + "epoch": 0.60119, + "grad_norm": 1.5977498846327578, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 60119 + }, + { + "epoch": 0.6012, + "grad_norm": 1.6861870272491573, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 60120 + }, + { + "epoch": 0.60121, + "grad_norm": 1.595514350066349, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 60121 + }, + { + "epoch": 0.60122, + "grad_norm": 1.8022146321089985, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 60122 + }, + { + "epoch": 0.60123, + "grad_norm": 1.713716933597866, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 60123 + }, + { + "epoch": 0.60124, + "grad_norm": 1.3348301069261697, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 60124 + }, + { + "epoch": 0.60125, + "grad_norm": 1.3140882939437772, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 60125 + }, + { + "epoch": 0.60126, + "grad_norm": 1.837892784756682, + "learning_rate": 0.003, + "loss": 4.057, + "step": 60126 + }, + { + "epoch": 0.60127, + "grad_norm": 1.1683483671893669, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 60127 + }, + { + "epoch": 0.60128, + "grad_norm": 1.6788362528080543, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 60128 + }, + { + "epoch": 0.60129, + "grad_norm": 1.3634295208815022, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 60129 + }, + { + "epoch": 0.6013, + "grad_norm": 1.563181863081111, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 60130 + }, + { + "epoch": 0.60131, + "grad_norm": 1.6653013042713602, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 60131 + }, + { + "epoch": 0.60132, + "grad_norm": 1.5486108502955074, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 60132 + }, + { + "epoch": 0.60133, + "grad_norm": 1.4818794747099064, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 60133 + }, + { + "epoch": 0.60134, + "grad_norm": 2.0512269931306077, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 60134 + }, + { + "epoch": 0.60135, + "grad_norm": 1.456714071255909, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 60135 + }, + { + "epoch": 0.60136, + "grad_norm": 1.6716950551218328, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 60136 + }, + { + "epoch": 0.60137, + "grad_norm": 1.343628854343338, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 60137 + }, + { + "epoch": 0.60138, + "grad_norm": 1.572701323708233, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 60138 + }, + { + "epoch": 0.60139, + "grad_norm": 1.4535888561481825, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 60139 + }, + { + "epoch": 0.6014, + "grad_norm": 1.308538042562216, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 60140 + }, + { + "epoch": 0.60141, + "grad_norm": 1.5889533819594799, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 60141 + }, + { + "epoch": 0.60142, + "grad_norm": 1.5006084174306977, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 60142 + }, + { + "epoch": 0.60143, + "grad_norm": 1.8156131116931578, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 60143 + }, + { + "epoch": 0.60144, + "grad_norm": 1.4799249589474275, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 60144 + }, + { + "epoch": 0.60145, + "grad_norm": 1.7036817303925873, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 60145 + }, + { + "epoch": 0.60146, + "grad_norm": 1.2217354114948147, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 60146 + }, + { + "epoch": 0.60147, + "grad_norm": 1.5735180402895053, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 60147 + }, + { + "epoch": 0.60148, + "grad_norm": 1.4188548277665074, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 60148 + }, + { + "epoch": 0.60149, + "grad_norm": 1.934315816043006, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 60149 + }, + { + "epoch": 0.6015, + "grad_norm": 1.2239684764248404, + "learning_rate": 0.003, + "loss": 4.019, + "step": 60150 + }, + { + "epoch": 0.60151, + "grad_norm": 1.5639574035748443, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 60151 + }, + { + "epoch": 0.60152, + "grad_norm": 1.3360487054894918, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 60152 + }, + { + "epoch": 0.60153, + "grad_norm": 1.5400644834602681, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 60153 + }, + { + "epoch": 0.60154, + "grad_norm": 1.4167691446670296, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 60154 + }, + { + "epoch": 0.60155, + "grad_norm": 1.9124843356169894, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 60155 + }, + { + "epoch": 0.60156, + "grad_norm": 1.6122260484944542, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 60156 + }, + { + "epoch": 0.60157, + "grad_norm": 1.9327618334144956, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 60157 + }, + { + "epoch": 0.60158, + "grad_norm": 1.610841220244119, + "learning_rate": 0.003, + "loss": 4.048, + "step": 60158 + }, + { + "epoch": 0.60159, + "grad_norm": 1.3898941330496442, + "learning_rate": 0.003, + "loss": 4.031, + "step": 60159 + }, + { + "epoch": 0.6016, + "grad_norm": 1.6633916464127647, + "learning_rate": 0.003, + "loss": 4.042, + "step": 60160 + }, + { + "epoch": 0.60161, + "grad_norm": 1.5216790369446764, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 60161 + }, + { + "epoch": 0.60162, + "grad_norm": 1.5658364522919166, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 60162 + }, + { + "epoch": 0.60163, + "grad_norm": 1.6442805130599076, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 60163 + }, + { + "epoch": 0.60164, + "grad_norm": 1.4107896528474901, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 60164 + }, + { + "epoch": 0.60165, + "grad_norm": 1.558189406963106, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 60165 + }, + { + "epoch": 0.60166, + "grad_norm": 1.3729706824646302, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 60166 + }, + { + "epoch": 0.60167, + "grad_norm": 1.6166097437404137, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 60167 + }, + { + "epoch": 0.60168, + "grad_norm": 1.3856179028175868, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 60168 + }, + { + "epoch": 0.60169, + "grad_norm": 1.5665258497696324, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 60169 + }, + { + "epoch": 0.6017, + "grad_norm": 1.2047559447170013, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 60170 + }, + { + "epoch": 0.60171, + "grad_norm": 1.966160389209346, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 60171 + }, + { + "epoch": 0.60172, + "grad_norm": 1.507328823232444, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 60172 + }, + { + "epoch": 0.60173, + "grad_norm": 2.0361038111992222, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 60173 + }, + { + "epoch": 0.60174, + "grad_norm": 1.4933369825202942, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 60174 + }, + { + "epoch": 0.60175, + "grad_norm": 1.356428001251298, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 60175 + }, + { + "epoch": 0.60176, + "grad_norm": 1.93332523715531, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 60176 + }, + { + "epoch": 0.60177, + "grad_norm": 1.5588851031403268, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 60177 + }, + { + "epoch": 0.60178, + "grad_norm": 1.4723426535558488, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 60178 + }, + { + "epoch": 0.60179, + "grad_norm": 1.577381429640157, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 60179 + }, + { + "epoch": 0.6018, + "grad_norm": 1.7963328733485073, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 60180 + }, + { + "epoch": 0.60181, + "grad_norm": 1.3116658885302335, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 60181 + }, + { + "epoch": 0.60182, + "grad_norm": 1.8764041589255083, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 60182 + }, + { + "epoch": 0.60183, + "grad_norm": 1.468125345206343, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 60183 + }, + { + "epoch": 0.60184, + "grad_norm": 1.9140461239604902, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 60184 + }, + { + "epoch": 0.60185, + "grad_norm": 1.6271991421564567, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 60185 + }, + { + "epoch": 0.60186, + "grad_norm": 1.6424070694005388, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 60186 + }, + { + "epoch": 0.60187, + "grad_norm": 1.380658982994187, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 60187 + }, + { + "epoch": 0.60188, + "grad_norm": 1.7061712798333635, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 60188 + }, + { + "epoch": 0.60189, + "grad_norm": 1.1766635583332583, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 60189 + }, + { + "epoch": 0.6019, + "grad_norm": 1.8207350561951847, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 60190 + }, + { + "epoch": 0.60191, + "grad_norm": 1.1268139377025133, + "learning_rate": 0.003, + "loss": 4.039, + "step": 60191 + }, + { + "epoch": 0.60192, + "grad_norm": 1.9622209167936888, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 60192 + }, + { + "epoch": 0.60193, + "grad_norm": 1.4444306593689054, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 60193 + }, + { + "epoch": 0.60194, + "grad_norm": 1.9377735050099971, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 60194 + }, + { + "epoch": 0.60195, + "grad_norm": 1.3606316374294858, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 60195 + }, + { + "epoch": 0.60196, + "grad_norm": 1.6768002547958019, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 60196 + }, + { + "epoch": 0.60197, + "grad_norm": 1.396093231764986, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 60197 + }, + { + "epoch": 0.60198, + "grad_norm": 1.4322058292509379, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 60198 + }, + { + "epoch": 0.60199, + "grad_norm": 1.580903833205013, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60199 + }, + { + "epoch": 0.602, + "grad_norm": 1.4692200026597448, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 60200 + }, + { + "epoch": 0.60201, + "grad_norm": 1.6105414487509062, + "learning_rate": 0.003, + "loss": 4.057, + "step": 60201 + }, + { + "epoch": 0.60202, + "grad_norm": 1.494927609567162, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 60202 + }, + { + "epoch": 0.60203, + "grad_norm": 1.9486747285294272, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 60203 + }, + { + "epoch": 0.60204, + "grad_norm": 1.2471367632985597, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 60204 + }, + { + "epoch": 0.60205, + "grad_norm": 1.7066249294580234, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 60205 + }, + { + "epoch": 0.60206, + "grad_norm": 1.3619929558123252, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 60206 + }, + { + "epoch": 0.60207, + "grad_norm": 1.6455834545747179, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 60207 + }, + { + "epoch": 0.60208, + "grad_norm": 1.9107252574985747, + "learning_rate": 0.003, + "loss": 4.028, + "step": 60208 + }, + { + "epoch": 0.60209, + "grad_norm": 1.3378348866045604, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 60209 + }, + { + "epoch": 0.6021, + "grad_norm": 1.3828336186205188, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 60210 + }, + { + "epoch": 0.60211, + "grad_norm": 1.491708875159125, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 60211 + }, + { + "epoch": 0.60212, + "grad_norm": 1.4606964316712254, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60212 + }, + { + "epoch": 0.60213, + "grad_norm": 1.518600013969181, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 60213 + }, + { + "epoch": 0.60214, + "grad_norm": 1.7287368936374257, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 60214 + }, + { + "epoch": 0.60215, + "grad_norm": 1.7046803570042008, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 60215 + }, + { + "epoch": 0.60216, + "grad_norm": 1.612718071038228, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 60216 + }, + { + "epoch": 0.60217, + "grad_norm": 1.4606727523123126, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60217 + }, + { + "epoch": 0.60218, + "grad_norm": 1.5105919559170127, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 60218 + }, + { + "epoch": 0.60219, + "grad_norm": 1.4775052286131627, + "learning_rate": 0.003, + "loss": 4.039, + "step": 60219 + }, + { + "epoch": 0.6022, + "grad_norm": 1.5862832581329336, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 60220 + }, + { + "epoch": 0.60221, + "grad_norm": 1.6270304152036834, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 60221 + }, + { + "epoch": 0.60222, + "grad_norm": 1.477368078877971, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 60222 + }, + { + "epoch": 0.60223, + "grad_norm": 1.5457834871767266, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 60223 + }, + { + "epoch": 0.60224, + "grad_norm": 1.3972711334233734, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 60224 + }, + { + "epoch": 0.60225, + "grad_norm": 1.6188201344120952, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 60225 + }, + { + "epoch": 0.60226, + "grad_norm": 1.3003193083671158, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 60226 + }, + { + "epoch": 0.60227, + "grad_norm": 1.6133181780604617, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 60227 + }, + { + "epoch": 0.60228, + "grad_norm": 1.3377519450541262, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 60228 + }, + { + "epoch": 0.60229, + "grad_norm": 1.739193007128017, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 60229 + }, + { + "epoch": 0.6023, + "grad_norm": 1.2366584869642319, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 60230 + }, + { + "epoch": 0.60231, + "grad_norm": 2.276460036888227, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 60231 + }, + { + "epoch": 0.60232, + "grad_norm": 1.6756090627663993, + "learning_rate": 0.003, + "loss": 4.014, + "step": 60232 + }, + { + "epoch": 0.60233, + "grad_norm": 2.0217934729438958, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 60233 + }, + { + "epoch": 0.60234, + "grad_norm": 1.530339801320359, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 60234 + }, + { + "epoch": 0.60235, + "grad_norm": 1.375056642070038, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 60235 + }, + { + "epoch": 0.60236, + "grad_norm": 1.5604210130950193, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 60236 + }, + { + "epoch": 0.60237, + "grad_norm": 1.870690190464013, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 60237 + }, + { + "epoch": 0.60238, + "grad_norm": 1.5035185364282286, + "learning_rate": 0.003, + "loss": 4.031, + "step": 60238 + }, + { + "epoch": 0.60239, + "grad_norm": 1.4118198770028805, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 60239 + }, + { + "epoch": 0.6024, + "grad_norm": 1.6124069377351746, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 60240 + }, + { + "epoch": 0.60241, + "grad_norm": 2.0242341707133003, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 60241 + }, + { + "epoch": 0.60242, + "grad_norm": 1.446818149571553, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 60242 + }, + { + "epoch": 0.60243, + "grad_norm": 1.5539387432238085, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 60243 + }, + { + "epoch": 0.60244, + "grad_norm": 1.7212644653857514, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 60244 + }, + { + "epoch": 0.60245, + "grad_norm": 1.5201089722689147, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 60245 + }, + { + "epoch": 0.60246, + "grad_norm": 1.3526975281442468, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 60246 + }, + { + "epoch": 0.60247, + "grad_norm": 1.5330312401071517, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 60247 + }, + { + "epoch": 0.60248, + "grad_norm": 1.597593861138464, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 60248 + }, + { + "epoch": 0.60249, + "grad_norm": 1.6771087313848456, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 60249 + }, + { + "epoch": 0.6025, + "grad_norm": 1.293601882887356, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 60250 + }, + { + "epoch": 0.60251, + "grad_norm": 1.8465583222050947, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 60251 + }, + { + "epoch": 0.60252, + "grad_norm": 1.266227708294293, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 60252 + }, + { + "epoch": 0.60253, + "grad_norm": 1.7444402837789792, + "learning_rate": 0.003, + "loss": 4.035, + "step": 60253 + }, + { + "epoch": 0.60254, + "grad_norm": 1.3286198347614455, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 60254 + }, + { + "epoch": 0.60255, + "grad_norm": 1.9655341984963128, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60255 + }, + { + "epoch": 0.60256, + "grad_norm": 1.8245542442289688, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 60256 + }, + { + "epoch": 0.60257, + "grad_norm": 1.6301403101591228, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 60257 + }, + { + "epoch": 0.60258, + "grad_norm": 1.471411089720896, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 60258 + }, + { + "epoch": 0.60259, + "grad_norm": 1.56583188223914, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 60259 + }, + { + "epoch": 0.6026, + "grad_norm": 1.3213253174693116, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 60260 + }, + { + "epoch": 0.60261, + "grad_norm": 1.601307980737887, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 60261 + }, + { + "epoch": 0.60262, + "grad_norm": 1.4688533775952106, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 60262 + }, + { + "epoch": 0.60263, + "grad_norm": 1.5803397566497273, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 60263 + }, + { + "epoch": 0.60264, + "grad_norm": 1.6414431274802894, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 60264 + }, + { + "epoch": 0.60265, + "grad_norm": 1.6127753273250731, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 60265 + }, + { + "epoch": 0.60266, + "grad_norm": 1.5489345315130627, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 60266 + }, + { + "epoch": 0.60267, + "grad_norm": 1.4583501011567037, + "learning_rate": 0.003, + "loss": 4.021, + "step": 60267 + }, + { + "epoch": 0.60268, + "grad_norm": 1.3522174984284525, + "learning_rate": 0.003, + "loss": 4.029, + "step": 60268 + }, + { + "epoch": 0.60269, + "grad_norm": 1.8155421449518008, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 60269 + }, + { + "epoch": 0.6027, + "grad_norm": 1.4869583352402893, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 60270 + }, + { + "epoch": 0.60271, + "grad_norm": 1.3996092125714679, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 60271 + }, + { + "epoch": 0.60272, + "grad_norm": 1.5539164693609648, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 60272 + }, + { + "epoch": 0.60273, + "grad_norm": 1.6186780416018773, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 60273 + }, + { + "epoch": 0.60274, + "grad_norm": 1.3876297086218514, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 60274 + }, + { + "epoch": 0.60275, + "grad_norm": 1.8017435764628404, + "learning_rate": 0.003, + "loss": 4.052, + "step": 60275 + }, + { + "epoch": 0.60276, + "grad_norm": 1.2378714378375393, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 60276 + }, + { + "epoch": 0.60277, + "grad_norm": 1.6795803015196185, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 60277 + }, + { + "epoch": 0.60278, + "grad_norm": 1.6751172734379365, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 60278 + }, + { + "epoch": 0.60279, + "grad_norm": 1.3595971446497481, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 60279 + }, + { + "epoch": 0.6028, + "grad_norm": 1.3403992444512975, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 60280 + }, + { + "epoch": 0.60281, + "grad_norm": 1.5538011146196036, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60281 + }, + { + "epoch": 0.60282, + "grad_norm": 1.7447650819943128, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 60282 + }, + { + "epoch": 0.60283, + "grad_norm": 1.3224814548998587, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 60283 + }, + { + "epoch": 0.60284, + "grad_norm": 1.5556271506719572, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 60284 + }, + { + "epoch": 0.60285, + "grad_norm": 1.4393114331734975, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 60285 + }, + { + "epoch": 0.60286, + "grad_norm": 1.605798615886779, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 60286 + }, + { + "epoch": 0.60287, + "grad_norm": 1.3432230512375622, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 60287 + }, + { + "epoch": 0.60288, + "grad_norm": 1.5792520622100814, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 60288 + }, + { + "epoch": 0.60289, + "grad_norm": 1.6067963318429888, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 60289 + }, + { + "epoch": 0.6029, + "grad_norm": 1.9491725078704614, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 60290 + }, + { + "epoch": 0.60291, + "grad_norm": 1.3471962486724838, + "learning_rate": 0.003, + "loss": 4.032, + "step": 60291 + }, + { + "epoch": 0.60292, + "grad_norm": 1.5584411552000865, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 60292 + }, + { + "epoch": 0.60293, + "grad_norm": 1.7049970640013024, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 60293 + }, + { + "epoch": 0.60294, + "grad_norm": 1.4737526816698299, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60294 + }, + { + "epoch": 0.60295, + "grad_norm": 1.5504857757238069, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 60295 + }, + { + "epoch": 0.60296, + "grad_norm": 1.5816709272183884, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 60296 + }, + { + "epoch": 0.60297, + "grad_norm": 1.2248945945684964, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 60297 + }, + { + "epoch": 0.60298, + "grad_norm": 1.8119012392994893, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 60298 + }, + { + "epoch": 0.60299, + "grad_norm": 1.457359711327919, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 60299 + }, + { + "epoch": 0.603, + "grad_norm": 1.5271722659515639, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 60300 + }, + { + "epoch": 0.60301, + "grad_norm": 1.7289090081600345, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 60301 + }, + { + "epoch": 0.60302, + "grad_norm": 1.2580321412311415, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 60302 + }, + { + "epoch": 0.60303, + "grad_norm": 2.1363000165018082, + "learning_rate": 0.003, + "loss": 4.06, + "step": 60303 + }, + { + "epoch": 0.60304, + "grad_norm": 1.4280189328016635, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 60304 + }, + { + "epoch": 0.60305, + "grad_norm": 1.931452296698799, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 60305 + }, + { + "epoch": 0.60306, + "grad_norm": 1.6007892523911085, + "learning_rate": 0.003, + "loss": 4.063, + "step": 60306 + }, + { + "epoch": 0.60307, + "grad_norm": 1.4207138824280796, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 60307 + }, + { + "epoch": 0.60308, + "grad_norm": 1.541240837685976, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 60308 + }, + { + "epoch": 0.60309, + "grad_norm": 1.8450043985248838, + "learning_rate": 0.003, + "loss": 4.029, + "step": 60309 + }, + { + "epoch": 0.6031, + "grad_norm": 1.3680010098214572, + "learning_rate": 0.003, + "loss": 4.063, + "step": 60310 + }, + { + "epoch": 0.60311, + "grad_norm": 1.4625729351260022, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 60311 + }, + { + "epoch": 0.60312, + "grad_norm": 1.5765378883835242, + "learning_rate": 0.003, + "loss": 4.047, + "step": 60312 + }, + { + "epoch": 0.60313, + "grad_norm": 1.5695501600787043, + "learning_rate": 0.003, + "loss": 4.034, + "step": 60313 + }, + { + "epoch": 0.60314, + "grad_norm": 1.582316407710527, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 60314 + }, + { + "epoch": 0.60315, + "grad_norm": 1.7192645764248378, + "learning_rate": 0.003, + "loss": 4.039, + "step": 60315 + }, + { + "epoch": 0.60316, + "grad_norm": 1.3361029713033974, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 60316 + }, + { + "epoch": 0.60317, + "grad_norm": 1.9809733055043484, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 60317 + }, + { + "epoch": 0.60318, + "grad_norm": 1.2250956464383316, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 60318 + }, + { + "epoch": 0.60319, + "grad_norm": 1.894828638515938, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 60319 + }, + { + "epoch": 0.6032, + "grad_norm": 1.5575459521142694, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 60320 + }, + { + "epoch": 0.60321, + "grad_norm": 1.7624420363969155, + "learning_rate": 0.003, + "loss": 4.046, + "step": 60321 + }, + { + "epoch": 0.60322, + "grad_norm": 1.382647128322774, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 60322 + }, + { + "epoch": 0.60323, + "grad_norm": 1.495589647343768, + "learning_rate": 0.003, + "loss": 4.019, + "step": 60323 + }, + { + "epoch": 0.60324, + "grad_norm": 1.5188233624066658, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 60324 + }, + { + "epoch": 0.60325, + "grad_norm": 1.7369323528576655, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 60325 + }, + { + "epoch": 0.60326, + "grad_norm": 1.2572564507706494, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 60326 + }, + { + "epoch": 0.60327, + "grad_norm": 1.8151475828648853, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 60327 + }, + { + "epoch": 0.60328, + "grad_norm": 1.4665192993601202, + "learning_rate": 0.003, + "loss": 4.035, + "step": 60328 + }, + { + "epoch": 0.60329, + "grad_norm": 1.7269219426420945, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 60329 + }, + { + "epoch": 0.6033, + "grad_norm": 1.4984688496262333, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 60330 + }, + { + "epoch": 0.60331, + "grad_norm": 1.5413214980834953, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60331 + }, + { + "epoch": 0.60332, + "grad_norm": 1.6870704311848526, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 60332 + }, + { + "epoch": 0.60333, + "grad_norm": 1.575231619409952, + "learning_rate": 0.003, + "loss": 4.026, + "step": 60333 + }, + { + "epoch": 0.60334, + "grad_norm": 1.9210867484825558, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60334 + }, + { + "epoch": 0.60335, + "grad_norm": 1.3342914668409394, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 60335 + }, + { + "epoch": 0.60336, + "grad_norm": 1.823029615336102, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 60336 + }, + { + "epoch": 0.60337, + "grad_norm": 1.6898500314573854, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 60337 + }, + { + "epoch": 0.60338, + "grad_norm": 1.144002767899715, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 60338 + }, + { + "epoch": 0.60339, + "grad_norm": 1.8171245355875683, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 60339 + }, + { + "epoch": 0.6034, + "grad_norm": 1.2042463444595464, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 60340 + }, + { + "epoch": 0.60341, + "grad_norm": 1.8586388621643213, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 60341 + }, + { + "epoch": 0.60342, + "grad_norm": 1.7507849557583124, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 60342 + }, + { + "epoch": 0.60343, + "grad_norm": 1.3818615247306716, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 60343 + }, + { + "epoch": 0.60344, + "grad_norm": 1.9639120922062567, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 60344 + }, + { + "epoch": 0.60345, + "grad_norm": 1.4931197084766181, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 60345 + }, + { + "epoch": 0.60346, + "grad_norm": 1.3192999978264879, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 60346 + }, + { + "epoch": 0.60347, + "grad_norm": 1.6561421509469942, + "learning_rate": 0.003, + "loss": 4.019, + "step": 60347 + }, + { + "epoch": 0.60348, + "grad_norm": 1.4987995187394632, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 60348 + }, + { + "epoch": 0.60349, + "grad_norm": 1.4697972702267663, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 60349 + }, + { + "epoch": 0.6035, + "grad_norm": 1.7346402328958819, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 60350 + }, + { + "epoch": 0.60351, + "grad_norm": 1.5673987199832686, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 60351 + }, + { + "epoch": 0.60352, + "grad_norm": 1.437875364531013, + "learning_rate": 0.003, + "loss": 4.01, + "step": 60352 + }, + { + "epoch": 0.60353, + "grad_norm": 1.8274693484290874, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60353 + }, + { + "epoch": 0.60354, + "grad_norm": 1.1635343319711275, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 60354 + }, + { + "epoch": 0.60355, + "grad_norm": 1.9894027505899006, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 60355 + }, + { + "epoch": 0.60356, + "grad_norm": 1.3748432808908628, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 60356 + }, + { + "epoch": 0.60357, + "grad_norm": 1.4349670612605507, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 60357 + }, + { + "epoch": 0.60358, + "grad_norm": 1.376384201841875, + "learning_rate": 0.003, + "loss": 4.036, + "step": 60358 + }, + { + "epoch": 0.60359, + "grad_norm": 1.4918118397440976, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 60359 + }, + { + "epoch": 0.6036, + "grad_norm": 1.8285067885938429, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 60360 + }, + { + "epoch": 0.60361, + "grad_norm": 1.4124597078912413, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 60361 + }, + { + "epoch": 0.60362, + "grad_norm": 1.576534161811763, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 60362 + }, + { + "epoch": 0.60363, + "grad_norm": 1.541580294389111, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 60363 + }, + { + "epoch": 0.60364, + "grad_norm": 1.5313025468645378, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 60364 + }, + { + "epoch": 0.60365, + "grad_norm": 1.7410017484283853, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 60365 + }, + { + "epoch": 0.60366, + "grad_norm": 1.4959636310999633, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 60366 + }, + { + "epoch": 0.60367, + "grad_norm": 1.4949635388114673, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 60367 + }, + { + "epoch": 0.60368, + "grad_norm": 1.4265512692656783, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 60368 + }, + { + "epoch": 0.60369, + "grad_norm": 1.71135405964895, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 60369 + }, + { + "epoch": 0.6037, + "grad_norm": 1.2544273688750156, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 60370 + }, + { + "epoch": 0.60371, + "grad_norm": 1.6954243701802536, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 60371 + }, + { + "epoch": 0.60372, + "grad_norm": 1.5108876217951397, + "learning_rate": 0.003, + "loss": 4.018, + "step": 60372 + }, + { + "epoch": 0.60373, + "grad_norm": 1.9628521298740667, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 60373 + }, + { + "epoch": 0.60374, + "grad_norm": 1.5928394493801483, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 60374 + }, + { + "epoch": 0.60375, + "grad_norm": 1.5894288947822617, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 60375 + }, + { + "epoch": 0.60376, + "grad_norm": 1.3348041143472624, + "learning_rate": 0.003, + "loss": 4.004, + "step": 60376 + }, + { + "epoch": 0.60377, + "grad_norm": 1.7097613442876607, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 60377 + }, + { + "epoch": 0.60378, + "grad_norm": 1.6543711367034448, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 60378 + }, + { + "epoch": 0.60379, + "grad_norm": 1.433683563926754, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 60379 + }, + { + "epoch": 0.6038, + "grad_norm": 1.6259531784708003, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 60380 + }, + { + "epoch": 0.60381, + "grad_norm": 1.3200085682274803, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 60381 + }, + { + "epoch": 0.60382, + "grad_norm": 1.6240955014549776, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 60382 + }, + { + "epoch": 0.60383, + "grad_norm": 1.530795213068341, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 60383 + }, + { + "epoch": 0.60384, + "grad_norm": 1.6697000955621109, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 60384 + }, + { + "epoch": 0.60385, + "grad_norm": 2.0853249883971827, + "learning_rate": 0.003, + "loss": 4.055, + "step": 60385 + }, + { + "epoch": 0.60386, + "grad_norm": 1.239068658921872, + "learning_rate": 0.003, + "loss": 4.012, + "step": 60386 + }, + { + "epoch": 0.60387, + "grad_norm": 1.5964121440911365, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 60387 + }, + { + "epoch": 0.60388, + "grad_norm": 1.553143679492303, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 60388 + }, + { + "epoch": 0.60389, + "grad_norm": 1.9131766989531571, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 60389 + }, + { + "epoch": 0.6039, + "grad_norm": 1.0742833194744326, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 60390 + }, + { + "epoch": 0.60391, + "grad_norm": 1.6206975666600434, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 60391 + }, + { + "epoch": 0.60392, + "grad_norm": 1.5562142524607991, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 60392 + }, + { + "epoch": 0.60393, + "grad_norm": 1.6281162451820772, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 60393 + }, + { + "epoch": 0.60394, + "grad_norm": 1.7104875660548167, + "learning_rate": 0.003, + "loss": 4.002, + "step": 60394 + }, + { + "epoch": 0.60395, + "grad_norm": 1.3179453560302858, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60395 + }, + { + "epoch": 0.60396, + "grad_norm": 1.8008584641810064, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 60396 + }, + { + "epoch": 0.60397, + "grad_norm": 1.2527776244471183, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 60397 + }, + { + "epoch": 0.60398, + "grad_norm": 1.8317965940124026, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 60398 + }, + { + "epoch": 0.60399, + "grad_norm": 1.671955654441998, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 60399 + }, + { + "epoch": 0.604, + "grad_norm": 1.3866844314979991, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 60400 + }, + { + "epoch": 0.60401, + "grad_norm": 1.7120597754253595, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 60401 + }, + { + "epoch": 0.60402, + "grad_norm": 1.2729837360851195, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 60402 + }, + { + "epoch": 0.60403, + "grad_norm": 1.6894238744724348, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 60403 + }, + { + "epoch": 0.60404, + "grad_norm": 1.2474571382229478, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 60404 + }, + { + "epoch": 0.60405, + "grad_norm": 1.805910961912089, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 60405 + }, + { + "epoch": 0.60406, + "grad_norm": 1.4893079358351464, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 60406 + }, + { + "epoch": 0.60407, + "grad_norm": 1.7800382954639726, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 60407 + }, + { + "epoch": 0.60408, + "grad_norm": 1.2981340144930038, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 60408 + }, + { + "epoch": 0.60409, + "grad_norm": 1.9328597164718655, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 60409 + }, + { + "epoch": 0.6041, + "grad_norm": 1.5768855818685512, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 60410 + }, + { + "epoch": 0.60411, + "grad_norm": 1.5185777419495992, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 60411 + }, + { + "epoch": 0.60412, + "grad_norm": 1.6911394905593198, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 60412 + }, + { + "epoch": 0.60413, + "grad_norm": 1.43477848228579, + "learning_rate": 0.003, + "loss": 4.057, + "step": 60413 + }, + { + "epoch": 0.60414, + "grad_norm": 1.609646178613382, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 60414 + }, + { + "epoch": 0.60415, + "grad_norm": 1.5760527173268888, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 60415 + }, + { + "epoch": 0.60416, + "grad_norm": 1.586888697096598, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 60416 + }, + { + "epoch": 0.60417, + "grad_norm": 1.6127411789113761, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 60417 + }, + { + "epoch": 0.60418, + "grad_norm": 1.5062245029218806, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 60418 + }, + { + "epoch": 0.60419, + "grad_norm": 1.6657760661417051, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 60419 + }, + { + "epoch": 0.6042, + "grad_norm": 1.3240054607544063, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 60420 + }, + { + "epoch": 0.60421, + "grad_norm": 1.7191262212157499, + "learning_rate": 0.003, + "loss": 4.01, + "step": 60421 + }, + { + "epoch": 0.60422, + "grad_norm": 1.401757664385896, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 60422 + }, + { + "epoch": 0.60423, + "grad_norm": 2.1702091938336268, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 60423 + }, + { + "epoch": 0.60424, + "grad_norm": 1.126966921610662, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 60424 + }, + { + "epoch": 0.60425, + "grad_norm": 1.8132744395860338, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 60425 + }, + { + "epoch": 0.60426, + "grad_norm": 1.6508076881904108, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 60426 + }, + { + "epoch": 0.60427, + "grad_norm": 1.7862941695926344, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 60427 + }, + { + "epoch": 0.60428, + "grad_norm": 1.3105784066914052, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 60428 + }, + { + "epoch": 0.60429, + "grad_norm": 1.507935059332748, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 60429 + }, + { + "epoch": 0.6043, + "grad_norm": 1.6508135411132785, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 60430 + }, + { + "epoch": 0.60431, + "grad_norm": 1.2443461705056553, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 60431 + }, + { + "epoch": 0.60432, + "grad_norm": 1.7688479185484296, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 60432 + }, + { + "epoch": 0.60433, + "grad_norm": 1.4133850723807393, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 60433 + }, + { + "epoch": 0.60434, + "grad_norm": 1.2610881066310824, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 60434 + }, + { + "epoch": 0.60435, + "grad_norm": 1.8727909906769218, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 60435 + }, + { + "epoch": 0.60436, + "grad_norm": 1.7216904668910347, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 60436 + }, + { + "epoch": 0.60437, + "grad_norm": 1.6504623014366897, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 60437 + }, + { + "epoch": 0.60438, + "grad_norm": 2.173589149530071, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 60438 + }, + { + "epoch": 0.60439, + "grad_norm": 1.305124057179889, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 60439 + }, + { + "epoch": 0.6044, + "grad_norm": 1.6069646991679079, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 60440 + }, + { + "epoch": 0.60441, + "grad_norm": 1.4837523788010702, + "learning_rate": 0.003, + "loss": 4.054, + "step": 60441 + }, + { + "epoch": 0.60442, + "grad_norm": 1.866552624370429, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 60442 + }, + { + "epoch": 0.60443, + "grad_norm": 1.6996554131329529, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 60443 + }, + { + "epoch": 0.60444, + "grad_norm": 1.3118240742997496, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 60444 + }, + { + "epoch": 0.60445, + "grad_norm": 1.4013677124236665, + "learning_rate": 0.003, + "loss": 4.033, + "step": 60445 + }, + { + "epoch": 0.60446, + "grad_norm": 1.4320937008082026, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 60446 + }, + { + "epoch": 0.60447, + "grad_norm": 1.573563033598747, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 60447 + }, + { + "epoch": 0.60448, + "grad_norm": 1.3761186215429384, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 60448 + }, + { + "epoch": 0.60449, + "grad_norm": 1.729077604420574, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 60449 + }, + { + "epoch": 0.6045, + "grad_norm": 1.5762896472949335, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 60450 + }, + { + "epoch": 0.60451, + "grad_norm": 1.7454355175811902, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 60451 + }, + { + "epoch": 0.60452, + "grad_norm": 1.2193152683745194, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 60452 + }, + { + "epoch": 0.60453, + "grad_norm": 1.8310967871496036, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 60453 + }, + { + "epoch": 0.60454, + "grad_norm": 2.0554127033350027, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 60454 + }, + { + "epoch": 0.60455, + "grad_norm": 1.5364521110142912, + "learning_rate": 0.003, + "loss": 4.039, + "step": 60455 + }, + { + "epoch": 0.60456, + "grad_norm": 1.568593118146274, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 60456 + }, + { + "epoch": 0.60457, + "grad_norm": 1.3463928901784452, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 60457 + }, + { + "epoch": 0.60458, + "grad_norm": 1.6994182013239683, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 60458 + }, + { + "epoch": 0.60459, + "grad_norm": 1.4323678710802141, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 60459 + }, + { + "epoch": 0.6046, + "grad_norm": 1.62509652288702, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 60460 + }, + { + "epoch": 0.60461, + "grad_norm": 1.680258052299365, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 60461 + }, + { + "epoch": 0.60462, + "grad_norm": 1.6439839121644388, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 60462 + }, + { + "epoch": 0.60463, + "grad_norm": 1.5396473023555326, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 60463 + }, + { + "epoch": 0.60464, + "grad_norm": 1.4606802947003248, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 60464 + }, + { + "epoch": 0.60465, + "grad_norm": 1.4843471682550282, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 60465 + }, + { + "epoch": 0.60466, + "grad_norm": 1.5183905105266098, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 60466 + }, + { + "epoch": 0.60467, + "grad_norm": 1.6401978148435463, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 60467 + }, + { + "epoch": 0.60468, + "grad_norm": 1.510201701536288, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 60468 + }, + { + "epoch": 0.60469, + "grad_norm": 1.2972130344075419, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 60469 + }, + { + "epoch": 0.6047, + "grad_norm": 1.6230681686728181, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 60470 + }, + { + "epoch": 0.60471, + "grad_norm": 1.578022476447156, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 60471 + }, + { + "epoch": 0.60472, + "grad_norm": 1.6191247348772597, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 60472 + }, + { + "epoch": 0.60473, + "grad_norm": 1.5652392092770688, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 60473 + }, + { + "epoch": 0.60474, + "grad_norm": 1.3761419119008755, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 60474 + }, + { + "epoch": 0.60475, + "grad_norm": 1.631686246561932, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 60475 + }, + { + "epoch": 0.60476, + "grad_norm": 1.5047513376851933, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 60476 + }, + { + "epoch": 0.60477, + "grad_norm": 1.6579183285222525, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 60477 + }, + { + "epoch": 0.60478, + "grad_norm": 1.6206551262345361, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 60478 + }, + { + "epoch": 0.60479, + "grad_norm": 1.8007200564225476, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 60479 + }, + { + "epoch": 0.6048, + "grad_norm": 1.9740626092121951, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 60480 + }, + { + "epoch": 0.60481, + "grad_norm": 1.2480359656999755, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 60481 + }, + { + "epoch": 0.60482, + "grad_norm": 1.727569175285943, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 60482 + }, + { + "epoch": 0.60483, + "grad_norm": 1.632540825259016, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 60483 + }, + { + "epoch": 0.60484, + "grad_norm": 1.8198177749237228, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 60484 + }, + { + "epoch": 0.60485, + "grad_norm": 1.5027809539154822, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 60485 + }, + { + "epoch": 0.60486, + "grad_norm": 1.354625798587052, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 60486 + }, + { + "epoch": 0.60487, + "grad_norm": 1.519395906991894, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 60487 + }, + { + "epoch": 0.60488, + "grad_norm": 1.8466886382275578, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 60488 + }, + { + "epoch": 0.60489, + "grad_norm": 1.2746816657167233, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 60489 + }, + { + "epoch": 0.6049, + "grad_norm": 1.5972590486116074, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 60490 + }, + { + "epoch": 0.60491, + "grad_norm": 1.521168872483663, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 60491 + }, + { + "epoch": 0.60492, + "grad_norm": 1.5019231948898486, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 60492 + }, + { + "epoch": 0.60493, + "grad_norm": 1.4873049699013743, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 60493 + }, + { + "epoch": 0.60494, + "grad_norm": 1.4229946984166761, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 60494 + }, + { + "epoch": 0.60495, + "grad_norm": 1.6437801408094443, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 60495 + }, + { + "epoch": 0.60496, + "grad_norm": 1.6927302354144422, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 60496 + }, + { + "epoch": 0.60497, + "grad_norm": 1.4433246585088304, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 60497 + }, + { + "epoch": 0.60498, + "grad_norm": 1.778329225975521, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 60498 + }, + { + "epoch": 0.60499, + "grad_norm": 1.8196292487066892, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 60499 + }, + { + "epoch": 0.605, + "grad_norm": 1.528178994796705, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 60500 + }, + { + "epoch": 0.60501, + "grad_norm": 1.6285860376555652, + "learning_rate": 0.003, + "loss": 4.032, + "step": 60501 + }, + { + "epoch": 0.60502, + "grad_norm": 1.6289895408955315, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 60502 + }, + { + "epoch": 0.60503, + "grad_norm": 1.6234805707048774, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60503 + }, + { + "epoch": 0.60504, + "grad_norm": 1.4565076525590688, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 60504 + }, + { + "epoch": 0.60505, + "grad_norm": 1.41896151310583, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 60505 + }, + { + "epoch": 0.60506, + "grad_norm": 1.5966109355024214, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 60506 + }, + { + "epoch": 0.60507, + "grad_norm": 1.5063158702313513, + "learning_rate": 0.003, + "loss": 3.985, + "step": 60507 + }, + { + "epoch": 0.60508, + "grad_norm": 1.57803066810288, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 60508 + }, + { + "epoch": 0.60509, + "grad_norm": 1.30759673176069, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 60509 + }, + { + "epoch": 0.6051, + "grad_norm": 2.081344659481782, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 60510 + }, + { + "epoch": 0.60511, + "grad_norm": 1.119683555470446, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 60511 + }, + { + "epoch": 0.60512, + "grad_norm": 2.5121895703143946, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 60512 + }, + { + "epoch": 0.60513, + "grad_norm": 1.6128670699585412, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 60513 + }, + { + "epoch": 0.60514, + "grad_norm": 1.8767564742156424, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 60514 + }, + { + "epoch": 0.60515, + "grad_norm": 1.7020835268937784, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 60515 + }, + { + "epoch": 0.60516, + "grad_norm": 1.3680449972085589, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 60516 + }, + { + "epoch": 0.60517, + "grad_norm": 1.3357623430475851, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 60517 + }, + { + "epoch": 0.60518, + "grad_norm": 1.7643645689943064, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 60518 + }, + { + "epoch": 0.60519, + "grad_norm": 1.8736034685511531, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 60519 + }, + { + "epoch": 0.6052, + "grad_norm": 1.3839289027357196, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 60520 + }, + { + "epoch": 0.60521, + "grad_norm": 1.7468907529734437, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 60521 + }, + { + "epoch": 0.60522, + "grad_norm": 1.4509268682219827, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 60522 + }, + { + "epoch": 0.60523, + "grad_norm": 1.2752364600763568, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 60523 + }, + { + "epoch": 0.60524, + "grad_norm": 1.612766696119078, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 60524 + }, + { + "epoch": 0.60525, + "grad_norm": 1.8952779457542344, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 60525 + }, + { + "epoch": 0.60526, + "grad_norm": 1.2726413952929903, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 60526 + }, + { + "epoch": 0.60527, + "grad_norm": 1.7443929886784197, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 60527 + }, + { + "epoch": 0.60528, + "grad_norm": 1.30686306781476, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 60528 + }, + { + "epoch": 0.60529, + "grad_norm": 1.88844003662641, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 60529 + }, + { + "epoch": 0.6053, + "grad_norm": 1.177700719174732, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 60530 + }, + { + "epoch": 0.60531, + "grad_norm": 1.7797223803496387, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 60531 + }, + { + "epoch": 0.60532, + "grad_norm": 1.3903738149367588, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 60532 + }, + { + "epoch": 0.60533, + "grad_norm": 1.6571192194764928, + "learning_rate": 0.003, + "loss": 4.057, + "step": 60533 + }, + { + "epoch": 0.60534, + "grad_norm": 1.6284652007115492, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 60534 + }, + { + "epoch": 0.60535, + "grad_norm": 1.3648825648082044, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 60535 + }, + { + "epoch": 0.60536, + "grad_norm": 1.6711694677052056, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 60536 + }, + { + "epoch": 0.60537, + "grad_norm": 1.6020272191646643, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 60537 + }, + { + "epoch": 0.60538, + "grad_norm": 1.4215744997950817, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 60538 + }, + { + "epoch": 0.60539, + "grad_norm": 1.694753118965917, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 60539 + }, + { + "epoch": 0.6054, + "grad_norm": 1.6385087268893297, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 60540 + }, + { + "epoch": 0.60541, + "grad_norm": 1.9543760858471557, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 60541 + }, + { + "epoch": 0.60542, + "grad_norm": 1.191043694138369, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 60542 + }, + { + "epoch": 0.60543, + "grad_norm": 1.8956802363095837, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 60543 + }, + { + "epoch": 0.60544, + "grad_norm": 1.2547096137689744, + "learning_rate": 0.003, + "loss": 4.053, + "step": 60544 + }, + { + "epoch": 0.60545, + "grad_norm": 1.6691518769182232, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 60545 + }, + { + "epoch": 0.60546, + "grad_norm": 1.5135401269903592, + "learning_rate": 0.003, + "loss": 4.031, + "step": 60546 + }, + { + "epoch": 0.60547, + "grad_norm": 1.6739717202161568, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 60547 + }, + { + "epoch": 0.60548, + "grad_norm": 1.7366484047269974, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 60548 + }, + { + "epoch": 0.60549, + "grad_norm": 1.500569995417147, + "learning_rate": 0.003, + "loss": 4.039, + "step": 60549 + }, + { + "epoch": 0.6055, + "grad_norm": 1.640585014634771, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 60550 + }, + { + "epoch": 0.60551, + "grad_norm": 1.4187891716898848, + "learning_rate": 0.003, + "loss": 4.01, + "step": 60551 + }, + { + "epoch": 0.60552, + "grad_norm": 1.6448919935489594, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 60552 + }, + { + "epoch": 0.60553, + "grad_norm": 1.3865445124513986, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 60553 + }, + { + "epoch": 0.60554, + "grad_norm": 1.6439943459129913, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 60554 + }, + { + "epoch": 0.60555, + "grad_norm": 1.443672139564008, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 60555 + }, + { + "epoch": 0.60556, + "grad_norm": 1.5328139952124282, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 60556 + }, + { + "epoch": 0.60557, + "grad_norm": 1.6233296644877449, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 60557 + }, + { + "epoch": 0.60558, + "grad_norm": 1.541096564729103, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 60558 + }, + { + "epoch": 0.60559, + "grad_norm": 1.7432575138841768, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 60559 + }, + { + "epoch": 0.6056, + "grad_norm": 1.4653650812075931, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 60560 + }, + { + "epoch": 0.60561, + "grad_norm": 1.5006954069944214, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 60561 + }, + { + "epoch": 0.60562, + "grad_norm": 1.400918661203025, + "learning_rate": 0.003, + "loss": 4.024, + "step": 60562 + }, + { + "epoch": 0.60563, + "grad_norm": 1.707173467256706, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 60563 + }, + { + "epoch": 0.60564, + "grad_norm": 1.3830810912731402, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60564 + }, + { + "epoch": 0.60565, + "grad_norm": 1.4347215931773798, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 60565 + }, + { + "epoch": 0.60566, + "grad_norm": 1.5335543091299355, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 60566 + }, + { + "epoch": 0.60567, + "grad_norm": 1.4803497174435045, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 60567 + }, + { + "epoch": 0.60568, + "grad_norm": 2.086862397730219, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 60568 + }, + { + "epoch": 0.60569, + "grad_norm": 1.4636019476272388, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 60569 + }, + { + "epoch": 0.6057, + "grad_norm": 1.882825432652314, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 60570 + }, + { + "epoch": 0.60571, + "grad_norm": 1.552136506337553, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 60571 + }, + { + "epoch": 0.60572, + "grad_norm": 1.7114186467919732, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 60572 + }, + { + "epoch": 0.60573, + "grad_norm": 1.4529516877914361, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 60573 + }, + { + "epoch": 0.60574, + "grad_norm": 1.9937772336365516, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 60574 + }, + { + "epoch": 0.60575, + "grad_norm": 1.1766803992892632, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 60575 + }, + { + "epoch": 0.60576, + "grad_norm": 1.751145448100837, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 60576 + }, + { + "epoch": 0.60577, + "grad_norm": 1.4305235512084202, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 60577 + }, + { + "epoch": 0.60578, + "grad_norm": 1.511994014296872, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 60578 + }, + { + "epoch": 0.60579, + "grad_norm": 1.7590845709197325, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 60579 + }, + { + "epoch": 0.6058, + "grad_norm": 1.2669611350107906, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 60580 + }, + { + "epoch": 0.60581, + "grad_norm": 2.0150272354390655, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 60581 + }, + { + "epoch": 0.60582, + "grad_norm": 1.5086451105876477, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 60582 + }, + { + "epoch": 0.60583, + "grad_norm": 2.670289181961722, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 60583 + }, + { + "epoch": 0.60584, + "grad_norm": 1.616138005369312, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 60584 + }, + { + "epoch": 0.60585, + "grad_norm": 3.0564730807536122, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 60585 + }, + { + "epoch": 0.60586, + "grad_norm": 2.1893266034041314, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 60586 + }, + { + "epoch": 0.60587, + "grad_norm": 3.0236736755571467, + "learning_rate": 0.003, + "loss": 4.1061, + "step": 60587 + }, + { + "epoch": 0.60588, + "grad_norm": 1.4247472989143326, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 60588 + }, + { + "epoch": 0.60589, + "grad_norm": 1.9086449840720345, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 60589 + }, + { + "epoch": 0.6059, + "grad_norm": 1.8083428346180914, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 60590 + }, + { + "epoch": 0.60591, + "grad_norm": 1.8898600700090935, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 60591 + }, + { + "epoch": 0.60592, + "grad_norm": 1.8293399445864331, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 60592 + }, + { + "epoch": 0.60593, + "grad_norm": 1.6433168927542234, + "learning_rate": 0.003, + "loss": 4.067, + "step": 60593 + }, + { + "epoch": 0.60594, + "grad_norm": 1.8658929049604265, + "learning_rate": 0.003, + "loss": 4.079, + "step": 60594 + }, + { + "epoch": 0.60595, + "grad_norm": 1.5076372015598938, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 60595 + }, + { + "epoch": 0.60596, + "grad_norm": 1.6840271871322343, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 60596 + }, + { + "epoch": 0.60597, + "grad_norm": 1.7264792254419496, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 60597 + }, + { + "epoch": 0.60598, + "grad_norm": 2.135150168257741, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 60598 + }, + { + "epoch": 0.60599, + "grad_norm": 1.5079583887885686, + "learning_rate": 0.003, + "loss": 4.059, + "step": 60599 + }, + { + "epoch": 0.606, + "grad_norm": 1.6234477943835797, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 60600 + }, + { + "epoch": 0.60601, + "grad_norm": 1.8799877525799427, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 60601 + }, + { + "epoch": 0.60602, + "grad_norm": 1.9091619231800872, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 60602 + }, + { + "epoch": 0.60603, + "grad_norm": 1.4068749123144049, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 60603 + }, + { + "epoch": 0.60604, + "grad_norm": 1.9601203087825787, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 60604 + }, + { + "epoch": 0.60605, + "grad_norm": 1.469984435447686, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 60605 + }, + { + "epoch": 0.60606, + "grad_norm": 1.2770200317700227, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 60606 + }, + { + "epoch": 0.60607, + "grad_norm": 1.4760637846502664, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 60607 + }, + { + "epoch": 0.60608, + "grad_norm": 1.4037075267883918, + "learning_rate": 0.003, + "loss": 4.035, + "step": 60608 + }, + { + "epoch": 0.60609, + "grad_norm": 1.6063116277080485, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 60609 + }, + { + "epoch": 0.6061, + "grad_norm": 1.9746968566508702, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 60610 + }, + { + "epoch": 0.60611, + "grad_norm": 1.5370196771104265, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 60611 + }, + { + "epoch": 0.60612, + "grad_norm": 1.2384149920229892, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 60612 + }, + { + "epoch": 0.60613, + "grad_norm": 1.7416031687625253, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 60613 + }, + { + "epoch": 0.60614, + "grad_norm": 1.2800184962373833, + "learning_rate": 0.003, + "loss": 4.055, + "step": 60614 + }, + { + "epoch": 0.60615, + "grad_norm": 1.7777788168482707, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 60615 + }, + { + "epoch": 0.60616, + "grad_norm": 1.402341364653781, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 60616 + }, + { + "epoch": 0.60617, + "grad_norm": 2.021383174853676, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 60617 + }, + { + "epoch": 0.60618, + "grad_norm": 1.3922384208804912, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 60618 + }, + { + "epoch": 0.60619, + "grad_norm": 1.772402509510021, + "learning_rate": 0.003, + "loss": 4.085, + "step": 60619 + }, + { + "epoch": 0.6062, + "grad_norm": 1.5107302440786148, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 60620 + }, + { + "epoch": 0.60621, + "grad_norm": 1.598292682993857, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 60621 + }, + { + "epoch": 0.60622, + "grad_norm": 1.7448609211036024, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 60622 + }, + { + "epoch": 0.60623, + "grad_norm": 1.7751622062892505, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 60623 + }, + { + "epoch": 0.60624, + "grad_norm": 1.1873765817016428, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 60624 + }, + { + "epoch": 0.60625, + "grad_norm": 1.4456871392100734, + "learning_rate": 0.003, + "loss": 4.025, + "step": 60625 + }, + { + "epoch": 0.60626, + "grad_norm": 1.5112162745735733, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 60626 + }, + { + "epoch": 0.60627, + "grad_norm": 1.2781288520499052, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 60627 + }, + { + "epoch": 0.60628, + "grad_norm": 1.7540368853788242, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 60628 + }, + { + "epoch": 0.60629, + "grad_norm": 1.42817972599088, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 60629 + }, + { + "epoch": 0.6063, + "grad_norm": 1.933005420043289, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 60630 + }, + { + "epoch": 0.60631, + "grad_norm": 1.6913980439338039, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 60631 + }, + { + "epoch": 0.60632, + "grad_norm": 1.3312070859650649, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 60632 + }, + { + "epoch": 0.60633, + "grad_norm": 1.9093619250180225, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 60633 + }, + { + "epoch": 0.60634, + "grad_norm": 1.4525548616592459, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 60634 + }, + { + "epoch": 0.60635, + "grad_norm": 1.2911333684338664, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 60635 + }, + { + "epoch": 0.60636, + "grad_norm": 1.61371157323621, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 60636 + }, + { + "epoch": 0.60637, + "grad_norm": 1.2854524869649717, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 60637 + }, + { + "epoch": 0.60638, + "grad_norm": 1.7038259445652315, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 60638 + }, + { + "epoch": 0.60639, + "grad_norm": 1.2300502206742387, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 60639 + }, + { + "epoch": 0.6064, + "grad_norm": 2.036160406558273, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 60640 + }, + { + "epoch": 0.60641, + "grad_norm": 1.3981972784271182, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 60641 + }, + { + "epoch": 0.60642, + "grad_norm": 1.4580993254527903, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 60642 + }, + { + "epoch": 0.60643, + "grad_norm": 1.541718422273681, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 60643 + }, + { + "epoch": 0.60644, + "grad_norm": 1.5067211813913086, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 60644 + }, + { + "epoch": 0.60645, + "grad_norm": 1.8187068213022164, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 60645 + }, + { + "epoch": 0.60646, + "grad_norm": 2.0768167252787646, + "learning_rate": 0.003, + "loss": 4.044, + "step": 60646 + }, + { + "epoch": 0.60647, + "grad_norm": 1.5741434524100122, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 60647 + }, + { + "epoch": 0.60648, + "grad_norm": 1.3538007925786928, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 60648 + }, + { + "epoch": 0.60649, + "grad_norm": 1.4336777153097287, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 60649 + }, + { + "epoch": 0.6065, + "grad_norm": 1.7194205433816199, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 60650 + }, + { + "epoch": 0.60651, + "grad_norm": 1.4534550673110072, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 60651 + }, + { + "epoch": 0.60652, + "grad_norm": 1.6096269804470018, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 60652 + }, + { + "epoch": 0.60653, + "grad_norm": 1.6659255560034374, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 60653 + }, + { + "epoch": 0.60654, + "grad_norm": 1.4113737173490413, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 60654 + }, + { + "epoch": 0.60655, + "grad_norm": 1.7566060455516448, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 60655 + }, + { + "epoch": 0.60656, + "grad_norm": 1.536364303729078, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 60656 + }, + { + "epoch": 0.60657, + "grad_norm": 1.3211602022025049, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 60657 + }, + { + "epoch": 0.60658, + "grad_norm": 1.6259533625131437, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 60658 + }, + { + "epoch": 0.60659, + "grad_norm": 1.5103966322292113, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 60659 + }, + { + "epoch": 0.6066, + "grad_norm": 1.6268780364176074, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 60660 + }, + { + "epoch": 0.60661, + "grad_norm": 1.2885647459897323, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 60661 + }, + { + "epoch": 0.60662, + "grad_norm": 1.7678781856732135, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 60662 + }, + { + "epoch": 0.60663, + "grad_norm": 1.1382046810168192, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 60663 + }, + { + "epoch": 0.60664, + "grad_norm": 1.5974559536553452, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 60664 + }, + { + "epoch": 0.60665, + "grad_norm": 1.3769381602578048, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 60665 + }, + { + "epoch": 0.60666, + "grad_norm": 1.9265495878952565, + "learning_rate": 0.003, + "loss": 4.067, + "step": 60666 + }, + { + "epoch": 0.60667, + "grad_norm": 1.5086593521630653, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 60667 + }, + { + "epoch": 0.60668, + "grad_norm": 1.892744355445419, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 60668 + }, + { + "epoch": 0.60669, + "grad_norm": 1.5427315068387293, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 60669 + }, + { + "epoch": 0.6067, + "grad_norm": 1.5439604438980803, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 60670 + }, + { + "epoch": 0.60671, + "grad_norm": 1.4503192271443184, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 60671 + }, + { + "epoch": 0.60672, + "grad_norm": 1.6525511245997615, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 60672 + }, + { + "epoch": 0.60673, + "grad_norm": 1.4987977311782277, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 60673 + }, + { + "epoch": 0.60674, + "grad_norm": 1.3341239105408924, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 60674 + }, + { + "epoch": 0.60675, + "grad_norm": 1.7664155830360166, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 60675 + }, + { + "epoch": 0.60676, + "grad_norm": 1.6549683851983497, + "learning_rate": 0.003, + "loss": 4.024, + "step": 60676 + }, + { + "epoch": 0.60677, + "grad_norm": 1.5585126389934687, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 60677 + }, + { + "epoch": 0.60678, + "grad_norm": 1.5276114165275025, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 60678 + }, + { + "epoch": 0.60679, + "grad_norm": 1.7821478500145977, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 60679 + }, + { + "epoch": 0.6068, + "grad_norm": 1.0825372144093695, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 60680 + }, + { + "epoch": 0.60681, + "grad_norm": 1.521807052664329, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 60681 + }, + { + "epoch": 0.60682, + "grad_norm": 1.810875052763839, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 60682 + }, + { + "epoch": 0.60683, + "grad_norm": 1.3940537235817538, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 60683 + }, + { + "epoch": 0.60684, + "grad_norm": 1.5446531715116578, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 60684 + }, + { + "epoch": 0.60685, + "grad_norm": 1.4415718601065481, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 60685 + }, + { + "epoch": 0.60686, + "grad_norm": 1.539012668494897, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 60686 + }, + { + "epoch": 0.60687, + "grad_norm": 1.6049084917663616, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 60687 + }, + { + "epoch": 0.60688, + "grad_norm": 1.4568569559830526, + "learning_rate": 0.003, + "loss": 4.036, + "step": 60688 + }, + { + "epoch": 0.60689, + "grad_norm": 1.3514432911511478, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 60689 + }, + { + "epoch": 0.6069, + "grad_norm": 1.4551586989109089, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 60690 + }, + { + "epoch": 0.60691, + "grad_norm": 1.5444653035559324, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 60691 + }, + { + "epoch": 0.60692, + "grad_norm": 1.3537439219369694, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 60692 + }, + { + "epoch": 0.60693, + "grad_norm": 1.83104607775455, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 60693 + }, + { + "epoch": 0.60694, + "grad_norm": 1.4316231843682974, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 60694 + }, + { + "epoch": 0.60695, + "grad_norm": 1.7050922740745518, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 60695 + }, + { + "epoch": 0.60696, + "grad_norm": 1.6052793477541756, + "learning_rate": 0.003, + "loss": 4.028, + "step": 60696 + }, + { + "epoch": 0.60697, + "grad_norm": 1.4357460257704484, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 60697 + }, + { + "epoch": 0.60698, + "grad_norm": 1.9781204517334297, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 60698 + }, + { + "epoch": 0.60699, + "grad_norm": 1.4090567064595476, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 60699 + }, + { + "epoch": 0.607, + "grad_norm": 1.5462461175901294, + "learning_rate": 0.003, + "loss": 4.045, + "step": 60700 + }, + { + "epoch": 0.60701, + "grad_norm": 1.5603869837835866, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 60701 + }, + { + "epoch": 0.60702, + "grad_norm": 1.2893984397043237, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 60702 + }, + { + "epoch": 0.60703, + "grad_norm": 1.5705412235093317, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 60703 + }, + { + "epoch": 0.60704, + "grad_norm": 1.5004003818133664, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 60704 + }, + { + "epoch": 0.60705, + "grad_norm": 1.4551641558045614, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 60705 + }, + { + "epoch": 0.60706, + "grad_norm": 1.6680723043511152, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 60706 + }, + { + "epoch": 0.60707, + "grad_norm": 1.5442139213200972, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 60707 + }, + { + "epoch": 0.60708, + "grad_norm": 1.7112298794958014, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 60708 + }, + { + "epoch": 0.60709, + "grad_norm": 1.616140290365925, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 60709 + }, + { + "epoch": 0.6071, + "grad_norm": 1.4681329852320144, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 60710 + }, + { + "epoch": 0.60711, + "grad_norm": 1.3025992261707429, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 60711 + }, + { + "epoch": 0.60712, + "grad_norm": 1.7141817842304556, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 60712 + }, + { + "epoch": 0.60713, + "grad_norm": 1.3040808423141665, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 60713 + }, + { + "epoch": 0.60714, + "grad_norm": 1.724172982650295, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 60714 + }, + { + "epoch": 0.60715, + "grad_norm": 1.6763492322646396, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 60715 + }, + { + "epoch": 0.60716, + "grad_norm": 1.544281486286585, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 60716 + }, + { + "epoch": 0.60717, + "grad_norm": 1.651726599473654, + "learning_rate": 0.003, + "loss": 4.022, + "step": 60717 + }, + { + "epoch": 0.60718, + "grad_norm": 1.676507730330943, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 60718 + }, + { + "epoch": 0.60719, + "grad_norm": 1.421537663828435, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 60719 + }, + { + "epoch": 0.6072, + "grad_norm": 1.7360376308923695, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 60720 + }, + { + "epoch": 0.60721, + "grad_norm": 1.17369574756414, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 60721 + }, + { + "epoch": 0.60722, + "grad_norm": 1.8067218627646215, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 60722 + }, + { + "epoch": 0.60723, + "grad_norm": 1.3548444547126046, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 60723 + }, + { + "epoch": 0.60724, + "grad_norm": 1.773291929065965, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 60724 + }, + { + "epoch": 0.60725, + "grad_norm": 1.4637609281764579, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 60725 + }, + { + "epoch": 0.60726, + "grad_norm": 1.5002309332151806, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 60726 + }, + { + "epoch": 0.60727, + "grad_norm": 1.6186213204076678, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 60727 + }, + { + "epoch": 0.60728, + "grad_norm": 1.740899168444796, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 60728 + }, + { + "epoch": 0.60729, + "grad_norm": 1.6443657112741714, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 60729 + }, + { + "epoch": 0.6073, + "grad_norm": 1.4895296659425568, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 60730 + }, + { + "epoch": 0.60731, + "grad_norm": 1.4030874337976964, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 60731 + }, + { + "epoch": 0.60732, + "grad_norm": 1.62297086907696, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 60732 + }, + { + "epoch": 0.60733, + "grad_norm": 1.4812429649797911, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 60733 + }, + { + "epoch": 0.60734, + "grad_norm": 1.362105129281826, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 60734 + }, + { + "epoch": 0.60735, + "grad_norm": 1.9502849434551524, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 60735 + }, + { + "epoch": 0.60736, + "grad_norm": 1.1331047279282451, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 60736 + }, + { + "epoch": 0.60737, + "grad_norm": 1.6390205715274695, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 60737 + }, + { + "epoch": 0.60738, + "grad_norm": 1.7848332024012596, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 60738 + }, + { + "epoch": 0.60739, + "grad_norm": 1.6903359863542033, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 60739 + }, + { + "epoch": 0.6074, + "grad_norm": 1.15369261304015, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 60740 + }, + { + "epoch": 0.60741, + "grad_norm": 1.7759273044178807, + "learning_rate": 0.003, + "loss": 4.054, + "step": 60741 + }, + { + "epoch": 0.60742, + "grad_norm": 1.2922796312094917, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 60742 + }, + { + "epoch": 0.60743, + "grad_norm": 1.5371381891019822, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 60743 + }, + { + "epoch": 0.60744, + "grad_norm": 1.4801588965843193, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 60744 + }, + { + "epoch": 0.60745, + "grad_norm": 1.9158273397365977, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 60745 + }, + { + "epoch": 0.60746, + "grad_norm": 1.436880914472193, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 60746 + }, + { + "epoch": 0.60747, + "grad_norm": 1.5748313897925432, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 60747 + }, + { + "epoch": 0.60748, + "grad_norm": 1.5666814293784082, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 60748 + }, + { + "epoch": 0.60749, + "grad_norm": 1.421722997800244, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 60749 + }, + { + "epoch": 0.6075, + "grad_norm": 1.530406553470381, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 60750 + }, + { + "epoch": 0.60751, + "grad_norm": 1.3800791547977536, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 60751 + }, + { + "epoch": 0.60752, + "grad_norm": 1.7698693678556259, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 60752 + }, + { + "epoch": 0.60753, + "grad_norm": 1.469711885680388, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 60753 + }, + { + "epoch": 0.60754, + "grad_norm": 1.8223608068653052, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 60754 + }, + { + "epoch": 0.60755, + "grad_norm": 1.927214326005512, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 60755 + }, + { + "epoch": 0.60756, + "grad_norm": 1.4533727170655508, + "learning_rate": 0.003, + "loss": 4.036, + "step": 60756 + }, + { + "epoch": 0.60757, + "grad_norm": 1.6399794207585252, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 60757 + }, + { + "epoch": 0.60758, + "grad_norm": 1.3190463214941774, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 60758 + }, + { + "epoch": 0.60759, + "grad_norm": 1.4984268948367778, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 60759 + }, + { + "epoch": 0.6076, + "grad_norm": 1.7608595816673631, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 60760 + }, + { + "epoch": 0.60761, + "grad_norm": 1.4943913109641547, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 60761 + }, + { + "epoch": 0.60762, + "grad_norm": 1.6071279304046568, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 60762 + }, + { + "epoch": 0.60763, + "grad_norm": 1.5535062832102653, + "learning_rate": 0.003, + "loss": 4.005, + "step": 60763 + }, + { + "epoch": 0.60764, + "grad_norm": 1.404822582725817, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 60764 + }, + { + "epoch": 0.60765, + "grad_norm": 1.6879296410452607, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 60765 + }, + { + "epoch": 0.60766, + "grad_norm": 1.5985435618149724, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 60766 + }, + { + "epoch": 0.60767, + "grad_norm": 1.7288132481475516, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 60767 + }, + { + "epoch": 0.60768, + "grad_norm": 1.4933224752649306, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 60768 + }, + { + "epoch": 0.60769, + "grad_norm": 1.6302067423307833, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 60769 + }, + { + "epoch": 0.6077, + "grad_norm": 1.3812229601131727, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 60770 + }, + { + "epoch": 0.60771, + "grad_norm": 1.5088804760808971, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 60771 + }, + { + "epoch": 0.60772, + "grad_norm": 1.5998784213973487, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 60772 + }, + { + "epoch": 0.60773, + "grad_norm": 1.2846960468675883, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 60773 + }, + { + "epoch": 0.60774, + "grad_norm": 1.6604101223844812, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 60774 + }, + { + "epoch": 0.60775, + "grad_norm": 1.4438048122885219, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 60775 + }, + { + "epoch": 0.60776, + "grad_norm": 1.6839419103323068, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 60776 + }, + { + "epoch": 0.60777, + "grad_norm": 1.4499116048837786, + "learning_rate": 0.003, + "loss": 4.017, + "step": 60777 + }, + { + "epoch": 0.60778, + "grad_norm": 1.6350449544647245, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 60778 + }, + { + "epoch": 0.60779, + "grad_norm": 1.237758496032182, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 60779 + }, + { + "epoch": 0.6078, + "grad_norm": 1.577100966341933, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 60780 + }, + { + "epoch": 0.60781, + "grad_norm": 1.6376612739647494, + "learning_rate": 0.003, + "loss": 4.023, + "step": 60781 + }, + { + "epoch": 0.60782, + "grad_norm": 1.5322251824418172, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 60782 + }, + { + "epoch": 0.60783, + "grad_norm": 1.4371735277945195, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 60783 + }, + { + "epoch": 0.60784, + "grad_norm": 1.6091076882335413, + "learning_rate": 0.003, + "loss": 4.033, + "step": 60784 + }, + { + "epoch": 0.60785, + "grad_norm": 1.4876190885837226, + "learning_rate": 0.003, + "loss": 4.016, + "step": 60785 + }, + { + "epoch": 0.60786, + "grad_norm": 1.5759028859258395, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 60786 + }, + { + "epoch": 0.60787, + "grad_norm": 1.5259233422685277, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 60787 + }, + { + "epoch": 0.60788, + "grad_norm": 1.6683200462194925, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 60788 + }, + { + "epoch": 0.60789, + "grad_norm": 1.6160397319833277, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 60789 + }, + { + "epoch": 0.6079, + "grad_norm": 1.3433119443735095, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 60790 + }, + { + "epoch": 0.60791, + "grad_norm": 1.6612603656419835, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 60791 + }, + { + "epoch": 0.60792, + "grad_norm": 1.530069861275774, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 60792 + }, + { + "epoch": 0.60793, + "grad_norm": 1.5837855018759777, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 60793 + }, + { + "epoch": 0.60794, + "grad_norm": 1.6332418529616382, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 60794 + }, + { + "epoch": 0.60795, + "grad_norm": 1.4202238122427466, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 60795 + }, + { + "epoch": 0.60796, + "grad_norm": 1.9306923509459588, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 60796 + }, + { + "epoch": 0.60797, + "grad_norm": 1.2501782556402325, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 60797 + }, + { + "epoch": 0.60798, + "grad_norm": 1.6421026553541287, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 60798 + }, + { + "epoch": 0.60799, + "grad_norm": 1.5452479099421366, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 60799 + }, + { + "epoch": 0.608, + "grad_norm": 1.605608487524007, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 60800 + }, + { + "epoch": 0.60801, + "grad_norm": 1.614265109231178, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 60801 + }, + { + "epoch": 0.60802, + "grad_norm": 2.138802014152242, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 60802 + }, + { + "epoch": 0.60803, + "grad_norm": 1.458126109466291, + "learning_rate": 0.003, + "loss": 4.022, + "step": 60803 + }, + { + "epoch": 0.60804, + "grad_norm": 1.4784561901632316, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 60804 + }, + { + "epoch": 0.60805, + "grad_norm": 1.823406290336125, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 60805 + }, + { + "epoch": 0.60806, + "grad_norm": 1.4162240742305434, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 60806 + }, + { + "epoch": 0.60807, + "grad_norm": 1.896743806921869, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 60807 + }, + { + "epoch": 0.60808, + "grad_norm": 1.5055708509745467, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 60808 + }, + { + "epoch": 0.60809, + "grad_norm": 1.708971081445693, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 60809 + }, + { + "epoch": 0.6081, + "grad_norm": 1.4844050288893578, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 60810 + }, + { + "epoch": 0.60811, + "grad_norm": 1.6520546412325439, + "learning_rate": 0.003, + "loss": 3.967, + "step": 60811 + }, + { + "epoch": 0.60812, + "grad_norm": 1.7344448047636627, + "learning_rate": 0.003, + "loss": 4.036, + "step": 60812 + }, + { + "epoch": 0.60813, + "grad_norm": 1.455155644517016, + "learning_rate": 0.003, + "loss": 4.027, + "step": 60813 + }, + { + "epoch": 0.60814, + "grad_norm": 1.3443803814819286, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 60814 + }, + { + "epoch": 0.60815, + "grad_norm": 1.4704923400617986, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 60815 + }, + { + "epoch": 0.60816, + "grad_norm": 1.9959539667707167, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 60816 + }, + { + "epoch": 0.60817, + "grad_norm": 1.3511011740302687, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 60817 + }, + { + "epoch": 0.60818, + "grad_norm": 1.7696234038772225, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 60818 + }, + { + "epoch": 0.60819, + "grad_norm": 1.6119706301422834, + "learning_rate": 0.003, + "loss": 4.022, + "step": 60819 + }, + { + "epoch": 0.6082, + "grad_norm": 1.6812823262149454, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 60820 + }, + { + "epoch": 0.60821, + "grad_norm": 1.457932596429411, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 60821 + }, + { + "epoch": 0.60822, + "grad_norm": 1.507517066561633, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 60822 + }, + { + "epoch": 0.60823, + "grad_norm": 1.4692088109211676, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 60823 + }, + { + "epoch": 0.60824, + "grad_norm": 1.4555024137748647, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 60824 + }, + { + "epoch": 0.60825, + "grad_norm": 1.4914278729545698, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 60825 + }, + { + "epoch": 0.60826, + "grad_norm": 1.6074501552558986, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 60826 + }, + { + "epoch": 0.60827, + "grad_norm": 1.5227441947754392, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 60827 + }, + { + "epoch": 0.60828, + "grad_norm": 1.3478508950336883, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 60828 + }, + { + "epoch": 0.60829, + "grad_norm": 1.5208047650517857, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 60829 + }, + { + "epoch": 0.6083, + "grad_norm": 1.6258613709907495, + "learning_rate": 0.003, + "loss": 4.028, + "step": 60830 + }, + { + "epoch": 0.60831, + "grad_norm": 1.7599529202151358, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 60831 + }, + { + "epoch": 0.60832, + "grad_norm": 1.5292177247271541, + "learning_rate": 0.003, + "loss": 4.037, + "step": 60832 + }, + { + "epoch": 0.60833, + "grad_norm": 1.5634144612303982, + "learning_rate": 0.003, + "loss": 4.03, + "step": 60833 + }, + { + "epoch": 0.60834, + "grad_norm": 1.593881954988526, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 60834 + }, + { + "epoch": 0.60835, + "grad_norm": 2.054687799474015, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 60835 + }, + { + "epoch": 0.60836, + "grad_norm": 1.4538739598732588, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 60836 + }, + { + "epoch": 0.60837, + "grad_norm": 1.7290509266514364, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 60837 + }, + { + "epoch": 0.60838, + "grad_norm": 1.4045263851184078, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 60838 + }, + { + "epoch": 0.60839, + "grad_norm": 1.8555735024576994, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 60839 + }, + { + "epoch": 0.6084, + "grad_norm": 1.4269624980075133, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 60840 + }, + { + "epoch": 0.60841, + "grad_norm": 1.5771250304429847, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 60841 + }, + { + "epoch": 0.60842, + "grad_norm": 1.4673231322810412, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 60842 + }, + { + "epoch": 0.60843, + "grad_norm": 1.794659090209894, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 60843 + }, + { + "epoch": 0.60844, + "grad_norm": 1.6674720843421464, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 60844 + }, + { + "epoch": 0.60845, + "grad_norm": 1.4756882453407403, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 60845 + }, + { + "epoch": 0.60846, + "grad_norm": 1.7467317799468836, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 60846 + }, + { + "epoch": 0.60847, + "grad_norm": 1.458137186412003, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 60847 + }, + { + "epoch": 0.60848, + "grad_norm": 1.4692661169731553, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 60848 + }, + { + "epoch": 0.60849, + "grad_norm": 1.5305456113208802, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 60849 + }, + { + "epoch": 0.6085, + "grad_norm": 1.4718082420513343, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 60850 + }, + { + "epoch": 0.60851, + "grad_norm": 1.9081725757091208, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 60851 + }, + { + "epoch": 0.60852, + "grad_norm": 1.5022700190951987, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 60852 + }, + { + "epoch": 0.60853, + "grad_norm": 1.4726350546598155, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 60853 + }, + { + "epoch": 0.60854, + "grad_norm": 1.7102437422962844, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 60854 + }, + { + "epoch": 0.60855, + "grad_norm": 1.599587741592764, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 60855 + }, + { + "epoch": 0.60856, + "grad_norm": 1.5471167785993816, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 60856 + }, + { + "epoch": 0.60857, + "grad_norm": 1.5571451451924778, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 60857 + }, + { + "epoch": 0.60858, + "grad_norm": 1.5183520611496677, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 60858 + }, + { + "epoch": 0.60859, + "grad_norm": 1.590744757341263, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 60859 + }, + { + "epoch": 0.6086, + "grad_norm": 1.5516394747487166, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 60860 + }, + { + "epoch": 0.60861, + "grad_norm": 1.7377417720558546, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 60861 + }, + { + "epoch": 0.60862, + "grad_norm": 1.8430107685497394, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 60862 + }, + { + "epoch": 0.60863, + "grad_norm": 1.269098643616746, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 60863 + }, + { + "epoch": 0.60864, + "grad_norm": 1.565425467178809, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 60864 + }, + { + "epoch": 0.60865, + "grad_norm": 1.509994345870754, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 60865 + }, + { + "epoch": 0.60866, + "grad_norm": 1.5289303145756323, + "learning_rate": 0.003, + "loss": 4.037, + "step": 60866 + }, + { + "epoch": 0.60867, + "grad_norm": 1.577593367446125, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 60867 + }, + { + "epoch": 0.60868, + "grad_norm": 1.5468575420712067, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 60868 + }, + { + "epoch": 0.60869, + "grad_norm": 1.6426691118913113, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 60869 + }, + { + "epoch": 0.6087, + "grad_norm": 1.550014456690369, + "learning_rate": 0.003, + "loss": 4.071, + "step": 60870 + }, + { + "epoch": 0.60871, + "grad_norm": 1.6729224337370696, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 60871 + }, + { + "epoch": 0.60872, + "grad_norm": 1.4024683400137143, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 60872 + }, + { + "epoch": 0.60873, + "grad_norm": 1.79655967632854, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 60873 + }, + { + "epoch": 0.60874, + "grad_norm": 1.5974234929872744, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 60874 + }, + { + "epoch": 0.60875, + "grad_norm": 1.8830999302740061, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 60875 + }, + { + "epoch": 0.60876, + "grad_norm": 1.603334202469924, + "learning_rate": 0.003, + "loss": 4.048, + "step": 60876 + }, + { + "epoch": 0.60877, + "grad_norm": 1.7736612499218913, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 60877 + }, + { + "epoch": 0.60878, + "grad_norm": 1.6965196870208394, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 60878 + }, + { + "epoch": 0.60879, + "grad_norm": 1.3812217414492205, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 60879 + }, + { + "epoch": 0.6088, + "grad_norm": 1.6317109911369347, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 60880 + }, + { + "epoch": 0.60881, + "grad_norm": 1.5947573145732639, + "learning_rate": 0.003, + "loss": 4.007, + "step": 60881 + }, + { + "epoch": 0.60882, + "grad_norm": 1.4735919721903772, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 60882 + }, + { + "epoch": 0.60883, + "grad_norm": 1.819066202081932, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 60883 + }, + { + "epoch": 0.60884, + "grad_norm": 1.3639913867422575, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 60884 + }, + { + "epoch": 0.60885, + "grad_norm": 1.6362828185204168, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 60885 + }, + { + "epoch": 0.60886, + "grad_norm": 1.243171477717963, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 60886 + }, + { + "epoch": 0.60887, + "grad_norm": 2.069639011192132, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 60887 + }, + { + "epoch": 0.60888, + "grad_norm": 1.4108067604410708, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 60888 + }, + { + "epoch": 0.60889, + "grad_norm": 1.8995006296936687, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 60889 + }, + { + "epoch": 0.6089, + "grad_norm": 1.8459013650927878, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 60890 + }, + { + "epoch": 0.60891, + "grad_norm": 1.384519155099402, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 60891 + }, + { + "epoch": 0.60892, + "grad_norm": 1.6062565663400934, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 60892 + }, + { + "epoch": 0.60893, + "grad_norm": 1.6587759395985788, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 60893 + }, + { + "epoch": 0.60894, + "grad_norm": 1.3905036290181898, + "learning_rate": 0.003, + "loss": 4.014, + "step": 60894 + }, + { + "epoch": 0.60895, + "grad_norm": 1.6229142695817014, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 60895 + }, + { + "epoch": 0.60896, + "grad_norm": 1.535139690763007, + "learning_rate": 0.003, + "loss": 4.045, + "step": 60896 + }, + { + "epoch": 0.60897, + "grad_norm": 1.4704836842760887, + "learning_rate": 0.003, + "loss": 4.042, + "step": 60897 + }, + { + "epoch": 0.60898, + "grad_norm": 1.5142983594143307, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 60898 + }, + { + "epoch": 0.60899, + "grad_norm": 1.505861796745128, + "learning_rate": 0.003, + "loss": 4.04, + "step": 60899 + }, + { + "epoch": 0.609, + "grad_norm": 1.5501423011480646, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 60900 + }, + { + "epoch": 0.60901, + "grad_norm": 1.3486350563744771, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 60901 + }, + { + "epoch": 0.60902, + "grad_norm": 1.6904357039561313, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 60902 + }, + { + "epoch": 0.60903, + "grad_norm": 1.7556072045440507, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 60903 + }, + { + "epoch": 0.60904, + "grad_norm": 1.4231721766984142, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 60904 + }, + { + "epoch": 0.60905, + "grad_norm": 1.6852851231402908, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 60905 + }, + { + "epoch": 0.60906, + "grad_norm": 1.267784545631039, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 60906 + }, + { + "epoch": 0.60907, + "grad_norm": 1.97530969393408, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 60907 + }, + { + "epoch": 0.60908, + "grad_norm": 1.835252585515544, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 60908 + }, + { + "epoch": 0.60909, + "grad_norm": 1.565407893849739, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 60909 + }, + { + "epoch": 0.6091, + "grad_norm": 1.5872838471522266, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 60910 + }, + { + "epoch": 0.60911, + "grad_norm": 1.7033833068329411, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 60911 + }, + { + "epoch": 0.60912, + "grad_norm": 1.1567079856797886, + "learning_rate": 0.003, + "loss": 4.0, + "step": 60912 + }, + { + "epoch": 0.60913, + "grad_norm": 1.8712795558250745, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 60913 + }, + { + "epoch": 0.60914, + "grad_norm": 1.137464079306673, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 60914 + }, + { + "epoch": 0.60915, + "grad_norm": 1.9678747375430674, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 60915 + }, + { + "epoch": 0.60916, + "grad_norm": 1.327546769573824, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 60916 + }, + { + "epoch": 0.60917, + "grad_norm": 2.1028919312795185, + "learning_rate": 0.003, + "loss": 4.041, + "step": 60917 + }, + { + "epoch": 0.60918, + "grad_norm": 1.628864780692834, + "learning_rate": 0.003, + "loss": 4.05, + "step": 60918 + }, + { + "epoch": 0.60919, + "grad_norm": 1.7506590860201539, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 60919 + }, + { + "epoch": 0.6092, + "grad_norm": 1.5926853966730536, + "learning_rate": 0.003, + "loss": 4.03, + "step": 60920 + }, + { + "epoch": 0.60921, + "grad_norm": 1.456340634697302, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 60921 + }, + { + "epoch": 0.60922, + "grad_norm": 1.705504281732269, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 60922 + }, + { + "epoch": 0.60923, + "grad_norm": 1.3819858258611202, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 60923 + }, + { + "epoch": 0.60924, + "grad_norm": 1.5589833234069967, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 60924 + }, + { + "epoch": 0.60925, + "grad_norm": 1.6726970381853137, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 60925 + }, + { + "epoch": 0.60926, + "grad_norm": 1.8010944828505737, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 60926 + }, + { + "epoch": 0.60927, + "grad_norm": 1.3484986645722534, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 60927 + }, + { + "epoch": 0.60928, + "grad_norm": 1.3854580301797845, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 60928 + }, + { + "epoch": 0.60929, + "grad_norm": 1.4453822612902103, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 60929 + }, + { + "epoch": 0.6093, + "grad_norm": 1.623792870340504, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 60930 + }, + { + "epoch": 0.60931, + "grad_norm": 1.613729549896835, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 60931 + }, + { + "epoch": 0.60932, + "grad_norm": 1.9624744503664784, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 60932 + }, + { + "epoch": 0.60933, + "grad_norm": 1.2775361673169476, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 60933 + }, + { + "epoch": 0.60934, + "grad_norm": 1.829880139748238, + "learning_rate": 0.003, + "loss": 4.054, + "step": 60934 + }, + { + "epoch": 0.60935, + "grad_norm": 1.4894717882511888, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 60935 + }, + { + "epoch": 0.60936, + "grad_norm": 1.6179799628583487, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 60936 + }, + { + "epoch": 0.60937, + "grad_norm": 1.436919058061551, + "learning_rate": 0.003, + "loss": 4.052, + "step": 60937 + }, + { + "epoch": 0.60938, + "grad_norm": 1.6543219351763312, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 60938 + }, + { + "epoch": 0.60939, + "grad_norm": 1.366732639033105, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 60939 + }, + { + "epoch": 0.6094, + "grad_norm": 1.8530465842632757, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 60940 + }, + { + "epoch": 0.60941, + "grad_norm": 1.4992258083702015, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 60941 + }, + { + "epoch": 0.60942, + "grad_norm": 1.6984822114597642, + "learning_rate": 0.003, + "loss": 4.02, + "step": 60942 + }, + { + "epoch": 0.60943, + "grad_norm": 1.3927380237011266, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 60943 + }, + { + "epoch": 0.60944, + "grad_norm": 1.7639972347860886, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 60944 + }, + { + "epoch": 0.60945, + "grad_norm": 1.4329576362933967, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 60945 + }, + { + "epoch": 0.60946, + "grad_norm": 1.7711039464174667, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 60946 + }, + { + "epoch": 0.60947, + "grad_norm": 1.2210432754964342, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 60947 + }, + { + "epoch": 0.60948, + "grad_norm": 2.320739501257921, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 60948 + }, + { + "epoch": 0.60949, + "grad_norm": 1.2994495351418827, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 60949 + }, + { + "epoch": 0.6095, + "grad_norm": 1.7061074679278874, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 60950 + }, + { + "epoch": 0.60951, + "grad_norm": 1.5009956521083583, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 60951 + }, + { + "epoch": 0.60952, + "grad_norm": 1.6123113332389234, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 60952 + }, + { + "epoch": 0.60953, + "grad_norm": 1.682350973087455, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 60953 + }, + { + "epoch": 0.60954, + "grad_norm": 1.4151235105566164, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 60954 + }, + { + "epoch": 0.60955, + "grad_norm": 1.4031451547499323, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 60955 + }, + { + "epoch": 0.60956, + "grad_norm": 1.6827941040698429, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 60956 + }, + { + "epoch": 0.60957, + "grad_norm": 1.2783861783062491, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 60957 + }, + { + "epoch": 0.60958, + "grad_norm": 1.7598271984116889, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 60958 + }, + { + "epoch": 0.60959, + "grad_norm": 1.3408610811554114, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 60959 + }, + { + "epoch": 0.6096, + "grad_norm": 1.7111800504093841, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 60960 + }, + { + "epoch": 0.60961, + "grad_norm": 1.6518035546995529, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 60961 + }, + { + "epoch": 0.60962, + "grad_norm": 1.7752901368323035, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 60962 + }, + { + "epoch": 0.60963, + "grad_norm": 1.406824205672557, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 60963 + }, + { + "epoch": 0.60964, + "grad_norm": 1.829239116770237, + "learning_rate": 0.003, + "loss": 4.003, + "step": 60964 + }, + { + "epoch": 0.60965, + "grad_norm": 1.5090606894700704, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 60965 + }, + { + "epoch": 0.60966, + "grad_norm": 1.363140657669908, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 60966 + }, + { + "epoch": 0.60967, + "grad_norm": 1.3630917052922669, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 60967 + }, + { + "epoch": 0.60968, + "grad_norm": 1.5364934838866897, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 60968 + }, + { + "epoch": 0.60969, + "grad_norm": 1.75270797376046, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 60969 + }, + { + "epoch": 0.6097, + "grad_norm": 1.4878989267730052, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 60970 + }, + { + "epoch": 0.60971, + "grad_norm": 1.501212839291879, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 60971 + }, + { + "epoch": 0.60972, + "grad_norm": 1.5247386682034152, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 60972 + }, + { + "epoch": 0.60973, + "grad_norm": 1.902400935506535, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 60973 + }, + { + "epoch": 0.60974, + "grad_norm": 1.3667179241587926, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 60974 + }, + { + "epoch": 0.60975, + "grad_norm": 1.7468462136635816, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 60975 + }, + { + "epoch": 0.60976, + "grad_norm": 1.5876669141386768, + "learning_rate": 0.003, + "loss": 4.036, + "step": 60976 + }, + { + "epoch": 0.60977, + "grad_norm": 1.9077590900868062, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 60977 + }, + { + "epoch": 0.60978, + "grad_norm": 1.2445930902121571, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 60978 + }, + { + "epoch": 0.60979, + "grad_norm": 1.43101557400094, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 60979 + }, + { + "epoch": 0.6098, + "grad_norm": 1.6575612288017991, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 60980 + }, + { + "epoch": 0.60981, + "grad_norm": 1.4776699875250072, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 60981 + }, + { + "epoch": 0.60982, + "grad_norm": 1.3745013503871621, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 60982 + }, + { + "epoch": 0.60983, + "grad_norm": 1.3964930013433101, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 60983 + }, + { + "epoch": 0.60984, + "grad_norm": 1.5383188856102084, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 60984 + }, + { + "epoch": 0.60985, + "grad_norm": 1.4625853767594128, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 60985 + }, + { + "epoch": 0.60986, + "grad_norm": 1.8078034747534084, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 60986 + }, + { + "epoch": 0.60987, + "grad_norm": 1.839145754823219, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 60987 + }, + { + "epoch": 0.60988, + "grad_norm": 1.6755799995450018, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 60988 + }, + { + "epoch": 0.60989, + "grad_norm": 1.529042117295203, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 60989 + }, + { + "epoch": 0.6099, + "grad_norm": 1.411584714115826, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 60990 + }, + { + "epoch": 0.60991, + "grad_norm": 1.8787635033502335, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 60991 + }, + { + "epoch": 0.60992, + "grad_norm": 1.6091927760224103, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 60992 + }, + { + "epoch": 0.60993, + "grad_norm": 1.3642068484346679, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 60993 + }, + { + "epoch": 0.60994, + "grad_norm": 1.616224759738396, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 60994 + }, + { + "epoch": 0.60995, + "grad_norm": 1.4151866403341482, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 60995 + }, + { + "epoch": 0.60996, + "grad_norm": 1.4633732544588216, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 60996 + }, + { + "epoch": 0.60997, + "grad_norm": 1.4531116077031534, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 60997 + }, + { + "epoch": 0.60998, + "grad_norm": 1.620211721919402, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 60998 + }, + { + "epoch": 0.60999, + "grad_norm": 1.4080505816594202, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 60999 + }, + { + "epoch": 0.61, + "grad_norm": 1.999965963675852, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 61000 + } + ], + "logging_steps": 1, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.418262214180864e+18, + "train_batch_size": 1024, + "trial_name": null, + "trial_params": null +}